SECURITIES AND EXCHANGE COMMISSION
Washington, D.C. 20549



FORM 6-K

Report of Foreign Private Issuer
Pursuant to Rule 13a-16 or 15d-16 of the
Securities Exchange Act of 1934

For the month of January, 2018

Commission File Number 1-34129



CENTRAIS ELÉTRICAS BRASILEIRAS S.A. - ELETROBRÁS
(Exact name of registrant as specified in its charter)



BRAZILIAN ELECTRIC POWER COMPANY
(Translation of Registrant's name into English)



Avenida Presidente Vargas, 409 - 13th floor,
Edifício Herm. Stoltz - Centro, CEP 20071-003,
Rio de Janeiro, RJ, Brazil
(Address of principal executive office)

Indicate by check mark whether the registrant files or will file annual reports under cover Form 20-F or Form 40-F. 

Form 20-F ___X___ Form 40-F _______

Indicate by check mark whether the registrant by furnishing the information contained in this Form is also thereby furnishing the information to the Commission pursuant to Rule 12g3-2(b) under the Securities Exchange Act of 1934.

Yes _______ No___X____


 

 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

DFP/DFF/DDE Joint Technical Information

 

Title  Analysis of the sale of Eletrobras’ shareholding in the distribution
companies Amazonas Distribuidora de Energia (AmE D), Boa Vista
Energia, Ceal, Cepisa, Ceron and Eletroacre. 
 
 
Subject  Assessment of the BNDES modeling and CPPI Resolution no.
20/2017 for the sale of the shareholding in the distribution
companies of Eletrobras. 
 
 

Authors
 
Felipe Baptista da Silva DFP;
Marcos Barreto de Faria Pinho DFPI;
Flávia Xavier Cirilo de Sá DFPI;
Pedro Paulo da Cunha DFF;
Frederico Pinto Eccard DDEF;
Luciana Pereira de Souza DDEF. 
 
 
 
 
Company  Eletrobras, Amazonas Distribuidora de Energia (AmE D), Boa Vista
Energia, Ceal, Cepisa, Ceron, Eletroacre 
 
 
Category  Technical Information. 
 
Key-words  Amazonas Energia (AmE D), Boa Vista Energia, Ceal, Cepisa, Ceron,
Eletroacre, Sale, Transfer, Liquidation. 
 

 

Page 1


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

Contents 

 

 
1.  objective  4 
2.  Disclaimer  4 
3.  PDNG 2017 - 2021  5 
4.  HISTORY  6 

4.1 

ACQUISITION OF the distribution companies CONTROLLED BY ELETROBRAS  6 

4.2

 Economic, financial and operational performance of energy distribution companies 8 

4.3 

DECISION TO NOT EXTEND ELETROBRAS’ electricity distribution CONCESSIONS   10 

4.4 

Temporary provision of distribution service  17 
5.  CURRENT STATUS Of distribution companies  20 

5.1 

rights and obligations with CCC Fund  25 
6.  Management of the process of privatization of the distribution companies  29 

6.1 

Costs with the privatization process  29 

6.2 

Remuneration of BNDES  30 
7.  Conditions for the DISPOSAL of controlling STAKE  31 
8.  Disposal modeling of the controlling stake  34 

8.1 

Methodology adopted for the economic and financial assessment  35 
9.  Disposal modeling of the controlling stake  36 

9.1 

Adjustments to the Equity Value of the distribution companies  39 

9.1.1

 Adjustments to contemplate the updates in the balance sheet until June 2017  40 

9.1.2

Adjustment of the flexibilization of regulatory parameters 

41 

9.2 

stock offering to employees and retirees 

41 

9.3 

Shareholding option of Eletrobras 

42 

9.4 

Requirements for completing the disposal of the controlling stake 

43 
10. 

Analysis of Results per distribution company 

45 

 

Page 2


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
10.1  relevant points  50 
10.1.1  Unbundling of Amazonas Energia  50 
10.1.2  employment litigation Bresser Plan - Ceal  52 
10.1.3  Boa Vista Modeling CERR Indemnity  53 
10.1.4  Fixed Assets underway  54 
11. Impacts on Eletrobras  55 
11.1  Financial Impact Cash Flow  55 
11.2  Economic Impact ACCOUNTING Effect  59 
12. Conclusion  65 
12.1  Reason for the non-continuance of Eletrobras in the control of the distribution companies    66 
12.2  Conditions proposed for the disposal operation of the distribution companies  66 
12.3  Financial Impact on Eletrobras to execute the proposal of CPPI  67 
12.4  ceal’S Specific concern ings  68 
12.5  amazonas distribuidora’s Specificconcernings  68 
12.6  Scenario of Liquidation of the Distribution companies  68 

 

Appendices

Tariff Level

Analysis and Results of Amazonas Energia Analysis and Results of Boa Vista Energia Analysis and Results of Ceal Analysis and Results of Cepisa Analysis and Results of Ceron Analysis and Results of Eletroacre

Page 3


 

 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

1. OBJECTIVE

This Technical Information is intended to support the decision of Eletrobras’ Management concerning the sale of controlling interest of the distribution companies - Amazonas Energia Distribuidora S.A. Amazonas Distribuidora de Energia (AmE D), Boa Vista Energia S.A Boa Vista Energia, Companhia Energética de Alagoas S.A. Ceal, Companhia Energética do Piauí S.A. Cepisa, Centrais Elétricas de Rondônia S.A. Ceron, Companhia de Eletricidade do Acre S.A. Eletroacre, in privatization auction according to modeling approved by the Board of the Investment Partnership Program CPPI, by means of Resolution 20, dated November 8, 2017, amended by Resolution 28, dated November 22, 2017, based on studies performed by BNDES. The privatization process will result in the adoption by such companies of a legal regime similar to the one of other non-state agents operating in the national energy supply segment.

In this vein, the 2015-2030 Strategic Plan, approved on November 26, 2014, by means of RES/DEL 775/2014, defines the generation and transmission sectors as a main part of its business portfolio or core business, excluding the distribution segment, rendering clear the need for repositioning Eletrobras in this segment.

2. DISCLAIMER

It should be noted that, for all the distribution companies, the basic data and the assumptions adopted in the economic and financial projections, and in the business modeling, were defined by BNDES - the Brazilian Development Bank, in its own independent process, without the interference of Eletrobras.

The first version of the BNDES’ study was sent to Eletrobras on September 27, 2017, through the AD Official Letter no. 22/2017 of BNDES. Subsequently, a new version was submitted to Eletrobras on October 19, 2017, through AD/DEADE3 Official Letter no. 18/2017 and, finally, on November 01, 2017, BNDES presented the final version of the study through the AD/DEADE3 Official Letter no. 20/2017, which was approved by CPPI

Resolution 20/2017.

Said CPPI Resolution was enacted based on Decree 9192, dated

November 6, 2017, and was amended by CPPI Resolution 28/2017.

The modeling proposed by BNDES includes the sale of Amazonas Distribuidora de Energia S.A. without Amazonas Geração e Transmissão de Energia S.A., i.e., already unbundled.

Page 4


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017


001/2017

Therefore, the realization of the sale of AmE D depends on the conclusion of the so-called unbundling process established in Paragraph 5, Article 4 of Law 9074, dated July 7, 1995, which defines that the concessionaires, the permit holders and the authorized electric energy public utility companies that operate under the National Interconnected System -SIN cannot develop activities of generation and transmission of electric energy, among others.

3. PDNG 2017 - 2021

In November 2016, the Board of Directors of Eletrobras CAE, by means of DEL 224/2016, approved the Master Plan of Business and Management of Eletrobras Companies - PDNG 2017 - 2021, which encompasses targets and projects for accomplishing the strategic goals of the companies of Eletrobras System and keep alignment with the Corporate Identity of Eletrobras Mission, Vision for 2030 and Values.

2017-2021 PDNG is centered upon three cornerstones or strategies, as shown below:

(i) Governance and compliance, by means of the strategic alignment, risk management and compliance;

(ii) Financial discipline, aiming to reduce its net and gross debt, by means of privatizations, demobilization of operating and non-operating assets;

(iii) Operating excellence, by means of the reduction of their costs with respect to the current costs and increase of its current revenues, upon operating restructuring, Encouraged Retirement Plan - PAI, Center of Shared Service and automation.

Thus, the privatization of the distribution companies is part of the scenario of economic and financial balance of Eletrobras, since it will avoid large disbursements of funds necessary to finance the immediate cash deficit and the large investments established in the new concession cycles of each of these distribution companies.

It should be noted that the disposal of the distribution companies’ controlling stake will greatly help achieving the target of reducing the net debt/EBITDA indicator, thus reducing the future needs for Eletrobras’ fund raising.

Page 5


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

It should also be pointed out that, since they are federal state-owned companies, the transfer of the shareholding controls of these distribution companies shall abide by the rules of PPI - Program of Partnerships and Investments, created by Provisional Measure no. 727/2016, converted into Law 13334, dated May 12, 2016, in order to facilitate the privatization.

In this sense, the inclusion of the distribution companies of Eletrobras in PPI happened upon the Resolution 3/2016 of CPPI, which, among other conditions, defined that the privatization of the distribution companies should happen over the 2nd half of 2017. This decision of CPPI was legitimized with the issuance of Decree 8893, dated November 1, 2016, which defined the PPI projects which will be treated as a national priority in the energy and mining industries and, recently, by Decree 9192, dated November 6, 2017, which provides, pursuant to Law No. 12783/2013, for the bidding of distribution and transmission concessions associated with the transfer of control of a legal entity providing public electricity service.

4.     

HISTORY

 

4.1      ACQUISITION OF THE DISTRIBUTION COMPANIES CONTROLLED BY

ELETROBRAS

Eletrobras acquired the equity control of a set of state-owned utility service companies in the North and Northeast regions in order to promote its economic and financial balance and aiming at the privatization of such companies. In this sense, Law 9619 / 98 authorized Eletrobras to purchase the respective controlling stake of the state electricity distribution utility service companies: Ceal, Cepisa, Ceron and Eletroacre, with funds of the General Reversal Reserve - RGR. The privatization process of such companies was legally supported by their inclusion in the National Privatization Program - PND. Later, by means of Provisional Measure no. 1985-25/2000 (transformed in MP no. 2181-45/2001), Eletrobras was also authorized to acquire the controlling stake of Companhia Energética do Amazonas - CEAM, which was included in the PND on January 22, 2007 by means of Decree 6026/2007. This company was dissolved by merger into Manaus Energia S.A., originating Amazonas Energia, currently Amazonas Distribuidora de Energia S.A.

Page 6


 

 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

Law 9619 of April 2nd, 1998 also provides that, once effected the acquisition control of the aforementioned companies, they would be included in the PND, and it would fall upon Eletrobras to implement any and all economic, financial and administrative measures required to privatize the company, as per Law 9491 of September 9th, 1997.

After assuming the respective controlling stake, Eletrobras was in charge of managing such energy distribution companies, as well as making investments required to foster the continuity of the distribution utility services provided by them. It is worth mentioning that the influence of Eletrobras should be temporary and would end with the effective privatization of the energy distribution companies. BNDES, at the time and in acting as privatization process manager, was in charge of conducting such processes, which were indeed initiated. At that occasion, the referred financial institution hired technical advisories in order to evaluate companies and proceeded to follow up such services.

Due to several reasons, the Privatization Program was not implemented in the 1990s, and the distribution companies, which initially would be managed by Eletrobras for a short period of time, became permanent investments. In a parallel process, focusing on the operation of its business group in the generation and transmission segments, Eletrobras split up Eletronorte, forming with the supply assets from Manaus and Boa Vista, respectively, Manaus Energia and Boa Vista Energia, independent companies also intended to be privatized. Later, CEAM was incorporated by Manaus Energia, whose process finished in 2008. In the following year, the name of the supplier was altered and became Amazonas Energia S.A.

It is important to highlight that during this whole time the federal-owned distribution utility service companies were not removed from the PND, even with the change of the sector management policy that took place after 2002. This fact is even more relevant considering that Eletrobras and its subsidiaries Eletrosul, Eletronorte, Chesf, Cgtee and Furnas were removed from the PND by Act 10848 of March 15th, 2004, which revoked article 5 of Law 9.648, dated May 27, 1998; The table hereunder presents the controlling shares of Eletrobras in the capital of the distribution companies aimed in this process.

 

Page 7


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
Table 1.  Current equity of Eletrobras in the capital stock of the energy suppliers 
 
COMPANIES NOT NEGOTIATED IN STOCK EXCHANGE Total Amount of Issued
Shares
Capital Stock (R$
Thousand)
Shares owned by
ELETROBRAS
Total
COM.  PREF. 
AMAZONAS DISTRIBUIDORA DE ENERGIA S.A.  6,276,666,628  4,610,171  100.00%  0.00%  100.00% 
BOA VISTA ENERGIA S.A  320,742,889  684.204  100.00%  0.00%  100.00% 
COMPANHIA ENERGÉTICA DE ALAGOAS CEAL  734,753,835  734.754  100.00%  100.00%  100.00% 
COMPANHIA ENERGÉTICA DO PIAUÍ CEPISA  779,223,552  1,272,747  100.00%  100.00%  100.00% 
CENTRAIS ELÉTRICAS DE RONDÔNIA CERON  1,212,156,653  1,325,369  100.00%  0.00%  100.00% 
COMPANHIA DE ELETRICIDADE DO ACRE ELETROACRE  108,271,416,523  475.789  97.71%  94.04%  96.71% 
Source: Financial Statements Dec/2016, Eletrobras. Own preparation.       

 

4.2 ECONOMIC, FINANCIAL AND OPERATIONAL PERFORMANCE OF

ENERGY DISTRIBUTION COMPANIES

Despite the contributions made by Eletrobras around R$ 22.9 billion (amounts adjusted by SELIC until Jun./17) since the year 2000, estimates show that the energy suppliers still need large sums of resources not only to finance expansion, but also to cover operating deficits and working capital, particularly by debt service, purchase of energy, at high indices of technical and commercial losses and the unbalances in the transfers of Fuel Consumption Account CCC to insolated systems.

Table 2.  Investments made in energy suppliers by Eletrobras (R$) 
Companies  HISTORY  Updated SELIC 
Amazonas Energia  BRL 2,688,321,197.10  BRL 7,159,690,493.17 
Boa Vista Energia  BRL 80,089,295.37  BRL 87,058,647.95 
Ceal  BRL 860,153,755.76  BRL 2,610,616,440.89 
Cepisa  BRL 1,536,692,720.96  BRL 5.267.936.674.59 
Ceron  BRL 1,655,015,610.24  BRL 6,531,674,654.86 
Eletroacre  BRL 392,140,618.49  BRL 1,230,955,731.94 
TOTAL  BRL 7,212,413,197.92  BRL 22,887,932,643.40 
Source: AFI Eletrobras System   

 

In addition to the disbursed amounts, Eletrobras made the payment of R$ 638.2 million for the state governments to acquire the distribution companies, with funds withdrawn from RGR Fund, which, adjusted until Jun./2017 at the rate of 5% per annum (rate that adjusts the debt balance of RGR fund for inflation), reaches the amount of R$ 1.4 billion1. Thus, it can be considered that, concerning updated amounts, Eletrobras has already spent R$

1 It is worth emphasizing that the historical amount of R$ 173.1 million paid to acquire Manaus Energia is not part of the sum paid to State Governments, as the former was acquired from Eletronorte, currently undergoing a corporate restructuring.

Page 8


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

distribution activity with six energy distribution companies, an amount that does not consider R$ 5.3 billion of the debt balance (Jun./2017) from loans granted with ordinary resources from Eletrobras.

Table 3.  Total paid by energy suppliers (BRL thousand) Eletrobras position 
Companies HISTORY Updated by 5% p.a.
Up to 06/2017 
CEPISA  120,003,368  274,709,096 
CERON  118,377,654  266,417,870 
CEAL  168,700,000  382,938,037 
ELETROACRE  28,000,000  63,654,389 
CEAM  203,167,080  430,141,473 
Manaus Energia (*)  173,139,000  503,634,737 
TOTAL  811,387,102  1,921,495,601 
Source: Eletrobras     

 

One may confidently affirm that the conditions of the service provided to local populations have improved over the years, such as, for instance, the increase of distribution network extension and the reduction of DEC and FEC indicators. However, it was not possible to reach the operating and quality indexes in any of the companies set out by the Brazilian Electricity Regulatory Agency - ANEEL. Additionally, even though there was no success in reverting the negative financial situation of most of those companies and no financial return of the invested capital, some earnings were relevant for the served populations and for the local economies.

In an analysis made in 2008 concerning the issues faced by energy distribution companies and the causes behind the unsatisfactory performance, factors of regulatory and organizational nature were identified. The first is related to the fact that changes to the Brazilian Power Sector model prioritized the interconnected system, while specific rules for the insolated systems were left to be analyzed in a future moment.

Higher loss and default indices, and in many cases, higher operating costs compared to private utility service companies, in view of legal requirements for contracting goods and services, contributed to the observed economic and financial imbalances.

Not only the difficulties and issues resulted from the specific regime of state-owned companies aggravated the scenario of operational and financial difficulty currently faced by

Page 9


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

energy distribution companies ,but also other peculiar factors, the vast and poorly populated concession area (approximately 30% of the Brazilian territory with only 6.5% of the Brazilian population), the geographic dispersion of the managed assets, social aspects of the areas where companies operate, and in some cases, with outdated tariffs2, where the peculiarity of the isolated systems, and mainly of the Amazon rainforest, all imposed more operating costs that are currently not reflected in the adopted tariffs.

It is also worth mentioning that energy distribution companies have faced difficulties in reaching the goals set out by ANEEL, mainly concerning default of clients, nontechnical losses, service interruption indexed and operating costs. Part of the problems is due to the precarious nature of the distribution utility system, considering the entire amount of resources invested in these companies, which increases maintenance costs of the network and will result in the significant needs for investments by the new controller.

4.3 DECISION TO NOT EXTEND ELETROBRAS’ ELECTRICITY DISTRIBUTION CONCESSIONS

 

Following the awarding of a sizeable number of concessions related generation, transmission and distribution of electric power in 2015, the Federal Government amended Act 12783, of January 11th, 2013, conversion of the Provisional Measure 579 of September 11th 2012, which allowed the extension of such concessions, at the discretion of the granting power, a single time, for the term of up to 30 (thirty) years, in order to ensure the continuity, efficiency of the service provided and the lower tariff value.

 

Extension of concessions related to generation and transmission of electric energy was anticipated to January 2013, as per Decrees 7805 of September 4th, 2012 and Decree 7850 of November 30th, 2012, which regulated Provisional Measure 579/2012.

Concerning the distribution utility concessions, considering that such activity is already strongly regulated by ANEEL, there was no need to anticipate the extension of such concessions. This way, the regulation of Article 7 of Law 12783/2013 was implemented with the homologation of Decree 8461, dated June 2, 2015.

2 For Cepisa and AmE D, assets considered in the shielded remuneration base are well beyond the totality of unitized and operating assets of such companies, which results in a low remuneration of the invested capital and in low tariffs, contributing to operating deficits presented by them.

Page 10


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Article 1 of Decree 8461/2015 allowed the extension of distribution utility concessions for 30 (thirty) years considering the compliance with the following criteria:

(i) efficiency with relation to the quality of the service provided, calculated by means of indicators “Equivalent Interruption Duration of Internal Origin per Consuming Unit” (DECi) and “Equivalent Interruption Frequency of Internal Origin per Consuming Unit”
(FECi);

(ii) efficiency with relation to the economic and financial management, calculated by means of EBITDA and indebtedness level;

(iii)      operational and economic rationality; and 
 
(iv)      lower tariff value.

On June 9th, 2015, ANEEL, with the purpose of obtaining subsidies and additional information and of improving the amendment term draft to the distribution utility concession contract, in accordance with Decree 8461/2015, summoned Public Hearing AP 38/2015, by means of documental interexchange, between June 10th and July 13th, 2015. This Public Hearing was reopened between September 25th and October 5th, 2015 with the purpose of collecting subsidies and additional information solely concerning the subclause included in the amendment term draft to the contract, to apply the objective criteria set forth in Decree 8461/2015, enticers of the concession extinction, related to the assessment of the quality of the service and of the economic and financial management as per the terms of determination 9.7.2 from Court Decision 2253/2015 of the Federal Court of Accounts TCU.

The result of AP 38/2015 was the publication of ANEEL Dispatch 3540/2015, which approved the draft of the amendment term or concession contract of electric energy distribution companies, with the purpose of extending them and submitted the referred processes to the Ministry of Mines and Energy.

Considering the need to extend theterm to conclude studies related to the draft of the amendment term or concession contract of electric energy distribution companiess made available by ANEEL, by the interested parties, in order to allow concessionaires to foster all businessrelated arrangements for decision-making and to make possible a detailed assessment by its controllers, it was then homologated Provisional Measure 706 of December 28th, 2015, converted in Act 13299 of June 21st, 2016, which granted term of

Page 11


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

up to 210 (two-hundred and ten) days to allow the concessionaire to sign a concession contract or amendment term following decision of the granting power for the respective extension.

Considering the enlargement of the term for decision-making concerning the extension of energy distribution concessions, Eletrobras hired an independent consulting company, Deloitte Touche Tohmatsu, to perform the economic and financial assessment of its suppliers with the purpose of providing subsidies to its decision-making bodies.

At first, it is worth emphasizing that the studies prepared by Deloitte had as main premises the transfer of the controlling stake of energy distribution companies in 2017 and the benefits resulting from PLC 11/2016. Additionally, it was performed an analysis of the amendment term draft to the concession contracts.

ANEEL, aiming to establish annual targets in compliance with the provisions under Decree 8461/2015, ascertained the methodology in Technical Note 335/2015-SCT-SFE-SFF-SRD-SEM/ANEEL for the period ending in September 2014. The results are included in the chart below. For the calculation of interest, Selic was considered at 12.75% p.a. The 111% spread implies a cost of 14.15% p.a.

Table 4.  ANEEL: Economic and financial situation of the distribution 
  companies with contracts to be expired until 2017. 
Distribution
company 
Net Debt  Adjusted
EBITDA 
Replacement
Investment (QRR) 
EBITDA-QRR EBITDA - QRR - Interest
CEPISA  1,019.90  60.7  25.9  29.3 x  -109.5 
Boa Vista  157.6  -9.7  7.7  Negative EBITDA  -89.7 
CEAL  499.6  -29.6  45.3  Negative EBITDA  -145.6 
CERON  851.3  -97.6  26.8  Negative EBITDA  -283.1 
Eletroacre  199.7  -82.4  14.5  Negative EBITDA  -125.2 
Amazonas Energia  2,211.80  -377.1  13.7  Negative EBITDA  -703.8 

Source: ANEEL, Technical Note 0335/2015-SCT-SFE-SFF-SRD-SEM/ANEEL.
Note: values in BRL million.

Additionally, Deloitte assessed the EBITDA and the indebtedness forecasted for the distribution companies of Eletrobras, as well as the need for disbursements of capital and the amounts of the debts of these distribution companies of ordinary funds (RO) of

Page 12


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Eletrobras based on the position of June 30, 2016, in such a way to check if they would meet the targets indicated by ANEEL. The results are included in the tables below.

Table 5.  EBITDA: with the benefits of PLC 11/20163     
Distribuidora  2016  2017  2018  2020  2021  2021  2022  2023  2024  2025 
CEPISA  219,2  96  110,6  188,2  353,3  255,2  277,2  305,9  296,1  314,7 
Boa Vista  -21  -6,9  18,8  39,2  76,6  82,8  75,6  86,4  94,9  100,6 
(com PLC 11/2016                     
CEAL  97,1  34,1  128,2  179,7  339,9  235  244,7  292  306,7  303,4 
CERON  4,5  -27,5  -47,6  38,5  119  273,4  244,9  272,1  310,5  372,2 
Eletroacre  1,6  0,7  66,2  101,6  214,2  157  163,3  185,5  199,1  218,3 
Amazonas Energia                     
(com PLC 11/2016 neutralidade  -965,8  -476,9  273,1  413,5  752,7  671,2  792,5  858,6  925  -934,8 
do contrato de gás)                     
Total  -664,49  -380,46  548,37  960,67 1.855,77 1.674,60 1.798,21 2.000,41 2.132,21 2.243,95 
Amazonas Energia                     
(com PLC 11/2016 com impacto  -1.607,70  -578,9  76,1  265,2  424,3  100,3  180,3  202,5  230,3  -371,5 
do contrato de gás)                     
 
Source: Deloitte Assessment.                   

 

Table 6.  Need for Disbursements of Capital: with the benefits of PLC 
11/2016.
Distribution
company 
2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Total
CEPISA  234.80  270.30  250.80  154.50  0.00  0.00  0.00  22.80  26.10  15.60  974.90 
Boa Vista                       
  59.90  39.70  0.00  0.00  0.00  0.00  0.00  0.00  0.00  0.00  99.60 
(with PLC 11/2016)                       
CEAL  85.60  103.40  50.10  0.00  0.00  0.00  0.00  0.00  0.00  0.00  239.10 
CERON  234.60  215.40  147.20  103.10  80.20  0.00  0.00  0.00  0.00  0.00  780.50 
Eletroacre  98.50  182.10  41.70  4.70  0.00  0.00  0.00  0.00  6.40  0.00  333.40 
Amazonas Energia                       
(with PLC 11/2016   2,289.10  1,681.10  811.80  568.40  76.40  0.00  113.80  71.00  43.30  0.00  5,654.90 
neutrality of the gas                       
contract)                       
Total  3,002.40  2,492.10  1,301.50  830.70  156.70  0.00  113.80  93.80  75.80  15.60  8,082.40 
 
Source: Deloitte Assessment.                   

 

3 PLC 11/2016 Bill for Conversion of the Provisional Measure 706/2015, which ended up with the sanction of Law 13299/2016.

Page 13


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
Table 7.  Need for Disbursements of Capital (c/ PLC) + RO Capitalization 
  Distribution
company 
Total   RO Capitalization    Total with RO
Capitalization 
     
     
  CEPISA  974.90    966.00    1,940.90 
  Boa Vista           
  (with PLC 11/2016)  99.60    44.00    143.60 
  CEAL  239.10    1,050.00    1,289.10 
  CERON  780.50    604.00    1,384.50 
  Eletroacre  333.40    243.00    576.40 
  Amazonas           
  Energia  5,654.90    1,149.00    6,803.90 
  (with PLC 11/2016 -           
  Total  8,082.40    4,056.00    12,138.40 
  Source: Deloitte Assessment.   

 

Table 8. Need for Disbursements of Capital: without the benefits of PLC
11/2016.
 
Distribution company  2016 2017 2018 2019 2020 2021    2022  2023  2024  2025 Total
Boa Vista
(without PLC 11/2016) 
86.90 87.80 0.00 0.00      0.00  0.00  0.00  0.00  0.00  0.00  174.70
Amazonas Energia
(without PLC 11/2016 neutrality of the gas contract) 
4,204.30 4,447.80 1,825.90 1,680.60 939.00  719.50  873.70  746.50  693.70  208.10  16,339.10
OTHER DISTRIBUTION COMPANIES  653.40  771.30  489.70  262.30  80.20  0.00  0.00  22.80   32.50  15.60  2,327.80 
Total  4,944.60  5,306.90  2,315.60  1,942.90  1,019.20  719.50  873.70  769.30  726.20  223.70  18,841.60 
 
Source: Deloitte Assesst.             

 

Page 14


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
Table 9.  Need for Disbursements of Capital (c/ PLC) + RO Capitalization 
 
                 
  Distribution company Total   RO Capitalization   Total with RO
Capitalization 
     
     
  Boa Vista
(without PLC 11/2016) 
174.70   44.00   218.70
     
  Amazonas
Energia
(with PLC 11/2016 neutrality of
the gas contract) 
16,339.10   1,149.00   17,488.10
     
     
  Other Distribution
companies 
2,327.70   2,863.00   5,190.70
     
  Total  8,082.40    4,056.00    12,138.40 
 
 
  Source: Deloitte Assessment.     

 

The tables above show that, in order to meet the economic and financial sustainability goals defined by ANEEL, it would be necessary to capitalize on the companies to make possible the payment of all RO’s debts to Eletrobras, as well as new capital contributions up to 2025, which would total R$ 12.1 billion, on the date of the studies, considering the scenario with the benefits of PLC 11/2016 and the neutrality of the natural gas contract signed with Cigás/Petrobras, existing in AmE D. Out of this amount, R$ 7.1 billion would be in 2016 and R$ 4.0 billion would be for the payment of RO’s debts, and R$ 3.0 billion for investments and other. In the alternative scenario, without the benefits of PLC 706/2016 and with the neutrality of the gas contract of AmE D, the need for disbursements of capital is expected to increase up to BRL 22.9 billion. Out of this amount, BRL 9.0 billion would be earmarked for 2016, out of which, BRL 4.0 billion should be used to pay the RO debts and BRL 4.9 billion would be allocated in investments and other disbursements.

Thus, the studies pointed out that the extension of Eletrobras’ distribution utility service concessions would only be viable if the Federal Government contributed capital directly to the distribution companies, through the assignment of the preemptive rights by Eletrobras, in compliance with Law 6,404/1976 and Decree 1,091/1994, in the total amount required to meet the targets set by ANEEL, until the sale of the controlling interest of the distribution companies.

In addition to this, not only the high level of indebtedness of the distribution companies is noteworthy, but also their litigations. The table below shows, in a simplified form, the

Page 15


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

assets and liabilities of the distribution companies in December 2015, which were taken into consideration in Deloitte’s assessments.

Table 10.  Utilities: Assets and Liabilities on 12/31/2015. 
Distribution company Assets (CP+LP) LIABILITIES
(CP+LP) 
Equity
Net amount 
Cepisa  1,556,502.00  2,257,652.00  -701,150.00 
Boa Vista  623,360.00  961,003.00  -337,643.00 
Ceal  1,802,221.00  2,049,877.00  -247,656.00 
Ceron  5,322,189.00  5,778,747.00  -456,558.00 
Eletroacre  915,320.00  1,045,003.00  -129,683.00 
Amazonas Energia  13,902,698.00  18,266,296.00  -4,363,598.00 
Total  24,122,290.00  30,358,578.00  -6,236,288.00 

 

Source: Eletrobras Financial Statements as of December 31, 2015

In this regard, the following should be underlined: i) the unavailability of Eletrobras to dispose of or arrange funds to make the said disbursements of capital given its level of indebtedness and the country's adverse macroeconomic situation; ii) the efforts to achieve the targets defined by ANEEL were relatively severe; and iii) PLC 11/2016 could not be converted into Law, increasing the risks for Boa Vista and AmE D, and generating risks and uncertainties about the feasibility of maintaining the concessions of these distribution companies.

In view of the foregoing, on July 22, 2016, the Eletrobras’ shareholders, through the 165th Extraordinary General Meeting, decided, by majority vote, not to extend the concessions of the Eletrobras’ distribution companies, pursuant to Decree No. 8461, dated June 2, 2015, of ANEEL Order no. 3540, dated October 20, 2015, of Provisional Measure no. 706/2015, dated December 28, 2015, of the Ministry of Mines and Energy Order (without number), dated December 28, 2015, and the draft contractual instrument issued by ANEEL.

In addition to this, the shareholders approved, by majority vote, the liquidation of the distribution companies Cepisa, Ceal, Eletroacre, Ceron, Boa Vista Energia and AmE D, allowing the transfer of controlling shareholder, as long as it happened until December 31,

Page 16


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

2017, under the terms of Paragraph 1-A of Article 8, Law no. 12783/2013, with the new wording given by Provisional Measure No. 735/2016, converted into Law 13360/2016, provided that, until the transfer of the distribution company to the new controller, the distribution company received, directly from the Federal Government or through a tariff, all the resources and remuneration necessary to operate, maintain and make the investments which are connected to the public utilities of the respective distribution company, maintaining the economic and financial balance of the Distribution company, without any contribution of resources, at any title, by Eletrobras.

On that occasion, Eletrobras' shareholders decided to approve that, should the transfer of control not occur until the date of December 31, 2017, all necessary steps should be taken to liquidate these electricity distribution utility service companies, under the terms of item 11 of the 165th EGM’s draft,, as transcribed below:

11. Approving, by majority vote, that the concession of distribution companies Companhia Energética de Piauí - CEPISA; Companhia Energética de Alagoas - CEAL; Companhia de Eletricidade do Acre -ELETROACRE; Centrais Elétricas de Rondônia S.A - CERON; Boa Vista Energia S.A; and Amazonas Distribuidora de Energia S.A be returned at any time and that the provisions of its liquidation are adopted, in the following events: (i) the transfer of controlling interest referred to in Item 9 above is not performed until December 31, 2017;"

4.4 TEMPORARY PROVISION OF DISTRIBUTION SERVICE

Law 12783/2013, in its Article 9 established that, if there was no extension of term of concession and aiming to ensure the continuance of the service provision, the holder may, after maturity remain in charge of the electricity services provision until the new utility service company assumes the responsibility, abiding by the conditions established under the Law.

Considering the decision of the shareholders of Eletrobras for the non-extension of its distribution concessions, the Ministry of Mines and Energy enacted the MME Ordinance 388, dated July 26, 2017, which establishes the terms and conditions for the Provision of the Public Utility of Distribution of Electric Energy by and Agency or Department of the Federal Government, under the terms of Article 9, paragraph 1 of Law 12783/2013. The

Page 17


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

said Service Provision shall be applicable to the areas whose concession has not been extended, under the terms of the said Law, Decree no. 7805/2012 and Decree no. 8461/2015, with the purpose of ensuring the service continuance.

In this context, by means of the MME Ordinances numbers 420, 421, 422, 423, 424 and 425, the distribution companies of Eletrobras Amazonas Energia SA, Companhia de Eletricidade do Acre, Centrais Elétricas de Rondônia SA - CERON, Companhia Energética of the Piauí - CEPISA, Companhia Energética de Alagoas - CEAL, and Boa Vista Energia SA were appointed as Responsible for the Provision of the Public Service of Distribution of Electric Energy, in order to ensure the continuance of the service in the established areas.

Additionally, by means of MME Ordinance no. 425, dated August 3, 2016, and by means of Order of Minister of Mines and Energy, on August 11, 2016, the concession that was owned by Companhia Energética de Roraima - CERR was terminated, and Boa Vista Energia became the temporary provider of the distribution service, effective as of January 1. 2017, in the area previously served by CERR.

On September 14, 2016, ANEEL opened the Public Hearing 063/2016, with the purpose of obtaining subsidies for the improvement of the regulation of MME Ordinance no. 388/2016 referred to above.

Additionally, as a result of the need for immediate application of the provisions proposed in the said Public Hearing, under the penalty of hindering the continuance and adequacy of the service provided, ANEEL provisionally published Order 2447/2016. The said hearing ended up with the publication of the Normative Resolution 748/2016 and Homologation Resolution 2184/2016. In this last Resolution, the monthly value to be received by each distribution company as loan of RGR was homologated4, and referred to as appropriate remuneration. This amount was calculated by ANEEL to eliminate the monthly cash deficits of these companies, based on financial information for 2015, since Eletrobras was precluded, by decision of its shareholders, to disburse new funds to the distribution companies.

4 Initially, the funds of RGR would be loans that the new controlling shareholder was expected to pay in 48 months after the execution of the concession contract, out of which, 12 months would stand as grace period for interests and principal, and 36 months would be for amortization purposes, at the cost of 111% of Selic. However, the current understanding of ANEEL and MME is that the payment of these RGR loans would significantly burden the future controlling shareholder of the distributors, reducing the attractiveness of the companies, and may frighten interested parties into the auction. Thus, by means of Technical Note 149/2017, a ANEEL, according to MME, there is positioning for the loan of RGR to start being paid by the consumer, through the tariff, between the 6th and the 30th year of the new concession.

Page 18


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

In October 2016, ANEEL forwarded to the distribution companies of Eletrobras the Instrument of Commitment and Temporary Distribution Service Provision Plan. The Instrument of Commitment is an instrument whereby the shareholders of the distribution companies undertake the commitment to abide by the conditions for the continuity and appropriateness of the service provided, especially with respect to the following parameters: (i) compliance with the industry rules; (ii) loss of electricity; (iii) operating costs; (iv) Equivalent Duration of Interruption per Consumer Unit (DEC); (v) Equivalent Frequency of Interruption per Consumer Unit (FEC); (vi) Quality of information provided to ANEEL; and (vii) compliance with ANEEL’s inspection regulations. The Temporary

Distribution Service Provision Plan establishes the conditions to ensure the continuity and the appropriate provision of the distribution service until the bidding and assumption of the new utility service company, for approval by Eletrobras. The instruments above were approved by the Board of Executive Officers of Eletrobras by means of Resolution 636/2016.

It should be noted that during the precarious concession period of the Distribution companies, both MME and ANEEL participated actively in the management of the distribution companies appointed for the provision of services. In MME Ordinance no. 388/2016, it was established, in its chapter IV, that the provision of the service would be monitored, inspected and regulated by ANEEL.

Aiming this purpose, ANEEL, by means of Normative Resolution no. 748/2016, ruled down, among other measures, that the senior management of the appointed distribution companies should monthly attend ANEEL, as of January 2017, which has been happening, to render accounts concerning the execution of the agreed on Temporary Service Plan, in order to present the results achieved concerning each of the indicators followed, as well as the actions implemented and under implementation to comply with said plan.

In addition to this, reports, certified by the respective Fiscal Boards, regarding the compliance with the Temporary Provision of the Service until its actual conclusion, are submitted quarterly to ANEEL, based on the income of the fourth quarter of 2016, as a condition for receiving the transfers of resources of CCC, CDE, RGR, and the application of tariff adjustments and reviews.

In this sense, a working group was created in the Agency by ANEEL Ordinance No. 4416 / 2017, to monthly monitor the compliance with the Temporary Service Plans.

Page 19


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

In addition to this working group, another group was created, within the scope of the

MME’s Electric Energy Office, through the SEE/MME Ordinance no. 4, dated September 22,

2016, for monthly monitoring of the provision of public energy distribution utility. This group includes members of the Executive Office/MME, Office of Electric Energy/MME, Office of Planning and Energy Development/MME and Economic Advisory/MME, as well as the Power Commercialization Chamber(CCEE) and Eletrobras.

5. CURRENT STATUS OF DISTRIBUTION COMPANIES

Taking into account the position of June 30, 2017, the accrued net equity of all six distribution companies was negative by R$ 14.8 billion, already considering past and successive conversions of debt into capital. The total debt of these distribution companies is about BRL 30 billion. Out of this total, R$ 468.10 million are debts with the RGR fund, R$ 5.5 billion are debts with Eletrobras Ordinary Resources or RO, R$ 15.8 billion with Petrobras, R$ 4.2 billion of obligation of reimbursement to the CCC Fund, and the remaining balance are debts with banking institutions and related parties.

Table 11.Summary of Loans and Financings (Jun./2017) R$ million

SOURCE  RO + AFAC WORLD BANK   RGR   REIMBURSEMENT RELATED PARTIES  PETROBRAS  Total 
AVERAGE COST  12.00%  6,59%  5,00%  OBLIGATIONS CCC-  7.5% (SELIC)  7.5% (SELIC) 
AmD  1,483.25  138.50  107.50  3,910.09  406.00  12,472.30  18,517.64 
Boa Vista Energia  139.51  15.48  2.62  76.99  587.00  335.30  1,156.89 
Ceal  1,401.33  133.12  54.11  -  53.00  -  1,641.56 
Cepisa  1,423.42  128.39  233.35  -  31.00  -  1,816.16 
Ceron  700.32  105.99  63.18  164.56  2,369.00  2,546.60  5,949.64 
Eletroacre  357.89  49.70  7.36  81.61  -  461.80  958.36 
TOTAL  5,505.72  571.18  468.12  4,233.25  3,446.00  15,816.00  30,040.26 

 

Source: Eletrobras

Regarding the total loans and financings of the distribution companies with Petrobras (which includes debts connected to BR Distribuidora), the distribution companies Amazonas Energia, Boa Vista Energia, Eletroacre and Ceron executed, with Petrobras/BR, debt settlement agreement - DSA, which are based on the supply of fuel by Petrobras and BR Distribuidora. In these CCDs, Eletrobras is guarantor and, on the base date of 06/30/2017, the guaranteed balance was nearly R$ 3 billion.

Page 20


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

As it will be demonstrated in the tables below, Eletrobras’ energy distribution companies have been showing recurring losses in the last five years. Except for the year 2014, where Eletroacre, Cepisa and Ceron made a profit, in all other years, all companies presented losses, with 2016 being the worst year in history, with a total loss of R$ 7.0 billion. Out of this amount, 40% is explained only by the financial result that, year on year, has been aggravating the performance of companies, which ranged from R$ 950 million, in 2013, to R$ 2.8 billion, in 2016, as income. The financial result is highly impacted by the stock of debts of RO and RGR, but mainly due to the adjustment for inflation of the debt of fuel (oil and gas) with Petrobras/BR Distribuidora, which, in August 2017, amounted to R$ 15.8 billion.

Analyzing EBITDA, once again excepting 2014, in the other years, the set of six distribution companies obtained a negative cash generation. In other words, revenues generated from operating activities are not even enough to cover the costs and expenses connected to the companies’ operation. Thus, it can be concluded that the operational activity of the distribution companies does not generate a positive margin for the Eletrobrasdistribution companies, although there are subsidies from the CDE CCC Fund

Accrued losses over the last years have brought consequences to the net equity of the companies. Since 2011, Eletrobras’ distribution companies have been showing a negative net equity, even with the constant contributions made by Eletrobras, as already referred to. In December 2016, the equity of the six distribution companies was negative at R$ 13.3 billion; in June 2017, it already accumulates R$ 14.8 billion, which shows its constant deterioration.

The tables below show the evolution of some economic and financial indicators, considering the consolidated and individual position of the six distribution companies:

Table 12.  Consolidated financial highlights of the energy suppliers (R$ million) 
Financial Highlights  2013  2014  2015  2016  Jun./17 
ROL on Construction Rec.  5,234.5  7,297.1  6,185.2  6,617.3  3,383.4 
PMSO    1,945.2  1,648.8  1,932.1  1,881.3  1,082.9 
EBITDA    (327.1)  1,408.0  (1,235.8)  (822.6)  (360.4) 
Financial Income  (950.7)  (1,643.5)  (1,964.5)  (2,778.5)  (1,724.9) 
Net Earnings    (2,344.8)  (133.2)  (4,173.1)  (7,045.3)  (1,477.0) 
Net Equity    (2,323.5)  (2,082.3)  (6,236.2)  (13,309.1)  (14,805.0) 
Total Debt    3,275.5  5,151.4  4,854.6  6,579.2  8,214.3 
Source: Suppliers Financial Statements (individualized and audited).       

 

Page 21


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
  Table 13. Financial highlights of the energy suppliers (R$ million)   
R$ Million  Financial  2013  2014  2015  2016  Jun./17 
  ROL on Construction Rec.  287.5  327.2  364.0  369.7  226.4 
  PMSO  114.0  111.8  180.3  135.7  64.6 
  EBITDA  (13.5)  87.3  (55.9)  (9.5)  (3.3) 
ELETROACRE  Financial Income  (46.2)  (59.4)  (68.4)  (138.7)  (113.8) 
  Net Earnings  (154.7)  25.9  (184.8)  (143.9)  (133.6) 
  Net Equity  (201.2)  54.9  (129.9)  (273.8)  (407.4) 
  Total Financial Debt  156.3  233.6  283.2  369.4  439.9 
 
  ROL on Construction Rec.  1,722.9  3,220.1  2,339,3  2,386.1  1,219.3 
  PMSO  720.3  744.0  733.6  652.3  373.5 
AMAZONAS  EBITDA  38.4  458.5  (771.9)  (347.3)  54.0 
ENERGIA  Financial Income  (768.9)  (1,218.4)  (1,210.6)  (1,957.4)  (1,085.8) 
  Net Earnings  (1,364.5)  (342.7)  (2,344.0)  (4,967.8)  (768.7) 
  Net Equity  (1,676.8)  (2,019.4)  (4,363.6)  (9,334.6)  (10,103.3) 
  Total Financial Debt  1,187.8  2,164.4  1,327.2  1,987.2  2,583.5 
 
  ROL on Construction Rec.  817.0  1.248.9  1.263.8  1,089.8  675.8 
  PMSO  338.0  241.4  280.1  282.4  158.5 
  EBITDA  (66.4)  116.4  (1.9)  (106.8)  (19.2) 
CEAL  Financial Income  (25.7)  (81.4)  (104.5)  (118.9)  (128.5) 
  Net Earnings  (137.1)  (95,4)  (252,6)  (321,5)  (122.1) 
  Net Equity  (21.4)  (11.1)  (247.7)  (573.8)  (695.8) 
  Total Financial Debt  629.8  958.3  1,174.4  1,462.1  1,753.2 
 
  ROL on Construction Rec.  1,377.1  1,165.4  1,123.7  1,198.3  613.2 
  PMSO  376.2  304.0  389.0  383.0  266.5 
  EBITDA  (237.6)  258.3  (206.5)  (154.0)  (95.2) 
CEPISA  Financial Income  (69.7)  (169.3)  (144.6)  (222.8)  (213.8) 
  Net Earnings  (427.9)  37.9  (563.0)  (506.8)  (48.6) 
  Net Equity  (219.5)  (141.1)  (701.2)  (1,221.7)  (1,270.3) 
  Total Financial Debt  794.6  1,058.6  1,281.4  1,681.2  2,077.4 
 
  ROL on Construction Rec.  854.3  1,143.1  851.2  1,285.6  491.6 
  PMSO  304.4  297.9  256.7  328.2  157.5 
  EBITDA  (36.8)  506.6  (163.4)  (73.3)  (250.0) 
CERON  Financial Income  (44.3)  (103.1)  (365.4)  (250.6)  (149.4) 
  Net Earnings  (313.9)  292.7  (560.8)  (835.3)  (303.3) 
  Net Equity  (188.7)  104.1  (456.3)  (1,295.9)  (1,599.2) 
  Total Financial Debt  481.8  692.1  734.7  964.0  1,179.2 
 
  ROL on Construction Rec.  175.7  192.3  243.2  287.8  157.0 
  PMSO  92.3  90.0  92.4  99.9  62.3 
BOA VISTA  EBITDA  (11.2)  (19.1)  (36.2)  (131.7)  (46.8) 
ENERGIA  Financial Income  4.1  (11.9)  (71.0)  (90.2)  (33.6) 
  Net Earnings  53.3  (51.7)  (267.9)  (270.0)  (100.7) 
  Net Equity  (15.9)  (69.7)  (337.6)  (609.3)  (728.9) 
  Total Financial Debt  25.2  44.4  53.7  115.2  181.0 
Source: Individualized and Audited Financial Statements         

 

Page 22


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Global level of investments of Eletrobras’ energy distribution companies from the period 2009-2016 shows a growing curve in the first four years, up to 2012. After the publication of Provisional Measure 579/2012, converted into Law 12783/2013 and the associated reduction of the revenues of Eletrobras, the level of investment of the distribution companies entered a decreasing curve as for the financial incapacity of Eletrobras to maintain the contributions of resources in these companies. It also contributed to the shrinkage of the investment levels.

However, investments in the distribution business posted a strong increase in 2015, and continued with this trend in 2016, when the amount invested reached R$ 1.13 billion (highest nominal value in the period), R$ 861.2 million out of which in expansion, and R$ 274.8 million in maintenance of the electric distribution network, with funds coming from IBRD, connected to the Project Energia + and RGR, notably in the so-called post-designation period, in the wake of the above MME Ordinances.


According to this sense, the historical data of the last seven years on the operating indicators of Eletrobras’ distribution companies may be checked below and, as can be seen, out of the four indicators (DEC, FEC, Energy Losses and Active Default), only the index of loss on injected energy did not show a downward trend, remaining at levels close to those of 2010, greatly influenced by the performance of AmE D, which has the highest weight among the indicators. Losses of AmE D on injected energy starts the series by 42.5% in 2010, drops towards 37,6% in 2014, rising again to 43.1% in 2016.

Page 23


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017


Source: Suppliers Eletrobras

Table 16.  Info about Eletrobras’ energy distribution market   
Items 2010 2011 2012 2013 2014 2015 2016 Annual
Average
Growth 
Sold Energy (GWh)  12.781  13.579  15.216  16.092  17.112  17.513  17.312  5.3% 
Number of Clients (mil)  3.303  3.490  3.602  3.804  3.942  4.058  4.265  4.4% 
Distribution Network Extension    177.9  226.4  248.4  253.4  254.1  263.1  8,6% 
Source: Eletrobras Management Reports             

 

Based on the data presented, it can be concluded that regardless of the investments made around R$ 870 million (annual) to expand and improve the electricy distribution system, contributing to an average annual growth of 8.6% of the distribution network extension

Page 24


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

(2011-2016) and of 4.4% in the number of clients, it was not possible to improve the economic and financial results of such companies.

It is important to emphasize that since 2003, it has been observed a concern of Eletrobras’ management with the losses caused by the maintenance of companies under its control. However, it was not possible to solve this issue. As shown by the graphs above, progress was made, but the speed of improvements is not in line with the goals set out by the regulator or even with the amounts practiced by comparable companies, thus worsening the economic and financial crises of Eletrobras’ energy distribution companies..

5.1 RIGHTS AND OBLIGATIONS WITH CCC FUND

The financial and economic unbalance of the distribution companies operating in the Northern region of Brazil over the last years was found to be strongly impacted by the generation cost of oil-powered thermal power plants. Law 12111 of December 9, 2009, which provides for the electric energy services in Isolated Systems, sets forth that part of such costs5 is reimbursed by the Fuel Consumption Account - CCC. Following the homologation of Act 12783 of January 11th, 2013, conversion of Provisional Measure 579 of September 11th, 2012, it was established that the granting to CCC would be made by the Energy Development Account - CDE. However, resources from CCC/CDE were not sufficient to achieve the financial and economic balance of such distribution companies. In view of this, there was a growing accumulation of past due debts since the distribution companies did not have the financial conditions to pay the fuel purchased from Petrobras and BR Distribuidora, considerably increasing the gas and oil debt of these companies over recent years.

The debt of AmE D, Boa Vista Energia, Eletroacre and Ceron with Petrobras and BR Distribuidora updated as of August 2017 reached the figure of R$ 15.8 billion. Out of this total, R$ 10,7 billion are negotiated under debt settlement agreement (DSA) with Petrobras and BR Distribuidora, signed in December 2014, in the payment conditions of 120 installments adjusted based on Selic, which currently, according to the reference date, has 89 installments not matured. Part of this fuel debt is backed by CCC receivables, also

5 Total cost of thermal generation in the isolated system above the average energy prince in the interconnected system is reimbursed by CCC. Concerning this right to reimbursement, ANEEL adopts a cutting factor on the difference between the effective energy losses of the company and the level of regulatory losses.

Page 25


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

substantiated in CCD, also signed in 2014, between AmE D, Ceron, Eletroacre and Boa Vista and the CDE/CCC Fund, based on express normative authorization.

Table 17. Debt of distribution companies with Petrobras/BR (BRL thousand)

Distribution company CCD Debts
Contracted 
Open Debt Total
Amazonas Energia  -9,022,500.00  -3,449,800.00  -12,472,300.00 
Ceron  -1,241,500.00  -1,305,100.00  -2,546,600.00 
Eletroacre  -301,600.00  -160,200.00  -461,800.00 
Boa Vista  -162,700.00  -172,600.00  -335,300.00 
Total  -10,728,300.00  -5,087,700.00  -15,816,000.00 

Source: Eletrobras

It should be noted that ANEEL is under the process of inspecting the amounts released by CCC to the distribution companies during the period from 2009 to 2016, with unfavorable results already presented to the Company, which has a direct impact on the CCD contracts executed between the CDE / CCC Fund and distribution companies Amazonas Energia, Ceron, Eletroacre and Boa Vista, as can be seen in ANEEL Order 2504/2017, rectified on 10/27/2017 (AmE D) and Technical Notes SFF-SFG-SRG/ANEEL no. 110/2017 (Ceron), no. 90/2017 (Eletroacre) and no. 177/2017 (Boa Vista).

Table 18.  CCC: ANEEL Supervised Amount x Balance Sheet Record Jun./2017 

 

Distribution company  ANEEL Supervised
Amount1 
Balance Sheet
Record 2 
Amazonas Energia  -2,906,095.46  4,055,549.00 
Boa Vista Energia  5,666.71  278,360.00 
Ceron  -731,570.40  3,847,293.00 
Eletroacre  -276,550.90  296,167.00 
Total  -3,908,550.05  8,477,369.00 
Source: 1 ANEEL Technical Notes; 2 Intermediary Financial Statements of suppliers as of 30/06/2017. 

 

Based on the table above, it can be observed that ANEEL's inspection established a value to be refunded to the CCC/CDE fund by all distribution companies, except for Boa Vista Energia, which had its debt claim diminished and, if this understanding prevails, the value currently booked as debt claim in the distribution companies’ balance sheets, can become

Page 26


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

a liability of R$ 4 billion in the sum of the four distribution companies, resulting in a potential loss of up to R$ 12.4 billion.

In this sense, it should be noted that the debt claims, which may be assumed by Eletrobras, at the discretion of the General Meeting, according to the modeling of BNDES approved by CPPI, are under discussion within the Administrative Proceedings at ANEEL, AmE D (Proceeding nº 48500.004972/2016-34), Ceron (Proceeding no. 48500.005104/2016-71), Eletroacre (Proceeding no. 48500.005103/2016) and Boa Vista (Proceedings no. 48500.000588/2017 and 48500.000587/2017), with the dispute of the results of the inspections performed by the Agency with respect to the movement of funds of the Fuel Consumption Account CCC earmarked for the said companies. Therefore, its realization depends on the decision of ANEEL, with the possibility of judicial treatment in the case of administrative setbacks with ANEEL.

It must be observed that, in the hypothetical event that the Regulatory Agency’s understanding is not reformed, while maintaining the inspection’s conclusion that the companies are indebted to the CCC Fund, except for Boa Vista Energia, such debt claims may be written off from Eletrobras’ Assets, and it is also necessary to recognize a liability of R$ 4 billion, which is not recognized in connection with the Administrative Appeal, namely:

(a)      AmE D - BRL 2,906,095.46;
(b)      Ceron - BRL 731,570.40; and
(c)      Eletroacre - BRL 276,550.90.

For EletrobrasThe riskconnected to the assumption of these debt claims with the CCC Funds is the occurrence ofa potential loss of up to R$ 12.4 billion, impacting the Company's income and equity, when and in case the administrative and/or judicial discussion has an unfavorable outcome to Eletrobras.

Despite the information described above, it is worth mentioning, according to the legal opinion prepared by the law firm Décio Freire and Associados, dated September 18, 2017, referring to Administrative Proceeding no. 48500.004972/2016-34 above (AmE D), that, to the best of its knowledge, while there is no decision of the ANEEL’s Board wholly or partially ratifying the conclusions set forth in ANEEL Order No. 2.504/2017, there is no administrative act that implies an obligation to reimburse the fund. Additionally, the opinion considers that there is a solid legal basis for the review of ANEEL’s rulings and

Page 27


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

ultimately states that it is too early to provide for any potential loss connected to said Administrative Proceeding or to disclose the matter in questions in explanatory notes to the Financial Statements, as reproduced below:

Under all of the circumstances described above, we believe there is a solid legal basis for reviewing the provisions under the Technical Notes numbers 52/2017-SFF-SFG-SRG and 141/2017-SFF-SFG-SRG, as well as a manifest failure to comply with decisions rendered by Federal Regional Court of the 1st Region, in the records of Writ of Mandamus no. 0026107-81.2012.4.01.3400.

This way, we understand that it is too early to provide for any potential loss connected to the Administrative Proceeding in question, or to disclose the matter in explanatory notes to the Financial Statements.”

It should be noted that CPPI Resolution 28/2017 rectified CPPI Resolution no. 20/2017, in its Article 3, Paragraph 1, which shall read as follows:

"In addition to the adjustments provided for in the head provision of this article, Eletrobras may, at the discretion of the General Meeting, undertake the rights and obligations of the distribution companies, referring to the Fuel Consumption Account - CCC and the Energy Development Account -CDE, recognized in the Financial Statements of the Distribution companies at the base date of the studies considering the adjustments up to June 30, 2017, including the rights connected to Item IX of Article 13 of Law 10438, dated April 26, 2002, with the new wording given by Law 13299, dated June 21, 2016, and Eletrobras shall, on the other hand, undertake the rights and/or obligations in an equivalent amount.” (emphasis added)

Accordingly, CPPI authorized the General Meeting of Eletrobras to decide whether or not to assume the rights and obligations connected to the CCC/CDE recognized in the distribution companies’ Financial Statements, dated June 30, 2017, amounting to R$ 8.47 billion, detailed in tables 17 and 18. And, therefore, the debts that may be assumed are limited to such value.

Page 28


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

6. MANAGEMENT OF THE PROCESS OF PRIVATIZATION OF THE DISTRIBUTION COMPANIES

BNDES, by means of Article 2 of Decree 8893/2016, was designated as responsible for the performance and follow-up of the process of privatization of the distribution companies of Eletrobras, with the following competences:

(i) disclosing and providing, as the case may be, the relevant information exclusively connected to the process of privatization referred to in the head provision, including for meeting the demand of the Board of the Partnership and Investment Program of the Presidency of the Republic and the other competent powers;

(ii) carrying out the engagement of consulting company, audit and other specialized services necessary for the performance of privatizations;

(iii) promoting the articulation with the system of distribution of securities and the stock exchange;

(iv) choosing and registering companies of renowned repute and traditional operation in the trading of capital, transfer of controlling interest, same and lease of assets; and

(v) preparing the documentation of processes of privatization for appraisal of the Federal Court of Accounts.

6.1 COSTS WITH THE PRIVATIZATION PROCESS

Decree No. 2,594, of May 15, 1998, which puts forth regulations relating to Law No. 9,491/1997, establishes that, abiding by the standards approved by CNB, expenses with third-party services incurred by the management body directly or indirectly responsible for executing and monitoring the privatization process shall be reimbursed by Eletrobras, the holder of the Brazilian Depositary Receipts ("BDR").

BNDES, in keeping with the provisions under Decree 8893/2016, by means of SUP/AD Letter 01/2017, has submitted, for appraisal by Eletrobras, a contract draft covering the remuneration and reimbursement owed to the Bank, as responsible for the performance and follow-up of the process of privatization of the distribution companies of Eletrobras.

Page 29


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

The execution of an agreement about the remuneration and reimbursement owed to BNDES was approved by the Board of Executive Officers and the Board of Directors of Eletrobras by Resolution 54, dated January 24, 2017, and Deliberation 28, dated January 25, 2017, respectively.

6.2 REMUNERATION OF BNDES

According to the article 25 of Decree No. 2,594/1998, for carrying out the role of manager, BNDES, Manager of the FND, shall be entitled to remuneration of two tenths one percent (0.2%) of the net value calculated on the sale of the shares, minus expenses incurred with third parties to cover its operating costs.

According to the remuneration and reimbursement agreement executed between BNDES and Eletrobras, the compensation of BNDES (Manager of FND) shall be paid when the privatization process of the last distribution company is settled. In addition to this, Eletrobras is responsible for paying the necessary services engaged by BNDES for the privatization process, amounting up to BRL 34,935,570.42. Lastly, in the event of the sale of any of the distribution companies, Eletrobras shall pay BNDES the amount of BRL 318,000.00, adjusted by IPCA as of the date of execution of the remuneration contract, with regard to each distribution company that has not been sold.

Therefore, it should be noted that Resolution 20/2017 of CPPI, which approves the transfer of the controlling interest of the distribution companies held by Eletrobras, in its Article 18 established that the remuneration of BNDES, as well as the reimbursement of expenses performed with third party, shall be as agreed with Eletrobras.

However, it’s worth mentioning that Paragraph 2, Article 11 of Decree 9192/ 2017 establishes that:

“Article 11. BNDES shall pay a remuneration for the performance of its activities, in accordance with the provisions of this Decree, as well as reimbursement of expenses incurred with third party.

Paragraph 1 The payments referred to in the caput shall be the responsibility of the controlling shareholder, as the legal entity responsible for the service of electric energy distribution or transmission.

Page 30


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Paragraph 2 The notice may establish that the bidding winner referred to in Article 1 and Article 9 makes the payments referred to in the caput.”

7. CONDITIONS FOR THE DISPOSAL OF CONTROLLING STAKE

As provided for in Article 15 of Law 9491/1997, the general conditions for privatization and the minimum price shall be defined by CND. Given that CPPI succeeded CND, the conditions and minimum price shall be defined by CPPI.

The minimum prince, according to Article 31 of Decree 2594/1998, is based on economic and financial assessments prepared by independent consulting firms engaged by BNDES. Note that, in the case of difference between the values ascertained by the consulting companies in excess of 20%, CPPI is entitled to determine the engagement of an outsourced assessor, to state its position, in up to sixty days, about the assessments, event in which the respective assessment shall also base the ascertainment of the minimum price.

Therefore, CPPI published Resolution no. 20, dated November 8, 2017, as amended by Resolution no. 28, dated November 22, 2017, defining the conditions of sale, as well as the minimum price, which in this case will be at the set price. The most relevant items include the following:

(i) Sale of common and/or preferred shares held by Eletrobras of at least 1(one) common share of the capital of the distribution company, at the total amount of BRL 50,000.00, in each of the six Eletrobras’ distribution companies ;

(ii) Eletrobras shall remain the holder of one (1) common share in each distribution company;

(iii) The payment for the shares held by Eletrobras shall be made in cash and in local currency;

(iv) The Eletroacre’s minority shareholders shall be entitled to sell their shares to the new controlling shareholder, under the same conditions and prices paid for Eletrobrasshares;

 

Page 31


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

(v) Eletrobras shall, before the transfer of controlling interest, make adjustments in the distribution companies, upon conversion of capital or assumption of the debts of the distribution companies with Eletrobras and/or third parties, according to the numbers presented in Table below;

Table 19.  Investments made in distribution companies by Eletrobras 
Distribution company  Adjustments (BRL) 
AmE D  8,911,866,558.94 
Boa Vista Energia  342,120,486.20 
Ceal  50,000.00 
Cepisa  50,000.00 
Ceron  1,872,522,463.42 
Eletroacre  113,779,871.99 
Total  11,240,389,380.55 

 

(vi) At the discretion of the General Meeting, Eletrobras may undertake the rights and obligations of the distribution companies, related to the Fuel Consumption Account - CCC and the Energy Development Account - CDE, recognized in the Financial Statements of the Distribution companies on the base date of the studies, considering the adjustments up to June 30, 2017, including the rights connected to item IX of article 13 of Law 10438, dated April 26, 2002, with the new wording given by Law 13299, dated June 21, 2016, and, in such respect, Eletrobras should undertake, on the other hand, in a commutative manner, the rights and/or obligations at an equivalent value;

(vii) AmE D maysettle the debts with Eletrobras upon transfer of all the shares issued by Amazonas Geração e Transmissão S.A Amazonas GT in favor of Eletrobras and/or third party, with this value deducted from the adjustment for AmE

D, defined in Item (v);

(viii)      The unbundling of Amazonas GT shall be completed until March 2, 2018; 
 
(ix)      In the event of the Ceal’s privatization, the publication of the public notice will

also be conditional on the execution and judicial homologation of an agreement concerning the payment of salary differences arising from the Bresser Plan, based on

Page 32


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

from the sale of the distribution companies’ shares, after the applicable deductions;

(x) The notice of privatization shall define the amount of shares of each distribution company to be offered in the auction of privatization, and each lot shall be acquired in full by the same purchaser;

(xi) The new controlling shareholder shall perform a new increase of capital of the distribution companies, upon subscription and payment, according to the minimum amounts presented in the next Table;

Table 20. Increase of capital on the distribution companies by the new controlling shareholder

Distribution company  Capital Increase (R$) 
AmE D  491,370,787.84 
Boa Vista Energia  175,999,185.71 
Ceal  545,770,485.33 
Cepisa  720,915,595.51 
Ceron  241,099,855.91 
Eletroacre  238,805,729.30 
Total  2,413,961,639.60 

 

(xii) Offering to employees and retirees, qualified according to notice of auction, of at least ten percent (10%) of all shares held by Eletrobras prior to the sale to the winning bidder, with an approximate discount of ten percent (10%) of the minimum price per share;

(xiii) The new controlling shareholder shall compensate Eletrobras for the amount of the discount of the shares purchased by the employees and retirees;

(xiv) The shares not purchased by the employees and retirees shall be compulsorily purchased by the new controller, so that Eletrobras will remain with only one (1) share;

(xv) The approval of the transfer of controlling interest by ANEEL and by the Administrative Council for Economic Defence - CADE shall be conditioned to:

Page 33


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

  a.      perform adjustments as provided for in Item (v), except for the adoption of the measures stated in Item (vii); 
 
  b.      financial settlement of the transfer of the controlling interest of the distribution companies by the winning bidder; 
 
  c.      subscription and payment of the increase of minimum capital of the distribution company by the new controlling shareholder; and 
 
  d.      financial settlement of the purchase of shares offered by Eletrobras to the employees and retirees of each distribution company. 
 
(xvi)      Opening of a data room before publication of the notice; 
 
(xvii)      Option for Eletrobras to increase its interest in the distribution companies

capital at up to 30%, which may be through the conversion of the debts that Eletrobras still holds against the distribution companies, in capital, according to the values in Table 30. The term for exercising the option shall be up to six (6) months, as of the date of execution of the purchase agreement; and

(xviii) Remuneration of BNDES under Decree no. 8893/2016 and Decree no. 9192, dated November 6, 2017, as well as reimbursement of expenses incurred with third party, in the manner contracted with Eletrobras.

8. DISPOSAL MODELING OF THE CONTROLLING STAKE

As already referred to, BNDES was responsible for engaging the consulting, audit and other specialized services required for performing the privatizations.

Thus, two independent consulting firms were engaged to perform the assessments of economic and financial, technical, legal, sale structure modeling thematic areas, without limitation, which were the following: Ceres Inteligência Financeira Ltda. (Ceres), which performed the economic and financial assessment (service A), and the Mais Energia B Consortium (Service B), comprising the companies Price Waterhouse Coopers Corporate Finance & Recovery Ltda., Siglasul Consultoria Ltda. and the law firm Loeser e Portela Advogados, which performed the other assessments, including the economic and financial and the modeling of the structure of sale.

Page 34


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

8.1 METHODOLOGY ADOPTED FOR THE ECONOMIC AND FINANCIAL ASSESSMENT

 

The methodology applied in the economic and financial assessments was the Discounted Cash Flow (DCF), as defined in Article 30 of Decree 2594, dated May 15, 1998. This methodology is based on the hypothesis of the value of a project to depend on its capacity of generating wealth in the future. Estimates of revenues, costs, expenses, investments and other working capital needs, in addition to all items that affect the cash flow of the project, for a certain period of time.

As the values of the generated cash flow are given in different times, they should be summed and compared at the same instant of time (present value). Thus, the net balances calculated in each period are brought to present value by a discount rate that reflects the risks inherent to the business added to the opportunity cost of risk-averse investors, that is to say, a rate of attractiveness that will reflect the opportunity cost of the business capital providers.

The discount rate adopted by the consultants is explained in the studies and was ascertained by parameters of the values generally offered by government bonds , which incorporate a specific risk of the activity and is calculated based on the model of assessment of assets, the Capital Asset Pricing Model (CAPM). By means of this model, the cost of own capital (Ke) of the project is calculated.

For project value calculations, the cost of equity (KE) and the cost of third-party capital (Kd) are weighted in proportion to the capital structure used in the project, resulting in the WACC (Weighted Average Cost of Capital) of the project. Thus, the assessments adopted the criteria of nominal Discounted Cash Flow (DCF), having the December 31st, 2016, as the base date. The discount rate used was WACC (Weighted Average Cost of Capital).

The assessments were based on the grant of a new concession over the period of 30 (thirty) years, as of March 1, 2018, until February 29, 2048.

Besides that, the two consulting companies engaged performed the economic and financial assessments of the distribution companies based on the regulatory framework for the distribution sector, which encompasses the tariff reviews, with five-year periodicity, and the annual tariff adjustments between cycles, and on the economic cornerstone of the actual forecasts, especially those connected to the costs incurred and the invested

Page 35


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

capitals. Also the costs and expenses, taxes, associated working capital, estimated indemnities and contingencies and actual disbursements in the composition of the business flow were considered. The assessments have also considered the results of the accounting and legal due diligences performed by Service B.

From the regulatory point of view, as a result of the Public Hearings 094/2016 and 032/2017 of ANEEL and the MME Public Query no. 37/2017, in the economic and financial assessments, several relaxations of regulatory parameter were introduced, thus allowing conditions to rebalance the concession. Such relaxations, as well as the characteristics and peculiarities of the Brazilian Electricity Sector, were also considered in the sale modeling.

It’s worth mentioning that the flexibilization of these parameters will be applied in the first five years of the concession, including adjustments in the regulatory curves of losses and PMSO, and after the fifth year, the amortization of loans with RGR resources obtained during the designation period shall be considered as well.

In addition to the economic and financial assessments, the consulting companies also performed the assessment by market multiples, which is used to improve the robustness of the analyses about the sum ascertained, based on the relation between the market values or those of the similar transactions and operating indicators. This approach is based on the idea that similar assets have similar future returns.

9. DISPOSAL MODELING OF THE CONTROLLING STAKE

The modeling of privatization approved by CPPI, through the Resolution no. 20/2017, and established in “Product 11: Proposal of Modeling of Privatization of each Distribution Company, attached to this IT, combines corporate aspects, the manner of capitalization, the process of sale, aiming to increase the chance of success in attracting investors, as well as offer the maximum competitiveness to arrange bests conditions of offer to be received.

The modeling, in accordance with the paragraph 1 of Article 31 of Decree 2.594, dated May 15, 1998, is based on the assumption that the difference between the enterprise value calculated by the two consulting companies is smaller than 20%, as, if that weren’t so, CND would be entitled to order the outsourcing of an assessor to state its positioning, in up to sixty days, about the assessments, event in which the respective assessment shall

Page 36


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

also serve as basis for the ascertainment of the minimum price, which shall be a fixed price, in this case.

For the calculation of the percentage difference, the following metric was used. The table below shows the amounts of percentages for each distribution company.


Table 21.  Enterprise value and difference Services A and B 
Distribution company Enterprise Value  Difference (%)
Service A  Service B 
Amazonas Energia  2,687,563,915.43  2,223,734,395.42  17.26% 
Boa Vista Energia  565,617,124.98  611,726,317.51  8.15% 
Ceal  1,994,373,551.48  2,446,562,300.47  18.48% 
Cepisa  2,476,185,262.79  2,880,719,840.98  14.04% 
Ceron  1,524,512,239.58  1,264,573,655.87  20.56% 
Eletroacre  921,169,419.22  944,913,418.27  2.51% 

Source: Eletrobras preparation, based on the economic and financial assessments, Service A and Service B.

Given the foregoing assumption, according to the assessments in Products 11 attached, the base value of the valuation is calculated from the simple average of the equity values resulting from the economic and financial assessments performed by Service A and Service B. The table below shows the starting equity value of each distribution company.

Table 22.  Equity value and average of assessments Services A and B 
Distribution company  Equity Value  Average of
Assessments 
Service A  Service B 
Amazonas Energia  -8,431,847,736.90  -8,895,677,256.91  -8,663,762,496.91 
Boa Vista Energia  -328,272,159.71  -282,172,937.18  -305,222,563.44 
Ceal  -1,116,190,768.34  -664,002,019.35  -890,096,393.85 
Cepisa  91,400,895.59  495,935,473.78  293,668,184.69 
Ceron  -1,636,209,285.18  -1,896,147,868.89  -1,766,178,577.04 
Eletroacre  -159,107,674.07  -135,363,675.02  -147,235,674.55 

Source: Eletrobras preparation, based on the economic and financial assessments, Service A and Service B.

Page 37


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Besides this, the modeling, aiming to achieve an equity value equal to zero, makes certain adjustments, which are in the individual analysis of each distribution company, which are Appendices to this IT.

Another premise adopted by the modeling is that the distribution companies will have the sale value set at R$ 50,000.00, for Eletrobras.

In general, the modeling proposed consists in the holding of auction jointly with the grant of a new concession and the disposal of the controlling stake of the distribution company. This way, a new concession is granted for the public service of electricity distribution to the distribution company itself, and the controlling interest is transferred to the new controlling shareholder.

The auction, per distribution company and sequential, shall be performed in two stages, as follows:


Source: Product 11: Proposal of Modeling of Privatization, Service B.

Page 38


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

In stage 1, in order to enable the privatization and increase the attractiveness of the business, the capital structure of the distribution companies should be adjusted, so that the shares have at least a symbolic value in their transference. Additionally, the sale of shares of the distribution companies to the investor is made together with the offering to employees and retirees, which shall be discussed further on. In this stage, Eletrobras remains with one share, for future decision (stage 2) of continuing or not in the partnership.

In stage 2, there is estimate for the subscription of shares by the new investor, through capital contribution to be made on a date concurrent to the assumption of control. It should be noted that, at this stage, Eletrobras is given the possibility to continue as a shareholder of the distribution company, increasing its interest at up to 30% of the total shares, within a period of six months after the purchase and sale contract. If Eletrobras chooses not to increase its shareholding, after the term of six months, this 1 share is sold to the new investor.

In addition to this, the auctions of the distribution companies shall be sequential, and shall be performed in B3 stock exchange, which is the substitute of BM&FBOVESPA. In the first two auctions, there will be a Right of Participation in the second stage of subsequent auctions. As of the 2nd until the 5th auction, the investors that have won at least one preceding auction, may remove their envelopes from the auction before opened.

Lastly, the auction winner shall be the investor that places the bid with the highest deduction of the tariff provided for each distribution company. If the reduction is higher than one hundred percent (100%), the difference between the bid and one hundred percent (100%) shall be considered as bonus of grant, to be paid by the winning bidder to the Federal Government.

9.1 ADJUSTMENTS TO THE EQUITY VALUE OF THE DISTRIBUTION COMPANIES

According to the modeling reports of Service B, the modeling of sale was performed based on the results of the due diligence reports and the economic and financial assessments performed by the consultants. However, for the final proposal of capital and corporate structuring of the distribution company, additional adjustments, already included in the modeling, were necessary. The adjustments made follow the stages below.

Page 39


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Table 23. 

Adjustments performed to the equity value 

 
 = Average of Services A and B 
 (+/-) Consolidated Adjustments 

Balance Adjustment 

 

(+/-) Adjustments of Assets and Liabilities 

 

(+) Reincorporation of RGR PPST 

 

(-) Re lassifi ation of AFAC’s as De t 

 

(-) Tax adjustments (Tax Loss and Negative Base) 

= Adjusted Equity Value 
(-) Adjustment of the reduction of tariff relaxation 
= Final Equity Value 
Source: Product 11: Proposal of Modeling of Privatization, Service B. 

 

9.1.1 ADJUSTMENTS TO CONTEMPLATE THE UPDATES IN THE BALANCE SHEET UNTIL JUNE 2017

The sale modeling considers, as a way to mitigate any financial divergences between the base date of valuation (December 2016) and the last balance available of the distribution companies (June 2017), updates connected to the balance accounts connected to the indebtedness and the working capital, in addition to specific accounting adjustments.

The breakdown of the adjustments is under Products 11: Proposal of Modeling of Privatization, attached to this IT.

a) Advance for Future Capital Increase - AFAC

For the case that the distribution company has AFAC, another adjustment is made, and the amounts connected to the AFAC’s considered obligations with Eletrobras. The table below shows the AFAC’s, per distribution company, adjusted until October 2017.

Table 24.  AFAC values per distribution company 
Distribution company  AFAC (BRL) 
Amazonas Energia  128,147,702.47 
Boa Vista Energia  89,095,828.38 
Ceal  178,053,088.37 
Cepisa  344,361,704.77 
Ceron  - 
Eletroacre  82,452,860.46 
Total  822,111,184.45 

 

Page 40


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

b) Tax adjustments tax loss and negative base

The economic and financial assessments of the distribution companies considered the use of part of the Tax Losses (Income Tax IRPF) and Negative Base (Social Contribution on the Net Profit CSLL) in the first half of 2017 for settling debts, admitted under the scope of the Tax Regularity Program PRT. For the modeling of sale, such values were deducted from the fiscal balances existing in December 2016, base date used in the economic and financial assessment.

9.1.2 ADJUSTMENT OF THE FLEXIBILIZATION OF REGULATORY PARAMETERS

In the case of the distribution companies Ceal, considering the scenario of the agreement of the Bresser Plan adopted in CPPI Resolution 20/2017, and Cepisa, the equity value was positive due to the regulatory relaxations granted, reason why the relaxed parameters were adjusted so that equity value of these distribution companies was equal to zero. Therefore, this excludes the possibility that consumers in the concession areas in question will have to bear special tariff adjustments in addition to those necessary for the companies holding the concession to operate in financial balance.

9.2 STOCK OFFERING TO EMPLOYEES AND RETIREES

The stock offering to employees and retirees follows the provisions of Article 38 of Decree 2594/1998, which ensures to employees and retirees the offering of part of the shares representing the capital of the companies included in PND. Nonetheless, in accordance with the sole paragraph of this article, the said offering shall be at least ten percent of the shares of the capital directly or indirectly held by the Federal Government.

Thus, the sales modeling proposed contemplates such legal provision, and the offering is performed concomitantly to the auction of privatization, however, its implementation is conditioned to winning the bidding procedure, i.e., upon actual transfer of controlling interest.

The modeling also establishes that, to negotiate the stocks, the employees and retirees shall need to be qualified, and engage a custody agent, institution authorized and accredited at B3, the new name of BM&FBOVESPA. This definition is based on Article 39 of Decree 2594/1998, which establishes that the interest of employees in the purchase of

Page 41


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

shares shall be optional, by means of the investment club referred to in Decree 2430, dated December 17, 1997, established to legally representing them, including as substitutes in proceedings.

In addition to this and in accordance with the paragraph 4 of Article 30 of said Decree, the modeling of sale provides for that the shares to employees and retirees and public in general may have prices and conditions different from those object of the disposal of controlling stake. This way, employees and retirees will have the prerogative to purchase the shares at a price lower than the price to be paid by the new controlling shareholder of the company, under the scope of the auction.

After the offering stage to employees and retirees, the discount offered to employees and retirees shall be compensated to Eletrobras by the investor as well asstocks that have not been sold.

9.3 SHAREHOLDING OPTION OF ELETROBRAS

As mentioned above, the modeling of sale of the distribution companies provides for the possibility that Eletrobras increases its interest up to 30% of the total of shares within a term up to six months after the auction. Such prerogative is aimed at allowing that Eletrobras recovers part of the investments made through the receipt of the dividends or the future increase of value and subsequent sale of its shareholding.

For this case, it is suggested that Eletrobras has the right to increase its shareholding in the distribution company after the privatization of the company, with the alternative not only of capital disbursements, but also the additional conversion of debt into shares. Therefore, Eletrobras does not necessarily need to disburse capital to the distribution company if it intends to exercise such option, using the remaining debt credits that it may have against the privatized company.

In the event that Eletrobras has an interest in increasing its interest, legal and usual procedures for capitalization of the company should be followed, considering the rights of minority shareholders including employees and retirees that purchase shares of the distribution company. The financial conditions of the capitalization, as such, for the value of market per share, shall be the same as those of stage 2 and performance of auction.

Page 42


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

It should be noted that, should Eletrobras increase its interest in the distribution company, the modeling recommends that the Holding maintains minimal roles in the management and governance of the distribution company. Thus, the attractiveness of the sale of the distribution company would not be reduced as a result of the sharing of its management, allowing the widest degree of freedom to the new investor.

It should be noted that, at this time, there are no elements to support the decision regarding the continuity of Eletrobras in these companies with an interest of up to 30% of the new capital, as provided for in CPPI Resolution no. 20/2017, given that the ascertainment of the forecast of the return that the new shareholding may warrant to Eletrobras depends on variables not explained in the modeling performed BNDES.

Additionally, there are determinant variables for the decision-making process connected to the new disbursement of capital, such as identification of member and which its efforts and terms for economic recovery of the company shall be, in such a way that Eletrobras’ decision to make new disbursements, upon conversion of the remaining debts in each one of the companies shall be subject to new analysis, to be timely performed.

Even though, the values to be capitalized by Eletrobras in stage 2 should be emphasized, as pointed out in the assessments conducted by BNDES, in the case of increase of its interest in the distribution companies to 30%. The numbers are presented in Table 30 hereunder.

9.4 REQUIREMENTS FOR COMPLETING THE DISPOSAL OF THE CONTROLLING STAKE

The requirements for completing the operation of sale of distribution companiesstocks are as follows:

(i) approval of the transaction by Eletrobras’ General Meeting, pursuant to Article 41, I of Eletrobras' Bylaws, and according to the provisions authorized by the respective CPPI concerning the conditions of privatization of distribution companies;

(ii) abidance by the terms set forth in Decree 9192/2017, within fifteen days from its publication, which was rendered official by Eletrobras through Letter CTA-PR 558/2017, dated November 21, 2017, sent to the Ministry of Mines and Energy;

Page 43


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

(iii)      submission of the Federal Court of Accounts - TCU; 
 
(iv)      publication of Notice of auction; 
 
(v)      execution of the auction, under the terms of CPPI Resolution and in keeping with

the approval by the General Meeting of Eletrobras and TCU; homologation of the result of the auction; and adjudication on the object of the auction by the Ministry of Mines and Energy;

(vi) receipt of the governmental and regulatory authorizations necessary for the acquisition by the winning bidder of the controlling interest in the distribution company, namely: a) from the Administrative Council for Economic Defense and b) ANEEL;

(vii) execution of the Share Purchase Agreement of the distribution company, payment of the price by the winning bidder, and transfer of the shares in the respective corporate books of the distribution company;

(viii) performance of stock option to the employees and retirees of the distribution company, under the terms of Law 9491/1997 and CPPI Resolution; and

(ix) After the transfer of the shares, submission to ANEEL of the documents provided for in article 27 of Law No. 8987/1995 and ANEEL Normative Resolution 484/2012.

In addition to this, CPPI Resolution 20/2017 establishes that the new controller, as mentioned above, shall, concurrently with the purchase of the controlling interest of the distribution company, make a new capital increase in the distribution companies, through subscription and payment, according to the following minimum values:

(i)      Eletroacre: R$ 238.8 Million; 
 
(ii)      Ceron: R$ 241.1 Million; 
 
(iii)      Cepisa: R$ 720.9 Million; 
 
(iv)      Ceal: R$ 545.8 Million; 
 
(v)      Boa Vista: R$ 176.0 million; and 
 
(vi)      Amazonas Distribuidora: R$ 491.4 Million.

Page 44


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

10. ANALYSIS OF RESULTS PER DISTRIBUTION COMPANY

This item presents, in a summary form, the results of the economic and financial assessments by the distribution company, after the adjustments referred to in Item 8.3 of this IT were made, according to the case of each distribution company.

The table below, prepared based on the modeling proposed by Service B Product 11:

The Proposal of Modeling of Privatization” of each distribution company show the basic valuation amounts, the adjustments, and the ending valuation.

Table 25.  Basic valuation, adjustments and ending valuation 
Distribution company  Basic valuation (BRL)  Adjustments (BRL)  Final Valuation (BRL) 
Amazonas Energia  -8,663,762,491.91  -248,054,062.03  -8,911,816,558.94 
Boa Vista Energia  -305,222,563.44  -36,847,922.75  -342,070,486.20 
Ceal Scenario 1   -890,096,393.84  890,096,393.84  0.00 
Ceal Scenario 2  -3,563,184.05  -893,659,577.89 
Cepisa  293,668,184.68  -293,668,184.68  0.00 
Ceron  -1,766,178,577.04  -106,293,886.39  -1,872,472,463.42 
Eletroacre  -147,235,674.55  33,505,802.55  -113,729,871.99 

Source: Own preparation, based on Product 11: Proposal of Modeling of Privatization of each distribution company, Service B.

The final Valuation value shown in the table above shows the adjustment value required by Eletrobras to bring the Equity Value to zero. However, since the set value of the distribution companies’ sales is defined at R$ 50,000.00, the total adjustment amount per distribution company, except CCC debt claims and debts, according to CPPI resolution, was as follows:

Page 45


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

As can be observed, the BNDES study considered two scenarios for CEAL, one considering the agreement in grieving regarding the Bresser Plan, and another not considering it The PPI, according to CPPI Resolution 20/2017, opted for the scenario with the agreement, thus defining the need for an adjustment of R$ 50,000.00 in CEAL. This subject will be further addressed hereunder.

The following tables present the impact on Eletrobras arising from the capitalization and assumption of debts, in the following CCC debt claim scenarios: (i) included in the distribution companies’ balance sheet as of 06/30/2017; and (ii) resulting from the inspection carried out by ANEEL:

Table 26.  Capitalization/Assumption of debts Stage 1 (BRL million balance 
sheet
Distribution Companies Capitalization/
Assumption of
Debts 
Assumption of
Debt Claims of
CCC 
Impact on
Eletrobras 
Amazonas Energia  12,967.42  4,055.55  8,911.87 
Boa Vista Energia  620.48  278.36  342.12 
Ceal  0.05  0  0.05 
Cepisa  0.05  0  0.05 
Ceron  5,719.82  3,847.29  1,872.52 
Eletroacre  409.95  296.17  113.78 
Total  19,717.76  8,477.37  11,240.39 

Source: Own preparation, based on Product 11: Proposal of Modeling of Privatization of each distribution company, Service B.

Page 46


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
Table 27.  Capitalization/Assumption of debts Stage 1 - R$ million ANEEL 
Inspection
Distribution Companies Capitalization/
Assumption of
Debts 
Assumption of
Debt Claims of
CCC 
Impact on
Eletrobras
Amazonas Energia  12,967.42  -2,906.10  15,966.20 
Boa Vista Energia  620.48  5.66  614.77 
Ceal  0.05  0  0.05 
Cepisa  0.05  0  0.05 
Ceron  5,719.82  -731.57  6,451.36 
Eletroacre  409.95  -276.55  686.50 
Total  19,717.76  -3,908.55  23,719.07 

Source: Own preparation, based on Product 11: Proposal of Modeling of Privatization of each distribution company, Service B.

Considering (i) the amounts in Table 26 above, (ii) that CPPI Resolution 20/2017 defined the adjustment in Ceal amounting to R$ 50,000.00, and (iii) that the provision for any potential loss connected to the Administrative Proceedings with ANEEL, regarding the inspection of the transfer of CCC resources, is premature, and (iv) the CCC debt claims in the balance sheets of the distribution companies are good, the impact on Eletrobras is R$ 11.24 billion

On the other hand, considering the amounts in Table 27 and the definition of CPPI Resolution 20/2017 of adjustment in Ceal amounting to R$ 50,000.00, it may be noted that, if the administrative appeals filed are dismissed and still to be filed before ANEEL, Eletrobras will have to reimburse the CDE CCC Fund at an amount which may amount to R$ 4.0 billion, in addition to having to reverse the CCC's debt claim values included in the distribution companies’ balance sheets, at R$ 8.47 billion, with the impact on Eletrobras in this scenario possibly reaching 23.7 billion.

Notwithstanding the foregoing, as already mentioned in the previous item, it should be noted that CPPI Resolution 20/2017, as amended by CPPI Resolution No. 28/2017, in its Article 3, paragraph 1, provides that, in addition to adjustments of R$ 11.24 billion, "Eletrobras may, at the discretion of the General Meeting, undertake the rights and obligations of the distribution companies, referring to the Fuel Consumption Account - CCC and the Energy Development Account - CDE, recognized in the Financial Statements of the Distribution companies at the base date of the studies considering the adjustments up to

Page 47


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

June 30, 2017, including the rights connected to Item IX of Article 13 of Law 10.438, dated April 26, 2002, with the new wording given by Law 13.299, dated June 21, 2016, and Eletrobras shall, on the other hand, undertake the rights and/or obligations in an equivalent amount.” (emphasis added)

Thus, assuming the prerogative above, of not assuming the CCC debt claims included in the distribution companies’ balance sheets, which amount to R$ 8.47 billion, the impact on Eletrobras will be R$ 11.24 billion, referring to the adjustments to achieve the set price of R$ 50,000.00, in each distribution company.

The table below shows the breakdown of the capitalization by Eletrobras in the distribution companies, based on the source of the appeals, in order to achieve the set price. It should be noted that, in this scenario, preference was given to the conversion of RO’s debts and to the World Bank in capital, although common resources had a higher cost (12.00%), since debts with Petrobras were partially connected to CCC debt claims, which will not be transferred to Eletrobras in this case.

Table 28. Conversion of debts in capital, per distribution company R$ million

Distribution Company Equity Debt Conversion Total
RO + AFAC  BIRD  RGR  CCC  Related Parties  Petrobras 
AmD  1,272.66  138.50  -  -  406.00  7,094.71  8,911.87 
Boa Vista Energia  64.08  15.48  -  -  262.56  -  342.12 
Ceal  0.05  -  -  -  -  -  0.05 
Cepisa  0.05  -  -  -  -  -  0.05 
Ceron  596.99  105.99  -  -  1,169.54  -  1,872.52 
Eletroacre  64.08  49.70  -  -  -  -  113.78 
TOTAL  1,997.91  309.67  0.00  0.00  1,838.10  7,094.71  11,240.38 

 

In the hypothetical event that the General Meeting of Eletrobras decides to assume the debt claims of CCC and CDE recorded in the distribution companies’ balance sheets, the table below establishes the amounts of debt, per source of funds, to be capitalized to reach the set price, per distribution company, as well as the assumption of debts in the same amount of CCC and CDE debt claims. In this context, the debts with Petrobras are

Page 48


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

prioritized, since most of these debts are already backed up by receivables of CCC, substantiated in Contracts of Admission of Debt, wherein Eletrobras stands as a guarantor.

Table 29.  Conversion of debts in capital and Assumption of Debt Claims of CCC 
and CDE, per distribution company
R$ million
Distribution Company  Equity Debt Conversion    Total 
 RO + AFAC   BIRD  RGR  CCC  Related Parties Petrobras  
AmD  -  138.50  -  2,752.20  -  6,021.17  8,911.87 
Boa Vista Energia  -  15.48  -  -  -  326.64  342.12 
Ceal  0.05  -  -  -  -  -  0.05 
Cepisa  0.05  -  -  -  -  -  0.05 
Ceron  -  105.99  -  -  -  1,766.53  1,872.52 
Eletroacre  56.72  49.70  -  -  -  7.36  113.78 
TOTAL  56.82  309.67  -  2,752.20  -  8,121.70  11,240.39 
 
  Assumption of Debt Claims    Total 
 Distribution Company  RO + AFAC  BIRD  RGR  CCC  Related Parties Petrobras  
AmD  -  -  107.50  1,157.89  -  2,790.16  4,055.55 
Boa Vista Energia  -  -  2.62  76.99  190.09  8.66  278.36 
Ceal  -  -  -  -  -  -  - 
Cepisa  -  -  -  -  -  -  - 
Ceron  470.49  -  63.18  164.56  2,369.00  780.07  3,847.29 
Eletroacre  -  -  7.36  81.61  -  207.20  296.17 
TOTAL  470.49  -  180.65  1,481.05  2,559.09  3,786.08  8,477.37 

 

If Eletrobras timely chooses, at Stage 2, to raise its interest up to 30% of the capital of each distribution company, the Holding shall have to perform an additional disbursement of funds, by means of conversion of new debts, as presented in the table below.

Table 30.  Capitalization of Investor and Eletrobras Stage 2 
Distribution company Disburseuent of Investor’s
Appeals (R$) 
Eletrobras’ Disburseuents
in Stage 2
of Interest (%) 
AmD  491,370,787.84  210,587,480.50 
Boa Vista Energia  175,999,185.71  75,428,222.45 
CEAL  545,770,485.33  233,901,636.57 
CEPISA  720,915,595.51  308,963,826.65 
CERON  241,099,855.91  103,328,509.68 
Eletroacre  238,805,729.30  102,345,312.56 
TOTAL  2,413,961,639.60  1,034,554,988.40 

Source: Own preparation, based on Product 11: Proposal of Modeling of Privatization of each distribution company, Service B.

Page 49


 


  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
In addition to this, the modeling assessments performed by Service B analyzed the case of liquidation of the distribution companies. The results show that liquidation, in general, from a financial perspective, and considering that the liquidator is Eletrobras, is more costly than the sale of the controlling interest connected to the grant of a new concession.

The Table below shows the liquidation costs versus the need for contribution of capital by Eletrobras, for the sale of the distribution companies.

Table 31.  Liquidation Cost by Eletrobras x Sale 
Distribution company  Liquidation Cost (BRL)  Capitalization of Eletrobras
Stage 1 (R$) 
Amazonas Energia  12,060,245,607.63  8,911,866,558.94 
Boa Vista Energia  465,696,747.78  342,120,486.20 
Ceal  569,913,103.09  50,000.00 
Cepisa  740,112,396.38  50,000.00 
Ceron  2,398,260,429.38  1,872,522,463.43 
Eletroacre  402,348,970.72  113,779,872.00 
Total  16,636,577,254.98  11,240,389,380.57 
Source: Product 11: Proposal of Modeling of Privatization of each distribution company, Service B. 
 
     

10.1 RELEVANT POINTS 

10.1.1UNBUNDLING OF AMAZONAS ENERGIA

With the full interconnection of the Manaus system to the National Interconnected System SIN on May 5, 2015, according to ANEEL Order 1365/2015, and aiming at the fulfillment of the provisions of Paragraph 5, Article 4 of Law 9,074/1995, the unbundling process of AmE D was started.

On July 1, 2015, Amazonas GT - AmGT - was created, by means of the transfer of the assets of Generation and Transmission that belonged to AmE D to the new company incorporated, which remained as a wholly-owned subsidiary of AmE D.

In order to complete the unbundling process, AmE D will transfer the shares of AmGT to Eletrobras, in exchange for which Eletrobras will assume/capitalize AmE D debts in the amount of AmGT value, which will be ruled by independent consulting firm.

Page 50


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

In October 2016, ANEEL issued Official Letter 574/2016-SFF, in which it determined that the partial assignment of gas contract number OC-1902 of Amazonas Energia to Amazonas-GT should be made in two steps: (i) immediately, about 60% of the contract for the consumption of Aparecida and Mauá 3 is transferred to GT, and (ii) in 2025, on the occasion of the expiration of the contracts of the 5 Independent Power Producers (PIE) operating in Manaus, the remaining 40% of the contract will be transferred. It should be noted that AmE D would continue to own 200,000 m3/s connected to the consumption of its PIE’s uptown the state.

Furthermore, as of December 2015, the Generation Regulation Department (SRG) of ANEEL started using a new interpretation, defining that ship-or-pay and take-or-pay would be limited to the installed capacity of thermal power plants, likely to consume the engaged daily amount of natural gas, which is smaller than that established under Contract OC-1902/2006. Such fact is dismissed by Eletrobras and is being disputed by an administrative proceeding.

It should be noted that Petrobras and Cigás, as the other parties to the Eletrobras companies in the commercial relationship concerning gas consumption in Amazonas, should also agree to the assignment of the gas contract, and Petrobras has so far shown little inclination to do so, while the debt settlement between the distribution companies of Eletrobras with it does not occur.

CPPI Resolution No. 20/2017 establishes in article 3, paragraph 4, that the unbundling of AmE D should be completed by March 2, 2018. The modeling proposed by BNDES and approved by CPPI already considers the AmE D unbundled in its studies.

Therefore, in order to complete the unbundling until the date set in the CPPI’s resolution, a solution should be reached regarding the current understanding of ANEEL with respect to the gas contract, and a consent for assigning the contract shall be obtained with Petrobras and Cigás.

In view of the foregoing, it should be noted that the compulsory unbundling for the AmE D auction to take place is a risk posed to the bidding procedure. In the event that it does not occur, the Granting Power will hold a bidding process relating to the electric energy distribution concession, without transfer of the corporate control of the legal entity responsible for the provision of the electric energy distribution service, according to Article 5 of Decree 9,192/2017, thus taking into account the liquidation of AmE D.

Page 51


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

10.1.2EMPLOYMENT LITIGATION BRESSER PLAN - CEAL

The Federal General Attorney (AGU), by means of Order W/N, dated September 1, 2017, approved the feasibility of executing an agreement between Eletrobras, Ceal and the Union of Workers of the Urban Industries in the State of Alagoas - STIUEA, amounting to R$ 129.7 million.

In turn, the Minister of State for Mines and Energy, through Order w/n, dated September 8, 2017, authorized the execution of the agreement in accordance with the Order of the Federal General Attorney above.

Through CTA-PR-054/2017 Letter, Ceal presented to STIUEA a proposal of agreement amounting to R$ 178 million, with the difference being nearly R$ 49 million owing to the update between the base date of the value calculated by AGU and the base date of the proposal.

According to mail STIUEA OF. No. 412/2017, dated November 6, 2017, the Union considered the value proposed by Ceal as insufficient and refused the agreement. In the same mail, the Union presented a counteroffer with values much higher than the one authorized by AGU, requesting the answer of Ceal until November 14, 2017.

The CPPI Resolution 20/2017 defined that the privatization of Ceal is conditioned to the execution and judicial homologation of an agreement regarding the payment of salary differences arising from the Bresser Plan. And, in turn, Decree 9192/2017, in its Article 5, establishes that a bidding process for grant of distribution shall happen, without the transfer of controlling interest, if the bidding process fails to achieve its original purpose.

In view of the foregoing, owing to the failure of the agreement with the Union of Ceal’s workers, the company is subject to liquidation.

Page 52


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

10.1.3BOA VISTA MODELING CERR INDEMNITY

It should also be noted that the economic and financial assessment studies of Services A and B considered, in the Regulatory Remuneration Basis of Boa Vista Energia, the assets of CERR, as well as the indemnification for the assets not fully depreciated and amortized of CERR, amounting to R$ 73 million, based on ANEEL Technical Note no. 293, dated September 20, 2017 (this being not a decision of ANEEL), being paid by Eletrobras.

Note that Law 12783/2013, in Paragraph 2 of its Article 8 establishes that the "calculation of the value of the indemnity corresponding to the installments of investments connected to reversible assets, not yet amortized or not depreciated, will be founded on the new replacement value methodology, according to criteria established in a granting authority regulation."

Decree No. 8,461/2015, which regulated the extension of electric energy distribution concessions, defined, in its Article 5, that the unpaid concessions will be subject to a bidding process and the "indemnity to be paid to the former utility service company, based on the value of the investments of the reversible assets not yet depreciated, shall be calculated by ANEEL based on the New Replacement Value - VNR and will consider the depreciation cumulative value from the date of commissioning of the facility, in accordance with the criteria of the Manual of Accounting of the Electric Industry - MCSE.”

Decree 9192/2017, in paragraph 15 if its Article 4, establishes that “the new utility service company shall, according to rules and deadlines to be defined in a public notice, purchases the assets and the reversible facilities connected to the provision of the service, which are separately owned by the company that will have its corporate control transferred through the bidding process provided for under the head provision, corresponding to the part of the investments not amortized and/or not depreciated in connection with them, valued based on the methodology of the New Replacement Value - VNR.”

Therefore, it should be noted that the particulars of the BNDES modeling, establishing the provision for of the liquidation of CERR's assets by Eletrobras in the manner calculated in the studies, is in conflict with the terms of Article 5 of Decree 8461/2015 and with the terms of Paragraph 15, Article 4 of Decree 9192/2017, are referred to above. Thus, it can be observed that the sales modeling proposed for Boa Vista Energia is in conflict with the provisions of Decrees 8461/2015 and 9192/2017.

Page 53


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

10.1.4FIXED ASSETS UNDERWAY

The studies performed by Services A and B did not incorporate the positions of Fixed Assets in Use (AIC) into the Regulatory Remuneration Base - BBR since, according to the Services, the assessors' report about the Remuneration Base does not take into account the specific assessment of the condition and characteristic of these assets concerning: (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the reassessment process. However, in the 2019 tariff review, at least part of these investments will be incorporated into BRR, increasing the value of the distribution companies.

The table below shows the Fixed Assets in progress, as stated in the assessment for Service B, for each distribution company. As can be seen, the amounts disregarded are significant, and their consideration would substantially improve the results of the economic and financial assessments performed by the Services, and consequently reduce the amount to be contributed by Eletrobras to reach the set value of R$ 50,000.00, as defined in CPPI Resolution no. 20/2017.

Table 32.  Fixed Assets in use not considered 
Distribution company  AIC (R$ million) 
AmE D  1,409.91 
Boa Vista Energia  57.37 
CEAL  177.42 
CEPISA  168.13 
CERON  501.74 
Eletroacre  139.29 
TOTAL  2,453.87 

 

Lastly, it’s worth mentioningthat the detailed analysis by distribution company may be found in the appendices to this IT.

Page 54


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

11.      IMPACTS ON ELETROBRAS 
 
11.1      FINANCIAL IMPACT CASH FLOW

Considering the model approved by CPPI, based on studies conducted by BNDES, Eletrobras' cash flow was projected in order to demonstrate the impact of the sale of the control of the distribution companies, in view of the obligations of debt assumption and capitalization, as detailed in Tables 28 and 29.

It should be noted that a simulation of this nature contains estimates or projections of future events and trends which are subject to several natural risks and uncertainties of this type of exercise, increased as being long-term estimates made considering the best information available and some assumptions that depend directly on third party’s negotiations and acceptance.

In this sense, and given the uncertainties regarding the realization of the debt claims with the CCC CDE Fund, dealt with in Item 5.1 of this IT, and the privileges provided for in Article 3, paragraph 1, of CCPI Resolution 20/2017, amended by CPPI Resolution 28/2017, two cash flow scenarios of Eletrobras were projected, namely: (i) without the assumption of CCC and CDE debt claims; and (ii) with the assumption of debt claims and debts in a corresponding amount.

The main assumptions used to forecast the inflows and outflows of cash flow are as follows:

a)      Premises common to both scenarios 
 
·      The receivables connected to the principal and charges of the financing agreements granted were projected in accordance with the maturity curve of the contracts, and considering the completion, until 01/31/2018, of the transfer of 100% of the assets received as payment in kind for amortization of the debts of the subsidiaries Furnas, Eletrosul, Eletronorte and Chesf, under the terms of CAE Deliberation 170/2017, dated 06/30/2017; 
 
·      Considering the sale until 12/31/2018 of the assets of the subsidiaries object of operations of payment in kind. The expected cash inflow was projected at R$ 3.4 billion, which corresponds to the book value of the assets ascertained on

Page 55


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

06/30/2017. This assumption considers obtaining the necessary approvals for the transfer of 100% of the assets involved in operations of payment in kind considered in CAE Deliberation 170/2017, excluding SPE BMTE and IE Madeira, which had transfers not approved by BNDES;
· Projected the receipt of dividends from affiliates and subsidiaries, with the dividends of affiliated companies being estimated for the year 2018 based on the average of the amounts received in the last 3 years and, with respect to the subsidiaries’ dividends, the premise was a growth of 50% compared to 2017. As of

2019, the projection considered the value projected for 2018 with annual adjustment by IPCA;
· The proceeds from capital income and the difference in US inflation in the Itaipu contracts were projected based on the amounts received in 2017; in 2018, in turn, the amounts approved by the MME (Ordinance no. 3/2017), and for subsequent years, applied to the receipt curve of the financing contracts to Itaipu;
· The projection considers, in the year 2018, the rollover of the bonus with maturity in 2019, amounting to US$ 1.0 billion and, in year 2020, the rollover of the bonus maturing in 2021, amounting to US$ 1.7 billion. The premise is the term of payment of 10 years (bullet) and coupon of 6.75% p.a.;
· The MSO expenses (materials, outsourced services and other expenses) were projected in 2018 based on the amount predicted in the budget proposal. From 2019 onwards, the annual updating by IPCA was performed;
· The expenses with the Personnel sheet and the cost of the PAE (Extraordinary Retirement Program) were reported by DJP;
· The expenses with payment of shares of the compulsory loan were projected based on the amount predicted in the budget proposal of 2018 and, from 2019 onwards, annually adjusted based on the IPCA;
· The debt service of loans obtained by Eletrobras was projected based on the curve of the current contracts and the new debt due to the bonus rollover premise of 2019 and 2021;
· The dividends payable by Eletrobras consider the payment of a minimum value required to the preferred shareholders;
· No provision was made for the granting of funds by Eletrobras to its subsidiaries, as well as the execution of restated contracts with the subsidiaries from 2018 onwards;

Page 56


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

b) Premise of scenario (ii)

c) Premise of scenario (ii)

Thus, based on the assumptions adopted above, the following results are observed concerning the final cash availability of Eletrobras at the end of each year, for the period from 2018 to 2022.

Table 33.  Cash Flow of Eletrobras for Scenario (i) - 2018 2022, in R$ million 
 
2018 2019 2020 2021 2022 Total
2018 a
2022 
Previous
balance 
3,457  4,373  5,006  5,033  4,920  3,457 
Cash inflows  12,121  6,978  12,894  6,346  4,330  42,668 
Cash outflows  (11,205)  (6,345)  (12,867)  (6,458)  (6,201)  (43,076) 
Ending Balance  4,373  5,006  5,033  4,920  3,049  3,049 

 

Page 57


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Table 34. Cash Flow of Eletrobras for Scenario (ii) - 2018 2022, in R$ million

2018 2019 2020 2021 2022 Total 2018
a 2022 
Previous             
balance  3,457  3,197  2,803  1,474  577  3,457 
Cash inflows  12,237  8,275  14,035  7,419  5,430  47,396 
Cash outflows  (12,497)  (8,669)  (15,364)  (8,316)  (7,322)  (52,169) 
Ending Balance  3,197  2,803  1,474  577  (1,315)  (1,315) 

 

Therefore, based on the analyses above and considering the adjustments of distribution companies’ capital conversion and/or debt assumption before Eletrobras and/or third party, provided for in BNDES modeling and studies and highlighted by CPPI Resolution 20/2017, it should be noted that, in scenario (ii) - with the assumption of CCC and CDE debt claims - the cash flow was not sustainable over the period, accumulating a deficit of about R$ 1.3 billion. It should be noted that the cash projected in scenario (ii) does not present a negative value as early as 2018, owing to the adoption of a premise of success in the sale of 100% of the SPE assets received from the subsidiaries as payment in kind, in accordance with CAE Resolution 170/2017, excluding SPE BMTE and IE Madeira, whose transfers were not approved by BNDES, which represents an expected cash inflow amounting to R$ 3.4 billion. In the case this operation shows no success, the cash may be negative already in 2018, in scenario (ii), at R$ 0.2 billion.

Page 58


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

11.2 ECONOMIC IMPACT ACCOUNTING EFFECT

CPPI Resolution 20/2017, amended by Resolution 28/2017, entitled the Eletrobras’ General Meeting the right to assume de debt claims of CCC and CDE recorded in the balance sheets of the distribution companies on June 30, 2016. Thus, as already mentioned, two scenarios for the accounting representation were devised: (i) without the assumption of these debt claims; and (ii) with the assumption of debt claims, according to the decision of EGM.

The accounting booking of the assumption of assets, liabilities and capital subscription in the distribution companies foreseen by the CPPI modeling does not have an effect on the result, given that the amounts are based on the equity amounts recorded herein, however, as foreseen for the CPPI modeling, both in the scenario without assumption of CCC and CDE debt claims as in the scenario with assumption of CCC and CDE debt claims, the sale of these post-modeling companies will result in a gain on the sale of approximately R$ 3.53 billion considering the amount of R$ 50 thousand Brazilian Reais for each distribution company.

The balance sheet, with the base date of June 2017, is presented below, simulating the scenario without assumption of CCC and CDE debt claims, according to the modeling approved by CPPI, pre-sale and post-sale and ascertainment of gain on sale.

Page 59


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
Table 35.  Change to the Assets without the assumption of CCC and CDE debt claims
CENTRAIS ELÉTRICAS BRASILEIRAS S.A - ELETROBRAS
SIMULATED BALANCE SHEET AS OF JULY 30, 2017
( in thousands of reais )
ASSETS 06/30/2017 CPPI adjustments Simulation
after
adjustments 
Sales
adjustments
Simulation
after sale
 
CURRENT           
Cash and cash equivalent  531,487  -  531,487  300  531,787 
Restricted cash  1,311,129  -  1,311,129  -  1,311,129 
Marketable Securities  4,289,359  -  4,289,359  -  4,289,359 
Accounts Receivable, net  364,784  -  364,784  -  364,784 
Financing and loans  7,787,594  -  7,787,594  -  7,787,594 
Dividends receivables  637,470  -  637,470  -  637,470 
Recoverable taxes  308,426  -  308,426  -  308,426 
Income tax and social contribution  785,634  -  785,634  -  785,634 
Reimbursement rights  736,697  -  736,697  -  736,697 
Inventories  162  -  162  -  162 
Others  819,824  -  819,824    -  819,824 
TOTAL CURRENT ASSETS  17,572,566  -  17,572,566  300  17,572,866 
 
NON-CURRENT           
LONG-TERM ASSETS           
Financing and loans  27,644,084  -1,997,909  25,646,175  -  25,646,175 
Accounts Receivable, net  53,508  -  53,508  -  53,508 
Marketable Securities  253,215  -  253,215  -  253,215 
Income tax and social contribution  1,488,158  -  1,488,158  -  1,488,158 
Guarantees and restricted deposits  3,159,025  -  3,159,025  -  3,159,025 
Financial assets - Concessions and Itai  2,527,517  -  2,527,517  -  2,527,517 
Advances for future capital increase  1,450,914  -  1,450,914  -  1,450,914 
Others  2,308,765  -  2,308,765  -  2,308,765 
  38,885,186  -1,997,909  36,887,277  -  36,887,277 
 
 
INVESTMENTS           
Accounted for-by the equity method  62,338,476  -  62,338,476  -  62,338,476 
Maintained at fair value  1,172,739  -  1,172,739  -  1,172,739 
  63,511,215  -  63,511,215  -  63,511,215 
 
FIXED ASSETS, NET  192,995    192,995    192,995 
 
 
TOTAL NON-CURRENT ASSETS  102,589,396    -1,997,909  100,591,487  -  100,591,487 
 
TOTAL ASSETS  120,161,962  -1,997,909    118,164,053  300    118,164,353 

 

Page 60


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Table 36. Change to the Liabilities without the assumption of CCC and CDE debt claims

 

CENTRAIS ELÉTRICAS BRASILEIRAS S.A - ELETROBRAS
SIMULATED BALANCE SHEET AS OF JULY 30, 2017
( in thousands of reais )
 
  06/30/2017 CPPI
adjustments 
Simulation after
adjustments 
Sales
adjustments 
Simulation
after sale 
LIABILITIES ANS EQUITY 
   
 
CURRENT LIABILITIES           
Financing and loans  3,145,174  -  3,145,174  -  3,145,174 
Debentures  -  -  -  -  - 
Financial Liabilities  -  -  -  -  - 
Compulsory Loan  46,001  -  46,001  -  46,001 
Suppliers  528,922  -  528,922  -  528,922 
Advances from clients  553,688  -  553,688  -  553,688 
Taxes payable  101,344  -  101,344  -  101,344 
Income tax and social contribution  742,747  -  742,747  -  742,747 
Shareholders' compensation  458,706  -  458,706  -  458,706 
Financial Liabilities - Concessions and Itaipu  1,231,603  -  1,231,603  -  1,231,603 
Estimated Obligations  141,588  -  141,588  -  141,588 
Reimbursement Obligations  2,062,165  -  2,062,165  -  2,062,165 
Post-employment Benefits  15,972  -  15,972  -  15,972 
Provisions for Litigation  700,339  -  700,339  -  700,339 
Liabilities associated to assets held for sale  5,633  -  5,633  -  5,633 
Others  85,535  -  85,535  -  85,535 
TOTAL CURRENT LIABILITIES  9,819,417  -  9,819,417  -  9,819,417 
 
NON-CURRENT LIABILITIES           
Financing and Loan  22,115,700  9,242,480  31,358,180  -  31,358,180 
Compulsoty Loan  465,509  -  465,509  -  465,509 
Provision for Litigation  14,581,432  -  14,581,432  -  14,581,432 
Post-employment benefits  408,454.00  -  408,454  -  408,454 
Excess of liabilities over assets (provision)  21,945,542  -11,240,389  10,705,153  -3,532,398  7,172,755 
Reimbursement Obligations  -  -  -  -  - 
Advance for Future Capital Increase  3,497,587  -  3,497,587  -  3,497,587 
Taxes payable  2,222  -  2,222  -  2,222 
Income tax and social contribution  321,853  -  321,853  -  321,853 
Others  1,040,324  -  1,040,324  -  1,040,324 
TOTAL NON-CURRENT LIABILITIES  64,378,623  -1,997,909  62,380,714  -3,532,398  58,848,316 
 
EQUITY           
Capital stock  31,305,331  -  31,305,331  -  31,305,331 
Capital reserves  13,867,170  -  13,867,170  -  13,867,170 
Profit reserves  3,018,680  -  3,018,680  -  3,018,680 
Adjustments to equity  24,367  -  24,367  -  24,367 
Accumulated losses  1,733,711  -  1,733,711  3,532,698  5,266,409 
Accumulated other comprehensive loss  -3,985,337  -  -3,985,337  -    -3,985,337 
TOTAL EQUITY  45,963,922  -  45,963,922  3,532,698  49,496,620 
 
TOTAL LIABILITIES + EQUITY  120,161,962    -1,997,909    118,164,053  300  118,164,353 

 

Page 61


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Table 37. Accounting profit or loss per company without the assumption of CCC and CDE debt claims

 

Sale's ascertainment  EDALAGOAS  EDRONDONIA  EDPIAU  EDACRE  EDRORAIMA  AMAZONAS D 
 
Excess of liabilities over assets/ Equity in the Parent Company  -695,835  -1,599,224  -1,270,321  -393,969  -710,095  -10,103,343 
Capitalization Adjustments  50  1,872,522  50  113,780  342,120  8,911,867 
Excess of liabilities over assets (ajusted after capitalization)  -695,785  273,298  -1,270,271  -280,189  -367,975  -1,191,476 
Sale Value  50  50  50  50  50  50 
Possible profit or loss from sale/ revenue (expense)  695,835  -273,248  1,270,321  280,239  368,025  1,191,526 
Sale Consolidated Profit or Loss             
  3,532,698           

 

The following tables present Eletrobras’ balance sheet, dated June 2017, simulating the scenario with the assumption of CCC and CDE debt claims, pre-sale and after-sales and ascertainment of the gain on sale.

Page 62


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Table 38. Change to the Assets with the assumption of CCC and CDE debt claims

 

CENTRAIS ELÉTRICAS BRASILEIRAS S.A - ELETROBRAS
SIMULATED BALANCE SHEET AS OF JULY 30, 2017
( in thousands of reais )
 
  06/30/2017 CPPI
adjustments 
Simulation after
adjustments 
Sales
adjustments 
Simulation
after sale 
ASSETS 
 
 
CURRENT           
Cash and cash equivalent  531,487  -  531,487  300  531,787 
Restricted cash  1,311,129  -  1,311,129  -  1,311,129 
Marketable Securities  4,289,359  -  4,289,359  -  4,289,359 
Accounts Receivable, net  364,784  -  364,784  -  364,784 
Financing and loans  7,787,594  -  7,787,594  -  7,787,594 
Dividends receivables  637,470  -  637,470  -  637,470 
Recoverable taxes  308,426  -  308,426  -  308,426 
Income tax and social contribution  785,634  -  785,634  -  785,634 
Reimbursement rights  736,697  -  736,697  -  736,697 
Inventories  162  -  162  -  162 
Others  819,824  -  819,824  -  819,824 
TOTAL CURRENT ASSETS  17,572,566  -  17,572,566  300  17,572,866 
 
NON-CURRENT           
LONG-TERM ASSETS           
Reimbursement rights  -  8,477,369  8,477,369  -  8,477,369 
Financing and loans  27,644,084  -707,964  26,936,120  -  26,936,120 
Accounts Receivable, net  53,508  -  53,508  -  53,508 
Marketable Securities  253,215  -  253,215  -  253,215 
Income tax and social contribution  1,488,158  -  1,488,158  -  1,488,158 
Guarantees and restricted deposits  3,159,025  -  3,159,025  -  3,159,025 
Financial assets - Concessions and Itaip  2,527,517  -  2,527,517  -  2,527,517 
Advances for future capital increase  1,450,914  -  1,450,914  -  1,450,914 
Others  2,308,765  -  2,308,765  -  2,308,765 
  38,885,186  7,769,405  46,654,591  -  46,654,591 
 
 
INVESTMENTS           
Accounted for-by the equity method  62,338,476  -  62,338,476  -  62,338,476 
Maintained at fair value  1,172,739  -  1,172,739  -  1,172,739 
  63,511,215  -  63,511,215  -  63,511,215 
 
FIXED ASSETS, NET  192,995    192,995    192,995 
 
 
TOTAL NON-CURRENT ASSETS  102,589,396  7,769,405  110,358,801  -  110,358,801 
 
TOTAL ASSETS  120,161,962  7,769,405  127,931,367  300  127,931,667 

 

Page 63


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Table 39. Change to the Liabilities with the assumption of CCC and CDE debt
claims

 
CENTRAIS ELÉTRICAS BRASILEIRAS S.A - ELETROBRAS
SIMULATED BALANCE SHEET AS OF JULY 30, 2017
( in thousands of reais )

 
      CPPI  Simulation after  Sales  Simulation 
LIABILITIES ANS EQUITY  06/30/2017         
      adjustments  adjustments  adjustments  after sale 
 
CURRENT LIABILITIES           
Financing and loans  3,145,174  -  3,145,174  -  3,145,174 
Debentures    -  -  -  -  - 
Financial Liabilities    -    -   
Compulsory Loan    46,001  -  46,001  -  46,001 
Suppliers    528,922  -  528,922  -  528,922 
Advances from clients  553,688  -  553,688  -  553,688 
Taxes payable    101,344  -  101,344  -  101,344 
Income tax and social contribution  742,747  -  742,747  -  742,747 
Shareholders' compensation  458,706  -  458,706  -  458,706 
Financial Liabilities - Concessions and Itaipu  1,231,603  -  1,231,603  -  1,231,603 
Estimated Obligations  141,588  -  141,588  -  141,588 
Reimbursement Obligations  2,062,165  -  2,062,165  -  2,062,165 
Post-employment Benefits  15,972  -  15,972  -  15,972 
Provisions for Litigation  700,339  -  700,339  -  700,339 
Liabilities associated to assets held for sale  5,633  -  5,633  -  5,633 
Others    85,535  -  85,535  -  85,535 
TOTAL CURRENT LIABILITIES  9,819,417  -  9,819,417  -  9,819,417 
 
NON-CURRENT LIABILITIES           
Financing and Loan  22,115,700  14,776,542  36,892,242  -  36,892,242 
Compulsoty Loan    465,509  -  465,509  -  465,509 
Provision for Litigation  14,581,432  -  14,581,432  -  14,581,432 
Post-employment benefits  408,454.00  -  408,454  -  408,454 
Excess of liabilities over assets (provision)  21,945,542  -11,240,389  10,705,153  -3,532,398  7,172,755 
Reimbursement Obligations  -  4,233,252  4,233,252  -  4,233,252 
Advance for Future Capital Increase  3,497,587  -  3,497,587  -  3,497,587 
Taxes payable    2,222  -  2,222  -  2,222 
Income tax and social contribution  321,853  -  321,853  -  321,853 
Others    1,040,324  -  1,040,324  -  1,040,324 
TOTAL NON-CURRENT LIABILITIES  64,378,623  7,769,405  72,148,028  -3,532,398  68,615,630 
 
EQUITY             
Capital stock    31,305,331  -  31,305,331  -  31,305,331 
Capital reserves    13,867,170  -  13,867,170  -  13,867,170 
Profit reserves    3,018,680  -  3,018,680  -  3,018,680 
Adjustments to equity  24,367  -  24,367  -  24,367 
Accumulated losses  1,733,711  -  1,733,711  3,532,698  5,266,409 
Accumulated other comprehensive loss  -3,985,337  -  -3,985,337  -  -3,985,337 
TOTAL EQUITY    45,963,922  -  45,963,922  3,532,698  49,496,620 
 
TOTAL LIABILITIES + EQUITY  120,161,962  7,769,405  127,931,367  300  127,931,667 

 

Page 64


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Table 40. Accounting profit or loss per company with the assumption of CCC and CDE debt claims

 

Sale's ascertainment  EDALAGOAS  EDRONDONIA  EDPIAU  EDACRE  EDRORAIMA  AMAZONAS D 
 
Excess of liabilities over assets/ Equity in the Parent Company  -695,835  -1,599,224  -1,270,321  -393,969  -710,095  -10,103,343 
Capitalization Adjustments  50  1,872,522  50  113,780  342,120  8,911,867 
Excess of liabilities over assets (ajusted after capitalization)  -695,785  273,298  -1,270,271  -280,189  -367,975  -1,191,476 
Sale Value  50  50  50  50  50  50 
Possible profit or loss from sale/ revenue (expense)  695,835  -273,248  1,270,321  280,239  368,025  1,191,526 
Sale Consolidated Profit or Loss             
3,532,698

 

12. CONCLUSION

Initially, the conclusion on the sale of the distribution companies’ controlling interest is based on the decision of the shareholders of Eletrobras at the 165t EGM held on July 22, 2016, namely:

(i)      decision, by majority vote, for the non-extension of the concessions of Eletrobras distribution companies;
(ii)      decision to transfer the controlling interest until December 31, 2017, of the distribution companies Cepisa, Ceal, Eletroacre, Ceron, Boa Vista Energia and AmE
  D, under the terms of Paragraph 1-A of Article 8, of Law 12783/2013; and

(iii)decision, by majority vote, for the adoption of measures to liquidate the distribution companies, in the cases of transfer of controlling interest not being performed until December 31, 2017 or the respective distribution companies fail to receive directly from the Federal Government or by a tariff, until its transfer to the new controlling shareholder, all the resources and remuneration required to operate, maintain and make investments that are connected to the public utilities of the respective distribution company, maintaining the economic and financial balance of the Distribution company, without any contribution by Eletrobras.

In this sense, the Granting Power, by means of the Ministry of Mines and Energy, appointed the distribution companies of Eletrobras to temporarily act as electric energy public utility service companies until the new utility service company assumes this activity.

Page 65


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

The Federal Government, by means of CPPI, approved the modeling of transfer of controlling stocks of the distribution companies associated to the grant of a new concession.

Therefore, it may be concluded that, with respect to the modeling of disposal approved by CPPI, the following points should be observed by the senior management of Eletrobras in their decision-making:

12.1 REASON FOR THE NON-CONTINUANCE OF ELETROBRAS IN THE

CONTROL OF THE DISTRIBUTION COMPANIES

According to PDNG 2017-2021, the non-continuance of Eletrobras in the distribution sector implies the sale of the distribution companies AmE D, Ceal, CEPISA, Ceron, Boa Vista Energia and Eletroacre. This is an important measure for remedying the financial situation of Eletrobras, since the distribution companies have been a source of destruction of value for the Company’s shareholders, as mentioned in Item 4.2 of this document, which establishes that the Holding Company has already allocated resources of R$ 30.1 billion (including Ordinary Resources currently owed by distribution companies to Eletrobras), at June 2017 prices, to the six distribution companies, with no return up to date.

The current situation of the distribution companies is critical, wholly presenting recurring losses along the last five years. In the last fiscal year, in 2016, the worst historical year occurred for the industry, with a total loss of R$ 7.0 billion. In 2017, until the first half of the year, the result was negative at R$ 934 million.

Therefore, it can be noted that the operation to divest the controlling interest of these companies controlled by Eletrobras is necessary and will significantly contribute to Eletrobras’ financial balance, since the current scenario is unsustainable and, with this operation, shareholders shall stop loss.

 

12.2 CONDITIONS PROPOSED FOR THE DISPOSAL OPERATION OF THE

DISTRIBUTION COMPANIES

The proposal is that the privatization of distribution companies controlled by Eletrobras happens through the sale of controlling stake. This modality was a choice of CPPI among the various modalities of privatization provided for in Law 9,491/1997. Likewise, the

Page 66


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

conditions of sale and the sale price of the control were defined by CPPI Resolution 20/2017, amended by CPPI Resolution 28/2017. Therefore, it is necessary to evaluate the conditions approved by CPPI based on the studies carried out by BNDES, given that, by deliberation of its shareholders mad at the 165th Extraordinary General Meeting, Eletrobras is no longer able to maintain the controlling interest of these companies.

With respect to the expected assumption of debt claims and debts connected to the CCC/CDE funds, this document, in its Item 5.1, already indicates that the value at risk for Eletrobras is up to R$ 12.4 billion, with impact on the income and equity of the Company, when and in case the administrative and/or judicial discussion has an unfavorable outcome for Eletrobras.

However, it should be noted that, in its Resolution No. 28/2017, CPPI warranted to Eletrobras’ General Meeting the decision on whether to assume or not the rights and obligations connected to CCC/CDE recognized in the Financial Statements of the distribution companies, dated June 30, 2017, amounting to R$ 8.47 billion, detailed in tables 17 and 18. And, therefore, the debts that may be assumed are limited to such value.

12.3 FINANCIAL IMPACT ON ELETROBRAS TO EXECUTE THE PROPOSAL OF CPPI

 

Based on the analysis shown in Item 10.1, it may be concluded that, considering the adjustments provided for by CPPI Resolution 20/2017, as amended by CPPI Resolution No. 28/2017, the Eletrobras' cash flow projected in scenario (ii) - with assumption of CCC and CDE debt claims - is not sustainable in the prospective period, between 2018 and 2022, accumulating a relevant deficit of R$ 1.3 billion.

It should also be noted that the cash projected in scenario (ii) does not present a negative value as early as 2018, owing to the adoption of a premise of success in the sale of 100% of the SPE assets received from the subsidiaries, in payment in kind, pursuant to CAE Resolution 170/2017, excluding the SPE BMTE and IE Madeira, whose transfers were not approved by BNDES, which accounts for an expected cash inflow of R$ 3.4 billion. In the case this operation shows no success, the cash may be negative already in 2018, at R$ 0.2 billion.

Page 67


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Therefore, from a financial perspective, Eletrobras does not have the capacity to assume the rights and obligations of the distribution companies, connected to CCC and CDE, recognized in the Financial Statements of the distribution companies on June 30, 2017.

For scenario

(i) without the assumption of credits, therefore, with the capitalization/assumption of R$ 11.2 billion necessary to bring the value of the distribution companies to R$ 50 thousand, the projected cash flow is positive between 2018 and 2022, considering all the assumptions made in Item 10.1.

12.4 CEAL’S SPECIFIC CONCERNINGS

With respect to the sale of Ceal, as described in Item 9.1.2, it should be noted that the non-remediation of the grievance, connected the Bresser Plan, may lead to the distribution company’s liquidation.

12.5 AMAZONAS DISTRIBUIDORA’S SPECIFICCONCERNINGS

As already mentioned in Item 9.1.1, the unbundling of Amazonas Energia is an impediment to its sale in the approved CPPI modeling, leading, therefore, to liquidation.

12.6 SCENARIO OF LIQUIDATION OF THE DISTRIBUTION COMPANIES

The studies of BNDES show the distribution companies’ liquidation scenario, considering Eletrobras as liquidator. The table below illustrates the results obtained:

 Table 41.  Liquidation x Sale   
Distribution company   Liquidation Cost (BRL) Sale Cost (BRL) 
Amazonas Energia  12,060,245,607.63  8,911,866,558.94 
Boa Vista Energia  465,696,747.78  342,120,486.20 
Ceal  569,913,103.09  50,000.00 
Cepisa  740,112,396.38  50,000.00 
Ceron  2,398,260,429.38  1,872,522,463.43 
Eletroacre  402,348,970.72  113,779,872.00 
Total  16,636,577,254.98  11,240,389,380.57 

 

Page 68


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Based on the table above, which briefly represents the BNDES' modeling with respect to the cost of the distribution companies’ liquidations, it is possible to affirm that the least expensive option for Eletrobras is to sell the companies.

It is important to note, however, that the scenario considered for the assessment of the liquidation above assumes the principle that Eletrobras will be the liquidator, and, therefore, is responsible for all rights and obligations previously belonged to the distribution companies.

However, such positioning requires a deep analysis of the limits of the responsibilities of Eletrobras, the Distribution companies and the Federal Government.

 

 

Felipe Baptista da Silva DFP Pedro Paulo da Cunha - DFF
 

 

Marcos B. F. Pinho DFPI Frederico Pinto Eccard - DDEF

 

 

 
Flávia Xavier Cirilo de Sá DFPI

Luciana Pereira de Souza - DDEF

 

Agreed:

Armando Casado de Araujo Chief Financial Officer and Investor Relations Officer

 

Luiz Henrique Hamann Distribution Officer

 

Page 69


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Appendix to the DFP/DFF/DDE Joint Technical

Information

Rate Level

Page 1


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
Contents   
1.  INTRODUCTION  3 
2.  PERIODIC RATE REVISION  4 
3.  SPECIAL RATE REVISION  5 

 

Page 2


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

1. INTRODUCTION

The updating of the rate of a Distributor usually comprises the following mechanisms: the annual rate adjustment (IRT) and the periodical rate revision (RTP), which usually happens every 4 or 5 years. There is also the special rate revision (RTE), which may occur at any time due to an economic-financial imbalance in the concession agreement.

The annual readjustment consists of a monetary updating of the rate. Within this mechanism, the so-called manageable (Portion B) and the non-manageable costs (Portion A) are given different treatments. Manageable costs are those that depend on the actions and on the control of companies, whereas non-manageable costs are unrelated to the management of the agent. Within the logic of incentive regulation, changes in non-manageable costs are usually passed on to consumers, guaranteeing that the expenses are neutral with the result of the regulated agent. In turn, manageable costs are updated based on the inflation deducted from the X Factor, passing on part of the productivity gain to the consumer.

The equation below exemplifies a calculation of the annual rate adjustment index (IRT), where Portion A values are recalculated and Portion B is updated based on inflation, deducted from the X Factor.

Where:

VPA1: Value of Portion A at the date of the adjustment in processing (DRP);

VPB0: Value of Portion B on the previous reference date (DRA), defined as RA0 VPA0; RA0: Annual Revenue with previous rates applied to the market and checked in a 12-month period; VPA0: Value of Portion A on the DRA;

IVI:      Inflation Variation Index; and
X:      Regulatory productivity.

Page 3


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Where:

VPA: Value of Portion A;

CE: Cost of purchase of electric power and own generation; CT: Cost with connection and use of transmission systems; ES: Sector Charges.

2. PERIODIC RATE REVISION

A new balance rate is established in the periodic rate revision, ensuring an appropriate compensation for prudent investments and the passing on of efficient operating costs. This rate, calculated based on the required revenue (RR), is composed of Portion A (non-manageable costs) and Portion B (manageable costs). Since the primary goal of the RTP is to analyze the economic-financial balance of the concession during the whole rate cycle, this process involves, in addition to the calculation of the rate repositioning, the establishment of the X Factor, a mechanism responsible for passing on efficiency gains to consumers, detected among the RTPs. The Value of Portion B (VPB) is calculated in the rate review process, according to the equation below:

Where:

CAOM: Cost of Administration, Operation and Maintenance;
CAA: Annual Cost of Assets;
Pm:Market Adjustment Factor;
M1Q: Incentive Mechanism for Quality Improvement; and
OE: Other Revenues

 

The Cost of Administration, Operation and Maintenance (CAOM) is composed of the following items:

Page 4


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Where:

CO: Operating Costs; and RI: Sunk Revenue.

The Annual Cost of Assets (CAA) is calculated by adding the components below:

Where:

RC: Remuneration of the capital;

QRR: Regulatory Reintegration Quota; and

CAIMI: Annual Cost of Movable and Stationary Facilities

3. SPECIAL RATE REVISION

In addition to the annual updates and periodic revisions, the Brazilian Electricity Regulatory Agency (ANEEL) may, at the request of the regulated agent or the granting authority, carry out an Special Rate Revision (RTE) at any time, when an event causes a significant economic-financial imbalance in the concession.

These regulatory rules are consolidated in the Rate Regulation Procedures (PRORET) which have a normative character and consolidate the regulation on the rate-related processes.

Concerning the electric energy distribution, there is a set of rate-related procedures that include specifics for each contractual situation, as detailed below:

(i) For distribution companies that had their concession agreements extended under Decree 8,461/2015, and signed contract additions with new economic clauses, or signed contract additions for full compliance with the new concession model;

Page 5


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

(ii) For distribution concessionaires that have not yet had their concession agreements extended; and

(iii) For the group of distribution companies that are responsible for the provision of energy distribution services in concessions that were not extended (the so-called designated distribution companies), the rules specified in Normative Resolution 748/2016 and Approving Resolution 2,184/2016 will be enforced.

For the designated dealers, confirming the conditions addressed in Technical Note 331/2016, REN 748/2016 and REH 2,184/2016 definitively established the additional conditions to be applied to designated dealers. Specifically, REN Nr. 748/2016 sets terms and conditions for the provision of services by designated dealers, related to revenue, RGR loans, management, delinquency, use of CCC resources and Special Sanctions Regime. In turn, REH 2,148/2016 approves regulatory limits for losses in rate-related processes up to 2025 for CERR, Amazonas Energia and Boa Vista Energia, and determines, for the group of designated dealers, the limits of losses, operating costs and continuity indicators to be reached by the end of 2017, in addition to the amount of the Appropriate Reference Compensation to be used in the RGR loans. This approving resolution also establishes that

RGR’s resources will be used to ensure the minimum sustainability conditions for the service, according to Ordinance 388/2016-MME, with terms, deficiencies and rates indicated in REN Nr. 748/2016, considering that the obligations assumed will be passed on to the new utility provider. Finally, the resolution establishes the exceptional regime of regulatory sanctions, prioritizing the exclusively directive and/or determining nature, without imposing penalties.

According to the result of public hearing nr. 094/2016 given in Technical Note nr. 182/2017, transitional provisions were considered regarding the use of compensation for breach of quality limits regarding the continuity of the service and the level of tension for investments. There are also limitations to the economic-financial management, which must follow certain criteria, as presented in Clause One.



 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017


Where the Regulatory Reintegration Quota (QRR) can be considered as regulatory depreciation and would be treated as the investments necessary to replenish the existing asset base.

In case of non-compliance with the proposed criteria for two consecutive years, the distribution of benefits and the operations of related parties will be limited. From the seventh calendar year following the signing of the agreement, failure to comply with these efficiency criteria will lead to the opening of the forfeiture procedure.

In May 2017, ANEEL put Technical Note 088/2017 through a public hearing, aiming to collect subsidies to improve the calculation of the regulatory parameters and service quality indicators that will be the foundation for the next rate revision processes of the dealers controlled by Eletrobras. The Technical Note determines that the regulatory parameters for the rate-related processes of the designated dealers, defined in the terms of Normative Resolution Nr. 748/2016 and of limits for collective continuity indicators DEC and FEC. It is proposed that the value of the T component of the X Factor be equal to 0 (zero) for every approved rate-related process until the first rate revision process subsequent to the signing of the Concession Agreement, while for the Pd component, it is recommended that the value in force is maintained until the first rate-related process after the signing of the Concession Agreement, when its value will be zero.

In July 2017, ANEEL announced Technical Note 351/2017 in response to a request from the Ministry of Mines and Energy to assess the conditions of balancing of the concessions of the designated dealers and the technical feasibility of re-balancing prior to its sale/public bidding process. This Technical Note identifies that the new utility providers would find a scenario of serious imbalance when they took on the new concession, mainly in relation to the operating costs and non-technical losses. As a solution, ANEEL considers the possibility of recognizing, in the rate, a portion exceeding the regulatory one for nontechnical losses and operating costs. This exceeding portion would correspond to the average between the regulatory values and carried out for such indicators.

Also in Technical Note 351/2017, the impacts of loosening the conditions for payment of RGR loans were assessed, as well as of allowing dealers to pass on their expenses on their rates, which would favor the bidding process and the assumption of a new utility provider.

Page 7


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

The impact of the rate was simulated considering a 5% interest per year and 25 years to be paid. According to ANEEL, the loan would start being paid in the sixth year after the signing of the agreement and would go until it expires (25 years for full payment). Considering that the loan will now be recognized in the rate, it will be neutral for the new utility provider in terms of the payment of the amounts involved. Initially, it was expected that the RGR’s resources would be loans that the new controller would have to pay in 48 months after the signing of the concession agreement, with 12 months grace period of interest and principal and 36 months of amortization, at the cost of 111% the SELIC (the main interest rate in Brazil).

After the issuance of Technical Note 351/2017, ANEEL, through regular mail to the Brazilian National Bank for Economic and Social Development (BNDES), presented clarifications, among which the following stand out: (i) that the rate increases related to non-technical losses and operating costs would be maintained until the first ordinary rate revision (2023); (ii) these increases would be applied to the 2017 annual rate-related processes; and (iii) in the additional revision expected to occur during the first cycle, before 2023, it should not change the relaxation pointed out and will exclusively deal with the rate recognition of the reassessed asset base. This additional revision will give the controller the option to open the armored asset base of the companies.

Due to Technical Note 351/2017, the Ministry of Mines and Energy proposed amendments to the draft of the Concession Agreement for the Electric Power Distribution Utility Service through Technical Note 247/2017. The main changes proposed relate to (i) the recognition, in Portion A, of the revenue required to pay RGR’s loans; and (ii) the definition, as a special revision or rate-related process, that the new controller may request in the five-year period of the new concession, basically to consider the volume of investments in the asset base.

On September 8th, 2017, ANEEL reopened Public Hearing 032/2017, through Technical Note Nr. 149/2017, in which the regulator presented his proposal to relax some of the regulatory parameters non-technical losses, operating costs and X Factor and of the limits for the collective continuity indicators DEC and FEC until the first general rate revision of the new controller of the designated companies, to be hired through a public bidding process.

Page 8


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

 was to maintain the Pd Component in force until the new utility provider enters. Concerning the value equal to zero in the first two cycles of this X Factor component, this topic will be addressed in the concession agreement, subject of Public Consultation 37/2017 of the Ministry of Mines and Energy. In addition, since the rules of the Rate Regulation Procedure (PRORET) will not apply to operating costs until the first ordinary rate revision (2023), which means that the T Component of the X Factor will be zero during that period.

In approving the draft of the concession agreement to be used by the Ministry of Mines and Energy in the public bidding process, ANEEL deemed it opportune to allow the temporary relaxation of the parameters of non-technical losses and operating costs. The reason for the relaxation is that any new utility provider, however efficient he may be, will need some time to adjust the level of losses and operating costs to Aneel’s regulatory benchmarks. Considering that, the loss of value in the first years could make the bidding process unfeasible.

Through Ordinance 346/2017 of the Ministry of Mines and Energy, the decision of the Granting Authority to relax the regulatory parameters regarding the operating costs and non-technical losses affects the composition between RGR’s rates and loans to be used for the remainder of the designated period. Rates are raised in order to re-balance the concession for the feasibility of the public bidding process and, consequently, the need for RGR loans is reduced.

Concerning the levels of non-technical losses, for two of the designated distribution companies (AmD and Boa Vista Energia), there has already been a legislative amendment (Law 13,299/2016) to allow them to pass on the rate transfer of the realized level of losses, with a path for the regulatory benchmarks in 10 years. This legislative amendment was regulated by ANEEL and REH 2,184/2016 approves the regulatory benchmarks of technical and non-technical losses of such dealers until 2025. Due to this, ANEEL considered that no additional amendment was applicable to the regulatory levels of these two distribution companies.

For the other designated distribution companies, due to the large gap between the real performance of these companies and the regulatory efficiency level, ANEEL suggested that the regulatory level of non-technical losses should remain stationary until the first general

Page 9


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

rate revision and be defined as the midpoint between the actual loss and the regulatory loss. The premise used is that the new utility provider would take five years, or one rate-related cycle, to converge to the regulatory benchmarks. In the first two years, therefore, the non-technical losses would still outweigh the rate recognition. The benchmarks would be equal during the third year, and in the last two years of the cycle, the new utility provider would have losses lower than the regulatory ones, compensating for the losses from the first two years.

The same relaxation rationale was used for the regulatory level of operating costs. During the designation period, however, an improvement in operating cost indices is being required, so that the one resulting from the reduction target of 5% per year, in actual terms, will be the one used. These figures are also included in REH 2,184/2016 and were updated based on the Extended National Consumer Price Index (IPCA) for the date of the 2016 rate adjustment. Thus, the average between the reduction target of operating costs and regulatory operating costs was adopted in the 2017 rate adjustment. As with the nontechnical losses, the premise used is that the new utility provider would take five years to converge to the regulatory benchmarks.

The figure below consolidates the rate-related impacts resulting from the relaxation of operating costs and losses:

Graph 1. Rate-related impacts resulting from the relaxation of operating costs

Page 10


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

The following table shows the original regulatory benchmarks of the 2016 rate-related process, the relaxed values and the difference between both

Table 1. Original regulatory benchmarks of the 2016 rate-related process,
the relaxed values and the difference between them

  Non-technical losses % Low Tension Market    Operating Costs   
Company   2016 Rate-Related   Relaxed regulatory   Difference   2016 Rate-Related Relaxed regulatory   Difference 
  Process  level    Process  level   
AMAZONAS  REH 2,184/2016  REH 2,184/2016    348,814,032  627,198,364  278,384,332 
BOA VISTA  REH 2,184/2016  REH 2,184/2016    37,541,148  69,062,498  31,521,350 
CEAL  15.67%  27.18%  11.51%  271,691,345  336,655,364  64,964,020 
CEPISA  13.93%  25.41%  11.48%  360,719,593  420,070,979  59,351,386 
CERON  8.33%  23.09%  14.76%  234,012,852  279,853,142  45,840,291 
ELETROACRE  11.28%  18.79%  7.51%  93,484,480  128,543,635  35,059,155 

 

As a way of preserving the continuity of the service provided, it was defined that, during the designation period, the rates used would be the ones previously applied by the local utility provider, updated according to an adjustment rule, and the difference in order to reach the minimum sustainability condition would be ensured through RGR’s loans. With the change provided for in Ordinance 346/2017 from the Ministry of Mines and Energy, and the relaxation of regulatory parameters, RGR’s loans should be reduced corresponding to the increase in revenue that will be passed on to consumers through rates. The table below shows the estimated amounts to be deducted from the loans.

Table 2. Estimated amounts to be deducted from loans
 
  Annual rise  Annual rise   Annual rise  Monthly Reduction of  Base  Monthly Amount   Reduction 
Company  PMSO (in Reais)  Losses (in Reais)  Consolidated (in  RGR’s Loans (in Reais)  Date  Currently Loaned (in  (%) 
      Reais)      Reais)   
AMAZONAS  278,384,333  -  278,384,333  23,198,694  nov/16  68,188,825  34% 
BOA VISTA  31,521,350  -  31,521,350  2,626,779  nov/16  17,885,981  15% 
CEAL  64,964,020  36,732,508  101,696,528  8,474,711  sep/16  26,658,844  32% 
CEPISA  59,351,386  46,987,751  106,339,138  8,861,595  sep/16  39,326,152  23% 
CERON  45,840,291  56,616,468  102,456,759  8,538,063  nov/16  23,985,177  36% 
ELETROACRE  35,059,155  7,050,742  42,109,897  3,509,158  nov/16  8,948,741  39% 

 

Page 11


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

ANEEL also considers appropriate the relaxation of the DEC and FEC continuity limits of the dealers prior to the public bidding process, thus, proposing the starting point to be the average between the value calculated (moving average from July 2016 to June 2017) and the limit in force in 2017. It is also proposed that these limits remain unchanged until the fifth calendar year after the signing of new contracts, so that the first general rate revision of the new contract establishes the limits for the next cycle.

Both Services A and B used the same assumptions concerning rate-related adjustments/revisions:

- In 2017 will see an adjustment (IRT) to the anniversary date of the old contracts of the designated companies. At the time of this IRT, there will be an additional regulatory recognition of non-technical losses and operating costs, as set forth in Technical Note Nr. 351/2017;

- In 2018 will see the signing of the new contracts, which date was estimated for the second quarter of 2018. In the same year, an IRT will be carried out on the anniversary date of the old contracts;

- In 2019 will see a special revision on the anniversary date of the old contracts. In this revision, the rate-related coverage of the Annual Cost of Assets (CAA), consisting in Capital Compensation, Regulatory Reintegration Quota and Annual Cost of Moving and Stationary Facilities will be reassessed, with the opening of the armored asset base. Metrics of non-technical losses, operating costs, delinquency and X Factor will not be revised;

contract, where regulatory parameters will be reestablished (such as non-technical losses, operating costs, delinquency, X Factor, etc.);

residue of the difference between Portion A and the verified revenue. Finally, the values of OR, EDand ERwill be expurgated and the values will be added with ONS;

Page 12


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

- Item Sunk Revenues” was removed from Portion B and moved to Portion A, with its percentage by consumption class defined by Aneel at the time of the rate revision;

- Neutralities are now calculated for all components of Portion A (including all other financial components, neutrality financial components and financial components of CVA balance to be offset), not only for the taxes of the sectors, as it happens under the current rules. The concept of neutrality remains unchanged, such as the difference between billed amounts and amounts contemplated in the rate-related processes, neutralizing differences arising from market changes.

Page 13


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Appendix to the DFP/DFF/DDE Joint Technical

Information

Analysis and Results of AMAZONAS ENERGIA (AmE D)

Page 1


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
Contents     
1.  objective  4 
2.  Assumptions of Economic and Financial Modeling  4 
2.1  Market and consumer units  4 
2.1.1  Evolution of Captive Market  4 
2.1.2  Market Projections prepared by Service A  5 
2.1.3  Market Projections prepared by Service B  9 
2.1.4  Assumptions used and points of diverge nce of Services A and B  14 
2.1.5  Evolution of Consumer Units  16 
2.1.6  Projections of Consumer Units prepared by Service A  17 
2.1.7  Projections of Consumer Units prepared by Service B  18 
2.1.8  Comparison of Services A and B projections  19 
2.1.9  Energy procurement  19 
2.2  Network extension (km)  20 
2.3  Electrical losses  21 
2.3.1  Projections of Electrical Losses prepared by Service A  22 
2.3.2  Projections of Electrical Losses prepared by Service B  25 

 

2.3.3      ASSUMPTIONS USED AND DIVERGENCE POINTS FROM SERVICE A AND
SERVICE B nce of Services A and B  27 
2.4  Default and Unrecoverable revenues  28 
2.5  PMSO (personnel, material, services and others)  30 
2.5.1  PMSO incurred (2012 2016)  30 
2.5.2  PMSO for Service A  32 
2.5.3  PMSO for Service B:  38 

 

Page 2


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 
  2.5.4  Forecasts and comparisons of the PMSO of Services A and B  41 
  2.6  Investment  46 
  2.7  Regulatory remuneration base  49 
  2.8  X Factor  53 
  2.9  WACC  58 
3.    The result of economic and financial assessments and by multiples  60 
  3.1  Amazonas energia unbundling  61 
  3.2  AFAC  62 
4.    Contingencies  62 
  4.1  Legal Contingencies  62 
  4.2  Accounting Contingencies  63 
5.    capital structure  65 
  5.1  Capitalization by Eletrobras and by the Investor  67 
  5.1.1 Eletrobras  67 
  5.1.2 Investor  67 
6.    Necessary adjustments at Eletrobras  67 
  6.1  Fixed Assets in use  69 
7.    Liquidation Costs  70 
8.    Analysis of the benefits and risks of disposal  71 
  8.1  Benefits  71 
  8.2  Risks  72 
9.    Final Remarks  73 

 

Page 3


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

1. OBJECTIVE

The purpose of this appendix is to include a detailed presentation of the results of the economic and financial assessments of Amazonas Energia - AmE performed by Services A and B, contracted by BNDES, as well as to detail the internal adjustments that Eletrobras needs to implement in order to enable its sale at the target price determined by the model approved under CPPI Resolution 20, dated November 8, 2017, amended by CPPI Resolution 28, dated November 22, 2017.

2.      ASSUMPTIONS OF ECONOMIC AND FINANCIAL MODELING
2.1      MARKET AND CONSUMER UNITS
2.1.1      EVOLUTION OF CAPTIVE MARKET

In the last ten years, from 2007 to 2016, AmE D had average annual growth of 3.9%, as can be seen in the tables below. Residential, Rural, Public Power and Public Lighting classes contributed positively to this development, with an annual average growth of 8.0%, 13.7%, 5.6% and 6.1%, respectively. In a negative way, notably in the last year, there was a strong reduction of the industrial class with the migration of consumers to the free market.

Table 1.  Evolution of consumption classes in MWh     
        ENERGY SEGREGATED BY CLASS (MWh)       
  2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
RESIDENTIAL  1,071,,001  1,122,,958  1,197,,666  1,317,610  1,385,603  1,555,129  1,783,151  2,011,401  2,190,231  2,124,625 
INDUSTRIAL  1,584,037  1,663,360  1,480,762  1,651,843  1,775,360  1,816,288  1,783,887  1,786,509  1,661,468  1,365,430 
COMMERCIAL  738,818  787,775  854,978  945,273  1,010,473  1,180,408  1,236,086  1,310,082  1,373,966  1,259,733 
RURAL  24,666  29,115  35,342  46,231  58,696  68,853  75,940  80,525  81,846  74,457 
GOVERNMENT  368,370  381,394  406,402  443,403  452,917  497,846  551,179  612,175  635,672  592,961 
PUBLIC LIGHTING  100,389  105,112  112,091  115,958  120,941  138,263  192,467  166,097  168,079  160,059 
PUBLIC SERVICE  166,391  173,768  186,474  195,962  198,365  220,590  220,747  198,526  176,012  150,338 
OWN CONSUMPTION  81,986  92,474  88,144  99,107  82,991  118,952  54,582  25,801  22,171  56,725 
TOTAL CONSUMPTION  4,135,657  4,355,956  4,361,859  4,815,387  5,085,346  5,596,329  5,898,040  6,191,116  6,309,446  5,784,328 

 

Page 4


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
  Table 2.    Composition of Consumption Class       
          COMPOSITION OF CLASSES         
  2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
RESIDENTIAL  25.9%  25.8%  27.5%  27.4%  27.2%  27.8%  30.2%  32.5%  34.7%  36.7% 
INDUSTRIAL  38.3%  38.2%  33.9%  34.3%  34.9%  32.5%  30.2%  28.9%  26.3%  23.6% 
COMMERCIAL  17.9%  18.1%  19.6%  19.6%  19.9%  21.1%  21.0%  21.2%  21.8%  21.8% 
RURAL  0.6%  0.7%  0.8%  1.0%  1.2%  1.2%  1.3%  1.3%  1.3%  1.3% 
GOVERNMENT  8.9%  8.8%  9.3%  9.2%  8.9%  8.9%  9.3%  9.9%  10.1%  10.3% 
PUBLIC LIGHTING  2.4%  2.4%  2.6%  2.4%  2.4%  2.5%  3.3%  2.7%  2.7%  2.8% 
PUBLIC SERVICE  4.0%  4.0%  4.3%  4.1%  3.9%  3.9%  3.7%  3.2%  2.8%  2.6% 
OWN CONSUMPTION  2.0%  2.1%  2.0%  2.1%  1.6%  2.1%  0.9%  0.4%  0.4%  1.0% 
TOTAL CONSUMPTION  100%  100%  100%  100%  100%  100%  100%  100%  100%  100% 

 

  Table 3.  Annual Growth Rate in the Period     
          ANNUAL GROWTH RATE       
                    ANNUAL 
  2008  2009  2010  2011  2012  2013  2014  2015  2016  GROWTH 
                    AVERAGE 
RESIDENTIAL  4.9%  6.7%  10.0%  5.2%  12.2%  14.7%  12.8%  8.9%  -3.0%  8.0% 
INDUSTRIAL  5.0%  -11.0%  11.6%  7.5%  2.3%  -1.8%  0.1%  -7.0%  -17.8%  -1.2% 
COMMERCIAL  6.6%  8.5%  10.6%  6.9%  16.8%  4.7%  6.0%  4.9%  -8.3%  6.3% 
RURAL  18.0%  21.4%  30.8%  27.0%  17.3%  10.3%  6.0%  1.6%  -9.0%  13.7% 
GOVERNMENT  3.5%  6.6%  9.1%  2.1%  9.9%  10.7%  11.1%  3.8%  -6.7%  5.6% 
PUBLIC LIGHTING  4.7%  6.6%  3.4%  4.3%  14.3%  39.2%  -13.7%  1.2%  -4.8%  6.1% 
PUBLIC SERVICE  4.4%  7.3%  5.1%  1.2%  11.2%  0.1%  -10.1%  -11.3%  -14.6%  -0.7% 
OWN CONSUMPTION  12.8%  -4.7%  12.4%  -16.3%  43.3%  -54.1%  -52.7%  -14.1%  155.9%  9.2% 
TOTAL CONSUMPTION  5.3%  0.1%  10.4%  5.6%  10.0%  5.4%  5.0%  1.9%  -8.3%  3.9% 

 

2.1.2 MARKET PROJECTIONS PREPARED BY SERVICE A

The consumption of electrical energy in Brazil can be segregated into eight distinct classes, which are: Residential, Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service and Own Consumption. In order to capture the possible heterogeneities among the analyzed groups, Service A separately modeled the Residential, Commercial, Industrial, Rural and Other sectors (which compose the classes of Public Power, Public Lighting, Public Service and Own Consumption). This decision was based on the similar behavior of

Page 5


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

consumption among them and to minimize the effect of the volatility incurred in low consumption classes.

To project energy demand for the next 30 years, ending in February/2048, the consultancy first made a historical analysis of the evolution of energy consumption and its distributions between classes, using an observable period of consumption from 2007 to 2016. The results are presented in the following tables:


  Table 4.  Geometric Mean of Consumption 2007-2016   
  RESIDENTIAL  INDUSTRIAL  COMMERCIAL  RURAL  OTHER  TOTAL 
Period  Consumption  Consumption  Consumption  Consumption  Consumption  Consumption 
  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a 
2007 - 2016  7.92%  -1.29%  6.44%  13.22%  2.97%  3.91% 

 

a. Residential Class Projection

For Residential class, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Ordinary Least Squares (OLS) methodology, with a dependent variable in the model, the residential per capita energy consumption variation, and as explanatory variables a trend line and the annual GDP of Brazil were used. From the estimated forecasts, consumption grows on average 3.74% p.a. between 2017 and 2047.

Page 6


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

b. Industrial Class Projection

A model was designed to project the Industrial consumption of energy through the GDP

(Industrial) and an exponential smoothing factor Ê = 96%, where it is assumed that Ê has effect from 2022. The estimated GDP coefficient is ± = 1.33. The following is the model definition:

. = Industrial Consumption.

From the estimated forecasts, consumption grows on average 2.75% p.a. between 2017 and 2047.

c. Commercial Class Projection

To design the Commercial consumption of energy, a model of time series SARIMA was elaborated.

From the estimated forecasts, consumption grows on average 2.20% p.a. between 2017 and 2047.

d. Rural Class Projection

To design the rural energy consumption, an ARIMA time series model was estimated.

From the estimated forecasts, consumption grows on average 3.72% p.a. between 2017 and 2047.

e. Other Class Projection

The model created to design the Other class was based on the premise that there is a linear dependency relationship between it and the Residential consumption class. The model can be defined as:

. = Other Consumption; c.r = Residential Consumption; and

Page 7


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Where ± (t) is the proportion that others represent of residential consumption at time t.Wherein:

 


Based on the estimated forecasts, the consumption of the classes Public Power, Public Lighting, Public Service and Own Consumption, classified in the Others class, grew by an average of 2.68% p.a. between 2017 and 2047.

f. Global Projection Results by Consumption Class

After elaborating the projection of each sector separately, the results were grouped on a consolidated basis. Below are graphs of the projection of aggregate energy consumption, as well as a general table of growth per class. There is also an increase in the annual variation of consumption in the first years, reaching the highest value in 2019 (above 4.0%), slowly reducing the annual growth rate to 2.5% in 2041.

Page 8


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

  Table 5.  Geometric Mean of Consumption 2017-2047   
  RESIDENTIAL  INDUSTRIAL  COMMERCIAL  RURAL  OTHER  TOTAL 
Period  Consumption  Consumption  Consumption  Consumption  Consumption  Consumption 
  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a 
2017 - 2047  3.74%  2.75%  2.20%  3.72%  2.68%  3.05% 

 

g. Projection of Demand by Voltage Level

For the projection of demand by Voltage Level, the historical average of low voltage of the last 5 years of the Company was used. This metric was based on the idea of stability in the composition of energy consumption between classes. Low voltage consumption was defined and the residual percentage was divided between medium and high voltage according to the SPARTA report. The set percentages for 2017-2047 are 51% for LV, 32% for AV and 17% for TA.

2.1.3 MARKET PROJECTIONS PREPARED BY SERVICE B

In order to know the reality experienced by the Distribution Company and to use the historical values for the necessary projections, Service B initially collected consumption

Page 9


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

information made by the utility from 2013 to 2016. The evolution of the Market by level of voltage in the period is presented in the table below:

Table 6.  Market Evolution by Voltage Level   
  Market per Voltage Level(MWh)      Variation Rate 
Voltage/Year  2013  2014  2015  2016  Period  Geometric 
          (2013/2016)  (p.a.) 
SUPPLY  5,773,899  6,046,889  6,309,520  6,079,604  5%  2% 
A3  1,008,689  1,049,049  1,053,644  950,666  -6%  -2% 
A3a  -  -  -  -  -  - 
A4  1,905,060  1,843,860  1,927,323  1,824,456  -4%  -1% 
LV  2,860,149  3,153,980  3,328,553  3,304,482  16%  5% 
LOW INCOME  81,420  204,978  299,034  281,484  246%  51% 
SUPPLY  -  -  -  -  -  - 
OTHER FREE  -  -  -  -  -  - 
DISTRIBUTION  -  -  -  -  -  - 
TOTAL  5,773,899  6,046,890  6,309,521  6,079,605  5%  2% 

 

In order to subsidize the various tariff calculations and the flow of the company, Service B elaborated market simulations by: (i) consumption class (Residential, Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service, Own Consumption); and (ii) voltage level (High Voltage - A1, A2 and A3 -, Medium Voltage - A3a and A4 - and Low Voltage - LV).

a. Market Projection by Voltage Level

The market projection by voltage level took into account the evolution of the share of voltage levels in the total market since 2012, according to information made available by the utilities. From the analysis of the behavior of these shares, it was possible to construct an evolution curve, based on linear extrapolation, generating the market projections for each voltage (High, Medium and Low). The evolution of the share of each voltage level is presented below:

Page 10


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Graph 4. Participation of the level of voltage in the total market (%) -


b. Projection of the Market by Classes of Consumption

In each consumption class, the models may differ depending on the distribution company. After numerous tests, the auxiliary variables used in the models by consumption classes are presented below. The model adjustments followed a hierarchical logic prioritizing the variables that appear first in the "Tested" column of the Table below.

Table 7.  Variables used in market forecasts by class 
  Variables 
Class     
  Tested  Used 
Residential  Population or UC RESIDENTIAL  Population 
Industrial  GDP and Employment  Employment 
COMMERCIAL  GDP  GDP 
Rural  GDP and Rural Population  GDP 
Government  Population and/or GDP  Population 
Public Lighting  Univariate  - 
Public service  Population  Population 
Own Consumption  Univariate  - 

 

Page 11


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

b.1 Residential Class Result

For the Residential Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Population as an auxiliary variable and presence of seasonality, using the observed consumption series from January 2010 to November 2016. From the estimated forecasts, consumption grows on average 3.42% p.a. between 2017 and 2048.

b.2 Industrial Class Result

For the Industrial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with the number of formal jobs, trend component and seasonality as the auxiliary variables, using the series observed from January 2009 to November 2016. From the estimated forecasts, consumption grows by an average of 3.05% p.a. between 2017 and 2048.

b.3 Commercial Class Result

For the Commercial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with the State GDP, trend component and presence of seasonality, as auxiliary variables using the observed consumption series from January 2009 to November 2016. From the estimated forecasts, consumption grows on average 3.55% p.a. between 2017 and 2048.

b.4 Result of the Rural class

For the Rural Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with the State GDP and momentary dummy in May 2016, as auxiliary variables, using the observed consumption series from January 2012 to November 2016. From the estimated forecasts, consumption grows on average 2.61% p.a. between 2017 and 2048.

b.5 Result of the Public Power class

For the Public Power class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Population and presence of seasonality as auxiliary variables, using the observed consumption series from January 2010 to November 2016. From the estimated forecasts, consumption grows on average 3.15% p.a. between 2017 and 2048.

Page 12


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

b.6 Result of Public Lighting class

For the Public Lighting class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the ETS methodology, using the observed consumption series from January 2010 to November 2016. From the estimated forecasts, consumption grows on average 1.04% p.a. between 2017 and 2048.

b.7 Result of the Public Service class

For the Public Service class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Population as auxiliary variable, structural break in February 2013 onwards and momentary dummies in August and September of 2015, in addition to the presence of seasonality, using the observed consumption series from January 2004 to November 2016. From the estimated forecasts, consumption grows on average 2.36% p.a. between 2017 and 2048.

b.8 Result of class Own Consumption

For the Own Consumption class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the ETS methodology, using the observed consumption series from January 2010 to November 2016. From the estimated forecasts, consumption grows on average 0.58% p.a. between 2017 and 2048.

b.9 Global Projection Result

The table below shows the market projected by consumption class broken down every ten years between 2016 and 2046. The consumption in MWh per class is presented through its percentage participation in relation to the total consumption and the percentage growth between the ten-year periods. The graph below illustrates the composition of consumption classes in 2046.

Page 13


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
Table 8.  Market Projected for each Decade by Consumption Class     
  MWh Participation  Average Growth p.a. 
CLASS                       
  2016  2026  2036  2046  2016  2026  2036  2046  2017/2026  2027/2036  2037/2046 
Residential  2,132,352  3,761,090  4,899,817  6,069,049  36.49%  39.96%  39.26%  38.83%  5.84%  2.68%  2.16% 
Industrial  1,405,186  2,145,984  2,888,361  3,546,515  24.04%  22.80%  23.14%  22.69%  4.33%  3.02%  2.07% 
Commercial  1,297,228  1,979,395  2,818,259  3,751,829  22.20%  21.03%  22.58%  24.00%  4.32%  3.60%  2.90% 
Rural  75,284  101,504  128,671  162,649  1.29%  1.08%  1.03%  1.04%  3.03%  2.40%  2.37% 
Government  598,876  959,071  1,235,091  1,543,543  10.25%  10.19%  9.90%  9.88%  4.82%  2.56%  2.25% 
Public Lighting  159,353  191,484  206,620  219,265  2.73%  2.03%  1.66%  1.40%  1.85%  0.76%  0.60% 
Public service  154,012  248,513  278,926  311,018  2.64%  2.64%  2.23%  1.99%  4.90%  1.16%  1.09% 
Own  21,780  25,769  25,781  25,792  0.37%  0.27%  0.21%  0.17%  1.70%  0.00%  0.00% 
Total  5,844,071  9,412,810  12,481,526  15,629,660  100.00%  100.00%  100.00%  100.00%  4.88%  2.86%  2.27% 
 
  Graph 5. Participation of Consumer Classes - 2046       

 

2.1.4      ASSUMPTIONS USED AND POINTS OF DIVERGENCE OF SERVICES A AND B
1.      The assessment reports lead to the conclusion that Service A used annual consumption figures for the projections, while Service B signals in its premises that it used monthly consumption figures for the projection calculations.
2.      For the forecast of the demand, Service A elaborated simulations of projection only by class of consumption, distributed in five categories: Residential, Industrial, Commercial, Rural and Others (composed of Public Power, Public Lighting, Public Service and Own Consumption). For the Projection of demand by Voltage Level, the

Page 14


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

 

 

historical average of low voltage of the last 5 years of the Company was used. Service B has elaborated two types of simulations for forecasting demand. A projection by consumption classes, distributed in eight categories: Residential, Industrial, Commercial, Rural, Public Power, Public Lighting, Public Service and Own Consumption. And another projection per voltage level, consisting of: High Voltage (A1, A2 and A3), Medium Voltage (A3a and A4) and Low Voltage (LV).

 

3.     

The historical consumption period of the distribution company, used by Service A, is the years 2007 to 2016. Service B, on the other hand, used periods beginning in 2004, 2009, 2010 or 2012, depending on the projected consumption class and always ending in November 2016.

 

4.     

Service A obtained average annual growth projections of 3.05% for the distribution company consumption in the period from March 2018 to February 2048. For Service B, the average annual growth of the distribution company’s consumption was around 3.41% for the period from 2017 to 2048.

 

 

Based on the Figure that describes the Projection of the Energy Consumption of Service A, we see that the consumption projected by that of Service A is around 14,000 GWh in 2047 (in Service A report, the absolute value of the consumption projection was not observed). Service B projects a consumption of 15,630 GWh in 2046. In absolute terms, with an approximate difference of 1,630 GWh higher for Service B, this is due to a lower annual average growth of 0.36% of Service A compared to that of Service B.

 

  The average annual growth of the consumption classes of Service A and B are presented in the table below:
 
Table 9.  Geometric Mean of Consumption of Services A and B for the period 
Geometric Mean of Consumption of Services A and B
  RESIDENTIAL  INDUSTRIAL  COMMERCIAL  RURAL  GOVERNMENT  PUBLIC  PUBLIC   OWN  TOTAL 
Services  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a   LIGHTING S  ERVICE  ”% p.a  ”% p.a 
            ”% p.a  ”% p.a     
A  3.74%  2.75%  2.20%  3.72%  2.68%  2.68%  2.68%  2.68%  3.05% 
B  3.48%  3.05%  3.55%  2.61%  3.15%  1.04%  2.36%  0.58%  3.41% 
Differences (A-B)  0.26%  -0.30%  -1.35%  1.11%  -0.47%  1.64%  0.32%  2.10%  -0.36% 

 

Page 15


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

2.1.5 EVOLUTION OF CONSUMER UNITS

The tables below show the evolution of the Consumer Units (UC) of consumption classes in recent years, from 2009 to 2016. The company distribution had an average annual growth of 4.6% of its UC. Of particular note are the Residential and Commercial classes, which together account for 94% of the AmE D UCs.

Table 10.  UCs Evolution by Consumption Classes   
      CONSUMER UNITS BY CLASS       
  2009  2010  2011  2012  2013  2014  2015  2016 
RESIDENTIAL  587,996  590,191  629,085  655,587  698,281  731,759  767,553  815,171 
INDUSTRIAL  3,328  3,205  3,108  3,177  3,145  3,056  3,018  2,917 
COMMERCIAL  57,682  60,981  65,858  67,548  70,150  71,821  73,439  74,649 
RURAL  22,515  36,158  38,076  40,498  40,795  41,079  41,212  42,320 
GOVERNMENT  6,488  6,838  7,462  8,046  8,693  8,974  9,366  9,448 
PUBLIC LIGHTING  324  329  381  643  661  661  664  664 
PUBLIC SERVICE  835  888  899  917  935  937  878  888 
OWN CONSUMPTION  290  392  385  387  398  399  353  439 
TOTAL CONSUMPTION  679,458  698,982  745,254  776,803  823,058  858,686  896,483  946,496 

 

Table 11.  UCs Composition by Consumption Classes     
        COMPOSITION OF CLASSES       
  2009  2010  2011  2012  2013  2014  2015  2016 
RESIDENTIAL  86.5%  84.4%  84.4%  84.4%  84.8%  85.2%  85.6%  86.1% 
INDUSTRIAL  0.5%  0.5%  0.4%  0.4%  0.4%  0.4%  0.3%  0.3% 
COMMERCIAL  8.5%  8.7%  8.8%  8.7%  8.5%  8.4%  8.2%  7.9% 
RURAL  3.3%  5.2%  5.1%  5.2%  5.0%  4.8%  4.6%  4.5% 
GOVERNMENT  1.0%  1.0%  1.0%  1.0%  1.1%  1.0%  1.0%  1.0% 
PUBLIC LIGHTING  0.0%  0.0%  0.1%  0.1%  0.1%  0.1%  0.1%  0.1% 
PUBLIC SERVICE  0.1%  0.1%  0.1%  0.1%  0.1%  0.1%  0.1%  0.1% 
OWN CONSUMPTION  0.0%  0.1%  0.1%  0.0%  0.0%  0.0%  0.0%  0.0% 
TOTAL CONSUMPTION  100%  100%  100%  100%  100%  100%  100%  100% 

 

Page 16


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017
 
Table 12.  Annual Growth Rate in the Period
  ANNUAL GROWTH RATE  
  2009  2010  2011  2012  2013  2014  2015  2016  ANNUAL GROWTH
 AVERAGE
 
RESIDENTIAL  2.6%  0.4%  6.6%  4.2%  6.5%  4.8%  4.9%  6.2%  4.5% 
INDUSTRIAL  -0.4%  -3.7%  -3.0%  2.2%  -1.0%  -2.8%  -1.2%  -3.3%  -1.7% 
COMMERCIAL  1.5%  5.7%  8.0%  2.6%  3.9%  2.4%  2.3%  1.6%  3.5% 
RURAL  27.3%  60.6%  5.3%  6.4%  0.7%  0.7%  0.3%  2.7%  13.0% 
GOVERNMENT  1.6%  5.4%  9.1%  7.8%  8.0%  3.2%  4.4%  0.9%  5.1% 
PUBLIC LIGHTING  2.9%  1.5%  15.8%  68.8%  2.8%  0.0%  0.5%  0.0%  11.5% 
PUBLIC SERVICE  19.5%  6.3%  1.2%  2.0%  2.0%  0.2%  -6.3%  1.1%  3.3% 
OWN CONSUMPTION  -0.3%  35.2%  -1.8%  0.5%  2.8%  0.3%  -11.5%  24.4%  6.2% 
TOTAL CONSUMPTION  3.2%  2.9%  6.6%  4.2%  6.0%  4.3%  4.4%  5.6%  4.6% 

 

2.1.6      PROJECTIONS OF CONSUMER UNITS PREPARED BY SERVICE A
A      model was used based on two parameters, the relation Population/UC Residential and

the average growth of Residential UC for the last 6 years. The same projected residential UC growth was extrapolated to the other classes. As a result of the projections, there was a growth of UCs in the first years, with initial variation of 9.0%, with a gradual reduction of growth until reaching close to 0% growth from the 32nd year of the new concession, consolidating around 2,000,000 UC. In the graph below the results of the projection follow:

Page 17


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017


2.1.7 PROJECTIONS OF CONSUMER UNITS PREPARED BY SERVICE B

Due to the differentiated behavior of the growth of the historical series of Residential UCs and other UCs, Service B chose to separately project the two series and then aggregate them, forming the series of Total UCs.

The evolution of Residential UCs is based on the growth of residential households and the electrical energy coverage that these households have over the years 2017 to 2048. For this, data from the National Household Sample Survey - PNAD (IBGE) were used for the construction of the historical household series and the historical series of Population.

The Other UCs series is based on the projection of average consumption that consists of the quotient between the consumption in MWh of classes (except residential) divided by total UC (except residential). The projection of average consumption was based on the trend of linear growth, converging to the national average consumption 1,749 kWh per year (ANEEL, 2016, considering only the distribution utilities), in up to 30 years.

The result obtained from the UC projection, with historical series up to 2048, is shown in the graph below. It can be noticed that the UC growth is constant until 2030, with a slight drop in growth as of this year. The average growth registered in the period from 2016 to 2048 was 2.85% per year.

Page 18


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Graph 7. UC Projection


2.1.8 COMPARISON OF SERVICES A AND B PROJECTIONS

Based on the trajectory of the graphs of the UCs of Services A and B we have a comparison between their evolutions over the period. It is noted that in the first 10 years the growth rate of Service A is higher than that of B, the opposite is the case in the last 20 years.

  Table 13.  Annual Growth Rate in the Period     
    Performed        Forecasted     
Services      2015/2016        2017/2026  2027/2036  2037/2046 
  2015  2016    2026  2036  2046       
      (% p. a.)        (% p. a.)  (% p. a.)  (% p. a.) 
A   896,483  946,496  5.58% 1,500,000
1,469,516 
1,800,000 2,000,000  4.71%
4.50% 
1.84%
2.51% 
1.06%
1.77% 
B  1,882,608    2,244,727 
Differences (A-B)  -  -  -  30,484  - 82.608 - 244.727  0.21%  -0.67%  -0.72% 

 

2.1.9 ENERGY PROCUREMENT

Regarding the energy procurement balance, Service B considered the existing energy contracts, the current overcontracting level and explained the treatment given, as well as the projection of the PLD. Service A has used, as an assumption, the non-treatment of overcontracting from its starting point. According to the Disclaimer of the consulting firm’s work: "In this assessment were not considered penalties in the overcontracting of energy

Page 19


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

by the assessed enterprise, since it was assumed the assumption that the new utility will be able to liquidate its positions at market prices". That is, the revenue from the sale of surplus energy in the short-term market would be equal to the distribution company’s average energy purchase cost.

2.2 NETWORK EXTENSION (KM)

The annual variation of the extension of the distribution network has an impact on the Distribution company’s projection of investments, as well as being part of the projection metric of the calculation of regulatory operating costs and the PMSO carried out by the Distributor.

The total extension of the network can be segregated into four voltage groups: low, medium and high voltage and “Luz para Todos” program.

In service A, the network extension projection started from the low, medium and high voltage network extension history between the years 2001 and 2016. Added to this history was the network extension foreseen by the Decennial Plans and PDDs of the distribution companies, from which the network projected for the years 2017 to 2021 for low and medium voltage, 2017 and 2018 for “Luz para Todos” program and high voltage from the years 2017 to 2026.

Service A presented the following projection of network extension until the year 2027:

Table 14. Network Extension Projection (km) - Service A

Service A                       
Network Projection (Km)                       
Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
LV/AV  29,389  29,952  30,577  30,887  31,110  31,540  31,970  32,400  32,830  33,260  33,690 
AT  3,342  3,700  3,999  4,379  4,379  4,385  4,484  4,570  4,619  4,910  5,016 
PLPT  2,820  6,579  6,579  6,579  6,579  6,579  6,579  6,579  6,579  6,579  6,579 
Total  35,551  40,231  41,155  41,845  42,068  42,504  43,033  43,549  44,028  44,749  45,285 

 

Page 20


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

Which equals to one of percentage growth per year, according to the respective five-year period, according to the table below:

Table 15. Percentage growth of network extension, per five-year period
Service A
Service A               
  1st Five-year  2nd Five-year  3rd Five-year  4th Five-year  5 th Five-year  6 th Five-year   
Growth annual (%)  period  period  period  period  period  period   Average 
  2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
  3.7%  1.3%  1.2%  1.1%  1.0%  1.0%  1.5% 

 

Service B did not show Network projection numbers, only presented the econometric model used to estimate the growth projection of the network extension. The total network extension of each company was projected according to a calculated elasticity for the total number of consumer units in the concession area. To determine this elasticity, panel data from the six Eletrobras distribution companies were used for the period between 2007 and 2016.

Regarding the starting point, that is, the network extension for the year 2017, according to the information contained in the assessment reports of Services A and B, the Consulting Firms started with very different information:

2.3 ELECTRICAL LOSSES

The losses of the electric distribution system are divided into two categories: Losses in the Basic Network and Losses in the Distribution Network. In turn, losses in the Distribution Network are divided into Technical Losses (PT) and Non-Technical Losses (NTP).

Losses in the Basic Network are those admitted by the SIN. The real values are calculated monthly by the CCEE, and the regulatory value is given through the average of the

Page 21


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

measurements of the last 12 months. The value of the 2016 tariff adjustment was used for the assessment report.

PTs are the loss of transport and electrical equipment. These are calculated by ANEEL in the tariff review period and remain constant for the entire tariff cycle.

PNTs are the manageable losses where ANEEL's reduction curve is estimated at each tariff review and takes into account three factors: the target, the trajectory to the target and the limits of reduction of regulatory loss. The goal is calculated based on the analysis of the comparative model of benchmarking, in which a company is established that will serve as a basis for comparison, taking into account the size of the distribution company and the socioeconomic complexity of the region in which the distribution company is located. According to the metric specified in PRORET submodule 2.6. It should be reminded that the percentage levels of PNT are related to the Low Voltage (LV) market of the utility, while the PT and Loss in RB indices are measured in relation to the Injected Energy, in accordance with the regulatory methodology.

Below are presented the percentage values of Losses, according to the Sparta worksheet of the AmE D of 2016, as well as the realized values of 2016 of the distribution company.

Table 16.  Calculation of losses - SPARTA AmE D 2016 

 

Description Regulatory  Performed 
2016
Non-Technical Percentage (w/o Low Voltage)  98.93%  117.45% 
Technical Percentage (w/o input market)  7.77%  7.77% 
% Basic Grid (w/o Input Market)  0.94%  0.94% 

 

2.3.1 PROJECTIONS OF ELECTRICAL LOSSES PREPARED BY SERVICE A

a. Projection of Losses in the Basic Network and Technical Losses

For Service Losses in the Basic Network and PT, Service A defined that for both indicators the regulatory and real values will remain constant and equal throughout the period, since the values of these losses are not related to the activity of the distribution company and its management, but rather physical processes and characteristics of the equipment used.

Page 22


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017


 

Table 17.  Level of Technical Losses and Losses in Basic Network 
% Losses / Energy Injected
    Losses in Basic 
  Technical Losses   Network
  7.77%  0.94% 

 

b. Projection of Non -Regulatory Technical Losses

Regarding the trajectory of Regulatory NTP, Service A defined in two different moments. For the period from 2017 to 2025, regulatory levels were defined based on Law 13.299 and Process Vote 48500.004245/2016-77, through Homologation Resolution No. 2.184, which stipulated specific levels of losses for the AmE D.

As of 2026, the method of calculation of the curve of regulatory losses occurred according to the definitions described in the submodule 2.6 of PRORET, taking into account three previously mentioned factors. After the comparative analysis of the listed companies with the AmE D it was defined that their benchmarking of the PNT goal would be LIGHT.

Upon definition of the target for the next year of the tariff review, the analysis of the annual reduction of regulatory losses that can have two forms of reduction for each cycle of tariff revision, is carried out. The trajectory will be the limit of reduction of regulatory losses stipulated in PRORET, in case the effort to reach the goal of benchmarking is more challenging. Otherwise, the reduction curve will have a linear trajectory until it loses benchmarking at the end of each tariff review cycle.

The trajectory of reduction of PNTs continues until reaching the level of real PNTs practiced by LIGHT, of 22.38%. Service A considered that this level of losses would be efficient, maintaining the trajectory of the real and regulatory curves of AmE D close to the end of the concession.

c. Projection of Non -Technical Real Losses

For the projection of real PNTs, three perspectives were observed: starting point, trajectory and target.

As a starting point, the PNT was considered in 2016 of 117.45%, obtained from the energy balance of the distribution company. The defined target is the level of PNT practiced by the benchmark company considered in the calculation of the target of regulatory losses, in this

Page 23


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

case, the levels practiced by LIGHT which are of 22.38%. The value of the loss of the benchmark company was obtained through the SAMP database.

For the annual PNT reduction trajectory, the observed trajectory of an efficient utility was used. The benchmark company selected was CEMAR - Companhia Energética do Maranhão. The choice of the company was based on the fact that CEMAR already has an efficient level of Non-Technical losses and has recently undergone the privatization process.

Table 18.  Starting Point and Target of Non-Technical Losses 
 
% Losses/BT Market
Non-technical losses
in 2016
117.45% 
Non-technical losses
Benchmark in 2016
22.38% 
Year in which it
reaches the regulatory
level
2019 

 

Graph 8. Evolution of Non-Technical Losses (%/BT Market)


Page 24


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

2.3.2 PROJECTIONS OF ELECTRICAL LOSSES PREPARED BY SERVICE B

a. Projection of Technical Losses (PT)

The PT projection consists of two distinct moments. For the period from 2017 to 2025,regulatory levels were defined based on Law 13.299 and Process Vote 48500.004245/2016-77, through Homologation Resolution No. 2.184, which stipulated specific levels of losses for the AmE D. The level of losses remains the same for the years of 2026 and 2027.

As of 2028, the PT projection methodology consists of evaluating the technical loss indices of each supply voltage segment (IPTS) of AmE D (NT 175/2013-SRD/ANEEL) and comparing them with average indices of compatible companies. The lowest per-segment indices (AmE D or weighted average of compatible companies) were used as the basis for an efficient PT target in 2047.

Based on the method described, a linear trajectory of the PT level in 2019 (7.62%) was traced to the PT level projected for 2047 (6.16%). In order to have a modeling that represents each tariff period, the indices that coincide with each tariff review were used, and these remain constant during the cycle. The graph below shows the projection of the percentage of regulatory technical losses.

Regarding Losses in the Basic Network, Service B does not present any projection model for the period from 2017 to 2047.

Page 25


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017


b. Projection of Non -Regulatory Technical Losses

The projections of Regulatory PNTs for designated companies were carried out based on the successive application until 2047 of the mechanisms for setting goals, starting points and trajectories detailed in NT 106/2015-ANEEL. The models considered for the projection of the trajectory of loss targets were those estimated with the database defined at the time.

For the projection of the regulatory PNTs targets, 5-year review cycles were considered, in accordance with the targets stipulated in Homologation Resolution No. 2.184/2016, applicable to the distribution companies AmE D, Boa Vista and CEA. For the years 2017 to 2023, the Regulatory PNTs of the last review cycle (3CRTP) were considered. This level is updated upon the 1st RTP of the new utility (2023).

c. Projection of Non -Technical Real Losses

During the Temporary Provision scheme, in force in 2017 and possibly at the beginning of 2018, it was considered that economic, financial and technical-managerial difficulties will limit efficiency gains in the combat against NWP. Therefore, it was defined that in the six Eletrobras distribution companies, the current level of real PNT (relative to the year 2016) will be in force until the start of the new administration in 2018.

Page 26


 
  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

From 2019 it is considered that a new phase in the combat against PNT is initiated with successive gains of efficiency and effectiveness of these activities. For AmE D, the projection foresees a reduction of the strongest PNT in the first 5 years, given the high incentive to combat such differences for the new administration. This additional reduction is projected based on the fall calculated by the equation that relates the rate of reduction and the loss threshold of NT No. 106/2015. In addition to this reduction in the first 5 years, projections were made based on the utilization of the real PNT reduction performance of companies identified, CELPA, CEMAR and CELPE. The results of the projection are reported in the graph below, prepared by Service B.


2.3.3 ASSUMPTIONS USED AND DIVERGENCE POINTS FROM SERVICE A AND

SERVICE BNCE OF SERVICES A AND B

1. Service A considered the projection of losses in the Basic Network to be 0.94% of the injected energy of the company for the entire period from 2017 to 2048. Service B did not mention these losses.

2. For Technical Losses (PT), Service A also projected them with the constant value of 7.77% of the energy injected during the concession. Service B initially used the levels established in Homologation Resolution No. 2184/2016. From 2028, Service

Page 27


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

 

 

B used as a parameter, PTs by the voltage segment of compatible companies to model an efficient PT curve, taking into account the tariff cycles. The reduction trajectory starts with PT of 7.77% in 2017, arriving in 2047 at the value of 6.42%.

 

3.     

Modeling of Service A Regulatory Non-Technical Losses (NTP) was organized in two stages. At first, from 2016 to 2026, regulatory levels describe a linear decrease reaching 2026 at 41.54%. It remained constant in 2027, and in later years, a gradual decrease to reach the level of 22.38% in 2037 and maintaining this level until the end of the concession.

 

 

In Service B, the curve of Regulatory PNTs was performed based on 2 periods: 1) From 2017 to 2025, the levels of regulatory losses stipulated in RH 2.184/2016; and 2) from 2026 to 2047, the modeling addresses the application of the mechanisms detailed in NT 106/2015-ANEEL. The reduction trajectory starts with 93.18% in 2017, reaching 25% in 2047 on the LV Market.

 

4.     

The trajectory of real PNTs prepared by Service A starts from the value in 2016 of 117.45%, the curve reduces until 2031, when it reaches the value realized by LIGHT (benchmarking) of 22.38% of the market, remaining at this level until the end of the concession. For Service B, in the years 2017 and 2018, the real PNT indices will remain constant at 117%, as of 2019, a new phase in the efficient combat against PNT will begin. The trajectory shows a marked reduction of losses in the first 5 years and thereafter the curve decreases gradually, taking as a reference the reductions in Real PNT of the companies CELPA, CEMAR and CELPE.

 

  Thus, at the end of 2047, the projection of the real NTP index is 22.0%.

2.4 DEFAULT AND UNRECOVERABLE REVENUES

The Projected Unrecoverable Revenues and Default is the median of billed value that was not received until the base date of December 2016, the stabilization point of the unbilled revenue curve, from the 49th to the 60th month.

The default assumption has great relevance in the Valuation analysis because it directly affects the collection, consequently, the company's cash flow, negatively impacting its Valuation.

Analyzing the two Services, there are some divergences, which are listed below:

Page 28


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

It should be noted that Service B numbers were estimated in a linear decline reaching the regulatory level in 10 years, according to text contained in the Assessment Report, without presenting the annual values used.

Below are tables with the numbers of the default up to 2027 of the two Services:

  Table 19.  Default until 2027 - Services A and B     
 
Service B                       
Unrecoverable revenue                     
Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
Residential  2.95%  2.84%  2.74%  2.63%  2.52%  2.42%  2.31%  2.20%  2.09%  1.99%  1.88% 
Commercial  2.92%  2.73%  2.53%  2.34%  2.14%  1.95%  1.76%  1.56%  1.37%  1.17%  0.98% 
Industrial  5.88%  5.44%  5.00%  4.57%  4.13%  3.69%  3.25%  2.81%  2.38%  1.94%  1.50% 
Rural  6.21%  5.79%  5.36%  4.94%  4.52%  4.10%  3.67%  3.25%  2.83%  2.40%  1.98% 
Government  2.95%  2.68%  2.42%  2.15%  1.88%  1.62%  1.35%  1.08%  0.81%  0.55%  0.28% 
Public lighting  1.74%  1.57%  1.40%  1.23%  1.06%  0.90%  0.73%  0.56%  0.39%  0.22%  0.05% 
Public Service  0.78%  0.71%  0.64%  0.57%  0.50%  0.43%  0.36%  0.29%  0.22%  0.15%  0.08% 
Service A                       
Unrecoverable revenue                     
Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
Residential  5.22%  5.00%  4.78%  4.57%  3.92%  3.36%  2.88%  2.47%  2.36%  2.36%  2.36% 
Commercial  2.25%  2.25%  2.24%  2.24%  1.92%  1.64%  1.41%  1.21%  1.03%  0.89%  0.78% 
Industrial  2.18%  2.14%  2.10%  2.06%  1.76%  1.69%  1.69%  1.69%  1.69%  1.69%  1.69% 
Rural  4.90%  4.90%  4.90%  4.90%  4.90%  4.90%  4.90%  4.90%  4.90%  4.90%  4.90% 
Government  1.15%  1.06%  0.98%  0.90%  0.78%  0.67%  0.57%  0.49%  0.43%  0.37%  0.31% 
Public lighting  0.90%  0.82%  0.75%  0.69%  0.57%  0.47%  0.38%  0.32%  0.31%  0.31%  0.31% 
Public Service  5.64%  5.40%  5.17%  4.94%  4.18%  3.53%  2.99%  2.52%  2.13%  1.80%  1.52% 

 

In Service A there is not a specific year to reach the regulatory level, as performed by Service B, below. A target was established, above the regulatory level, and an annual

Page 29


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

reduction curve based on the rate of change proposed by ANEEL in 4CRTP, 3CRTP and 2CRTP. With this assumption, the regulatory level is not reached during the thirty years of concession.

Service A raised R$ 77.417 million in Unrecoverable Revenues for the year 2016.

Service B did not show any value.

The year-on-year comparison of the projection of default rates by each consulting was hampered because Service B did not disclose the annual amounts in a table, showing only the starting value and the year that the regulatory level is reached.

2.5      PMSO (PERSONNEL, MATERIAL, SERVICES AND OTHERS)
2.5.1      PMSO INCURRED (2012 2016)

For the analysis of the PMSO item contained in the Services A and B reports, it is necessary to look, in advance, at the history of this item over the last few years. For the case, the PMSO was analyzed between the period from 2012 to 2016 (same period used by Service A). Below are the values in the aforementioned period for AmE D:

Table 20.  PMSO incurred from 2012 - 2016 - R$ million   
PMSO  2012  2013  2014  2015  2016 
Personnel  327.2  354.5  356.4  333.5  357.8 
PID  0.0  25.9  0.0  0.0  0.0 
Material  46.8  38.8  37.4  35.4  25.4 
Third-party services  251.7  249.8  248.6  246.4  235.4 
Others  29.0  51.2  101.6  40.4  33.7 
Total  654.7  720.3  744.0  655.7  652.3 

 

For the composition of each item above, the Financial Statements of the respective years were used as support material. In item "O" of the PMSO, the following items were considered: Leases and Rentals; Taxes; Miscellaneous indemnities; Judicial costs and fees; Own energy consumption; Inspection Fee and Others. It is also worth noting that the PID (Resignation Incentive Plan) was considered for the years 2013 and 2014.

Page 30


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

In order to show consistency in the comparison with the Assessment Reports presented by the Consulting Firms, the PMSO values were updated by the IPCA for December 2016.

Table 21.  PMSO incurred from 2012 - 2016 updated to Dec/16 - R$ million 
PMSO  2012  2013  2014  2015  2016 
Personnel  438.1  448.7  423.9  359.6  357.8 
PID  0.0  32.8  0.0  0.0  0.0 
Material  62.7  49.2  44.5  38.2  25.4 
Third-party services  337.0  316.2  295.7  265.7  235.4 
Others  38.9  64.8  120.8  43.6  33.7 
Total  876.6  911.6  884.8  707.1  652.3 

 

In relation to regulatory operating costs, it is possible to observe the estimated values for AmE D in the years 2013 (last tariff review) until 2016, brought to the base date of December 2016.

Regarding AmE D, it can be observed that the regulatory PMSO is well below compared to that realized. It should be noted that only the P item of the PMSO incurred is already equivalent to the total of the regulatory PMSO. This fact can be seen in the graph below:


Page 31


 

  Date: 30/11/2017

IT DFP/DFF/DDE
001/2017

 

2.5.2      PMSO FOR SERVICE A
-      Historical part:

For the composition of the historical part, the consultancy, based on the accounting statements provided by the distribution company, made a grouping of similar accounts in terms of type of expenses and also made an adjustment of the values for the base date of December 2016. Below is the table containing the opening of the PMSO for AmE D as well as some comments:

Table 22.  PMSO values calculated by the consultancy (base date - Dec/16) - 
R$ million
PMSO  2012  2013  2014  2015  2016 
Personnel  254.7  278.9  262.7  247.7  282.6 
PID  0.0  0.0  0.0  0.0  0.0 
Material  13.5  10.2  10.0  9.3  5.3 
Third-party Services  270.1  238.6  230.6  232.9  191.1 
Others  36.0  45.5  125.0  67.9  112.5 (*) 
Total  574.3  573.2  628.2  557.7  591.4 (*) 

 

(*) The amount of R$ 112.5 million shown in the table for item Other does not match the rationale described in the consulting report. The correct amount is R$ 2,615.8 million, which generates a total PMSO of R$ 3,094.8 million. This fact deserves to be highlighted, since the sum of subitem Other that is within the opening of item Other of the PMSO was not considered. It may be the result of a waste, but there was no mention of it in the report.

- Methodology used in Service A: (part common to all Distribution companies)

The analysis of the PMSO as well as the projection of the values was divided into two parts, namely:

Part 1 - Characterization of Analytical Accounts (Divided in 5 phases):

  • Phase 1 - Grouping of analytical accounts:
      o      Adjustment for inflation of the accounts for the base date of December 2016;
      o      Grouping of accounts according to similarity of expense type;

    Page 32


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

  • Phase 2 - Analysis of the nature of the account in fixed and variable:
      o      Analysis of Fixed and Variable Accounts;
      o      The Accounts identified as fixed, when projected, were calculated as the average of the expenses of the sample period, excluding the outliers;
      o     

    For the projection of the variable accounts, an indicator of this account was calculated first as a function of Consumer Units, Network Expansion (Km) and Consumption, in order to later carry out the projection. The outliers were also excluded.

     

  • Phase 3 - Identification of outliers:
      o      The values are considered outliers when:
        §      They are outside the established limits of normality of the sample, i.e.:
         
  • Out Upper Limit => (sample mean) + (standard deviation of sample); Out Lower Limit => (sample mean) - (standard deviation of sample).
      o      After this procedure, the nature of the expense and its periodicity in the last 3
        years     

    were assessed in order to evaluate whether or not it would be excluded.

     

  • Phase 4 - Assessment of the best correlations for the variable accounts:
      o     

    At this phase, a study was made to verify which variables were more related like the expenses of PMSO. It was concluded that Network Extension (km) and Consumption in MWh obtained higher ratio;

     

  • Phase 5 - Identification of regulatory accounts (considered by and paid by
      ANEEL):     
      o      Consists of the classification of the accounting accounts forming the PMSO according to the characterization defined by ANEEL for regulatory PMSO. Thus, for each component account of the analytical PMSO, a classification between Effective
        PMSO      (total), Regulatory PMSO and other PMSO (Effective PMSO - Regulatory
        PMSO) is provided).

    Page 33


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Part 2 - Projection of Effective and Regulatory PMSO Grouping (Divided into 4 phases):

    · Phase 1 - Projection of Expense for 2017:

    O It considers a variation of expense in line with the company's history, taking into account the fixed or variable nature of the expenses that make up the PMSO; O Variable Expenses: are transformed into indicators in which the denominator is the parameter with which the variable had the best correlation (it can be network km or consumption in MWh).

         Example: Personnel = R$/Km of network => item P will increase as the network Km increases; O Fixed Expenses: fixed expenses were calculated as the average expenses for the sample period, excluding outliers.

    · Phase 2 - Operating Costs variation rate: This phase is divided into three steps:

    O Step 1 - Definition of Operational Cost Benchmarks;

    § For Northern Companies (CERON, Boa Vista Energia, Eletroacre and AmE

    D) => Celpa and Celtins;

    § For the Northeast Companies (Cepisa and Ceal) => Celpe, Cemar,

    Coelce, Energisa Paraíba, Energisa Sergipe, Coelba and Cosern.

    O Step 2 - Rate Variation Definition and Target between 2018 and 2022:

    § It is considered the DEA (Data Envelopment Analysis in Portuguese) database of efficient costing from Aneel for 4CRTP;
    § The PMSO expense of each company is parameterized by the km of Network Extension;
    § Based on the indicators, the average of the group of each distribution company according to the Region is calculated;
    § After this step, proceed as follows:

    Ø Distribution companies with the indicator greater than the

    group mean adjust their cost in terms of indicator until the next Tariff Review in 2023.

    Page 34


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

       
  • On the other hand, if the indicator is less than the mean of the group, the Distribution companys plot the indicator level between 2023 and 2027 as target, which is lower than the group mean.
    o      Step 3 - Rate Variation Definition and Target between 2023 and 2027:
      §      It consists of calculating the mean indicator of the privatized Benchmarks of the groups;
      §      Upon reaching the indicator of R$/km of network of the cycles target, the distribution companies stabilize operating cost indicators parameterized per consumer unit.
  • Phase 3 - Identification and Treatment of Accounts with Behavior Linked to
      the      Distribution company’s Improvements:
      o      It consists of identifying accounts that influence the actual cost of PMSO, but which vary according to the distribution company’s operational improvements.
  • Phase 4 - Projection of the Concession Period (2018 to 2047): It consists of
      estimating      the levels of PMSO expenses of Distribution companies over the
      concession      period.
      o      Based on the targets set previously, the Distribution companies reach the operational cost benchmarks until the end of 2027;
      o      Thereafter, the network PMSO/km indicator stabilizes itself, as a consequence, the absolute value of the PMSO varies in real terms through the variation of the Network km Extension.

    Efficient Costs:

    The elaboration of the estimate of the efficient operational costs was based on Annex I -Efficiency Intervals of the PRORET sub-module 2.2 V2. Operating costs tend to align with their respective benchmarks in two moments, the first ending 2022 and the second in 2027.

    Page 35


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Calculation Method:

    - First moment => simple mean of each of these limits for distribution companies according to their regional groups, which are:

    - Second moment => mean of the privatized of the regional groups mentioned above.

    It is important to note that, as Technical Note 149/2017-ANEEL defines the operational costs in force from 2017 until 2022, the efficient cost intervals defined in this section will only impact the tariff from the Tariff Review of 2023.

    - Effective PMSO vs. Regulatory PMSO (Service A):

    Based on the values of the effective PMSO projection allocated above and with the values calculated for the regulatory PMSO, the following graph was obtained:

    Page 36


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Graph 12. Effective PMSO vs. Regulatory PMSO Projection


    Important points:

    - It is important to point out that for the composition of the effective PMSO graph, the expenses with the PDV were considered in the years 2018 and 2019, therefore, the values for the P item contained in above table are different from the values used to compose the graph. For information the values are: R$ 260 million against R$ 251.1 million (2018) and R$ 275 million against R$ 264 million (2019);

    - By means of the graph it can be observed that the projections of the regulatory PMSO are greater between 2018 and 2023, then they are smaller in the period from 2024 to 2033 and ends up being greater until the end of the projection.

    - Another point worth highlighting is that the values of Item Other for the years 2025, 2030, 2035, 2040, 2045, 2046 and 2047 contained in the table on page 175 of the assessment report, which were used to compose the graph above, of the report, are in disagreement with the values indicated in the projection, pages 78 and 79 of the document.

    Page 37


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    - The difference between the VPL of the two items, in real terms, discounted by the WACC rate of the model, considering, in a simplified form, a tax shield of 34%, results in a negative value of R$ 39.1 million, that is, for AmE D, the PMSO destroys the value for the economic-financial assessment;

    - Another important point is that the projection of the regulatory PMSO values took into account the technical note 149/2017 - ANEEL.

    2.5.3      PMSO FOR SERVICE B:
    -      Historical part:

    Service B did not present the rational to compose its database from 2012 to 2016 as did Service A. The report shows the projections of P, M, S and O values directly for the period from 2017 to 2047.

  • Methodology used in Service B (common to all Distribution companies):
     
  • Database: Calculation of Efficiency for the 4CRTP - 4th Cycle of Periodic Tariff Reviews (PMSO, Market, Consumers, network, losses, etc. data) => 2001-2013;
     
  • Clustering of Companies: Automatic grouping of similarity second degree data. The result of this process can be seen in the table below:
    Table 23.  Clusters composition and efficiencies - AmE D   
     
    Companies Efficiency
    Indicator of
    4CRTP
    efficiency 
    Efficiency
    Indicator 2011
    Efficiency
    Indicator
    2013 
    %
    (p.a.)
    RGE  100%  0.58  0.66  -6.4% 
    Celtins  100%  0.44  0.51  -8.5% 
    AES Sul  83%  0.68  0.64  2.9% 
    EPB  82%  0.73  0.69  3.1% 
    Light  78%  0.86  0.87  -0.9% 
    Caiuá  74%  0.86  0.78  4.8% 
    National  68%  0.99  0.80  10.0% 
    CEE  42%  1,21  1,67  -17.4% 
    Ame  31%  2.27  1,72  13.1% 

     

    Page 38


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Understanding the indicator: For example, AmE D in 2011 obtained an indicator of 2.27 and in 2013 this valor was 1.72. This means that the company has improved from 2011 to 2013 as it needs 1.72 (2013) PMSO units to offer a set of services (weighted combination of Market, Network, Consumer, Non-technical Losses).

    Projection of the trajectory of the indicator:

    1.      Phase 1: 2017-2018 => No Efficiency gain
    2.      Phase 2: 2019-2020 => It follows the rhythm of the utility with the greatest gain of efficiency of the Cluster until reaching the median of the indicators of the group
    3.      Phase 3: 2021-2048 => Convergence for the mean of the Companies of the 1st Portion of the Group

    Thus, the AmE D efficiency curve as well as the projection for the regulatory PMSO are presented in the graph and table below:

    Graph 13.

    AmE D efficiency curve


    From the efficiency curve and projected values for the products under consideration, it is possible to determine the estimated value of regulatory PMSO for each year. With this, the values in the table below are obtained:

    Page 39


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 24.  Projection of regulatory PMSO AmE D 
    Period  Regulatory PMSO (R$ million 2016) Var. % p.a. 
    2016  332   
    2017  350  5.2% 
    2018  362  3.5% 
    2019  334  -7.7% 
    2020  308  -7.7% 
    2021  289  -6.3% 
    2022  272  -5.9% 
    Average 2023 - 2027  256  -0.1% 
    Average 2028 - 2032  289  2.1% 
    Average 2033 - 2037  309  1,0% 
    Average 2038 - 2042  325  1,0% 
    Average 2043 - 2047  341  0.9% 

     

    Although Service B shows an Efficiency curve that should be followed in the document, it is not evident the comparison with the effective PMSO values projected by the consultancy. Therefore, it is not possible to state when the company is able to reach the level of regulatory PMSO, as well as to estimate if this trajectory will generate or destroy value for the Company.

    In addition, it is important to take into account the issue of relaxation, since according to Technical Note No. 149/2017 - ANEEL, which deals with the flexibility of some regulatory parameters, the new values for the Regulatory PMSO that go until the first ordinary tariff review of the new controller of the designated companies, are:

    Table 25.  Flexibility of the PMSO (Aneel Technical Note No. 149/2017) - R$ 
    million

     

    Distribution
    Company
    Tariff Process
    2016
    Flexible
    Regulatory
    Level 
    Difference Difference
    (%)
    AME  348.8  627.2  278.4  80% 
    Boa Vista  37.5  69.1  31.5  84% 
    CEAL  271.7  336.7  65.0  24% 
    CEPISA  360.7  420.1  59.4  16% 
    CERON  234.0  279.9  45.8  20% 
    ELETROACRE  93.5  128.5  35.1  38% 
    CEA  67.8  124.8  56.9  84% 

     

    Page 40


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    2.5.4      FORECASTS AND COMPARISONS OF THE PMSO OF SERVICES A AND B
    -      AmE D projections until 2047:
    Table 26.  PMSO Projection Data (Service A) - R$ Million 
     
    Item  2017  2018  2019  2020  2021  2022  2023  2024 
    Personnel  248.4  251.1  264.0  276.7  288.8  301.0  301.8  302.9 
    Material  8.8  10.2  10.5  10.7  10.8  10.9  11.0  11.2 
    Third Party Service  221.1  227.4  234.2  240.8  247.4  253.9  269.6  286.3 
    Other  71.8  73.2  73.5  73.7  73.7  73.9  74.1  74.4 
    Total PMSO  550.1  561.8  582.1  601.9  620.6  639.7  656.5  674.7 
    Item  2025  2026  2027  2028  2033  2038  2043  2047 
    Personnel  304.4  314.9  325.6  330.1  373.0  452.3  548.6  640.7 
    Material  11.3  11.4  11.6  10.2  6.9  9.1  11.9  14.7 
    Third Party Service  304.0  323.1  343.3  362.6  487.2  665.7  897.3  1,128.4 
    Other  74.8  75.5  76.1  72.7  66.5  78.8  95.6  112.6 
    Total PMSO  694.5  724.9  756.6  775.6  933.6  1,205.9  1,553.3  1,896.4 

     

    In addition, Service A estimated a value for the Voluntary Resignation Plan (PDV), defining the target public (employees with at least 20 years of company), the success rate, its cost (R$ 18.6 million) and the date of resignation of employees, June/18 and June/19.

    Table 27.  PMSO Projection Data (Service B) - R$ Million   
    Item  2017  2018  2019  2020  2021  2022  2023  2024 
    Personnel  206.6  223.7  196.0  164.3  145.8  134.1  126.7  121.4 
    Material  17.7  19.1  16.8  14.1  12.5  11.5  10.8  10.4 
    Third Party Service  142.7  154.5  135.4  113.4  100.7  92.6  87.5  83.8 
    Other  233.3  252.6  221.4  185.5  164.6  151.5  143.1  137.1 
    Total PMSO  600.2  650.0  569.5  477.3  423.6  389.7  368.2  352.7 
    Item  2025  2026  2027  2028  2033  2038  2043  2047 
    Personnel  129.6  140.3  151.0  161.3  215.7  278.0  357.7  436.2 
    Material  11.1  12.0  12.9  13.8  18.5  23.8  30.6  37.3 
    Third Party Service  89.5  96.9  104.3  111.4  149.0  192.0  247.1  301.3 
    Other  146.4  158.5  170.5  182.2  243.6  313.9  404.0  492.6 
    Total PMSO  376.7  407.7  438.6  468.8  626.8  807.7  1,039.4  1,267.4 

     

    Page 41


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The graph below shows the evolution of Service A PMSO compared to Service B.

    Graph 14. Projection of the PMSO for Service A and Service B

    From the above graph analysis, one can see how far a projection is from one another. Service A reaches a final PMSO of R$ 1,896.4 million against a value of R$ 1,267.4 million for Service B. This difference results in a gap of R$ 629 million at the end of the projection (2047). It is also noticed that the projection of the PMSO for service B starts higher and after 2019, starts to fall and so it goes on with smaller values of projection until the end of the period. However, it is important to remember that this difference in values has no significant impact on valuation if the company's PMSO is within the regulatory level.

    Therefore, in order to further explore the reasons that led to this discrepancy in the projection, we will show how the evolution of each item behaves until the end of 2047.

    The items will be shown in pairs and in sequence, some comments on the main differences will be made.

    Page 42


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Personnel and Material

    Graph 15. Projection of Personnel and Material items for Services A and

    From the observation of the projection graphs for the Personnel and Materials items, it can be verified for item P that the projection of Service A follows a linear trend from the first

    Page 43


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    years, reaching a value of R$ 640.7 million end of the period. On the other hand, the projection of item for Service B shows a downward trend from 2018 up to its minimum projection value, R$ 121.4 million in the year 2024 and thereafter shows a linear behavior up to reach the amount of R$ 436.2 million in 2047. In relation to the Materials item, both Services have shown an irregular projection over the years. The final value for each Service is R$ 14.7 million (Service A) and R$ 37.3 million (Service B).

    Third Party Service and Other

    Graph 16. Projection of Third Parties and Other Services items referring to Services A and B

     

    Page 44


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The Third Party Service item shows a very similar behavior to that of Personnel item. For Service A, the values grow respecting almost an exponential regime reaching the final value in 2047 of R$ 1,128.4 million. For Service B, it declines from 2018 to 2024, where it reaches its minimum projection value, R$ 83.8 million, and then increases linearly until reaching a value of R$ 301.3 million, resulting in a gap of R$ 827.2 million. Finally, for Other item, as was the case with other Companies, the projection of Service B is higher throughout the period and has the same behavior as in item P. At the end of the period, the value of the discrepancy is R$ - 380 million (comparing Service A with B).

    found, it is noticed that there is a very marked difference in relation to the projections. This fact may be due to the basis that each Consultancy used to make the PMSO projection and also according to the projection model considered. In relation to the first item mentioned above, only Service A presented the rationale for the initial composition of the PMSO.

    expenses, it is concluded that Service B adopted a much more projection than that of Service A. Highlighting again, this difference in values has no significant impact on Valuation if the company's PMSO is within the regulatory level.

    Table 28. Comparison between Services - R$ Million

      Service A  % Service B   
    Company          % 
      2017  2047  2017  2047   
    Eletroacre  112.5  684.9  509%  133.5  327.6  145% 
    CEAL  288.8  1,126.5  290%  296.7  686.0  131% 
    AmE D  550.1  1,896.4  245%  600.2  1,267.4  111% 
    Cepisa  383.7  1,757.7  358%  452.6  1,122.8  148% 
    CERON  340.1  1,534.8  351%  373.3  866.6  132% 
    Boa Vista  105.6  334.1  216%  216.3  371.7  72% 
    Total  1,780.8  7,334.4  312%  2,072.5  4,642.1  124% 

     

    Page 45


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    - Finally, regarding the documents presented by the consultancies, Service A detailed the methodology much more than Service B, presenting in its text more coherence in the presentation of the information.

    2.6 INVESTMENT

    Both services contracted for economic and financial assessment adopted different assumptions for the projection of the investment notably from the period 2023 to 2047, since both adopted the Five-Year Plan for the period from 2018 to 2022 of the company. This plan was presented in Product 7 “AmE D Operational Technical Assessment Report" and was prepared based on the "Expansion Plan for the Electrical System (2017-2026 period)", the "Development Plan for Distribution - PDD for the year 2017", the "Results Plan for the Improvement of Distribution Services, 2015" and the "Temporary Provision Plan for the Electricity Distribution Service, 2017". In addition, information obtained through the "Field Visit Report" (Annex I of the Technical and Operational Assessment Report) and the incremental investment needs projected by the company's planning area, obtained during face-to-face meetings and conference call carried out. The only caveat made for this period is that Service A for the investment in “Luz para Todos” program considered only 10% of the value since 90% is subsidized by CDE, and therefore treated as Special Obligations. In addition, approximately R$ 57.4 million in 2019 and R$ 172.4 million in 2020 of the investments in expansion AT were also considered as Special Obligations, since they are of origin of subrogation.

    In relation to the other five-year period, the main differences are due to the fact that Service A included, in addition to the projections made for implementation and renovation projected from the Five Year Plan, other analyzes to define the investment in maintenance and replacement of partially and totally depreciated assets . Whilst Service B was restricted only to the first analysis.

    For the projection from 2023 of the investments in implantation and renovation, Service A adopted as basic premise for the expansion of AT, AV and LV, as well as for the improvements some metrics1 depending of the type of investment depending on the last

    1 The methodology for projecting the investment from 2023 varies between companies for Service A. The expansion in AT had the information until 2026 through the Development Plan for Distribution 2017 (PDD 2017) and, then, it was adopted for the other years between the years 2022 and 2026. As for the investment in AV and

    Page 46


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    years contemplated in the plans of the companies and that value was kept constant until the end of the concession period. Service B, for electric investments (expansion and improvement), estimated the second five-year period of the concession through the mean of the period from 2018 to 2022 and expurgated values outside the range between the average and +/- a standard deviation. Thus, in order to obtain the amount of investment required for the five-year period, the investments of the years considered typical were added up and multiplied by the factor 5/No. typical years. In the case of AmE D, since there was information up to 2026 for the expansion of HV and AV, these values were used and the year 2027 was estimated from the average investments of the last 4 years. For the other five-year period, they adopted as base the previous five-year period with the caveat that for the expansion in AT, it is considered that every 10 years the need for investments doubles in relation to the previous five-year period. For non-electric investments (infrastructure), Service B did not project them since the value projected for 2018 would meet the needs of improvement of distribution management. Service A, however, did not project them for any distribution company.

    For renovation investments (maintenance), since 2023, service A considered that investments of this type are equivalent to the value of the assets depreciation. To define this value, the depreciation reference of ANEEL's Normative Resolution 674/20152, was used for the current asset base plus the depreciation of the new investments of the company, calculated from the average depreciation of the asset base of the Assets Control Report. These investments were accumulated and concentrated in the two years before each tariff review according to the proportion of the reference companies3. Service B, however, first calculated the share of the amount destined to the investment in renovation within the period of the Five-Year Plan 2018-2022. Subsequently, it applied this percentage for all other years.

    For investments in replacement, only service A estimated them. Based on the assets base of 2016, the estimated investments for implementation and renovation from 2017 to 2022 were added, as well as the inflation adjustments and application of the depreciation of each item, based on Aneel Normative Resolution 674/2015, both for the assets base of

    LV expansion, the mean of expenses from 2018 to 2021 was used. On the other hand, the investment in improvement was defined as the minimum between the years of 2018 and 2022.

    2 Resolution that approved the review of the Assets control manual of the Electric Sector - MCPSE.

    3 Cemar, Coelce, Coelba and Celpe.

    Page 47


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2016 and for the new investments projected up to 2022. In this way, the company's asset base of 2022 was formed. Therefore, the indicator Value of Market in Use (VMU) / New Value of Replacement (VNR) wars calculated for the company and for the benchmark4 companies so that the company distribution in analysis reaches the same percentage of the private reference distributed between 2023 and 2027. As the new AmE D base exceeded the reference value, it was not necessary to estimate investments for asset replacement in 2022.

    Given the assumptions and their differences, we find the following values for the two consultancies for AmE D.

    Table 29.  Investment comparison - Services A and B   
    Service A              R$ Million 
      1st Five-Year  2nd Five-Year  3rd Five-Year  4th Five-Year  5th Five-Year  6th Five-Year   
                  Sum for the 
    Investment  Period  Period  Period  Period  Period  Period   
                  Period 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
    AT Extension  1,183  237  433  433  433  433  3,152 
    AV/LV Extension  683  683  683  683  683  683  4,096 
    Improvement  468  200  200  200  200  200  1,468 
    Renewal (maintenance)  34  1,054  1,296  1,583  1,912  868  6,747 
    Luz para Todos (Light for Everyone)  73  -  -  -  -  -  73 
    Infrastructure and Support  1  -  -  -  -  -  1 
    Replacement  -  -  -  -  -  -  - 
    Total  2,442  2,173  2,612  2,899  3,228  2,184  15,536 
    Service B              R$ Million 
      1st Five-Year  2nd Five-Year  3rd Five-Year  4th Five-Year  5th Five-Year  6th Five-Year   
                  Sum for the 
    Investment  Period  Period  Period  Period  Period  Period   
                  Period 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
    AT Extension  1,429  170  339  170  339  170  2,617 
    AV/LV Extension  691  346  346  346  346  346  2,418 
    Improvement  474  248  248  248  248  248  1,713 
    Renewal (maintenance)  34  8  10  8  10  8  77 
    Luz para Todos (Light for Everyone)  735  -  -  -  -  -  735 
    Infrastructure and Support  1  -  -  -  -  -  1 
    Replacement  -  -  -  -  -  -  - 
    Total  3,364  771  942  771  942  771  7,561 

     

    The difference between the consortiums is R$ 8 billion, notably in the item Renovation (maintenance) in which Service A estimates that from the 2nd five-year period the investments are equivalent to the value of the depreciation of the assets and Service B calculates them from their participation in the Five-Year Plan from 2018 to 2022. It should be noted that this difference in the volume of projected investments does not have a significant impact on valuation, since the investment increases the Company's Remuneration Base, and consequently, it is reflected in a higher tariff.

    4 Cemar, Energisa Sergipe and Energisa Paraíba

    Page 48


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    By way of comparison, the value of the investment made from 2012 to 2016 to the values of December 2016 by AmE D is as follows:

      Table 30.  Investment made from 2012 to 2016     
                R$ Million 
    Investment  2012  2013  2014  2015  2016  Five-Year Period 
                2012-2016 
    Updated Values (Dec./16)  1,005  1,322  500  328  330  3,484 

     

    2.7 REGULATORY REMUNERATION BASE

    Both consortiums used the data contained in the valuation reports of the assets as reference to the date of February 28, 2017 to prepare the projection of the Regulatory Remuneration Base (BRR). In these reports, both closed BRR and the incremental BRR from 3CRTP to 4CRTP were assessed. In the case of AmE D, the following reports were used:

    Both service A and B adopted different projection criteria for the asset base unitized through Feb/17 (closed BRR - starting point for 4CRTP) and for unitizations occurring after Feb/17 and new investments (incremental BRR).

    Regarding the starting point for 4CRTP, both consortia use the full assessment report (delivered on 06/30/2017), however, they present different criteria for the disallowance of Fixed Assets in Use (AIS).

    For service A, the disallowance is treated in two ways: i) Disallowance over the reviewed closed base (Closed Base of Full Report - Closed Base of the Incremental Report) in the value defined in 5,73% for the companies of the North and 5.01% for the Northeast, based on the average adjustment (disallowance) of the distribution group assessed in the 3CRTP by region, i.e., this disallowance is made only in the AIS of the difference found in

    Page 49


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    the opening of the closed base compared to that approved by Aneel in 3CRTP monetarily corrected; and ii) The incremental basis between 3CRTP and 4CRTP - the lowest value was applied between the average adjustment of the distribution companies group assessed in 3CRTP by region and that verified by the company, without showing the amount applied in the Assessment Report.

    Unlike Service A, Service B used only a disallowance methodology to reach the Remuneration Base of Feb/17 of the Distribution company. The disallowances were performed between the difference in AIS of the two assessment reports of the Closed Base and the Incremental Base. This difference would be the opening of the Closed Base. That is, the disallowance for Service B was performed only at the opening of the Closed Base and occurred in the AIS. The percentage of disallowance was calculated from the relation between the final AIS (with disallowances) and the initial AIS (without disallowances) of the incremental BRR of the distribution companies 3RTP of Eletrobras group, obtaining a mean of 9.2%. If the company has a disallowance lower than this mean, it is considered the mean of disallowance percentage itself (9.2%). Otherwise, the mean between the percentage of the company and the maximum percentage of disallowances (20.1%) is used. In the case of AmE D, it falls within the second criterion, since it presented a 20.1% disallowance of the AIS in the 3CRTP, then, adopting a disallowance of 20.1 for the Closed Base. This disallowance percentage is applied only in the difference between the AIS of the two reports and added to the value of the base of the incremental report. That is, the disallowance of 20.1% is not applied in the total AIS of the Complete Base, it is calculated only for the increment of the closed base.

    Page 50


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Therefore, the following information was obtained for the Closed base, in the position of February 2017, for both services:

    Table 31.  Closed base, in the position of February 2017 
     
          R$ Million 
        Armored Base   
      Description     
        Service A  Service B 
    (1) Property, Plant And Equipment in Service (New Replacement Value)  3,139  2,855 
    (2) Index of Full Utilization  3  3 
    (3) Gross Special Obligations  1,043  1,043 
    (4) Fully Depreciated Assets  385  385 
    (5) Gross Compensation Reference = (1) - (2) - (3) - (4)  1,708  1,424 
    (6) Accumulated Depreciation  1,332  1,332 
    (7) Net AIS (Market Value In Use)  1,807  1,523 
    (8) Index of Depreciated Utilization  2  2 
    (9) Compensation Reference Value (VBR)  1,804  1,521 
    (10) Warehouse in Operation  37  37 
    (11) Deferred Assets    -  - 
    (12) Special Net Obligations  945  945 
    (13) Lands and Easements  40  40 
     
    (14) Base of Total Net Compensation = (1) - (6) - (8) + (10) + (11) - (12) +  936  652 
    (13)       
    (15) RGR PLPT Balance    -  - 
    (16) RGR Balance Other Investments  217  188 
    (17) Depreciation Rate    4%  4% 
    (18) Share of Regulatory Reintegration = (5) * (17)  63  51 
    (19) Actual WACC before Taxes  11%  ND 
    (20) RGR PLPT Rate    1%  ND 
    (21) RGR Rate of Other Investments  4%  ND 
    (22) Capital Compensation (RC) = (15) * (20) + (16) * (21) + [(14) - (15) -     
    (16)] * (19)    89  ND 

     

    For the years subsequent to the 4CRTP Starting Base, both Services apply adjustment for inflations for the IPCA, additions and write-offs of the assets that were considered in the valuation until the end of the concession period.

    However, it should be noted that both Services did not incorporate the positions of Fixed Assets in use into the Remuneration Base, since, according to the services, the assessors’ report of the Remuneration Base does not take into account the specific assessment of the condition and characteristic of these assets concerning (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the reassessment process (such as the calculation of VNR, depreciation, utilization rate and assessment of write-offs). In regards to the case of AmE D, it should be noted that the Full Assessment Report

    “Eletrobras Amazonas Energia – Project no. 3174-17752,” performed by Levin, includes as the position of February 2017 the balance of R$ 1,409.9 million for the Fixed Assets in Use.

    Page 51


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    In addition, Service B segments the assumptions for the projection of BRR transactions into two groups. The first group operated the closed BRR by the IPCA adjustment for inflation and the discount of the depreciation and write-off of the assets that went into operation until February 17 and in a second group operates the incremental BRR (new investments) through the additions of assets projected after Feb/2017, less the losses related to these assets.

    Thus, from 2018, service A monthly corrects, through the IPCA, the items Fixed Assets in Use, Warehousing in operation and land and easements and increases them according to their projection of the investment. The gross special obligations item is also updated by the IPCA and by the increase in projected special obligations. On this incremental Base, the average adjustment of privatized benchmark distribution companies was applied as disallowance by region obtained through the 3CRTP Incremental Base. For the Northeast companies, Celpe, Energisa Paraíba and Energisa Sergipe were considered, giving a value disallowance of 2.09%. For the North companies, the disallowance value was 2.83%, using Celpa and Cemar as benchmark. In regards to the depreciation rate, the same verified in 3CRTP was maintained for each distribution company assessed. In relation to asset write-offs, the ratio of fully depreciated assets to fixed assets in use observed in the 4RTP BRR resulted in an equilibrium rate calculated as a simple mean between Eletrobras' companies. Thus, starting from the company's ratio, it varies at a constant monthly rate until reaching that equilibrium rate of 12.54% at the end of the concession.

    Service B, as mentioned earlier, operated BRR with different premises for the base that served as a starting point and the one that incorporates the new investments. The transaction of the closed BRR was made by monetary restatement by the IPCA and by the discount of depreciation and write-offs. The depreciation rate, unlike service A, is not constant, the rate falls from the average level calculated in Feb/17, as the assets are being written-off. In relation to asset write-offs, the average write-off/AIS ratio was applied to the AIS of each month based on the information of the Assessment Report prepared by Levin (delivered on 06/19/2017) in the period between 3CRTP and Feb/17. The incremental BRR transaction (new investments) was made based on the projection of the investments rejected in 5%. The depreciation rate used was obtained by the weighted mean of the investments of the Five-Year Plan 2018 to 2022 per group of assets and their respective depreciation rates, obtaining 3.59% of the AIS. This rate was considered constant throughout the concession. Regarding the write-offs, until Feb/27, they were considered null in view of the reduced probability of assets being written off in the first 10

    Page 52


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    years. After this period, 1/3 of the BRR Closed/AIS write-off ratio was applied in a staggered manner every 5 year until reaching the value of the integral ratio as of Mar/42.

    Accordingly, given the differences in the investment projections, the criteria of reversal, depreciation and write-offs, we obtained the following Net Remuneration Base for AmE D at the end of the concession (Feb./2048) according to the projections of the two services:

      Table 32.  Net Remuneration Basis for AmE D - Services A and B 
            R$ Million 
      Description    Service A  Service B 
      Net Remuneration Base (Feb./48)  38,047  5,688 
     
     
    2.8  X FACTOR       

     

    The primary objective of the X-Factor is to ensure that the balance established in the tariff review between efficient revenues and expenses is maintained in subsequent tariff repositioning. This occurs through the transfer to the consumer of potential productivity gains from the distribution segment.

    The approach adopted by Aneel adds to the X Factor a mechanism to encourage the improvement of the technical and commercial quality of services provided to the consumer. It also introduces a component that seeks to establish the gradual transition of the operating costs of the utility towards its efficient costs.

    The X Factor adjusts the value of Installment B, thus affecting the tariff and, consequently, the revenue and the collection of the Distribution company. Therefore, the X factor value inversely impacts the value of the VPL (the higher the X factor, the lower the revenue and the VPL).

    The X Factor is composed of the sum of three components: 

    Fator X = Pd + Q + T

     

    Page 53


     

     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

         account the variation of seven indicators and the compliance with the quality standards established by ANEEL;· Component T: adjusts, over a defined period, the observed operating costs of each utility to the efficient operational cost.

    The table below shows X factor and its components calculated by Services A and B in years of tariff reviews5:

    Table 33.  X Factor and its components calculated by Services A and B 
      Service A         
        X Factor  Pd  T  Q 
      2019  0.09%  0.00%  0.00%  0.09% 
      2023  0.68%  0.97%  -0.29%  0.00% 
      2028  -0.71%  0.96%  -1.59%  -0.08% 
      2033  1.16%  0.97%  0.19%  0.00% 
      2038  1.18%  0.97%  0.21%  0.00% 
      2043  0.98%  0.97%  0.01%  0.00% 
      2048*  0.98%  0.98%  0.00%  0.00% 
     
     
      Service B         
        X Factor  Pd  T  Q 
      2019  -0.11%  0.00%  0.00%  -0.11% 
      2023  -0.51%  1.44%  -1.26%  -0.69% 
      2028  0.56%  1.32%  -0.48%  -0.28% 
      2033  0.05%  1.31%  -0.92%  -0.34% 
      2038  -0.06%  1.28%  -0.92%  -0.42% 
      2043  0.06%  1.25%  -0.72%  -0.47% 
      2047*  0.05%  1.23%  -0.72%  -0.46% 

     

    As can be seen in the table above, the estimated values for the X Factor are quite divergent, especially for the Q and T components, affecting, as a whole, the X-Factor as a whole.

    5 It should be noted that Service B provided the value of the component Q for all thirty years of concession, while Service A presented the values only in the years of tariff reviews.

    Page 54


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Table 34.  Factor X - Services A and B comparison 
     
    Service A - Service B       
      X Factor  Pd  T  Q 
    2019  0.20%  0.00%  0.09%  0.11% 
    2023  1.19%  -0.47%  1.26%  0.40% 
    2028  -1.27%  -0.36%  0.40%  -1.31% 
    2033  1.11%  -0.34%  0.92%  0.53% 
    2038  1.24%  -0.31%  0.92%  0.63% 
    2043  0.92%  -0.28%  0.72%  0.48% 

     

    1 - Component Pd - Productivity gains from distribution activity

    The formula of Component Pd:


    PTF: average productivity of the distribution segment, 1.53% p.a.

    MWh(i): Average annual variation of the utility market, between the previous tariff review and the tariff review being processed;

    Average MWh: Average annual variation of company distributions' market, of 4.65% p.a.;

    The main highlights on accounting differences are:

    · Both considered Component Pd, during the years of 2018 and 2022, equal to zero, following the proposal of the Aneel´s Official Letter No. 113/2017.

    · On the evolution of the PTF, average ”MWh and average ”UC, Service A did not project variation in these indicators. As Service B added that it considered appropriate to maintain throughout the period under review the format and parameters of the approved formula for 4CRTP due to the absence of the updated database for the estimation of total costs and market of all the Distribution utilities.

    2 - Component Q - Technical and commercial quality of the service provided to the consumer

    · The component Q of the X-Factor consists of a technical quality installment (DEC and FEC) and a commercial performance installment comprising five indicators,

    Page 55


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    FER, IASC, INS, IAB and ICO, with the technical part having 70% weighting in the component Q and the commercial performance part having 30%.

    · It should be noted that as of 2019, only the DEC index will correspond alone to 50% of the component Q of the X Factor;· Service B used as a premise for all the years of the new concession the formula below, which will only come into effect as of Apr/19.

    Component X Factor:


    DEC - Equivalent Duration of Interruption per Consumption Unit FEC - Equivalent Frequency of Interruption per Consumption Unit FER - Equivalent frequency of complaint IASC - Aneel Consumer Satisfaction Index INS - Service Level Index IAB - Abandonment Index ICO - Busy Call Index

    Service A considered the proportions of the table below, following the provisions of PRORET. For the years after 2020, the ratio defined between Apr/19 to March/20 was used.

    Page 56


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 35.  Methodology used - Service A   
     
      3rd CRTP    New Methodology   
      Methodology         
    Indicator  Apr/15 to  Apr/16 to  Apr/17 to  Apr/18 to  Apr/19 to 
      Mar/16  Mar/17  Mar/18  Mar/19  Mar/20 
    DEC  50%  30%  37.5%  45%  50% 
    FEC  50%  30%  30.0%  27%  20% 
    INS      0.8%  2%  4% 
    ICO      0.4%  1%  3% 
    IAb      0.4%  1%  3% 
    FER      3.0%  7%  10% 
    IASC      3.0%  7%  10% 
    Total  100%  60%  75%  90%  100% 

     

    3 - Component T - Trajectory of operating costs

    Component T of the X-Factor is applied to Installment B so that the tariff increases or decreases so that the recognized regulatory operating costs reach the efficiency threshold defined by Aneel.

    In defining efficient operating costs, the distribution company’s costs, the efficient level of costs and the characteristics of the concession area are observed. Aneel uses a model whose objective is to define a feasible production/cost frontier as a way of identifying the level of efficiency of companies.

    In a simplified way, the identification of the efficient level of costs is obtained by comparing the distribution companies by means of a benchmarking method that takes into account the attributes of each utility. From these conditions, a target of efficient operational costs is established to be reached throughout the tariff cycle. At the time of the tariff review, the regulatory goal is compared to the operational costs coverage present in the utility´s tariff, called operating costs revenue. From the difference between the regulatory goal and the operating costs revenue, a regulatory trajectory will be calculated. Part of the difference will be incorporated at the time of the tariff review and the remaining installment will be considered for the purpose of calculating component T of the X Factor.

    Services A and B, following the provisions of Aneel's Technical Note No. 149/2017, used a zero value for component T in the period between the date of signature of the agreement and the first ordinary tariff review, that is, from 2018 to 2022. From 2023, Services A and B found very divergent values for Component T of X Factor.

    Page 57


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.9 WACC

    Services A and B adopted the method of Capital Asset Pricing Model - CAPM to estimate the return that the shareholders expect to arrange with the capital employed in the business, for discounting the current value of the distribution company’s cash flow. It should be noted that this method is well-established and widely used, either by the market at large, and by Eletrobras.

    In the calculation of the Weighted Average Cost of Capital WACC, Services A and B applied different approaches of the CAPM method, thus allowing quite different results.

    Below are the main differences found in the calculation of WACC by both Services.

    Capital Structure

    Beta

    · Service A: used comparable domestic companies. Unlevered Beta: 0.68;

    · Service B: considered comparable domestic companies, with a larger sample. Unlevered Beta: 0.43.

     

     

     
     

    Ke

    · Service A: applied the CAPM method and a “foreign exchange risk” of 4.17% at nominal Ke in Dollars to achieve a nominal Ke in Brazilian Reais. Therefore, this “foreign exchange risk,” which derives from the “Future DI” rate and “Fraction of exchange rate coupon,” works as a differential of inflation. To achieve such rate in real terms, it was deflated by IPCA (4.23%);

     

    · Service B: applied the CAPM method.

     

     

    Page 58


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    The table below shows the result for Ke under real terms:         
     
      Table 36.  Profit or Loss of Real Ke Services A and B     
    Actual Ke  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  7.95%  7.81%  7,53%  7.21%  7,02%  6.85%  6.70%  6.55%  6.43%  6.31%  6.31% 
    Service B  11.73%  10.16%  10.18%  10.42%  10.42%  10.42%  10.42%  10.42%  10.84%  10.84%  10.79% 

     

    Kd

    The table below shows the Kd result in real terms after taxes.         
     
      Table 37.  Kd in real terms after taxes Services A and B     
    Actual                       
      2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Kd                       
    Service A  4.25%  4.09%  4,27%  4.19%  4.19%  4.19%  4.19%  4.19%  3.25%  3.25%  3.25% 
    Service B  4.44%  3,34%  3.36%  3.29%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25% 
    WACC                       

     

    Based on the values of Beta, Ke and Kd above, Services A and B calculated WACC in real terms after taxes, according to the results included in the table below.

          Table 38.  WACC Services A and B       
    WACC  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  8.09%  7.20%  7.30%  7.39%  7.39%  7.39%  7.39%  7.39%  7.60%  7.60%  7.58% 
    Service B  5.29%  4.80%  4.89%  4.87%  4.93%  5.00%  5.06%  5.12%  5.19%  5.26%  5.26% 

     

    Out of the results stated above, it should be noted that the factor with a higher impact on the rate presented by Service A was the use of the foreign exchange risk (4.17%) in a heterogeneous manner to the market practices, as a differential of inflation, as this is applied on a Ke (nominal in USD), in a scenario where the CPI US inflation is 2.30% and IPCA, 4.23%. The result of this is WACC, in real terms, which is higher compared to the number obtained in calculation by Service B.

    Page 59


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    3. THE RESULT OF ECONOMIC AND FINANCIAL ASSESSMENTS AND

    BY MULTIPLES

    Service A's assessment calculated the enterprise value from R$ 2,687,563,915.43 to AmE D, considering, as already mentioned, the granting of a new concession. After deducting the debts, liabilities with suppliers and contingencies, AmE D´s equity value calculated by Service A is negative R$ 8,431,847,736.90, as shown below. To obtain these values, the consultancy used the WACC of 7.57%.

    Table 39.  Enterprise value Service A   
        Free Cash Flow to Firm (FCFF) 
    Base date of Enterprise Value    Feb./2018  Dec./2016 
    Adjustment on Enterprise Value    -  IPCA 
    Enterprise Value (BRL)  A  2,828,419,335.80  2,687,563,915.43 
    Cash (BRL)  B  99,391,000.00  99,391,000.00 
    Gross Debt (BRL)  C  (10,058,418,395.23)  (10,058,418,395.23) 
    Net Debt (BRL)  B+C  (9,959,027,395.23)  (9,959,027,395.23) 
    Tax, Labor and Social Security Contingencies (BRL)  D  (603,104,000.00)  (603,104,000.00) 
    Legal Contingencies (BRL)  AND  (373,230,168.53)  (373,230,168.53) 
    Environmental Adjustments (BRL)  F  445,310.00  445,310.00 
    Environmental Adequations (R$)  G  (184,495,398.57)  (184,495,398.57) 
    Valuation (R$)  A+B+C+D+E+F+G  (8,290,992,316.53)  (8,431,847,736.90) 
    Source: Assessment Report of Eletrobras Distribuição AmE, Service A.   

     

    Service B calculated the enterprise value of R$ 2,223,734,395.42 for AmE D, in view of the granting of a new concession. After deducting the debts, liabilities with suppliers and

    contingencies, AmE D equity value calculated 8,895,677,256.91, as shown below.

    by Service

    B is negative R$

    Table 40.  Enterprise value - Service B 
    Result  Dec./2016 
    Operation NPV  1,876,305,503.24 
    NPV of Terminal Values  347,428,892.17 
    Enterprise Value  2,223,734,395.42 
    (-) Net Revenue  (9,959,027,395.23) 
    (-) Contingencies  (975,888,858.53) 
    (-) Environmental Adjustments  (184,495,398.57) 
    Equity Value  (8,895,677,256.91) 

     

    Source: Product 10: AmE D Conclusive Economic and Financial Assessment Report, Service B.

    In order to obtain these values, Service B used rolling WACC, which considers a movable (leveraged) capital structure, as follows.

    Page 60


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 41.  Rolling WACC - Mais Energia B Consortium     
     
    Parameters  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  > 2026 
    Rate Free from Risk (Rf)  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78% 
    Country Risk  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53% 
    Market Premium  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90% 
    Non-Leveraged Beta (²u)  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43 
    % Debt (Kd) - [D/(D+E)]  54.0%  50.7%  47.3%  44.0%  40.6%  37.3%  33.9%  30.6%  27.3%  23.9%  23.9% 
    % Own Capital (Ke) - [E/(D+E)]  46.0%  49.3%  52.7%  56.0%  59.4%  62.7%  66.1%  69.4%  72.7%  76.1%  76.1% 
    Taxes - IRPJ / CSLL  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0% 
    Leveraged Beta (²L)  0.76  0.72  0.68  0.65  0.62  0.59  0.57  0.55  0.53  0.52  0.52 
    Premium Size  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00% 
    Inflation - Brazil  4.33%  4.43%  4.36%  4.31%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23% 
    Inflation - USA  2.40%  2.30%  2.30%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40% 
    Cost of Own Capital  12.62% 12.55% 12.22% 11.83% 11.55% 11.37% 11.21% 11.06% 10.93% 10.81%10.81% 
    Selic  10.71%  9.07%  8.87%  8.87%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73% 
    Spread  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64% 
    Cost of Debt  11.35%  9.71%  9.67%  9.54%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37% 
    Debt Cost - Net of Taxes  7.49%  6.41%  6.38%  6.28%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18% 
    WACC in current BRL (Post-Tax)  9.85%  9.44%  9.46%  9.39%  9.37%  9.44%  9.50%  9.57%  9.64%  9.71%  9.71% 

     

    Source: Product 10: AmE D Conclusive Economic and Financial Assessment Report, Service B.

    With respect to the multi-market valuation, Service A, in its Assessment Report of Eletrobras Distribuição AmE D, attached to this Appendix, obtained AmE D enterprise value equal to R$ 2,384,424.000, which represents a difference of 13% in relation to the value of the economic and financial assessment.

    Service B, in Product 10: AmE D Conclusive Economic and Financial Assessment Report, attached to this Appendix, did not obtain the value of enterprise value, but performed comparative analyzes of different implicit multiples to the result of the Discounted Cash Flow assessment.

    3.1 AMAZONAS ENERGIA UNBUNDLING

    It should be noted that the results of the economic and financial assessments consider that AmE D is already vertical divested, that is, in both studies, it is not considered an impact on the value of the company in the implementation of the operation that will result in the separation of the generation and transmission from the distribution business. Operation is in compliance with the provisions of § 5 of article 4 of Law 9074/1995. It is important to emphasize the mandatory nature of such separation occurring before the auction for the sale of share control and the granting of a new concession. Currently, the generation and transmission activities are already within the company Amazonas Energia GT - AmGT, however, this is still a wholly owned subsidiary of AmE D.

    Page 61


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    AmGT is in the unbundling process and preliminary studies indicate that the company's value is positive. Accordingly, upon being vertical divested, AmE D will transfer AmGT to Eletrobras and, as a counterpart, Eletrobras will capitalize or assume AmE D debts in the amount corresponding to AmGT. Thus, the conclusion of the process will lead to an improvement in the value of AmE D in relation to that presented in the consultancy studies contracted by BNDES.

    The deadline for completing the unbundling for the auction of AmE D to happen before the other distribution companies of Eletrobras is 03/02/2018, as defined in CPPI Resolution 20/2017. If the unbundling does not take place in a timely manner, it will be at the discretion of the Granting Power to carry out an auction only for the granting of the concession, and Eletrobras, in this case, is the extreme solution to liquidate the AmE D company.

    3.2 AFAC

    The study presented considers that all Advances for Future Capital Increase AFCA’s made by Eletrobras at AmE D and which are accounted for will be converted into debt or loan. On October 17, 2017, AmE D had R$ 128,15 million in AFAC accounted for. The AFAC contracts already provide for the conditions to be considered in case of conversion into debt. Accordingly, Eletrobras must convert the AFACs into debt, according to a loan agreement executed under the following conditions:

    4.      CONTINGENCIES
    4.1      LEGAL CONTINGENCIES

    AmE D´s legal due diligence, performed by Service B and listed in Product 4: Due Diligence Legal Report - AMAZONAS, pointed out some critical points, although such points

    Page 62


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    do not impede the process of privatization of the distribution company, the new investor should be aware of them so that it can take the applicable measures.

    4.2 ACCOUNTING CONTINGENCIES

    ANEEL carried out inspections at the distribution companies of Eletrobras - AmE D, Eletroacre, Ceron and Boa Vista Energia, regarding the lending of CCC's funds, from June 2009 to June 2016, which aims to cover generation costs in current regulations, namely: Law 12.111/2009, Decree No. 7.246/2010, ANEEL Normative Resolution 427/2011 and Normative Resolution ANEEL No. 597/2013.

    The inspections pointed out that the management of CCC, at the time in charge of Eletrobras, was not applying a series of rules set forth in ANEEL Normative Resolution 427/2011 for the monthly calculation of the financial criteria that involve the total cost of generation of the beneficiaries.

    In the case of AmE D, the ANEEL’s decisions are unfavorable to the distribution company, since the Agency, by means of Order no. 2504/2017, rectified on 27/20/2017 (Case no. 48500.004972/2016-34), understands that Eletrobras, as manager of CCC at the time, has to reimburse CCC in the value of R$ 3.0 billion, connected to the over transfer to AmE D. Besides this, considering that, in the Balance Sheet of AmE D dated 06/30/2017, there is a debt claim receivable of CCC of R$ 4.0 billion, keeping the Agency’s position, there is an accounting risk about R$ 7.0 billion.

    It is understood that this issue is still under discussion between the interested parties and that, therefore, ANEEL decisions unfavorable to AmE D can be reversed. However, the potential reversal of an unfavorable decision should not occur in a timely manner from the completion of this document or even within the timeline of privatization planned for the distribution company. If AmE D is not successful in administrative appeals with ANEEL, it will bring legal action to try to succeed in this matter.

    It is noteworthy that both AmE D and Eletrobras appealed against this decision, and the Agency decided to deny the suspensory effect to the two companies, according to ANEEL Order No. 2.902/2017 and No. 2.903/2017, respectively.

    This issue was pointed out as being of great risk to the eventual buyer of AmE D, and may even render the auction unfeasible, since the amount under discussion may cause an

    Page 63


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    impact of R$ 7.0 billion. In this sense, the proposed model for privatization seeks to neutralize potential effects on the valuation due to uncertainty in the amount effectively recognized of the accounting difference of CCC credits. Such a proposal increases the assertiveness of AmE D's value to the privatization.

    The modeling proposal foresees that AmE D will transfer to the Eletrobras registered CCC credit in the amount of R$ 4.0 billion and debts (which AmE D has with Petrobras) in the same amount, which are already backed by this credit. Accordingly, the accounting effect of such a transfer in AmE D would be zero, and the credit risk is transferred from AmE D to Eletrobras.

    Although AmE D has recognized this amount as a credit in its balance sheet, in the event that ANEEL maintains its position on the matter, even though AmE D has filed a judicial appeal, it is probable, based on accounting rules, that the credit currently recorded becomes a provision. That being so, there is an operation that would originally be neutral for Eletrobras to become a deficit operation in large sums. In this case, Eletrobras, in addition to assuming a debt of R$ 4.0 billion, would record in its result a provision of R$ 3.0 billion with the CCC, thus reaching a possible negative impact of R$ 7.0 billion on its balance sheet.

    Table 42.  Accounting Impact Risk BRL Thousand 
          Possible 
    Value  Balance Value  Accounting 
    ANEEL  (Jun/2017)  Impact 
    -2,906,095.46  3,968,639.00  -6.874.734,46 

     

    Notwithstanding the foregoing, it is worth mentioning, according to the legal opinion prepared by the law firm Décio Freire and Associados, dated September 18, 2017, referring to Administrative Proceeding no. 48500.004972/2016-34 above (AmE D), that while there is no decision of the ANEEL’s Board wholly or partially ratifying the conclusions set forth in ANEEL Order No. 2.504/2017, there is no administrative act that implies an obligation to reimburse the fund. Additionally, the opinion considers that there is a solid legal basis for the review of ANEEL’s rulings and ultimately states that it is too early to provide for any potential loss connected to said Administrative Proceeding or to disclose the matter in questions in explanatory notes to the Financial Statements, as reproduced below:

    Page 64


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    "Under all of the circumstances described above, we believe there is a solid legal basis for reviewing the provisions under the Technical Notes numbers 52/2017-SFF-SFG-SRG and 141/2017-SFF-SFG-SRG, as well as a manifest failure to comply with decisions rendered by Federal Regional Court of the 1st Region, in the records of Writ of Mandamus no. 0026107- 81.2012.4.01.3400.

    This way, we understand that it is too early to provide for any potential loss connected to the Administrative Proceeding in question, or to disclose the matter in explanatory notes to the Financial Statements.”

    5. CAPITAL STRUCTURE

    The structure developed is aimed at enabling the sale of the Eletrobras distribution companies in the business model associated with the granting of a new concession. The following is the current AmE D corporate structure.

    Table 43.  AmE D's capital structureon 12/31/2016 
    Total Shares        6,276,666,628 
    Capital    R$  4,610,171,000.00 
    Nominal Value of Shares  R$    0.73449353 
     
    Interest      Shares  (%) 
    Eletrobras      6,276,656,028  100.00% 
    Minority      0  - % 
     
    Classes of Shares    Shares  (%) 
    Common      6,276,666,628  100.00% 
    Preferred      0  - % 

     

    Source: Product 11: AmE D Privatization Modeling Proposal, Service B.

    The economic value of the company's shares was determined as of 12/31/2016. Then, the definition of values was performed, considering the mean values between Services A and B, as shown in the table below.

    Page 65


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 44.  Valuation of AmE D Mean of Services A and B 
    Service A    R$  (8,431,847,736.90) 
    Service B    R$  (8,895,677,256.91) 
    Average of Services A and B  R$  (8,663,762,496.91) 

    Source: Product 11: AmE Privatization Modeling Proposal, Service B.

    As referred to in Item 8.3 of IT, adjustments to the equity value of AmE D were necessary. The adjustments made were intended to contemplate the update of the balance sheet up to June 2017, which totaled negative R$ 248,054,062.03. Thus, the new value of equity value is R$ 8,911,816,558.94.

     

    Table 45.  Equity Value Final 
     
    Average of Services A and B  R$  (8,663,762,496.91) 
    Consolidated Adjustments  R$  (248,054,062.03) 
    Equity Value Adjusted  R$  (8,911,816,558.94) 

    Source: Product 11: AmE Privatization Modeling Proposal, Service B.

     

    At the end of the privatization process, AmE should present the capital structuredescribed below, considering the total adhesion of employees and retirees to the stock offering:

    Table 46.  AmE D's capital structure after sale   
    Total Shares      201,043,851,652,925 
    Capital  R$    14,068,005,149.60 
    Nominal Value of Shares  R$      0.00006997 
     
    Interest  Common Shares Preferred Shares     Total Shares  (%) 
    Investor  180,939,450,373.560  0 180,939,450,373,560  90.0000% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retirees  20,104,401,279,364  0  20,104,401,279,364  10.0000% 
    Other Minority  0  0  0  - % 
    Shares per Class      Total Shares  (%) 
    Common      201,043,851,652,925  100.0000% 
    Preferred      0  - % 
     
    Source: Product 11: AmE D Privatization Modeling Proposal, Service B.       

     

    Page 66


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    5.1      CAPITALIZATION BY ELETROBRAS AND BY THE INVESTOR
    5.1.1      ELETROBRAS

    Considering the minimum amount of R$ 50,000.00 for AmE D, with the granting of a new concession, the conversion of debt into equity and the assumption of AmE D's debt by Eletrobras in exchange for actions in Stage 1 of the modeling of sales. Accordingly, there is a need to capitalize Eletrobras in the amount of R$ 8,911,866,558.94, as detailed below.

    If Eletrobras chooses to increase its stake in up to 30% in Stage 2, as detailed in item 6.5 of IT, it will have to capitalize the distribution company in the amount of R$ 210.5 million, which may occur through capital contribution or conversion debt.

    5.1.2 INVESTOR

    As anticipated in Stage 2 of the modeling, the investor should make a capital contribution with subscription of new shares in the company, in the amount of R$ 491,370,787.84.

    6. NECESSARY ADJUSTMENTS AT ELETROBRAS

    According to the proposed model to reach the minimum value of R$ 50,000.00, Eletrobras must capitalize R$ 8.9 billion in AmE D, in addition to assuming an additional R$ 4.0 billion, which today are backed by a CCC credit of the same amount, which will also be transferred to Eletrobras, as explained in item 4.2.

    Thus, Eletrobras should prioritize the cheapest debts to be capitalized/assumed, since other debts will be settled by the new Controller. The table below shows AmE D's debt amount by type of creditor and its associated costs:

      Table 47.  Debt amount per creditor and its costs   
     
     
      RO + AFAC World
    Bank 
    Obligation
    RGR
    CCC: 
    Related
    Parties 
    Petrobras Total
    Average
    Cost 
    12.00% 6.59% 5.00% 8.75%
    (Selic) 
    8.75%
    (Selic) 
    AmE D  1,483.25  138.50  107.50 3,910.09  406.00  12,472.30  18,517.64 

     

    Page 67


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    As can be seen from the table above, AmE D's cheapest debts, which will be assumed by Eletrobras to be fully or partially capitalized, are as follows: i) RGR, which has a cost of 5% per year; ii) World Bank, with a cost of 6.59%; iii) Petrobras, at SELIC cost (8.75%); and RO, with a cost of 12.59%. It is important to note that ANEEL must authorize Eletrobras to assume RGR's debt without having to remove all of it at the time of the transfer.

    Debts with the World Bank refer to the raising of funds from this bank for the implementation of the "Projeto Energia +". The RGR debts refer to financing of investments obtained before the designation period as temporary public service provider and RGR resources received during the temporary provision. The Ordinary Resources are the transfers of Eletrobras' own funds to the distribution companies through loan, in order to cover operating deficits. Related parts debts were contracted with other distribution companies or Eletrobras companies, except the Holding Company. Among the main parties related to outstanding credits are, for example, Eletronorte, Furnas and Chesf. Debts with third parties refer to debts due mainly to Petrobras and Cigás.

    The table below shows the detailing of the capitalization by Eletrobras into AmE D, according to the source of funds, to achieve the set price of R$ 50,000.00. It should be underlined that, in this scenario, the preferred option was to convert the RO debts and before the World Bank in capital, although the common resources have a higher cost

    (12.00%), as the debts with Petrobras are posted as bonds under the CCD’s executed.

    Table 48.  Conversion of debts in capital, per distribution company R$ million 
      Debt  Capitalization 
      RO + AFAC  1,272.66 
      IBRD  138.50 
      Related Parties  406.00 
      Petrobras  7,094.71 
      Total  8,911.87 

     

    If the Shareholders’ Meeting of Eletrobras decides to assume the debt claims of CCC and

    CDE recorded in the balance sheets of the distribution companies, under the terms of Paragraph 1, Article 3 of CCPI Resolution 20/2017, as amended by CPPI Resolution no. 28/2017, the table below sets forth the amounts of debts, by source of funds, to be capitalized to achieve the set price in AmE D, as well as the assumption of debts in the

    Page 68


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    same amount of CCC and CDE debt claims. In this context, the debts with Petrobras are prioritized, since most of these debts are already backed up by receivables of CCC, substantiated in Contracts of Admission of Debt, wherein Eletrobras stands as a guarantor.

    Table 49.  Conversion of debts in capital and Assumption of Debt Claims of CCC 
    and CDE, per distribution company R$ million
    Debt    Capitalization  Assumption  Total 
    IBRD    138.50  -  138.50 
    RGR    -  107.5  107.50 
    Petrobras    6,021.17  2,790.16  8,811.32 
    CCC Reimbursement  2,752.20  1,157.89  3,910.09 
    Total    8,911.87  4,055.55  12,967.42 

     

    The table below summarizes the impact on Eletrobras in the event that the company succeeds in its administrative / judicial litigation and in case it cannot reverse the position of ANEEL. It is important to note that the result of the administrative / judicial funds may be in intermediate amounts to those presented in the table below.

    Table 50.  Impact Scenarios at Eletrobras - R$ million 
    SCENARIOS Assumption of Debts Assumption of CCC
    Credits 
    Impact on Eletrobras
    Current position  12,967.42  4,055.55  -8,911.87 
    ANEEL Inspection  12,967.42  -2,906.10  -15,873.52 

     

    6.1 FIXED ASSETS IN USE

    As referred to in Item 2.7 of this Appendix, it should be pointed out that the studies performed by Services A and B did not incorporate the positions of Fixed Assets in Use (AIC) into the Remuneration Base since, according to the Services, the assessors' report about the Remuneration Base does not take into account the specific assessment of the condition and characteristic of these assets concerning: (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the reassessment process. The table below shows the Fixed Assets in use, ascertained based on data of interim accounting balance sheets for each distribution company. The consideration of these

    Page 69


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    assets, amounting to R$ 1,409.9 million at Feb./2017, would improve the results of the economic and financial assessments performed by the Services, and consequently, reduce the amount to be contributed by Eletrobras in the distribution company to reach the minimum amount of R$ 50,000.00, as defined in CPPI Resolution 20/2017.

    7. LIQUIDATION COSTS

    In the 165th EGM, held in July 2016, Eletrobras' shareholders decided, by majority vote, to approve: (i) the non-extension of the concessions of its distribution companies; (ii) the transfer of the ownership control of the distribution companies until December 31, 2017, pursuant to Paragraph 1-A of article 8, of Law 12.783/2013; and (iii) if there is no transfer of control under the terms of the previous item, the adoption of provisions for liquidation thereof.

    Thus, the modeling study carried out by Service B calculated the cost with the liquidation of AmE D, in the case of non-occurrence of item (ii) above or not accepted by Eletrobras of the adjustments made to reach the equity value equal to zero.

    For these cases, Service B developed a comparative analysis between the distribution company’s sales value associated with the grant of a new concession and the occasional liquidation value of the company. Thus, it was possible to evaluate if the sale would be more attractive to the state-owned company than the simple liquidation.

    Since the sale value considers adjustments to the equity value (equity value equal to zero), the adjustments were also considered in the liquidation value estimate. However, there were exceptions to the Tax and Negative Adjustments, since these do not impact on the liquidation value, and for the AFACs, since these can be paid in the event that the distribution companies are not sold.

    As shown below, in the case of AmE D, the best economic option is to sell the company, considering the value of the assets and the concession. The balance between selling and liquidating is favorable for the first option, given the following table:

    Page 70


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 51.  Comparison between sale and liquidation of AmE D 
    Final Equity Value    R$  (8,911,816,558.94) 
    Value of Liquidation (Prior to Adjustments)  R$  (11,929,637,545.60) 
    + Compensation    R$  838,871,261.00 
    (-) Net Debt    R$  (9,959,027,395.23) 
    (-) Contingencies    R$  (975,888,858.53) 
    (+/-) Payments/Receipts  R$  (1,833,592,552.84) 
    Liquidation Amount (After Adjustments)  R$  (12,060,246,607.63) 
    =Value of Liquidation (Before Adjustments)  R$  (11,929,637,545.60) 
    (+/-) Balance Adjustments  R$  (1,131,204,000.00) 
    (+) RGR PPST Reincorporation  R$  1,000,595,937.97 
    Ending Equity Value Liquidation Amount (After Adjustments)  R$  3,148,429,048.70 
    Source: Product 11: AmE D Privatization Modeling Proposal, Service B.     

     

    The results presented above corroborate that the sale of the distribution company, associated to the granting of a new concession, is the option that is less onerous to Eletrobras. If Eletrobras had to simply liquidate the distribution company under the scope of Eletrobras, the latter would have to pay a loss of R$ 12,060,245,607.63, an amount greater than R$ 3,148,429,048.70 when compared to the final equity value of negative R$ 8,911,816,558.84.

    In relation to the CCC credits in discussion with ANEEL, the effects shown in item 6 impact the company in both cases, of sale and liquidation. Thus, it has no direct impact on the comparison of values, so that the economic decision remains unchanged in the event of unsuccessful lawsuits.

    In addition, in the event of distribution company’s liquidation, Eletrobras would not be entitled to the benefits arising from the relaxation of regulatory parameters, which generated a reduction in costs in the order of R$ 1.5 billion for AmE D.

    8.      ANALYSIS OF THE BENEFITS AND RISKS OF DISPOSAL
    8.1      BENEFITS
      (i)      The disposal of AmE D shares will result in the adoption by the distribution
      company      of a legal regime similar to that of other non-state agents operating in the
      distribution      segment, allowing for significant operating gains;

    Page 71


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    (ii) Improvement of Eletrobras' Ebitda and elimination of the need for disbursement of funds by Eletrobras to finance the cash deficit and investments of AmE D, contributing to the reduction strategy of the net debt / ebitda indicator;

    (iii) It avoids the need for capitalization by Eletrobras at the distribution company to meet the financial sustainability targets imposed by ANEEL for the extension of the concession, and for investments in the expansion and improvement of service quality;

    (iv) The sale together with the granting of concession allows the use of the relaxation of the regulatory parameters approved for the period from 2018 to 2023, when the tariff review of the new concession contract will occur; and

    (v) The said flexibility provided an improvement in the distribution company’s equity value of the order of R$ 1.5 billion.

    8.2      RISKS
      (i)      Non-completion of the unbundling of AmE D, with the segmentation of the
      distribution      business of the generation and transmission businesses, which would be in
      charge      of AmE D GT, today a wholly owned subsidiary of AmE D;
      (ii)      Frustration of the disposal of the shares or non-performance of the auction, due to
      several      factors, notably due to possible judicial, political, economic-financial,
      regulatory,      governmental or administrative impediments;
      (iii)      Absence of investors and / or companies interested in purchasing;
      (iv)      Reduced competitiveness in the bidding procedure due to the current economic
      situation      of the country; and
      (v)      Amendment of the minimum conditions for sale by the TCU.

    Page 72


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    9. FINAL REMARKS

    Lastly, it should be noted that the economic and financial assessment methodology was found to be widely known and adopted by the market, which is the Free Cash Flow to Firm - FCFF, in nominal terms. In addition, the calculated discount rates were based on the CAPM - Capital Asset Pricing Model method, which estimates the return that the shareholders expect to obtain from the company, given this is a well-established method widely used either by the market at large or by Eletrobras.

    Additionally, some points need to be highlighted. Among them, there is the need for approval of the minimum amount and the adjustments necessary for achieving such amount by the Executive Board and Board of Directors of Eletrobras, in addition to the need to submit the item to the general shareholders' meeting.

    Another point to be highlighted is that the modeling, as well as the economic-financial assessments prepared by the consultancies, considers the unbundling of AmE D, with segmentation of the distribution activities of the generation and transmission activities, before the sale of the control of the distribution company and grant of new concession. In this sense, the unbundling of the AmGT of AmE D is mandatory for the auction to take place, and, according to CPPI Resolution no. 20/2017, should be completed until March 2, 2018.

    It should also be noted that, while under an unbundling process, AmE D will probably receive, via debt discount, the value corresponding to AmGT. As this value should be significant, it will affect AmE D's value, reducing Eletrobras' need for capitalization of debt. Such fact is already provided for under the CPPI resolution.

    With regard to the need for adjustments, the study establishes the need for capitalization by Eletrobras of the debts in exchange of shares at AmE D in the value of R$ 8,911.8 million, in order for the company’s value to achieve the minimum value of R$ 50,000.00.

    Additionally, CPPI allows Eletrobras to assume R$ 4,055.5 million in debts under the debt claim corresponding entry, with CCC/CDE, at the same amount.

    It should be noted that the studies do not consider the fixed assets in use - AIC, amounting to R$ 1,409.9 million, referenced back to values of Feb./2017. If, in the modeling, AIC’s were considered unitized, even if set aside, they would be able to reduce the value to be disbursed in the distribution company by Eletrobras.

    Page 73


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Furthermore, it should be noted that in the proposed sales model there is the possibility that Eletrobras to remain with one share of AmE D, with the sole purpose of increasing its shareholding in the distribution company by up to 30%, after sale of the controlling interest in the company to a third party. The increase of interest, if adopted by Eletrobras, should be implemented in up to six months after the auction date.

    It should also be reiterated that the modeling entertained the possibility that AmE D has values of CCC previously stated as amounts receivable in the balance sheet, which were currently reverted into obligations payable, owing to technical decisions of ANEEL. The difference of understanding between AmE D and ANEEL achieves amounts as large as R$ 7 billion. Therefore, in view of the relevance of the value discussed for the valuation of AmE D and the unpredictability of the deadline for a final decision on this matter, it was recommended by Service B that the responsibility for such debt claims - including possible benefits and burdens - is transferred from AmE D for Eletrobras’ Holding. Thus, the effect of the current questioning would be neutral on the assessment of the market value of the distribution company and, therefore, on its sales modeling. However, this recommendation requires the consent of the regulators and should be accompanied by considerations of the company's external auditors. Furthermore, the uncertainties existing within the scope of AmE D should also be taken into consideration, as this operating will make them be incurred by Eletrobras. Therefore, this point should be considered by the managers of the Holding.

    It is worth noting that Service B performed a comparative analysis between the option of divesting the shareholder control of the distribution company with the granting of a new concession and the simple liquidation option of the distribution company, as deliberated by the shareholders at the 165th EGM of Eletrobras. With this analysis, one may conclude that the liquidation of the distribution company, under the scope of Eletrobras, as established under the modeling proposed by BNDES, is the most onerous option to Eletrobras, given that, for such purpose, it would have to bear an additional cost of R$ 3.15 billion connected to the sale scenario.

    Lastly, it should be noted that in the case of the distribution company’s liquidation, Eletrobras would not benefit from the relaxation of regulatory parameters, which contributed to an improvement in the value of Boa Vista Energias equity value of the order of R$ 1.5 billion.

    Page 74


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Appendix to the DFP/DFF/DDE Joint Technical

    Information

    Analysis and Results of BOA VISTA ENERGIA

    Page 1


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Contents     
    1.  objective  4 
    2.  Assumptions of economic and financial modeling  4 
    2.1  Market and consumer units  4 
    2.1.1  Evolution of Captive Market  4 
    2.1.2  Market Projections prepared by Service A  5 
    2.1.3  Market Projections prepared by Service B  9 
    2.1.4  Assumptions used and points of divergence of Services A and B  15 
    2.1.5  Evolution of Consumer Units  16 
    2.1.6  Projections of Consumer Units prepared by Service A  17 
    2.1.7  Projections of Consumer Units prepared by Service B  18 
    2.1.8  Comparison of Services A and B projections  19 
    2.1.9  Energy procurement  20 
    2.2  Network EXTENSION (Km)  20 
    2.3  Electrical losses  22 
    2.3.1  Projections of Electrical Losses prepared by Service A  23 
    2.3.2  Projections of Electrical Losses prepared by Service B  25 
    2.3.3  Assumptions used and points of divergence of Services A and B  28 
    2.4  Default and Unrecoverable revenues  29 
    2.5  PMSO (personnel, material, services and others)  32 
    2.5.1  PMSO incurred (2012 - 2016)  32 
    2.5.2  PMSO for Service A  33 
    2.5.3  PMSO for Service B:  39 

     

    Page 2


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
     

    2.5.4 Forecasts and comparisons of the PMSO of Services A and B 

    43 
      2.6  Investment  48 
      2.7  Regulatory Remuneration Base  51 
      2.8  X Factor  56 
      2.9  WACC  61 
    3.    The result of economic and financial assessments and by multiples  63 
      3.1  AFAC  65 
    4.    Contingencies  66 
      4.1  Legal Contingencies  66 
      4.2  Accounting Contingencies  66 
    5.    Capital structure  68 
      5.1  Capitalization by Eletrobras and by the Investor  69 
      5.1.1 ELETROBRAS  69 
      5.1.2 Investor  70 
    6.    Necessary adjustments at Eletrobras  70 
      6.1  Fixed Assets in use  72 
      6.2  Indemnity of the CERR’s assets  73 
    7.    liquidation Costs  74 
    8.    Analysis of the benefits and risks of disposal  76 
      8.1  Benefits  76 
      8.2  Risks  76 
    9.    Final remarks  77 

     

    Page 3


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    1. OBJECTIVE

    The purpose of this appendix is to include a detailed presentation of the results of the economic and financial assessments of Boa Vista Energia performed by Services A and B, contracted by BNDES, as well as to detail the internal adjustments that Eletrobras needs to implement in order to enable its sale at the target price determined by the model approved under CPPI Resolution 20, dated November 8, 2017, amended by CPPI Resolution 28, dated November 22, 2017.

    2.      ASSUMPTIONS OF ECONOMIC AND FINANCIAL MODELING
    2.1      MARKET AND CONSUMER UNITS
    2.1.1      EVOLUTION OF CAPTIVE MARKET

    In the last ten years, from 2007 to 2016, Eletrobras Distribuição Roraima had average annual growth of 8.7%, as can be seen in the tables below. Residential, Rural, Public Power and Public Lighting classes contributed positively to this development, with an annual average growth of 10.3%, 9.5%, 11.7% and 9.6%, respectively.

      Table 1.    Evolution of consumption classes in MWh     
     
              ENERGY SEGREGATED BY CLASS (MWh)       
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  175,787  190,409  214,998  238,781  265,784  297,843  321,545  372,966  406,600  422,544 
    INDUSTRIAL  13,233  13,704  13,422  12,988  12,530  14,238  14,015  16,310  16,904  19,450 
    COMMERCIAL  83,579  92,106  99,345  108,942  120,777  138,374  145,342  161,168  181,358  187,631 
    RURAL  5,999  6,150  7,036  6,883  7,441  12,830  14,427  16,585  13,413  13,641 
    GOVERNMENT  63,326  63,564  68,800  70,259  73,535  79,157  80,625  84,281  83,356  85,140 
    PUBLIC ILUMINATION  18,867  19,571  20,732  20,523  20,708  23,447  24,126  28,132  31,504  41,674 
    PUBLIC SERVICE  13,708  13,330  14,557  16,984  17,437  18,355  19,218  19,553  21,536  21,951 
    OWN CONSUMPTION  1,125  1,145  1,149  1,068  1,049  1,175  1,176  1,173  1,178  1,177 
    TOTAL CONSUMPTION  375,624  399,979  440,039  476,428  519,261  585,419  620,476  700,167  755,848  793,208 
     
     
     
     
      Table 2.  Composition of Consumption Classes     
     
              COMPOSITION OF CLASSES         
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  46.8%  47.6%  48.9%  50.1%  51.2%  50.9%  51.8%  53.3%  53.8%  53.3% 
    INDUSTRIAL  3.5%  3.4%  3.1%  2.7%  2.4%  2.4%  2.3%  2.3%  2.2%  2.5% 
    COMMERCIAL  22.3%  23.0%  22.6%  22.9%  23.3%  23.6%  23.4%  23.0%  24.0%  23.7% 
    RURAL  1.6%  1.5%  1.6%  1.4%  1.4%  2.2%  2.3%  2.4%  1.8%  1.7% 
    GOVERNMENT  16.9%  15.9%  15.6%  14.7%  14.2%  13.5%  13.0%  12.0%  11.0%  10.7% 
    PUBLIC LIGHTING  5.0%  4.9%  4.7%  4.3%  4.0%  4.0%  3.9%  4.0%  4.2%  5.3% 
    PUBLIC SERVICE  3.6%  3.3%  3.3%  3.6%  3.4%  3.1%  3.1%  2.8%  2.8%  2.8% 
    OWN CONSUMPTION  0.3%  0.3%  0.3%  0.2%  0.2%  0.2%  0.2%  0.2%  0.2%  0.1% 
    TOTAL CONSUMPTION  100%  100%  100%  100%  100%  100%  100%  100%  100%  100% 

     

    Page 4


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 3.  Annual Growth Rate in the Period     
     
            ANNUAL GROWTH RATE        ANNUAL 
                        GROWTH 
      2008  2009  2010  2011  2012  2013  2014  2015  2016  AVERAGE 
    RESIDENTIAL  8.3%  12.9%  11.1%  11.3%  12.1%  8.0%  16.0%  9.0%  3.9%  10.3% 
    INDUSTRIAL  3.6%  -2.1%  -3.2%  -3.5%  13.6%  -1.6%  16.4%  3.6%  15.1%  4.7% 
    COMMERCIAL  10.2%  7.9%  9.7%  10.9%  14.6%  5.0%  10.9%  12.5%  3.5%  9.5% 
    RURAL  2.5%  14.4%  -2.2%  8.7%  72.4%  12.4%  15.0%  -19.1%  1.7%  11.7% 
    GOVERNMENT  0.4%  8.2%  2.1%  4.7%  7.6%  1.9%  4.5%  -1.1%  2.1%  3.4% 
    PUBLIC LIGHTING  3.7%  5.9%  -1.0%  0.9%  13.2%  2.9%  16.6%  12.0%  32.3%  9.6% 
    PUBLIC SERVICE  -2.8%  9.2%  16.7%  2.7%  5.3%  4.7%  1.7%  10.1%  1.9%  5.5% 
    OWN CONSUMPTION  1.8%  0.3%  -7.0%  -1.8%  12.0%  0.1%  -0.3%  0.5%  -0.1%  0.6% 
    TOTAL CONSUMPTION  6.5%  10.0%  8.3%  9.0%  12.7%  6.0%  12.8%  8.0%  4.9%  8.7% 

     

    2.1.2 MARKET PROJECTIONS PREPARED BY SERVICE A

    The consumption of electricity in Brazil can be segregated into eight distinct classes: Residential, Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service and Personal Consumption. In order to capture the possible heterogeneities among the analyzed groups, Service A separately modeled the Residential, Commercial, Industrial, Rural and Other sectors (which compose the classes of Public Power, Public Lighting, Public Service and Personal Consumption). This decision was based on the similar behavior of consumption among them and to minimize the effect of the volatility incurred in low consumption classes.

    To forecast the energy demand for the next 30 (thirty) years, ending in February/2048, the consultant first made a historical analysis of the evolution of energy consumption and its distributions between classes, using an observable period of consumption from 2007 to 2016. The results are presented in the following tables:

    Page 5


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Gráfico 1. History of Consumption by Classes

    RESIDENTIAL  INDUSTRIAL  COMMERCIAL  RURAL  OTHER 
    Consumption  Consumption  Consumption  Consumption  Consumption 

     

      Table 4.  Geometric Average of Consumption 2007-2016   
      RESIDENTIAL  INDUSTRIAL  COMMERCIAL  RURAL  OTHER  TOTAL 
    Period  Consumption  Consumption  Consumption  Consumption  Consumption  Consumption 
      ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a 
    2007 - 2016  10.22%  4.34%  9.40%  9.51%  5.00%  8.66% 
     
     
    a . Residential Class Projection         

     

    For Residential class, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Ordinary Least Squares (OLS) methodology, with a dependent variable in the model, the residential per capita energy consumption variation, and as explanatory variables a trend line and the annual GDP of Brazil were used. From the estimated forecasts, consumption grows on average 3.67% p.a. between 2017 and 2047.

    b. Industrial Class Projection

    A deterministic linear model was used in which time consumption (t + 1) is equal to consumption at time t plus a parameter Ë defined as the average of the unit change in

    MWh between year t (t + 1).

    c.(t+1) = c.i(t) * u

     

    Page 6


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    . = Industrial Consumption.

    From the estimated forecasts, consumption grows on average 3.40% p.a. between 2017 and 2047.

    c. Commercial Class Projection

    To project the Commercial consumption of energy, a model of time series SARIMA was elaborated.

    From the estimated forecasts, consumption grows on average 3.92% p.a. between 2017 and 2047.

    d. Rural Class Projection

    A model was designed to project the Rural consumption of energy through the GDP and an exponential smoothing factor Ê = 96%, where it is assumed that Ê has effect from 2022.

    The following is the model definition:


    From the estimated forecasts, consumption grows on average 1.68% p.a. between 2017 and 2047.

    e. Other Class Projection

    The model created to design the class Others was based on the premise that there is a linear dependency relation between it and the Residential consumption class. The model can be defined as:


    Page 7


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017


    Based on the estimated forecasts, the consumption of the classes Public Power, Public Lighting, Public Service and Own Consumption, classified in the Others class, grew by an average of 3.19% p.a. between 2017 and 2047.

    f. Global Projection Results by Consumption Class

    After preparing the projection of each sector separately, the results were grouped on a consolidated basis. Below are graphs of the projection of aggregate energy consumption, as well as a general table of growth per class. There is also an increase in the annual variation of consumption in the first years, reaching the highest value in 2019 (around 5.5%), slowly reducing the annual growth rate to 2% in 2047.

    Page 8


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

      Table 5.  Geometric Mean of Consumption 2017-2047   
     
      RESIDENTIAL  INDUSTRIAL  COMMERCIAL  RURAL  OTHER  TOTAL 
    Period  Consumption  Consumption  Consumption  Consumption  Consumption  Consumption 
      ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a 
    2017 - 2047  3.67%  3.40%  3.92%  1.68%  3.19%  3.50% 

     

    g. Projection of Demand by Voltage Level

    For the projection of demand by Voltage Level, the historical average of low voltage of the last 5 years of the Company was used. This metric was based on the idea of stability in the composition of energy consumption between classes. Low voltage consumption was defined and the residual percentage was divided between medium and high voltage according to the SPARTA report. The set percentages for 2017-2047 are 75% for LV, 25% for AV.

    2.1.3 MARKET PROJECTIONS PREPARED BY SERVICE B

    In order to know the reality experienced by the distribution company and use the historical values for the necessary projections, Service B initially collected consumption information

     Page 9


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    from the utility in the period from 2013 to 2016. The evolution of the Market by level of voltage in the period is presented in the table below:

      Table 6.  Market evolution by Voltage Level - Capital   
     
      Market by Voltage Level (MWh)      Variation Rate 
    Voltage / Year  2013  2014  2015  2016  Period  Geometric 
              (2013/2016)  (p.a.) 
    PROVISION  608,049  686,074  742,397  790,750  30%  9% 
    A3  -  -  -  -  -  - 
    A3a  -  -  -  -  -  - 
    A4  149,077  162,830  178,971  188,181  26%  8% 
    LV  458,971  523,244  563,426  602,569  31%  10% 
    LOW INCOME  18,326  39,763  36,953  31,145  70%  31% 
    SUPPLY  163,938  186,597  203,091  221,728  35%  11% 
    OTHER FREE  -  -  -  -  -  - 
    DISTRIBUTION  -  -  -  -  -  - 
    TOTAL  771,987  872,672  945,489  1,012,479  31%  9% 

     

    For CERR, the only available public information about the CERR market that dates from 2014 through Technical Note No. 366/2014 - SRE/ANEEL, which is presented below, was used:

    Table 7.  Number of consumers and Interior Market (2014) 
     
    Voltage Level  Consumers    Market (MWh)   
    HV  122  0.3%  10,581  11.1% 
    LV  39,555  99.7%  84,736  88.9% 
    LOW INCOME  21,122  53.2%  16,491  17.3% 
    TOTAL  39,677  100.0%  95,317  100.0% 

     

    In order to expand the analysis, it was estimated for the regulatory years of 2015 and 2016, the Interior Market by applying the market weight of CERR in relation to the level of supply of Boa Vista in the level of supply verified in the capital, obtaining the following values:

    Table 8.  Estimated evolution of the billed energy market in the Interior (MWh)
     
    Region/Year  2014  2015  2016  Period  Geometric 
            (2014/2016)  (p.a.) 
    Interior  95,318  103,373  113,263  19%  9% 

     

    In order to subsidize the various tariff calculations and the flow of the company, Service B prepared for the capital market simulations by: (i) consumption class (Residential,

    Page 10


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service, Own Consumption); and (ii) voltage level (High Voltage - A1, A2 and A3 -, Medium Voltage -A3a and A4 - and Low Voltage - LV), as well as a projection to the Interior.

    a. Market Projection by Voltage Level

    The market projection by voltage level was made only for the capital and took into account the evolution of the share of voltage levels in the total market since 2012, according to information provided by the utilities. From the analysis of the behavior of these shares, it was possible to construct an evolution curve, based on linear extrapolation, generating the market projections for each voltage (High, Medium and Low). The evolution of the share of each tension level is presented below:

    Gráfico 4. Share of the voltage level in the total market (%) - Projections


    b. Projection of the Market by Consumption Classes

    In each consumption class the models may differ depending on the distribution company. After numerous tests, the auxiliary variables used in the models by consumption classes are presented below. The model adjustments followed a hierarchical logic prioritizing the variables that appear first in the "Tested" column of the Table below.

    Page 11


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 9.  Variables used in market forecasts by class 
     
    Class  Variables 
    Tested  Used 
    Residential  Population or UC Residential  Population 
    Industrial  GDP    GDP 
    Commercial  GDP and Population  GDP and Population 
    Rural  GDP and Rural Population  GDP 
    Government  Population, GDP and/or Univariate  Univariate 
    Public Lighting  Univariate  - 
    Public Service  Population  Population 
    Own consumption  Univariate  - 

     

    For the projection of the domestic market, it was adopted the assumption provided by the company for growth of the energy market between 2017 and 2026 (5% p.a.) and from 2027, the rate of growth of each consumption class of Boa Vista was applied (capital) to the classes of the interior.

    For segregation by consumption classes, it was made considering the share of each class over total consumption in October 2014, the latest information available on the Aneel website.

    b.1 Residential Class Result

    For the Residential Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Population as an auxiliary variable and presence of seasonality, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 4.32% p.a. between 2017 and 2048.

    b.2 Industrial Class Result

    For the Industrial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with State GDP as an auxiliary variable and presence of seasonality, using the observed consumption series from January 2012 to December 2016. From the estimated forecasts, consumption grows on average 4.05% p.a. between 2017 and 2048.

    b.3 Commercial Class Result

    For the Commercial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology,

    Page 12


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    with Population and State GDP as auxiliary variables, in addition to the seasonality, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 2.67% p.a. between 2017 and 2048.

    b.4 Result of the Rural class

    For the Rural Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with GDP as an auxiliary variable, presence of seasonality and structural break in January 2012 onwards, using the observed consumption series from January 2004 to December 2016. From the estimated forecasts, consumption grows on average 3.20% p.a. between 2017 and 2048.

    b.5 Result of the Public Power class

    For the Public Power class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the ETS methodology, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 1.89% p.a. between 2017 and 2048.

    b.6 Result of Public Lighting class

    For the Public Lighting class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with component of stochastic trend, using the observed consumption series from January 2012 to December 2016. From the estimated forecasts, consumption grows on average 1.07% p.a. between 2017 and 2048.

    b.7 Result of the Public Service class

    For the Public Service class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Population as an auxiliary variable and presence of seasonality, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 1.12% p.a. between 2017 and 2048.

    b.8 Result of Own Consumption class

    For the Own Consumption class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with structural break as of January 2012 and presence of seasonality, using the observed

    Page 13


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 0.49% p.a. between 2017 and 2048.

    b.9 Global Projection Result

    The table below shows the market projected for the capital and interior by consumption class broken down every ten years between 2016 and 2046. The consumption in MWh per class is presented through its percentage participation in relation to the total consumption and the percentage growth between the ten-year periods. The following graphs illustrate the composition of consumption classes in 2046 for both the capital and the interior.

    Table 10.  Market Designed every Decade by Consumption Class - Capital 
     
        MWh        Participation    Average Growth p.a.   
    CLASS                       
      2016  2026  2036  2046  2016  2026  2036  2046  2017/2026  2027/2036  2037/2046 
    Residential  422,544  663,985  999,496  1,504,756  53.17%  55.97%  58.50%  61.47%  4.62%  4.17%  4.18% 
    Industrial  187,631  296,770  436,339  623,723  23.61%  25.02%  25.54%  25.48%  4.69%  3.93%  3.64% 
    Commercial  19,450  26,869  35,419  43,208  2.45%  2.26%  2.07%  1.76%  3.28%  2.80%  2.01% 
    Rural  13,641  18,760  26,105  35,054  1.72%  1.58%  1.53%  1.43%  3.24%  3.36%  2.99% 
    Public Authority  85,140  108,786  129,744  150,703  10.71%  9.17%  7.59%  6.16%  2.48%  1.78%  1.51% 
    Public Ilumination  21,951  25,116  27,879  30,384  2.76%  2.12%  1.63%  1.24%  1.36%  1.05%  0.86% 
    Public Service  41,674  42,837  50,318  57,070  5.24%  3.61%  2.95%  2.33%  0.28%  1.62%  1.27% 
    Own Consumption  2,746  3,185  3,203  3,204  0.35%  0.27%  0.19%  0.13%  1.49%  0.06%  0.00% 
    Total  794,777  1,186,308  1,708,503  2,448,102  100.00%  100.00%  100.00%  100.00%  4.09%  3.72%  3.66% 

     

    Page 14


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Table 11. Market Projected Each Decade by Consumption Class - Interior 
     
        MWh      Participation    Average Growth p.a. 
    CLASS   2016  2026  2036  2046  2016  2026  2036  2046  2017/2026  2027/2036  2037/2046 
    Residential  51,679  74,230  111,726  168,205  45.5%  44.5%  47.7%  51.5%  3.7%  4.2%  4.2% 
    Industrial  4,106  5,994  7,905  9,645  3.6%  3.6%  3.4%  3.0%  3.9%  2.8%  2.0% 
    COMMERCIAL  15,717  22,715  33,392  47,731  13.9%  13.6%  14.3%  14.6%  3.8%  3.9%  3.6% 
    Rural  19,108  29,908  41,616  55,867  16.8%  17.9%  17.8%  17.1%  4.6%  3.4%  3.0% 
    Public power  10,560  15,922  18,992  22,061  9.3%  9.5%  8.1%  6.8%  4.2%  1.8%  1.5% 
    Public Lighting  6,954  10,148  11,920  13,519  6.1%  6.1%  5.1%  4.1%  3.9%  1.6%  1.3% 
    Public Service  5,254  7,693  8,540  9,307  4.6%  4.6%  3.6%  2.9%  3.9%  1.0%  0.9% 
    Own                       
    Consumption  93  136  136  136  0.1%  0.1%  0.1%  0.0%  3.9%  0.0%  0.0% 
    Total  113,471  166,746  234,227  326,471  100.0%  100.0%  100.0%  100.0%  3.9%  3.5%  3.4% 

     

    2.1.4 ASSUMPTIONS USED AND POINTS OF DIVERGENCE OF SERVICES A AND B

    1. The assessment reports lead to the conclusion that Service A used annual consumption figures for the projections, while Service B used as assumption monthly consumption figures for the projection calculations.

    2. To forecast the demand, Service A developed projection simulations by consumption class, distributed in five categories: Residential, Industrial, Commercial, Rural and Others (composed of Public Power, Public Lighting, Public Service and Own Consumption). For the Projection of demand by Voltage Level, the historical average of low voltage of the last 5 years of the Company was used.

    Page 15


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     

    Service B has elaborated two types of simulations for forecasting demand. A projection by consumption classes, distributed in eight categories: Residential, Industrial, Commercial, Rural, Government, Public Lighting, Public Service and Personal Consumption. And another projection by voltage level, composed of: High Voltage (A1, A2 and A3), Medium Voltage (A3a and A4) and Low Voltage (LV).

     

    3.     

    The historical consumption period of the Distribution Company, used by Service A, is the years 2007 to 2016. On the other hand, Service B used periods beginning in 2004, 2009, 2010 or 2012, depending on the projected consumption class and always ending in December 2016.

     

    4.     

    Only Service B made projections for the domestic market.

     

    5.      Service A obtained average annual growth projections of 3.50% for the capital's consumption in the period from March 2018 to February 2048. For Service B, the average annual growth of the distribution company's consumption was around 3.83% for the period from 2017 to 2048.

    The average annual growth of the consumption classes of Service A and B are presented in the table below:

    Table 12.  Geometric Mean of Consumptions of Services A and B for the period - Capital
     
    Geometric Average of Consumptions of Services A and B
                PUBLIC  PUBLIC     
      RESIDENTIAL  INDUSTRIAL COMMERCIAL  RURAL  GOVERNMENT      OWN  TOTAL 
    Services  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a  LIGHTING  SERVICE  ”% p.a  ”% p.a 
                ”% p.a  ”% p.a     
    A  3.67%  3.40%  3.92%  1.68%  3.19%  3.19%  3.19%  3.19%  3.50% 
    B  4.32%  4.05%  2.67%  3.20%  1.89%  1.07%  1.12%  0.49%  3.83% 
    Differences (A-B)  -0.65%  -0.65%  1.25%  -1.52%  1.30%  2.12%  2.07%  2.70%  -0.33% 

     

    2.1.5 EVOLUTION OF CONSUMER UNITS

    The tables below show the evolution of the Consumer Units (UC) of consumption classes in recent years, from 2009 to 2016 for the Capital. The company distribution had an average annual growth of 5.2% of its UCs. Of particular note are the Residential and Commercial classes, which together account for 96% of Eletrobras Distribuição Roraima UCs.

    Page 16


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 13. UCs Evolution by Consumption Classes
     
      2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  69,603  72,247  76,654  80,245  84,591  89,353  93,288  98,715 
    INDUSTRIAL  313  309  324  353  337  326  342  343 
    COMMERCIAL  6,864  7,286  8,121  8,762  8,800  8,937  9,166  9,430 
    RURAL  1,829  1,803  2,463  2,824  2,814  2,786  2,848  3,003 
    GOVERNMENT  758  758  768  766  795  787  791  814 
    PUBLIC LIGHTING  31  31  38  54  82  96  95  102 
    PUBLIC SERVICE  77  62  63  63  68  79  77  76 
    OWN CONSUMPTION  7  7  8  17  17  16  13  15 
    TOTAL CONSUMPTION  79,482  82,503  88,439  93,084  97,504  102,380  106,620  112,498 
     
     
     
     
      Table 14.  UCs Composition by Consumption Classes     
     
     
      2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  87.6%  87.6%  86.7%  86.2%  86.8%  87.3%  87.5%  87.7% 
    INDUSTRIAL  0.4%  0.4%  0.4%  0.4%  0.3%  0.3%  0.3%  0.3% 
    COMMERCIAL  8.6%  8.8%  9.2%  9.4%  9.0%  8.7%  8.6%  8.4% 
    RURAL  2.3%  2.2%  2.8%  3.0%  2.9%  2.7%  2.7%  2.7% 
    PUBLIC POWER  1.0%  0.9%  0.9%  0.8%  0.8%  0.8%  0.7%  0.7% 
    PUBLIC LIGHTING  0.0%  0.0%  0.0%  0.1%  0.1%  0.1%  0.1%  0.1% 
    PUBLIC SERVICE  0.1%  0.1%  0.1%  0.1%  0.1%  0.1%  0.1%  0.1% 
    OWN CONSUMPTION  0.0%  0.0%  0.0%  0.0%  0.0%  0.0%  0.0%  0.0% 
    TOTAL CONSUMPTION  100%  100%  100%  100%  100%  100%  100%  100% 
     
     
     
     
    Table 15. Annual Growth Rate in the Period
     
                    ANNUAL 
                    GROWTH 
      2009  2010  2012  2013  2014  2015  2016  AVERAGE 
    RESIDENTIAL  6.0%  3.8%  4.7%  5.4%  5.6%  4.4%  5.8%  5.2% 
    INDUSTRIAL  -1.9%  -1.3%  9.0%  -4.5%  -3.3%  4.9%  0.3%  1.0% 
    COMMERCIAL  4.4%  6.1%  7.9%  0.4%  1.6%  2.6%  2.9%  4.7% 
    RURAL  9.5%  -1.4%  14.7%  -0.4%  -1.0%  2.2%  5.4%  8.2% 
    PUBLIC POWER  1.2%  0.0%  -0.3%  3.8%  -1.0%  0.5%  2.9%  1.1% 
    PUBLIC LIGHTING  63.2%  0.0%  42.1%  51.9%  17.1%  -1.0%  7.4%  25.4% 
    PUBLIC SERVICE  4.1%  -19.5%  0.0%  7.9%  16.2%  -2.5%  -1.3%  0.8% 
    OWN CONSUMPTION  -22.2%  0.0%  112.5%  0.0%  -5.9%  -18.8%  15.4%  11.9% 
    TOTAL CONSUMPTION  5.9%  3.8%  5.3%  4.7%  5.0%  4.1%  5.5%  5.2% 

     

    2.1.6      PROJECTIONS OF CONSUMER UNITS PREPARED BY SERVICE A
    A      model was used based on two parameters, the Population/Residential UC ratio and the

    average growth of Residential UC for the last 6 years. The same projected residential UC growth was extrapolated to the other classes. As a result of the projections, there was a growth of UCs in the first years, with initial variation of 4.5%, with a gradual reduction of growth until reaching zero growth in 2047, consolidating around 275,000 UCs. In the graph below the results of the projection follow:

    Page 17


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017


    2.1.7 PROJECTIONS OF CONSUMER UNITS PREPARED BY SERVICE B

    Due to the differentiated behavior of the growth of the historic series of Residential UCs and other UCs, Service B, for the capital, chose to separately project the two series and then aggregate them, forming the series of Total UCs.

    The evolution of Residential UCs is based on the growth of residential households and the electrical energy coverage that these households have over the years 2017 to 2048. For such purpose, data from the National Household Sample Survey - PNAD (IBGE) were used for the construction of the historical household series and the historical series of the population.

    Series Other CUs is based on the projection of average consumption that consists of the quotient between the consumption in MWh of classes (except residential) divided by total UC (except residential). The projection of average consumption was based on the trend of linear growth, converging to the national average consumption 1,749 kWh per year (ANEEL, 2016, considering only the distribution utilities), in up to 30 years.

    Page 18


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The result obtained from the UC projection, with historical series up to 2048, is shown in the graph below. It is noted that the growth of UCs is constant and increasing until the end of the concession. The average growth registered in the period from 2016 to 2048 was 3.66% per year


    For the projection of the consumer units of Boa Vista (Interior), the same growth rate of the total market was assumed for increments of consumer units, obtaining as a result the table below:

      Table 16.  Projection of Interior UCs   
     
    2016  2026  2036  2046  2017/2026  2027/2036  2037/2046 
              (% p. a.)  (% p. a.)  (% p. a.) 
    46,469  64,714  90,909  126,711  3.40%  3.50%  3.40% 

     

    2.1.8 COMPARISON OF SERVICES A AND B PROJECTIONS

    Service A did not make estimates for the evolution of the number of Consumer Units for the Interior. Thus, based on the trajectory of the graphs of the UCs of Services A and B, we have a comparison between their evolutions over the period for the Capital. It is noted that in the first 10 years the growth rate of Service A is higher than that of B, the opposite is the case in the last 20 years.

    Page 19


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
        Table 17.  Annual Growth Rate in the Period     
     
        Incurred        Projected     
    Services      2015/2016        2017/2026  2027/2036  2037/2046 
      2015  2016    2026  2036  2046       
          (% p. a.)        (% p. a.)  (% p. a.)  (% p. a.) 
    A        210,000  250,000  275,000  6.44%  1.76%  0.96% 
      106,620  112,498  5.51%             
    B        171,359  239,414  329,757  4.30%  3.40%  3.25% 
    Differences (A-B)  -  -  -  38,641  10,586 -  54,757  2.14%  -1.64%  -2.30% 

     

    2.1.9 ENERGY PROCUREMENT

    In relation to the energy procurement balance, Service B considered the existing power purchase agreements, the current level of overcontracting and explained the treatment given, as well as the projection of the PLD. Service A has assumed the non-treatment of overcontracting from the starting point. According to the Disclaimer of the consultancy work: "In this assessment were not considered penalties in the overcontracting of energy by the assessed enterprise, since it was assumed the assumption that the new utility will be able to liquidate its positions at market prices". That is, the revenue from the sale of excess energy in the short-term market would be equal to the distribution company's average energy purchase cost.

    2.2 NETWORK EXTENSION (KM)

    The annual variation of the extension of the distribution network has an impact on the distribution company's projection of investments, and is part of the projection metric for the calculation of regulatory operating costs and PMSO carried out by the distribution company.

    The total extension of the network can be segregated into four voltage groups: low, medium and high voltage and “Luz para Todos” program.

    It is worth mentioning that in January 2017, Boa Vista Energia started to serve the interior of the state (former CERR's area of activity). Thus, Services A and B projected the physical advance taking into account the new characteristics of the concession area.

    In service A, the network extension projection started from the low, medium and high voltage network extension history between the years 2001 and 2016. Added to this history was the network extension foreseen by the Decennial Plans and PDDs of the distribution companies, from which the network projected for the years 2017 to 2021 for low and

    Page 20


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    medium voltage, 2017 and 2018 for “Luz para Todos” program and high voltage from the years 2017 to 2026.

    Service A presented the following projection of network extension until the year 2027:

      Table 18.  Network Extension Projection 2027 - Service A     
     
    Network Projection (Km)                     
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    LV/AV  10,322  10.66  10,680  10,697  10,697  10,747  10,797  10,847  10,897  10,947  10,997 
    HV  266  316  316  316  316  316  316  316  325  334  348 
    PLPT  0  713  713  713  713  713  713  713  713  713  713 
    Total  10,588  11,690  11,709  11,726  11,726  11,776  11,826  11,876  11,935  11,994  12,058 

     

    That is equivalent to one of percentage growth per year, according to the respective five-year period, according to the table below:

    Table 19.  Percentage growth of network extension, per five-year period - Service A
     

     
      1st Five-year  2nd Five-year  3rd Five-year  4th Five-year  5th Five-year  6th Five-year   
    Annual growth (%)  period  period  period  period  period  period  Total Mean 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
      2.2%  0.5%  0.5%  0.5%  0.5%  0.5%  0.8% 
     
     
     
    Service B did  not show Network projection numbers, only presented the econometric 

     

    model used to estimate the growth projection of the network extension. The total network extension of each company was projected according to a calculated elasticity for the total number of consumer units in the concession area. To determine this elasticity, panel data from the six Eletrobras distribution companies were used for the period between 2007 and 2016.

    Regarding the starting point, that is, the network extension for the year 2017, according to the information contained in the assessment reports of Services A and B, the Consulting Firms started with very different information:

    Service A reported for the year 2017 the network extension numbers of 10,322 km in LV/AV, 226 km in AT and zero for PLPT, not showing whether it was only the network extension of the city of Boa Vista or all the State of Roraima, including the assets of the CERR concession area. From the figures presented, it is deduced that the whole State was considered.

    Page 21


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Service B reported the figures only to the capital Boa Vista: 1,307 km for LV, 1,967 km for AV, 708 km for AT. No information on PLPT km was cited.

    2.3 ELECTRICAL LOSSES

    The losses of the electric distribution system are divided into two categories: Losses in the Basic Network and Losses in the Distribution Network. In turn, losses in the Distribution Network are divided into Technical Losses (PT) and Non-Technical Losses (PNT).

    The losses in the Basic Network are those admitted by the SIN. The real values are calculated monthly by the CCEE, and the regulatory value is a result from the average of the measurements of the last 12 months. The value of the 2016 tariff adjustment was used for the assessment report.

    The PT is losses caused in transport and electrical equipment. They are calculated by ANEEL in the tariff review period and remain constant for the entire tariff cycle.

    PNTs are the manageable losses where ANEELs reduction curve is estimated at each tariff review and takes into account three factors: the target, the path to the target and the limits of reduction of regulatory loss. The goal is calculated based on the analysis of the comparative model of benchmarking, in which a company is established that will serve as a basis for comparison, taking into account the size of the distribution company and the socioeconomic complexity of the region in which the distribution company is located. According to the metric specified in submodule 2.6 of PRORET. It should be reminded that the percentage levels of PNT are related to the Low Voltage (LV) market of the utility, while the PT and Loss in RB indices are measured in relation to the Injected Energy, in accordance with the regulatory methodology.

    Below are presented the Percentage values of the Losses, according to the Boa Vista Energia worksheet of 2016, as well as the realized values of 2016 of the distribution company, according to NT 149/2017-ANEEL.

    Page 22


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Table 20.  Calculation of losses - SPARTA Boa Vista Energia 2016 
     
    Description Regulatory Performed
    2016 
    Non-Technical Percentage (w/o Low Voltage)  11.05%  10.55% 
    Technical Percentage (w/o input market)  7.04%  7.04% 
    % Basic Grid (w/o Input Market)  0.00%  0.00% 

     

    2.3.1 PROJECTIONS OF ELECTRICAL LOSSES PREPARED BY SERVICE A

    a. Projection of Losses in the Basic Network and Technical Losses

    For Losses in the Basic Network and PT, Service A established that for both indicators the regulatory and real values will remain constant and equal throughout the period, since the values of these losses are not related to the activity of the distribution company and its management, but rather physical processes and characteristics of the equipment used.

    Level ofTechnical Losses and Losses in the Basic Network 
    % Loss / Injected energy 
     Losses in Basic 
     Technical Losses  Network 
    7.04%  0.00% 

     

    b. Projection of Non-Regulatory Technical Losses

    Regarding the trajectory of Regulatory NTP, Service A defined in two different moments. For the period from 2017 to 2025, regulatory levels were defined based on Law 13.299 and Process Vote 48500.004245/2016-77, through Homologation Resolution No. 2.184, which stipulated specific levels of losses for the Boa Vista Energia.

    As of 2026, the method of calculation of the curve of regulatory losses occurred according to the definitions described in the submodule 2.6 of PRORET, taking into account three previously mentioned factors. After comparing the companies listed with Boa Vista Energia, it was defined that their benchmarking of the PNT target would be UHENPAL -Nova Palma Energia.

    Upon definition of the target for the next year of the tariff review, the analysis of the annual reduction of regulatory losses that can have two forms of reduction for each cycle of tariff revision, is carried out. The trajectory will be the limit of reduction of regulatory

    Page 23


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    losses stipulated in PRORET, in case the effort to reach the goal of benchmarking is more challenging. Otherwise, the reduction curve will have a linear trajectory until it loses benchmarking at the end of each tariff review cycle.

    The trajectory of reduction of PNTs continues until reaching the level of 2.5% according to PRORET's determination.

    c. Projection of Non -Technical Real Losses

    For the projection of real PNTs, three perspectives were observed: starting point, trajectory and target.

    As a starting point, the PNT was considered in 2016 of 10.55%, obtained from the energy balance of the distribution company. The defined target is the level of PNT practiced by the benchmark company considered in the calculation of the target of regulatory losses, in this case, the levels practiced by UHENPAL which are of 0.32%. The value of the loss of the benchmark company was obtained through the SAMP database.

    For the annual PNT reduction trajectory, the observed trajectory of an efficient utility was used. The benchmark company selected was CEMAR - Companhia Energética do Maranhão. The choice of the company was based on the fact that CEMAR already has an efficient level of Non-Technical losses and has recently undergone the privatization process.

    Table 21.  Starting Point and Target of Non-Technical Losses 
     
    % Losses / LV Market
    Non-Technical Losses
    in 2016
    10.55% 
    Non-technical losses
    Benchmark in 2016
    0.32% 
    Year in which it
    reaches the regulatory
    level
    2017 

     

    Page 24


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Gráfico 9. Evolution of Non-Technical Losses (%/LV Market)

    2.3.2 PROJECTIONS OF ELECTRICAL LOSSES PREPARED BY SERVICE B

    a. Projection of Technical Losses (PT)

    The PT projection consists of two distinct moments. For the period from 2017 to 2025, regulatory levels were defined based on Law 13.299 and Process Vote 48500.004245/2016-77, through Homologation Resolution No. 2.184, which stipulated joint specific levels of losses for the companies Boa Vista Energia and CERR in 7.62%. The level of losses remains the same for the years of 2026 and 2027.

    As of 2028, the PTs projection methodology consists of evaluating the technical loss indices of each supply voltage segment (IPTS) of Boa Vista Energia (NT 159/2013-SRD/ANEEL) together with those of CERR (NT 0035/2016- SGT/ANEEL) and to compare them with the average indices of compatible companies. The lowest indices by segment (from the Roraima utilities or the weighted average of the compatible companies) were used as the basis for an efficient PT target in 2047.

    Based on the method described, a linear trajectory of the PT level in 2019 (7.60%) was traced to the PT level projected for 2047 (6.92%). In order to have a modeling that

    Page 25


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    represents each tariff period, the indices that coincide with each tariff review were used, and these remain constant during the cycle. The graph below shows the projection of the percentage of regulatory technical losses.

    Regarding Losses in the Basic Network, Service B does not present any projection model for the period from 2017 to 2047.


    b. Projection of Non -Regulatory Technical Losses

    Since there is no complexity index to combat PNT calculated for the whole of the state of Roraima, it was chosen to calculate the Regulatory PNTs of the Capital. The modeling of the Regulatory PNT curve, which covers both Capital and the Interior, is based on two situations. The first, from 2017 to 2025, the PNT level of ANEEL Homologation Resolution 2.184/2016 is admitted. The homologated level continues constant in the years of 2026 and 2027. From 2028, the PNT Total Regulatory trajectory has a behavior proportional to the PNT Regulatory curve projected for the capital Boa Vista until 2047.

    The projections of the Regulatory PNTs for the capital Boa Vista, from 2028 to 2047, were based on the successive application of the mechanisms for setting goals, starting points and detailed trajectories in NT 106/2015-ANEEL. The models considered for the projection of the trajectory of loss targets were those estimated with the database defined at the time.

    Page 26


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    c. Projection of Non -Technical Real Losses

    As the real PNT trajectories also depend on social complexity, real PNT is calculated separately for the Capital and Interior of the state.

    c.1 Projection of the Real PNTs of the Capital

    During the Temporary Provision scheme, in force in 2017 and possibly at the beginning of 2018, it was considered that economic, financial and technical-managerial difficulties will limit efficiency gains in the combat against PNT. Therefore, it was defined that at Boa Vista Energia, the current level of Real PNTs (relative to the year 2016) will be in force until the start of the new administration in 2018.

    From 2019 it is considered that a new phase in the combat against PNT is initiated with successive gains of efficiency and effectiveness of these activities. For Boa Vista Energia, the real losses reduction trajectory will be equal to the average reductions (p.p. per year) in real PNTs in historical five-year periods (from 2003 to 2016) of SANTA MARIA and FORCEL companies. Based on this real performance, reductions were projected for Boa Vista Energia until it reached the simulated target in the regulatory model. Upon reaching the regulatory target before 2047, a new target is calculated from the mean of the lowest real PNT of benchmarks companies, using the linear interpolation technique in the intermediate years.

    c.2 Projection of the Real PNTs of the Interior

    In order to estimate the level of real losses for the interior of Roraima, the percentage of regulatory PNTs of 2016 (113.78%) was adopted as the percentage of Real PNT of 2015, as set forth in Homologation Resolution No. 2184/2016. This constant level was maintained until 2018.

    Starting in 2019, the projected trajectory for Amazonas Energia was adopted as the base, which presents the same levels of Real PNTs. However, it should be considered that the interior of the state of Roraima has a complexity index to combat PNT (0.181), which is half of that obtained for the AmE (0.364). In this way, we simulate that the trajectory of Real PNT in the interior falls 50% more than in the concession area of AmE between 2019 and 2023. Then, over the next ten years, the real PNT falls to match the levels of Boa Vista (Capital). From there, both the Capital and Interior areas have the same trajectory.

    Page 27


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017


    Based on the percentages of Real PNT’s and applying to the projected LV market (for the

    Interior and Capital), the value of Total Real PNT in MWh is recorded year by year. These values were summed and, finally, the percentage of the Real PNT was calculated on the LV market of every state of Roraima. The consolidated result of the projections prepared by Service B is reported in the following figure.


    2.3.3 ASSUMPTIONS USED AND POINTS OF DIVERGENCE OF SERVICES A AND B

    1. From what was observed, Service A did not consider the participation of CERR -Companhia Energética de Roraima in its loss projections, both PT and PNT. Thus, the results of Service A projections were quite different from those prepared by Service B, especially when observing the PNT starting points of both consultancies.

    2. Service A considered in its projections that the company does not have Losses in the Basic Network during the period from 2017 to 2048. Service B did not mention these losses.

    3. For Technical Losses (PT), Service A also projected them with the constant value of 7.04% of the energy injected during the concession. Service B initially used the levels established in Homologation Resolution No. 2184/2016. From 2028, Service B used as a parameter PTs by the voltage segment of compatible companies to

     Page 28


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

      model an efficient PT curve of the Roraima concessionaires, taking into account the tariff cycles. The reduction trajectory starts with PT of 7.62% in 2017, arriving in 2047 at the value of 7.04%.
    4.      Modeling of Service A Regulatory Non-Technical Losses (NTP) was organized in two stages. At first, from 2016 to 2026, regulatory levels describe a linear decrease reaching 2026 at 7.19%. It remained constant in 2027, and in later years, a gradual decrease to reach the level of 2.5% in 2032 and maintaining this level until the end of the concession.
      In Service B, the curve of Regulatory PNTs was performed based on 2 periods: 1) From 2017 to 2025, the levels of regulatory losses stipulated in RH 2.184/2016; and 2) From 2026 to 2047, the modeling addresses the reflection of the application of the detailed mechanisms in NT 106/2015-ANEEL in the capital Boa Vista for the entire state. The reduction trajectory starts with 24.49% in 2017, reaching 3.10% in 2047 on the LV Market.
    5.      The trajectory of real PNTs prepared by Service A starts from the value in 2016 of 9.44%, the curve reduces until 2047, when it reaches the value of 0.34% of the LV market. For Service B, in the years 2017 and 2018, the real PNT indices will remain constant at 23.10%, as of 2019, a new phase in the efficient combat against PNT will begin. The trajectory shows a marked reduction of losses in the first 5 years and then the curve gradually decreases, having as a reference the reductions of Real PNTs of SANTA MARIA and FORCEL companies in the modeling of Boa Vista Energia and AmE in the modeling of CERR. Thus, at the end of 2047, the projection of the real NTP index is 2.40%.

    2.4 DEFAULT AND UNRECOVERABLE REVENUES

    The Projected Unrecoverable Revenues and Default is the median of billed value that was not received until the base date of December 2016, the stabilization point of the unbilled revenue curve, from the 49th to the 60th month.

    The default assumption has great relevance in the Valuation analysis because it directly affects the collection, consequently, the company's cash flow, negatively impacting its Valuation.

    Analyzing the two Services, there are some divergences, which are listed below:

    Page 29


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Below are tables with the numbers of the default up to 2027 of the two Services:

        Table 22.  Default 2027 - Services A and B       
     
    Service B                       
    Unrecoverable Revenues                     
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    Residential  2.95%  2.66%  2.38%  2.09%  1.80%  1.52%  1.23%  0.94%  0.65%  0.37%  0.08% 
    COMMERCIAL  2.92%  2.63%  2.35%  2.06%  1.78%  1.49%  1.20%  0.92%  0.63%  0.35%  0.06% 
    Industrial  5.88%  5.29%  4.70%  4.12%  3.53%  2.94%  2.35%  1.76%  1.18%  0.59%  0.00% 
    Rural  6.21%  5.59%  4.97%  4.35%  3.73%  3.12%  2.50%  1.88%  1.26%  0.64%  0.02% 
    Government  2.95%  2.66%  2.36%  2.07%  1.77%  1.48%  1.18%  0.89%  0.59%  0.30%  0.00% 
    Public lighting  1.74%  1.57%  1.39%  1.22%  1.04%  0.87%  0.70%  0.52%  0.35%  0.17%  0.00% 
    Public Service  0.78%  0.70%  0.62%  0.55%  0.47%  0.39%  0.31%  0.23%  0.16%  0.08%  0.00% 
    Service A*                       
    Unrecoverable Revenues                     
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    Residential  1.39%  1.39%  1.39%  1.39%  1.39%  1.39%  1.39%  1.39%  1.39%  1.39%  1.39% 
    COMMERCIAL  1.10%  1.08%  1.06%  1.04%  0.92%  0.82%  0.78%  0.78%  0.78%  0.78%  0.78% 
    Industrial  2.73%  2.32%  1.97%  1.97%  1.69%  1.69%  1.69%  1.69%  1.69%  1.69%  1.69% 
    Rural  1.15%  1.15%  1.15%  1.15%  1.15%  1.15%  1.15%  1.15%  1.15%  1.15%  1.15% 
    Government  13.20%  12.16%  11.20%  10.32%  8.72%  7.37%  6.23%  5.27%  4.45%  3.76%  3.18% 
    Public lighting  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00% 
    Public Service  93.47%  89.51%  85.72%  82.09%  67.84%  56.07%  46.34%  38.30%  31.65%  26.16%  21.62% 

     

    In Service A there is not a specific year to reach the regulatory level, as performed by Service B, below. A target was established, above the regulatory level, and an annual

    Page 30


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    reduction curve based on the rate of change proposed by ANEEL in 4CRTP, 3CRTP and 2CRTP. With this assumption, the regulatory level is not reached during the thirty years of concession.

    Service A included in its preparation the "Others" Market. We did not include it because it would make it difficult to compare the tables.

    Table 23.  Capital and Interior Default, Dec/2016 - Service B 
     
      Unrecoverable Revenues  Capital  Interior 
      Residential  2.95%  1.39% 
      COMMERCIAL  2.92%  3.21% 
      Industrial  5.88%  1.12% 
      Rural  6.21%  1.15% 
      Public power  2.95%  14.33% 
      Public lighting  1.74%  0.00% 
      Public Services  0.78%  97.60% 

     

    Page 31


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    2.5      PMSO (PERSONNEL, MATERIAL, SERVICES AND OTHERS)
    2.5.1      PMSO INCURRED (2012 - 2016)

    For the analysis of the PMSO item contained in the reports of Services A and B, it is necessary in advance to look at the history of this item over the last few years. For the case, the PMSO was analyzed between the period from 2012 to 2016 (same period used by Service A). Below are the values for Boa Vista at the abovementioned period:

    Table 24.  PMSO Incurred from 2012 to 2016 (Boa Vista)   
              R$ Million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  56.4  66.1  67.4  72.7  80.2 
    PID  0.0  4.8  0.0  0.0  0.0 
    Material  1.1  1.3  1.1  1.7  1.4 
    Third-party services  15.7  16.9  17.3  14.7  16.8 
    Other  -2.1  3.1  4.2  3.3  1.5 
    Total  71.1  92.3  90.0  92.4  99.9 

     

    For the composition of each item above, the Financial Statements of the respective years were used as support material. In heading "O" of the PMSO, the following items were considered: Arrangement and Rent; Taxes; Miscellaneous indemnities; Court costs and fees; personal energy consumption; Inspection Fee and Others. Note that the Severance Incentive Plan (Plano de Incentivo ao Desligamento PID) was considered for the years 2013 and 2014.

    In order to show consistency in the comparison with the Assessment Reports presented by the Consultancies, the PMSO values were updated by the IPCA for December 2016.

    Table 25.  PMSO Incurred from 2012 - 2016 updated to Dec/16 (Boa Vista) 
                R$ Million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  75.5  83.7  80.2  78.4  80.2 
    PID  0.0  6.1  0.0  0.0  0.0 
    Material  1.5  1.7  1.3  1.8  1.4 
    Third-party services  21.1  21.4  20.5  15.9  16.8 
    Other  -2.8  4.0  5.0  3.5  1.5 
    Total  95.2  116.8  107.0  99.6  99.9 

     

    Page 32


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    In relation to regulatory operating costs, it is possible to observe the estimated values for Boa Vista in the years 2013 (last tariff review) until 2016, brought to the base date of December 2016.

    In relation to the comparison of the PMSO incurred with the Regulatory to Roraima, it is noticed that the discrepancy is much accentuated. The PMSO performed is well above the regulatory value for all years analyzed. This fact can be seen in the graph below:

    2.5.2      PMSO FOR SERVICE A
    -      Historical part:

    For the composition of the historical part, the consultancy, based on the accounting statements provided by the distribution company, made a grouping of similar accounts in terms of type of expenses and also made an adjustment of the values for the base date of December 2016. Below is the table containing the opening of the PMSO for Boa Vista:

    Page 33


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 26.  PMSO values calculated by the consultancy - Boa Vista (base date - 
    Dec/16) - R$ million
              R$ Million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  77.3  91.1  81.4  80.6  80.1 
     PID  0.0  0.0  0.0  0.0  0.0 
    Material  1.5  1.7  1.3  1.9  1.4 
    Third-party services  25.5  21.7  20.8  16.3  17.1 
     Other 9.3  6.1  4.9  7.6  1.7 
     Total  113.7  120.6  108.4  106.3  100.4 

     

    - Methodology used in Service A: (part common to all Distribution Companies)

    The analysis of the PMSO as well as the projection of the values was divided into two parts, namely:

    Part 1 - Characterization of Analytical Accounts (Divide in 5 phases):

  • Phase 1 - Analytical Account Clustering: 
     
      o     

    Inflation adjustment of the accounts for the base date of December 2016; 

      o      Clustering of accounts according to similarity of the type of expenditure;
  • Phase 2 - Analysis of the nature of the account into fixed and variable:
    o      Analysis of Fixed and Variable Accounts;
    o      The Accounts identified as fixed, when projected, were calculated as the average of the expenses of the sample period, excluding the outliers;
    o      For the projection of the variable accounts, an indicator of this account was calculated first as a function of Consumer Units, Network Expansion (km) and Consumption, in order to later carry out the projection. The outliers were also excluded.

    Page 34


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

        §      They are outside the established limits of normality of the sample, i.e.: 
     
         
  • Upper limit out =>(sample mean) + (sample standard deviation); Lower limit out => (sample mean) - (standard deviation of the sample). 
     
      o      After this procedure, the nature of the expenditure and its periodicity in the last 3 years were assessed in order to assess whether or not it was excluded. 
     
  • Phase 4 - Assessment of the best correlations for the variable accounts: 
     
      o      At this stage, a study was made to verify which variables were more related to the expenses of PMSO. It was concluded that Network Extension (km) and Consumption in MWh obtained higher ratio; 
     
  • Phase 5 - Identification of regulatory accounts (deemed to be recurrent and paid  by ANEEL):
         
      o      Consists of the classification of the accounts forming the PMSO according to the characterization defined by ANEEL for regulatory PMSO. Thus, for each component account of the analytical PMSO a classification between Effective PMSO (total), Regulatory PMSO and other PMSO (Effective PMSO - Regulatory PMSO) is provided.

    Part 2 - Projection of Effective and Regulatory PMSO Clustering (Divided into 4 phases):

  • Phase 1 - Projection of Expenditure for 2017:
      o      Considers a variation of expenditure in line with the company's history, taking into account the fixed or variable nature of the expenses composing the PMSO; 
     
      o      Variable Expenses: turned into indicators in which the denominator is the parameter with which the variable had the best correlation (it can be network km or consumption in MWh). 
     
        §      Example: Personnel = R$/km of network => item P will increase as the network km increases;

    Page 35


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    O Fixed expenses:

    Fixed expenses were calculated as the average of expenditure in the sample period, excluding outliers.

  • Phase 2 - Operating Costs Variation Rate: This phase is divided into three steps: 
     
      o      Step 1 - Definition of Operational Cost Benchmarks l; 
     
        §      For Northern Companies (CERON, Boa Vista Energia, Eletroacre and Amazonas Energia) => Celpa and Celtins; 
     
        §      For the Northeast Companies (Cepisa and Ceal) => Celpe, Cemar, Coelce, Energisa Paraíba, Energisa Sergipe, Coelba and Cosern. 
     
      o      Step 2 - Definition of Rate of Change and Target between 2018 and 2022: 
     
        §      DEA (Data Envelopment Analysis) database of efficient costing from 
     
         
  • to 4CRTP is considered; 
     
        §      The PMSO expense of each company is parameterized by the km of Network Extension; 
     
        §      Based on the indicators, the average of the group of each distribution company according to the Region is calculated; 
     
        §      After this step, proceed as follows: 
     
         
  • distribution companies whose indicator is greater than the group mean, adjust their cost in terms of indicator until the next Tariff Review in 2023. 
     
         
  • On the other hand, if the indicator is less than the mean of the group, the Distribution companies plot the indicator level between 2023 and 2027 as target, which is lower than the group mean. 
     
      o      Step 3 - Rate Variation Definition and Target between 2023 and 2027: 
     
        §      Consists of calculating the mean indicator of the private benchmarks of the groups;

    Page 36


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    § When reaching the R$/network km indicator of the cycles' target, the distribution companies stabilize parameterized operational cost indicators per consumer unit.

  • Phase 3 - Identification and Treatment of Accounts Linked to distribution company Improvements: 
     
      o      Consists of identifying accounts that influence the actual cost of PMSO, but vary according to the distribution company's operational improvements. 
     
  • Phase 4 - Projection of the Concession Period (2018 to 2047): Consists of estimating the PMSO spending levels of distribution companies over the concession period. 
     
      o      Based on the targets set previously, the distribution companies reach the operational cost benchmarks by the end of 2027; 
     
      o      Thereafter, the PMSO/network km indicator is stabilized, as a consequence, the absolute value of the PMSO varies in real terms by means of the variation of the Network km extension. 
     
  • Costs:

    The estimation of efficient operating costs was based on Annex I - Efficiency Intervals of the submodule 2.2 V2 of PRORET. Operating costs tend to align with their respective benchmarks in two moments, the first ending 2022 and the second in 2027.

    Calculation Method:

    - First moment => simple average of each of these limits for distribution companies according to their regional groups, which are:

    Page 37


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    - In the second moment => average of the private companies from the regional groups mentioned above.

    It should be underlined that, similarly to Technical Note 149/2017 - ANEEL and Technical Note 293/2017 - ANEEL define the operating costs in force from 2017 to 2023 for the entire State of Roraima (capital and interior), the intervals of efficient costs set out in this section shall only have any impact on the tariff after the Tariff Review of 2023.

    - Effective PMSO vs. Regulatory PMSO (Service A):

    Based on the values of the effective PMSO projection allocated above and with the values calculated for the regulatory PMSO, the following graph was obtained:

    Gráfico 13. Effective PMSO Projection vs. Regulatory PMSO


    Important points:

    - It is important to point out that for the composition of the effective PMSO graph, the expenses with the PDV were considered in the years 2018 and 2019, therefore, the values for the P item contained in table 30 are different from the values used to compose the graph. For information the values are: R$ 99 million against R$ 84 million (2018) and R$ 91 million against R$ 87.4 million (2019);

    Page 38


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    - Another point worth highlighting is that the values of Item Other for the years 2025, 2030, 2035, 2040, 2045, 2046 and 2047 contained in the table on page 174 of the assessment report, which were used to compose the graph above, of the report, are in disagreement with the values indicated in the projection, pages 79 and 80 of the document;

    - By means of the graph it can be observed that the projections of the regulatory PMSO are smaller than those of the effective PMSO. However, these projections tend to equalize from 2045;

    - The difference between the VPL of the two items, in real terms, discounted by the WACC rate of the model, considering in simplified form, a tax shield of 34%, results in a negative value of R$ 254.9 million, that is, for Eletrobras Roraima the PMSO is not valuable for the economic and financial appraisal;

    by IPCA between Nov./2017 and Nov./2016 and by IGP-M and X Factor after Nov./2017), according to paragraph 43 of NT 149/2017- ANEEL, and R$ 15.1 million earmarked for the interior of the State of Roraima (considering values of 2016, being adjusted by IPCA between Nov./2017 and Nov./2016 and by IGP-M and X Factor after Nov./2017), according to paragraph 53 of NT 293/2017- ANEEL.

    2.5.3      PMSO FOR SERVICE B:
    -      History:

    Service B did not present the rationale to compose its database from 2012 to 2016 as did Service A. The report shows the projections of P, M, S and O values directly for the period from 2017 to 2047.

    Page 39


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    Table 27.  Clusters composition and efficiencies - Boa Vista 
     
    Boa Vista
    Companies
    Efficiency
    Indicator of
    4CRTP efficiency 
    Efficiency
    Indicator 2011
    Efficiency
    Indicator
    2013 
    % (p.a.)
    CPFL Piratininga  100%  0.59  0.55  3.5% 
    COELBA  96%  0.39  0.50  -13.4% 
    Cemat  77%  0.48  0.51  -2.2% 
    Bragantina  69%  0.92  0.80  6.8% 
    Sulgipe  66%  0.75  0.67  5.4% 
    CEB  53%  1.12  1.08  1.7% 
    ENF  53%  1.43  1.38  1.6% 
    Boa Vista  23%  2.91  2.57  6.1% 

     

    Understanding the indicator: For example, Boa Vista in 2011 obtained an indicator of 2.91 and in 2013, this value was 2.57. This means that the company has improved a little from 2011 to 2013 as it needs 2.57 (2013) PMSO units to offer a set of services (weighted combination of Market, Network, Consumer, Non-technical Losses).

    Projection of the trajectory of the indicator:

    1.      Phase 1: 2017-2018 => No Efficiency gain
    2.      Phase 2: 2019-2020 => follows the rhythm of the utility with the greater gain of efficiency of the Cluster until reaching the median of the indicators of the group
    3.      Phase 3: 2021-2048 => Convergence for the mean of the Companies of the 1st Portion of the Group

    Thus, the Boa Vista efficiency curve as well as the projection for the regulatory PMSO are presented in the graph and table below:

    Page 40


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Gráfico 14. Boa Vista Efficiency Curve


    From the efficiency curve and the projected values for the products under consideration, it is possible to determine the estimated value of regulatory PMSO for each year. Therefore, the values in the table below are obtained:

    Table 28.  Boa Vista Regulatory PMSO 
     
    Period Regulatory PMSO
    (R$ million 2016) 
    Var % pa
    2016  100  - 
    2017  162  61.6% 
    2018  176  9.0% 
    2019  165  -6.3% 
    2020  156  -5.8% 
    2021  147  -5.5% 
    2022  140  -5.2% 
    Average 2023 - 2027  116.76  -3.5% 
    Average 2028 - 2032  126.06  1.5% 
    Average 2033 - 2037  173.27  6.6% 
    Average 2038 - 2042  238.91  6.6% 
    Average 2043 - 2047  328.18  6.6% 

     

    Page 41


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Although Service B shows an Efficiency curve that should be followed in the document, it is not evident the comparison with the effective PMSO values projected by the consultancy. Therefore, it is not possible to state when the company is able to reach the level of regulatory PMSO, as well as to estimate if this trajectory will generate or destroy value for the Company.

    In addition, it is important to take into account the issue of relaxation, since according to Technical Note No. 149/2017, which deals with the flexibility of some regulatory parameters, the new values for the Regulatory PMSO that go until the first ordinary tariff review of the new controller of the designated companies, are:

    Table 29.  Flexibility of the PMSO (Aneel Technical Note No. 149/2017) 
     
     
    Distribution
    Company
    Tariff Process
    2016
    Flexible
    Regulatory
    Level 
    Difference Difference
    (%)
     AME  349  627.2  278.4  80% 
    Boa Vista  38  69.1  31.5  84% 
     CEAL  272  336.7  65.0  24% 
    CEPISA  361  420.1  59.4  16% 
    CERON  234  279.9  45.8  20% 
    ELETROACRE  94  128.5  35.1  38% 
     CEA  68  124.8  56.9  84% 

     

    Page 42


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    2.5.4      FORECASTS AND COMPARISONS OF THE PMSO OF SERVICES A AND B
    -      Projections of Boa Vista Energia until 2047:

    Table 30. PMSO Projection Data (Service A)

                    R$ million 
    Item  2017  2018  2019  2020  2021  2022  2023  2024 
    Personnel  79.2  84.0  87.4  91.0  94.6  98.4  96.2  94.7 
    Material  1.5  1.7  1.8  1.9  2.1  2.2  2.3  2.5 
    Third Party Service  19.0  20.1  21.2  22.5  23.7  25.0  25.8  26.8 
    Other  5.8  6.4  6.7  7.0  7.3  7.6  7.8  8.1 
    Total PMSO  105.6  112.1  117.1  122.3  127.6  133.2  132.2  132.1 
    Item  2025  2026  2027  2028  2033  2038  2043  2047 
    Personnel  93.9  93.6  93.8  97.7  119.4  145.8  178.3  209.4 
    Material  2.6  2.8  3.0  3.2  4.5  6.3  8.6  10.9 
    Third Party Services  27.8  29.0  30.3  32.1  42.6  56.3  73.9  91.3 
    Others  8.4  8.7  9.0  9.4  11.9  14.9  18.8  22.5 
    Total PMSO  132.8  134.1  136.1  142.4  178.4  223.4  279.5  334.1 

     

    In addition, Service A estimated a value for the Voluntary Resignation Plan (PDV), defining the target public (employees with at least 20 years of company), the success rate, its cost (R$ 17.6 million) and the date of resignation of employees, June/18 and June/19. (However, the report includes a value of 0 for the Expenses with the PDV).

      Table 31.  PMSO Projection Data (Service B)     
                    R$ Million 
    Item  2017  2018  2019  2020  2021  2022  2023  2024 
    Personnel  98.8  107.7  93.7  80.0  71.6  65.9  60.9  57.0 
    Material  2.1  2.2  2.0  1.7  1.5  1.4  1.3  1.2 
    Third Party Service  112.0  122.1  106.2  90.7  81.2  74.7  69.1  64.6 
    Others  3.4  3.7  3.3  2.8  2.5  2.3  2.1  2.0 
    Total PMSO  216.3  235.8  205.0  175.2  156.8  144.3  133.3  124.8 
    Item  2025  2026  2027  2028  2033  2038  2043  2047 
    Personnel  53.5  50.2  47.0  50.2  69.3  95.6  131.3  169.8 
    Material  1.1  1.0  1.0  1.0  1.4  2.0  2.7  3.5 
    Third Party Service  60.6  56.9  53.3  56.9  78.6  108.4  148.8  192.5 
    Other  1.9  1.7  1.6  1.7  2.4  3.3  4.6  5.9 
    Total PMSO  117.1  109.8  103.0  110.0  151.8  209.3  287.4  371.7 

     

    The graph below shows the evolution of Service A PMSO compared to Service B.

    Page 43


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017


    From the graphical analysis, a somewhat atypical behavior can be perceived in relation to the projection of Service B. The larger gaps occur in the initial years and have a peak in the year 2018, when the value of the difference reaches R$ - 123,7 million (comparing Service A with Service B). From this point, Service B projection has a downward trajectory where it crosses Service A projection in the year 2023 and goes to its minimum PMSO value in the year 2027 (R$ 103 million). From then on, the projections follow the same trend, ending the period with a difference of R$ - 37.5 million, again (comparing Service A with Service B). It is important to remember that this difference in values has no significant impact on valuation if the company's PMSO is within the regulatory level.

    Therefore, in order to further explore the reasons that led to this discrepancy in the projection, we will show how the evolution of each item behaves until the end of 2047.

    Page 44


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Personnel and Material


    Again, the same behavior observed in the PMSO can be observed in items P and M. For both, the projection of Service B starts higher in relation to Service A. After it has a downward trajectory, it reaches a minimum value and from this point, follow the same trend until the end of the projection when they reach a difference of R$ 39.7 million for item Personnel and R$ 7.4 million for item Materials.

    Page 45


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Third Party Service and Other

    Gráfico 17. Projection of Third Party and Other Services items referring to Services A and B

    Undoubtedly, the item that shapes the projection trend of Service B projection is the Third Party Service. Looking at the graph above, this behavior can be seen with ease. Whilst the projection of Service A for this Item is practically linear. Finally, as observed in the other Companies, for Item Other, the projections are different from the beginning of the

    Page 46


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    projection period, however in the case of Boa Vista, Projection of Service A has values higher than the projection of Service B, a fact that was not observed in the other companies.

    found, it is noticed that there is a very marked difference in relation to the projections. This fact may be due to the basis that each Consultancy used to make the PMSO projection and the projection model considered. In relation to the first item mentioned above, only Service A presented the rationale for the initial composition of the PMSO.

    expenses, it is concluded that Service B adopted a much more projection than that of Service A. Highlighting again, this difference in values has no significant impact on valuation if the company's PMSO is within the regulatory level.

      Table 32.  Comparison between the Services   
                R$ 
                million 
      Service A    Service B  % 
    Company      %       
      2017  2047    2017  2047  2017 
    Eletroacre  112.5  684.9  509%  133.5  327.6  145% 
    CEAL  288.8  1,126.5  290%  296.7  686.0  131% 
    Amazonas  550.1  1,896.4  245%  600.2  1,267.4  111% 
    Cepisa  383.7  1,757.7  358%  452.6  1,122.8  148% 
    CERON  340.1  1,534.8  351%  373.3  866.6  132% 
    Boa Vista  105.6  334.1  216%  216.3  371.7  72% 
    Total  1,780.8  7,334.4  312%  2,072.5  4,642.1  124% 

     

    - Finally, regarding the documents presented by the consultants, Service A detailed the methodology much more than Service B, and was more coherent when presenting information.

    Page 47


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.6 INVESTMENT

    Both services contracted for economic and financial assessment adopted different assumptions for the projection of the investment notably from the period 2023 to 2047, since both adopted the Five-Year Plan for the period from 2018 to 2022 of the company. This plan was presented in Product 7 "Boa Vista Operational Technical Assessment Report (Capital and Interior)" and was elaborated based on the "Expansion Plan of the Electrical System (2014-2023 period)", the "Development Plan for Distribution - PDD for the year 2017" and the Temporary Rendering Plan for the interior of the state of Roraima presented by the company to Aneel in mid-March 2017. In addition, information obtained through the "Field Visit Report" (Annex I of the Technical and Operational Assessment Report) and the incremental investment needs projected by the company's planning area, obtained during face-to-face meetings and conference call carried out. The only caveat made for this period is that Service A for the investment in “Luz para Todos” program considered only 5% of the value since 95% is subsidized by CDE, and therefore treated as Special Obligations.

    In relation to the other five-year period, the main differences are due to the fact that Service A included, in addition to the projections made for implementation and renovation projected from the Five-Year Plan, other analyzes to define the investment in maintenance and replacement of partially and totally depreciated assets. While Service B was restricted only to the first analysis.

    For the projection from 2023 of investments in implementation and renewal, Service A adopted as a basic assumption for the expansion of HV, AV and LV as well as for the improvements the mean of the last years contemplated in the plans of the companies and this value was kept constant until the end of the concession period1. Service B, for electric investments (expansion and improvement), also estimated the second five-year period of the concession by the mean of the period from 2018 to 2022, however, it eliminated values outside the range between the mean and +/- a standard deviation. Thus, in order to obtain the amount of investment needed for the five-year period, the investments of

    1 The expansion in HV and AV/LV was available for information until 2026 by means of the Distribution Development Plan 2017 (PDD 2017) and then the average of the investments from 2018 to 2026 was adopted for the other years. As for the investment in improvements, the average of expenditures from 2018 to 2022 was used.

    Page 48


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    the years considered typical were added up and multiplied by factor 5/No. typical years. For the other five-year period, they adopted as base the previous five-year period with the caveat that for the expansion in AT, it is considered that every 10 years the need for investments doubles in relation to the previous five-year period. For non-electric investments (infrastructure), it was adopted as a premise that every five years, 25% of what was foreseen in the first five years is invested. Service A, however, did not project them.

    For renovation investments (maintenance), since 2023, service A considered that investments of this type are equivalent to the value of the depreciation of Boa Vista and CERR assets. To define this value, the depreciation reference of ANEEL's Normative Resolution 674/20152, was used for the current asset base plus the depreciation of the new investments of the company, calculated from the average depreciation of the asset base of the Assets Control Report. These investments were accumulated and concentrated in the two years before each tariff review according to the proportion of the reference companies3. Service B, however, first calculated the share of the amount allocated to the investment in renovation within the period of the 2018-2022 Five Year Plan. Subsequently, it applied this percentage for all other years.

    For investments in replacement, only service A estimated them. For the capital, based on the assets base of 2016, CERR was used as the basis for the Fixed Assets of the 2016 regulatory financial statement, which were added to the estimated investments for implementation and renovation from 2017 to 2022, as well as the inflation adjustments and application of the depreciation of each item, based on Aneel Normative Resolution 674/2015, both for the assets base of 2016 and for the new investments projected up to 2022. In this way, the company's asset base (capital + interior) of 2022 was formed. Therefore, the indicator Market Use Value (VMU)/New Value of Replacement (VNR) was calculated for the company and for the benchmark 4companies so that the distribution company under review reaches the same percentage of the private reference distributed

    2 Resolution that approved the review of the Assets control manual of the Electric Sector - MCPSE.

    3 Cemar, Coelce, Coelba and Celpe.

    4 Capital - Cemar, Energisa Sergipe and Energisa Paraíba.

    Interior Amazonas, Acre, Rondônia and Roraima

    Page 49


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    between 2023 and 2027 by the average proportion of investment of the reference companies5.

    Given the premises and their differences, we find the following values for the two consultancies for Boa Vista:


    The difference between the consortiums is R$ 741 million, notably in the item Renovation (maintenance) in which Service A estimates that from the 2nd five-year period the investments are equivalent to the value of the depreciation of the assets and Service B calculates them from their participation in the Five-Year Plan from 2018 to 2022. In addition, service A, in contrast to B, estimated the replacement investment in the value of R$ 41.2 million in order for Boa Vista (capital + interior) to reach the same ratio as the VMU/VNR indicator of the companies used as reference. Note that this difference in the

    5 Cemar, Coelce, Coelba and Celpe.

    Page 50


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    volume of projected investments does not have a significant impact on valuation, since the investment increases the Company's Remuneration Base, and consequently, it is contemplated in a higher tariff.

    By way of comparison, the value of the investment made from 2012 to 2016 at December 2016 values for Boa Vista capital:

    Table 34.  Comparison of Investment carried out from 2012 to 2016, at values of Dec / 2016
     
                  R$ Million 
                  Five-year period 
    Investment  2012  2013  2014  2015  2016   2012-2016 
    Values updated (Dec/16)  36  23  25  25  29  137 

     

    2.7 REGULATORY REMUNERATION BASE

    Both Services used the data contained in the valuation reports of the assets as reference to the date of February 28, 2017 to prepare the projection of the Regulatory Remuneration Base (BRR). In these reports, both the closed BRR and the incremental BRR from 3CRTP to 4CRTP were appraisal. In the case of Boa Vista, the following reports were used:

  • Closed BRR of Boa Vista (capital) in Feb/17 - corresponds to the asset base unitized
      until      Feb/17 for Boa Vista (capital). The following assessment reports were used as
      source of information:
      o      Assessment Report prepared by Levin (delivered on 06/30/2017) for Boa Vista, which considered the reopening of the unitized asset base until the base date of 3CRTP and the assessment of the unitized asset base between 3CRTP and Feb/17 - Integral report;
      o      Assessment Report prepared by Levin (delivered on 02/28/2017) for Boa Vista, which considered the asset base to 3CRTP in a closed (closed) manner and the assessment of the unitized asset base between 3CRTP and Feb/17 - Incremental report.
  • Incremental BRR: considers the unitizations occurred after Feb/17 and new investments planned over 30 years of concession.

    Page 51


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Both service A and B adopted different projection criteria for the asset base unitized through Feb / 17 (closed BRR - starting point for 4CRTP) and for unitization occurring after Feb/17 and new investments (incremental BRR).

    Regarding the starting point for 4CRTP, both Services use the full appraisal report (delivered on 06/30/2017), however, they present different criteria for the reversal of Fixed Assets in Use (AIS).

    For service A, the reversal is treated in two ways: i) reversal of the revised closed base (Closed Base of the Full Report - Closed Base of the Incremental Report) in the value set at 5.73% for the companies from the North and 5.01% for companies from the Northeast, based on the average adjustment (reversal) of the group of distribution companies appraised in the 3CRTP by region, i.e., this reversal is made only in the AIS of the difference found in the opening of the closed base compared to that approved by Aneel in 3CRTP adjusted for inflation; and ii) The incremental base between 3CRTP and 4CRTP -the lowest value was applied between the average adjustment of the group of distribution companies appraised in 3CRTP by region and that verified by the company, without evidencing in the Appraisal Report the invested amount.

    In contrast to Service A, Service B used only a disallowance methodology to reach the Remuneration Base of Feb./17 for the Distribution Company. The disallowances were between the difference in AIS of the two appraisal reports of the Closed Base and the Incremental Base. This difference would be the opening of the Closed Base. That is, the reversal for Service B was only at the opening of the Closed Base and occurred in the AIS. The percentage of disallowances was calculated from the relation between the final AIS (with reversals) and the initial AIS (without reversals) of the incremental BRR of the 3RTP of the distribution companies of the Eletrobras group, obtaining a mean of 9.2%. If the company's reversal is lower than this mean, the average percentage of disallowances is considered (9.2%). Otherwise, the mean between the percentage of the company and the maximum percentage of disallowances (20.1%) is used. In the case of Boa Vista, it falls within the second criterion, since it presented a 9.7% disallowance of the AIS in the 3CRTP, then, adopting a disallowance of 14.9% for the Closed Base. This disallowance percentage is applied only in the difference between the AIS of the two reports and added to the value of the base of the incremental report. That is, the disallowance of 15% is not applied in the total AIS of the Complete Base, it is calculated only for the increment of the closed base.

    Page 52


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    In the table below, the information for the closed base of the capital, in the position of February 2017, is presented for both services.

    Table 35.  Closed base in the position of February 2017 - Services A and B 
     
          R$ Million 
        Closed Base 
      Description     
      Service A    Service B 
    (1) Property, Plant And Equipment in Service (New Replacement Value)  440  422 
    (2) Index of Full Utilization  -  - 
    (3) Gross Special Obligations  46  46 
    (4) Fully Depreciated Assets  86  86 
    (5) Gross Compensation Reference = (1) - (2) -(3) - (4)  308  291 
    (6) Accumulated Depreciation  227  227 
    (7) Net AIS (Market Value In Use)  213  196 
    (8) Index of Depreciated Utilization  -  - 
    (9) Compensation Reference Value (VBR)  213  196 
    (10) Warehouse in Operation  3  3 
    (11) Deferred Assets  -  - 
    (12) Special Net Obligations  34  34 
    (13) Lands and Easements  1  1 
     
    (14) Base of Total Net Compensation = (1) - (6) - (8) + (10) + (11) - (12) +  183  166 
    (13)       
    (15) RGR PLPT Balance  -  - 
    (16) RGR Balance Other Investments  -  - 
    (17) Depreciation Rate  4%  3% 
    (18) Share of Regulatory Reintegration = (5) * (17)  12  11 
    (19) Actual WACC before Taxes  11%  ND 
    (20) RGR PLPT Rate  1%  ND 
    (21) RGR Rate of Other Investments  4%  ND 
    (22) Capital Compensation (RC) = (15) * (20) + (16) * (21) + [(14) - (15) -     
    (16)] * (19)  21  ND 

     

    For the pricing of the Starting BRR of the interior, both Service A and Service B used the Assessment Report surveyed by the independent audit company, included in NT no. 293/2017-SGT/ANEEL on the base date of November 2016. This way, in order for the capital and interior starting base to be on the same base date, the CERR’s base was inflation adjusted for February 2017. Additionally, an adjustment to the depreciation on historical level of Boa Vista was performed. After such adjustments, BRR of capital and interior are summed, and the increments of assets in the years subsequent to the distribution company’s starting BRR happen according to the Investment Projection.

    Thus, the Closed BRR of the interior in Nov/16 and Feb/17 was composed by the values presented below:

    Page 53


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 36.  Closed base of interior, Nov/2017 - Services A and B 
        Closed Base 
      Description     
        Nov-16  Feb-17 
    (1) Fixed Assets in Service (New Replacement Value)  327  332 
    (2) Total Application Rate  -  - 
    (3) Special gross obligations  216  219 
    (4) Fully depreciated assets  -  - 
    (5) Gross Remuneration Basis = (1) - (2) - (3) - (4)  111  112 
    (6) Accumulated depreciation  82  84 
    (7) AIS Net (Market value in use)  246  248 
    (8) Depreciated application rate  -  - 
    (9) Remuneration Base Value (VBR)  246  248 
    (10) Warehouse in operation  5  5 
    (11) Deferred asset  -  - 
    (12) Net special obligations  189  191 
    (13) Land and easements  11  11 
    (14) Total Net Remuneration Basis = (1) - (6) - (8) + (10) + (11) - (12) + (13)  73  73 
    (15) RGR PLPT Balance  -  - 
    (16) RGR Balance Other investments  -  - 
    (17) Depreciation rate  3%  3% 
    (18) Regulatory Reintegration Quota = (5) * (17)  4  4 
    (19) Real WACC before taxes  12%  12% 
    (20) RGR PLPT rate  1%  1% 
    (21) RGR rate other investments  4%  4% 
    (22) Capital Remuneration (RC) = (15) * (20) + (16) * (21) + [(14) - (15) -     
    (16)] * (19)    9  9 

     

    For the years subsequent to the 4CRTP Starting Base, both Services apply inflation adjustments according to the IPCA, additions and write-offs of the assets that were considered in the appraisal until the end of the concession period.

    However, it should be noted that both Services did not incorporate the positions of Fixed

    Assets in use into the Remuneration Base, since, according to the services, the assessors’ report of the Remuneration Base does not take into account the specific assessment of the condition and characteristic of these assets concerning (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the reassessment process (such as the calculation of VNR, depreciation, utilization rate and valuation of losses). In regards to the case of Boa Vista, it should be noted that the Full Assessment Report “Eletrobras

    Page 54


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Distribuição Roraima Project no. 3012-17862,” performed by Levin, includes as the position of February 2017 the balance of R$ 57.4 million for the Fixed Assets in Use.

    In addition, Service B divides into two groups the assumptions for the projection of BRR movements. In the first group, the closed BRR is under the IPC inflation adjustment and the discount of the depreciation and write-off of the assets that went into operation until February 17 and in a second group the incremental BRR (new investments) is through the additions of assets projected after Feb/2017, less the losses related to these assets.

    Thus, as from 2018, service A adjusts for inflation, on a monthly basis, according to the IPCA, the Fixed Assets in Use, Warehousing in operation and land and easements and increases them according to their projection of the investment. The gross special obligations are also adjusted according to the IPCA and the increase in projected special obligations. On said incremental base, the average adjustment of privatized benchmark distribution companies was applied as reversal by region obtained through the 3CRTP Incremental Base. For the companies from the Northeast, Celpe, Energisa Paraíba and Energisa Sergipe were considered, resulting in a reversal value of 2.09%. For the North companies, the disallowance value was 2.83%, using Celpa and Cemar as benchmark. In relation to the depreciation rate, the same verified in 3CRTP was maintained for each distribution company assessed. In relation to asset write-offs, the ratio of fully depreciated assets to fixed assets in use observed in the 4RTP BRR resulted in an equilibrium rate calculated as a simple mean between Eletrobras' companies. Thus, starting from the company's ratio, it varies at a constant monthly rate until reaching that equilibrium rate of 12.54% at the end of the concession.

    Service B, as mentioned earlier, operated BRR with different premises for the base that served as a starting point and the one that incorporates the new investments. The transaction of the closed BRR was made to the capital and to the interior separately by the monetary restatement by the IPCA and by the discount of the depreciations and write-offs. The rate of depreciation for capital and interior6, unlike service A, is not constant, the rate falls from the average level calculated in Feb/17, since the assets are being written-off. In relation to asset write-offs, the average write-off/AIS ratio was applied to the AIS of each month based on the information of the Assessment Report prepared by Levin (delivered on 02/28/2017) in the period between 3CRTP and Feb/17 for the Capital. And, for Interior,

    6 In order to estimate the average depreciation rate of the interior, the variation of the depreciation rate of the Closed BRR of Boa Vista (capital) was applied at the rate estimated by Aneel in the interior in Nov/14

    Page 55


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    the average ratio of write-offs in relation to the AIS of Boa Vista capital was used for the same period in the interior AIS.

    The incremental BRR transaction (new investments) was made based jointly for the capital and interior from the projection of the investments rejected in 5%. The depreciation rate used was obtained by the weighted mean of the investments of the Five-Year Plan 2018 to 2022 per group of assets and their respective depreciation rates, obtaining 3.82% of the AIS. This rate was considered constant throughout the concession. Regarding the write-offs, until Feb/27, they were considered null in view of the reduced probability of assets being written off in the first 10 years. After this period, 1/3 of the BRR Closed/AIS write-off ratio was applied in a staggered manner every 5 year until reaching the value of the integral ratio as of Mar/42.

    Therefore, given the differences in the investment projections, the disallowance, depreciation and write-off criteria and treatment of the interior in the formation of the Closed BRR, we obtained at the end of the concession (Feb/48) the following values of Net Remuneration Basis for Boa Vista (capital + interior) according to the projections of the two services:

      Table 37.  Net Remuneration Base - Services A and B 
            R$ Million 
      Description    Service A  Service B 
      Net Remuneration Base (Feb./48)  4,418  898 
     
     
     
    2.8  X FACTOR       

     

    The primary goal of X Factor is to ensure that the balance established in the tariff review between efficient revenues and expenses is maintained in subsequent tariff repositioning. This occurs upon transfer to the consumer of potential productivity gains from the distribution segment.

    The approach adopted by Aneel adds to X Factor a mechanism to encourage the improvement of the technical and commercial quality of the services provided to the consumer. A component that seeks to establish the gradual transition of the operating costs of the utility towards its efficient costs is also inserted.

    Page 56


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    X Factor adjusts the value of Portion B, thus affecting the tariff and, consequently, the revenue and the collection of the Distribution Company. Therefore, the factor X value inversely impacts the value of the VPL (the higher x factor, the lower the revenue and the VPL).

    X factor is composed of the sum of three components:

    Page 57


     

     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The table below shows X factor and its components calculated by Services A and B in years of tariff reviews7:

    Table 38.  X Factor and its components in year of RTP - Services A and B 
      Service A         
        Fator X  Pd  T  Q 
      2019  -0.27%  0.00%  0.00%  -0.27% 
      2023  -0.93%  1.01%  -1.92%  -0.02% 
      2028  -0.94%  1.00%  -1.59%  -0.35% 
      2033  -0.67%  1.00%  -1.67%  0.00% 
      2038  0.87%  1.00%  -0.13%  0.00% 
      2043  1.00%  1.00%  0.00%  0.00% 
      2048*  1.00%  1.00%  0.00%  0.00% 
     
      Service B         
        Fator X  Pd  T  Q 
      2019  0.21%  0.00%  0.00%  0.21% 
      2023  3.87%  1.43%  2.97%  -0.53% 
      2028  1.14%  1.39%  0.00%  -0.25% 
      2033  0.83%  1.38%  -0.34%  -0.21% 
      2038  0.51%  1.38%  -0.64%  -0.23% 
      2043  0.50%  1.38%  -0.70%  -0.18% 
      2047*  0.49%  1.38%  -0.70%  -0.19% 

     

    As can be seen in the table above, the estimated values for X factor are quite divergent, especially for the Q and T components, affecting, as a whole, X factor as a whole.

    Table 39. X Factor Differences between Services A and B

    Service A - Service B

     

      Fator X  Pd  T Q
    2019  -0.48%  0.00%  0.00%  -0.48% 
    2023  -4.80%  -0.42%  -4.89%  0.51% 
    2028  -20.80%  -0.39%  -1.59%  -0.10% 
    2033  -1.50%  -0.38%  -1.33%  0.21% 
    2038  0.36%  -0.38%  0.51%  0.23% 
    2043  0.50%  -0.38%  0.70%  0.18% 

     

    7 It should be noted that Service B provided the value of the component Q for all thirty years of concession, while Service A presented the values only in the years of tariff reviews.

    Page 58


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    1 - Component Pd - Productivity gains from distribution activity; 
    The formula of Component Pd: 
    2Q (V) = 26( + 0,14 × (”/J (V) ”MWh average) 0,04 × (”7%(V) ”7% average) 

     

    PTF: average productivity of the distribution segment, 1.53% pa

    MWh (i): Utility's average annual market variation i, between the previous tariff review and the tariff review being processed;

    Average MWh: Average annual variation of company distributions' market, of 4.65% p.a.;

    The main highlights on accounting differences are:

    2 - Component Q - Technical and commercial quality of the service provided to the consumer

    Page 59


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017


      Table 40.  Methodology used - Service A   
     
             
       3rd CRTP    New methodology   
      Methodology         
      Apr/15 to  Apr/16 to  Apr/17 to  Apr/18 to  Apr/19 to 
    Indicator           
      Mar/16  Mar/17  Mar/18  Mar/19  Mar/20 
    DEC  50%  30%  37.5%  45%  50% 
    FEC  50%  30%  30.0%  47.  20% 
    INS      0.8%  2%  4% 
    ICO      0.4%  1%  3% 
    IAb      0.4%  1%  3% 
    FER      3.0%  7%  10% 
    IASC      3.0%  7%  10% 
    Total  100%  60%  75%  90%  100% 

     

    the difference

    ·

    The DEC, FEC variation and other quality indicators, as well as in treatment and projections will be better explored in the item "Quality Indicators" of this report. It should be noted that Component Q of X Factor

    Page 60


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    varied considerably in all the years available in the comparison between the values calculated by Services A and B.

    3 - Component T - Trajectory of operating costs

    The T component of X factor is applied to Portion B so that the tariff increases or decreases for the recognized regulatory operating costs to reach the efficiency level established by Aneel.

    In defining efficient operating costs, the costs used by the distribution companies, the efficient level of costs and the characteristics of the concession area are observed. Aneel uses a model whose purpose is to define a feasible production/cost boundary as a way of identifying the level of efficiency of companies.

    In a simplified way, the identification of the efficient level of costs is obtained by comparing the distribution companies by means of a benchmarking method that takes into account the attributes of each utility. From these constraints, a target of efficient operational costs is established to be reached throughout the tariff cycle. At the time of the tariff review, the regulatory goal is compared to the operational costs coverage present in the utility's tariff, called operating costs revenue. From the difference between the regulatory goal and the operating costs revenue, a regulatory path will be calculated. Part of the difference will be incorporated at the time of the tariff review and the remaining portion will be considered for the purpose of calculating the T component of X factor.

    Services A and B, following Aneel's Technical Note No. 149 / 2017, used the value zero for T component in the period between the date of execution of the contract and the first common tariff review, that is, from 2018 to 2022. From 2023, Services A and B found very divergent values for T component of X factor.

    2.9 WACC

    Services A and B adopted the method of Capital Asset Pricing Model - CAPM to estimate the return that the shareholders expect to arrange with the capital employed in the business, for discounting the current value of the distribution company’s cash flow. It should be noted that this method is well-established and widely used, either by the market at large, and by Eletrobras.

    Page 61


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    In the calculation of the Weighted Average Cost of Capital WACC, Services A and B applied different approaches of the CAPM method, thus allowing quite different results.

    Below are the main differences found in the calculation of WACC by both Services.

    Capital Structure

    Beta                         
     
    ·  Service A: used comparable domestic companies. Unlevered Beta: 0.68;   
    ·  Service B: considered comparable domestic companies, with a larger sample. 
      Unlevered Beta: 0.43.                 
     
    Ke                         
     
    ·  Service A: applied the CAPM method and a “foreign exchange risk” of 4.17% at 
      nominal Ke in Dollars to achieve a nominal Ke in Brazilian Reais. Therefore, this 
      “foreign exchange risk,” which derives from the “Future DI” rate and “Fraction of 
      exchange rate coupon,” works as a differential of inflation. To achieve such rate in 
      real terms, it was deflated by IPCA (4.23%);           
    ·  Service B: applied the CAPM method.             
     
    The table below shows the result for Ke under real terms:         
     
        Table 41.  Profit or Loss of Real Ke Services A and B     
    Actual Ke  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  7.95%  7.81%  7,53%  7.21%  7,02%  6.85%  6.70%  6.55%  6.43%  6.31%  6.31% 
    Service B  11.73%  10.16%  10.18%  10.42%  10.42%  10.42%  10.42%  10.42%  10.84%  10.84%  10.79% 

     

    Kd

    Page 62


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    the amount of debt of each company, achieving an average weighted cost, which

    was invested in a projection for CDI (source: Bradesco).         
    · Service B: calculated Kd by the sum of SELIC and a spread of 0.64%.     
     
    The table below shows the Kd result in real terms after taxes.         
     
      Table 42.  Kd in real terms after taxes Services A and B     
     Actual  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Kd                       
    Service A  4.25%  4.09%  4,27%  4.19%  4.19%  4.19%  4.19%  4.19%  3.25%  3.25%  3.25% 
    Service B  4.44%  3,34%  3.36%  3.29%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25% 

     

    WACC

    Based on the values of Beta, Ke and Kd above, Services A and B calculated WACC in real terms after taxes, according to the results included in the table below.

      Table 43.  WACC Services A and B       
    WACC  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  8.09%  7.20%  7.30%  7.39%  7.39%  7.39%  7.39%  7.39%  7.60%  7.60%  7.58% 
    Service B  5.29%  4.80%  4.89%  4.87%  4.93%  5.00%  5.06%  5.12%  5.19%  5.26%  5.26% 

     

    Out of the results stated above, it should be noted that the factor with a higher impact on the rate presented by Service A was the use of the foreign exchange risk (4.17%) in a heterogeneous manner to the market practices, as a differential of inflation, as this is applied on a Ke (nominal in USD), in a scenario where the CPI US inflation is 2.30% and IPCA, 4.23%. The result of this is WACC, in real terms, which is higher compared to the number obtained in calculation by Service B.

    3. THE RESULT OF ECONOMIC AND FINANCIAL ASSESSMENTS AND

    BY MULTIPLES

    The assessment of Service A calculated enterprise value of R$ 565,617,124.98 for Boa Vista Energia, considering, as already mentioned, the granting of a new concession. Discounting debts, liabilities with suppliers and contingencies, Boa Vista Energia equity value calculated by Service A is negative R$ 328,272,159.71, as shown in the table below. To obtain these values, the consultancy used the WACC of 7.57%.

    Page 63


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    Table 44.  Enterprise value - Service A   
        Free Cash Flow to Firm (FCFF) 
    Base Date Enterprise Value    Feb/2018  Dec/2016 
    Adjustment on Enterprise Value    -  IPCA 
    Enterprise Value (BRL)  A  672,136,279.60  638,663,860.21 
    Cash (R$)  B  35,597,051.18  35,597,051.18 
    Gross Debt (BRL)  C  (924,445,234.99)  (924,445,234.99) 
    Net Debt (R$)  B+C  (888,848,183,81)  (888,848,183.81) 
    Tax, labor and social security contingencies (R$)  D  NQ  NQ 
    Legal Contingencies (BRL)  E  851,632.45  851,632.45 
    Actuarial Adjustments (R$)  F  449,282.02  449,282.02 
    Environmental Adequations (R$)  G  (6,342,015.35)  (6,342,015.35) 
    Interior BRR Payment (R$)  H  (73,046,735.23)  (73,046,735.23) 
    Valuation (R$)  A+B+C+D+E+F+G+H  (294,799,740.32)  (328,272,159.71) 

     

    Source: Assessment Report of Eletrobras Distribuição Roraima, Service A.

    Service B calculated the enterprise value of R$ 611,716,317.51 for Boa Vista Energia, in view of the granting of a new concession. Discounting debts, liabilities with suppliers and contingencies, Boa Vista Energia equity value calculated by Service B is negative R$ 282,172,967.18, as shown in the table below. In order to obtain these values, Service B used rolling WACC, which considers a movable (leveraged) capital structure.

    Table 45. Enterprise value - Service B
    Result    Dec-16 
    Operation VPL    634,524,170.81 
    Terminal Value VPL    50,452,807.62 
    (-)CERR Indemnity    -73,260,660.92 
    Enterprise Value    611,716,317.51 
    (-) Net Debt    -888,848,183.81 
    (-) Contingencies    1,300,914.47 
    (-) Enviromental Adequations    -6,342,015.35 
    Equity Value    -282,172,967.18 

    Source: Product 10: Boa Vista Conclusive Economic and Financial Assessment Report, Service B.

    Page 64


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 46.  Rolling WACC - Service B         
     
    Parameters  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  > 2026 
    Risk Free Rate (Rf)  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78% 
    Country Risk  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53% 
    Market Premium  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90% 
    Unlevered Beta (²u)  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43 
    % Debt (Kd) - [D/(D+E)]  54.0%  50.7%  47.3%  44.0%  40.6%  37.3%  33.9%  30.6%  27.3%  23.9%  23.9% 
    % Equity (Ke) - [E/(D+E)]  46.0%  49.3%  52.7%  56.0%  59.4%  62.7%  66.1%  69.4%  72.7%  76.1%  76.1% 
    Taxes - IRPJ / CSLL  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0% 
    Leveraged Beta (²L)  0.76  0.72  0.68  0.65  0.62  0.59  0.57  0.55  0.53  0.52  0.52 
    Size Premium  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00% 
    Inflation - Brazil  4.33%  4.43%  4.36%  4.31%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23% 
    Inflation - USA  2.40%  2.30%  2.30%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40% 
    Cost of Equity  12.62%  12.55% 

    12.22% 

    11.83%  11.55%   11.37%   11.21%   11.06%  10.93%  10.81%  10.81% 
    Selic  10.71%  9.07%  9.03%  8.87%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73% 
    Spread  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64% 
    Debt Cost  11.35%  9.71%  9.67%  9.51%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37% 
    Debt Cost - Taxes Net  7.49%  6.41%  6.38%  6.28%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18% 
    WACC in R$ (Post-Tax)  9.85%  9.44%  9.46%  9.39%  9.37%  9.44%  9.50%  9.57%  9.64%  9.71%  9.71% 

     

    Source: Product 10: Boa Vista Conclusive Economic and Financial Assessment Report, Service B.

    With respect to the multi-market valuation, Service A, in its Assessment Report of Eletrobras Distribuição Roraima, attached to this Appendix, obtained Boa Vista enterprise value equal to R$ 638.66 million, which represents a difference of 20% in relation to the value of the economic and financial assessment.

    Service B, in Product 10: Boa Vista Energia Conclusive Economic and Financial Assessment Report, attached to this Appendix, did not obtain the value of enterprise value, but performed comparative analyzes of different implicit multiples to the result of the Discounted Cash Flow assessment.

    3.1 AFAC

    The study presented considers that all Advances for Future Capital Increase - AFAC’s made by Eletrobras in Boa Vista Energia and which are accounted for will be converted into debt or loan. On October 17, 2017, Boa Vista Energia had R$ 89.1 million in AFAC accounted for. The AFAC contracts already provide for the conditions to be considered in case of conversion into debt. Accordingly, Eletrobras must convert the AFACs into debt, according to a loan agreement executed under the following conditions:

    Page 65


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

    4.      CONTINGENCIES
    4.1      LEGAL CONTINGENCIES

    Legal due diligence of Boa Vista Energia, carried out by Service B and contained in Product 4: Due Diligence Legal Report - Boa Vista, pointed out some critical points, although such points do not impede the process of privatization of the distribution company, the new investor should be aware of them so that it can take the applicable measures.

    4.2 ACCOUNTING CONTINGENCIES

    ANEEL carried out inspections at the distribution companies of Eletrobras - AmD, Eletroacre, Ceron and Boa Vista Energia, regarding the onlendings of CCC's funds, from June 2009 to June 2016, which aims to cover generation costs in current regulations, namely: Law 12.111/2009, Decree No. 7.246/2010, ANEEL Normative Resolution 427/2011 and Normative Resolution ANEEL No. 597/2013.

    The inspections pointed out that the management of CCC, at the time in charge of Eletrobras, was not applying a series of rules set forth in ANEEL Normative Resolution 427/2011 for the monthly calculation of the financial criteria that involve the total cost of generation of the beneficiaries.

    In the case of Boa Vista Energia, the preliminary decisions of ANEEL are favorable to the distribution company, since the Agency, according to Technical Note 177/2017-SFF-SFG-SRG/ANEEL, of 10/09/2017, understands that Boa Vista Energia has to receive from CCC R$ 5.7 million, at June/2017. In addition, considering that in the Balance Sheet of Boa Vista Energia, 06/30/2017, a receivable of CCC of R$ 278.36 million is registered and the position of the Agency is maintained, there will be an accounting difference of approximately R$ 272.7 million.

    This issue was pointed out as being of great risk to the eventual buyer of Boa Vista Energia, and may even render the auction unfeasible, since the amount under discussion may cause an impact of R$ 272.7 million. In this sense, the proposed model for privatization seeks to neutralize potential effects on the valuation due to uncertainty in the amount effectively recognized of the accounting difference of CCC credits.

    Page 66


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The modeling proposal foresees that Boa Vista Energia will transfer to the Eletrobras registered CCC credit in the amount of R$ 278.36 million and debts (which Boa Vista Energia has with Petrobras) in the same amount, which are already backed by this credit. Accordingly, the accounting effect of such a transfer in Boa Vista Energia would be zero, and the credit risk is transferred from Boa Vista Energia to Eletrobras.

    Although Boa Vista Energia has recognized this amount as a credit in its balance sheet, in the event that ANEEL maintains its position on the matter, even though Boa Vista Energia has filed a judicial appeal, it is probable, based on accounting rules, that the credit currently recorded becomes a provision. In that being so, there shall be an operation that would originally be neutral for Eletrobras to become a deficit operation in large sums. In this case, Eletrobras would assume a debit of R$ 278.36 million from Petrobras, which would result in a R$ 6.7 million loan from the CCC, resulting in a possible negative impact of R$ 272.7 million on its balance sheet.

    Table 47.  Accounting Impact Risk BRL Thousand 
    Value
    ANEEL 
    Balance Amount
    (Jun / 2017) 
    Possible
    Accounting
    Impact 
    5,666.71  278,360.00  -272,693.29 

     

    Notwithstanding the foregoing, it is worth mentioning, according to the legal opinion prepared by the law firm Décio Freire and Associados, dated September 18, 2017, referring to Administrative Proceeding no. 48500.004972/2016-34 of AmD, which, owing to its similarity, is being duplicated for Boa Vista, while there is no decision of the ANEEL’s

    Board wholly or partially ratifying the conclusions made, event in which there is no administrative act that implies an obligation to reimburse the fund. Additionally, the opinion considers that there is a solid legal basis for the review of ANEEL’s rulings and ultimately states that it is too early to provide for any potential loss connected to said Administrative Proceeding or to disclose the matter in questions in explanatory notes to the Financial Statements, as reproduced below:

    "Under all of the circumstances described above, we believe there is a solid legal basis for reviewing the provisions under the Technical Notes numbers 52/2017-SFF-SFG-SRG and 141/2017-SFF-SFG-SRG, as well as a manifest failure to comply with decisions rendered by Federal Regional Court of the

    Page 67


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    1st Region, in the records of Writ of Mandamus no. 0026107-81.2012.4.01.3400.

    This way, we understand that it is too early to provide for any potential loss connected to the Administrative Proceeding in question, or to disclose the matter in explanatory notes to the Financial Statements.”

    5. CAPITAL STRUCTURE

    The structure developed is aimed at enabling the sale of the Eletrobras distribution companies in the business model associated with the granting of the concession. The following is the current capital structure of Boa Vista Energia.

    Table 48.  Capital structure of Boa Vista Energia on 12/31/2016 
    Total Shares      320,742,889 
    Capital    R$  684,204,000.00 
    Par Value of Shares  R$    2,133,185.25 
     
    Participation    Shares  (%) 
    Eletrobras    320,742,889  100.00% 
    Minority      0  - % 
     
    Share Classes    Shares  (%) 
    Common    320,742,889  100.00% 
    Preferred    0  - % 

    Source: Product 11: Proposal for Modeling Privatization of Boa Vista, Service B.

    The economic value of the company's shares was determined as of 12/31/2016. Next, the definition of values was performed, considering the average between Services A and B values, as shown in the following table.

    Table 49.  Valuation of Boa Vista Energia - Average of Services A and B 
    Service A    R$  (328,272,159.71) 
    Service B    R$  (282,172,967.18) 
    Average of Services A and B  R$  (305,222,563.44) 
     Source: Product 11: Proposal for Modeling Privatization of Boa Vista, Service B. 

     

    Page 68 

     


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    As mentioned in item 6.3 of IT, adjustments to Boa Vista Energias equity value were required. The adjustments made were intended to contemplate the update of the balance sheet up to June 2017, which totaled negative R$ 36,847,922.75. Thus, the new equity value is negative R$ 342,025,486.20.

    Table 50.  Equity Value Final   
    Average of Services A and B  R$  (305,222,563.44) 
    Consolidated Adjustments  R$  (36,847,922.75) 
    Equity Value Adjusted  R$  (342,070,486.20) 

    Source: Product 11: Proposal for Modeling Privatization of Boa Vista, Service B.

    At the end of the privatization process, Boa Vista Energia should present the capital structure described below, considering the total adhesion of employees and retirees to the stock offering:

    Table 51.  Boa Vista Energia's capital structure after sale   
     
    Total shares        1,882,198,084,527 
    Capital Stock      R$  1,221,879,153.25 
    Par value of shares      R$  0.00064918 
    Interest  Common shares  Preferred Shares  Shares  (%) 
    Investor  1,693,978,135,238  0  1,693,978,135,238  90.0000% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retirees  188,219,949,288  0  188,219,949,288  10.0000% 
    Other Minority Shareholders  0  0  0  - % 
    Shares per class      Total Shares  (%) 
    Common      1,882,198,084,527  100.00% 
    Preferential      0  - % 
    Source: Product 11: Proposal for Modeling Privatization of Boa Vista, Service B.   

     

    5.1      CAPITALIZATION BY ELETROBRAS AND BY THE INVESTOR
    5.1.1      ELETROBRAS

    The scope of the proposed capital structure, considering the symbolic value of R$ 50,000.00, presupposes the discharge of the debts by Eletrobras in exchange for actions in Stage 1 of the sales modeling. Accordingly, there is a need to capitalize Eletrobras in the amount of R$ 342,120,486.20.

    Page 69


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    If Eletrobras chooses to increase its stake in up to 30% in Stage 2, as detailed in item 6.5 of IT, it will have to capitalize the distribution company in the amount of R$ 75.4 million, which may occur through capital contribution or conversion debt.

    5.1.2 INVESTOR

    As anticipated in Stage 2 of the modeling, the investor should make a capital contribution with subscription of new shares in the company, in the amount of R$ 175,999,185.71.

    6. NECESSARY ADJUSTMENTS AT ELETROBRAS

    According to the proposed model to reach the minimum value of R$ 50,000.00, Eletrobras must capitalize R$ 342.1 million at Boa Vista Energia, in addition to assuming an additional R$ 278.4 million, which today are backed by a CCC credit of the same amount, which will also be transferred to Eletrobras, as explained in item 4.2.

    Thus, Eletrobras should prioritize the cheapest debts to be capitalized/assumed, since the new Controlling Company will pay other debts. The table below shows the amount of debt of Boa Vista Energia by type of creditor and its associated costs:

    Table 52. Debt amount per creditor and its costs BRL millions

    RO + AFAC The
    World
    Bank 
    RGR CCC
    Obligations
    Related
    Parties
    Petrobras Total 
    Average
    cost 
    12.00% 6.59% 5.00% 8.75%
    (Selic) 
    8.75%
    (Selic) 
    Boa Vista
    Energia 
    139.51 15.48 2.62 76.99 587.00 335.30 1,156.89

     

    As can be seen from the table above, Boa Vista Energia´s cheapest debts, which will be assumed by Eletrobras to be fully or partially capitalized, are as follows: i) RGR, which has a cost of 5% per year; ii) World Bank, with a cost of 6.59%; iii) Petrobras, at SELIC cost (8.75%); and RO, with a cost of 12.59%. It is important to note that ANEEL must authorize Eletrobras to assume RGR's debt without having to remove all of it at the time of the transfer.

    Page 70


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Debts with the World Bank refer to the fund raising from this bank for the implementation of the "Energy+ Project". The RGR debts refer to financing of investments obtained before the period when it was appointed as temporary public service provider and funds of RGR received during the temporary provision. The Ordinary Resources are the pass-through of Eletrobras' own funds to the distribution company under loans, in order to cover operating deficits. Related party debts were contracted with other distribution companies or Eletrobras' companies, except the Holding Company. Among the main related parties with outstanding claims are, for example, Eletronorte, Furnas and Chesf.

    The table below shows the detailing of the capitalization by Eletrobras into Boa Vista, according to the source of funds, to achieve the set price of R$ 50,000.00. It should be underlined that, in this scenario, the preferred option was to convert the RO debts and before the World Bank in capital, although the common resources have a higher cost (12.00%), as the debts with Petrobras are posted as bonds under the CCD’s executed.

    Table 53.  Conversion of debts in capital, per distribution company R$ million 
      Debt  Capitalization 
      RO + AFAC  64.08 
      IBRD  15.48 
      Related Parties  262.56 
      Total  342.12 

     

    If the Shareholders’ Meeting of Eletrobras decides to assume the debt claims of CCC and

    CDE recorded in the balance sheets of the Distribution Companies, under the terms of Paragraph 1, Article 3 of CCPI Resolution 20/2017, as amended by CPPI Resolution no. 28/2017, the table below sets forth the amounts of debts, by source of funds, to be capitalized to achieve the set price in Boa Vista, as well as the assumption of debts in the same amount of CCC and CDE debt claims. In this context, the debts with Petrobras are prioritized, since most of these debts are already backed up by receivables of CCC, substantiated in Contracts of Admission of Debt, wherein Eletrobras stands as a guarantor.

    Page 71


     

     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

    Table 54.  Conversion of debts in capital and Assumption of Debt Claims of CCC 
    and CDE, per distribution company R$ million

    Debt  Capitalization   Assumption  Total 
    IBRD  15.48  -  15.48 
    RGR  -  2.62  2.62 
    CCC Reimbursement  -  76.99  76.99 
    Related Parties  -  190.09  190.09 
    Petrobras  326.64  8.66  335.30 
    Total  342.12  278.36  620.48 

     

    It should be noted that the amount corresponding to the assumption of debts of R$ 278.4 million, has as counterpart the credits of CCC registered in the Financial Statements of Boa Vista Energia, which, as explained in item 4.2, are subject to discussion with ANEEL, with the risk of not being incurred and becoming a much smaller credit, in the order of R$ 5.7 million.

    The table below summarizes the impact on Eletrobras in the event that the company succeeds in its administrative / judicial litigation and in case it cannot reverse the position of ANEEL. It is important to note that the result of the administrative / judicial funds may be in intermediate amounts to those presented in the table below.

    Table 55.  Impact Scenarios at Eletrobras - R$ million 
    SCENARIOS Assumption of Debts  Assumption of Debt
    Claims of CCC 
    Impact on Eletrobras
    Current position  620.48  278.36  -342.12 
    ANEEL Inspection  620.48  5.67  -614.85 

     

    6.1 FIXED ASSETS IN USE

    As referred to in Item 2.7 of this Appendix, it should be pointed out that the studies performed by Services A and B did not incorporate the positions of Fixed Assets in Use (AIC) into the Remuneration Base since, according to the Services, the assessors' report about the Remuneration Base does not take into account the specific assessment of the condition and characteristic of these assets concerning: (i) the eligibility to provide the

    Page 72


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the reassessment process. The table below shows the Fixed Assets in use, ascertained based on data of interim accounting balance sheets for each distribution company. The consideration of these assets, amounting to R$ 57.4 million at Feb./2017, would improve the results of the economic and financial assessments performed by the Services, and consequently, reduce the amount to be contributed by Eletrobras in the distribution company to reach the minimum amount of R$ 50,000.00, as defined in CPPI Resolution 20/2017.

    6.2 INDEMNITY OF THE CERR’S ASSETS

    It should also be noted that the economic and financial assessment studies of Services A and B considered, in the Regulatory Remuneration Basis of Boa Vista Energia, the assets of CERR, as well as the indemnification for the assets not fully depreciated and amortized of CERR, amounting to R$ 73 million, based on ANEEL Technical Note no. 293, dated September 20, 2017 (this being not a decision of ANEEL), being paid by Eletrobras.

    It should be noted that Law 12783/2013, in Paragraph 2 of its Article 8 establishes that the "calculation of the value of the indemnity corresponding to the installments of investments connected to reversible assets, not yet amortized or not depreciated, will be founded on the new replacement value methodology, according to criteria established in a granting authority regulation."

    Decree No. 8,461 / 2015, which regulated the extension of electric energy distribution concessions, defined, in its Article 5, that the unpaid concessions will be subject to a bidding process and the "indemnity to be paid to the former concessionaire, based on the value of the investments of the reversible assets not yet depreciated, shall be calculated by ANEEL based on the New Replacement Value - VNR and will consider the depreciation cumulative value from the date of commissioning of the facility, in accordance with the criteria of the Manual of Accounting of the Electric Industry - MCSE.”

    Decree 9192/2017, in paragraph 15 if its Article 4, establishes that “the new concessionaire shall, according to rules and deadlines to be defined in a public notice, purchases the assets and the reversible facilities connected to the provision of the service, which are separately owned by the company that will have its corporate control transferred through the bidding process provided for under the head provision,

    Page 73


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    corresponding to the part of the investments not amortized and/or not depreciated in connection with them, valued based on the methodology of the New Replacement Value -

    VNR.”

    Therefore, it should be noted that the particulars of the BNDES modeling, establishing the provision for of the liquidation of CERR's assets by Eletrobras in the manner calculated in the studies, is in conflict with the terms of Article 5 of Decree 8461/2015 and with the terms of Paragraph 15, Article 4 of Decree 9192/2017, are referred to above. Thus, it can be observed that the proposed sales modeling for Boa Vista Energia is in conflict with the provisions of Decrees 8461/2015 and 9192/2017.

    7. LIQUIDATION COSTS

    In the 165th EGM, held in July 2016, Eletrobras' shareholders decided, by majority vote, to approve: (i) the non-extension of the concessions of its distribution companies; (ii) the transfer of the ownership control of the distribution companies until December 31, 2017, pursuant to Paragraph 1-A of article 8, of Law 12.783/2013; and (iii) if there is no transfer of control under the terms of the previous item, the adoption of provisions for settlement thereof.

    Thus, the modeling study carried out by Service B calculated the cost with the settlement of Boa Vista Energia, in the case of non-occurrence of item (ii) above or not accepted by Eletrobras of the adjustments made to reach the equity value equal to zero.

    For these cases, Service B developed a comparative analysis between the distribution company's sales value associated with the grant of a new concession and the eventual settlement value of the company. Thus, it was possible to evaluate if the sale of the said state-owned company would be more attractive than its liquidation.

    Since the sale value considers adjustments in the equity value (equity value equal to zero), the adjustments were also considered in the estimated liquidation value. However, there were exceptions to the Tax and Negative Adjustments, since these do not impact on the settlement value, and for the AFACs, since these can be paid in the event that the distribution companies are not sold.

    Page 74


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    As shown below, in the case of Boa Vista Energia, the best economic option is to sell the company, considering the value of the assets and the concession. The balance between sell and settle is favorable for the first option, given the following table:

    Table 56.  Comparison between sale and liquidation of Boa Vista Energia 
     
    Equity Value Final  R$  -342,070,486.20 
    liquidation Amount (Pre-adjustments)  R$  -509,092,491.17 
    (+) Indemnity    R$  232,737,137.94 
    (-) Net Debt    R$  -888,848,183.81 
    (-) Contingencies  R$  1,300,914.47 
    (+/-) Payments / Receipts  R$  145,717,640.23 
    liquidation Amount (Post-adjustments)  R$  -465,696,717.78 

     


    Source: Product 11: Proposal for Modeling Privatization of Boa Vista, Service B.

    The results presented above support the scenario in which the distribution company sale, associated to the granting of a new concession, is the option that is less onerous to Eletrobras. If Eletrobras had to simply liquidate the distribution company under the scope of Eletrobras, the latter would have to pay a loss of R$ 465,696,717.78, an amount greater than R$ 123,626,231.59 when compared to the final equity value of negative R$ 342,070,786.20.

    Regarding the CCC claims in discussion with ANEEL, the effects shown in item 6 impact the company in both cases, disposal and liquidation. Thus, it has no direct impact on the comparison of values, so that the economic decision remains unchanged in the event of unsuccessful lawsuits.

    In addition, in the event of settlement of the distribution company, Eletrobras would not be entitled to the benefits arising from the relaxation of regulatory parameters, which generated a reduction in costs of R$ 150.00 million for Boa Vista Energia.

    Page 75


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    8.      ANALYSIS OF THE BENEFITS AND RISKS OF DISPOSAL
    8.1     

    BENEFITS

     

      (i) The disposal of Boa Vista Energia shares will result in the adoption by the company of a legal regime similar to that of other non-state agents
     in the sector, allowing for significant operating gains; 
     
     
     
      (ii) Improvement of Eletrobras' Ebitda and elimination of the need for disbursement of by Eletrobras to finance the cash deficit and investments of Boa Vista, to the reduction strategy of the net debt / ebitda indicator; 
     
     
     
      (iii) It avoids the need for capitalization by shareholders to meet the financial targets imposed by ANEEL for the extension of the concession, and for in the expansion and improvement of service quality; 
     
     
     
      (iv) The sale together with the granting of concession allows the use of the relaxation the regulatory parameters approved for the period from 2018 to 2023, when the review of the new concession contract will occur; and 
     
     
     
      (v) The said flexibility provided an improvement in the distribution company's equity of the order of R$ 150.00 million. 
     
     
    8.2      RISKS
      (i) Frustration of the disposal of the shares or non-performance of the auction, due to

    several factors, notably due to possible judicial, political, economic-financial, regulatory, governmental or administrative impediments;

    (ii)      Non-integration of capital and interior operations; 
     
    (iii)      Frustration of the disposal of the shares or non-performance of the auction, due to

    several factors, notably due to possible judicial, political, economic-financial, regulatory, governmental or administrative impediments;

    Page 76


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    (iv)      Absence of investors and/or companies interested in purchasing; 
     
    (v)      Reduced competitiveness in the bidding procedure due to the current economic

    situation of Brazil; and

    (vi) Amendment of the minimum conditions for disposal by the TCU.

    9. FINAL REMARKS

    Lastly, it should be noted that the economic and financial assessment methodology was found to be widely known and adopted by the market, which is the Free Cash Flow to Firm - FCFF, in nominal terms. In addition, the calculated discount rates were based on the CAPM - Capital Asset Pricing Model method, which estimates the return that the shareholders expect to obtain from the company, given this is a well-established method widely used either by the market at large or by Eletrobras.

    Additionally, some points need to be highlighted. Among them, there is the need for approval of the minimum amount and the adjustments necessary for achieving such amount by the Executive Board and Board of Directors of Eletrobras, in addition to the need to submit the item to the general shareholders' meeting.

    Another point to be highlighted is that the modeling, as well as the economic-financial assessments elaborated by the consulting firms, considers the integration of the operations of distribution of energy of the capital and interior of the State of Roraima, including assets and areas of provision of the service of distribution of electric energy. However, the CNPJ connected to Boa Vista Energia holds only the assets connected to the service of distribution of electric energy of the Capital of the State of Roraima, while CERR, which has as shareholder the State of Roraima, holds the assets connected to the provision of the service of distribution of electric energy to its interior.

    With regard to the need for adjustments, the study establishes the need for capitalization by Eletrobras of the debts in exchange of shares at Boa Vista Energia in the value of R$

    342.12 million, in order for the company’s value to achieve the minimum value of R$

    50,000.00. Additionally, Eletrobras shall assume R$ 278.36 million in debts under the debt claim corresponding entry, with CCC/CDE, at the same amount.

    Page 77


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    It should be noted that the studies do not consider the fixed assets in use - AIC, amounting to R$ 57.4 million, referenced back to values of Feb./2017. If, in the modeling, AIC’s were considered unitized, even if set aside, they would be able to reduce the value to be disbursed in the distribution company by Eletrobras.

    In the specific case of Boa Vista Energia, attention should be given to the indemnity of assets not amortized or depreciated of CERR, given that, in the modeling, while defining such value and providing for the settlement of these assets by Eletrobras, there is a conflict with the provisions of Decrees 8461/2015 and 9192/2017. Both define, respectively, that: i) ANEEL shall define the value of indemnity and; ii) the new utility company shall purchase non-amortized or depreciated assets of CERR by VNR.

    Furthermore, it should be noted that in the proposed sales model there is the possibility that Eletrobras to remain with one share of Boa Vista Energia, with the sole purpose of increasing its shareholding in the distribution company by up to 30%, after sale of the controlling interest in the company to a third party. The increase in interest, if it is the option adopted by Eletrobras, must be carried out within six months after the date of the auction.

    It should also be reiterated that the modeling entertained the possibility that Boa Vista Energia has values of CCC previously stated as amounts receivable in the balance sheet, which were currently reverted into obligations payable, owing to technical decisions of ANEEL. The difference of understanding between Boa Vista Energia and ANEEL achieves amounts as large as R$ 272,7 million. Therefore, in view of the relevance of the value discussed for the valuation of Boa Vista Energia and the unpredictability of the deadline for a final decision on this matter, it was recommended by Service B that the responsibility for such debt claims - including possible benefits and burdens - is transferred from Boa Vista Energia for Eletrobras’ Holding. Thus, the effect of the current questioning would be neutral on the assessment of the market value of the distribution company and, therefore, on its sales modeling. However, this recommendation requires the consent of the regulators and should be accompanied by considerations of the company's external auditors. Furthermore, the uncertainties existing within the scope of Boa Vista Energia should also be taken into consideration, as this operating will make them be incurred by Eletrobras. Therefore, this point should be considered by the managers of the Holding.

    It is worth noting that Service B performed a comparative analysis between the option of divesting the shareholder control of the distribution company with the granting of a new

    Page 78


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    concession and the simple liquidation option of the distribution company, as deliberated by the shareholders at the 165th EGM of Eletrobras. With this analysis, one may conclude that the liquidation of the distribution company, under the scope of Eletrobras, as established under the modeling proposed by BNDES, is the most onerous option to Eletrobras, given that, for such purpose, it would have to bear an additional cost of R$ 123.6 million connected to the sale scenario.

    Lastly, it should be noted that in the case of the distribution company’s liquidation, Eletrobras would not benefit from the relaxation of regulatory parameters, which contributed to an improvement in the value of Boa Vista Energia’s equity value of the order of R$ 150 million.

    Page 79


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Appendix to the DFP/DFF/DDE Joint Technical

    Information

    Analysis and Results of CEAL

    Page 1


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Contents     
    1.  Objective  4 
    2.  Assumptions of economic and financial modeling  4 
    2.1  Market and consumer units  4 
    2.1.1  Evolution of Captive Market  4 
    2.1.2  Market Projections prepared by Service A  5 
    2.1.3  Market Projections prepared by Service B  10 
    2.1.4  Assumptions used and points of divergence of Services A and B  14 
    2.1.5  Evolution of Consumer Units  16 
    2.1.6  Projections of Consumer Units prepared by Service A  17 
    2.1.7  Projections of Consumer Units prepared by Service B  18 
    2.1.8  Comparison of Services A and B projections  19 
    2.1.9  Energy procurement  19 
    2.2  Network extension (km)  19 
    2.3  Electrical losses  21 
    2.3.1  Projections of Electrical Losses prepared by Service A  22 
    2.3.2  Projections of Electrical Losses prepared by Service B  24 
    2.3.3  Assumptions used and points of divergence of Services A and B  26 
    2.4  Default and unrecoverable revenue  27 
    2.5  PMSO (personnel, material, services and others)  29 
    2.5.1  PMSO incurred (2012 2016)  29 
    2.5.2  PMSO for Service A  31 
    2.5.3  PMSO for Service B:  36 

     

    Page 2


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      2.5.4 Projections and Comparisons of the PMSO of Services A and B  40 
      2.6  Investment  45 
      2.7  Regulatory remuneration base  48 
      2.8  X Factor  52 
      2.9  WACC  57 
    3.    The result of economic and financial assessments and by multiples  59 
      3.1  AFAC  61 
    4.    Contingencies  61 
      4.1  Legal Contingencies  61 
      4.2  Accounting Contingencies  61 
    5.    Capital structure  62 
      5.1  Scenario 1  63 
      5.2  Scenario 2  64 
      5.3  Capitalization by Eletrobras and by the Investor  65 
      5.3.1 ELETROBRAS  65 
      5.3.2 Investor  65 
    6.    Necessary adjustments at Eletrobras  66 
      6.1  Fixed Assets in use  67 
      6.2  Plano Bresser AGREEMENT  68 
    7.    Liquidation costs  68 
    8.    Analysis of the benefits and risks of disposal  71 
      8.1  Benefits  71 
      8.2  Risks  71 
    9.    Final Remarks  72 

     

    Page 3


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    1. OBJECTIVE

    The purpose of this appendix is to include a detailed presentation of the results of the economic and financial assessments of Ceal, performed by Services A and B, contracted by BNDES, as well as to detail the internal adjustments that Eletrobras needs to implement in order to enable its sale at the target price determined by the model approved under CPPI Resolution 20, dated November 8, 2017, amended by CPPI Resolution 28, dated November 22, 2017.

    2.      ASSUMPTIONS OF ECONOMIC AND FINANCIAL MODELING
    2.1      MARKET AND CONSUMER UNITS
    2.1.1      EVOLUTION OF CAPTIVE MARKET

    In the last ten years, from 2008 to 2016, Eletrobras Distribuição Alagoas had average annual growth of 5.2%, as can be seen in the tables below. The Residential, Commercial, Industrial and Public Lighting classes contributed to this development, with an annual average growth of 6.6%, 6.1%, 4.8% and 6.4%, respectively.

      Table 1.  Evolution of consumption classes in MWh     
     
              POWER BROKEN DOWN PER CLASS (MWh)       
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  740,401  808,170  853,067  924,744  1,018,464  1,094,798  1,225,494  1,304,933  1,323,913  1,305,037 
    INDUSTRIAL  374,289  381,483  353,865  467,648  474,217  584,373  554,697  579,270  556,396  535,000 
    COMMERCIAL  431,658  464,756  487,051  530,211  563,924  633,334  680,412  731,982  734,369  730,449 
    RURAL  162,590  172,829  167,200  147,680  153,175  210,218  216,079  178,919  180,284  164,837 
    GOVERNMENT  117,802  122,374  122,396  127,958  131,717  138,799  145,272  150,566  159,606  161,107 
    PUBLIC LIGHTING  126,093  131,407  135,147  136,058  136,148  155,555  189,829  202,582  201,256  215,344 
    PUBLIC SERVICE  156,553  160,803  166,593  168,549  172,747  185,652  179,167  183,564  194,694  205,849 
    OWN CONSUMPTION  4,284  4,077  5,021  2,604  4,149  4,243  3,713  3,724  3,172  3,183 
    TOTAL CONSUMPTION  2,113,670  2,245,899  2,290,340  2,505,452  2,654,541  3,006,972  3,194,664  3,335,540  3,353,690  3,320,806 

     

    Page 4


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
      Table 2.  Composition of Consumption Class     
     
                COMPOSITION OF CLASSES       
        2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
        35.0%                   
    RESIDENTIAL      36.0%  37.2%  36.9%  38.4%  36.4%  38.4%  39.1%  39.5%  39.3% 
    INDUSTRIAL    17.7%  17.0%  15.5%  18.7%  17.9%  19.4%  17.4%  17.4%  16.6%  16.1% 
        20.4%                   
    COMMERCIAL      20.7%  21.3%  21.2%  21.2%  21.1%  21.3%  21.9%  21.9%  22.0% 
    RURAL    7.7%  7.7%  7.3%  5.9%  5.8%  7.0%  6.8%  5.4%  5.4%  5.0% 
    GOVERNMENT    5.6%  5.4%  5.3%  5.1%  5.0%  4.6%  4.5%  4.5%  4.8%  4.9% 
        6.0%                   
    PUBLIC LIGHTING      5.9%  5.9%  5.4%  5.1%  5.2%  5.9%  6.1%  6.0%  6.5% 
    PUBLIC SERVICE    7.4%  7.2%  7.3%  6.7%  6.5%  6.2%  5.6%  5.5%  5.8%  6.2% 
        0.2%                   
    OWN CONSUMPTION      0.2%  0.2%  0.1%  0.2%  0.1%  0.1%  0.1%  0.1%  0.1% 
    TOTAL CONSUMPTION    100%  100%  100%  100%  100%  100%  100%  100%  100%  100% 
     
     
     
      Table 3.  Annual Growth Rate in the Period     
     
              ANNUAL GROWTH RATE         
                          AVERAGE OF 
      2008  2009  2010  2011  2012  2013  2014  2015  2016  ANNUAL 
                          GROWTH 
    RESIDENTIAL  9.2%  5.6%  8.4%  10.1%  7.5%  11.9%  6.5%  1.5%  -1.4%  6.6% 
    INDUSTRIAL  1.9%  -7.2%  32.2%  1.4%  23.2%  -5.1%  4.4%  -3.9%  -3.8%  4.8% 
    COMMERCIAL  7.7%  4.8%  8.9%  6.4%  12.3%  7.4%  7.6%  0.3%  -0.5%  6.1% 
    RURAL  6.3%  -3.3%  -11.7%  3.7%  37.2%  2.8%  -17.2%  0.8%  -8.6%  1.1% 
    GOVERNMENT  3.9%  0.0%  4.5%  2.9%  5.4%  4.7%  3.6%  6.0%  0.9%  3.6% 
    PUBLIC LIGHTING  4.2%  2.8%  0.7%  0.1%  14.3%  22.0%  6.7%  -0.7%  7.0%  6.4% 
    PUBLIC SERVICE  2.7%  3.6%  1.2%  2.5%  7.5%  -3.5%  2.5%  6.1%  5.7%  3.1% 
    OWN CONSUMPTION  -4.8%  23.2%  -48.1%  59.3%  2.3%  -12.5%  0.3%  -14.8%  0.3%  0.6% 
    TOTAL CONSUMPTION  6.3%  2.0%  9.4%  6.0%  13.3%  6.2%  4.4%  0.5%  -1.0%  5.2% 

     

    2.1.2 MARKET PROJECTIONS PREPARED BY SERVICE A

    The consumption of electrical energy in Brazil can be segregated into eight distinct classes, which are: Residential, Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service and Own Consumption. In order to capture the possible heterogeneities among the analyzed groups, Service A separately modeled the Residential, Commercial, Industrial, Rural and Other sectors (which compose the classes of Public Power, Public Lighting, Public Service and Own Consumption).

    This decision was based on the similar behavior of consumption among them and to minimize the effect of the volatility incurred in low consumption classes.

    To project energy demand for the next 30 years, ending in February/2048, the consultancy first made a historical analysis of the evolution of energy consumption and its distributions between classes, using an observable period of consumption from 2007 to 2016. The results are presented in the following tables:

    Page 5


     



     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 5.  Geometric Mean of Consumption 2017-2047   
     
      Consumption  Consumption  Consumption  Consumption  Consumption  Consumption 
    Period  RESIDENTIAL  INDUSTRIAL  COMMERCIAL  RURAL  OTHERS  TOTAL 
      ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a 
    2017 - 2047  3.63%  2.64%  2.83%  1.68%  2.78%  3.08% 

     

    g. Projection of Demand by Voltage Level

    For the projection of demand by Voltage Level, the historical average of low voltage of the last 5 years of the Company was used. This metric was based on the idea of stability in the composition of energy consumption between classes. Low voltage consumption was defined and the residual percentage was divided between medium and high voltage according to the SPARTA report. The set percentages for 2017-2047 are 60% for LV, 27% for AV and 13% for TA.

    2.1.3 MARKET PROJECTIONS PREPARED BY SERVICE B

    In order to know the reality experienced by the Company distribution and to use the historical values for the necessary projections, Service B initially collected consumption information made by the utility from 2013 to 2016. The evolution of the Market by level of voltage in the period is presented in the table below:

      Table 6.  Market Evolution by Voltage Level     
     
      Market per Voltage Level (MWh)        Change Rate 
                Period  Geometric 
    Voltage/Year  2013  2014  2015  2016    (2013/2016)  (p.a.) 
    SUPPLY  3,144,209  3,283,014  3,389,576  3,299,290    5%  2% 
    A3  458,531  409,962  409,026  417,090    -9%  -3% 
    A3a  -  -  -  -  -    - 
    A4  880,332  895,395  906,768  888,530    1%  0% 
    LV  1,805,345  1,977,657  2,073,782  1,993,670    10%  3% 
    LOW INCOME  261,853  309,028  311,097  227,525    -13%  -5% 
    SUPPLY  -  -  -  -    -  - 
    OTHER FREE  148,363  195,982  195,132  159,001    7%  2% 
    DISTRIBUTION  18,942  19,317  20,233  20,221    7%  2% 
    TOTAL  3,311,514  3,498,314  3,604,942  3,478,512    5%  2% 

     

    In order to subsidize the various tariff calculations and the flow of the company, Service B elaborated market simulations by: (i) consumption class (Residential, Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service, Own Consumption); and (ii) voltage level (High Voltage - A1, A2 and A3 -, Medium Voltage - A3a and A4 - and Low Voltage - LV).

    Page 10


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 7.  Variables used in market forecasts by class 
    Class Variables
    Tested 
    Used 
    Residential  Population or UC Residential  Population 
    Industrial  GDP GDP 
    COMMERCIAL  GDP and one among: Population and UC Commercial  GDP and Population 
    Rural  GDP and Rural Population or UC and Univariate  Univariate 
    Government  Population and/or GDP  Population 
    Public Lighting  Univariate  - 
    Public service  Population  Population 
    Own Consumption  Univariate  - 

     

    b.1 Residential Class Result

    For the Residential Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Population as an auxiliary variable and presence of seasonality, using the observed consumption series from January 2012 to October 2016. From the estimated forecasts, consumption grows on average 2.83% p.a. between 2017 and 2048.

    b.2 Industrial Class Result

    For the Industrial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust), was based on the Dynamic Model, with the auxiliary variable of GDP, seasonality component, structural break as of January 2012 and seasonality, using the observed series of consumption from January 2009 to October 2016. From the estimated forecasts, consumption grows on average 2.35% p.a. between 2017 and 2048.

    b.3 Commercial Class Result

    For the Commercial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Population and State GDP as auxiliary variables, in addition to the seasonality, using the observed consumption series from January 2009 to October 2016. From the estimated forecasts, consumption grows on average 3.62% p.a. between 2017 and 2048.

    b.4 Result of the Rural class

    For the Rural class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the ETS methodology, using the observed

    Page 12


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    consumption series from January 2009 to October 2016. From the estimated forecasts, consumption grows on average 0.21% p.a. between 2017 and 2048.

    b.5 Result of the Public Power class

    For the class of Public Power of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with the variables of Population, seasonality and structural break as of July 2014, using the observed series of consumption from January 2011 to October 2016. From the estimated forecasts, consumption grows on average 1.30% p.a. between 2017 and 2048.

    b.6 Result of Public Lighting class

    For the Public Lighting class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the ETS methodology, using the observed consumption series from January 2009 to October 2016. From the estimated forecasts, consumption grows on average 0.99% p.a. between 2017 and 2048.

    b.7 Result of the Public Service class

    For the class of Public service of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with the auxiliary variable of Population, structural break as of December 2013 and momentary dummies in February 2012, using the observed series of consumption from January 2009 to October 2016. From the estimated forecasts, consumption grows on average 0.81% p.a. between 2017 and 2048.

    b.8 Result of class Own Consumption

    For the Public Lighting Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with a momentary dummy in May 2014 and seasonality, using the observed series of consumption from January 2011 to October 2016. From the estimated forecasts, consumption grows on average 0.02% p.a. between 2017 and 2048.

    b.9 Global Projection Result

    The table below shows the market projected by consumption class broken down every ten years between 2016 and 2046. The consumption in MWh per class is presented through its percentage participation in relation to the total consumption and the percentage growth

    Page 13


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

      Industrial, Commercial, Rural, Public Power, Public Lighting, Public Service and Own Consumption. And another projection per voltage level, consisting of: High Voltage (A1, A2 and A3), Medium Voltage (A3a and A4) and Low Voltage (LV).
    3.      The historical consumption period of the Distribution Company, used by Service A, is the years 2007 to 2016. Service B, on the other hand, used periods beginning in 2009, 2011, or 2012, depending on the projected consumption class and always ending in 2016.
    4.      Service A obtained average annual growth projections of 3.08% for the distribution company's consumption in the period from March 2018 to February 2048. For Service B, the average annual growth of the distribution company's consumption was around 2.61% for the period from 2017 to 2048.
      Based on the Figure that describes the Projection of the Energy Consumption of Service A, we see that the consumption projected by that of Service A is around 8,400 GWh in 2047 (in Service A report, the absolute value of the consumption projection was not observed). Service B projects a consumption of 7,195 GWh in 2046, as shown in Table 9. In absolute values, Service A shows a difference approximately 1,200 GWh greater than Service B, which is the result of an average annual growth 0.47% smaller for Service A compared to that admitted by Service B.

    Page 15


     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    2.1.8 COMPARISON OF SERVICES A AND B PROJECTIONS

    Based on the trajectory of the graphs of the UCs of Services A and B we have a comparison between their evolutions over the period. Note that in the first 10 years the growth rate of Services A and B are similar. In the 20 years after the projection, Service B has a higher annual growth rate. Also noting that Service A shows that the rate stabilized over the last 10 years.

        Table 13.  Annual Growth Rate in the Period     
     
        Performed    Forecasted
    Services 2015  2016  2015/2016
    (% p. a.)  
    2026  2036  2046  2017/2026
    (% p. a.) 
    2027/2036
    (% p. a.)  
    2037/2046
    (% p. a.)  
    A        1,230,000  1,280,000  1,285,000  0.97%  0.40%  0.04% 
     B  1,045,270  1,117,108  6.87%   1,223,030   1,316,609   1,420,163   0.91%   0.74%   0.76% 
    Differences (A-B)  -  -  -  6,970  - 36,609  - 135,163  0.06%  -0.34%  -0.72% 
     
     
     
     
    2.1.9 ENERGY PROCUREMENT             

     

    Regarding the energy procurement balance, Service B considered the existing energy contracts, the current overcontracting level and explained the treatment given, as well as the projection of the PLD. Service A has used, as an assumption, the non-treatment of overcontracting from its starting point. According to the Disclaimer of the consulting firm’s work: "In this assessment were not considered penalties in the overcontracting of energy by the assessed enterprise, since it was assumed the assumption that the new utility will be able to liquidate its positions at market prices". That is, the revenue from the sale of surplus energy in the short-term market would be equal to the distribution company's average energy purchase cost.

    2.2 NETWORK EXTENSION (KM)

    The annual variation of the extension of the distribution network has an impact on the Distribution company's projection of investments, as well as being part of the projection metric of the calculation of regulatory operating costs and the PMSO carried out by the Distribution company.

    The total extension of the network can be segregated into four voltage groups: low, medium and high voltage and “Luz para Todos” program.

    Page 19


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Regarding the starting point, that is, the network extension for the year 2017, according to the information contained in the assessment reports of Services A and B, the Consulting Firms started with very different information:

    Service A reported for the year 2017 the network extension numbers of 41,210 km in LV/AV, 2,128 km in AT and 2,160 km in PLPT.

    - Service B reported only the figure of 42,000 km of length for LV. The AT and PLPT figures were not referred to.

    2.3 ELECTRICAL LOSSES

    The losses of the electric distribution system are divided into two categories: Losses in the Basic Network and Losses in the Distribution Network. In turn, losses in the Distribution Network are divided into Technical Losses (PT) and Non-Technical Losses (NTP).

    Losses in the Basic Network are those admitted by the SIN. The real values are calculated monthly by the CCEE, and the regulatory value is given through the average of the measurements of the last 12 months. The value of the 2016 tariff adjustment was used for the assessment report.

    PTs are the loss of transport and electrical equipment. These are calculated by ANEEL in the tariff review period and remain constant for the entire tariff cycle.

    PNTs are the manageable losses where ANEEL's reduction curve is estimated at each tariff review and takes into account three factors: the target, the trajectory to the target and the limits of reduction of regulatory loss. The goal is calculated based on the analysis of the comparative model of benchmarking, in which a company is established that will serve as a basis for comparison, taking into account the size of the distribution company and the socioeconomic complexity of the region in which the distribution company is located. According to the metric specified in PRORET submodule 2.6. It should be reminded that the percentage levels of PNT are related to the Low Voltage (LV) market of the utility, while the PT and Loss in RB indices are measured in relation to the Injected Energy, in accordance with the regulatory methodology.

    Page 21


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The percentage values of Losses are presented below, according to Ceal's Sparta worksheet of 2016, as well as the values of 2016 of the distribution company, according to NT 149/2017-ANEEL.

    According to NT 149/2017-ANEEL, the company's new PNT regulatory level will be 27.18% of the LV market, and should remain fixed until the first common tariff review, which is defined as the midpoint between the loss in 2016 and regulatory loss. The premise used is that the new utility provider would take five years, or one rate-related cycle, to converge to the regulatory benchmarks.

    Table 16.  Calculation of losses - SPARTA Ceal 2016 
     
    Description  Regulatory  Performed
    2016 
    Non-Technical Percentage (on Low Voltage)  15.67%  38.70% 
    Technical Percentage (w/o input market)  10.34%  10.34% 
    % Basic Grid (w/o Input Market)  2.71%  2.71% 

     

    2.3.1 PROJECTIONS OF ELECTRICAL LOSSES PREPARED BY SERVICE A

    a. Projection of Losses in the Basic Network and Technical Losses

    For Service Losses in the Basic Network and PT, Service A defined that for both indicators the regulatory and real values will remain constant and equal throughout the period, since the values of these losses are not related to the activity of the distribution company and its management, but rather physical processes and characteristics of the equipment used.

    Table 17. Level of Technical Losses and Losses in Basic Network

      Technical Losses  Losses in the Basic Grid 
      10.34%  2.71% 
     
     
    b.  Projection of Non - Regulatory Technical Losses 

     

    Regarding the trajectory of Regulatory NTP, Service A defined in two different moments. The first period, from Sept./2017 to Sept./2023, the regulatory levels were maintained at the level of the regulatory losses indicated in NT 149/2017-ANEEL, equivalent to 27.18%.

    Page 22


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    As of Sep./2023, the method of calculation of the curve of regulatory losses occurred according to the definitions described in the submodule 2.6 of PRORET, taking into account three previously mentioned factors. After the comparative analysis of the listed companies with the Ceal it was defined that their benchmarking of the PNT goal would be COELBA.

    Upon definition of the target for the next year of the tariff review, the analysis of the annual reduction of regulatory losses that can have two forms of reduction for each cycle of tariff revision, is carried out. The trajectory will be the limit of reduction of regulatory losses stipulated in PRORET, in case the effort to reach the goal of benchmarking is more challenging. Otherwise, the reduction curve will have a linear trajectory until it loses benchmarking at the end of each tariff review cycle.

    The path of reduction of PNTs continues until reaching the level of 7.5% as determined by PRORET.

    c. Projection of Non -Technical Real Losses

    For the projection of real PNTs, three perspectives were observed: starting point, trajectory and target.

    As a starting point, the PNT was considered in 2016 of 38.18%, obtained from the energy balance of the distribution company. The defined target is the level of PNT used by the benchmark company considered in the calculation of the target of regulatory losses, in this case, the levels used by COELBA which are 5.05%. The value of the loss of the benchmark company was obtained through the SAMP database.

    For the annual PNT reduction trajectory, the observed trajectory of an efficient utility was used. The benchmark company selected was CEMAR - Companhia Energética do Maranhão. The choice of the company was based on the fact that CEMAR already has an efficient level of Non-Technical losses and has recently undergone the privatization process.

    Table 18.  Starting Point and Target of Non-Technical Losses 
    % of Losses/BT Market
    Non-technical Losses
    in 2016
    38.18% 
    Non-technical Losses
    Benchmark in 2016
    5.05% 
    Year in which the
    regulatory level was
    achieved
    2025 

     

    Page 23


     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

    3.      Modeling of Service A Regulatory Non-Technical Losses (NTP) was organized in two stages. At first, from 2017 to 2022, the regulatory levels were maintained at the level of 27.18%, based on the value defined in NT 149/2017 - ANEEL. With a sharp reduction in 2023, and in later years, there was a gradual decrease to 7.5% in 2032 and maintaining this level until the end of the concession.
      In Service B, the modeling of Regulatory PNTs was carried out based on the application of the mechanisms detailed in NT 106/2015-ANEEL. The reduction trajectory starts with 15.7% in 2017, reaching 7% in 2047 on the LV Market. Although Service B describes the application of NT 351/2017-ANEEL (which also addresses the same PNT parameters specified in NT 149/2017-ANEEL), the defined PNT levels were not observed in the graph of its report. Thus, Service was consulted and ensured that the 27.18% PNT level was used in the first five years, as determined by 149/2017-ANEEL. The trajectory of real PNT’s prepared by
      Service A starts from the value in 2016 of 38.18%, the curve reduces until 2034, when it reaches the value realized by COELBA (benchmarking) of 5.05% of the market, remaining at this level until the end of the concession. For Service B, in the years 2017 and 2018, the real PNT indices will remain constant at 41%, as of 2019, a new phase in the efficient combat against PNT will begin. The path of reduction of losses considers as reference the average of reductions of the Real PNT’s of the companies CELPE, CEMAR and COELBA. Thus, at the end of 2047, the projection of the real NTP index is 8%.
    4.      Note that the PNT carried out in Ceal in 2016 and used as assumption by Service A is 38.18% of the LV market is different from that presented in NT-149/2017-
    SRM/SGT/SRD/SFF/ANEEL,      which was 38.7% on the LV market.

    2.4 DEFAULT AND UNRECOVERABLE REVENUE

    The Projected Unrecoverable Revenues and Default is the median of billed value that was not received until the base date of December 2016, the stabilization point of the unbilled revenue curve, from the 49th to the 60th month.

    The default assumption has great relevance in the Valuation analysis because it directly affects the collection, consequently, the company's cash flow, negatively impacting its Valuation.

    Analyzing the two Services, there are some divergences, which are listed below:

    Page 27


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    It should be noted that Service B numbers were estimated in a linear decline reaching the regulatory level in 10 years, according to text contained in the Assessment Report, without presenting the annual values used.

    Below are tables with the numbers of the default up to 2027 of the two Services:

      Table 19.  Default until 2027 - Services A and B       
     
    Service B                       
    Sunk Cost                       
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    Residential  2.95%  2.80%  2.65%  2.50%  2.35%  2.20%  2.04%  1.89%  1.74%  1.59%  1.44% 
    Commercial  2.92%  2.70%  2.48%  2.26%  2.04%  1.82%  1.60%  1.38%  1.16%  0.94%  0.72% 
    Industrial  5.88%  5.40%  4.91%  4.43%  3.95%  3.47%  2.98%  2.50%  2.02%  1.53%  1.05% 
    Rural  6.21%  5.71%  5.20%  4.70%  4.20%  3.70%  3.19%  2.69%  2.19%  1.68%  1.18% 
    Government  2.95%  2.68%  2.41%  2.13%  1.86%  1.59%  1.32%  1.05%  0.77%  0.50%  0.23% 
    Public Lighting  1.74%  1.57%  1.40%  1.23%  1.06%  0.90%  0.73%  0.56%  0.39%  0.22%  0.05% 
    Public service  0.78%  0.71%  0.64%  0.56%  0.49%  0.42%  0.35%  0.28%  0.20%  0.13%  0.06% 
    Service A*                       
    Sunk Cost                       
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    Residential  2.95%  2.66%  2.38%  2.09%  1.80%  1.52%  1.23%  0.94%  0.65%  0.37%  0.08% 
    Commercial  2.92%  2.63%  2.35%  2.06%  1.78%  1.49%  1.20%  0.92%  0.63%  0.35%  0.06% 
    Industrial  5.88%  5.29%  4.70%  4.12%  3.53%  2.94%  2.35%  1.76%  1.18%  0.59%  0.00% 
    Rural  6.86%  6.76%  6.66%  6.56%  6.21%  5.87%  5.55%  5.25%  5.04%  5.04%  5.04% 
    Government  2.31%  2.13%  1.96%  1.81%  1.53%  1.29%  1.09%  0.93%  0.78%  0.66%  0.56% 
    Public Lighting  10.54%  9.67%  8.87%  8.14%  6.62%  5.38%  4.37%  3.56%  2.89%  2.35%  1.91% 
    Public service  0.14%  0.14%  0.14%  0.14%  0.14%  0.14%  0.14%  0.14%  0.14%  0.14%  0.14% 

     

    In Service A there is not a specific year to reach the regulatory level, as performed by Service B, below. A target was established, above the regulatory level, and an annual

    Page 28


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    reduction curve based on the rate of change proposed by ANEEL in 4CRTP, 3CRTP and 2CRTP. With this assumption, the regulatory level is not reached during the thirty years of concession.

    Service A raised R$ 18.221 million in Unrecoverable Revenues for the year 2016. Service B did not show any value.

    The year-on-year comparison of the projection of default rates by each consulting was hampered because Service B did not disclose the annual amounts in a table, showing only the starting value and the year that the regulatory level is reached.

    2.5      PMSO (PERSONNEL, MATERIAL, SERVICES AND OTHERS)
    2.5.1      PMSO INCURRED (2012 2016)

    For the analysis of the PMSO item contained in the Services A and B reports, it is necessary to look, in advance, at the history of this item over the last few years. For the case, the PMSO was analyzed between the period from 2012 to 2016 (same period used by Service A). Below are the values in the aforementioned period for Ceal:

    Table 20.  PMSO incurred from 2012 - 2016 (Ceal) - R$ million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  155.9  166.1  145.3  154.1  158.5 
    PID  0.0  78.6  0.0  0.0  0.0 
    Materials  3.4  1.8  1.5  1.9  1.7 
    Third-party Service  79.7  74.7  80.3  89.9  99.5 
    Other  23.5  37.0  17.4  37.5  24.2 
    Total  262.4  358.2  244.5  283.3  283.9 

     

    For the composition of each item above, the Financial Statements of the respective years were used as support material. In item "O" of the PMSO, the following items were considered: Leases and Rentals; Taxes; Miscellaneous indemnities; Judicial costs and fees; Own energy consumption; Inspection Fee and Others. It is also worth noting that the PID (Resignation Incentive Plan) was considered for the years 2013 and 2014.

    In order to show consistency in the comparison with the Assessment Reports presented by the Consultancies, the PMSO values were updated by the IPCA for December 2016.

    Page 29


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

    2.5.2      PMSO FOR SERVICE A
    -      Historical part:

    For the composition of the historical part, the consultancy, based on the accounting statements provided by the distribution company, made a grouping of similar accounts in terms of type of expenses and also made an adjustment of the values for the base date of December 2016. Below is the table containing the opening of the PMSO for Ceal:

    Table 22.  PMSO calculated by the base date of December/2016 (Ceal) - R$ 
    million
      R$ Million 

     

    PMSO  2012  2013  2014  2015  2016 
    Personnel  213.8  314.5  175.3  170.7  161.3 
    PID  0.0  0.0  0.0  0.0  0.0 
    Materials  4.6  2.3  1.9  2.1  1.7 
    Third-party Service  109.3  96.0  96.9  99.6  101.3 
    Other  7.9  13.2  18.9  23.1  10.9 
    Total  335.6  426.0  292.9  295.5  275.3 

     

    - Methodology used in Service A: (part common to all Distribution Companies)

    The analysis of the PMSO as well as the projection of the values was divided into two parts, namely:

    Part 1 - Characterization of Analytical Accounts (Divided in 5 phases):

  • Phase 1 - Grouping of analytical accounts:

     

      o     

    Adjustment for inflation of the accounts for the base date of December 2016;

     

      o     

    Grouping of accounts according to similarity of expense type.

     

  • Phase 2 - Analysis of the nature of the account in fixed and variable:

     

      o     

    Analysis of Fixed and Variable Accounts;

     

      o      The Accounts identified as fixed, when projected, were calculated as the average of the expenses of the sample period, excluding the outliers;

    Page 31


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

      o     

    For the projection of the variable accounts, an indicator of this account was calculated first as a function of Consumer Units, Network Expansion (km) and Consumption, in order to later carry out the projection. The outliers were also excluded.

     

  • Phase 3 - Identification of outliers:

     

      o     

    The values are considered outliers when:

     

        §     

    They are outside the established limits of normality of the sample, i.e.:

     

         
  • Out Upper Limit => (sample mean) + (standard deviation of sample); Out Lower Limit => (sample mean) - (standard deviation of sample).

     

      o     

    After this procedure, the nature of the expense and its periodicity in the last 3 years were assessed in order to evaluate whether or not it would be excluded.

     

  • Phase 4 - Assessment of the best correlations for the variable accounts:

     

      o     

    At this phase, a study was made to verify which variables were more related like the expenses of PMSO. It was concluded that Network Extension (km) and Consumption in MWh obtained higher ratio;

     

  • Phase 5 - Identification of regulatory accounts (considered by and paid by ANEEL): 
          
      o      Consists of the classification of the accounting accounts forming the PMSO according to the characterization defined by ANEEL for regulatory PMSO. Thus, for each component account of the analytical PMSO, a classification between Effective PMSO (total), Regulatory PMSO and other PMSO (Effective PMSO - Regulatory PMSO)  is provided).
           

    Part 2 - Projection of Effective and Regulatory PMSO Grouping (Divided into 4 phases):

    account the fixed or variable nature of the expenses that make up the PMSO;

    Page 32


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

      o      Variable Expenses: are transformed into indicators in which the denominator is the parameter with which the variable had the best correlation (it can be network km or consumption in MWh).
        Example: Personnel = R$/km of network => item P will increase as the network km increases;
      o      Fixed Expenses: fixed expenses were calculated as the average expenses for the sample period, excluding outliers.
  • Phase 2 - Operating Costs variation rate: This phase is divided into three
      steps:     
      o      Step 1 - Definition of Operational Cost Benchmarks;
        §      For Northern Companies (CERON, Boa Vista Energia, Eletroacre and Amazonas Energia) => Celpa and Celtins;
        §      For the Northeast Companies (Cepisa and Ceal) => Celpe, Cemar, Coelce, Energisa Paraíba, Energisa Sergipe, Coelba and Cosern.
      o      Step 2 - Rate Variation Definition and Target between 2018 and 2022:
        §      It is considered the DEA (Data Envelopment Analysis in Portuguese) database of efficient costing from Aneel for 4CRTP;
        §      The PMSO expense of each company is parameterized by the km of Network Extension;
        §      Based on the indicators, the average of the group of each distribution company according to the Region is calculated;
        §      After this step, proceed as follows:
         
  • Distribution companies with the indicator greater than the group mean adjust their cost in terms of indicator until the next Tariff Review in 2023.
         
  • On the other hand, if the indicator is less than the mean of the group, the Distribution companies plot the indicator level between 2023 and 2027 as target, which is lower than the group mean.

    Page 33


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

      o     

    Step 3 - Rate Variation Definition and Target between 2023 and 2027:

     

        §     

    It consists of calculating the mean indicator of the privatized Benchmarks of the groups;

     

        §     

    Upon reaching the indicator of R$/km of network of the cycles target, the distribution companies stabilize operating cost indicators parameterized per consumer unit.

     

  • Phase 3 - Identification and Treatment of Accounts with Behavior Linked to the Distribution Companies Improvements:

           
      o      It consists of identifying accounts that influence the actual cost of PMSO, but which varyaccording to the distribution company’s operational improvements.
               
  • Phase 4 - Projection of the Concession Period (2018 to 2047): It consists of estimating the levels of PMSO expenses of Distribution Companies over the concession period.
             
      o     

    Based on the targets set previously, the Distribution Companies reach the operational cost benchmarks until the end of 2027;

     

      o      Thereafter, the network PMSO/km indicator stabilizes itself, as a consequence, the absolute value of the PMSO varies in real terms through the variation of the Network km Extension.
  • Costs:

    The elaboration of the estimate of the efficient operational costs was based on Annex I -Efficiency Intervals of the PRORET sub-module 2.2 V2. Operating costs tend to align with their respective benchmarks in two moments, the first ending 2022 and the second in 2027.

    Calculation Method:

    - First moment => simple mean of each of these limits for distribution companies according to their regional groups, which are:

    Page 34


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    graph. For information the values are: R$ 191 million against R$ 181.9 million (2018) and R$ 196 million against R$ 189.2 million (2019);

    - By means of the graph note that the projections of the regulatory PMSO are larger than those of the actual PMSO.

    - The difference between the VPL of the two items, in real terms, discounted by the WACC rate of the model, considering in simplified form, a tax shield of 34%, results in a value of R$ 408.3 million, that is, for Ceron, PMSO does not add value for the economic and financial assessment;

    - Another important point is that the projection of the regulatory PMSO values took into account the technical note 149/2017 - ANEEL.

    2.5.3      PMSO FOR SERVICE B:
    -      Historical part:

    Service B did not present the rational to compose its database from 2012 to 2016 as did Service A. The report shows the projections of P, M, S and O values directly for the period from 2017 to 2047.

  • Methodology used in Service B (common to all Distribution Companies):
     
  • Database: Calculation of Efficiency for the 4CRTP - 4th Cycle of Periodic Tariff Reviews (PMSO, Market, Consumers, network, losses, etc. data) => 2001-2013;
     
  • Clustering of Companies: Automatic grouping of similarity second degree data. The result of this process can be seen in the table below:

    Page 36


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 23.  Clusters composition and efficiencies - Ceal   
     
    Companies Efficiency
    Indicator
    of
    4CRTP
    efficiency 
    Efficiency
    Indicator 2011
    Efficiency
    Indicator
    2013 
    %
    (p.a.)
    COELCE  100%  0.44  0.43  0.5% 
    Celtins  100%  0.44  0.51  -8.5% 
    CELPE  86%  0.74  0.88  -8.7% 
    Bandeirante  82%  0.90  0.92  -1.2% 
    Ampla  71%  1.03  0.90  6.2% 
    EDVP  70%  0.97  0.66  17.8% 
    CELG  69%  0.64  0.54  8.2% 
    CEEE  42%  1.21  1.67  -17.4% 
    CEAL  44%  1.87  1.58  8.0% 

     

    Understanding the indicator - For example, Ceal in 2011 obtained an indicator of 1.87 and in 2013, this value decreased to 1.58. This means that the company has improved slightly from 2011 to 2013 as it needs 1.58 (2013) PMSO units to offer a set of services (weighted combination of Market, Network, Consumer, Non-technical Losses).

    Projection of the trajectory :

    1.      Phase 1: 2017-2018 => No Efficiency gain
    2.      Phase 2: 2019-2020 => It follows the rhythm of the utility with the greatest gain of efficiency of the Cluster until reaching the median of the indicators of the group
    3.      Phase 3: 2021-2048 => Convergence for the mean of the Companies of the 1st Portion of the Group

    Thus, the Ceal efficiency curve as well as the projection for the regulatory PMSO are presented in the graph and table below:

    Page 37


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Although Service B shows an Efficiency curve that should be followed in the document, it is not evident the comparison with the effective PMSO values projected by the consultancy. Therefore, it is not possible to state when the company is able to reach the level of regulatory PMSO, as well as to estimate if this trajectory will generate or destroy value for the Company.

    In addition, it is important to take into account the issue of relaxation, since according to Technical Note No. 149/2017 - ANEEL, which deals with the flexibility of some regulatory parameters, the new values for the Regulatory PMSO that go until the first ordinary tariff review of the new controller of the designated companies, are:

    Table 25.  Flexibility of the PMSO (Aneel Technical Note No. 149/2017) 
            R$ Million 
    Distribution
    Company
    2016 Tariff Process Relaxed
    Regulatory
    Level 
    Difference Difference
    (%)
     AME  348.8  627.2  278.4  80% 
    Boa Vista  37.5  69.1  31.5  84% 
     CEAL 271.7  336.7  65.0  24% 
     Cepisa  360.7  420.1  59.4  16% 
     CERON  234.0  279.9  45.8  20% 
    ELETROACRE  93.5  128.5  35.1  38% 
     CEA 67.8  124.8  56.9  84% 

     

    Page 39


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

    2.5.4      PROJECTIONS AND COMPARISONS OF THE PMSO OF SERVICES A AND B
    -      Projections of CEAL until 2047:
    Table 26.  PMSO Projection Data (Service A)   
                    R$ Million 
    Item  2017  2018  2019  2020  2021  2022  2023  2024 
    Personnel  178.6  181.9  189.2  196.7  204.4  212.1  221.0  230.2 
    Material  2.0  2.5  2.6  2.7  2.8  2.9  3.0  3.2 
    Third Party Service  99.6  101.7  105.6  109.6  113.7  117.9  122.9  128.1 
    Other  8.6  8.8  9.1  9.4  9.8  10.1  10.5  10.9 
    Total PMSO  288.8  294.9  306.5  318.4  330.6  343.1  357.5  372.5 
    Item  2025  2026  2027  2028  2033  2038  2043  2047 
    Personnel  239.8  249.7  259.9  273.0  349.4  448.1  575.9  698.9 
    Material  3.3  3.5  3.6  3.9  5.5  7.6  10.7  13.7 
    Third Party Service  133.6  139.2  145.0  152.2  194.0  247.7  317.0  385.5 
    Other  11.3  11.8  12.3  12.8  15.8  19.5  24.0  28.4 
    Total PMSO  388.0  404.1  420.8  441.9  564.5  723.0  927.6  1,126.5 

     

    In addition, Service A estimated a value for the Voluntary Resignation Plan (PDV), defining the target public (employees with at least 20 years of company), the success rate, its cost (R$ 14,5 million) and the date of resignation of employees, June/18 and June/19.

    Table 27.  PMSO Projection Data (Service B)   
                    R$ Million 
    Item  2017  2018  2019  2020  2021  2022  2023  2024 
    Personnel  165.5  175.1  176.1  176.1  176.8  177.9  179.4  180.9 
    Material  1.9  2.0  2.0  2.0  2.0  2.0  2.1  2.1 
    Third Party Service  100.2  106.0  106.6  106.6  107.1  107.7  108.6  109.6 
    Other  29.1  30.8  30.9  30.9  31.1  31.3  31.5  31.8 
    Total PMSO  296.7  313.9  315.7  315.6  316.9  318.9  321.5  324.3 
    Item  2025  2026  2027  2028  2033  2038  2043  2047 
    Personnel  182.6  184.6  186.7  188.8  208.0  257.9  320.9  382.6 
    Material  2.1  2.1  2.1  2.2  2.4  3.0  3.7  4.4 
    Third Party Service  110.6  111.8  113.1  114.4  126.0  156.2  194.3  231.7 
    Other  32.1  32.4  32.8  33.2  36.5  45.3  56.4  67.2 
    Total PMSO  327.4  330.9  334.8  338.6  372.9  462.4  575.2  686.0 

     

    Page 40


     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Basis on the graph analysis, it can be seen that, concerning the Third Party Service Account, the two projections are virtually the same in the first years, i.e., until 2023, and, from this point onwards, the difference starts growing until the value of R$ 153.7 million is reached in 2047. As for other companies, the projection of the Account Other begins and ends differently for the two consulting firms, reaching the final gap of R$ -38.8 million in 2047. It should be noted that, also for this account, the projection of Service B is larger than that for Service A.

    found, it is noticed that there is a very marked difference in relation to the projections. This fact may be due to the basis that each Consultancy used to make the PMSO projection and also according to the projection model considered. In relation to the first item mentioned above, only Service A presented the rationale for the initial composition of the PMSO.

    expenses, it is concluded that Service B adopted a much more projection than that of Service A. Highlighting again, this difference in values has no significant impact on Valuation if the company's PMSO is within the regulatory level.

      Table 28.  Comparison between the Services   
                R$ Million 
      Service A  % Service B   
     Company  2017  2047  2017  2047   % 
    Eletroacre  112.5  684.9  509%  133.5  327.6  145% 
    CEAL  288.8  1,126.5  290%  296.7  686.0  131% 
    Amazonas  550.1  1,896.4  245%  600.2  1,267.4  111% 
    Cepisa  383.7  1,757.7  358%  452.6  1,122.8  148% 
    CERON  340.1  1,534.8  351%  373.3  866.6  132% 
    Boa Vista  105.6  334.1  216%  216.3  371.7  72% 
    Total  1,780.8  7,334.4  312%  2,072.5  4,642.1  124% 

     

    - Finally, regarding the documents presented by the consultancies, Service A detailed the methodology much more than Service B, presenting in its text more coherence in the presentation of the information.

    Page 44


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.6 INVESTMENT

    Both services contracted for economic and financial assessment adopted different assumptions for the projection of the investment notably from the period 2023 to 2047, since both adopted the Five-Year Plan for the period from 2018 to 2022 of the company. This plan was presented in Product 7 “Ceal Operational Technical Assessment Report” and was prepared based on the “Expansion Plan for the Electrical System (2016-2025 period),” the “Development Plan for Distribution - PDD for the year 2017,” the “Plan of Results for the Improvement of Distribution Services, 2015” and the “Temporary Provision Plan for the Electricity Distribution Service, 2017.” In addition to this, information obtained through the "Field Visit Report" (Annex I of the Technical and Operational Assessment Report of Ceal) and the incremental investment needs projected by the company's planning area, obtained during face-to-face meetings and conference call carried out. The only caveat made for this period is that Service A for the investment in “Luz para Todos” program considered only 30% of the value since 70% is subsidized by CDE, and therefore treated as Special Obligations.

    In relation to the other five-year period, the main differences are due to the fact that Service A included, in addition to the projections made for implementation and renovation projected from the Five Year Plan, other analyzes to define the investment in maintenance and replacement of partially and totally depreciated assets . Whilst Service B was restricted only to the first analysis.

    For the projection from 2023 of investments in implementation and renewal, Service A adopted as a basic assumption for the expansion of HV, AV and LV as well as for the improvements the mean of the last years contemplated in the plans of the companies and this value was kept constant until the end of the concession period1. Service B, for electric investments (expansion and improvement), also estimated the second five-year period of the concession by the mean of the period from 2018 to 2022, however, it eliminated values outside the range between the mean and +/- a standard deviation. Thus, in order to obtain the amount of investment required for the five-year period, the investments of

    1 The expansion in HV was available for information until 2026 by means of the Distribution Development Plan

    2017 (PDD 2017) and then the average of the investments from 2020 to 2026 was adopted for the other years. As for the investment in the expansion of AV and LV, the average of expenditures from 2019 to 2022 was applied.

    Page 45


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    the years considered typical were added up and multiplied by the factor 5/No. typical years. For the other five-year period, they adopted as base the previous five-year period with the caveat that for the expansion in AT, it is considered that every 10 years the need for investments doubles in relation to the previous five-year period. As for the non-electric investments (infrastructure), it was assumed that every 5 years it invests 50% of what was forecast in the first five years. Service A, however, did not project them.

    For renovation investments (maintenance), since 2023, service A considered that investments of this type are equivalent to the value of the assets depreciation. To define this value, the depreciation reference of ANEEL's Normative Resolution 674/20152, was used for the current asset base plus the depreciation of the new investments of the company, calculated from the average depreciation of the asset base of the Assets Control Report. These investments were accumulated and concentrated in the two years before each tariff review according to the proportion of the reference companies.3 Service B, however, first calculated the share of the amount destined to the investment in renovation within the period of the Five-Year Plan 2018-2022. Subsequently, it applied this percentage for all other years.

    For investments in replacement, only service A estimated them. Based on the assets base of 2016, the estimated investments for implementation and renovation from 2017 to 2022 were added, as well as the inflation adjustments and application of the depreciation of each item, based on Aneel Normative Resolution 674/2015, both for the assets base of 2016 and for the new investments projected up to 2022. In this way, the company's asset base of 2022 was formed. Therefore, the indicator Market Use Value (VMU)/New Value of Replacement (VNR) was calculated for the company and for the benchmark companies4 so that the distribution company under review reaches the same percentage of the private reference distributed between 2023 and 2027 by the average proportion of investment of the reference companies5.

    Given the assumptions and their differences, we find the following values for the two consultancies for Ceal.

    2 Resolution that approved the review of the Assets control manual of the Electric Sector - MCPSE.

    3 Ceal - Cemar, Coelce, Coelba e Celpe.

    Cemar; Energisa Sergipe and Energisa Paraíba

    5 Cemar, Coelce, Coelba and Celpe.

    Page 46


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 30.  Comparison of Investment carried out from 2012 to 2016, values of Dec./2016
     
                  R$ Million 
                  Five-Year Period 
    Investment    2012  2013  Cost of services  2015  2016   
                  2012-2016 
    Updated Values (Dec./16)  141  142  105  146  166  699 

     

    2.7 REGULATORY REMUNERATION BASE

    Both consortiums used the data contained in the valuation reports of the assets as reference to the date of February 28, 2017 to prepare the projection of the Regulatory Remuneration Base (BRR). In these reports, both closed BRR and the incremental BRR from 3CRTP to 4CRTP were assessed. In the case of Ceal, the following reports were used:

    Both service A and B adopted different projection criteria for the asset base unitized through Feb/17 (closed BRR - starting point for 4CRTP) and for unitizations occurring after Feb/17 and new investments (incremental BRR).

    Regarding the starting point for 4CRTP, both consortia use the full assessment report (delivered on 07/31/2017), however, they present different criteria for the definition of Fixed Assets in Use (AIS).

    For service A, the disallowance is treated in two ways: i) Disallowance over the reviewed closed base (Closed Base of Full Report - Closed Base of the Incremental Report) in the value defined in 5,73% for the companies of the North and 5.01% for the Northeast, based on the average adjustment (disallowance) of the distribution group assessed in the 3CRTP by region, i.e., this disallowance is made only in the AIS of the difference found in the opening of the closed base compared to that approved by Aneel in 3CRTP monetarily

    Page 48


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    corrected; and ii) The incremental basis between 3CRTP and 4CRTP - the lowest value was applied between the average adjustment of the distribution companies group assessed in 3CRTP by region and that verified by the company, without showing the amount applied in the Assessment Report.

    Unlike Service A, Service B used only a disallowance methodology to reach the Remuneration Base of Feb/17 of the Distribution Company. The disallowances were performed between the difference in AIS of the two assessment reports of the Closed Base and the Incremental Base. This difference would be the opening of the Closed Base. That is, the disallowance for Service B was performed only at the opening of the Closed Base and occurred in the AIS. The percentage of disallowance was calculated from the relation between the final AIS (with disallowances) and the initial AIS (without disallowances) of the incremental BRR of the distribution companies 3RTP of Eletrobras group, obtaining a mean of 9.2%. If the company has a disallowance lower than this mean, it is considered the mean of disallowance percentage itself (9.2%). Otherwise, the mean between the percentage of the company and the maximum percentage of disallowances (20.1%) is used. In the case of Ceal, it falls within the first criterion, since it presented a reversal of 2.2% of the AIS in the 3CRTP, adopting, then, the reversal of 9.2% for the Closed Base. This disallowance percentage is applied only in the difference between the AIS of the two reports and added to the value of the base of the incremental report. That is, the disallowance of 9.2% is not applied in the total AIS of the Complete Base, it is calculated only for the increment of the closed base.

    Therefore, the following information was obtained for the Closed base, in the position of February 2017, for both services:

    Page 49


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 31.  Closed base, in the position of Feb./2017 - Services A and B 
     
          R$ Million 
        Armored Base   
      Description     
        Service A  Service B 
    (1) Property, Plant And Equipment in Service (New Replacement Value)  2,211  2,195 
    (2) Index of Full Utilization  1  1 
    (3) Gross Special Obligations  550  550 
    (4) Fully Depreciated Assets  335  335 
    (5) Gross Compensation Reference = (1) - (2) - (3) - (4)  1,326  1,309 
    (6) Accumulated Depreciation  1,099  1,099 
    (7) Net AIS (Market Value In Use)  1,112  1,096 
    (8) Index of Depreciated Utilization  0  0 
    (9) Compensation Reference Value (VBR)  1,112  1,096 
    (10) Warehouse in Operation  6  6 
    (11) Deferred Assets  -  - 
    (12) Special Net Obligations  414  414 
    (13) Lands and Easements  8  8 
     
    (14) Base of Total Net Compensation = (1) - (6) - (8) + (10) + (11) - (12) +  712  695 
    (13)       
    (15) RGR PLPT Balance  68  57 
    (16) RGR Balance Other Investments  211  176 
    (17) Depreciation Rate  4%  4% 
    (18) Share of Regulatory Reintegration = (5) * (17)  53  48 
    (19) Actual WACC before Taxes  11%  ND 
    (20) RGR PLPT Rate  1%  ND 
    (21) RGR Rate of Other Investments  4%  ND 
    (22) Capital Compensation (RC) = (15) * (20) + (16) * (21) + [(14) - (15) -     
    (16)] * (19)    58  ND 

     

    For the years subsequent to the 4CRTP Starting Base, both Services apply adjustment for inflations for the IPCA, additions and write-offs of the assets that were considered in the valuation until the end of the concession period.

    However, it should be noted that both Services did not incorporate the positions of Fixed Assets in use into the Remuneration Base, since, according to the services, the assessors’ report of the Remuneration Base does not take into account the specific assessment of the condition and characteristic of these assets concerning (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the reassessment process (such as the calculation of VNR, depreciation, utilization rate and valuation of write-offs). In regards to the case of Ceal, it should be noted that the Full Assessment Report

    “Companhia Energética de Alagoas - CEAL project no. 2715-17745,” performed by Levin,

    Page 50


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    states, as the position of June 2017, the balance of R$ 177.4 million for the Fixed Assets in Use.

    In addition, Service B segments the assumptions for the projection of BRR transactions into two groups. The first group operated the closed BRR by the IPCA adjustment for inflation and the discount of the depreciation and write-off of the assets that went into operation until February 17 and in a second group operates the incremental BRR (new investments) through the additions of assets projected after Feb/2017, less the losses related to these assets.

    Thus, from 2018, service A monthly corrects, through the IPCA, the items Fixed Assets in Use, Warehousing in operation and land and easements and increases them according to their projection of the investment. The gross special obligations item is also updated by the IPCA and by the increase in projected special obligations. On this incremental Base, the average adjustment of privatized benchmark distribution companies was applied as disallowance by region obtained through the 3CRTP Incremental Base. For the Northeast companies, Celpe, Energisa Paraíba and Energisa Sergipe were considered, giving a value disallowance of 2.09%. For the North companies, the disallowance value was 2.83%, using Celpa and Cemar as benchmark. In relation to the depreciation rate, the same verified in 3CRTP was maintained for each distribution company assessed. In relation to asset write-offs, the ratio of fully depreciated assets to fixed assets in use observed in the 4RTP BRR resulted in an equilibrium rate calculated as a simple mean between Eletrobras' companies. Thus, starting from the company's ratio, it varies at a constant monthly rate until reaching that equilibrium rate of 12.54% at the end of the concession.

    Service B, as mentioned earlier, operated BRR with different premises for the base that served as a starting point and the one that incorporates the new investments. The transaction of the closed BRR was made by monetary restatement by the IPCA and by the discount of depreciation and write-offs. The depreciation rate, unlike service A, is not constant, the rate falls from the average level calculated in Feb/17, as the assets are being written-off. With respect to the asset write-offs, the average write-off/AIS ratio was applied to the AIS of each month based on the information of the incremental assessment report prepared by Levin in the period between 3CRTP and Feb./17. The incremental BRR transaction (new investments) was made based on the projection of the investments rejected in 5%. The depreciation rate used was obtained by the weighted mean of the investments of the Five-Year Plan 2018 to 2022 per group of assets and their respective depreciation rates, obtaining 3.15% of the AIS. This rate was considered constant

    Page 51


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    throughout the concession. Regarding the write-offs, until Feb/27, they were considered null in view of the reduced probability of assets being written off in the first 10 years. After this period, 1/3 of the BRR Closed/AIS write-off ratio was applied in a staggered manner every 5 year until reaching the value of the integral ratio as of Mar/42.

    Accordingly, given the differences in the investment projections, the criteria for disallowance, depreciation and write-offs, we obtained at the end of the concession (Feb / 48) the following values of Net Remuneration Basis for Ceal according to the projections of the two services:

    Table 32.  Net Remuneration - Services A and B 
          R$ Million 
    Description    Service A  Service B 
    Net Remuneration Base (Feb./48)  14,032  5,243 

     

    2.8 X FACTOR

    The primary objective of the X-Factor is to ensure that the balance established in the tariff review between efficient revenues and expenses is maintained in subsequent tariff repositioning. This occurs through the transfer to the consumer of potential productivity gains from the distribution segment.

    The approach adopted by Aneel adds to the X Factor a mechanism to encourage the improvement of the technical and commercial quality of services provided to the consumer. It also introduces a component that seeks to establish the gradual transition of the operating costs of the utility towards its efficient costs.

    The X Factor adjusts the value of Installment B, thus affecting the tariff and, consequently, the revenue and the collection of the Distribution Company. Therefore, the X factor value inversely impacts the value of the VPL (the higher the X factor, the lower the revenue and the VPL).

    Page 52


     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 35.  Methodology used - Service A 
     
      3rd CRTP    New Methodology   
      Methodology       
        Apr/16  Apr/17  Apr/18  Apr/19 
    Indicator  Apr/15 to  to  to  to  to 
      Mar/16  Mar/17  Mar/18  Mar/19  Mar/20 
    DEC  50%  30%  37.5%  45%  50% 
    FEC  50%  30%  30.0%  27%  20% 
    INS      0.8%  2%  4% 
    ICO      0.4%  1%  3% 
    IAb      0.4%  1%  3% 
    FER      3.0%  7%  10% 
    IASC      3.0%  7%  10% 
    Total  100%  60%  75%  90%  100% 

     

    3 - Component T - Trajectory of operating costs

    Component T of the X-Factor is applied to Installment B so that the tariff increases or decreases so that the recognized regulatory operating costs reach the efficiency threshold defined by Aneel.

    In defining efficient operating costs, the distribution company’s costs, the efficient level of costs and the characteristics of the concession area are observed. Aneel uses a model whose objective is to define a feasible production/cost frontier as a way of identifying the level of efficiency of companies.

    In a simplified way, the identification of the efficient level of costs is obtained by comparing the distribution companies by means of a benchmarking method that takes into account the attributes of each utility. From these conditions, a target of efficient operational costs is established to be reached throughout the tariff cycle. At the time of the tariff review, the regulatory goal is compared to the operational costs coverage present in the utility´s tariff, called operating costs revenue. From the difference between the regulatory goal and the operating costs revenue, a regulatory trajectory will be calculated. Part of the difference will be incorporated at the time of the tariff review and the remaining installment will be considered for the purpose of calculating component T of the X Factor.

    Page 56


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Services A and B, following the provisions of Aneel's Technical Note No. 149/2017, used a zero value for component T in the period between the date of signature of the agreement and the first ordinary tariff review, that is, from 2018 to 2022. From 2023, Services A and B found very divergent values for Component T of X Factor.

    2.9 WACC

    Services A and B adopted the method of Capital Asset Pricing Model - CAPM to estimate the return that the shareholders expect to arrange with the capital employed in the business, for discounting the current value of the distribution company cash flow. It should be noted that this method is well-established and widely used, either by the market at large, and by Eletrobras.

    In the calculation of the Weighted Average Cost of Capital WACC, Services A and B applied different approaches of the CAPM method, thus allowing quite different results.

    Below are the main differences found in the calculation of WACC by both Services.

    Capital Structure

    Beta

    ·Service A: used comparable domestic companies. Unlevered Beta: 0.68;

    ·Service B: considered comparable domestic companies, with a larger sample. Unlevered Beta: 0.43.

     

     
     

    Ke

    ·Service A: applied the CAPM method and a “foreign exchange risk” of 4.17% at nominal Ke in Dollars to achieve a nominal Ke in Brazilian Reais. Therefore, this “foreign exchange risk,” which derives from the “Future DI” rate and “Fraction of

     

     

    Page 57


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    exchange rate coupon,” works as a differential of inflation. To achieve such rate in

    real terms, it was deflated by IPCA (4.23%);           
    · Service B: applied the CAPM method.             
     
    The table below shows the result for Ke under real terms:         
     
      Table 36.  Profit or Loss of Real Ke Services A and B     
    Actual Ke  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  7.95%  7.81%  7.53%  7.21%  7.02%  6.85%  6.70%  6.55%  6.43%  6.31%  6.31% 
    Service B  11.73%  10.16%  10.18%  10.42%  10.42%  10.42%  10.42%  10.42%  10.84%  10.84%  10.79% 

     

    Kd

    The table below shows the Kd result in real terms after taxes.         
     
      Table 37.  Kd in real terms after taxes Services A and B     
                         
    Actual
    Kd 
    2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
                         
    Service A  4.25%  4.09%  4.27%  4.19%  4.19%  4.19%  4.19%  4.19%  3.25%  3.25%  3.25% 
    Service B  4.44%  3,34%  3.36%  3.29%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25% 

     

    WACC

    Based on the values of Beta, Ke and Kd above, Services A and B calculated WACC in real terms after taxes, according to the results included in the table below.

          Table 38.  WACC Services A and B       
    WACC  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  8.09%  7.20%  7.30%  7.39%  7.39%  7.39%  7.39%  7.39%  7.60%  7.60%  7.58% 
    Service B  5.29%  4.80%  4.89%  4.87%  4.93%  5.00%  5.06%  5.12%  5.19%  5.26%  5.26% 

     

    Out of the results stated above, it should be noted that the factor with a higher impact on the rate presented by Service A was the use of the foreign exchange risk (4.17%) in a

    Page 58


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    heterogeneous manner to the market practices, as a differential of inflation, as this is applied on a Ke (nominal in USD), in a scenario where the CPI US inflation is 2.30% and IPCA, 4.23%. The result of this is WACC, in real terms, which is higher compared to the number obtained in calculation by Service B.

    3. THE RESULT OF ECONOMIC AND FINANCIAL ASSESSMENTS AND

    BY MULTIPLES

    Service A's assessment calculated the enterprise value from R$ 1,994,373,551.48 to Ceal, considering, as already mentioned, the granting of a new concession. After deducting the debts, liabilities with suppliers and contingencies, Ceal´s equity value calculated by Service A is negative R$ 1,116,190,768.34, as shown below. To obtain these values, the consultancy used the WACC of 7.57%.

    Table 39.  Enterprise value Service A   
     
        Free Cash Flow to Firm (FCFF) 
    Base date of Enterprise Value    Feb./2018  Dec./2016 
    Adjustment on Enterprise Value    -  IPCA 
    Enterprise Value (BRL)  A  2,098,898,814.44  1,994,373,551.48 
    Cash (BRL)  B  36,484,281.39  36,484,281.39 
    Gross Debt (BRL)  C  (1,732,032,633.98)  (1,732,032,633.98) 
    Net Debt (BRL)  B+C  (1,695,548,352.59)  (1,695,548,352.59) 
    Tax, Labor and Social Security Contingencies (BRL)  D  (22,670,000.00)  (22,670,000.00) 
    Legal Contingencies (BRL)  AND  (1,364,643,680.61)  (1,364,643,680.61) 
    Actuarial Adjustments (BRL)  F  1,879,692.53  1,879,692.53 
    Environmental Adjustments (BRL)  G  (29,581,979.15)  (29,581,979.15) 
    Valuation (R$)  A+B+C+D+E+F+G  (1,011,665,505.38)  (1,116,190,768.34) 

    Source: Assessment Report of Eletrobras Distribuição Alagoas, Service A.

    Service B calculated the enterprise value of R$ 2,446,562,300.47 for Ceal, in view of the granting of a new concession. After deducting the debts, liabilities with suppliers and contingencies, Ceal equity value calculated by Service B is negative R$ 664,002,019.35, as shown below.

    Page 59


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 40.  Enterprise value - Service B   
    Result    Dec./16 
    Operation NPV    2,153,270,399.68 
    NPV of Terminal Values    293,291,900.79 
    Enterprise Value    2,446,562,300.47 
    (-) Net Revenue    -1.695.5487.352,59 
    (-) Contingencies    -1,385,433,988.08 
    (-) Environmental Adjustments    -29,581,979.15 
    Equity Value    -664,002,019.35 

     

    Source: Product 10: Ceal Conclusive Economic and Financial Assessment Report, Service B.

    In order to obtain these values, Service B used rolling WACC, which considers a movable (leveraged) capital structure, as follows.

      Table 41.  Rolling WACC - Service B         
     
    Parameters  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  > 2026 
    Rate Free from Risk (Rf)  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78% 
    Country Risk  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53% 
    Market Premium  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90% 
    Non-Leveraged Beta (²u)  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43 
    % Debt (Kd) - [D/(D+E)]  54.0%  50.7%  47.3%  44.0%  40.6%  37.3%  33.9%  30.6%  27.3%  23.9%  23.9% 
    % Own Capital (Ke) - [E/(D+E)]  46.0%  49.3%  52.7%  56.0%  59.4%  62.7%  66.1%  69.4%  72.7%  76.1%  76.1% 
    Taxes - IRPJ/CSLL  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0% 
    Leveraged Beta (²L)  0.76  0.72  0.68  0.65  0.62  0.59  0.57  0.55  0.53  0.52  0.52 
    Premium Size  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00% 
    Inflation - Brazil  4.33%  4.43%  4.36%  4.31%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23% 
    Inflation - USA  2.40%  2.30%  2.30%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40% 
    Cost of Own Capital  12.62% 12.55% 12.22% 11.83% 11.55% 11.37% 11.21% 11.06% 10.93% 10.81%10.81% 
    Selic  10.71%  9.07%  8.87%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73% 
    Spread  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64% 
    Cost of Debt  11.35%  9.71%  9.67%  9.54%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37% 
    Debt Cost - Net of Taxes  7.49%  6.41%  6.38%  6.28%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18% 
    WACC in current CRL (Post-Tax)  9.85%  9.44%  9.46%  9.39%  9.37%  9.44%  9.50%  9.57%  9.64%  9.71%  9.71% 

    Source: Product 10: Ceal Conclusive Economic and Financial Assessment Report, Service B.

    With respect to the multi-market valuation, Service A, in its Assessment Report of Eletrobras Distribuição Alagoas, attached to this Appendix, obtained a Ceal enterprise value equal to R$ 1,693,876,000, which represents a difference of 18% in connection with the value of the economic and financial assessment.

    Service B, however, in Product 10: Ceal Conclusive Economic and Financial Assessment Report, attached to this Appendix, did not obtain the value of enterprise value, but performed comparative analyzes of different implicit multiples to the result of the Discounted Cash Flow assessment.

    Page 60


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    3.1 AFAC

    The study presented considers that all Advances for Future Capital Increase - AFAC’s made by Eletrobras in Ceal and which are accounted for will be converted into debt or loan. On October 17, 2017, Ceal had R$ 178,0 million in AFAC accounted for. The AFAC contracts already provide for the conditions to be considered in case of conversion into debt. Accordingly, Eletrobras must convert the AFACs into debt, according to a loan agreement executed under the following conditions:

    4.      CONTINGENCIES
    4.1      LEGAL CONTINGENCIES

    Ceal´s legal due diligence, performed by Service B and listed in Product 4: Due Diligence Legal Report - Ceal, pointed out some critical points, although such points do not impede the process of privatization of the distribution company, the new investor should be aware of them so that it can take the applicable measures.

    4.2 ACCOUNTING CONTINGENCIES

    Among the accounting contingencies, special attention should be paid to the ongoing grievance brought against Ceal by the Union of Urban Industry Workers of the State of Alagoas, as a procedural substitute, to pay salary differences arising from the economic measures established by Decree Law No. 2335 / 87 (Bresser Plan).

    For this grievance, a high risk of loss (probable loss) was ascertained, amounting to about R$ 1,406,687,677.62 (restated up to December 2016). Owing to the amount and its risk of loss, the grievance represents a potential risk and is an impediment to Ceal’s privatization process. The management of Ceal made a proposal of agreement with the

    Page 61


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    union amounting to R$ 129 million.

    Thus, the sales modeling considers two scenarios,

    described as follows.

    Scenario 1 considers the execution of an agreement with respect to the Bresser Plan proposed, with ending cost amounting to R$ 129,738,947.39. Thus, the difference of R$ 1,279,460,689.21 obtained between the full amount and the adjusted amount is reincorporated into the distribution company’s equity value, since, in the economic and financial analyzes carried out by the consultants, the amount under discussion was fully reduced.

    In turn, Scenario 2 considers the full value of the grievance of R$ 1,406,687,677.62, and there are no adjustments in the results of economic and financial assessments.

    5. CAPITAL STRUCTURE

    The developed structure is aimed at enabling the sale of the distribution companies of Eletrobras in the business model associated with the grant of the concession.

    As mentioned in the item Contingencies, the Ceal’s sales modeling considers two scenarios: (i) Scenario 1, which implies an adjustment of the share value connected to the Bresser Plan, with the final cost amounting to R$ 129,738,947.39; and (ii) Scenario 2, which assumes the full value of the share of R$ 1,406,687,677.62, and there are no adjustments in the results of the economic and financial assessments.

    The table below shows the current structure of Ceal.

    Table 42.  CEAL's capital structure on 12/31/2016 
    Total Shares      734,753,835 
    Capital  R$  734,754,000.00 
    Par Value of Shares  R$    1.00000022 
     
    Interest    Shares  (%) 
    Eletrobras    734,753,835  100.00% 
    Minority    0  - % 
     
    Classes of Shares    Shares  (%) 
    Common    722,322,535  98.31% 
    Preferred    12,431,300  1.69% 

    Source: Product 11: Ceal Privatization Modeling Proposal, Service B.

    Page 62


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The economic value of the company's shares was determined as of 12/31/2016. Then, the definition of values was performed, considering the mean values between Services A and B, as shown in the table below.

    Table 43.  Valuation of CEAL Average of Services A and B 
    Service A    R$  -1,116,190,768.34 
    Service B    R$  -664,002,019.35 
    Average of Services A and B  R$  -890,096,393.84 

     

    Source: Product 11: Ceal Privatization Modeling Proposal, Service B.

    5.1 SCENARIO 1

    In this scenario, the equity value was adjusted, amounting to R$ 1,279,460,689.21, connected to the Bresser plan, reincorporated into Ceal’s equity value. In addition to this, the adjustments made in Item 6.3.1 of IT were performed, amounting to negative R$ 3.5 million. Thus, the new equity value is R$ 385,801,111.32.

    Table 44.  Adjusted equity value 
    Average of Services A and B  R$  -890,096,393.84 
    Reversal of Bresser Plan  R$  1,279,460,689.21 
    Consolidated Adjustments  R$  -3,563,184.05 
    Equity Value Adjusted  R$  385,801,111.32 

     

    Source: Product 11: Ceal Privatization Modeling Proposal, Service B.

    As already referred to in Item 8.3.2 of IT, in cases of positive equity values, the relaxation of regulatory parameters should be reduced to result in an equity value equal to zero. For this scenario, the percentage reduction in the relaxation of Ceal was 99.2%, according to the table below.

    Table 45.  Equity value equal to zero, owing to the adjustments in the 
    relaxation of parameters
    Equity Value Adjusted  R$  385,801,111.32 
    Adjustment of the reduction of tariff relaxation  R$  -385,801,111.32 
    Ending Equity Value  R$  - 
    Percent of reduction of tariff relaxation    99.2% 

     

    Page 63


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Source: Product 11: Ceal Privatization Modeling Proposal, Service B.

    Upon performance of the adjustments above, one has the Ceal's capital structure after sale.

    Table 46.  Ceal's capital structure after sale - Scenario 1   
    Total Shares      8,918,650,838,435 
    Capital    R$  1,341,626,910.40 
    Par Value of Shares    R$    0.00015043 
     
    Interest  Common Shares  Preferred Shares  Total Shares  (%) 
    Investor  7,886,765,772,863  134,580,015,440  8,021,345,788,303  89.9390% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retirees  880,990,412,856  16,314,637,275  897,305,050,131  10.0610% 
    Other Minority  0  0  0  - % 
    Shares per Class      Total Shares  (%) 
    Common      8,767,756,185,720  98.3081% 
    Preferred      150,894,652,715  1.6919% 
     
    Source: Product 11: Ceal Privatization Modeling Proposal, Service B.     

     

    5.2 SCENARIO 2

    As already reported, Scenario 2 considers all liabilities arising from the Bresser Plan, amounting to R$ 1,406,687,677.62, and it is not necessary to make valuation adjustments referring to this amount, since the valuation already provide for it. As in Scenario 1, the adjustments described in Item 8.3.1 of IT were also made, amounting to negative R$ 3.5 million. Thus, the new equity value is negative R$ 893,659,577.89.

    Table 47.  Adjusted equity value 
    Average of Services A and B  R$  -890,096,393.84 
    Reversal of Bresser Plan  R$  - 
    Consolidated Adjustments  R$  -3,563,184.05 
    Equity Value Adjusted  R$  -893,659,577.89 

     

    Source: Product 11: Ceal Privatization Modeling Proposal, Service B.

    For this case, in view of the negative valuation, the flexibility of the regulatory parameters was maintained, as in the economic and financial assessments carried out.

    The table below shows the capital structure of Ceal after the sale for Scenario 2.

    Page 64


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 48.  Ceal's capital structure after sale - Scenario 2   
    Total Shares      9,345,197,332,746 
    Capital    R$  1,915,346,597.26 
    Par Value of Shares    R$    0.00020496 
     
    Interest  Common Shares  Preferred Shares  Total Shares  (%) 
    Investor  8,268,054,904,033  140,040,558,222  8,408,095,462,255  89.9724% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retirees  919,031,035,519  18,070,834,968  937,101,870,487  10.0276% 
    Other Minority  0  0  0  - % 
    Shares per Class      Total Shares  (%) 
    Common      9,187,085,939,553  98.3081% 
    Preferred      158,111,393,190  1.6919% 
     
    Source: Product 11: Ceal Privatization Modeling Proposal, Service B.     

     

    5.3      CAPITALIZATION BY ELETROBRAS AND BY THE INVESTOR
    5.3.1      ELETROBRAS

    For Scenario 1, since the equity value is equal to zero, Eletrobras would have to capitalize the distribution company at R$ 50,000.00, in order to reach the minimum value.

    In Scenario 2, in turn, considering the minimum value of R$ 50,000.00, the discharge of the debts by Eletrobras is assumed to happen in exchange for shares, in Stage 1 of the sales modeling. Accordingly, there is a need to capitalize Eletrobras in the amount of R$ 893,709,577.89, as detailed below.

    If Eletrobras chooses to increase its interest in up to 30% in Stage 2, as detailed in Item 6.5 of IT, it will have to capitalize the distribution company at R$ 233.9 million in Scenario 1 and R$ 110.6 million in Scenario 2, which may happen through capital contribution or debt of conversion of debts in shares.

    5.3.2 INVESTOR

    As provided in Stage 2 of the modeling, the investor should make a capital contribution with subscription of new shares in the company. The value of the contribution depends on the scenario to be adopted, Scenario 1 or Scenario 2. Therefore, the estimated

    Page 65


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    contribution for the investor is R$ 545,770,485.33, in Scenario 1, and R$ 258,115,449.84, in Scenario 2.

    6. NECESSARY ADJUSTMENTS AT ELETROBRAS

    According to the modeling proposed, in Scenario 2, to achieve the minimum value of R$ 50,000.00, Eletrobras should capitalize R$ 893.7 million in Ceal.

    Thus, Eletrobras should prioritize the cheapest debts to be capitalized, since other debts will be settled by the new Controller. The table below shows the amount of debt eligible to be capitalized by Ceal per type of creditor and its associated costs:

    Table 49.  Debt amount per creditor and its costs BRL million 
     
    RO + AFAC World
    Bank 
    RGR Related parts Total
    Average Cost  12.00%  6.59%  5.00%  8.75%
    (Selic) 
    Ceal  1,401.33  133.12  54.11  53.00  1,641.56 

     

    As can be seen in the table above, the smallest debts of Ceal and that will be assumed by Eletrobras, to be subsequently capitalized, are connected to the following creditors: i) RGR, which has a cost of 5% per year; ii) World Bank, with a cost of 6.59% and; iii) RO, with an average cost of 12% p.a. Although the debt with related parties presents a cost lower than RO, in the case of Ceal, as it is not very significant, the decision was to capitalize part of the RO’s debt to the detriment of the debt with related parties, owing to

    its ease of operationalization.

    It is important to note that ANEEL must authorize

    Eletrobras to assume RGR's debt without having to remove all of it at the time of the transfer.

    Debts with the World Bank refer to the raising of funds from this bank for the implementation of the "Projeto Energia +". The RGR debts refer to financing of investments obtained before the designation period as temporary public service provider and RGR resources received during the temporary provision. The Ordinary Resources are the transfers of Eletrobras' own funds to the distribution companies through loan, in order to cover operating deficits. Related parts debts were contracted with other distribution companies or Eletrobras companies, except the Holding Company. Among the main parties related to outstanding credits are, for example, Eletronorte, Furnas and Chesf.

    Page 66


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The table below shows the amount of each debt that will be capitalized by Eletrobras, in scenarios 1 and 2.

    Table 50. Capitalization of Debts by Eletrobras - R$ million

    Capitalization  Total 
    With Bresser Plan Agreement  0.05 
    Without Bresser Plan
    Agreement 
    893.70
    RGR  54.10 
    RO + AFAC  706.48 
    IBRD  133.12 

     

    It should be noted that CPPI Resolution no. 20, dated November 8, 2017, considered scenario 1 to define, under Item IV of Article 3, that the adjustments to be performed in Ceal by Eletrobras, upon conversion of debt in capital or assumption of the debts of the distribution company, amount to R$ 50,000.00.

    6.1 FIXED ASSETS IN USE

    As referred to in Item 2.7 of this Appendix, it should be pointed out that the studies performed by Services A and B did not incorporate the positions of Fixed Assets in Use (AIC) into the Remuneration Base since, according to the Services, the assessors' report about the Remuneration Base does not take into account the specific assessment of the condition and characteristic of these assets concerning: (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the reassessment process. The table below shows the Fixed Assets in use, ascertained based on data of interim accounting balance sheets for each distribution company. The consideration of these assets, amounting to R$ 177.4 million at Jun./2017, would improve the results of the economic and financial assessments performed by the Services, and consequently, reduce the amount to be contributed by Eletrobras in the distribution company to reach the minimum amount of R$ 50,000.00, as defined in CPPI Resolution 20/2017.

    Page 67


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    6.2 PLANO BRESSER AGREEMENT

    The General Counsel for the Federal Government, by means of Order W/N, dated September 1, 2017, approved the feasibility of executing an agreement between Eletrobras, Ceal and the Union of Workers of the Urban Industries in the State of Alagoas -STIUEA, amounting to R$ 129.7 million.

    In turn, the Minister of State for Mines and Energy, through Order w/n, dated September 8, 2017, authorized the execution of the agreement in accordance with the Order of the General Counsel for the Federal Government above.

    Through CTA-PR-054/2017 Letter, Ceal presented to STIUEA a proposal of agreement amounting to R$ 178 million, with the difference being nearly R$ 49 million owing to the update between the base date of the value calculated by AGU and the base date of the proposal.

    According to mail STIUEA OF. No. 412/2017, dated November 6, 2017, the Union considered the value proposed by Ceal as insufficient and, on account of that, refused the agreement. In the same mail, the Union presented a counteroffer with values much higher than the one authorized by AGU, requesting the answer of Ceal until November 14, 2017.

    The CPPI Resolution 20/2017 defined that the privatization of Ceal is conditioned to the execution and judicial homologation of an agreement regarding the payment of salary differences arising from the Bresser Plan. And, in turn, Decree 9192/2017, in its Article 5, establishes that a bidding process for grant of distribution shall happen, without the transfer of controlling interest, if the bidding process fails to achieve its original purpose.

    In view of the foregoing, owing to the failure of the agreement with the Union of Ceal’s workers, the company is subject to liquidation.

    7. LIQUIDATION COSTS

    In the 165th EGM, held in July 2016, Eletrobras' shareholders decided, by majority vote, to approve: (i) the non-extension of the concessions of its distribution companies; (ii) the transfer of the ownership control of the distribution companies until December 31, 2017, pursuant to Paragraph 1-A of article 8, of Law 12.783/2013; and (iii) if there is no transfer

    Page 68


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    of control under the terms of the previous item, the adoption of provisions for liquidation thereof.

    Thus, the modeling study carried out by Service B calculated the cost with the liquidation of Ceal, in the case of non-occurrence of item (ii) above or not accepted by Eletrobras of the adjustments made to reach the equity value equal to zero.

    For these cases, Service B developed a comparative analysis between the distribution company's sales value associated with the grant of a new concession and the eventual liquidation value of the company. Thus, it was possible to evaluate if the sale of the said state-owned company would be more attractive than its liquidation.

    Since the sale value considers adjustments to the equity value (equity value equal to zero), the adjustments were also considered in the liquidation value estimate. However, there were exceptions to the Tax and Negative Adjustments, since these do not impact on the liquidation value, and for the AFACs, since these can be paid in the event that the distribution companies are not sold.

    As shown below, in the case of Ceal, the best economic option, either for Scenario 1 or Scenario 2, is to sell the company, considering the value of the assets and the concession. The balance between selling and liquidating is favorable for the first option, given the following tables:

    Table 51.  Comparison between the Sale and Liquidation of Ceal Scenario 1 
    Ending Equity Value  R$  - 
    Value of Liquidation (Prior to Adjustments)  R$  (763,188,817.53) 
    + Compensation  R$  684,931,672.05 
    (-) Net Debt  R$  (1,695,548,352.59) 
    (-) Contingencies  R$  (1,385,433,988.08) 
    (+/-) Payments/Receipts  R$  1,632,861,851.10 
    Value of Liquidation (After Adjustments)  R$  (569,913,103.09) 
    =Value of Liquidation (After Adjustments)  R$  (763,188,817.53) 
    (+/-) Balance Adjustments  R$  (143,028,000.00) 
    (+) RGR PPST Reincorporation  R$  336,303,714.44 
    Ending Equity Value (-) Value of Liquidation (After     
    Adjustments)  R$  569,913,103.09 
     
    Source: Product 11: Ceal Privatization Modeling Proposal, Service B.   

     

    Page 69


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The results presented above corroborate that the sale of the distribution company, associated to the granting of a new concession, is the option that is less onerous to Eletrobras. If there was a simple liquidation of the distribution company under the scope of Eletrobras, the latter would have to bear a loss of R$ 569,913,103.09, compared to the final equity value, which amounts to zero in Scenario 1.

    Table 52.  Comparison between the Sale and Liquidation of Ceal Scenario 2. 
    Ending Equity Value  R$  (893,659,577.89) 
    Value of Liquidation (Prior to Adjustments)  R$  (2,042,649,506.74) 
    + Compensation  R$  684,931,672.05 
    (-) Net Debt  R$  (1,695,548,352.59) 
    (-) Contingencies  R$  (1,385,433,988.08) 
    (+/-) Payments/Receipts  R$  353,401,161.89 
    Value of Liquidation (After Adjustments)  R$  (1,849,373,792.30) 
    =Value of Liquidation (After Adjustments)  R$  (2,042,649,506.74) 
    (+/-) Balance Adjustments  R$  (143,028,000.00) 
    (+) RGR PPST Reincorporation  R$  336,303,714.44 
    Ending Equity Value (-) Value of Liquidation (After     
    Adjustments)  R$  955,714,214.40 
     
    Source: Product 11: Ceal Privatization Modeling Proposal, Service B.     

     

    The results presented above support the assumption that the sale of the distribution company, associated to the granting of a new concession, is the option that is less onerous to Eletrobras, taking Scenario 2 into consideration. If Eletrobras had to simply liquidate the distribution company under the company’s scope, the latter would have to bear a loss of R$ 1,849,373,792.30, an amount greater than R$ 955,714,214.40 compared to the final equity value calculated for the distribution company in this scenario.

    In addition, in the event of liquidation of the distribution company, Eletrobras would not be entitled to the benefits arising from the relaxation of regulatory parameters, which generated a reduction in costs in the order of R$ 500 million for Ceal.

    Lastly, it should be noted that the sale modeling approved by CPPI Resolution no. 20/2017 considered scenario 1 to define, under Item IV of Article 3, that the adjustments to be performed in Ceal by Eletrobras, upon conversion of debt in capital or assumption of the debts of the distribution company, amount to R$ 50,000.00.

    Page 70


     

     

     

    Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    8.      ANALYSIS OF THE BENEFITS AND RISKS OF DISPOSAL
    8.1      BENEFITS
     
    (i)
    The sale of Ceal's shares will result in the adoption, by Ceal, of a legal regime similar to that of other non-state agents operating in the sector, allowing for significant operating gains; 
     
     
     
      (ii) Improvement of Eletrobras' Ebitda and elimination of the need for disbursement of funds by Eletrobras to finance the cash deficit and investments of Ceal, contributing to
    the reduction strategy of the net debt / ebitda indicator; 
     
     
     
      (iii) It prevents the need for capitalization by shareholders to meet the financial sustainability targets imposed by ANEEL for the extension of the concession, and for investments in the expansion and improvement of service quality; 
     
     
     
      (iv) The sale together with the granting of concession allows the use of the relaxation of the regulatory parameters approved for the period from 2018 to 2023, when the tariff review of the new concession contract will occur; and 
     
     
     
      (v) This flexibility has led to an improvement in the equity value of the distribution company of around R$ 500 million. 
     
     
    8.2      RISKS
      (i)      Possibility of loss in grievance connected to the Bresser Plan; 
     
      (ii)      Frustration of the disposal of the shares or non-performance of the auction, due to several factors, notably due to possible judicial, political, economic-financial, regulatory, governmental or administrative impediments;
     
    (iii)      Absence of investors and / or companies interested in purchasing; 
     
    (iv)      Reduced competitiveness in the bidding procedure due to the current economic situation of the country; and

     

    Page 71


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    (v) Amendment of the minimum conditions for sale by the TCU.

    9. FINAL REMARKS

    Lastly, it should be noted that the economic and financial assessment methodology was found to be widely known and adopted by the market, which is the Free Cash Flow to Firm - FCFF, in nominal terms. In addition, the calculated discount rates were based on the CAPM - Capital Asset Pricing Model method, which estimates the return that the shareholders expect to obtain from the company, given this is a well-established method widely used either by the market at large or by Eletrobras.

    Additionally, some points need to be highlighted. Among them, there is the need for approval of the minimum amount and the adjustments necessary for achieving such amount by the Executive Board and Board of Directors of Eletrobras, in addition to the need to submit the item to the general shareholders' meeting.

    Another point to be highlighted is that the approved sales modeling of CPPI Resolution no. 20/2017 considered Scenario 1, with adjustments by Eletrobras of R$ 50 thousand, and conditioned that the bidding process of Ceal be held upon the execution and judicial homologation of an agreement concerning the payment of the salary differences arising from the Bresser Plan. Nonetheless, there was no agreement between the concerned parties. If this status continues in place, based on the provisions under CPPI Resolution no. 20/2017 and Decree 9192, of 2017, the most probable scenario is the liquidation of Ceal.

    Concerning the need for adjustments, given that in the scenario considered by CPPI, with the tariff relaxation, Ceal presented a positive equity value, such relaxation was reduced for the equity value to be zero. This way, the study provides for the need for capitalization by Eletrobras of R$ 50,000.00, in order for the company’s value to achieve the minimum value of R$ 50,000.00.

    It should be noted that the studies do not consider the fixed assets in use - AIC, amounting to R$ 177.4 million, referenced back to values of Jun./2017. If, in the modeling, AIC’s were considered unitized, even if set aside, they would be able to reduce the distribution company’s value.

    Page 72


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Furthermore, it should be noted that in the proposed sales model there is the possibility that Eletrobras to remain with one share of Ceal, with the sole purpose of increasing its shareholding in the distribution company by up to 30%, after sale of the controlling interest in the company to a third party. The increase in interest, if it is the option adopted by Eletrobras, must be carried out within six months after the date of the auction.

    It is worth noting that Service B performed a comparative analysis between the option of divesting the shareholder control of the distribution company with the granting of a new concession and the simple liquidation option of the distribution company, as deliberated by the shareholders at the 165th EGM of Eletrobras. With this analysis, one may conclude that the liquidation of the distribution company, under the scope of Eletrobras, as established under the modeling proposed by BNDES, is the most onerous option to Eletrobras, given that, for such purpose, it would have to bear an additional cost of R$ 570.0 million connected to the sale scenario.

    Lastly, it should be noted that in the case of the distribution company’s liquidation, Eletrobras would not benefit from the relaxation of regulatory parameters, which contributed to an improvement in the value of Boa Vista Energia’s equity value of the order of R$ 500 million.

    Page 73


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Appendix to the DFP/DFF/DDE Joint Technical

    Information

    Analysis and Results of CEPISA

    Page 1


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Contents     
    1.  Objective  4 
    2.  Assumptions of economic and financial modeling  4 
    2.1  Market and consumer units  4 
    2.1.1  Evolution of Captive Market  4 
    2.1.2  Market Projections prepared by Service A  5 
    2.1.3  Market Projections prepared by Service B  10 
    2.1.4  Assumptions used and points of divergence of Services A and B  14 
    2.1.5  Evolution of Consumer Units  16 
    2.1.6  Projections of Consumer Units prepared by Service A  17 
    2.1.7  Projections of Consumer Units prepared by Service B  18 
    2.1.8  Comparison of Services A and B projections  19 
    2.1.9  energy procurement  19 
    2.2  Network extension (Km)  20 
    2.3  Electrical losses  22 
    2.3.1  Projections of Electrical Losses prepared by Service A  23 
    2.3.2  Projections of Electrical Losses prepared by Service B  25 
    2.3.3  Assumptions used and points of divergence of Services A and B  28 
    2.4  Default and Unrecoverable revenues  29 
    2.5  PMSO (personnel, material, services and others)  31 
    2.5.1  PMSO incurred (2012 2016)  31 

     

    Page 2


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

      2.5.2 PMSO for Service A  33 
      2.5.3 PMSO for Service B:  39 
      2.5.4  PROJECTIONS AND COMPARISONS OF THE PMSO OF SERVICES A AND B  43 
      2.6  Investment  48 
      2.7  Regulatory remuneration base  52 
      2.8  X Factor  56 
      2.9  WACC  61 
    3.    The result of economic and financial assessments and by multiples  64 
      3.1  AFAC  66 
    4.    Contingencies  66 
      4.1  Legal Contingencies  66 
    5.    capital structure  67 
      5.1  Capitalization by Eletrobras and by the Investor  69 
      5.1.1 ELETROBRAS  69 
      5.1.2 Investor  69 
      5.2  Fixed Assets in use  69 
    6.    liquidation Costs  70 
    7.    Analysis of the benefits and risks of disposal  71 
      7.1  Benefits  71 
      7.2  Risks  72 
    8.    Final Remarks  72 

     

    Page 3


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    1. OBJECTIVE

    The purpose of this appendix is to include a detailed presentation of the results of the economic and financial assessments of Cepisa performed by Services A and B, contracted by BNDES, as well as to detail the internal adjustments that Eletrobras needs to implement in order to enable its sale at the target price determined by the model approved under CPPI Resolution 20/2017, dated November 8, 2017, amended by CPPI Resolution 28, dated November 22, 2017.

    2.      ASSUMPTIONS OF ECONOMIC AND FINANCIAL MODELING
    2.1      MARKET AND CONSUMER UNITS
    2.1.1      EVOLUTION OF CAPTIVE MARKET

    In the last ten years, from 2007 to 2016, Eletrobras Distribuição Piauí had average annual growth of 7.51%, as can be seen in the tables below. Residential, Commercial and Rural classes contributed to this evolution, with average annual growth of 9.8%, 8.4% and 8.6%, respectively.

    An atypical growth was that of Industrial class, which remained at the same level of consumption of electricity, with an average annual growth of 0.9%. This is due to the lower consumption between the years 2011, 2012 and 2013, where demand in this class had a reduction of 23.2% over this period. This low growth over the last 10 years has meant that the weight of the industrial class in total electricity consumption in the Cepisa concession fell from 12.1% in 2007 to only 6.2%. On the other hand, the weight of the residential class went from 41.3% to 48.9% in the respective years.

    Page 4


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

      Table 1.  Evolution of consumption classes in MWh     
     
            POWER BROKEN DOWN PER CLASS (MWh)       
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  710,125  759,959  807,695  989,528  1,030,587  1,194,233  1,327,936  1,431,593  1,526,439  1,629,013 
    INDUSTRIAL  207,291  235,932  231,367  251,424  245,274  227,822  193,042  215,788  218,090  207,352 
    COMMERCIAL  350,847  369,849  388,716  453,775  491,320  572,242  609,616  659,465  702,029  740,754 
    RURAL  83,277  81,719  79,862  98,277  102,054  128,646  128,739  142,042  152,350  162,769 
    GOVERNMENT  135,404  145,906  151,355  173,530  172,104  203,878  214,938  221,152  225,927  243,826 
    PUBLIC LIGHTING  115,313  118,579  121,840  124,960  127,732  144,904  174,709  188,457  183,478  183,793 
    PUBLIC SERVICE  112,308  115,410  112,680  124,122  131,156  152,184  147,838  154,196  157,309  161,250 
    OWN CONSUMPTION  3,288  3,126  3,112  3,247  3,346  3,920  4,004  3,778  3,505  3,311 
    TOTAL CONSUMPTION  1,717,853  1,830,480  1,896,627  2,218,863  2,303,573  2,627,829  2,800,823  3,016,471  3,169,127  3,332,069 

     

      Table 2.  Composition of Consumption Class     
     
              COMPOSITION OF CLASSES         
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  41.3%  41.5%  42.6%  44.6%  44.7%  45.4%  47.4%  47.5%  48.2%  48.9% 
    INDUSTRIAL  12.1%  12.9%  12.2%  11.3%  10.6%  8.7%  6.9%  7.2%  6.9%  6.2% 
    COMMERCIAL  20.4%  20.2%  20.5%  20.5%  21.3%  21.8%  21.8%  21.9%  22.2%  22.2% 
    RURAL  4.8%  4.5%  4.2%  4.4%  4.4%  4.9%  4.6%  4.7%  4.8%  4.9% 
    GOVERNMENT  7.9%  8.0%  8.0%  7.8%  7.5%  7.8%  7.7%  7.3%  7.1%  7.3% 
    PUBLIC LIGHTING  6.7%  6.5%  6.4%  5.6%  5.5%  5.5%  6.2%  6.2%  5.8%  5.5% 
    PUBLIC SERVICE  6.5%  6.3%  5.9%  5.6%  5.7%  5.8%  5.3%  5.1%  5.0%  4.8% 
    OWN CONSUMPTION  0.2%  0.2%  0.2%  0.1%  0.1%  0.1%  0.1%  0.1%  0.1%  0.1% 
    TOTAL CONSUMPTION  100%  100%  100%  100%  100%  100%  100%  100%  100%  100% 

     

      Table 3.  Annual Growth Rate in the Period       
     
            ANNUAL GROWTH RATE        AVERAGE
    OF
    ANNUAL
    GROWTH 
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  6.8%  7.0%  6.3%  22.5%  4.1%  15.9%  11.2%  7.8%  6.6%  6.7%  9.8% 
    INDUSTRIAL  5.1%  13.8%  -1.9%  8.7%  -2.4%  -7.1%  -15.3%  11.8%  1.1%  -4.9%  0.9% 
    COMMERCIAL  5.7%  5.4%  5.1%  16.7%  8.3%  16.5%  6.5%  8.2%  6.5%  5.5%  8.4% 
    RURAL  13.1%  -1.9%  -2.3%  23.1%  3.8%  26.1%  0.1%  10.3%  7.3%  6.8%  8.6% 
    GOVERNMENT  1.5%  7.8%  3.7%  14.7%  -0.8%  18.5%  5.4%  2.9%  2.2%  7.9%  6.4% 
    PUBLIC LIGHTING  2.4%  2.8%  2.8%  2.6%  2.2%  13.4%  20.6%  7.9%  -2.6%  0.2%  5.2% 
    PUBLIC SERVICE  1.4%  2.8%  -2.4%  10.2%  5.7%  16.0%  -2.9%  4.3%  2.0%  2.5%  4.0% 
    OWN CONSUMPTION  -0.2%  -4.9%  -0.4%  4.3%  3.0%  17.2%  2.2%  -5.6%  -7.2%  -5.5%  0.3% 
    TOTAL CONSUMPTION  5.5%  6.6%  3.6%  17.0%  3.8%  14.1%  6.6%  7.7%  5.1%  5.1%  7.5% 

     

    2.1.2 MARKET PROJECTIONS PREPARED BY SERVICE A

    The consumption of electrical energy in Brazil can be segregated into eight distinct classes, which are: Residential, Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service and Own Consumption. In order to capture the possible heterogeneities among the analyzed groups, Service A separately modeled the Residential, Commercial, Industrial, Rural and Other sectors (which compose the classes of Public Power, Public Lighting, Public

    Page 5


     



     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.1.3 MARKET PROJECTIONS PREPARED BY SERVICE B

    In order to know the reality experienced by the Distribution Company and to use the historical values for the necessary projections, Service B initially collected consumption information made by the utility from 2013 to 2016. The evolution of the Market by level of voltage in the period is presented in the table below:

      Table 6.  Market Evolution by Voltage Level   
      Market per Voltage Level (MWh)      Change Rate 
    Voltage/Year  2013  2014  2015  2016  Period  Geometric 
              (2013/2016)  (p.a.) 
    SUPPLY  2,728,793  2,909,591  3,113,198  3,257,275  19%  6% 
    A3  59,516  78,545  95,188  109,276  84%  22% 
    A3a  13,470  11,605  14,162  16,745  24%  8% 
    A4  597,897  606,915  649,221  652,066  9%  3% 
    LV  2,057,910  2,212,526  2,354,627  2,479,188  20%  6% 
    LOW INCOME  372,805  402,376  392,734  334,522  -10%  -4% 
    SUPPLY  -  -  -  -  -  - 
    OTHER FREE  85,756  81,367  72,733  44,491  -48%  -20% 
    DISTRIBUTION  102,994  113,294  38,637  133,148  29%  9% 
    TOTAL  2,917,543  3,104,253  3,224,568  3,434,914  18%  6% 

     

    In order to subsidize the various tariff calculations and the flow of the company, Service B elaborated market simulations by: (i) consumption class (Residential, Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service, Own Consumption); and (ii) voltage level (High Voltage - A1, A2 and A3 -, Medium Voltage - A3a and A4 - and Low Voltage - LV).

    a. Market Projection by Voltage Level

    The market projection by voltage level took into account the evolution of the share of voltage levels in the total market since 2012, according to information made available by the utilities. From the analysis of the behavior of these shares, it was possible to construct an evolution curve, based on linear extrapolation, generating the market projections for each voltage (High, Medium and Low). The evolution of the share of each voltage level is presented below:

    Page 10


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    b.1 Residential Class Result

    For the Residential Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Population as an auxiliary variable and presence of seasonality and trend, using the observed consumption series from January 2010 to December 2016. From the estimated forecasts, consumption grows on average 4.81% p.a. between 2017 and 2048.

    b.2 Industrial Class Result

    For the Industrial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust), was based on the Dynamic Model, with the auxiliary variable of state GDP, seasonality component, structural break as of August 2012 and seasonality, using the observed series of consumption from January 2012 to December 2016. From the estimated forecasts, consumption grows on average 1.73% p.a. between 2017 and 2048.

    b.3 Commercial Class Result

    For the Commercial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Commercial Consumer Units and State GDP as auxiliary variables, in addition to the seasonality, using the observed consumption series from January 2012 to December 2016. From the estimated forecasts, consumption grows on average 2.26% p.a. between 2017 and 2048.

    b.4 Rural class Result

    For the Rural Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with State GDP as auxiliary variable, in addition to the seasonality and trend, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 5.80% p.a. between 2017 and 2048.

    Page 12


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    b.5 Public Power Class Result

    For the Public Power class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the ETS methodology, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 2.79% p.a. between 2017 and 2048.

    b.6 Public Lighting Class Result

    For the Public Lighting class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the ETS methodology, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 0.27% p.a. between 2017 and 2048.

    b.7 Public Service Class Result

    For the Public Service class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Consumption Unit of Public Service as an auxiliary variable, in addition to momentary dummies and the presence of seasonality and trend, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 1.94% p.a. between 2017 and 2048.

    a.1 Own Consumption Class Result

    For the Own Consumption class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Consumption Unit of Public Service as an auxiliary variable, in addition to momentary dummies and the presence of seasonality, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 0.07% p.a. between 2017 and 2048.

    a.2 Global Projection Result

    The table 9 shows the market projected by consumption class broken down every ten years between 2016 and 2046. The consumption in MWh per class is presented through its percentage participation in relation to the total consumption and the percentage growth

    Page 13


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     

    Commercial, Rural and Others (composed of Public Power, Public Lighting, Public Service and Own Consumption). For the Projection of demand by Voltage Level, the historical average of low voltage of the last 5 years of the Company was used. Service B has elaborated two types of simulations for forecasting demand. A projection by consumption classes, distributed in eight categories: Residential, Industrial, Commercial, Rural, Public Power, Public Lighting, Public Service and Own Consumption. And another projection per voltage level, consisting of: High Voltage (A1, A2 and A3), Medium Voltage (A3a and A4) and Low Voltage (LV).

     

    3.     

    The historical consumption period of the Distribution Company, used by Service A, is the years 2007 to 2016. Service B, on the other hand, used periods beginning in 2009, 2010 or 2012, depending on the projected consumption class and always ending in 2016.

     

    4.     

    Service A obtained average annual growth projections of 3.27% for the distribution company's consumption in the period from March 2018 to February 2048. For Service B, the average annual growth of the distribution company's consumption was around 3.89% for the period from 2017 to 2048.

     

      Based on the Figure that describes the Projection of the Energy Consumption of Service A, we see that the consumption projected by that of Service A is around 9,100 GWh in 2047 (in Service A report, the absolute value of the consumption projection was not observed). Service B projects a consumption of 10,403 GWh in 2046, as shown in Table 9. In absolute values, Service A shows a difference approximately 1,300 GWh smaller than Service B, which is the result of an average annual growth 0.67% smaller for Service A compared to that admitted by Service
    B.     

    The average annual growth of the consumption classes of Service A and B are presented in the table below:

    Page 15


     


    Date: 11/30/2017

    IT DFP/DFF/DDE-
    001/2017

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    Table 9.  Geometric Average of Consumption of Services A and B for the period
    Geometric Mean of the Consumptions of Services A and B
                PUBLIC  PUBLIC     
      RESIDENTIAL  INDUSTRIAL  COMMERCIAL  RURAL  GOVERNMENT      OWN  TOTAL 
    Services  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a  LIGHTING  SERVICE  ”% p.a  ”% p.a 
                ”% p.a  ”% p.a     
    A  3.71%  2.46%  3.44%  2.01%  3.19%  3.19%  3.19%  3.19%  3.27% 
    B  4.81%  1.73%  2.26%  5.80%  2.79%  0.27%  1.94%  0.07%  3.89% 
    Differences (A-B)  -1.10%  0.73%  1.18%  -3.79%  0.40%  2.92%  1.25%  3.12%  -0.62% 

     

    It should be noted that the consumption per class between 2007 and 2016 by the Distribution Company, presented in Graph 1 and used by Service A as premises, differed from the figures which are followed by the energy history produced by the Distribution Board (Table 1). With regard to Service B, although there are some differences, the consumption per voltage level (Table 6), they are close to those followed by the energy history of the Distribution Board.

    2.1.5 EVOLUTION OF CONSUMER UNITS

    The tables below show the evolution of the Consumer Units (UC) of consumption classes in recent years, from 2007 to 2016. The distribution company had an average annual growth of 4.7% of its UC. We highlight the Residential and Rural classes, which together account for around 95% of the UC of Eletrobras Distribuição Piauí.

      Table 10.  UCs Evolution by Consumption Classes     
     
              CONSUMING UNITS PER CLASS         
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  701,693  734,576  774,486  828,736  883,714  930,429  967,471  1,006,280  1,031,712  1,079,034 
    INDUSTRIAL  4,074  4,042  3,945  3,810  3,765  3,786  3,600  3,525  3,354  3,414 
    COMMERCIAL  65,278  67,136  69,510  70,258  74,566  77,547  80,131  83,474  86,151  90,976 
    RURAL  25,658  26,590  27,379  28,900  29,377  30,073  30,302  30,268  30,385  31,587 
    GOVERNMENT  12,394  13,077  13,671  13,432  13,769  14,248  14,309  14,279  15,004  15,550 
    PUBLIC LIGHTING  797  803  800  834  857  837  695  365  393  423 
    PUBLIC SERVICE  2,240  2,388  2,463  3,322  3,847  4,949  5,348  5,976  5,847  6,203 
    OWN CONSUMPTION  132  151  137  135  171  225  173  163  151  146 
    TOTAL CONSUMPTION  812,266  848,763  892,391  949,427  1,010,066  1,062,094  1,102,029  1,144,330  1,172,997  1,227,333 

     

    Page 16


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
      Table 11.  UCs Composition by Consumption Classes     
     
              COMPOSITION OF CLASSES         
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  86.4%  86.5%  86.8%  87.3%  87.5%  87.6%  87.8%  87.9%  88.0%  87.9% 
    INDUSTRIAL  0.5%  0.5%  0.4%  0.4%  0.4%  0.4%  0.3%  0.3%  0.3%  0.3% 
    COMMERCIAL  8.0%  7.9%  7.8%  7.4%  7.4%  7.3%  7.3%  7.3%  7.3%  7.4% 
    RURAL  3.2%  3.1%  3.1%  3.0%  2.9%  2.8%  2.7%  2.6%  2.6%  2.6% 
    GOVERNMENT  1.5%  1.5%  1.5%  1.4%  1.4%  1.3%  1.3%  1.2%  1.3%  1.3% 
    PUBLIC LIGHTING  0.1%  0.1%  0.1%  0.1%  0.1%  0.1%  0.1%  0.0%  0.0%  0.0% 
    PUBLIC SERVICE  0.3%  0.3%  0.3%  0.3%  0.4%  0.5%  0.5%  0.5%  0.5%  0.5% 
    OWN CONSUMPTION  0.0%  0.0%  0.0%  0.0%  0.0%  0.0%  0.0%  0.0%  0.0%  0.0% 
    TOTAL CONSUMPTION  100%  100%  100%  100%  100%  100%  100%  100%  100%  100% 
     
     
     
     
      Table 12.  Annual Growth Rate in the Period     
     
            ANNUAL GROWTH RATE       AVERAGE
    OF
    ANNUAL
    GROWTH 
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  5.1%  4.7%  5.4%  7.0%  6.6%  5.3%  4.0%  4.0%  2.5%  4.6%  4.9% 
    INDUSTRIAL  -1.8%  -0.8%  -2.4%  -3.4%  -1.2%  0.6%  -4.9%  -2.1%  -4.9%  1.8%  -1.9% 
    COMMERCIAL  6.3%  2.8%  3.5%  1.1%  6.1%  4.0%  3.3%  4.2%  3.2%  5.6%  4.0% 
    RURAL  6.4%  3.6%  3.0%  5.6%  1.7%  2.4%  0.8%  -0.1%  0.4%  4.0%  2.8% 
    GOVERNMENT  5.9%  5.5%  4.5%  -1.7%  2.5%  3.5%  0.4%  -0.2%  5.1%  3.6%  2.9% 
    PUBLIC LIGHTING  -23.5%  0.8%  -0.4%  4.3%  2.8%  -2.3%  -17.0%  -47.5%  7.7%  7.6%  -6.8% 
    PUBLIC SERVICE  3.4%  6.6%  3.1%  34.9%  15.8%  28.6%  8.1%  11.7%  -2.2%  6.1%  11.6% 
    OWN CONSUMPTION  1.5%  14.4%  -9.3%  -1.5%  26.7%  31.6%  -23.1%  -5.8%  -7.4%  -3.3%  2.4% 
    TOTAL CONSUMPTION  5.2%  4.5%  5.1%  6.4%  6.4%  5.2%  3.8%  3.8%  2.5%  4.6%  4.7% 

     

    2.1.6      PROJECTIONS OF CONSUMER UNITS PREPARED BY SERVICE A
    A      model was used based on two parameters, the relation Population/UC Residential and

    the average growth of Residential UC for the last 6 years. The same projected residential UC growth was extrapolated to the other classes. As a result of the projections, there was a growth of UCs in the first years, with initial variation of 3.5%, with a gradual reduction of growth until reaching close to 0% growth from the 25th year of the new concession, consolidating around 1,650,000 UC. In the graph below the results of the projection follow:

    Page 17


     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 14.    Projection of Network Extension until the year 2027 Service A 
    Class    2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    LV/AV                         
        89,648  90,618  91,259  91,997  92,808  93,546  94,284  95,022  95,760  96,498  97,236 
    HV                         
        3,342  3,700  3,999  4,379  4,379  4,385  4,484  4,570  4,619  4,910  5,201 
    PLPT                         
        5,373  6,002  9,002  9,002  9,002  9,002  9,002  9,002  9,002  9,002  9,002 
    Total                         
        98,363  100,320  104,260  105,378  106,189  106,933  107,770  108,594  109,381  110,410  111,439 

     

    Which equals to one of percentage growth per year, according to the respective five-year period, according to the table below:

    Table 15.  Percentage growth of network extension, per five-year period  
    Service A
      1st Five-Year  2nd Five-Year  3rd Five-Year  4th Five-Year  5th Five-Year  6th Five-Year   
    Annual Growth (%)  Period  Period  Period  Period  Period  Period  Total Average 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
      1.7%  0.8%  0.9%  0.9%  0.8%  0.8%  1.0% 

     

    Service B did not show Network projection numbers, only presented the econometric model used to estimate the growth projection of the network extension. The total network extension of each company was projected according to a calculated elasticity for the total number of consumer units in the concession area. To determine this elasticity, panel data from the six Eletrobras distribution companies were used for the period between 2007 and 2016.

    Regarding the starting point, that is, the network extension for the year 2017, according to the information contained in the assessment reports of Services A and B, the Consulting Firms started with very different information:

    - Service A reported for the year 2017 the network extension numbers of 89,648 km in LV/AV, 5,373 km in AT and 5,373 km in PLPT.

    - Service B reported only the figure of 71,000 km of length for LV. The HV and PLPT figures were not reported.

    Page 21


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.3 ELECTRICAL LOSSES

    The losses of the electric distribution system are divided into two categories: Losses in the Basic Network and Losses in the Distribution Network. In turn, losses in the Distribution Network are divided into Technical Losses (PT) and Non-Technical Losses (NTP).

    Losses in the Basic Network are those admitted by the SIN. The real values are calculated monthly by the CCEE, and the regulatory value is given through the average of the measurements of the last 12 months. The value of the 2016 tariff adjustment was used for the assessment report.

    PTs are the loss of transport and electrical equipment. These are calculated by ANEEL in the tariff review period and remain constant for the entire tariff cycle.

    PNTs are the manageable losses where ANEELs reduction curve is estimated at each tariff review and takes into account three factors: the target, the trajectory to the target and the limits of reduction of regulatory loss. The goal is calculated based on the analysis of the comparative model of benchmarking, in which a company is established that will serve as a basis for comparison, taking into account the size of the distribution company and the socioeconomic complexity of the region in which the distribution company is located. According to the metric specified in PRORET submodule 2.6. It should be reminded that the percentage levels of PNT are related to the Low Voltage (LV) market of the utility, while the PT and Loss in RB indices are measured in relation to the Injected Energy, in accordance with the regulatory methodology.

    The percentage values of Losses are presented below, according to Cepisa's Sparta worksheet of 2016, as well as the values of 2016 of the distribution company, according to NT 149/2017-ANEEL.

    According to NT 149/2017-ANEEL, the company's new PTN regulatory level will be 25.41% of the LV market, and should remain fixed until the first common tariff review, which is defined as the midpoint between the loss in 2016 and regulatory loss. The premise used is that the new utility provider would take five years, or one rate-related cycle, to converge to the regulatory benchmarks.

    Page 22


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 16.  Calculation of losses - SPARTA Cepisa 2016 
     
    Description  Regulatory  Performed
    2016 
    Non-Technical Percentage (w/o Low Voltage)   13.93%  36.90%
    Technical Percentage (w/o input market)   12.17%  12.17%
    % Basic Grid (w/o Input Market)  2.64%  2.64% 

     

    2.3.1 PROJECTIONS OF ELECTRICAL LOSSES PREPARED BY SERVICE A

    a. Projection of Losses in the Basic Network and Technical Losses

    For Service Losses in the Basic Network and PT, Service A defined that for both indicators the regulatory and real values will remain constant and equal throughout the period, since the values of these losses are not related to the activity of the distribution company and its management, but rather physical processes and characteristics of the equipment used.

    Table 17. Level of Technical Losses and Losses in Basic Network

    % of Losses/Injected Energy
    Technical Losses  Losses in the Basic Grid 
    12.17%  2.64% 
     
     
    b. Projection of Non - Regulatory Technical Losses 

     

    Regarding the trajectory of Regulatory NTP, Service A defined in two different moments. The first period, from Sep./2017 to Sep./2023, regulatory levels were maintained at the level of regulatory losses indicated in NT 149/2017-ANEEL, equivalent to 25.41%.

    As of Sep./2023, the method of calculation of the curve of regulatory losses occurred according to the definitions described in the submodule 2.6 of PRORET, taking into account three previously mentioned factors. After the comparative analysis of the listed companies with the Cepisa it was defined that their benchmarking of the PNT goal would be COELBA.

    Page 23


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Upon definition of the target for the next year of the tariff review, the analysis of the annual reduction of regulatory losses that can have two forms of reduction for each cycle of tariff revision, is carried out. The trajectory will be the limit of reduction of regulatory losses stipulated in PRORET, in case the effort to reach the goal of benchmarking is more challenging. Otherwise, the reduction curve will have a linear trajectory until it loses benchmarking at the end of each tariff review cycle.

    The path of reduction of PNTs continues until reaching the level of 7.5% as determined by PRORET.

    c. Projection of Non -Technical Real Losses

    For the projection of real PNTs, three perspectives were observed: starting point, trajectory and target.

    As a starting point, the PNT was considered in 2016 of 37.34%, obtained from the energy balance of the distribution company. The defined target is the level of PNT practiced by the benchmark company considered in the calculation of the target of regulatory losses, in this case, the levels practiced by COELBA which are of 5.05%. The value of the loss of the benchmark company was obtained through the SAMP database.

    For the annual PNT reduction trajectory, the observed trajectory of an efficient utility was used. The benchmark company selected was CEMAR - Companhia Energética do Maranhão. The choice of the company was based on the fact that CEMAR already has an efficient level of Non-Technical losses and has recently undergone the privatization process.

    Table 18.  Starting Point and Target of Non-Technical Losses 
     
    % of Losses/BT Market
    Non-technical Losses
    in 2016
    37.34% 
    Non-technical Losses
    Benchmark in 2016
    5.05% 
    Year in which the
    regulatory level was
    achieved
    2025 

     

    Page 24


     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    2.3.3      ASSUMPTIONS USED AND POINTS OF DIVERGENCE OF SERVICES A AND B
    1.      Service A considered the projection of losses in the Basic Network to be 2.64% of the injected energy of the company for the entire period from 2017 to 2048.
      Service B did not mention these losses.
    2.      For Technical Losses (PT), Service A also projected them with the constant value of 12.17% of the energy injected during the concession. Service B used as a parameter PTs per voltage segment of compatible companies to model an efficient PT curve, taking into account the tariff cycles. The reduction trajectory starts with PT of 12.17% in 2017, arriving in 2047 at the value of 8.20%.
    3.      Modeling of Service A Regulatory Non-Technical Losses (NTP) was organized in two stages. At first, from 2017 to 2022, the regulatory levels were maintained at the level of 25.41%, based on the value defined in NT 149/2017 - ANEEL. With a sharp reduction in 2023, and in later years, there was a gradual decrease to 7.5% in 2032 and maintaining this level until the end of the concession.
      In Service B, the modeling of Regulatory PNTs was carried out based on the application of the mechanisms detailed in NT 106/2015-ANEEL. The reduction trajectory starts with 13.9% in 2017, reaching 8.0% in 2047 on the LV Market.
      Although Service B describes the application of NT 351/2017-ANEEL (which also addresses the same PNT parameters specified in NT 149/2017-ANEEL), the defined PNT levels were not observed in the graph of its report. Thus, Service was consulted and ensured that the 25.41% PNT level was used in the first five years, as determined by 149/2017-ANEEL. The trajectory of real PNTs prepared by Service A starts from the value in 2016 of 37.34%, the curve reduces until 2034, when it reaches the value realized by COELBA (benchmarking) of 5.05% of the market, remaining at this level until the end of the concession. For Service B, in the years
    2017      and 2018, the real PNT indices will remain constant at 37%, as of 2019, a
    new      phase in the efficient combat against PNT will begin. The path of reduction of
    losses      considers as reference the average of reductions of the Actual PNT’s of the

    Page 28


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

      companies CELPE, CEMAR and COELBA. Thus, at the end of 2047, the projection of the real NTP index is 6%.
    4.      Note that the PNT carried out in Cepisa in 2016 and used as assumption by Service A is 37.34% of the LV market is different from that presented in NT-149/2017- SRM/SGT/SRD/SFF/ANEEL, which was 36.9% on the LV market.

    2.4 DEFAULT AND UNRECOVERABLE REVENUES

    The Projected Unrecoverable Revenues and Default is the median of billed value that was not received until the base date of December 2016, the stabilization point of the unbilled revenue curve, from the 49th to the 60th month.

    The default assumption has great relevance in the Valuation analysis because it directly affects the collection, consequently, the company's cash flow, negatively impacting its Valuation.

    Analyzing the two Services, there are some divergences, which are listed below:

    Page 29


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    It should be noted that Service B numbers were estimated in a linear decline reaching the regulatory level in 10 years, according to text contained in the Assessment Report, without presenting the annual values used.

    Below are tables with the numbers of the default up to 2027 of the two Services:

      Table 19.  Default until 2027 - Services A and B     
     
    Service B                       
    Sunk Costs                       
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    Residential  2.95%  2.78%  2.62%  2.45%  2.29%  2.12%  1.95%  1.79%  1.62%  1.46%  1.29% 
    Commercial  2.92%  2.68%  2.45%  2.21%  1.97%  1.74%  1.50%  1.26%  1.02%  0.79%  0.55% 
    Industrial  5.88%  5.39%  4.89%  4.40%  3.90%  3.41%  2.92%  2.42%  1.93%  1.43%  0.94% 
    Rural  6.21%  5.71%  5.20%  4.70%  4.20%  3.70%  3.19%  2.69%  2.19%  1.68%  1.18% 
    Government  2.95%  2.68%  2.40%  2.13%  1.86%  1.59%  1.31%  1.04%  0.77%  0.49%  0.22% 
    Public Lighting  1.74%  1.57%  1.40%  1.23%  1.06%  0.90%  0.73%  0.56%  0.39%  0.22%  0.05% 
    Public service  0.78%  0.71%  0.63%  0.56%  0.49%  0.42%  0.34%  0.27%  0.20%  0.12%  0.05% 
    Service A*                       
    Sunk Costs                       
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    Residential  4.64%  4.44%  4.25%  4.07%  3.61%  3.21%  2.85%  2.53%  2.36%  2.36%  2.36% 
    Commercial  2.62%  2.61%  2.61%  2.60%  2.39%  2.19%  2.01%  1.85%  1.70%  1.56%  1.43% 
    Industrial  4.26%  4.18%  4.10%  4.02%  3.57%  3.17%  2.81%  2.50%  2.22%  1.97%  1.75% 
    Rural  6.08%  5.99%  5.90%  5.82%  5.60%  5.39%  5.19%  5.04%  5.04%  5.04%  5.04% 
    Government  0.16%  0.16%  0.16%  0.16%  0.16%  0.16%  0.16%  0.16%  0.16%  0.16%  0.16% 
    Public Lighting  0.16%  0.16%  0.16%  0.16%  0.16%  0.16%  0.16%  0.16%  0.16%  0.16%  0.16% 
    Public service  0.03%  0.03%  0.03%  0.03%  0.03%  0.03%  0.03%  0.03%  0.03%  0.03%  0.03% 

     

    In Service A there is not a specific year to reach the regulatory level, as performed by Service B, below. A target was established, above the regulatory level, and an annual reduction curve based on the rate of change proposed by ANEEL in 4CRTP, 3CRTP and 2CRTP. With this assumption, the regulatory level is not reached during the thirty years of concession.

    Service A raised R$ 54.725 million in Unrecoverable Revenues for the year 2016. Service B did not show any value.

    The year-on-year comparison of the projection of default rates by each consulting was hampered because Service B did not disclose the annual amounts in a table, showing only the starting value and the year that the regulatory level is reached.

    Page 30


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

    2.5      PMSO (PERSONNEL, MATERIAL, SERVICES AND OTHERS)
    2.5.1      PMSO INCURRED (2012 2016)

    For the analysis of the PMSO item contained in the Services A and B reports, it is necessary to look, in advance, at the history of this item over the last few years. For the case, the PMSO was analyzed between the period from 2012 to 2016 (same period used by Service A). Below are the values in the aforementioned period for Cepisa:

    Table 20.  PMSO incurred 2012 2016 (Cepisa) R$ million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  12.6  11.2  12.5  14.7  20.7 
    PID  0.0  0.0  0.0  0.0  0.0 
    Material  0.4  0.3  0.3  0.1  0.1 
    Third-party Service  2.1  3.0  4.3  3.9  2.9 
    Others  57.6  22.7  26.4  96.7  39.5 
    Total  72.6  37.1  43.5  115.5  63.2 

     

    For the composition of each item above, the Financial Statements of the respective years were used as support material. In item "O" of the PMSO, the following items were considered: Leases and Rentals; Taxes; Miscellaneous indemnities; Judicial costs and fees; Own energy consumption; Inspection Fee and Others. It is also worth noting that the PID (Resignation Incentive Plan) was considered for the years 2013 and 2014.

    In order to show consistency in the comparison with the Assessment Reports presented by the Consultancies, the PMSO values were updated by the IPCA for December 2016.

    Page 31


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 21.  PMSO incurred from 2012 - 2016 updated to Dec./16 - R$ million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  219.8  225.1  196.0  195.8  223.2 
    PID  0.0  93.3  2.2  0.0  0.0 
    Material  9.0  6.4  5.2  6.4  7.8 
    Third-party Services  138.2  130.3  121.3  114.3  114.1 
    Others  77.0  28.7  31.4  104.3  39.5 
    Total  444.0  483.8  356.2  420.9  384.5 

     

    In relation to regulatory operating costs, it is possible to observe the estimated values for Cepisa in the years 2013 (last tariff review) until 2016, brought to the base date of December 2016.

    In the case of Cepisa, the largest difference between the realized and the regulatory values takes place in year 2013 (This happens because of the GDP value, but if this value were disregarded, this difference would fall to 9%). For the other years, the effective PMSO follows the same line of regulation, being, in the year 2014, below the regulatory value. This fact can be seen in the graph below:

    Page 32


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

     

    2.5.2      PMSO FOR SERVICE A
    -      Historical part:

    For the composition of the historical part, the consultancy, based on the accounting statements provided by the distribution company, made a grouping of similar accounts in terms of type of expenses and also made an adjustment of the values for the base date of December 2016. Below is the table containing the opening of the PMSO for Cepisa:

    Table 22.  PMSO values calculated by the consultancy (base date - Dec/16) 
                R$ Million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  225.2  321.2  201.0  200.9  227.2 
    PID  0.0  0.0  0.0  0.0  0.0 
    Material  12.9  6.6  5.5  6.7  7.9 
    Third-party Service  151.9  123.7  114.6  117.7  116.2 
    Others  47.7  13.1  13.8  76.0  26.9 
    Total  437.7  464.5  334.9  401.3  378.2 

     

    - Methodology used in Service A: (part common to all Distribution Companies)

    The analysis of the PMSO as well as the projection of the values was divided into two parts, namely:

    Part 1 - Characterization of Analytical Accounts (Divided in 5 phases):

    Phase 1 - Grouping of analytical accounts:

    o      Adjustment for inflation of the accounts for the base date of December 2016;
    o      Grouping of accounts according to similarity of expense type;

    Phase 2 - Analysis of the nature of the account in fixed and variable:

    o      Analysis of Fixed and Variable Accounts;

    Page 33


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    o      The Accounts identified as fixed, when projected, were calculated as the average of the expenses of the sample period, excluding the outliers;
    o      For the projection of the variable accounts, an indicator of this account was calculated first as a function of Consumer Units, Network Expansion (km) and Consumption, in order to later carry out the projection. The outliers were also excluded.

    Phase 3 - Identification of outliers:

    o      The values are considered outliers when:
      §      They are outside the established limits of normality of the sample, i.e.:
       
  • Out Upper Limit => (sample mean) + (standard deviation of sample); Out Lower Limit => (sample mean) - (standard deviation of sample).
    o      After this procedure, the nature of the expense and its periodicity in the last 3
      years      were assessed in order to evaluate whether or not it would be excluded.

    Phase 4 - Assessment of the best correlations for the variable accounts:

    o      At this phase, a study was made to verify which variables were more related like the expenses of PMSO. It was concluded that Network Extension (km) and Consumption in MWh obtained higher ratio.

    Phase 5 - Identification of regulatory accounts (considered by and paid by ANEEL):

    o      Consists of the classification of the accounting accounts forming the PMSO according to the characterization defined by ANEEL for regulatory PMSO. Thus, for each component account of the analytical PMSO, a classification between Effective PMSO (total), Regulatory PMSO and other PMSO (Effective PMSO - Regulatory PMSO) is provided).

    Page 34


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Part 2 - Projection of Effective and Regulatory PMSO Grouping (Divided into 4 phases):

    Phase 1 - Projection of Expense for 2017:

    o      It considers a variation of expense in line with the company's history, taking into account the fixed or variable nature of the expenses that make up the PMSO;
    o      Variable Expenses: are transformed into indicators in which the denominator is the parameter with which the variable had the best correlation (it can be network km or consumption in MWh).
      Example: Personnel = R$/km of network => item P will increase as the network km increases;
    o      Fixed Expenses: fixed expenses were calculated as the average expenses for the sample period, excluding outliers.
  • Phase 2 - Operating Costs variation rate: This phase is divided into three
      steps:     
      o      Step 1 - Definition of Operational Cost Benchmarks;
        §      For Northern Companies (CERON, Boa Vista Energia, Eletroacre and Amazonas Energia) => Celpa and Celtins;
        §      For the Northeast Companies (Cepisa and Ceal) => Celpe, Cemar, Coelce, Energisa Paraíba, Energisa Sergipe, Coelba and Cosern.
      o      Step 2 - Rate Variation Definition and Target between 2018 and 2022:
        §      It is considered the DEA (Data Envelopment Analysis in Portuguese) database of efficient costing from Aneel for 4CRTP;
        §      The PMSO expense of each company is parameterized by the km of Network Extension;

    Page 35


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

      §      Based on the indicators, the average of the group of each distribution company according to the Region is calculated;
      §      After this step, proceed as follows:
       
  • Distribution companies with the indicator greater than the group mean adjust their cost in terms of indicator until the next Tariff Review in 2023.
       
  • On the other hand, if the indicator is less than the mean of the group, the Distribution companies plot the indicator level between 2023 and 2027 as target, which is lower than the group mean.
    o      Step 3 - Rate Variation Definition and Target between 2023 and 2027:
      §      It consists of calculating the mean indicator of the privatized Benchmarks of the groups;
      §      Upon reaching the indicator of R$/km of network of the cycles target,
       
  • distribution companies stabilize operating cost indicators parameterized
       
  • consumer unit.

    Phase 3 - Identification and Treatment of Accounts with Behavior Linked to the Distribution Companies Improvements:

    o      It consists of identifying accounts that influence the actual cost of PMSO, but which vary according to the distribution company’s operational improvements.

    Phase 4 - Projection of the Concession Period (2018 to 2047): It consists of estimating the levels of PMSO expenses of Distribution Companies over the concession period.

    o      Based on the targets set previously, the Distribution Companies reach the operational cost benchmarks until the end of 2027;
    o      Thereafter, the network PMSO/km indicator stabilizes itself, as a consequence, the absolute value of the PMSO varies in real terms through the variation of the Network km Extension.

    Page 36


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Efficient Costs:

    The elaboration of the estimate of the efficient operational costs was based on Annex I -Efficiency Intervals of the PRORET sub-module 2.2 V2. Operating costs tend to align with their respective benchmarks in two moments, the first ending 2022 and the second in 2027.

    Calculation Method:

    - First moment => simple mean of each of these limits for distribution companies according to their regional groups, which are:

    - Second moment => mean of the privatized of the regional groups mentioned above.

    It is important to note that, as Technical Note 149/2017 - ANEEL defines the operational costs in force from 2017 until 2022, the efficient cost intervals defined in this section will only impact the tariff from the Tariff Review of 2023.

    - Effective PMSO vs. Regulatory PMSO (Service A):

    Based on the values of the effective PMSO projection allocated above and with the values calculated for the regulatory PMSO, the following graph was obtained:

    Page 37


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    - The difference between the VPL of the two items, in real terms, discounted by the WACC rate of the model, considering in simplified form, a tax shield of 34%, results in a value of R$ 288.7 million, that is, for Cepisa, PMSO does not add value for the economic and financial assessment;

    - Another important point is that the projection of the regulatory PMSO values took into account the technical note 149/2017 - ANEEL.

    2.5.3      PMSO FOR SERVICE B:
    -      Historical part:

    Service B did not present the rational to compose its database from 2012 to 2016 as did Service A. The report shows the projections of P, M, S and O values directly for the period from 2017 to 2047.

  • Methodology used in Service B (common to all Distribution Companies):
     
  • Database: Calculation of Efficiency for the 4CRTP - 4th Cycle of Periodic Tariff Reviews (PMSO, Market, Consumers, network, losses, etc. data) => 2001-2013;
     
  • Clustering of Companies: Automatic grouping of similarity second degree data. The result of this process can be seen in the table below:

    Page 39


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    Table 23.  Cluster composition and efficiency - Cepisa 
     
    Cepisa Cluster
    Companies
    Efficiency
    Indicator of
    4CRTP
    efficiency 
    Efficiency
    Indicator
    2011
    Efficiency
    Indicator
    2013
    % (p.a.)
    CSPE  100%  0.49  0.42  6.6% 
    COSERN  92%  0.59  0.67  -6.7% 
    Escelsa  72%  0.89  0.98  -5.1% 
    Bragantina  69%  0.92  0.80  6.8% 
    Sulgipe  66%  0.75  0.67  5.4% 
    COPEL  64%  0.73  0.67  3.7% 
    DME  42%  1.43  1.26  6.2% 
    CEPISA  59%  0.89  0.89  0.0% 

     

    Understanding the Indicator: For instance, Cepisa in 2011 obtained an indicator of 0.89 and. in 2013, this value continued the same. This means that the company needs 0.89 (2011 and 2013) PMSO units to offer a set of services (weighted combination of Market, Network, Consumer, Non-technical Losses).

    Projection of the trajectory of the indicator:

    1.      Phase 1: 2017-2018 => No Efficiency gain
    2.      Phase 2: 2019-2020 => It follows the rhythm of the utility with the greatest gain of efficiency of the Cluster until reaching the median of the indicators of the group
    3.      Phase 3: 2021-2048 => Convergence to the mean of the Companies of the 1st Portion of the Group

    Thus, the Cepisa efficiency curve as well as the projection for the regulatory PMSO are presented in the graph and table below:

    Page 40


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Although Service B shows an Efficiency curve that should be followed in the document, it is not evident the comparison with the effective PMSO values projected by the consultancy. Therefore, it is not possible to state when the company is able to reach the level of regulatory PMSO, as well as to estimate if this trajectory will generate or destroy value for the Company.

    In addition, it is important to take into account the issue of relaxation, since according to Technical Note No. 149/2017 - ANEEL, which deals with the flexibility of some regulatory parameters, the new values for the Regulatory PMSO that go until the first ordinary tariff review of the new controller of the designated companies, are:

    Page 42


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 25.  Flexibility of the PMSO (Aneel Technical Note No. 149/2017) 
            R$ Million 
    Distribution
    Company
    Tariff Process 2016 Flexible
    Regulatory
    Level 
    Difference Difference
    (%)
     AME  348.8  627.2  278.4  80% 
    Boa Vista  37.5  69.1  31.5  84% 
     CEAL  271.7  336.7  65.0  24% 
     Cepisa  360.7  420.1  59.4  16% 
     CERON  234.0  279.9  45.8  20% 
    ELETROACRE  93.5  128.5  35.1  38% 
     CEA  67.8  124.8  56.9  84% 

     

    2.5.4      PROJECTIONS AND COMPARISONS OF THE PMSO OF SERVICES A AND B
    Table 26.  CEPISA Projections until 2047: PMSO Projection Data (Service A)   
                      R$ Million 
    Account  2017  2018  2019  2020  2021  2022  2023  2024 
    Personnel    226.5  225.9  232.4  238.7  245.1  251.4  263.7  276.5 
    Material    7.3  7.4  7.7  7.9  8.1  8.4  8.8  9.2 
    Third-party Service  119.8  133.3  139.4  145.7  152.1  158.6  166.2  174.1 
    Others    30.2  31.8  33.9  36.0  38.3  40.7  43.5  46.4 
    Total PMSO    383.7  398.4  413.2  428.3  443.6  459.1  482.1  506.1 
    Account  2025  2026  2027  2028  2033  2038  2043  2047 
    Personnel    289.9  303.9  318.6  338.5  457.0  610.0  804.4  995.0 
    Material    9.7  10.2  10.7  11.4  15.4  20.7  27.3  33.9 
    Third-party Service  182.3  191.1  200.4  210.6  270.2  345.7  441.0  534.8 
    Others    49.5  52.8  56.4  60.4  84.2  115.2  154.9  194.0 
    Total PMSO    531.4  558.1  586.1  620.9  826.8  1,091.5  1,427.6  1,757.7 

     

    In addition, Service A estimated a value for the Voluntary Resignation Plan (PDV), defining the target public (employees with at least 20 years of company), the success rate, its cost (R$ 30.0 million) and the date of resignation of employees, June/18 and June/19.

    Page 43


     



     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 28.  Comparison between Services - R$ Million 
      Service A  % Service B  % 
    Company           
      2017  2047  2017  2047  2017 
    Eletroacre  112.5  684.9  509%  133.5  Eletroacre  112.5 
    CEAL  288.8  1,126.5  290%  296.7  CEAL  288.8 
    Amazonas  550.1  1,896.4  245%  600.2  Amazonas  550.1 
    CEPISA  383.7  1,757.7  358%  452.6  CEPISA  383.7 
    CERON  340.1  1,534.8  351%  373.3  CERON  340.1 
    Boa Vista  105.6  334.1  216%  216.3  371.7  72% 
    Total  1,780.8  7,334.4  312%  2,072.5  4,642.1  124% 

     

    - Finally, regarding the documents presented by the consultancies, Service A detailed the methodology much more than Service B, presenting in its text more coherence in the presentation of the information.

    2.6 INVESTMENT

    Both services contracted for economic and financial assessment adopted different assumptions for the projection of the investment notably from the period 2023 to 2047, since both adopted the Five-Year Plan for the period from 2018 to 2022 of the company.

    This plan was presented in Product 7 “Cepisa Operational Technical Assessment Report" and was prepared based on the "Expansion Plan for the Electrical System (2017-2026 period)", the "Development Plan for Distribution - PDD for the year 2017", the "Results Plan for the Improvement of Distribution Services, 2015" and the "Temporary Provision Plan for the Electricity Distribution Service, 2017". In addition, information obtained through the "Field Visit Report" (Cepisa Annex I of the Technical and Operational Assessment Report) and the incremental investment needs projected by the company's planning area, obtained during face-to-face meetings and conference call carried out.

    In relation to the other five-year period, the main differences are due to the fact that Service A included, in addition to the projections made for implementation and renovation projected from the Five Year Plan, other analyzes to define the investment in maintenance

    Page 48


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    and replacement of partially and totally depreciated assets . Whilst Service B was restricted only to the first analysis.

    For the projection from 2023 of the investments in implantation and renovation, Service A adopted as basic premise for the expansion of AT, AV and LV, as well as for the improvements some metrics1 depending of the type of investment depending on the last years contemplated in the plans of the companies and that value was kept constant until the end of the concession period. Service B, for electric investments (expansion and improvement), estimated the second five-year period of the concession through the mean of the period from 2018 to 2022 and expurgated values outside the range between the average and +/- a standard deviation. Thus, in order to obtain the amount of investment required for the five-year period, the investments of the years considered typical were added up and multiplied by the factor 5/No. typical years. In the case of Cepisa, since there was information up to 2026 for the expansion of HV and AV, these values were confirmed and the methodology above adopted for the estimate of the last year of 2027. For the other five-year period, they adopted as base the previous five-year period with the caveat that for the expansion in AT, it is considered that every 10 years the need for investments doubles in relation to the previous five-year period. As for the non-electric investments (infrastructure), it was assumed that every 5 years it invests 50% of what was forecast in the first five years. Service A, however, did not project them.

    For renovation investments (maintenance), since 2023, service A considered that investments of this type are equivalent to the value of the assets depreciation. To define this value, the depreciation reference of ANEEL's Normative Resolution 674/20152, was used for the current asset base plus the depreciation of the new investments of the company, calculated from the average depreciation of the asset base of the Assets Control

    1 The methodology for projecting the investment from 2023 varies between companies for Service A. The

    expansion in AT had the information until 2026 through the Development Plan for Distribution 2017 (PDD 2017) and, then, the average value between 2022 and 2026 was adopted for the other years. As for the investment in AV and LV expansion, the smallest of the expenses from 2018 to 2022 was used. In turn, the improvements were defined as the average of the amounts spent between 2018 and 2022.

    2 Resolution that approved the review of the Assets control manual of the Electric Sector - MCPSE.

    Page 49


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Report. These investments were accumulated and concentrated in the two years before each tariff review according to the proportion of the reference companies3. Service B, however, first calculated the share of the amount destined to the investment in renovation within the period of the Five-Year Plan 2018-2022. Subsequently, it applied this percentage for all other years.

    For investments in replacement, only service A estimated them. Based on the assets base of 2016, the estimated investments for implementation and renovation from 2017 to 2022 were added, as well as the inflation adjustments and application of the depreciation of each item, based on Aneel Normative Resolution 674/2015, both for the assets base of 2016 and for the new investments projected up to 2022. In this way, the company's asset base of 2022 was formed. Therefore, the indicator Market Use Value (VMU)/New Value of Replacement (VNR) was calculated for the company and for the benchmark 4companies so that the distribution company under review reaches the same percentage of the private reference distributed between 2023 and 2027 by the average proportion of investment of the reference companies5.

    Given the assumptions and their differences, we find the following values for the two consultancies for Cepisa:

    3 Cemar, Coelce, Coelba and Celpe.

    4 Cemar, Energisa Sergipe and Energisa Paraíba

    5 Cemar, Coelce, Coelba and Celpe.

    Page 50


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    Table 29. Comparison of Services A and B
     
    Service A              R$ Million 
      1st Five-Year  2nd Five-Year  3rd Five-Year  4th Five-Year  5th Five-Year  6th Five-Year   
                  Sum for the 
    Investment  Period  Period  Period  Period  Period  Period   
                  Period 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
    AT Extension  458  315  323  323  323  323  2,064 
    AV/LV Extension  519  145  145  145  145  145  1,244 
    Improvement  234  234  234  234  234  234  1,404 
    Renewal (maintenance)  160  717  848  982  1,134  500  4,341 
    Luz para Todos (Light for Everyone)  -  -  -  -  -  -  - 
    Infrastructure and Support  3  -  -  -  -  -  3 
    Replacement  -  159  -  -  -  -  159 
    Total  1,374  1,570  1,550  1,684  1,836  1,202  9,214 
    Service B              R$ Million 
      1st Five-Year  2nd Five-Year  3rd Five-Year  4th Five-Year  5th Five-Year  6th Five-Year   
                  Sum for the 
    Investment  Period  Period  Period  Period  Period  Period   
                  Period 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
    AT Extension  464  231  324  97  291  29  1,437 
    AV/LV Extension  524  256  256  256  256  256  1,802 
    Improvement  237  129  129  129  129  129  880 
    Renewal (maintenance)  160  80  92  63  88  54  537 
    Luz para Todos (Light for Everyone)  -  -  -  -  -  -  - 
    Infrastructure and Support  3  2  2  2  2  2  11 
    Replacement  -  -  -  -  -  -  - 
    Total  1,388  697  802  545  765  469  4,666 

     

    The difference between the consortiums is R$ 4.5 billion, notably in the item Renovation (maintenance) in which Service A estimates that from the 2nd five-year period the investments are equivalent to the value of the depreciation of the assets and Service B calculates them from their participation in the Five-Year Plan from 2018 to 2022. In addition to this, service A, in contrast to B, estimated the replacement investment amounting to R$ 158.6 million in order for Cepisa to reach the same ratio as the VMU/VNR indicator of the companies used as private reference. It should be noted that this difference in the volume of projected investments does not have a significant impact on valuation, since the investment increases the Company's Remuneration Base, and consequently, it is reflected in a higher tariff.

    By way of comparison, the value of the investment made from 2012 to 2016 to the values of December 2016 by Cepisa is as follows:

    Table 30.  Comparison of Investment carried out from 2012 to 2016, values of Dec./2016
                  R$ Million 
    Investment  2012  2013  2014  2015  2016  Five-Year Period 
                  2012-2016 
    Updated Values (Dec./16)  421  262  158  171  152  1,164 

     

    Page 51


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.7 REGULATORY REMUNERATION BASE

    Both consortiums used the data contained in the valuation reports of the assets as reference to the date of February 28, 2017 to prepare the projection of the Regulatory Remuneration Base (BRR). In these reports, both closed BRR and the incremental BRR from 3CRTP to 4CRTP were assessed. In the case of Cepisa, the following reports were used:

    Both service A and B adopted different projection criteria for the asset base unitized through Feb/17 (closed BRR - starting point for 4CRTP) and for unitizations occurring after Feb/17 and new investments (incremental BRR).

    Regarding the starting point for 4CRTP, both consortia use the full assessment report (delivered on 06/30/2017), however, they present different criteria for the definition of Fixed Assets in Use (AIS).

    For service A, the disallowance is treated in two ways: i) Disallowance over the reviewed closed base (Closed Base of Full Report - Closed Base of the Incremental Report) in the value defined in 5,73% for the companies of the North and 5.01% for the Northeast, based on the average adjustment (disallowance) of the distribution group assessed in the 3CRTP by region, i.e., this disallowance is made only in the AIS of the difference found in the opening of the closed base compared to that approved by Aneel in 3CRTP monetarily corrected; and ii) The incremental basis between 3CRTP and 4CRTP - the lowest value was applied between the average adjustment of the distribution companies group assessed in

    Page 52


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    3CRTP by region and that verified by the company, without showing the amount applied in the Assessment Report.

    Unlike Service A, Service B used only a disallowance methodology to reach the Remuneration Base of Feb/17 of the Distribution Company. The disallowances were performed between the difference in AIS of the two assessment reports of the Closed Base and the Incremental Base. This difference would be the opening of the Closed Base. That is, the disallowance for Service B was performed only at the opening of the Closed Base and occurred in the AIS. The percentage of disallowance was calculated from the relation between the final AIS (with disallowances) and the initial AIS (without disallowances) of the incremental BRR of the distribution companies 3RTP of Eletrobras group, obtaining a mean of 9.2%. If the company has a disallowance lower than this mean, it is considered the mean of disallowance percentage itself (9.2%). Otherwise, the mean between the percentage of the company and the maximum percentage of disallowances (20.1%) is used. In the case of Cepisa, it falls within the first criterion, since it presented a reversal of 6.9% of the AIS in the 3CRTP, adopting, then, the reversal of 9.2% for the closed base. This disallowance percentage is applied only in the difference between the AIS of the two reports and added to the value of the base of the incremental report. That is, the disallowance of 9.2% is not applied in the total AIS of the Complete Base, it is calculated only for the increment of the closed base.

    Therefore, the following information was obtained for the Closed base, in the position of February 2017, for both services:

    Page 53


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    Table 31.  Closed base, in the position of February 2017 - Services A and B 
     
          R$ Million 
        Armored Base   
      Description     
        Service A  Service B 
    (1) Property, Plant And Equipment in Service (New Replacement Value)  3,290  3,249 
    (2) Index of Full Utilization  11  11 
    (3) Gross Special Obligations  941  941 
    (4) Fully Depreciated Assets  957  957 
    (5) Gross Compensation Reference = (1) - (2) -(3) - (4)  1,381  1,339 
    (6) Accumulated Depreciation  1,599  1,599 
    (7) Net AIS (Market Value In Use)  1,691  1,650 
    (8) Index of Depreciated Utilization  7  7 
    (9) Compensation Reference Value (VBR)  1,684  1,643 
    (10) Warehouse in Operation  9  9 
    (11) Deferred Assets  -  - 
    (12) Special Net Obligations  739  739 
    (13) Lands and Easements  10  10 
    (14) Base of Total Net Compensation = (1) - (6) - (8) + (10) + (11) - (12) +     
    (13)    965  923 
    (15) RGR PLPT Balance  34  28 
    (16) RGR Balance Other Investments  384  320 
    (17) Depreciation Rate  4%  4% 
    (18) Share of Regulatory Reintegration = (5) * (17)  55  50 
    (19) Actual WACC before Taxes  11%  ND 
    (20) RGR PLPT Rate  1%  ND 
    (21) RGR Rate of Other Investments  4%  ND 
    (22) Capital Compensation (RC) = (15) * (20) + (16) * (21) + [(14) - (15) -  76  ND 
    (16)] * (19)     

     

    For the years subsequent to the 4CRTP Starting Base, both Services apply adjustment for inflations for the IPCA, additions and write-offs of the assets that were considered in the valuation until the end of the concession period.

    However, it should be noted that both Services did not incorporate the positions of Fixed

    Assets in use into the Remuneration Base, since, according to the services, the assessors’ report of the Remuneration Base does not take into account the specific assessment of the condition and characteristic of these assets concerning (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the reassessment process (such as the calculation of VNR, depreciation, utilization rate and valuation of losses). In regards to the case of Cepisa, it should be noted that the Full Assessment Report “Cepisa

    Eletrobras Distribuição Piauí Project no. 3082-18367,” performed by Levin, includes as the position of February 2017 the balance of R$ 168.1 million for the Fixed Assets in Use.

    Page 54


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    In addition, Service B segments the assumptions for the projection of BRR transactions into two groups. The first group operated the closed BRR by the IPCA adjustment for inflation and the discount of the depreciation and write-off of the assets that went into operation until February 17 and in a second group operates the incremental BRR (new investments) through the additions of assets projected after Feb/2017, less the losses related to these assets.

    Thus, from 2018, service A monthly corrects, through the IPCA, the items Fixed Assets in Use, Warehousing in operation and land and easements and increases them according to their projection of the investment. The gross special obligations item is also updated by the IPCA and by the increase in projected special obligations. On this incremental Base, the average adjustment of privatized benchmark distribution companies was applied as disallowance by region obtained through the 3CRTP Incremental Base. For the Northeast companies, Celpe, Energisa Paraíba and Energisa Sergipe were considered, giving a value disallowance of 2.09%. For the North companies, the disallowance value was 2.83%, using Celpa and Cemar as benchmark. In relation to the depreciation rate, the same verified in 3CRTP was maintained for each distribution company assessed. In relation to asset write-offs, the ratio of fully depreciated assets to fixed assets in use observed in the 4RTP BRR resulted in an equilibrium rate calculated as a simple mean between Eletrobras' companies. Thus, starting from the company's ratio, it varies at a constant monthly rate until reaching that equilibrium rate of 12.54% at the end of the concession.

    Service B, as mentioned earlier, operated BRR with different premises for the base that served as a starting point and the one that incorporates the new investments. The transaction of the closed BRR was made by monetary restatement by the IPCA and by the discount of depreciation and write-offs. The depreciation rate, unlike service A, is not constant, the rate falls from the average level calculated in Feb/17, as the assets are being written-off. With respect to the asset write-offs, the average write-off/AIS ratio recorded in the Assessment Report prepared by Delloite in the period between 3CRTP and Feb./2017. The incremental BRR transaction (new investments) was made based on the projection of the investments rejected in 5%. The depreciation rate used was obtained by the weighted mean of the investments of the Five-Year Plan 2018 to 2022 per group of assets and their respective depreciation rates, obtaining 3.13% of the AIS. This rate was

    Page 55


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    considered constant throughout the concession. Regarding the write-offs, until Feb/27, they were considered null in view of the reduced probability of assets being written off in the first 10 years. After this period, 1/3 of the BRR Closed/AIS write-off ratio was applied in a staggered manner every 5 year until reaching the value of the integral ratio as of Mar/42.

    Accordingly, given the differences in the investment projections, the criteria for disallowance, depreciation and write-offs, we obtained at the end of the concession (Feb/48) the following values of Net Remuneration Basis for Cepisa according to the projections of the two services:

      Table 32.  Net Remuneration Base - Services A and B 
            R$ Million 
      Description    Service A  Service B 
      Net Remuneration Base (Feb./48)  23,807  7,099 
     
     
    2.8  X FACTOR       

     

    The primary objective of the X-Factor is to ensure that the balance established in the tariff review between efficient revenues and expenses is maintained in subsequent tariff repositioning. This occurs through the transfer to the consumer of potential productivity gains from the distribution segment.

    The approach adopted by Aneel adds to the X Factor a mechanism to encourage the improvement of the technical and commercial quality of services provided to the consumer. It also introduces a component that seeks to establish the gradual transition of the operating costs of the utility towards its efficient costs.

    The X Factor adjusts the value of Installment B, thus affecting the tariff and, consequently, the revenue and the collection of the Distribution Company. Therefore, the X factor value inversely impacts the value of the VPL (the higher the X factor, the lower the revenue and the VPL).

    Page 56


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 33.  X factor and its components in RTP year - Services A and B 
      Service A         
        X Factor  Pd  T  Q 
      2019  -0.13%  0.00%  0.00%  -0.13% 
      2023  0.50%  0.98%  -0.70%  0.22% 
      2028  1.25%  0.98%  0.38%  -0.11% 
      2033  1.35%  0.98%  0.37%  0.00% 
      2038  1.06%  0.98%  0.08%  0.00% 
      2043  0.98%  0.98%  0.00%  0.00% 
      2048*  0.98%  0.98%  0.00%  0.00% 
     
     
     
      Service B         
        X Factor  Pd  T  Q 
      2019  0.34%  0.00%  0.00%  0.34% 
      2023  4.88%  1.55%  4.19%  -0.86% 
      2028  4.17%  1.44%  3.04%  -0.31% 
      2033  2.53%  1.50%  1.35%  -0.32% 
      2038  1.85%  1.52%  0.59%  -0.26% 
      2043  1.92%  1.54%  0.66%  -0.28% 
      2047*  1.92%  1.57%  0.66%  -0.31% 

     

    As can be seen in the table above, the estimated values for the X Factor are quite divergent, especially for the Q and T components, affecting, as a whole, the X-Factor as a whole.

    Table 34.  X Factor differences betweenServices A and B 
    Service A - Service B       
      X Factor  Pd  T  Q 
    2019  -0.47%  0.00%  -0.13%  -0.34% 
    2023  -4.38%  -0.57%  -3.97%  0.16% 
    2028  -2.92%  -0.46%  -3.15%  0.69% 
    2033  -1.18%  -0.52%  -1.35%  0.69% 
    2038  -0.79%  -0.54%  -0.59%  0.34% 
    2043  -0.94%  -0.56%  -0.66%  0.28% 

     

    Page 58


     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    3 - Component T - Trajectory of operating costs

    Component T of the X-Factor is applied to Installment B so that the tariff increases or decreases so that the recognized regulatory operating costs reach the efficiency threshold defined by Aneel.

    In defining efficient operating costs, the distribution company’s costs, the efficient level of costs and the characteristics of the concession area are observed. Aneel uses a model whose objective is to define a feasible production/cost frontier as a way of identifying the level of efficiency of companies.

    In a simplified way, the identification of the efficient level of costs is obtained by comparing the distribution companies by means of a benchmarking method that takes into account the attributes of each utility. From these conditions, a target of efficient operational costs is established to be reached throughout the tariff cycle. At the time of the tariff review, the regulatory goal is compared to the operational costs coverage present in the utility´s tariff, called operating costs revenue. From the difference between the regulatory goal and the operating costs revenue, a regulatory trajectory will be calculated. Part of the difference will be incorporated at the time of the tariff review and the remaining installment will be considered for the purpose of calculating component T of the X Factor.

    Services A and B, following the provisions of Aneel's Technical Note No. 149/2017, used a zero value for component T in the period between the date of signature of the agreement and the first ordinary tariff review, that is, from 2018 to 2022. From 2023, Services A and B found very divergent values for Component T of X Factor.

    2.9 WACC

    Services A and B adopted the method of Capital Asset Pricing Model - CAPM to estimate the return that the shareholders expect to arrange with the capital employed in the business, for discounting the current value of the distribution company’s cash flow. It

    Page 61


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    should be noted that this method is well-established and widely used, either by the market at large, and by Eletrobras.

    In the calculation of the Weighted Average Cost of Capital WACC, Services A and B applied different approaches of the CAPM method, thus allowing quite different results.

    Below are the main differences found in the calculation of WACC by both Services.

    Capital Structure

    Beta

    ·Service A: used comparable domestic companies. Unlevered Beta: 0.68;

    ·Service B: considered comparable domestic companies, with a larger sample. Unlevered Beta: 0.43

     

    Ke

    ·Service A: applied the CAPM method and a “foreign exchange risk” of 4.17% at nominal Ke in Dollars to achieve a nominal Ke in Brazilian Reais. Therefore, this

    “foreign exchange risk,” which derives from the “Future DI” rate and “Fraction of exchange rate coupon,” works as a differential of inflation. To achieve such rate in real terms, it was deflated by IPCA (4.23%);

    ·Service B: applied the CAPM method.

     

     

    The table below shows the result for Ke under real terms:

    Page 62


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 36.  Profit or Loss of Real Ke Services A and B     
    Actual Ke  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  7.95%  7.81%  7.53%  7.21%  7.02%  6.85%  6.70%  6.55%  6.43%  6.31%  6.31% 
    Service B  11.73%  10.16%  10.18%  10.42%  10.42%  10.42%  10.42%  10.42%  10.84%  10.84%  10.79% 

     

    Kd

    The table below shows the Kd result in real terms after taxes.         
     
      Table 37.  Kd in real terms after taxes Services A and B     
    Actual Kd  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  4.25%  4.09%  4.27%  4.19%  4.19%  4.19%  4.19%  4.19%  3.25%  3.25%  3.25% 
    Service B  4.44%  3.34%  3.36%  3.29%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25% 

     

    WACC

    Based on the values of Beta, Ke and Kd above, Services A and B calculated WACC in real terms after taxes, according to the results included in the table below.

          Table 38.  WACC Services A and B       
    WACC  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  8.09%  7.20%  7.30%  7.39%  7.39%  7.39%  7.39%  7.39%  7.60%  7.60%  7.58% 
    Service B  5.29%  4.80%  4.89%  4.87%  4.93%  5.00%  5.06%  5.12%  5.19%  5.26%  5.26% 

     

    Out of the results stated above, it should be noted that the factor with a higher impact on the rate presented by Service A was the use of the foreign exchange risk (4.17%) in a heterogeneous manner to the market practices, as a differential of inflation, as this is applied on a Ke (nominal in USD), in a scenario where the CPI US inflation is 2.30% and

    Page 63


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    IPCA, 4.23%. The result of this is WACC, in real terms, which is higher compared to the number obtained in calculation by Service B.

    3. THE RESULT OF ECONOMIC AND FINANCIAL ASSESSMENTS AND

    BY MULTIPLES

    Service A's assessment calculated the enterprise value from R$ 2,476,185,262.79 to Cepisa, considering, as already mentioned, the granting of a new concession. After deducting the debts, liabilities with suppliers and contingencies, Cepisa’s equity value calculated by Service A is negative R$ 91,400,895.59, as shown below. To obtain these values, the consultancy used the WACC of 7.57%.

    Table 39.  Enterprise value - Service A   
     
        Free Cash Flow to Firm (FCFF) 
    Base date of Enterprise Value    Feb./2018  Dec./2016 
    Adjustment on Enterprise Value    -  IPCA 
    Enterprise Value (BRL)  A  2,605,962,312.59  2,476,185,262.79 
    Cash (BRL)  B  4,791,748.50  4,791,748.50 
    Gross Debt (BRL)  C  -(2,244,464,437.23)  -(2,244,464,437.23) 
    Net Debt (BRL)  B+C  -(2,239,672,688.73)   -(2,239,672,688.73) 
    Tax, Labor and Social Security Contingencies (BRL)  D  -(53,301,000.00)  -(53,301,000.00) 
    Legal Contingencies (BRL)  AND  -(50,980,862.47)  -(50,980,862.47) 
    Environmental Adjustments (BRL)  F  -(40,829,816.00)  -(40,829,816.00) 
    Valuation (R$)  A+B+C+D+E+F  221,177,945.39  91,400,895.59 

    Source: Assessment Report of Eletrobras Distribuição Piauí, Service A.

    Service B calculated the enterprise value of R$ 2,880,719,840.98 for Cepisa, in view of the granting of a new concession. After deducting the debts, liabilities with suppliers and contingencies, Cepisa’s equity value calculated by Service B is negative R$

    495,935,473.78, as shown below.

    Page 64


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 40.  Enterprise value - Service B   
    Result    Dec./16 
    Operation NPV    2,497,948,936.57 
    NPV of Terminal Values    382,770,904.40 
    Enterprise Value    2,880,719,840.98 
    (-) Net Revenue    (2,239,672,688.73) 
    (-) Contingencies    (104,281,862.47) 
    (-) Environmental Adjustments    (40,829,816.00) 
    Equity Value    495,935,473.78 

    Source: Product 10: Cepisa Conclusive Economic and Financial Assessment Report, Service B.

    In order to obtain these values, Service B used rolling WACC, which considers a movable (leveraged) capital structure, as follows.

      Table 41.  Rolling WACC - Service B         
     
    Parameters  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  > 2026 
    Rate Free from Risk (Rf)  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78% 
    Country Risk  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53% 
    Market Premium  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90% 
    Non-Leveraged Beta (²u)  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43 
    % Debt (Kd) - [D/(D+E)]  54.0%  50.7%  47.3%  44.0%  40.6%  37.3%  33.9%  30.6%  27.3%  23.9%  23.9% 
    % Own Capital (Ke) - [E/(D+E)]  46.0%  49.3%  52.7%  56.0%  59.4%  62.7%  66.1%  69.4%  72.7%  76.1%  76.1% 
    Taxes - IRPJ/CSLL  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0% 
    Leveraged Beta (²L)  0.76  0.72  0.68  0.65  0.62  0.59  0.57  0.55  0.53  0.52  0.52 
    Premium Size  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00% 
    Inflation - Brazil  4.33%  4.43%  4.36%  4.31%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23% 
    Inflation - USA  2.40%  2.30%  2.30%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40% 
    Cost of Own Capital  12.62%    12.55%   12.22%    11.83%     11.55%  11.37%  11.21%  11.06%   10.93%  10.81%  10.81% 
    Selic  10.71%  9.07%  8.87%  8.87%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73% 
    Spread  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64% 
    Cost of Debt  11.35%  9.71%  9.67%  9.5%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37% 
    Debt Cost - Net of Taxes  7.49%  6.41%  6.38%  6.28%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18% 
    WACC in current CRL (Post-Tax)  9.85%  9.44%  9.46%  9.39%  9.37%  9.44%  9.50%  9.57%  9.64%  9.71%  9.71% 

    Source: Product 10: Cepisa Conclusive Economic and Financial Assessment Report, Service B.

    With respect to the multi-market valuation, Service A, in its Assessment Report of Eletrobras Distribuição Piauí, attached to this Appendix, obtained a Cepisa enterprise value

    Page 65


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    equal to R$ 2,046,865,000.00, which represents a difference of 20% in connection with the value of the economic and financial assessment.

    Service B, however, in Product 10: Cepisa Conclusive Economic and Financial Assessment Report, attached to this Appendix, did not obtain the value of enterprise value, but performed comparative analyzes of different implicit multiples to the result of the Discounted Cash Flow assessment.

    3.1 AFAC

    The study presented considers that all Advances for Future Capital Increase - AFAC’s made by Eletrobras in AmD and which are accounted for will be converted into debt or loan. On October 17, 2017, AmD had R$ 344,4 million in AFAC accounted for. The AFAC contracts already provide for the conditions to be considered in case of conversion into debt. Accordingly, Eletrobras must convert the AFACs into debt, according to a loan agreement executed under the following conditions:

    4.      CONTINGENCIES
    4.1      LEGAL CONTINGENCIES

    The due diligence of Cepisa, performed by Service B and included in the Product 4: Report of Legal Due Diligence Cepisa pointed out some critical points, although such points do not impede the process of privatization of the distribution company, the new investor should be aware of them so that it can take the applicable measures.

    Page 66


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    5. CAPITAL STRUCTURE

    The developed structure is aimed at enabling the sale of the distribution companies of Eletrobras in the business model associated with the grant of the concession. The following is the current Cepisa corporate structure.

    Table 42.  Capital structure of Cepisa on 12/31/2016 
    Total Shares        779,223,552 
    Capital    R$  1,272,747,000.00 
    Par Value of Shares  R$    1.63335284 
     
    Participation      Shares  (%) 
    Eletrobras      779,223,552  100.00% 
    Minority      0  - % 
     
    Classes of Shares    Shares  (%) 
    Common      744,131,334  95.50% 
    Preferred      35,092,218  4.50% 

    Source: Product 11: Privatization Modeling Offer of Cepisa, Service B.

    The economic value of the company's shares was determined as of 12/31/2016. Then, the definition of values was performed, considering the mean values between Services A and B, as shown in the table below.

    Table 43.  Valuation of Cepisa Average of Service A and B 
    Service A    R$  91,400,895.59 
    Service B    R$  495,935,473.78 
    Average of Services A and B  R$  293,668,184.68 

    Source: Product 11: Privatization Modeling Offer of Cepisa, Service B.

    As mentioned in item 6.3, adjustments to Cepisa´s equity value were required. The adjustments made were intended to contemplate the update of the balance sheet up to June 2017, which totaled negative R$ 10.2 million. Thus, the new equity value is R$ 283,420,710.76.

    Page 67


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 44.  Adjusted equity value 
    Average of Services A and B  R$  293,668,184.68 
    Consolidated Adjustments  R$  (10,247,473.93) 
    Equity Value Adjusted  R$  283,420,710.76 

    Source: Product 11: Privatization Modeling Offer of Cepisa, Service B.

    As already referred to in Item 6.3.2 of IT, in cases of positive equity values, the relaxation of regulatory parameters should be reduced to result in an equity value equal to zero. For Cepisa, the percentage reduction in the relaxation was 61.3%, according to the table below.

    Table 45.  Equity value equal to zero, owing to the adjustments in the relaxation of parameters
    Equity Value Adjusted  R$  283,420,710.76 
    Adjustment of the reduction of tariff relaxation  R$  (283,420,710.76) 
    Ending Equity Value  R$  - 
    Percent of reduction of tariff relaxation    61.3% 
    Source: Product 11: Privatization Modeling Offer of Cepisa, Service B.   

     

    At the end of the privatization process, Cepisa should present the corporate structure described below, considering the total adhesion of employees and retirees to the stock offering:

    Table 46. Cepisa's capital structure after sale
    Total Shares      12,492,680,402,577 
    Capital    R$  2,074,326,316.46 
    Par Value of Shares    R$    0.00016604 
     
    Interest  Common Shares  Preferred Shares  Total Shares  (%) 
    Investor  10,731,062,154,927  505,168,319,634  11,236,230,474,561  89.8425% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retirees  1,199,012,217,021  57,437,710,994  1,256,449,928,015  10.0575% 
    Other Minority  0  0  0  - % 
     
    Shares per Class      Total Shares  (%) 
    Common      11,930,074,371,949  95.4965% 
    Preferred      562,606,030,628  4.5035 % 
    Source: Product 11: Privatization Modeling Offer of Cepisa, Service B.     

     

    Page 68


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    5.1      CAPITALIZATION BY ELETROBRAS AND BY THE INVESTOR
    5.1.1      ELETROBRAS

    As the equity value of Cepisa, after the adjustments referred to in the previous Item, Eletrobras would have to capitalize the distribution company at R$ 50,000.00, in order to reach the minimum value.

    If Eletrobras chooses to increase its stake in up to 30% in Stage 2, as detailed in item 6.5 of IT, it will have to capitalize the distribution company in the amount of R$ 309 million, which may occur through additional conversion of debts in shares.

    5.1.2 INVESTOR

    As anticipated in Stage 2 of the modeling, the investor should make a capital contribution with subscription of new shares in the company, in the amount of R$ 720,915,595.51.

    5.2 FIXED ASSETS IN USE

    As referred to in Item 2.7 of this Appendix, it should be pointed out that the studies performed by Services A and B did not incorporate the positions of Fixed Assets in Use (AIC) into the Remuneration Base since, according to the Services, the assessors' report about the Remuneration Base does not take into account the specific assessment of the condition and characteristic of these assets concerning: (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the reassessment process. The table below shows the Fixed Assets in use, ascertained based on data of interim accounting balance sheets for each distribution company. The consideration of these assets, amounting to R$ 168.1 million at Feb./2017, would improve the results of the economic and financial assessments performed by the Services, and consequently, reduce

    Page 69


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    the amount to be contributed by Eletrobras in the distribution company to reach the minimum amount of R$ 50,000.00, as defined in CPPI Resolution 20/2017.

    6. LIQUIDATION COSTS

    In the 165th EGM, held in July 2016, Eletrobras' shareholders decided, by majority vote, to approve: (i) the non-extension of the concessions of its distribution companies; (ii) the transfer of the ownership control of the distribution companies until December 31, 2017, pursuant to Paragraph 1-A of article 8, of Law 12.783/2013; and (iii) if there is no transfer of control under the terms of the previous item, the adoption of provisions for liquidation thereof.

    Thus, the modeling study carried out by Service B calculated the cost with the liquidation of Cepisa, in the case of non-occurrence of item (ii) above or not accepted by Eletrobras of the adjustments made to reach the equity value equal to zero.

    For these cases, Service B developed a comparative analysis between the distribution company's sales value associated with the grant of a new concession and the eventual liquidation value of the company. Thus, it was possible to evaluate if the sale of the said state-owned company would be more attractive than its liquidation.

    Since the sale value considers adjustments to the equity value (equity value equal to zero), the adjustments were also considered in the liquidation value estimate. However, there were exceptions to the Tax and Negative Adjustments, since these do not impact on the liquidation value, and for the AFACs, since these can be paid in the event that the distribution companies are not sold.

    As shown below, in the case of Cepisa, the best economic option is to sell the company, considering the value of the assets and the concession. The balance between selling and liquidating is favorable for the first option, given the following tables:

    Page 70


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 47.  COMPARISON BETWEEN THE SALE AND LIQUIDATION OF CEPISA 
    Final Equity Value  R$  - 
    Value of Liquidation (Prior to Adjustments)  R$  (1,131,330,938.88) 
    + Compensation  R$  930,755,292.59 
    (-) Net Debt  R$  (2,239,672,688.73) 
    (-) Contingencies  R$  (104,281,862.47) 
    (+/-) Payments/Receipts  R$  (281,868,319.73) 
    Liquidation Amount (After Adjustments)  R$  (740,112,396.38) 
    =Value of Liquidation (Prior to Adjustments)  R$  (1,131,330,938.88) 
    (+/-) Balance Adjustments  R$  (92,404,000.00) 
    (+) RGR PPST Reincorporation  R$  483,622,542.50 
    Final Equity Value Liquidation Amount (After Adjustments)  R$  740,112,396.38 
    Source: Product 11: Privatization Modeling Offer of Cepisa, Service B.     

     

    The results presented above corroborate that the sale of the distribution company, associated to the granting of a new concession, is the best option for Eletrobras. If Eletrobras had to simply liquidate the distribution company under the scope of Eletrobras, the latter would have to pay a loss of R$ 740,112,396.38, an amount much compared to the final equity value of negative of the distribution company, which equals zero.

    In addition, in the event of liquidation of the distribution company, Eletrobras would not be entitled to the benefits arising from the relaxation of regulatory parameters, which generated a reduction in costs in the order of R$ 780 million for Cepisa.

    7.      ANALYSIS OF THE BENEFITS AND RISKS OF DISPOSAL
    7.1      BENEFITS
      (i) The sale of Cepisa's shares will result in the adoption, by Cepisa, of a legal regime similar to that of other non-state agents operating in the sector, allowing for significant operating gains;
     
     
      (ii) Improvement of Eletrobras' Ebitda and elimination of the need for disbursement of funds by Eletrobras to finance the cash deficit and investments of Cepisa, contributing to the reduction strategy of the net debt / ebitda indicator;
     
     

    Page 71


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    (iii) It prevents the need for capitalization by shareholders to meet the financial sustainability targets imposed by ANEEL for the extension of the concession, and for investments in the expansion and improvement of service quality;

    (iv) The sale together with the granting of concession allows the use of the relaxation of the regulatory parameters approved for the period from 2018 to 2023, when the tariff review of the new concession contract will occur; and

    (v) This flexibility has led to an improvement in the equity value of the distribution company of around R$ 780.0 million.

    7.2      RISKS
      (i) Frustration of the disposal of the shares or non-performance of the auction, due to several factors, notably due to possible judicial, political, economic-financial, regulatory, governmental or administrative impediments;
     
     
      (ii) Absence of investors and / or companies interested in purchasing; 
     
      (iii) Reduced competitiveness in the bidding procedure due to the current economic situation of the country; and
     
      (iv) Amendment of the minimum conditions for sale by the TCU.

    8. FINAL REMARKS

    Lastly, it should be noted that the economic and financial assessment methodology was found to be widely known and adopted by the market, which is the Free Cash Flow to Firm - FCFF, in nominal terms. In addition, the calculated discount rates were based on the CAPM - Capital Asset Pricing Model method, which estimates the return that the

    Page 72


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    shareholders expect to obtain from the company, given this is a well-established method widely used either by the market at large or by Eletrobras.

    Additionally, some points need to be highlighted. Among them, there is the need for approval of the minimum amount and the adjustments necessary for achieving such amount by the Executive Board and Board of Directors of Eletrobras, in addition to the need to submit the item to the general meeting.

    Concerning the need for adjustments, given that, with the tariff relaxation, Cepisa presented a positive equity value, such relaxation was reduced for the equity value to be zero. This way, the study provides for the need for capitalization by Eletrobras of R$ 50,000.00, in order for the company’s value to achieve the minimum value of R$ 50,000.00.

    It should be noted that the studies do not consider the fixed assets in use - AIC, amounting to R$ 168.4 million, referenced back to values of Feb./2017. If, in the modeling, AIC’s were considered unitized, even if set aside, they would be able to reduce the distribution company’s value.

    Furthermore, it should be noted that in the proposed sales model there is the possibility that Eletrobras to remain with one share of Cepisa, with the sole purpose of increasing its shareholding in the distribution company by up to 30%, after sale of the controlling interest in the company to a third party. The increase in interest, if it is the option adopted by Eletrobras, must be carried out within six months after the date of the auction.

    It is worth noting that Service B performed a comparative analysis between the option of divesting the shareholder control of the distribution company with the granting of a new concession and the simple liquidation option of the distribution company, as deliberated by the shareholders at the 165th EGM of Eletrobras. With this analysis, one may conclude that the liquidation of the distribution company, under the scope of Eletrobras, as established under the modeling proposed by BNDES, is the most onerous option to Eletrobras, given that, for such purpose, it would have to bear an additional cost of R$ 740.0 million connected to the sale scenario.

    Page 73


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Lastly, it should be noted that in the case of the distribution company’s liquidation, Eletrobras would not benefit from the relaxation of regulatory parameters, which contributed to an improvement in the value of Boa Vista Energias equity value of the order of R$ 780 million.

    Page 74


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Appendix to the DFP/DFF/DDE Joint Technical

    Information

    Analysis and Results of CERON

    Page 1


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Contents     
    1.  Objective  4 
    2.  Assumptions of economic and financial modeling  4 
    2.1  Market and consumer units  4 
    2.1.1  Evolution of Captive Market  4 
    2.1.2  Market Projections prepared by Service A  5 
    2.1.3  Market Projections prepared by Service B  9 
    2.1.4  Assumptions used and points of divergence of Services A and B  14 
    2.1.5  Evolution of Consumer Units  15 
    2.1.6  Projections of Consumer Units prepared by Service A  16 
    2.1.7  Projections of Consumer Units prepared by Service B  17 
    2.1.8  Comparison of Services A and B projections  18 
    2.1.9  Energy procurement  19 
    2.2  Network extension (Km)  19 
    2.3  Electrical Losses  21 
    2.3.1  Projections of Electrical Losses prepared by Service A  22 
    2.3.2  Projections of Electrical Losses prepared by Service B  24 
    2.3.3  Assumptions used and points of divergence of Services A and B  26 
    2.4  Default and unrecovarable revenue  27 
    2.5  PMSO (personnel, material, services AND OTHERS)  30 
    2.5.1  PMSO accomplished (2012 - 2016)  30 
    2.5.2  PMSO for Service A  31 
    2.5.3  PMSO for Service B:  37 

     

    Page 2


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

      2.5.4 Forecasts and comparisons of the PMSO of Services A and B  40 
      2.6  Investment  46 
      2.7  Regulatory Remuneration Base  49 
      2.8  X Factor  52 
      2.9  WACC  57 
    3.    The result of economic and financial assessments and by multiples  59 
    4.    Contingencies  61 
      4.1  Legal Contingencies  61 
      4.2  Accounting Contingencies  61 
    5.    capital structure  63 
      5.1  Capitalization by Eletrobras and the Investor  65 
      5.1.1 Eletrobras  65 
      5.1.2 Investor  65 
    6.    Necessary adjustments at Eletrobras  65 
      6.1  Fixed Assets in use  68 
    7.    Liquidation Costs  68 
    8.    Analysis of the benefits and risks of disposal  70 
      8.1  Benefits  70 
      8.2  Risks  70 
    9.    Final remarks  71 

     

    Page 3


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    1. OBJECTIVE

    The purpose of this appendix is to include a detailed presentation of the results of the economic and financial evaluations of Ceron performed by Services A and B, contracted by BNDES, as well as to detail the internal adjustments that Eletrobras needs to implement in order to enable its sale at the target price determined by the model approved under CPPI Resolution 20, dated November 8, 2017, amended by CPPI Resolution 28, dated November 22, 2017.

    2.      ASSUMPTIONS OF ECONOMIC AND FINANCIAL MODELING
    2.1      MARKET AND CONSUMER UNITS
    2.1.1      EVOLUTION OF CAPTIVE MARKET

    In the last ten years, from 2007 to 2016, Ceron had an average annual growth of 6.7%, according to the tables below. Residential, Rural and Public Lighting classes contributed to this evolution, with average annual growth of 8.1%, 8.1% and 8.4%, respectively.

      Table 1.  Evolution of consumption classes in MWh     
     
            ENERGY SEGREGATED BY CLASS (MWh)       
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  579,235  619,770  685,006  793,957  875,463  1,060,690  1,084,116  1,157,444  1,177,038  1,151,499 
    INDUSTRIAL  271,057  300,797  320,707  394,086  432,204  457,204  500,251  525,662  430,214  388,944 
    COMMERCIAL  371,994  387,880  426,766  480,267  527,469  603,134  598,550  632,376  645,799  627,581 
    RURAL  154,165  171,143  188,779  208,048  229,591  249,313  264,268  285,676  304,871  311,149 
    GOVERNMENT  134,238  138,986  151,603  163,489  172,348  195,237  199,189  208,490  223,951  220,113 
    PUBLIC LIGHTING  66,832  71,543  79,368  83,991  82,532  91,604  122,634  129,892  130,506  133,664 
    PUBLIC SERVICE  46,638  49,227  50,851  49,529  47,825  52,009  49,217  48,549  48,694  50,599 
    OWN CONSUMPTION  3,293  3,308  3,833  3,829  3,697  3,894  3,631  3,966  5,607  5,467 
    TOTAL CONSUMPTION  1,627,452  1,742,654  1,906,913  2,177,196  2,371,129  2,713,084  2,821,857  2,992,056  2,966,680  2,889,016 
     
     
     
      Table 2.  Composition of Consumption Classes     
     
              COMPOSITION OF CLASSES       
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  35.6%  35.6%  35.9%  36.5%  36.9%  39.1%  38.4%  38.7%  39.7%  39.9% 
    INDUSTRIAL  16.7%  17.3%  16.8%  18.1%  18.2%  16.9%  17.7%  17.6%  14.5%  13.5% 
    COMMERCIAL  22.9%  22.3%  22.4%  22.1%  22.2%  22.2%  21.2%  21.1%  21.8%  21.7% 
    RURAL  9.5%  9.8%  9.9%  9.6%  9.7%  9.2%  9.4%  9.5%  10.3%  10.8% 
    GOVERNMENT  8.2%  8.0%  8.0%  7.5%  7.3%  7.2%  7.1%  7.0%  7.5%  7.6% 
    PUBLIC LIGHTING  4.1%  4.1%  4.2%  3.9%  3.5%  3.4%  4.3%  4.3%  4.4%  4.6% 
    PUBLIC SERVICE  2.9%  2.8%  2.7%  2.3%  2.0%  1.9%  1.7%  1.6%  1.6%  1.8% 
    OWN CONSUMPTION  0.2%  0.2%  0.2%  0.2%  0.2%  0.1%  0.1%  0.1%  0.2%  0.2% 
    TOTAL CONSUMPTION  100%  100%  100%  100%  100%  100%  100%  100%  100%  100% 

     

    Page 4


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 3.  Annual Growth Rate in the Period     
     
            ANNUAL GROWTH RATE       ANNUAL
    AVERAGE
    GROWTH 
      2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  7.0%  10.5%  15.9%  10.3%  21.2%  2.2%  6.8%  1.7%  -2.2%  8.1% 
    INDUSTRIAL  11.0%  6.6%  22.9%  9.7%  5.8%  9.4%  5.1%  -18.2%  -9.6%  4.7% 
    COMMERCIAL  4.3%  10.0%  12.5%  9.8%  14.3%  -0.8%  5.7%  2.1%  -2.8%  6.1% 
    RURAL  11.0%  10.3%  10.2%  10.4%  8.6%  6.0%  8.1%  6.7%  2.1%  8.1% 
    GOVERNMENT  3.5%  9.1%  7.8%  5.4%  13.3%  2.0%  4.7%  7.4%  -1.7%  5.7% 
    PUBLIC LIGHTING  7.0%  10.9%  5.8%  -1.7%  11.0%  33.9%  5.9%  0.5%  2.4%  8.4% 
    PUBLIC SERVICE  5.6%  3.3%  -2.6%  -3.4%  8.7%  -5.4%  -1.4%  0.3%  3.9%  1.0% 
    OWN CONSUMPTION  0.5%  15.9%  -0.1%  -3.4%  5.3%  -6.8%  9.2%  41.4%  -2.5%  6.6% 
    TOTAL CONSUMPTION  7.1%  9.4%  14.2%  8.9%  14.4%  4.0%  6.0%  -0.8%  -2.6%  6.7% 
     
     
     
     
    2.1.2 MARKET PROJECTIONS PREPARED BY SERVICE A         

     

    The consumption of electricity in Brazil can be segregated into eight distinct classes: Residential, Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service and Own consumption. In order to capture the possible heterogeneities among the analyzed groups, Service A separately modeled the Residential, Commercial, Industrial, Rural and Other sectors (which compose the classes of Public Power, Public Lighting, Public Service and Own consumption). This decision was based on the similar behavior of consumption among them and to minimize the effect of the volatility incurred in low consumption classes.

    To forecast the energy demand for the next thirty years, ending in February/2048, the consultant first made a historical analysis of the evolution of energy consumption and its distributions between classes, using an observable period of consumption from 2007 to 2016. The results are presented in the following tables:

    Page 5


     



     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    from the utility in the period from 2013 to 2016. The table below presents the evolution of the market by level of voltage in the period:

      Table 6.  Market Evolution by Voltage Level     
     
      Market by Voltage Level (MWh)      Variation Rate 
    Voltage / Year  2013  2014  2015  2016  Period  Geometric 
              (2013/2016)  (p.a.) 
    SUPPLY  2,778,836  3,001,914  2,966,893  2,900,693  4%  1% 
    A3  62,324  94,137  39,860  23,268  -63%  -28% 
    A3a  36,037  40,917  34,501  29,165  -19%  -7% 
    A4  756,062  828,510  812,491  789,327  4%  1% 
    LV  1,924,412  2,038,350  2,080,041  2,058,933  7%  2% 
    LOW INCOME  139,289  148,569  127,303  107,048  -23%  -8% 
    SUPPLY  -  -  -  -  -  - 
    OTHERS FREE  -  35,535  31,472  22,999  -  - 
    DISTRIBUTION  -  -  -  -  -  - 
    TOTAL  2,778,836  3,037,450  2,998,366  2,923,692  5%  2% 

     

    In order to subsidize the various tariff calculations and the flow of the company, Service B prepared market simulations by: (i) consumption class (Residential, Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service, Own consumption) ; and (ii) voltage level (High Voltage - A1, A2 and A3 -, Medium Voltage - A3a and A4 - and Low Voltage - LV).

    a. Market Projection by Voltage Level

    The projection of the market by voltage level took into account the evolution of the share of voltage levels in the total market since 2012, according to information provided by the utilities. From the analysis of the behavior of such share, it was possible to develop an evolution curve, based on linear extrapolation, generating the market projections for each voltage (High, Medium and Low). The evolution of the share of each Voltage level is presented below:

    Page 10


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    b.1 Result of the Residential Class

    For the Residential Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Population and presence of seasonality as an auxiliary variable, using the observed series of consumption from January 2013 to January 2017. The estimated forecasts show that consumption grows on average 2.01% p.a. between 2017 and 2048.

    b.2 Result of the Industrial class

    For the Industrial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Dynamic Model, with the Number of formal jobs, seasonality component, as auxiliary variable using the observed series of consumption of January 2012 to January 2017. The estimated forecasts show that consumption grows on average 3.04% p.a. between 2017 and 2048.

    b.3 Result of the Commercial Class

    For the Commercial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with the state GDP as auxiliary variables, using the observed series of consumption from January 2012 to January 2017. The estimated forecasts show that consumption grows on average 1.89% p.a. between 2017 and 2048.

    b.4 Result of the Rural class

    For the Rural class of the Company, the model chosen for the performance of forecasts(which proved to be more robust) was Box & Jenkins, with population as an auxiliary variable, using the observed series of consumption from January 2012 to January 2017. The estimated forecasts show that consumption grows on average 1.63% p.a. between 2017 and 2048.

    b.5 Result of the Government class

    For the Public Power class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was Box & Jenkins , with population as an auxiliary variable, using the observed series of consumption from January 2012 to January 2017. The estimated forecasts show that consumption grows on average 1.42% p.a. between 2017 and 2048.

    Page 12


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    b.6 Result of Public Lighting class

    For the Public Lighting Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with the presence of seasonality, using the observed series of consumption from January 2013 to January 2017. The estimated forecasts show that consumption grows on average 1.65% p.a. between 2017 and 2048.

    b.7 Result of the Public Service class

    For the Public Service class of the Company, the model chosen for the performance of

    forecasts (which proved to be more robust) was

    Box & Jenkins , with population as an

    auxiliary variable, using the observed series of consumption from January 2013 to January 2017. The estimated forecasts show that consumption grows on average 0.66% p.a. between 2017 and 2048.

    b.8 Result of Own consumption Class

    For the Public Lighting Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with a structural break in September 2015 and seasonality, using the observed series of consumption from January 2012 to January 2017. The estimated forecasts show that consumption grows on average 0.19% p.a. between 2017 and 2048.

    b.9 Global Projection Result

    The table below shows the market projected by consumption class broken down every ten years between 2016 and 2046. The consumption in MWh per class is presented by means of its percentage share in relation to the total consumption and the percentage growth between the ten-year periods. The following graph illustrates the composition of the consumption class in 2046.

    Table 8.  Market Designed every Decade by Consumption Class.   
     
        MWh      Share    Average growth p.a. 
     CLASS  2016  2026  2036  2046  2016  2026  2036  2046  2017/2026  2027/2036  2037/2046 
    RESIDENTIAL  1,151,498  1,430,504  1,701,015  2,077,271  39.86%  38.58%  38.33%  39.26%  2.19%  1.75%  2.02% 
    Industrial  388,944  566,109  737,639  961,142  13.46%  15.27%  16.62%  18.16%  3.84%  2.68%  2.68% 
    COMMERCIAL  627,581  820,260  977,302  1,114,441  21.72%  22.12%  22.02%  21.06%  2.71%  1.77%  1.32% 
    Rural  311,150  394,600  458,691  511,961  10.77%  10.64%  10.33%  9.68%  2.40%  1.52%  1.10% 
    Government  220,113  267,561  302,901  337,967  7.62%  7.22%  6.82%  6.39%  1.97%  1.25%  1.10% 
    Public Lighting  133,664  167,923  196,108  220,721  4.63%  4.53%  4.42%  4.17%  2.31%  1.56%  1.19% 
    Public Service  50,598  55,595  58,908  61,916  1.75%  1.50%  1.33%  1.17%  0.95%  0.58%  0.50% 
    Own Consumption  5,467  5,790  5,790  5,790  0.19%  0.16%  0.13%  0.11%  0.58%  0.00%  0.00% 
    Total  2,889,015  3,708,342  4,438,354  5,291,209  100.00%  100.00%  100.00%  100.00%  2.53%  1.81%  1.77% 

     

    Page 13


     



     

     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    4.      Service A obtained average annual growth projections of 3.39% for the distribution company's consumption from March 2018 to February 2048. For Service B, the average annual growth of the distribution company's consumption was around 2.05% for the period from 2017 to 2048.

    Based on the Figure that describes the Projection of Energy Consumption of Service A, we see that the consumption projected by Service A is around 8,000 Gwh in 2047 (Service A's report did not contain the absolute value of the consumption projection). Service B projects a consumption of 5,291 GWh in 2046. In Absolute Values. Service A shows a difference of approximately 2,700 GWh greater than Service B, this is due to an average annual growth greater by 1.34% of Service A compared to that admitted by Service B.

    The Table below presents the average annual growth of the consumption classes of Service A and B:

    Table 9.  Geometric Average of Consumption of Services A and B for the period
     
    Geometric Average of Consumption of Service A and B
      RESIDENTIAL INDUSTRIAL COMMERCIAL  RURAL  GOVERNMENT   PUBLIC  PUBLIC   PERSONAL  TOTAL 
    Services  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a  LIGHTING  SERVICE  ”% p.a  ”% p.a 
                ”% p.a  ”% p.a     
    A  3.71%  2.84%  3.36%  3.18%  3.37%  3.37%  3.37%  3.37%  3.39% 
    B  2.01%  3.04%  1.89%  1.63%  1.42%  1.65%  0.66%  0.19%  2.05% 
    Differences (A-B)  1.70%  -0.20%  1.47%  1.55%  1.95%  1.72%  2.71%  3.18%  1.34% 
     
     
     
    2.1.5 EVOLUTION OF CONSUMER UNITS           

     

    The tables below show the evolution of the Consumer Units (UC) of consumption classes in recent years, from 2009 to 2016. The company distribution had an average annual growth of 4.2% of its UC. We highlight the Residential and Rural classes, which together account for around 92% of UC of Ceron.

    Page 15


     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.1.9 ENERGY PROCUREMENT

    In relation to the energy procurement balance, Service B considered the existing power purchase agreements, the current level of overcontracting and explained the treatment given, as well as the projection of the PLD. Service A has assumed the non-treatment of overcontracting from the starting point. According to the Disclaimer of the consultancy work: "This evaluation did not consider fines in the overcontracting of energy by the evaluated enterprise, since it was assumed that the new utility will be able to liquidate its positions at market prices." That is, the revenue from the sale of excess energy in the short-term market would be equal to the distribution company's average energy purchase cost.

    2.2 NETWORK EXTENSION (KM)

    The annual variation of the extension of the distribution network has an impact on the distribution company's projection of investments, and is part of the projection metric for the calculation of regulatory operating costs and PMSO carried out by the distribution company.

    The total extension of the network can be segregated into four voltage groups: low, medium and high voltage and Luz para Todosprogram.

    In service A, the network extension projection started from the low, medium and high voltage network extension history between 2001 and 2016. This history was added to the network extension provided by the Ten-year Plan and PDDs of the distribution companies, from which the network designed for the years 2017 to 2021 for low and medium voltage, 2017 and 2018 for the Light for All Program and high voltage from the years 2017 to 2026 was extracted.

    Page 19


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Service A presented the following projection of network extension until 2027:

      Table 14.  Network extension projection (km) - Service A     
    Service A                       
    Network Projection                       
    (km)                       
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    LV/MV                       
      61,996  62,377  62,782  63,090  63,422  63,432  63,443  63,453  63,464  63,475  63,485 
    HV                       
      1,462  2,105  2,156  2,259  2,455  2,515  2,585  2,691  2,722  2,792  2,822 
    PLPT                       
      599  1,233  1,233  1,233  1,233  1,233  1,233  1,233  1,233  1,233  1,233 
    Total                       
      64,057  65,715  66,171  66,582  67,110  67,180  67,261  67,377  67,419  67,500  67,540 

     

    That is equivalent to one of percentage growth per year, according to the respective five-year period, according to the table below:

    Table 15.  Percentage growth of network extension, per five-year period - Service A
    Service A               
      1st five-year  2nd five-year  3rd five-year  4th five-year  5th five-year  6th five-year   
    Annual growth (%)  period  period  period  period  period  period  Total Average 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
      1.0%  0.1%  0.1%  0.1%  0.1%  0.1%  0.2% 

     

    Service B did not show Network projection numbers, only presented the econometric model used to estimate the growth projection of the network extension. The total network extension of each company was projected according to a calculated elasticity for the total number of consumer units in the concession area. To determine this elasticity, panel data from the six Eletrobras distribution companies were used for the period between 2007 and 2016.

    Regarding the starting point, that is, the network extension for 2017, according to the information contained in the evaluation reports for Services A and B, the Consultants started with very different information:

  • Service A reported for 2017 the network extension numbers of 61,996 km in LV/AV, 1,462 km in HV and 599 for PLPT.
     
  • Service B reported only 52,000 km for LV/AV but did not mention the extension of HV or PLPT.
     
  • Service A did not report anything about the distribution company's isolated systems. Service B reported that there are some, but did not present any numbers.

    Page 20


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.3 ELECTRICAL LOSSES

    The losses of the electric distribution system are divided into two categories: Losses in the Basic Network and Losses in the Distribution Network. In turn, losses in the Distribution Network are divided into Technical Losses (PT) and Non-Technical Losses (PNT).

    The losses in the Basic Network are those admitted by the SIN. The real values are calculated monthly by the CCEE, and the regulatory value is a result from the average of the measurements of the last 12 months. The value of the 2016 tariff adjustment was used for the evaluation report.

    The PT are losses caused in transport and electrical equipment. They are calculated by ANEEL in the tariff review period and remain constant for the entire tariff cycle.

    PNTs are the manageable losses where ANEEL's reduction curve is estimated at each tariff review and takes into account three factors: the target, the path to the target and the limits of reduction of regulatory loss. The goal is calculated based on the analysis of the comparative model of benchmarking, in which a company is established that will serve as a basis for comparison, taking into account the size of the distribution company and the socioeconomic complexity of the region in which the distribution company is located. According to the metric specified in submodule 2.6 of PRORET. Note that the percentage levels of PNT are related to the Low Voltage (LV) market of the utility, while the PT and Loss in RB indexes are measured in relation to the Injected Energy, in accordance with the regulatory methodology.

    The percentage values of Losses are presented below, according to Ceron's Sparta worksheet of 2016, as well as the values of 2016 of the distribution company, according to NT 149/2017-ANEEL.

    According to NT 149/2017-ANEEL, the company's new PNT regulatory level will be 23.09% of the LV market, and should remain fixed until the first common tariff review, which is defined as the midpoint between the loss in 2016 and regulatory loss. The assumption is that the new utility would take five years, or a tariff cycle, to converge to regulatory benchmarking.

    Page 21


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Otherwise, the reduction curve will have a linear path until it reaches the benchmarking at the end of each tariff review cycle.

    The path of reduction of PNTs continues until reaching the level of 7.5% as determined by PRORET.

    c. Projection of the Real Non -Technical Losses

    For the projection of real PNT, three perspectives were observed: starting point, path and target.

    As a starting point the PNT in 2016 of 42.23% was considered, obtained in the energy balance of the distribution company. The defined target is the level of PNT used by the benchmark company considered in the calculation of the target of regulatory losses, in this case, the levels used by EBO which are 0%. The value of the loss used by the benchmark company was obtained from the SAMP database.

    For the PNT annual reduction path, the path from an efficient utility was used. The benchmark company selected was CEMAR - Companhia Energética do Maranhão. The choice of the company was based on the fact that CEMAR already has an efficient level of Non-Technical losses and has recently gone private.

    Table 18.Starting Point and Target of Non-Technical Losses

    % Losses / LV Market
    Non-technical Losses
    in 2016
    42.23% 
    Non-technical Losses
    Benchmark in 2016
    0.00% 
    Year in which it
    reaches the regulatory
    level
    2031 

     

    Page 23


     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    3. The modeling of the Regulatory Non-Technical Losses (NTP) of Service A was organized in two stages. At first, from 2017 to 2022, regulatory levels were maintained at 23.09%. With a sharp reduction in 2023, and in later years, there was a gradual decrease to 7.5% in 2032 and maintaining this level until the end of the concession.

    In Service B, the modeling of Regulatory PNTs was carried out based on the application of the mechanisms detailed in NT 106/2015-ANEEL. The reduction path starts with 8.3% in 2017, reaching 2047 in the value of 6.0% on the LV Market. Although Service B describes the application of NT 351/2017-ANEEL (which also addresses the same PNT parameters specified in NT 149/2017-ANEEL), the defined PNT levels were not observed in the graph. Thus, Service was consulted and ensured that the 23.09% PNT level was used in the first five years, as determined by 149/2017-ANEEL.

    4. The path of the real PNTs prepared by Service A starts from the value in 2016 of 42.23%, the curve is reduced until 2047, when it reaches the value of 1.36% of the LV market. For Service B, in the years 2017 and 2018, the real PNT indexes will remain constant at 39%, starting in 2019 a new phase in the efficient fight against PNT will begin. The path shows a marked reduction of losses in the first 5 years and then the curve gradually decreases, taking as reference the reductions of Real PNTs of the companies EBO and ESE. Thus, at the end of 2047 the projection of the real PNT index is at 5%.

     

    5. Note that the PNT carried out in Ceron in 2016 and assumed by Service A is 42.23% of the LV market, is different from that presented in NT 149/2017-ANEEL, which was 37.9% on the LV market.

    2.4 DEFAULT AND UNRECOVARABLE REVENUE

    The Projected Unrecovablel Revenue and Default is the median of the billed value that was not received until the base date of December 2016, which was the stabilization point of the unbilled revenue curve, from the 49th to the 60th month.

    Page 27


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Default is of great relevance in the analysis of the Valuation because it directly affects the collection, consequently, the company's cash flow, negatively impacting its NPV and

    Valuation.

    Analyzing the two Services, there are some divergences, which are listed below:

    The tables below show the numbers of default up to 2027 of the two Services and relevant Comments:

    Page 28


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 19.  Default until 2027 - Services A and B       
     
    Service B                       
    Unrecovarable revenue                       
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    Residential  2.95%  2.74%  2.53%  2.32%  2.11%  1.91%  1.70%  1.49%  1.28%  1.07%  0.86% 
    Commercial  2.92%  2.68%  2.43%  2.19%  1.94%  1.70%  1.46%  1.21%  0.97%  0.72%  0.48% 
    Industrial  5.88%  5.34%  4.80%  4.27%  3.73%  3.19%  2.65%  2.11%  1.58%  1.04%  0.50% 
    Rural  6.21%  5.67%  5.13%  4.59%  4.05%  3.51%  2.97%  2.43%  1.89%  1.35%  0.81% 
    Government  2.95%  2.67%  2.38%  2.10%  1.81%  1.53%  1.25%  0.96%  0.68%  0.39%  0.11% 
    Public Lighting  1.74%  1.57%  1.39%  1.22%  1.05%  0.88%  0.70%  0.53%  0.36%  0.18%  0.01% 
    Public service  0.78%  0.70%  0.63%  0.55%  0.47%  0.40%  0.32%  0.24%  0.16%  0.09%  0.01% 
    Service A*                       
    Unrecovarable revenue                       
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    Residential  5.61%  5.56%  5.52%  5.47%  5.61%  5.38%  5.34%  5.29%  5.25%  5.20%  5.16% 
    Commercial  2.02%  1.95%  1.89%  1.82%  1.64%  1.48%  1.34%  1.21%  1.09%  0.99%  0.89% 
    Industrial  1.56%  1.56%  1.56%  1.56%  1.56%  1.56%  1.56%  1.56%  1.56%  1.56%  1.56% 
    Rural  2.60%  2.60%  2.60%  2.60%  2.60%  2.60%  2.60%  2.60%  2.60%  2.60%  2.60% 
    Government  0.85%  0.78%  0.72%  0.67%  0.55%  0.46%  0.37%  0.32%  0.26%  0.22%  0.19% 
    Public Lighting  1.67%  1.53%  1.40%  1.29%  1.05%  0.85%  0.69%  0.56%  0.46%  0.37%  31.00% 
    Public service  21.33% 20.46% 19.63%  18.83%  15.34%  12.50%  10.19%  8.30%  6.77%  5.51%  4.49% 

     

    In Service A there is no specific year to reach the regulatory level, as performed by Service B. A target was established, above the regulatory level, and an annual reduction curve based on the rate of change proposed by ANEEL in 4CRTP, 3CRTP and 2CRTP. With this assumption, the regulatory level is not reached during the thirty years of concession.

    Service A included in its preparation the "Others" Market. We did not include it because it would make it difficult to compare the tables.

    Page 29


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    2.5      PMSO (PERSONNEL, MATERIAL, SERVICES AND OTHERS)
    2.5.1      PMSO ACCOMPLISHED (2012 - 2016)

    For the analysis of the PMSO item contained in the reports of Services A and B, it is necessary in advance to look at the history of this item over the last few years. For the case, the PMSO was analyzed between the period from 2012 to 2016 (same period used by Service A). Find below the values in the above-mentioned period Ceron:

      Table 20.  PMSO INCURRED (2012 - 2016)   
              R$ Million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  101.4  111.8  122.2  125.2  134.3 
    PID   0.0 23.8  0.0  0.0  0.0 
    Material   6.4  6.2  4.6  4.7  6.2 
    Third-party s.  102.7  116.6  124.2  123.5  132.0 
    Others  46.7  49.8  47.4  46.0  71.0 
    Total  257.2  308.2  298.4  299.4  343.5 

     

    For the composition of each item above, the Financial Statements of the respective years were used as support material. In heading "O" of the PMSO, the following items were considered: Arrangement and Rent; Taxes; Miscellaneous indemnities; Court costs and fees; personnel energy consumption; Inspection Fee and Others. Note that the Severance Incentive Plan (Plano de Incentivo ao Desligamento PID) was considered for the years 2013 and 2014.

    In order to show consistency in the comparison with the Appraisal Reports presented by the Consultants, the PMSO values were updated by the IPCA for December 2016.

    Table 21.  PMSO incurred from 2012 - 2016 updated to Dec / 16 
              R$ Million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  135.7  141.5  145.4  135.0  134.3 
    PID  0.0  30.1  0.0  0.0  0.0 
    Material  8.6  7.8  5.5  5.1  6.2 
    Third-party s.  137.5  147.6  147.7  133.2  132.0 
    Others  62.5  63.0  56.3  49.6  71.0 
    Total  344.4  390.1  354.9  322.9  343.5 

     

    Page 30


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 22.  PMSO values calculated by the consultancy, Dec / 2016 - Service A 
                R$ Million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  139.0  143.7  147.5  138.7  136.7 
    PID  0.0  0.0  0.0  0.0  0.0 
    Material  8.8  7.9  5.6  5.3  6.3 
    Third-party services  140.9  149.9  149.8  136.8  134.4 
    Others  37.5  49.9  112.7 (*)  3.6  56.6 
    Total  326.3  351.4  190.3 (*)  284.3  334.0 

     

    (*) The amount of R$ 112.7 Million shown in the table for item Other does not match the rationale described in the consulting report. The correct amount is R$ 225.3 Million, which generates a total PMSO of R$ 528.1 Million. Another point to note is that the sum of R$ 190.3 Million concerning the total of the 2014 PMSO is also incorrect, since the sum of the amounts described in the table is R$ 415.6 Million.

    - Methodology used in Service A: (common to all distribution companies)

    The analysis of the PMSO as well as the projection of the values were divided into two parts, namely:

    Part 1 - Characterization of Analytical Accounts (Divide in 5 phases):

  • Phase 1 - Analytical Account Clustering:
      o      Inflation adjustment of the accounts for the base date of December 2016;
      o      Clustering of accounts according to similarity of the type of expenditure;
  • Phase 2 - Analysis of the nature of the account into fixed and variable:
      o      Analysis of Fixed and Variable Accounts;
      o      The Accounts identified as fixed, when projected, were calculated as the average of the expenses of the sample period, excluding the outliers;
      o      For the projection of the variable accounts, an indicator of this account was calculated first as a function of Consumer Units, Network Expansion (km) and Consumption, in order to later carry out the projection. The outliers were also excluded.

    Page 32


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

  • Phase 3 - Identification of outliers:
      o      The values are considered outliers when:
        §      They are outside the established limits of normality of the sample, i.e.:
         
  • Upper limit out =>(sample mean) + (sample standard deviation); Lower limit out => (sample mean) - (standard deviation of the sample).
    o      After this procedure, the nature of the expenditure and its periodicity in the last 3 years were evaluated in order to assess whether or not it was excluded.
  • Phase 4 - Evaluation of the best correlations for the variable accounts:
      o      At this stage, a study was made to verify which variables were more related to the expenses of PMSO. It was concluded that Network Extension (km) and Consumption in MWh obtained higher ratio;
  • Phase 5 - Identification of regulatory accounts (deemed to be recurrent and
      paid      by ANEEL):
      o      Consists of the classification of the accounts forming the PMSO according to the characterization defined by ANEEL for regulatory PMSO. Thus, for each component account of the analytical PMSO a classification between Effective PMSO (total), Regulatory PMSO and other PMSO (Effective PMSO - Regulatory PMSO) is provided.

    Part 2 - Projection of Effective and Regulatory PMSO Clustering (Divided into 4 phases):

  • Phase 1 - Projection of Expenditure for 2017:
      o      Considers a variation of expenditure in line with the company's history, taking into account the fixed or variable nature of the expenses composing the PMSO;
      o      Variable Expenses: turned into indicators in which the denominator is the parameter with which the variable had the best correlation (it can be network km or consumption in MWh).

    Page 33


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Example: Personnel = R$/km of network => item P will increase as the network km increases;

    O Fixed expenses: Fixed expenses were calculated as the average of expenditure in the sample period, excluding outliers.

  • Phase 2 - Rate of Change of Operating Costsggg: This phase is divided into three steps:
      o      Step 1 - Definition of Operational Cost Benchmarks;
        §      For Companies from the North (Ceron, Boa Vista Energia, Eletroacre and Amazonas Energia) => Celpa and Celtins ;
        §      For Companies from the Northeast (Cepisa and Ceal) => Celpe, Cemar, Coelce, Energisa Paraíba, Energisa Sergipe, Coelba and Cosern .
      o      Step 2 - Definition of Rate of Change and Target between 2018 and 2022:
        §      DEA (Data Envelopment Analysis) database of efficient costing from
         
  • to 4CRTP is considered;
        §      The PMSO expense of each company is parameterized by the km of Network Extension;
        §      Based on the indicators, the average of the group of each distribution company according to the Region is calculated;
        §      After this step, proceed as follows:
         
  • distribution companies whose indicator is greater than the group mean, adjust their cost in terms of indicator until the next Tariff Review in 2023.
         
  • On the other hand, if the indicator is less than the group mean, the distribution companies set the indicator level between 2023 and 2027 as the target, lower than the group mean.

    Page 34


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

      o      Step 3 - Definition of Rate of Change and Target between 2023 and 2027:
        §      Consists of calculating the mean indicator of the private benchmarks of the groups;
        §      When reaching the R$/network km indicator of the cycles' target, the distribution companies stabilize parameterized operational cost indicators per consumer unit.
  • Phase 3 - Identification and Treatment of Accounts Linked to distribution company Improvements:
      o      Consists of identifying accounts that influence the actual cost of PMSO, but vary according to the distribution company's operational improvements.
  • Phase 4 - Projection of the Concession Period (2018 to 2047): Consists of estimating the PMSO spending levels of distribution companies over the concession period.
      o      Based on the targets set previously, the distribution companies reach the operational cost benchmarks by the end of 2027;
      o      Thereafter, the PMSO/network km indicator is stabilized, as a consequence, the absolute value of the PMSO varies in real terms by means of the variation of the Network km extension.
  • Costs:

    The estimation of efficient operating costs was based on Annex I - Efficiency Intervals of the submodule 2.2 V2 of PRORET. Operating costs tend to align with their respective benchmarks in two moments, the first ending 2022 and the second in 2027.

    Calculation Method:

    - First moment => simple average of each of these limits for distribution companies according to their regional groups, which are:

    Page 35


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    the amounts are: R$ 165 million against R$ 150.8 million (2018) and R$ 161 million against R$ 153 million (2019);

    - By means of the graph note that the projections of the regulatory PMSO are smaller than those of the effective PMSO.

    - The difference between the VPL of the two items, in real terms, discounted by the WACC rate of the model, considering in simplified form, a tax shield of 34%, results in a negative value of R$ 457.1 million, that is, for Ceron the PMSO is not valuable for the economic and financial appraisal;

    - Another important point is that the projection of the regulatory PMSO values took into account technical note 149/2017-ANEEL.

    2.5.3      PMSO FOR SERVICE B:
    -      Historical part:

    Service B did not present the rationale to compose its database from 2012 to 2016 as did Service A. The report shows the projections of P, M, S and O values directly for the period from 2017 to 2047.

    Page 37


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 23.  Cluster composition and efficiency - Ceron 
    Ceron Cluster
    Companies
    Efficiency in
    4CRTP
    efficiency
    indicator 
    Efficiency
    Indicator 2011
    Efficiency
    Indicator
    2013
    % (p.a.)
    COELBA  96%  0.39  0.50  -13.4% 
    EMG  83%  0.53  0.51  2.2% 
    Santa Cruz  77%  0.78  0.67  6.9% 
    EBO  73%  1.03  0.90  -2.6% 
    CELESC  62%  0.80  0.67  8.5% 
    CELPA  56%  0.72  0.98  -16.3% 
    CEB  53%  1.12  1.08  1.7% 
    Ceron  51%  0.98  0.97  0.4% 

     

    Understanding the Indicator For example, Ceron in 2011 obtained an indicator of 0.98 and in 2013, this value decreased to 0.97. This means that the company has improved slightly from 2011 to 2013 as it needs 0.97 (2013) PMSO units to offer a set of services (weighted combination of Market, Network, Consumer, Non-technical Losses).

    Indicator Projection

    1.      Phase 1: 2017-2018 => No Efficiency Gain
    2.      Phase 2: 2019-2020 => follows the rhythm of the utility with the greater gain of efficiency of the Cluster until reaching the median of the indicators of the group
    3.      Phase 3: 2021-2048 => Convergence to the mean of the Companies of the 1st Portion of the Group

    Thus, Ceron's efficiency curve as well as Projection for regulatory PMSO are shown in the chart and table below:

    Page 38


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Therefore, it is not possible to say when the company is able to reach the level of regulatory PMSO, as well as to estimate if this trajectory will generate value or result in impairment to the Company.

    In addition, it is important to take into account the issue of relaxation, since according to Technical Note No. 149/2017 - ANEEL, on the relaxation of some regulatory parameters, the new values for the Regulatory PMSO up to the first tariff review of the new controlling shareholder of the designated companies are:

    Table 25.  Flexibility of the PMSO (Aneel Technical Note No. 149/2017) 
      R$ Million 

     

    Distribution
    Company
    2016 Tariff
    Procedure
    Flexible
    Regulatory
    Level 
    Difference Difference
    (%)
    AME  349  627.2  278.4  80% 
    Boa Vista  38  69.1  31.5  84% 
    CEAL  272  336.7  65.0  24% 
    CEPISA  361  420.1  59.4  16% 
    Ceron  234.0  279.9  45.8  20% 
    ELETROACRE  94  128.5  35.1  38% 
    CEA  68  124.8  56.9  84% 

     

    2.5.4      FORECASTS AND COMPARISONS OF THE PMSO OF SERVICES A AND B
    -      Projections of Ceron until 2047:
    Table 26.  PMSO Projection Data (Service A)   
                    R$ Million 
    Item  2017  2018  2019  2020  2021  2022  2023  2024 
    Personnel  147.5  150.8  153.0  155.2  157.5  159.2  164.7  170.3 
    Material  6.4  7.3  7.8  8.4  8.9  9.5  10.2  11.0 
    Third party service  141.3  149.9  159.8  170.3  181.3  192.8  207.1  222.5 
    Others  44.9  44.5  45.9  47.2  48.6  50.0  51.8  53.6 
    Total of PMSO  340.1  352.5  366.5  381.1  396.4  411.5  433.8  457.4 
    Item  2025  2026  2027  2028  2033  2038  2043  2047 
    Personnel  175.9  181.8  187.8  193.7  225.8  262.5  304.7  343.2 
    Material  11.9  12.8  13.8  14.8  21.3  29.9  41.3  52.8 
    Third party service  238.9  256.4  274.9  295.6  419.8  585.2  802.1  1,021.0 
    Others  55.5  57.5  59.5  61.6  73.1  86.7  102.8  117.8 
    Total of PMSO  482.2  508.4  536.0  565.8  740.1  964.4  1,250.9  1,534.8 

     

    Page 40


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    In addition, Service A estimated a value for the Voluntary Dismissal Plan (PDV), defining the target public (employees with at least 20 years of company), the success rate, its cost (R$ 21.3 million) and the date of dismissal of employees, Jun./18 and Jun./19.

    Table 27.Data Projection of PMSO (Service B) - R$ Million

    Item  2017  2018  2019  2020  2021  2022  2023  2024 
    Personnel  133.4  143.0  136.7  129.8  131.0  133.4  136.3  139.4 
    Material  5.5  5.9  5.6  5.3  5.4  5.5  5.6  5.7 
    Third party service  131.5  140.9  134.7  127.9  129.1  131.5  134.3  137.9 
    Others  102.9  110.3  105.4  100.2  101.1  103.0  105.2  107.6 
    Total of PMSO  373.3  400.1  382.4  363.2  366.7  373.3  381.3  390.6 
    Item  2025  2026  2027  2028  2033  2038  2043  2047 
    Personnel  142.5  143.2  147.9  152.9  183.0  220.7  266.2  309.7 
    Material  5.8  5.9  6.1  6.3  7.5  9.0  10.9  12.7 
    Third party service  140.4  141.1  145.8  150.6  180.3  217.5  262.3  305.2 
    Others  110.0  110.5  114.2  117.9  141.2  170.3  205.4  239.0 
    Total of PMSO  398.8  400.7  413.9  427.7  512.0  617.5  744.9  866.6 

     

    The graph below shows the evolution of the PMSO of Service A compared to the PMSO of Service B.

    Page 41


     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The graphical analysis shows that for the projection of the item "Other", Service A was more conservative than Service B, the only item that was below in the PMSO. Furthermore, at the end of the projection, there is a difference of R$ 121.2 million when the projection of service A and B is compared. Finally, unlike the other items, the projection discrepancy has already been observed since the first year (2017) and so on until the end in 2047. This may be related to the composition of the item "Other" in the projection basis.

    found, there is a very marked difference in relation to the projections. This fact may be due to the basis that each Consultancy used to make the PMSO projection and the projection model considered. In relation to the first item mentioned above, only Service A presented the rationale for the initial composition of the PMSO.

    expenses, we conclude that Service B adopted a bolder projection than that of Service A. Note again that this difference in values has no significant impact on valuation if the company's PMSO is within the regulatory level.

      Table 28.  Comparison between the Services   
      Service A  % Service B  %
     Company 2017  2047  2017  2047 
    Eletroacre  112.5  684.9  509%  133.5  327.6  145% 
    CEAL  288.8  1,126.5  290%  296.7  686.0  131% 
    Amazonas  550.1  1,896.4  245%  600.2  1,267.4  111% 
    Cepisa  383.7  1,757.7  358%  452.6  1,122.8  148% 
    Ceron  340.1  1,534.8  351%  373.3  866.6  132% 
    Boa Vista  105.6  334.1  216%  216.3  371.7  72% 
    Total  1,780.8  7,334.4  312%  2,072.5  4,642.1  124% 

     

    - Finally, regarding the documents presented by the consultants, Service A detailed the methodology much more than Service B, and was more coherent when presenting information.

    Page 45


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.6 INVESTMENT

    The two services hired for economic and financial appraisal adopted different assumptions for the investment projection, notably from 2023 to 2047, since both adopted the Five-Year Plan for the period from 2018 to 2022 of the company. This plan was presented in Product 7 "Operational Technical Appraisal Report of Ceron" and prepared based on the "Expansion Plan for the Electrical System (horizon 2017-2026)", the "Development Plan for Distribution - PDD for the year 2017 ", the "Results Plan for the Improvement of Distribution Services, 2015 "and the "Plan for the Temporary Provision of the Electricity Distribution Service, 2017 ". In addition, information obtained by means of the "Field Visit Report" (Annex I of the Technical and Operational Assessment Report) and the incremental investment needs projected by the company's planning area, obtained during face-to-face meetings and conference call. The only reservation to be made for this period is that Service A for the investment in Luz Para Todosconsidered only 10% of the amount since 90% is subsidized by CDE, and therefore treated as Special Obligations. In addition, approximately R$ 173 million of the investments in HV expansion and R$ 58 million of investments in AV expansion in the year 2018 were also considered as Special Obligations, since they result from subrogation.

    In relation to the other five-year periods, the main differences were due to the fact that Service A included, in addition to the projections made for implementation and renovation projected from the Five-Year Plan, other analyzes to define the investment in maintenance and replacement of partially and totally depreciated assets. While Service B was restricted only to the first analysis.

    For the projection from 2023 of investments in implementation and renewal, Service A adopted as a basic assumption for the expansion of HV, AV and LV as well as for the improvements the mean of the last years contemplated in the plans of the companies and this value was kept constant until the end of the concession period1. Service B, for electric investments (expansion and improvement), also estimated the second five-year period of the concession by the mean of the period from 2018 to 2022, however, it eliminated

    1 The expansion in HV and AV/LV was available for information until 2026 by means of the Distribution

    Development Plan 2017 (PDD 2017) and then the average of the investments from 2018 to 2026 was adopted for the other years. As for the investment in improvements, the average of expenditures from 2018 to 2022 was used.

    Page 46


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    values outside the range between the mean and +/- a standard deviation. Thus, in order to obtain the amount of investment needed for the five-year period, the investments of the years considered typical were added up and multiplied by factor 5/No. typical years. For the other five-year period, they adopted as base the previous five-year period with the reservation that for the expansion in HV it is considered that every 10 years the need for investments doubles in relation to the previous five-year period. As for the non-electric investments (infrastructure), it was assumed that every 5 years it invests 50% of what was forecast in the first five years. Service A, however, did not project them.

    For renewal (maintenance) investments, as from 2023, service A considered that investments of this type are equivalent to the value of the depreciation of the assets. To define this value, the depreciation reference of ANEEL's Normative Resolution 674/20152, was used for the current asset base plus the depreciation of the new investments of the company, calculated from the average depreciation of the asset base of the Assets Control Report. These investments were accumulated and concentrated in the two years before each tariff review according to the proportion of the reference companies.3 Service B, however, first calculated the share of the amount allocated to the investment in renovation within the period of the 2018-2022 Five Year Plan. Subsequently, it applied this percentage for all other years.

    For investments in replacement, only service A estimated them. According to the asset base of 2016, the investments estimated for implementation and renovation from 2017 to 2022, as well as the inflation adjustments and application of depreciation of each item were added up, based on Aneel Normative Resolution 674/2015, both for the asset base of 2016 and the new investments projected until 2022. Thus, the company's asset base of 2022 was formed. Therefore, the indicator Market Use Value (VMU)/New Value of Replacement (VNR) was calculated for the company and for the benchmark 4companies so that the distribution company under review reaches the same percentage of the private reference distributed between 2023 and 2027 by the average proportion of investment of the reference companies5.

    2 Resolution that approved the revision of the Equity Control Manual of the Electric Sector MCPSE. 3 Cemar, Coelce, Coelba and Celpe.

    4 Cemar, Energisa Sergipe and Energisa Paraíba.

    5 Cemar, Coelce, Coelba and Celpe.

    Page 47


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Given the assumptions and their differences we find the following values for the two consultants for Ceron:

    Table 29. Investments - Services A and B
     
    Service A              R$ Million 
      1st five-year  2nd five-year  3rd five-year  4th five-year  5th five-year  6th five-year   
                  Sum of the 
    Investment  period  period  period  period  period  period   
                  period 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
    HV Expansion  147  40  93  93  93  93  560 
    MV/LV Expansion  161  46  103  103  103  103  620 
    Improvement  124  124  124  124  124  124  744 
    Renewal (maintenance)  165  428  482  550  627  275  2,527 
    Luz para Todos  20  -  -  -  -  -  20 
    Infrastructure and Support  31  -  -  -  -  -  31 
    Replacement  -  22  -  -  -  -  22 
    Total  648  659  803  871  948  596  4,524 
    Service B              R$ Million 
      1st five-year  2nd five-year  3rd five-year  4th five-year  5th five-year  6th five-year   
                  Sum of the 
    Investment  period  period  period  period  period  period   
                  period 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
    HV Expansion  323  46  92  46  92  46  643 
    MV/LV Expansion  222  191  191  191  191  191  1,179 
    Improvement  125  79  79  79  79  79  520 
    Renewal (maintenance)  168  62  70  62  70  62  493 
    Luz para Todos  201  -  -  -  -  -  201 
    Infrastructure and Support  31  15  15  15  15  15  106 
    Replacement  -  -  -  -  -  -  - 
    Total  1,069  393  447  393  447  393  3,142 

     

    Note that the difference between the consortiums is R$ 1.4 billion, notably in item Renovation (maintenance) in which Service A estimates that from the 2nd five-year period the investments are equivalent to the value of the depreciation of the assets and Service B calculates them from their share in the 2018-2022 Five Year Plan. In addition, there is an important difference in the projection of the company's expansion program, since in order to project the next five-year periods, service B considers the full investment of the Five-Year Plan, while service A eliminates those concerning subrogation of the CCC. Note that this difference in the volume of projected investments does not have a significant impact on valuation, since the investment increases the Company's Remuneration Base, and consequently, it is contemplated in a higher tariff.

    As a comparison, find below the value of the investment made from 2012 to 2016 to the values of December 2016 by Ceron:

    Table 30.  Investment made from 2012 to 2016, Dec / 2017 
                R$ Million 
                2012-2016 five- 
    Investment  2012  2013  2014  2015  2016   
                year period 
    Updated Values (Dec/16)  281  219  189  148  142  978 

     

    Page 48


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.7 REGULATORY REMUNERATION BASE

    Both consortiums used the data contained in the valuation reports of the assets as the reference source for the projection of the Regulatory Remuneration Base (BRR), based on the date of February 28, 2017. In these reports, both the closed BRR and the incremental BRR from 3CRTP to 4CRTP were appraisal. In the case of Ceron, the following reports were used:

    Both service A and B adopted different projection criteria for the asset base unitized through Feb/17 (closed BRR - starting point for 4CRTP) and for unitization occurring after Feb/17 and new investments (incremental BRR).

    Regarding the starting point for 4CRTP, both consortiums use the full appraisal report (delivered on 07/27/2017), however, they present different criteria for the reversal of Fixed Assets in Use (AIS).

    For service A, the reversal is treated in two ways: i) reversal of the revised closed base (Closed Base of the Full Report - Closed Base of the Incremental Report) in the value set at 5.73% for the companies from the North and 5.01% for companies from the Northeast, based on the average adjustment (reversal) of the group of distribution companies appraised in the 3CRTP by region, i.e., this reversal is made only in the AIS of the difference found in the opening of the closed base compared to that approved by Aneel in 3CRTP adjusted for inflation; and ii) The incremental base between 3CRTP and 4CRTP -the lowest value was applied between the average adjustment of the group of distribution companies appraised in 3CRTP by region and that verified by the company, without evidencing in the Appraisal Report the invested amount.

    In contrast to Service A, Service B used only a disallowance methodology to reach the Remuneration Base of Feb./17 for the Distribution company. The disallowances were

    Page 49


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    between the difference in AIS of the two appraisal reports of the Closed Base and the Incremental Base. This difference would be the opening of the Closed Base. That is, the disallowances for Service B was only at the opening of the Closed Base and occurred in the AIS. The percentage of reversal was calculated from the relation between the final AIS (with disallowances) and the initial AIS (without reversals) of the incremental BRR of the 3RTP of the distribution companies of the Eletrobras group, obtaining a mean of 9.2%. If the company's reversal is lower than this mean, the average percentage of disallowances is considered (9.2%). Otherwise, the mean between the percentage of the company and the maximum percentage of disallowances (20.1%) is used. In the case of Ceron, it falls within the first criterion, since it presented a reversal of 6.5% of the AIS in the 3CRTP, adopting, then, the reversal of 9.2% for the closed base. This percentage of reversal is applied only to the difference between the AIS of the two reports and added to the value of the base of the incremental report. That is, the reversal of 9.2% is not applied to the total AIS of the Full Base, it is calculated only for the increment of the closed base.

    Therefore, the following information was obtained for the closed base, in the position of February 2017, for both services:

    Table 31.  Closed base in the position Feb/2017 - Services A and B 
     
          R$ Million 
        Closed Base   
      Description     
        Service A  Service B 
    (1) Property, Plant And Equipment in Service (New Replacement Value)  2,084  1,964 
    (2) Index of Full Utilization  31  31 
    (3) Gross Special Obligations  609  609 
    (4) Fully Depreciated Assets  287  233 
    (5) Gross yield Base = (1) - (2) - (3) - (4)  1,157  1,091 
    (6) Accumulated Depreciation  948  863 
    (7) Net AIS (Market Value In Use)  1,137  1,101 
    (8) Index of Depreciated Utilization  19  19 
    (9) Compensation Reference Value (VBR)  1,118  1,082 
    (10) Warehouse in Operation  7  7 
    (11) Deferred Assets  -  - 
    (12) Special Net Obligations  503  503 
    (13) Lands and Easements  14  9 
    (14) Base of Total Net Compensation = (1) - (6) - (8) + (10) + (11) - (12) + (13)  635  595 
    (15) RGR PLPT Balance  48  42 
    (16) RGR Balance Other Investments  103  89 
    (17) Depreciation Rate  4%  4% 
    (18) Share of Regulatory Reintegration = (5) * (17)  44  41 
    (19) Actual WACC before Taxes  11%  ND 
    (20) RGR PLPT Rate  1%  ND 
    (21) RGR Rate of Other Investments  4%  ND 
    (22) Yield (RC) = (15) * (20) + (16) * (21) + [(14) - (15) - (16)] * (19)  59  ND 

     

    Page 50


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    For the years subsequent to the 4CRTP Starting Base, both Services apply inflation adjustments according to the IPCA, additions and write-offs of the assets that were considered in the appraisal until the end of the concession period.

    However, it should be noted that both Services did not incorporate the positions of Fixed

    Assets in use into the Remuneration Base, since, according to the services, the assessors’ report of the Remuneration Base does not take into account the specific evaluation of the condition and characteristic of these assets concerning (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the revaluation process (such as the calculation of VNR, depreciation, utilization rate and valuation of losses). In regards to the case of Ceron, it should be noted that the Full Evaluation Report “Centrais Elétricas de

    Rondônia - Ceron,” performed by Deloitte, states, as the position of June 2017, the balance of R$ 501.7 million for the Fixed Assets in Use.

    In addition, Service B divides into two groups the assumptions for the projection of BRR movements. In the first group, the closed BRR is under the IPC inflation adjustment and the discount of the depreciation and write-off of the assets that went into operation until February 17 and in a second group the incremental BRR (new investments) is through the additions of assets projected after Feb/2017, less the losses related to these assets.

    Thus, as from 2018, service A adjusts for inflation, on a monthly basis, according to the IPCA, the Fixed Assets in Use, Warehousing in operation and land and easements and increases them according to their projection of the investment. The gross special obligations are also adjusted according to the IPCA and the increase in projected special obligations. On said incremental base, the average adjustment of privatized benchmark distribution companies was applied as reversal by region obtained through the 3CRTP Incremental Base. For the companies from the Northeast, Celpe, Energisa Paraíba and Energisa Sergipe were considered, resulting in a reversal value of 2.09%. For companies from the North, the value of the reversal was 2.83%, using as a benchmark Celpa and Cemar. In relation to the depreciation rate, the same verified in 3CRTP for each distribution company was kept. In relation to asset write-offs, from the ratio of fully depreciated assets to fixed assets in use observed in the 4RTP BRR, an equilibrium rate was calculated as a simple mean among Eletrobras companies. Thus, starting from the company's ratio, it varies at a constant monthly rate until reaching that equilibrium rate of 12.54% at the end of the concession.

    Page 51


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Service B, as previously mentioned, operated the BRR with different assumptions for the base that served as a starting point and the one that incorporates the new investments. The closed BRR was operated by inflation adjustment according to the IPCA and the deduction of depreciation and write-offs. The rate of depreciation, unlike service A, is not constant, the rate falls from the average level calculated in Feb / 17, as the assets are being written-off. In relation to asset write-offs, the average write-off/AIS ratio recorded in the Assessment Report prepared by American Appraisal (delivered on 06/22/2017) in the period between 3CRTP and Feb / 17 was applied in the AIS of each month. The incremental BRR (new investments) was operated based on the projection of the reversed investments in 5%. The depreciation rate used was obtained by the weighted average of the investments of the 2018-2022 Five Year Plan per group of assets and their respective depreciation rates, obtaining 3.22% of the AIS. This rate was considered constant throughout the concession. Regarding Write-offs, until Feb/27 it was deemed to be null in view of the reduced probability of assets being written off in the first 10 years. After this period, 1/3 of the write-off ratio/AIS of the closed BRR/AIS was applied in a layered manner every 5 years until reaching the value of the full ratio as of Mar / 42.

    Accordingly, given the differences in the investment projections, the reversal criteria, depreciation and write-offs, we obtained the following Net Remuneration Base for Ceron at the end of the concession (Feb / 48) according to the projections of the two services:

    Table 32.  Net Remuneration Base, Feb./2048 - Services A and B 
          R$ Million 
    Description    Service A  Service B 
    Net yield base (Feb / 48)  10,462  3,472 

     

    2.8 X FACTOR

    The primary goal of X Factor is to ensure that the balance established in the tariff review between efficient revenues and expenses is maintained in subsequent tariff repositioning. This occurs upon transfer to the consumer of potential productivity gains from the distribution segment.

    The approach adopted by Aneel adds to X Factor a mechanism to encourage the improvement of the technical and commercial quality of the services provided to the

    Page 52


     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 35.  Methodology used - Service A   
     
      Methodology    New Methodology   
      3rd CRTP         
    Indicator  Apr / 15 to  Apr / 16 to  Apr / 17 to  Apr / 18 to  Apr / 19 to 
      March 16  March 17  March 18  Mar / 19  March 20 
    DEC  50%  30%  37.5%  45%  50% 
    FEC  50%  30%  30.0%  27%  20% 
    INS      0.8%  2%  4% 
    ICO      0.4%  1%  3% 
    IAb      0.4%  1%  3% 
    FER      3.0%  7%  10% 
    IASC      3.0%  7%  10% 
    Total  100%  60%  75%  90%  100% 
     
    3 - T Component - Trajectory of operating costs     

     

    The T component of X factor is applied to Portion B so that the tariff increases or decreases for the recognized regulatory operating costs to reach the efficiency level established by Aneel.

    In defining efficient operating costs, the costs used by the distribution companies, the efficient level of costs and the characteristics of the concession area are observed. Aneel uses a model whose purpose is to define a feasible production/cost boundary as a way of identifying the level of efficiency of companies.

    In a simplified way, the identification of the efficient level of costs is obtained by comparing the distribution companies by means of a benchmarking method that takes into account the attributes of each utility. From these constraints, a target of efficient operational costs is established to be reached throughout the tariff cycle. At the time of the tariff review, the regulatory goal is compared to the operational costs coverage present in the utility's tariff, called operating costs revenue. From the difference between the regulatory goal and the operating costs revenue, a regulatory path will be calculated. Part of the difference will be incorporated at the time of the tariff review and the remaining portion will be considered for the purpose of calculating the T component of X factor.

    Services A and B, following Aneel's Technical Note No. 149 / 2017, used the value zero for T component in the period between the date of execution of the contract and the first

    Page 56


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    common tariff review, that is, from 2018 to 2022. From 2023, Services A and B found very divergent values for T component of X factor.

    2.9 WACC

    Services A and B adopted the method of Capital Asset Pricing Model - CAPM to estimate the return that the shareholders expect to arrange with the capital employed in the business, for discounting the current value of the Distribution company’s cash flow. It should be noted that this method is well-established and widely used, either by the market at large, and by Eletrobras.

    In the calculation of the Weighted Average Cost of Capital WACC, Services A and B applied different approaches of the CAPM method, thus allowing quite different results.

    Below are the main differences found in the calculation of WACC by both Services.

    Capital Structure

    Beta

    ·Service A: used comparable domestic companies. Unlevered Beta: 0.68;

    ·Service B: considered comparable domestic companies, with a larger sample. Unlevered Beta: 0.43.

     

    Ke

    ·Service A: applied the CAPM method and a “foreign exchange risk” of 4.17% at nominal Ke in Dollars to achieve a nominal Ke in Brazilian Reais. Therefore, this “foreign exchange risk,” which derives from the “Future DI” rate and “Fraction of exchange rate coupon,” works as a differential of inflation. To achieve such rate in real terms, it was deflated by IPCA (4.23%);

    ·Service B: applied the CAPM method.

     

     

    Page 57


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    The table below shows the result for Ke under real terms:         
     
      Table 36.  Profit or Loss of Real Ke Services A and B     
    Actual Ke  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  7.95%  7.81%  7.53%  7.21%  7.02%  6.85%  6.70%  6.55%  6.43%  6.31%  6.31% 
    Service B  11.73%  10.16%  10.18%  10.42%  10.42%  10.42%  10.42%  10.42%  10.84%  10.84%  10.79% 

     

    Kd

    The table below shows the Kd result in real terms after taxes.         
     
      Table 37.  Kd in real terms after taxes Services A and B     
    Actual Kd  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  4.25%  4.09%  4.27%  4.19%  4.19%  4.19%  4.19%  4.19%  3.25%  3.25%  3.25% 
    Service B  4.44%  3.34%  3.36%  3.29%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25% 

     

    WACC

    Based on the values of Beta, Ke and Kd above, Services A and B calculated WACC in real terms after taxes, according to the results included in the table below.

          Table 38.  WACC Services A and B       
    WACC  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  8.09%  7.20%  7.30%  7.39%  7.39%  7.39%  7.39%  7.39%  7.60%  7.60%  7.58% 
    Service B  5.29%  4.80%  4.89%  4.87%  4.93%  5.00%  5.06%  5.12%  5.19%  5.26%  5.26% 

     

    Out of the results stated above, it should be noted that the factor with a higher impact on the rate presented by Service A was the use of the foreign exchange risk (4.17%) in a heterogeneous manner to the market practices, as a differential of inflation, as this is applied on a Ke (nominal in USD), in a scenario where the CPI US inflation is 2.30% and

    Page 58


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    IPCA, 4.23%. The result of this is WACC, in real terms, which is higher compared to the number obtained in calculation by Service B.

    3. THE RESULT OF ECONOMIC AND FINANCIAL ASSESSMENTS AND

    BY MULTIPLES

    The appraisal of service A, made by Ceres, calculated the enterprise value of R $ 1,524,512,239.58 for Ceron, considering, as already mentioned, the grant of a new concession. After deducting the debts, liabilities with suppliers and contingencies, the equity value of Ceron calculated by Service A is negative R$ 1,636,209,285.18, as shown in the table below. To obtain these values, Service A used the WACC of 7.57%.

    Table 39.  Enterprise value - Service A   
     
        Free Cash Flow to Firm (FCFF) 
    Data base Enterprise Value    Feb/2018  December/2016 
    Adjustment on Enterprise Value    -  IPCA 
    Enterprise Value (R$)  A  1,604,412,036.98  1,524,512,239.58 
    Cash (R$)  B  20,745,000.00  20,745,000.00 
    Gross Debt (R$)  C  (2,642,243,793.15)  (2,642,243,793.15) 
    Net Debt (BRL)  B+C  (2,621,498,793.15)  (2,621,498,793.15) 
    Tax, Labor and Social Security Contingencies (BRL)  D  (367,667,000.00)  (367,667,000.00) 
    Legal Contingencies (BRL)  E  (132,838,526.86)  (132,838,526.86) 
    Environmental Adjustments (BRL)  F  (38,717,204.75)  (38,717,204.75) 
    Valuation (R$)  A+B+C+D+E+F  (1,556,309,487.78)  (1,636,209,285.18) 

    Source: Assessment Report of Eletrobras Distribuição Rondônia, Service A.

    Service B, carried out by Consortium Mais Energia B, obtained an enterprise value of R$ 1,264,573,655.87 for Ceron, in view of the grant of a new concession. After deducting the debts, liabilities with suppliers and contingencies, the equity value of Ceron calculated by Service B is negative R$ 1,896,147,868.89, according to the table below.

    Page 59


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 40.  Enterprise value - Service B 
    Result    Dec/16 
    NPV of the transaction    1,130,687,074.35 
    NPV of the Terminal Value  203,978,481.19 
    (-) Termonorte contract  (70,091,899.67) 
    Enterprise value    1,264,573,665.87 
    (-) Net Debt    (2,621,498,793.15) 
    (-) Contingencies    (500,505,526.86) 
    (-) Environmental Adjustments  (38,717,204.75) 
    Equity value    (1,896,147,868.89) 

     

    Source: Product 10: Conclusive Economic and Financial Assessment Report of Ceron, Consortium Mais Energia B.

    In order to obtain these values, Service B used the rolling WACC, which is considered a mobile (leveraged) capital structure, according to the table below.

    Table 41.  Rolling WACC - Consortium Mais Energia B     
     
    Parameters  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  > 2026 
    Risk Free Rate (Rf)  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78% 
    Country Risk  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53% 
    Market Premium  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90% 
    Unleveraged Beta (²u)  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43 
    % Debt (Kd) - [D/(D+E)]  54.0%  50.7%  47.3%  44.0%  40.6%  37.3%  33.9%  30.6%  27.3%  23.9%  23.9% 
    % Stockholders Equity (Ke) - [E / (D +  46.0%  49.3%  52.7%  56.0%  59.4%  62.7%  66.1%  69.4%  72.7%  76.1%  76.1% 
    E)]                       
    Taxes - IRPJ / CSLL  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0% 
    Leveraged Beta (²L)  0.76  0.72  0.68  0.65  0.62  0.59  0.57  0.55  0.53  0.52  0.52 
    Size Premium  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00% 
    Inflation - Brazil  4.33%  4.43%  4.36%  4.31%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23% 
    Inflation - USA  2.40%  2.30%  2.30%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40% 
    Cost of Stockholders Equity  12.62%     12.55%    12.22%   11.83%  11.55%  11.37%   11.21%    11.06%  10.93%  10.81%  10.81% 
    Selic  10.71%  9.07%  9.03%  8.87%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73% 
    Spread  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64% 
    Cost of Debt  11.35%  9.71%  9.67%  9.51%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37% 
    Cost of Debt - Net of Taxes  7.49%  6.41%  6.38%  6.28%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18% 
    WACC in R$ (Post-Tax)  9.85%  9.44%  9.46%  9.39%  9.37%  9.44%  9.50%  9.57%  9.64%  9.71%  9.71% 

    Source: Product 10: Conclusive Economic and Financial Assessment Report of Ceron, Consortium Mais Energia B.

    With regard to the appraisal by market multiples, Service A, in its Assessment Report of Eletrobras Distribuição Rondônia, attached hereto , obtained the enterprise value equal to R$ 1,543,247,000.00, which represents a difference of -1% in relation to the value of the economic and financial appraisal.

    Service B, in Product 10: Conclusive Economic and Financial Assessment Report, attached hereto, did not obtain the value of enterprise value, but made comparative analyzes of different implicit multiples to the result of the Discounted Cash Flow assessment.

    Page 60


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    4.      CONTINGENCIES
    4.1      LEGAL CONTINGENCIES

    The legal due diligence of Ceron, carried out by Service B and contained in Product 4: Legal Due Diligence Report - Ceron pointed out some critical issues, although such issues do not impede the process of privatization of the distribution company, the new investor should be aware of them so that it may take the appropriate measures.

    4.2 ACCOUNTING CONTINGENCIES

    ANEEL carried out inspections at the distribution companies of Eletrobras - AmD, Eletroacre, Ceron and Boa Vista Energia, related to the pass-through of CCC's funds, from June 2009 to June 2016, aiming to cover generation costs in compliance with the regulations in force, namely: Law 12.111/2009, Decree No. 7.246/2010, ANEEL Normative Resolution 427/2011 and Normative Resolution ANEEL No. 597/2013.

    The inspections pointed out that the management of CCC, at the time in charge of Eletrobras, was not applying a series of rules set forth in ANEEL Normative Resolution 427/2011 for the monthly calculation of the financial criteria that involve the total cost of generation of the beneficiaries.

    In the case of Ceron, ANEEL's decisions per hour are unfavorable to the distribution company, since the Agency, by means of Aneel Technical Note No. 110/2017-SFF-SFG-SRG/ANEEL of July 4, 2017, understands that Eletrobras, as the manager of CCC at the time, has to reimburse CCC in the amount of approximately R$ 733.2 million in May / 17, which, adjusted for inflation up to June/2017, amount to R$ 731.6 million, related to the excessive pass-through to Ceron. In addition, considering that in Ceron's Balance Sheet, dated 06/30/2017, a receivable from CCC of approximately R$ 3.85 billion is registered and the position of the Agency is maintained, there is a risk of an accounting difference of approximately R$ 4.6 billion.

    This issue is still under discussion between the interested parties and, therefore, ANEEL's decisions unfavorable to Ceron can be reversed. However, the potential reversal of an unfavorable decision should not occur in a timely manner for the completion of this document or even within the timeline of privatization planned for the distribution

    Page 61


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    company. If Ceron does not succeed in administrative appeals with ANEEL, it will file legal appeals to try to succeed in this matter.

    This issue was pointed out as being of great risk to the possible buyer of Ceron, and may even render the auction unfeasible, given that the amount under discussion may cause an impact of R$ 4.6 billion. In this regard, the proposed model for privatization seeks to neutralize potential effects on the appraisal due to uncertainty in the amount effectively recognized of the accounting difference of CCC's claims. Such an offer increases the assertiveness of the value for Ceron to go private.

    The modeling offer establishes that Ceron will transfer to Eletrobras the recorded claim of CCC of R$ 3.85 billion and debts, which Ceron owns, in the same amount. Thus, initially, the accounting effect of such a transfer in Ceron would be zero, and the claim risk is transferred from Ceron to Eletrobras.

    Although Ceron has recognized this amount as a claim in its balance sheet, in the event that ANEEL maintains its position on the matter, even though Ceron has filed an appeal, it is probable, based on accounting rules, that the current claim becomes a provision. That being so, a transaction that would originally be neutral for Eletrobras becomes a deficit transaction in large sums. In this case, Eletrobras, in addition to assuming a debt of R$ 3.85 billion with Ceron, would record in its result a provision of R$ 731 million from CCC, thus reaching a possible negative impact of R$ 4.6 billion on its balance sheet.

    Table 42.  Accounting impact risk 
        R$ thousand 
    Value
    ANEEL
    Balance Amount
    (Jun / 2017)
    Possible
    Accounting
    Impact 
    -731,570.40  3,847,293.00  -4,578,863.40 

     

    Notwithstanding the foregoing, it is worth mentioning, according to the legal opinion prepared by the law firm Décio Freire and Associados, dated September 18, 2017, referring to Administrative Proceeding no. 48500.004972/2016-34 of AmD, which, owing to its similarity, is being duplicated for CERON, while there is no decision of the ANEEL’s

    Board wholly or partially ratifying the conclusions made, event in which there is no administrative act that implies an obligation to reimburse the fund. Additionally, the opinion considers that there is a solid legal basis for the review of ANEEL’s rulings and

    Page 62


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    ultimately states that it is too early to provide for any potential loss connected to said Administrative Proceeding or to disclose the matter in questions in explanatory notes to the Financial Statements, as reproduced below:

    "Under all of the circumstances described above, we believe there is a solid legal basis for reviewing the provisions under the Technical Notes numbers 52/2017-SFF-SFG-SRG and 141/2017-SFF-SFG-SRG, as well as a manifest failure to comply with decisions rendered by Federal Regional Court of the 1st Region, in the records of Writ of Mandamus no. 0026107- 81.2012.4.01.3400.

    This way, we understand that it is too early to provide for any potential loss connected to the Administrative Proceeding in question, or to disclose the matter in explanatory notes to the Financial Statements.”

    5. CAPITAL STRUCTURE

    The developed structure is aimed at enabling the sale of the distribution companies of Eletrobras in the business model associated with the grant of the concession. The following is the current corporate structure of Ceron.

    Table 43.  Capital structure of Ceron on 12/31/2016 
    Total Shares        1,212,156,653 
    Capital Stock    R$  1,325,369,000.00 
    Par value of shares  R$    1.09339746 
     
    Interest      Shares  (%) 
    Eletrobras      1,212,156,653  100.00% 
    Minority Shareholders    0  - % 
     
    Classes of Stock    Shares  (%) 
    Common      1,212,156,653  100.00% 
    Preferential      0  - % 

    Source: Product 11: Privatization Modeling Offer of Ceron, Service B.

    Page 63


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The economic value of the company's shares was determined as of 12/31/2016. Then, the values were defined, considering the mean between the values of Services A and B, as shown in the following table.

    Table 44. Valuation of Ceron - Average of Services A and B

    Service A  R$  (1,636,209,285.18) 
    Service B  R$  (1,896,147,868.89) 
    Average of Services A and B  R$  (1,766,178,577.04) 
    Source: Product 11: Privatization Modeling Offer of Ceron, Service B.   

     

    As mentioned in item 6.3 of the IT, adjustments in the equity value of Ceron were necessary. The adjustments made were intended to contemplate the update of the balance sheet up to June 2017, which totaled negative R$ 106,293,886.39. Thus, the new equity value is negative R$ 1,872,472,463.42.

        Table 45.  Equity Value Final 
        Mean of Services A and B  R$  (1,766,178,577.04) 
        Consolidated Adjustments  R$  (106,293,886.39) 
        Adjusted Equity value  R$  (1,872,472,463.42) 
        Source: Product 11: Privatization Modeling Offer of Ceron, Service B. 
     
    At  the  end of the privatization process, Ceron must present the corporate structure 

     

    described below, considering the total adhesion of employees and retirees to the stock offering:

    Table 46. Ceron's capital structure after sale   
    Total Shares      15,672,967,033,247 
    Capital Stock  R$    3,465,780,235.99 
    Par value of shares  R$      0.00022113 
     
    Share  Common Shares Preferred shares    Total Shares  (%) 
    Investor  14,105,668,025,112  0  14,105,668,025,112  90.0000% 
    Eletrobras  1  0  1  0,000% 
    Employees and Retirees  1,567,299,008,134  0  1,567,299,008,134  10.0000% 
    Other Minority Shareholders  0  0  0  - % 
    Shares per class      Total Shares  (%) 
    Common      15,672,967,033,247  100.0000% 
    Preferred      0  - % 
    Source: Product 11: Privatization Modeling Offer of Ceron, Service B.       

     

    Page 64


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    5.1      CAPITALIZATION BY ELETROBRAS AND THE INVESTOR
    5.1.1      ELETROBRAS

    The scope of the proposed corporate structure, considering the symbolic value of R$ 50,000.00, assumes the conversion of Ceron's debts by Eletrobras in exchange for shares in Stage 1 of the sales model. Accordingly, there is a need to capitalize Eletrobras in the amount of R$ 1,872,522,463.42. As shown in item 6, the adjustments required by Eletrobras are not limited to the capitalization value.

    If Eletrobras chooses to increase its stake in up to 30% in Stage 2, as detailed in item 6.5 of IT, it will have to capitalize the distribution company in the amount of R$ 103.3 million, which may occur through capital contribution or conversion debt.

    5.1.2 INVESTOR

    As provided in Stage 2 of the modeling, the investor should make a capital contribution with subscription of new shares in the company. Therefore, the estimated contribution to the investor, whose amount must be paid in currency, will be R$ 241,099,855.91.

    6. NECESSARY ADJUSTMENTS AT ELETROBRAS

    According to the proposed model to reach the minimum value of R$ 50,000.00, Eletrobras must capitalize R$ 1.87 billion in Ceron, in addition to assuming a further R$ 3.85 billion, which today are backed by a CCC's claim of the same amount, which will also be transferred to Eletrobras, as explained in item 4.2.

    Thus, Eletrobras should prioritize the cheapest debts to be capitalized/assumed, since the new Controlling Company will pay other debts. The table below shows the amount of Ceron's debt by type of creditor and its associated costs:

    Page 65


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 47.  Debt amount per creditor and its costs   
     
    RO +
    AFAC
    The
    World
    Bank 
    RGR Obligation  Petrobras Total 
    CCC Related
    Parties
    Average
    cost 
    12.00% 6.59% 5.00% 8.75%
    (Selic) 
    8.75%
    (Selic) 
    Ceron  700.32  105.99  63.18  164.56  2,369.00  2,546.60  5,949.64 

     

    As can be seen in the table above, the cheapest debts of Ceron and that will be assumed by Eletrobras to be subsequently fully or partially capitalized, are as follows: i) RGR, which has a cost of 5% per year; ii) World Bank, with a cost of 6.59% and; iii) Petrobras, at SELIC cost (8.75% p.a.); iv) Related parties (debt with Eletronorte), at SELIC cost (8.75% p.a.) and RO, with an average cost of 12% p.a Note that ANEEL must authorize Eletrobras to assume RGR's debt without having to pay all of it at the time of the transfer.

    Debts with the World Bank refer to the fund raising from this bank for the implementation of the "Energy+ Project". The RGR debts refer to financing of investments obtained before the period when it was appointed as temporary public service provider and funds of RGR received during the temporary provision. The Ordinary Resources are the pass-through of Eletrobras' own funds to the distribution company under loans, in order to cover operating deficits. Related party debts were contracted with other distribution companies or Eletrobras' companies, except the Holding Company. Among the main related parties with outstanding claims are, for example, Eletronorte, Furnas and Chesf.

    The table below shows the detailing of the capitalization by Eletrobras into Ceron, according to the source of funds, to achieve the set price of R$ 50,000.00. It should be underlined that, in this scenario, the preferred option was to convert the RO debts and before the World Bank in capital, although the common resources have a higher cost (12.00%), as the debts with Petrobras are posted as bonds under the CCD’s executed.

    Table 48.  Conversion of debts in capital, per Distribution company R$ million 
      Debt  Capitalization 
      RO + AFAC  596.99 
      IBRD  105.99 
      Related Parties  1,169.54 
      Total  1,872.52 

     

    Page 66


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    If the Shareholders’ Meeting of Eletrobras decides to assume the debt claims of CCC and

    CDE recorded in the balance sheets of the distribution companies, under the terms of Paragraph 1, Article 3 of CCPI Resolution 20/2017, as amended by CPPI Resolution no. 28/2017, the table below sets forth the values of Ceron D, as well as the assumption of debts in the same amount of CCC and CDE debt claims. In this context, the debts with Petrobras are prioritized, since most of these debts are already backed up by receivables of CCC, substantiated in Contracts of Admission of Debt, wherein Eletrobras stands as a guarantor.

    Table 49.  Conversion of debts in capital and Assumption of Debt Claims of CCC 
    and CDE, per Distribution company R$ million

    Debt  Capitalization  Assumption  Total 
    RO + AFAC  -  470.49  470.49 
    IBRD  105.99  -  105.99 
    RGR  -  63.18  63.18 
    CCC Reimbursement  -  164.56  164.56 
    Related Parties  -  2,369.00  2,369.00 
    Petrobras  1,766.53  780.07  2,546.60 
    Total  1,872.52  3,847.30  5,719.82 

     

    It should be noted that the amount corresponding to the assumption of debts of R$ 3.85 billion, has as counterparty the CCC's claims recorded in the Ceron's Financial Statements, which, as explained in item 4.2, are under discussion with ANEEL, with the risk of not realizing and becoming a debt of R$ 731 million.

    The table below summarizes the impact on Eletrobras in the event that the company succeeds in its Administrative/judicial litigation and in case it cannot reverse the position of ANEEL.

    Table 50.  Impact Scenarios at Eletrobras - R$ million 
    SCENARIOS  Assumption of Debts   Assumption of CCC's  Impact on Eletrobras 
        claims   
    Current position  5,719.81  3,847.29  -1,872.52 
    ANEEL Inspection  5,719.81  -731.57  -6,451.38 

     

    Page 67


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    6.1 FIXED ASSETS IN USE

    As referred to in Item 2.7 of this Appendix, it should be pointed out that the studies performed by Services A and B did not incorporate the positions of Fixed Assets in Use (AIC) into the Remuneration Base since, according to the Services, the assessors' report about the Remuneration Base does not take into account the specific evaluation of the condition and characteristic of these assets concerning: (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the revaluation process. The table below shows the Fixed Assets in use, ascertained based on data of interim accounting balance sheets for each Distribution company. The consideration of these assets, amounting to R$ 501.7 million at Jun./2017, would improve the results of the economic and financial assessments performed by the Services, and consequently, reduce the amount to be contributed by Eletrobras in the Distribution company to reach the minimum amount of R$ 50,000.00, as defined in CPPI Resolution 20/2017.

    7. LIQUIDATION COSTS

    At the 165th EGM, held in July 2016, Eletrobras' shareholders decided, by majority vote, to approve: (i) the non-extension of the concessions of its distribution companies; (ii) the transfer of the ownership control of the distribution companies until December 31, 2017, pursuant to Paragraph 1-A of article 8, of Law 12,783/2013; and (iii) if the transfer of control does not occur as set forth in the previous item, measures will be taken for liquidation thereof.

    Thus, the modeling study made by Service B, calculated the cost with the liquidation of Ceron, in case item (ii) above does not occur or is not accepted by Eletrobras of the adjustments made in the scope of the equity value equal to zero.

    For these cases, Service B developed a comparative analysis between the distribution company's sales value associated with the grant of a new concession and the eventual settlement value of the company. Thus, it was possible to evaluate if the sale of the said state-owned company would be more attractive than its liquidation.

    Since the sale value considers adjustments in the equity value (equity value equal to zero), the adjustments were also considered in the estimated liquidation value. However,

    Page 68


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    an exception was made to the Tax Adjustments and Negative Adjustments, since they do not impact on the liquidation value and for the AFACs, since they can be paid in case the distribution company is not sold.

    As shown below, in Ceron's case, the best economic option is to sell the company, considering the value of the assets and the concession. The balance between sell and settle is favorable for the first option, according to the following table:

    Table 51.  COMPARISON BETWEEN THE SALE AND SETTLEMENT OF CERON 
    Final Equity Value  R$  (1,872,472,463.42) 
    Settlement Amount (Pre-adjustments)  R$  (2,298,110,140.19) 
    + Indemnity  R$  618,244,312.33 
    (-) Net Debt  R$  (2,621,498,793.15) 
    (-) Contingencies  R$  (500,505,526.86) 
    (+/-) Payments / Receipts  R$  205,649,867.49 
    Settlement Amount (Post-adjustments)  R$  (2,398,260,429.38) 
    =settlement Amount (Pre-adjustments)  R$  (2,298,110,140.18) 
    (+/-) Balance Adjustments  R$  (394,775,000.00) 
    (+) Reinstatement RGR PPST  R$  294,624,710.81 
    Final Equity Value (-) settlement Amount (Post-Adjustment)  R$  525,787,965.96 
     
    Source: Product 11: Privatization Modeling Offer of Ceron, Service B.     

     

    The results presented above corroborate that the sale of the Distribution company, associated to the granting of a new concession, is the option that is less onerous to Eletrobras. If Eletrobras had to simply liquidate the Distribution company under the scope of Eletrobras, the latter would have to pay a loss of R$ 2,398,260,429.38, an amount greater than R$ 525,787,965.96 when compared to the final equity value of negative R$ 1,872,472,463.42.

    Regarding the CCC claims in discussion with ANEEL, the effects shown in item 6 impact the company in both cases, disposal and liquidation. Thus, it has no direct impact on the comparison of values, so that the economic decision remains unchanged in the event of unsuccessful lawsuits.

    In addition, in the event of liquidation of the distribution company, Eletrobras would not be entitled to the benefits arising from the relaxation of regulatory parameters, which generated a reduction in costs of around R$ 500 million for Ceron.

    Page 69


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    8.      ANALYSIS OF THE BENEFITS AND RISKS OF DISPOSAL
    8.1      BENEFITS
      (i) The sale of Ceron's shares will result in the adoption of a legal regime similar to that of other non-state agents operating in the sector, allowing for significant operating gains;
     
     
      (ii) Improvement of Eletrobras' Ebitda and elimination of the need for disbursement of funds by Eletrobras to finance Ceron's cash and investment deficits, contributing to the achievement of the strategy to reduce the net debt/ebitda indicator;
     
     
      (iii) It prevents the need for capitalization by shareholders to meet the financial sustainability targets imposed by ANEEL for the extension of the concession, and for
     nvestments in the expansion and improvement of service quality;
     
     
      (iv) The disposal, along with the grant of concession allows the use of relaxation of the regulatory parameters approved in the period from 2018 to 2023, when the tariff review of the new concession contract will occur; and
      (v) This flexibility has led to an improvement in the equity value of the distribution company of around R$ 500 million.
    8.2      RISKS
      (i)      Failure in the disposal of the shares or non-occurrence of the auction, due to several factors, notably due to possible judicial, political, economic-financial, regulatory, governmental or administrative obstacles;
      (ii)      Failure in the disposal of the shares or non-occurrence of the auction, due to several factors, notably due to possible judicial, political, economic-financial, regulatory, governmental or administrative obstacles;

    (iii)Absence of investors and/or companies interested in purchasing;

    Page 70


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    (iv)Reduced competitiveness in the bidding procedure due to the current economic situation of Brazil; and

    (v) Amendment of the minimum conditions for disposal by the TCU.

    9. FINAL REMARKS

    Lastly, it should be noted that the economic and financial evaluation methodology was found to be widely known and adopted by the market, which is the Free Cash Flow to Firm - FCFF, in nominal terms. In addition, the calculated discount rates were based on the CAPM - Capital Asset Pricing Model method, which estimates the return that the shareholders expect to obtain from the company, given this is a well-established method widely used either by the market at large or by Eletrobras.

    Additionally, some points need to be highlighted. Among them, there is the need for approval of the minimum amount and the adjustments necessary for achieving such amount by the Executive Board and Board of Directors of Eletrobras, in addition to the need to submit the item to the general shareholders' meeting.

    With regard to the need for adjustments, the study establishes the need for capitalization by Eletrobras of the debts in exchange of shares at Ceron in the value of R$ 1,872.52 million, in order for the company’s value to achieve the minimum value of R$ 50,000.00.

    Additionally, Eletrobras shall assume R$ 3,847.29 million in debts under the debt claim corresponding entry, with CCC/CDE, at the same amount.

    It should be noted that the studies do not consider the fixed assets in use - AIC, amounting to R$ 501.7 million, referenced back to values of Jun/2017. If, in the modeling,

    AIC’s were considered unitized, even if set aside, they would be able to reduce the value to be disbursed in the Distribution company by Eletrobras.

    Furthermore, it should be noted that, in the sales modeling proposed, there is the possibility that Eletrobras remains with one share of Ceron, with the sole purpose of increasing its shareholding in the Distribution company by up to 30%, after sale of the controlling interest in the company to a third party. The increase in interest, if it is the option adopted by Eletrobras, must be carried out within six months after the date of the auction.

    Page 71


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    It should also be reiterated that the modeling entertained the possibility that Ceron has values of CCC previously stated as amounts receivable in the balance sheet, which were currently reverted into obligations payable, owing to technical decisions of ANEEL. The difference of understanding between Ceron and ANEEL achieves amounts as large as R$ 4,6 million. Therefore, in view of the relevance of the value discussed for the valuation of Ceron and the unpredictability of the deadline for a final decision on this matter, it was recommended by Service B that the responsibility for such debt claims - including possible benefits and burdens - is transferred from Ceron for Eletrobras´ Holding. Thus, the effect of the current questioning would be neutral on the evaluation of the market value of the Distribution company and, therefore, on its sales modeling. However, this recommendation requires the consent of the regulators and should be accompanied by considerations of the company's external auditors. Furthermore, the uncertainties existing within the scope of Ceron should also be taken into consideration, as this operating will make them be incurred by Eletrobras. Therefore, this point should be considered by the managers of the Holding.

    It is worth noting that Service B performed a comparative analysis between the option of divesting the shareholder control of the Distribution company with the granting of a new concession and the simple liquidation option of the Distribution company, as deliberated by the shareholders at the 165th EGM of Eletrobras. With this analysis, one may conclude that the liquidation of the Distribution company, under the scope of Eletrobras, as established under the modeling proposed by BNDES, is the most onerous option to Eletrobras, given that, for such purpose, it would have to bear an additional cost of R$ 525.8 million connected to the sale scenario.

    It should also be noted that in the case of the simple liquidation of the distribution company, Eletrobras would not benefit from the relaxation of regulatory parameters, which contributed to an improvement in the equity value of Ceron of R$ 500 million.

    Page 72


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Appendix to the DFP/DFF/DDE Joint Technical

    Information

    Analysis and Results of ELETROACRE

    Page 1


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    Contents     
    1.  Objective  4 
    2.  Assumptions of economic and financial modeling  4 
    2.1  Market and consumer units  4 
    2.1.1  Evolution of Captive Market  4 
    2.1.2  Market Projections prepared by Service A  5 
    2.1.3  Market Projections prepared by Service B  10 9 
    2.1.4  Assumptions used and points of divergence of Services A and B  14 
    2.1.5  Evolution of Consumer Units  16 15 
    2.1.6  Projections of Consumer Units prepared by Service A  17 16 
    2.1.7  Projections of Consumer Units prepared by Service B  18 17 
    2.1.8  Comparison of Services A and B projections  19 18 
    2.1.9  Energy procurement  19 18
    2.2  Network extension (km)  19 
    2.3  Electrical losses  21 20 
    2.3.1  Projections of Electrical Losses prepared by Service A  22 21 
    2.3.2  Projections of Electrical Losses prepared by Service B  24 23 
    2.3.3  Assumptions used and points of divergence of Services A and B  27 26 
    2.4  Default and Unrecoverable revenues  28 27 
    2.5  Pmso (personnel, material, services and others)  29 
    2.5.1  PMSO incurred (2012 2016)  29 
    2.5.2  PMSO for Service A  31 30 
    2.5.3  PMSO for Service B:  37 36 

     

    Page 2


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      2.5.4 Forecasts and comparisons of the PMSO of Services A and B  40 38 
      2.6  Investment  45 43 
      2.7  Regulatory remuneration base  48 47 
      2.8  X Factor  53 50 
      2.9  WACC  57 55 
    3.    The result of economic and financial assessments and by multiples  60 57 
      3.1  AFAC  62 58 
    4.    Contingencies  62 59 
      4.1  Legal Contingencies  62 59 
      4.2  Accounting Contingencies  63 59 
    5.    capital structure  65 61 
      5.1  Capitalization by Eletrobras and by the Investor  66 63 
      5.1.1 ELETROBRAS  66 63 
      5.1.2 Investor  67 63 
    6.    Necessary adjustments at Eletrobras  67 63 
      6.1  Fixed Assets in use  69 66 
    7.    Settlement Costs  70 66 
    8.    Analysis of the benefits and risks of disposal  72 68 
      8.1  Benefits  72 68 
      8.2  Risks  72 69 
    9.    Final Remarks  73 69 

     

    Page 3


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    1. OBJECTIVE

    The purpose of this appendix is to include a detailed presentation of the results of the economic and financial assessments of Eletroacre performed by Services A and B, contracted by BNDES, as well as to detail the internal adjustments that Eletrobras needs to implement in order to enable its sale at the target price determined by the model approved under CPPI Resolution 20, dated November 8, 2017, amended by CPPI Resolution 28, dated November 22, 2017.

    2.      ASSUMPTIONS OF ECONOMIC AND FINANCIAL MODELING
    2.1      MARKET AND CONSUMER UNITS
    2.1.1      EVOLUTION OF CAPTIVE MARKET

    In the last ten years, from 2007 to 2016, Eletrobras Distribuição Acre had average annual growth of 7.0%, as can be seen in the tables below. Residential, Rural, Public Power and Public Service classes contributed positively to this development, with an annual average growth of 7.7%, 6.9%, 6.8% and 15.5%, respectively.

      Table 1.    Evolution of consumption classes in MWh     
     
            POWER BROKEN DOWN PER CLASS (MWh)       
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  231,109  251,802  272,038  309,735  331,922  362,026  373,116  400,241  431,340  447,637 
    INDUSTRIAL  30,372  32,480  33,013  35,795  37,941  43,823  37,296  36,906  38,150  39,564 
    COMMERCIAL  115,370  123,952  130,740  143,669  159,860  188,643  182,922  198,497  213,734  207,913 
    RURAL  26,059  28,886  30.211  32,582  35,418  38,943  39,766  41,521  45,793  46,745 
    GOVERNMENT  74,516  77,135  85,376  92,392  103,168  125,536  116,910  117,278  122,049  124,137 
    PUBLIC LIGHTING  28,768  30,113  30,819  32,276  31,967  37,128  46,302  46,095  47,254  48,183 
    PUBLIC SERVICE  18,342  19,828  21,929  22,991  22,990  14,533  27,178  33,747  45,631  48,379 
    OWN CONSUMPTION  1,031  1,003  1,085  1,075  1,030  1,070  1,184  1,186  974  769 
    TOTAL CONSUMPTION  525,566  565,198  605,211  670,515  724,297  811,702  824,673  875,470  944,926  963,327 
     
     
     
     
      Table 2.  Composition of Consumption Class     
     
              COMPOSITION OF CLASSES         
      2007  2008  2009  2010  2011  2012  2013  2014  2015  2016 
    RESIDENTIAL  44.0%  44.6%  44.9%  46.2%  45.8%  44.6%  45.2%  45.7%  45.6%  46.5% 
    INDUSTRIAL  5.8%  5.7%  5.5%  5.3%  5.2%  5.4%  4.5%  4.2%  4.0%  4.1% 
    COMMERCIAL  22.0%  21.9%  21.6%  21.4%  22.1%  23.2%  22.2%  22.7%  22.6%  21.6% 
    RURAL  5.0%  5.1%  5.0%  4.9%  4.9%  4.8%  4.8%  4.7%  4.8%  4.9% 
    GOVERNMENT  14.2%  13.6%  14.1%  13.8%  14.2%  15.5%  14.2%  13.4%  12.9%  12.9% 
    PUBLIC LIGHTING  5.5%  5.3%  5.1%  4.8%  4.4%  4.6%  5.6%  5.3%  5.0%  5.0% 
    PUBLIC SERVICE  3.5%  3.5%  3.6%  3.4%  3.2%  1.8%  3.3%  3.9%  4.8%  5.0% 
    OWN CONSUMPTION  0.2%  0.2%  0.2%  0.2%  0.1%  0.1%  0.1%  0.1%  0.1%  0.1% 
    TOTAL CONSUMPTION  100%  100%  100%  100%  100%  100%  100%  100%  100%  100% 

     

    Page 4


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
      Table 3.  Annual Growth Rate in the Period     
     
              ANNUAL GROWTH RATE     AVERAGE OF 
      2008  2009  2010  2011  2012  2013  2014  2015  2016  ANNUAL 
                        GROWTH 
    RESIDENTIAL  9.0%  8.0%  13.9%  7.2%  9.1%  3.1%  7.3%  7.8%  3.8%  7.7% 
    INDUSTRIAL  6.9%  1.6%  8.4%  6.0%  15.5%  -14.9%  -1.0%  3.4%  3.7%  3.3% 
    COMMERCIAL  7.4%  5.5%  9.9%  11.3%  18.0%  -3.0%  8.5%  7.7%  -2.7%  6.9% 
    RURAL  10.8%  4.6%  7.8%  8.7%  10.0%  2.1%  4.4%  10.3%  2.1%  6.8% 
    GOVERNMENT  3.5%  10.7%  8.2%  11.7%  21.7%  -6.9%  0.3%  4.1%  1.7%  6.1% 
    PUBLIC LIGHTING  4.7%  2.3%  4.7%  -1.0%  16.1%  24.7%  -0.4%  2.5%  2,0%  6.2% 
    PUBLIC SERVICE  8.1%  10.6%  4.8%  0.0%  -36.8%  87.0%  24.2%  35.2%  6.0%  15.5% 
    OWN CONSUMPTION  -2.8%  8.2%  -0.9%  -4.2%  3.8%  10.7%  0.2%  -17.8%  -21.0%  -2.7% 
    TOTAL CONSUMPTION  7.5%  7.1%  10.8%  8.0%  12.1%  1.6%  6.2%  7.9%  1.9%  7.0% 

     

    2.1.2 MARKET PROJECTIONS PREPARED BY SERVICE A

    The consumption of electrical energy in Brazil can be segregated into eight distinct classes, which are: Residential, Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service and Own Consumption. In order to capture the possible heterogeneities among the analyzed groups, Service A separately modeled the Residential, Commercial, Industrial, Rural and Other sectors (which compose the classes of Public Power, Public Lighting, Public Service and Own Consumption). This decision was based on the similar behavior of consumption among them and to minimize the effect of the volatility incurred in low consumption classes.

    To project energy demand for the next 30 (thirty) years, ending in February/2048, the consultancy first made a historical analysis of the evolution of energy consumption and its distributions between classes, using an observable period of consumption from 2007 to 2016. The results are presented in the following tables:

    Page 5


     



     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.1.3 MARKET PROJECTIONS PREPARED BY SERVICE B

    In order to know the reality experienced by the Distribution Company and to use the historical values for the necessary projections, Service B initially collected consumption information made by the utility from 2013 to 2016. The evolution of the Market by level of voltage in the period is presented in the table below:

      Table 6.  Market Evolution by Voltage Level   
     
      Market per Voltage Level (MWh)      Change Rate 
    Voltage/Year  2013  2014  2015  2016  Period  Geometric 
              (2013/2016)  (p.a.) 
    SUPPLY  817,746  866,360  905,341  928,684  14%  4% 
    A3  -  -  -  -  -  - 
    A3a  4,183  4,973  26,171  28,167  573%  89% 
    A4  177,109  202,813  186,801  178,225  1%  0% 
    LV  636,453  658,574  692,369  722,292  13%  4% 
    LOW INCOME  36,651  69,252  84,145  69,180  89%  24% 
    SUPPLY  393  417  -  -  -  - 
    OTHER FREE  8,587  10,874  21,410  15,508  81%  22% 
    DISTRIBUTION  -  -  -  -  -  - 
    TOTAL  826,726  877,652  926,752  944,193  14%  5% 

     

    It is important to emphasize that, for year 2016, the resulting value of the market per level of voltage in the work of Service B is not in keeping with the sum of the supply and free consumers, resulting in a value of 948,110 MWh.

    In order to subsidize the various tariff calculations and the flow of the company, Service B elaborated market simulations by: (i) consumption class (Residential, Commercial, Industrial, Rural, Public Power, Public Lighting, Public Service, Own Consumption); and (ii) voltage level (High Voltage - A1, A2 and A3 -, Medium Voltage - A3a and A4 - and Low Voltage - LV).

    a. Market Projection by Voltage Level

    The market projection by voltage level took into account the evolution of the share of voltage levels in the total market since 2012, according to information made available by the utilities. From the analysis of the behavior of these shares, it was possible to construct an evolution curve, based on linear extrapolation, generating the market projections for each voltage (High, Medium and Low). The evolution of the share of each voltage level is presented below:

    Page 10


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    COMMERCIAL  GDP and one among: Population and UC Commercial  GDP and Population 
    Rural  GDP, Rural Population or Consuming Unit, Univariate  Univariate 
    Government  Population and GDP  Population 
    Public Lighting  Univariate  - 
    Public service  Population  Population 
    Own Consumption  Univariate  - 

     

    b.1 Residential Class Result

    For the Residential Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Residential Consumer Units as an auxiliary variable and presence of seasonality, using the observed consumption series from January 2012 to December 2016. From the estimated forecasts, consumption grows on average 4.09% p.a. between 2017 and 2048.

    b.2 Industrial Class Result

    For the Industrial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Dynamic Model, with the Number of formal jobs, seasonality component, as auxiliary variable, and structural break as of April 2012. using the observed series of consumption of January 2009 to December 2016. From the estimated forecasts, consumption grows on average 4.40% p.a. between 2017 and 2048.

    b.3 Commercial Class Result

    For the Commercial Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with Population and State GDP as auxiliary variables, in addition to the seasonality, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 3.23% p.a. between 2017 and 2048.

    b.4 Result of the Rural class

    For the Rural Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, and trend component, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 4.56% p.a. between 2017 and 2048.

    Page 12


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    b.5 Result of the Public Power class

    For the class of Public Power of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, with the variables of Population, seasonality and structural break as of February 2015, using the observed series of consumption from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 2.88% p.a. between 2017 and 2048.

    b.6 Result of Public Lighting class

    For the Public Lighting Class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, momentary dummy in December 2010, in addition to the presence of seasonality, using the observed series of consumption from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 0.63% p.a. between 2017 and 2048.

    b.7 Result of the Public Service class

    For the Public Service class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology, structural break dummies in July 2013 and, after March 2015 onwards, in addition to Population and presence of seasonality, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 1.71% p.a. between 2017 and 2048.

    b.8 Result of class Own Consumption

    For the Own Consumption class of the Company, the model chosen for the performance of forecasts (which proved to be more robust) was based on the Box & Jenkins methodology and momentary dummies in March 2013, April 2013 and April 2015, and presence of seasonality, using the observed consumption series from January 2009 to December 2016. From the estimated forecasts, consumption grows on average 1.17% p.a. between 2017 and 2048.

    b.9 Global Projection Result

    The table below shows the market projected by consumption class broken down every ten years between 2016 and 2046. The consumption in MWh per class is presented by means of its percentage share in relation to the total consumption and the percentage growth

    Page 13


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    2.     

    For the forecast of the demand, Service A elaborated simulations of projection only by class of consumption, distributed in five categories: Residential, Industrial, Commercial, Rural and Others (composed of Public Power, Public Lighting, Public Service and Own Consumption). For the Projection of demand by Voltage Level, the historical average of low voltage of the last 5 years of the Company was used.

     

     

    Service B has elaborated two types of simulations for forecasting demand. A projection by consumption classes, distributed in eight categories: Residential, Industrial, Commercial, Rural, Public Power, Public Lighting, Public Service and Own Consumption. And another projection per voltage level, consisting of: High Voltage (A1, A2 and A3), Medium Voltage (A3a and A4) and Low Voltage (LV).

     

    3.     

    The historical consumption period of the Distribution Company, used by Service A, is the years 2007 to 2016. Service B, on the other hand, used periods beginning in 2009 or 2012, depending on the projected consumption class and always ending in 2016.

     

    4.     

    Service A obtained average annual growth projections of 3.47% for the distribution company's consumption in the period from March 2018 to February 2048. For Service B, the average annual growth of the distribution company's consumption was around 3.72% for the period from 2017 to 2048.

     

      Based on the Figure that describes the Projection of the Energy Consumption of Service A, we see that the consumption projected by that of Service A is around 2,750 GWh in 2047 (in Service A report, the absolute value of the consumption projection was not observed). Service B projects a consumption of 2,827 GWh in 2046. In absolute terms, with an approximate difference of 75 GWh higher for Service B, this is due to a lower annual average growth of 0.25% of Service A compared to that of Service B.

    The average annual growth of the consumption classes of Service A and B are presented in the table below:

    Table 9.  Geometric Average of Consumption of Services A and B for the period 
     
        Geometric Average of the Consumptions of Services A and B      
      RESIDENTIAL  INDUSTRIAL  COMMERCIAL  RURAL  GOVERNMENT   PUBLIC  PUBLIC  OWN  TOTAL 
    Services  ”% p.a  ”% p.a  ”% p.a  ”% p.a  ”% p.a  LIGHTING  SERVICE  ”% p.a  ”% p.a 
                ”% p.a  ”% p.a     
    A  3.84%  2.79%  3.30%  3.14%  3.31%  3.31%  3.31%  3.31%  3.47% 
    B  4.09%  4.40%  3.23%  4.56%  2.88%  0.63%  1.71%  1.17%  3.72% 

     

    Page 15


     



     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    In service A, the network extension projection started from the low, medium and high voltage network extension history between the years 2001 and 2016. Added to this history was the network extension foreseen by the Decennial Plans and PDDs of the distribution companies, from which the network projected for the years 2017 to 2021 for low and medium voltage, 2017 and 2018 for 'Luz para Todos' program and high voltage from the years 2017 to 2026.

    Service A presented the following projection of network extension until the year 2027:

    Table 14.  Projection of Network Extension until the year 2027 Service A 
    Service A                       
    Forecast of Grid (Km)                       
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    LV/AV  19,669  19,798  19,941  20,088  20,256  20,416  20,575  20,735  20,894  21,054  21,214 
    HV  441  717  770  814  829  949  976  1,008  1,035  1,062  1,089 
    PLPT  904  2,042  2,042  2,042  2,042  2,042  2,042  2,042  2,042  2,042  2,042 
    Total  21,014  22,557  22,753  22,944  23,127  23,407  23,593  23,785  23,971  24,158  24,345 

     

    Which equals to one of percentage growth per year, according to the respective five-year period, according to the table below:

    Table 15.  Percentage growth of network extension, per five-year period 
     
    Service A               
     
      1st Five-  2nd Five-  3rd Five-  4th Five-  5th Five-  6th Five-   
                  Total 
    Annual Growth (%)  Year Period  Year Period  Year Period  Year Period  Year Period  Year Period   
                  Average 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
     
      2.2%  0.8%  0.8%  0.7%  0.7%  0.7%  1.0% 

     

    Service B did not show Network projection numbers, only presented the econometric model used to estimate the growth projection of the network extension. The total network extension of each company was projected according to a calculated elasticity for the total number of consumer units in the concession area. To determine this elasticity, panel data from the six Eletrobras distribution companies were used for the period between 2007 and 2016.

    Regarding the starting point, that is, the network extension for the year 2017, according to the information contained in the assessment reports of Services A and B, the Consulting Firms started with very different information:

    Page 20


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Service A reported for the year 2017 the network extension numbers of 19,669 km in LV/AV, 441 km in AT and 904 Km in PLPT.

    - Service B reported only 19,000 km for LV/AV, but did not mention the extension of HV or PLPT.

    - Service B reported that the Acre AV Network has been pioneer, since 2014, with the use of multiplexed cables in its service lines. Service A, in turn, did not refer to none of this.

    2.3 ELECTRICAL LOSSES

    The losses of the electric distribution system are divided into two categories: Losses in the Basic Network and Losses in the Distribution Network. In turn, losses in the Distribution Network are divided into Technical Losses (PT) and Non-Technical Losses (NTP).

    Losses in the Basic Network are those admitted by the SIN. The real values are calculated monthly by the CCEE, and the regulatory value is given through the average of the measurements of the last 12 months. The value of the 2016 tariff adjustment was used for the assessment report.

    PTs are the loss of transport and electrical equipment. These are calculated by ANEEL in the tariff review period and remain constant for the entire tariff cycle.

    PNTs are the manageable losses where ANEEL's reduction curve is estimated at each tariff review and takes into account three factors: the target, the trajectory to the target and the limits of reduction of regulatory loss. The goal is calculated based on the analysis of the comparative model of benchmarking, in which a company is established that will serve as a basis for comparison, taking into account the size of the distribution company and the socioeconomic complexity of the region in which the distribution company is located. According to the metric specified in PRORET submodule 2.6. It should be reminded that the percentage levels of PNT are related to the Low Voltage (LV) market of the utility, while the PT and Loss in RB indices are measured in relation to the Injected Energy, in accordance with the regulatory methodology.

    The percentage values of Losses are presented below, according to ELETROACRE's Sparta worksheet of 2016, as well as the values of 2016 of the distribution company, according to NT 149/2017-ANEEL.

    Page 21


     



     



     



     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    2.3.42.3.3 Assumptions used and points of divergence of Services A and B

    1.      Service A considered the projection of losses in the Basic Network to be 2.48% of the injected energy of the company for the entire period from 2017 to 2048.
      Service B did not mention these losses.
    2.      For Technical Losses (PT), Service A also projected them with the constant value of 9.85% of the energy injected during the concession. Service B used as a parameter PTs per voltage segment of compatible companies to model an efficient PT curve, taking into account the tariff cycles. The reduction trajectory starts with PT of 9.85% in 2017, arriving in 2047 at the value of 7.26%.
    3.      Modeling of Service A Regulatory Non-Technical Losses (NTP) was organized in two stages. At first, from 2017 to 2022, the regulatory levels were maintained at the level of 18.79%, based on the value defined in NT 149/2017 - ANEEL. With a sharp reduction in 2023, and in later years, there was a gradual decrease to 7.5% in 2032 and maintaining this level until the end of the concession.
      In Service B, the modeling of Regulatory PNTs was carried out based on the application of the mechanisms detailed in NT 106/2015-ANEEL. The reduction trajectory starts with 11.3% in 2017, reaching 1% in 2047 on the LV Market.
      Although Service B describes the application of NT 351/2017-ANEEL (which also addresses the same PNT parameters specified in NT 149/2017-ANEEL), the defined PNT levels were not observed in the graph. Thus, Service was consulted and ensured that the 18.79% PNT level was used in the first five years, as determined by 149/2017-ANEEL.
    4.      The trajectory of real PNTs prepared by Service A starts from the value in 2016 of 25.57%, the curve reduces until 2031, when it reaches the value realized by COELBA (benchmarking) of 5.05% of the market, remaining at this level until the end of the concession. For Service B, in the years 2017 and 2018, the real PNT indices will remain constant at 26%, as of 2019, a new phase in the efficient combat against PNT will begin. The path of reduction of losses considers as reference  the average of reductions of the Real PNT’s of the companies CELPE, CEMAR and COELBA. Thus, at the end of 2047, the projection of the real NTP index is 0.46%.
       
         

    Page 27


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    5.      Note that the PNT carried out in ELETROACRE in 2016 and used as assumption by Service A is 25.57% of the LV market is different from that presented in NT- 149/2017-SRM/SGT/SRD/SFF/ANEEL, which was 26.3% on the LV market.

    2.4 DEFAULT AND UNRECOVERABLE REVENUES

    The Projected Unrecoverable Revenues and Default is the median of billed value that was not received until the base date of December 2016, the stabilization point of the unbilled revenue curve, from the 49th to the 60th month.

    The default assumption has great relevance in the Valuation analysis because it directly affects the collection, consequently, the company's cash flow, negatively impacting its Valuation.

    Analyzing the two Services, there are some divergences, which are listed below:

    It should be noted that Service B numbers were estimated in a linear decline reaching the regulatory level in 10 years, according to text contained in the Assessment Report, without presenting the annual values used.

    Below are tables with the numbers of the default up to 2027 of the two Services:

    Table 19. Default Figures until 2027 - Services A and B

    Page 28


     

     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    Service B                       
    Unrecoverable Revenue                     
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    Residential  2.95%  2.77%  2.59%  2.41%  2.23%  2,05%  1.86%  1.68%  1,50%  1.32%  1.14% 
    COMMERCIAL  2.92%  2.68%  2.45%  2.21%  1.97%  1.74%  1,50%  1.26%  1.02%  0.79%  0.55% 
    Industrial  5.88%  5.37%  4.86%  4.35%  3.84%  3.33%  2.82%  2,31%  1.80%  1.29%  0.78% 
    Rural  6.21%  5.70%  5.18%  4.67%  4.15%  3.64%  3.13%  2.61%  2,10%  1.58%  1.07% 
    Government  2.95%  2.68%  2.40%  2.13%  1.85%  1.58%  1.30%  1.03%  0.75%  0.48%  0.20% 
    Public Lighting  1.74%  1.57%  1.39%  1.22%  1.05%  0.88%  0.70%  0.53%  0.36%  0,18%  0.01% 
    Public service  0.78%  0.70%  0.63%  0.55%  0.48%  0.40%  0.32%  0.25%  0.17%  0.10%  0.02% 
    Service A*                       
    Unrecoverable Revenue                     
    Class  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2027 
    Residential  4.05%  3.88%  3.71%  3.55%  3.16%  2.81%  2.51%  2.36%  2.36%  2.36%  2.36% 
    COMMERCIAL  2.03%  2.03%  2.02%  2.02%  1.85%  1.70%  1,56%  1.44%  1.32%  1.21%  1.11% 
    Industrial  4.13%  4.05%  3.97%  3.89%  3.81%  3.74%  3.66%  3.59%  3.52%  3.45%  3.38% 
    Rural  6.86%  6.76%  6.66%  6.56%  6.21%  5.87%  5.55%  5.25%  5.04%  5.04%  5.04% 
    Government  2.31%  2.13%  1.96%  1.81%  1.53%  1.29%  1.09%  0.93%  0.78%  0.66%  0.56% 
    Public Lighting  10.54%  9.67%  8.87%  8.14%  6.62%  5.38%  4.37%  3.56%  2.89%  2.35%  1.91% 
    Public service  0.14%  0.14%  0.14%  0.14%  0.14%  0.14%  0.14%  0.14%  0.14%  0.14%  0.14% 

     

    In Service A there is not a specific year to reach the regulatory level, as performed by Service B, below. A target was established, above the regulatory level, and an annual reduction curve based on the rate of change proposed by ANEEL in 4CRTP, 3CRTP and 2CRTP. With this assumption, the regulatory level is not reached during the thirty years of concession.

    Service A raised R$ 18.221 million in Unrecoverable Revenues for the year 2016. Service B did not show any value.

    The year-on-year comparison of the projection of default rates by each consulting was hampered because Service B did not disclose the annual amounts in a table, showing only the starting value and the year that the regulatory level is reached.

    2.5      PMSO (PERSONNEL, MATERIAL, SERVICES AND OTHERS)
    2.5.1      PMSO INCURRED (2012 2016)

    For the analysis of the PMSO item contained in the Services A and B reports, it is necessary to look, in advance, at the history of this item over the last few years. For the case, the PMSO was analyzed between the period from 2012 to 2016 (same period used by Service A). Below are the values in the aforementioned period for Eletroacre:

    Table 20. PMSO incurred (2012 - 2016)

    Page 29


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
              R$ Million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  40.0  46.8  38.8  42.4  45.7 
    PID  0.0  14.6  1.2  0.0  0.0 
    Material  1.2  1.0  1.1  1.0  1.2 
    Third-party Service  35.4  45.3  48.7  52.8  53.9 
    Others  1.9  22.1  22.5  25.4  34.9 
    Total  78.5  129.8  112.3  121.5  135.7 

     

    For the composition of each item above, the Financial Statements of the respective years were used as support material. In item "O" of the PMSO, the following items were considered: Leases and Rentals; Taxes; Miscellaneous indemnities; Judicial costs and fees; Own energy consumption; Inspection Fee and Others. It is also worth noting that the PID (Resignation Incentive Plan) was considered for the years 2013 and 2014.

    In order to show consistency in the comparison with the Appraisal Reports presented by the Consultants, the PMSO values were updated by the IPCA for December 2016.

    Table 21.  PMSO incurred from 2012 - 2016 updated to Dec / 16 
              R$ Million 
    PMSO  2012  2013  2014  2015  2016 
    Personnel  53.5  59.2  46.2  45.8  45.7 
    PID  0.0  18.5  1.4  0.0  0.0 
    Material  1.6  1.2  1.3  1.1  1.2 
    Third-party Service  47.4  57.3  57.9  56.9  53.9 
    Others  2.6  28.0  26.7  27.3  34.9 
    Total  105.1  164.2  133.6  131.1  135.7 

     

    In relation to regulatory operating costs, it is possible to observe the estimated values for Eletroacre in the years 2013 (last tariff review) until 2016, brought to the base date of December 2016.

    By comparing the PMSO incurred in the period 2013-2016 with the Regulatory, one can verify how far from the Regulatory PMSO the Company is. This fact can be seen in the graph below:

    Gráfico 16.Gráfico 11. PMSO incurred x PMSO regulatory 2013-2016

    Page 30


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    - Methodology used in Service A: (part common to all Distribution Companies)

    The analysis of the PMSO as well as the projection of the values was divided into two parts, namely:

    Part 1 - Characterization of Analytical Accounts (Divided in 5 phases):

  • Phase 1 - Grouping of analytical accounts:

     

      o      Adjustment for inflation of the accounts for the base date of December 2016;
      o     

    Grouping of accounts according to similarity of expense type;

     

  • Phase 2 - Analysis of the nature of the account in fixed and variable:

     

      o      Analysis of Fixed and Variable Accounts;
      o      The Accounts identified as fixed, when projected, were calculated as the average of the expenses of the sample period, excluding the outliers;
      o     

    For the projection of the variable accounts, an indicator of this account was calculated due to Consumer Units, Network Extension (Km) and Consumption (Km), and then the projection was carried out. The outliers were also excluded.

     

  • Phase 3 - Identification of outliers:

     

      o      The values are considered outliers when:
        §      They are outside the established limits of normality of the sample, i.e.:
         
  • Out Upper Limit => (sample mean) + (standard deviation of sample); Out Lower Limit => (sample mean) - (standard deviation of sample).
      o      After this procedure, the nature of the expense and its periodicity in the last 3 years were assessed in order to evaluate whether or not it would be excluded.

    Page 32


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

  • Phase 4 - Assessment of the best correlations for the variable accounts:

     

      o     

    At this phase, a study was made to verify which variables were more related like the expenses of PMSO. It was concluded that Network Extension (km) and Consumption in MWh obtained higher ratio;

     

  • Phase 5 - Identification of regulatory accounts (considered by and paid by ANEEL):
           
      o      Consists of the classification of the accounting accounts forming the PMSO according to the characterization defined by ANEEL for regulatory PMSO. Thus, for each component account of the analytical PMSO, a classification between Effective PMSO (total), Regulatory PMSO and other PMSO (Effective PMSO - Regulatory PMSO) is provided).

    Part 2 - Projection of Effective and Regulatory PMSO Grouping (Divided into 4 phases):

  • Phase 1 - Projection of Expense for 2017:

     

      o      It considers a variation of expense in line with the company's history, taking into account the fixed or variable nature of the expenses that make up the PMSO;
      o      Variable Expenses: are transformed into indicators in which the denominator is the parameter with which the variable had the best correlation (it can be network km or consumption in MWh).
        Example: Personnel = R$/Km of network => item P will increase as the network Km increases;
      o     

    Fixed Expenses: fixed expenses were calculated as the average expenses for the sample period, excluding outliers.

     

  • Phase 2 - Operating Costs variation rate: This phase is divided into three steps:     
     
      o      Step 1 - Definition of Operational Cost Benchmarks;
        §      For Northern Companies (CERON, Boa Vista Energia, Eletroacre and Amazonas Energia) => Celpa and Celtins;

    Page 33


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    §      For the Northeast Companies (Cepisa and Ceal) => Celpe, Cemar, Coelce, Energisa Paraíba, Energisa Sergipe, Coelba and Cosern.

    Step 2 - Rate Variation Definition and Target between 2018 and 2022:

    §      DEA (Data Envelopment Analysis) database of efficient costing from Aneel to 4CRTP is considered;
    §      The PMSO expense of each company is parameterized by the Network Km Extension;
    §      Based on the indicators, the average of the group of each distribution company according to the Region is calculated;
    §      After this step, proceed as follows:
     
  • Distribution companies with the indicator greater than the group mean adjust their cost in terms of indicator until the next Tariff Review in 2023.
     
  • On the other hand, if the indicator is less than the mean of the group, the Distribution Companies plot the indicator level between 2023 and 2027 as target, which is lower than the group mean.

    Step 3 - Rate Variation Definition and Target between 2023 and 2027:

    §      It consists of calculating the mean indicator of the privatized Benchmarks of the groups;
    §      Upon reaching the indicator of R$/km of network of the cycles target, the distribution companies stabilize operating cost indicators parameterized per consumer unit.
  • Phase 3 - Identification and Treatment of Accounts with Behavior Linked to
      the      Distribution Companies Improvements:
      o      It consists of identifying accounts that influence the actual cost of PMSO, but which vary according to the distribution company’s operational improvements.

    Page 34


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

  • Phase 4 - Projection of the Concession Period (2018 to 2047): It consists of
      estimating      the levels of PMSO expenses of Distribution Companies over the
      concession      period.
      o      Based on the targets set previously, the Distribution Companies reach the operational cost benchmarks until the end of 2027;
      o      Thereafter, the network PMSO/km indicator stabilizes itself, as a consequence, the absolute value of the PMSO varies in real terms through the variation of the Network Km Extension.
  • Costs:

    The elaboration of the estimate of the efficient operational costs was based on Annex I -Efficiency Intervals of the PRORET sub-module 2.2 V2. Operating costs tend to align with their respective benchmarks in two moments, the first ending 2022 and the second in 2027.

    Calculation Method:

    - First moment => simple mean of each of these limits for distribution companies according to their regional groups, which are:

    - Second moment => mean of the privatized of the regional groups mentioned above.

    It is important to note that, as Technical Note 149/2017 - ANEEL defines the operational costs in force from 2017 until 2022, the efficient cost intervals defined in this section will only impact the tariff from the Tariff Review of 2023.

    Page 35


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    2.5.3      PMSO FOR SERVICE B:
    -      Historical part:

    Service B did not present the rational to compose its database from 2012 to 2016 as did Service A. The report shows the projections of P, M, S and O values directly for the period from 2017 to 2047.

  • Methodology used in Service B (common to all Distribution Companies):
     
  • Database: Calculation of Efficiency for the 4CRTP - 4th Cycle of Periodic Tariff Reviews (PMSO, Market, Consumers, network, losses, etc. data) => 2001-2013;
     
  • Clustering of Companies: Automatic grouping of similarity second degree data. The result of this process can be seen in the table below:
  • Table 23.  Clusters composition and efficiencies - ELETROACRE 
    Companies of
    Eletroacre Cluster
    Efficiency
    Indicator
    in 4CRTP
    efficiency 
    2011 Efficiency
    Indicator
    2013
    Efficiency
    Indicator 
    % (p.a.)
    CSPE  100%  0.49  0.42  6.6% 
    COSERN  92%  0.59  0.67  -6.7% 
    Escelsa  72%  0.89  0.98  -5.1% 
    Bragantina  69%  0.92  0.80  6.8% 
    COPEL  64%  0.73  0.67  3.7% 
    CELESC  62%  0.80  0.67  8.5% 
    DME  42%  1.43  1.26  6.2% 
    Eletroacre  52%  1.02  1.22  -9.6% 

     

    Understanding the indicator: For example, Eletroacre in 2011 obtained an indicator of 1.02 and in 2013, this value decreased to 1.22. This means that the company has undergone a worsening from 2011 to 2013 as it needs 1.02 (2011) PMSO units to offer a set of services (weighted combination of Market, Network, Consumer, Non-technical Losses) and started needing 1.22.

    Page 37


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Table 24.  Projection of regulatory PMSO Eletroacre 
    Period Regulatory
    PMSO (BRL million
    2016) 
    Percentage
    change p.a. 
    2016  97  - 
    2017  101  3.5% 
    2018  105  4.2% 
    2019  100  -4.8% 
    2020  95  -4.9% 
    2021  91  -4.8% 
    2022  86  -4.8% 
    Average 2023 - 2027  89  1.3% 
    Average 2028 - 2032  93  0.4% 
    Average 2033 - 2037  94  0.3% 
    Average 2038 - 2042  95  0.3% 
    Average 2043 - 2047  97  0.3% 

     

    Although Service B shows an Efficiency curve that should be followed in the document, it is not evident the comparison with the effective PMSO values projected by the consultancy. Therefore, it is not possible to state when the company is able to reach the level of regulatory PMSO, as well as to estimate if this trajectory will generate or destroy value for the Company.

    In addition, it is important to take into account the issue of relaxation, since according to Technical Note No. 149/2017 - ANEEL, which deals with the flexibility of some regulatory parameters, the new values for the Regulatory PMSO that go until the first ordinary tariff review of the new controller of the designated companies, are:

    Table 25. Flexibility of the PMSO (Aneel Technical Note No. 149/2017)

            R$ Million 
    Distribution
    Company
    2016 Tariff
    Process
    Relaxed
    Regulatory
    Level 
    Difference Difference
    (%)
    AME  348.8  627.2  278.4  80% 
    Boa Vista  37.5  69.1  31.5  84% 
    CEAL  271.7  336.7  65.0  24% 
    CEPISA  360.7  420.1  59.4  16% 
    CERON  234.0  279.9  45.8  20% 

     

    Page 39


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     
    ELETROACRE  93.5  128.5  35.1  38% 
    CEA  67.8  124.8  56.9  84% 

     

    2.5.4      FORECASTS AND COMPARISONS OF THE PMSO OF SERVICES A AND B
    -      Eletroacre Projections until 2047:

    Table 26. PMSO Projection Data (Service A)

                    R$ Million 
    Account  2017  2018  2019  2020  2021  2022  2023  2024 
    Personnel  46.6  56.2  60.0  64.0  68.3  72.8  76.8  81.0 
    Material  1.1  1.3  1.4  1.4  1.5  1.6  1.7  1.8 
    Third-party Service  53.3  56.8  60.4  64.2  68.2  72.4  76.0  79.8 
    Others  11.5  5.3  5.7  6.3  6.8  7.4  7.9  8.5 
    Total PMSO  112.5  119.6  127.5  135.9  144.8  154.2  162.4  171.2 
    Account  2025  2026  2027  2028  2033  2038  2043  2047 
    Personnel  85.4  90.1  95.0  100.9  137.1  189.2  257.3  325.7 
    Material  1.9  2.0  2.2  2.3  3.1  4.3  5.8  7.4 
    Third-party Service  83.8  88.1  92.5  98.0  131.5  179.0  240.8  302.6 
    Others  9.2  9.9  10.6  11.6  17.8  26.3  37.7  49.2 
    Total PMSO  180.4  190.1  200.4  212.8  289.4  398.8  541.6  684.9 

     

    In addition, Service A estimated a value for the Voluntary Resignation Plan (PDV), defining the target public (employees with at least 20 years of company), the success rate, its cost (R$ 7,1 million) and the date of resignation of employees, June/18 and June/19.

      Table 27.  PMSO Projection Data (Service B)     
                    R$ Million 
    Account  2017  2018  2019  2020  2021  2022  2023  2024 
    Personnel  39.3  42.9  40.8  38.0  36.0  34.7  35.1  37.5 
    Material  1.0  1.1  1.0  1.0  0.9  0.9  0.9  0.9 
    Third-party Service  47.2  51.5  48.9  45.6  43.3  41.7  42.1  45.0 
    Others  46.0  50.2  47.7  44.4  42.2  40.6  41.0  43.8 
    Total PMSO  133.5  145.6  138.4  128.9  122.4  117.8  119.1  127.2 
    Account  2025  2026  2027  2028  2033  2038  2043  2047 
    Personnel  40.1  42.2  44.2  46.3  57.4  71.2  88.4  105.2 
    Material  1.0  1.1  1.1  1.2  1.4  1.8  2.2  2.6 
    Third-party Service  48.1  50.7  53.0  55.5  68.9  85.5  106.1  126.3 

     

    Page 40


     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    From the analysis of the graph showing the projection of Item P (Personnel), we can see that the projection of the line is not so different between the consultants. Note that the amount of the difference increases from 2018 up to 2047, when it reaches the amount of R$ 220.5 million. The same behavior can be observed in the item Material account, in which the difference is increasing from 2018, and reaches the highest value in 2047, R$ 4.7 million.

    Third Party Service and Other

    Gráfico 21.Gráfico 16. Projection of Third Party and Other Services items

    referring to Services A and B

    Page 43


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    - When comparing the methodologies used by Services A and B, and especially the values found, it is noticed that there is a very marked difference in relation to the projections. This fact may be due to the basis that each Consultancy used to make the PMSO projection and also according to the projection model considered. In relation to the first item mentioned above, only Service A presented the rationale for the initial composition of the PMSO.

    Expenses, it is concluded that Service B adopted a much more projection than that of Service A. Highlighting again, this difference in values has no significant impact on valuation if the company's PMSO is within the regulatory level.

      Table 28.  Comparison between the Services   
                R$ Million 
    Company Service A  % Service B  % 
           
    2017  2047  2017  2047   
    Eletroacre  112.5  684.9  509%  133.5  327.6  145% 
    CEAL  288.8  1,126.5  290%  296.7  686.0  131% 
    Amazonas  550.1  1,896.4  245%  600.2  1,267.4  111% 
    Cepisa  383.7  1,757.7  358%  452.6  1,122.8  148% 
    CERON  340.1  1,534.8  351%  373.3  866.6  132% 
    Boa Vista  105.6  334.1  216%  216.3  371.7  72% 
    Total  1,780.8  7,334.4  312%  2,072.5  4,642.1  124% 

     

    - Finally, regarding the documents presented by the consultancies, Service A detailed the methodology much more than Service B, presenting in its text more coherence in the presentation of the information.

    2.6 INVESTMENT

    Both services contracted for economic and financial assessment adopted different assumptions for the projection of the investment notably from the period 2023 to 2047, since both adopted the Five-Year Plan for the period from 2018 to 2022 of the company. This plan was presented in Product 7 “Eletroacre Operational Technical Assessment Report" and was prepared based on the "Expansion Plan for the Electrical System (2016-2025 period)", the "Development Plan for Distribution - PDD for the year 2017", the

    Page 45


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    "Results Plan for the Improvement of Distribution Services, 2015" and the "Temporary Provision Plan for the Electricity Distribution Service, 2017". In addition to this, information obtained through the "Field Visit Report" (Annex I of the Technical and Operational Assessment Report of Eletroacre) and the incremental investment needs projected by the company's planning area, obtained during face-to-face meetings and conference call carried out. The only caveat made for this period is that Service A for the investment in 'Luz Para Todos' program considered only 10% of the value since 90% is subsidized by CDE, and therefore treated as Special Obligations. In addition to this, approximately R$ 65.8 million of the investments in HV expansion in the year 2018 were also considered as Special Obligations, since they result from subrogation.

    In relation to the other five-year period, the main differences are due to the fact that Service A included, in addition to the projections made for implementation and renovation projected from the Five Year Plan, other analyzes to define the investment in maintenance and replacement of partially and totally depreciated assets . Whilst Service B was restricted only to the first analysis.

    For the projection from 2023 of the investments in implantation and renovation, Service A adopted as basic premise for the expansion of AT, AV and LV, as well as for the improvements some metrics1 depending of the type of investment depending on the last years contemplated in the plans of the companies and that value was kept constant until the end of the concession period. Service B, for electric investments (expansion and improvement), estimated the second five-year period of the concession through the mean of the period from 2018 to 2022 and expurgated values outside the range between the average and +/- a standard deviation. Thus, in order to obtain the amount of investment required for the five-year period, the investments of the years considered typical were added up and multiplied by the factor 5/No. typical years. In the case of Eletroacre, since there was information up to 2024 for the expansion of HV, these values were confirmed and the methodology above adopted for the estimate of the last year from 2025 to 2027. For the other five-year period, they adopted as base the previous five-year period with the

    1 The methodology for projecting the investment from 2023 varies between companies for Service A. The expansion in AT had the information until 2024 through the Development Plan for Distribution 2017 (PDD 2017) and, then, the smallets value between 2017 and 2024 was adopted for the other years. As for the investment in AV and LV expansion, the mean of expenses from 2018 to 2022 was used. In turn, the improvements were defined as 50% of the 2019 investment.

    Page 46


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    caveat that for the expansion in AT, it is considered that every 10 years the need for investments doubles in relation to the previous five-year period. As for the non-electric investments (infrastructure), it was assumed that every 5 years it invests 50% of the average of what was forecast in the first five years. Service A, however, did not project them.

    For renovation investments (maintenance), since 2023, service A considered that investments of this type are equivalent to the value of the assets depreciation. To define this value, the depreciation reference of ANEEL's Normative Resolution 674/20152, was used for the current asset base plus the depreciation of the new investments of the company, calculated from the average depreciation of the asset base of the Assets Control Report. These investments were accumulated and concentrated in the two years before each tariff review according to the proportion of the reference companies.3 Service B, however, first calculated the share of the amount destined to the investment in renovation within the period of the Five-Year Plan 2018-2022. Subsequently, it applied this percentage for all other years.

    For investments in replacement, only service A estimated them. Based on the assets base of 2016, the estimated investments for implementation and renovation from 2017 to 2022 were added, as well as the inflation adjustments and application of the depreciation of each item, based on Aneel Normative Resolution 674/2015, both for the assets base of 2016 and for the new investments projected up to 2022. In this way, the company's asset base of 2022 was formed. Therefore, the indicator Value of Market in Use (VMU) / New Value of Replacement (VNR) wars calculated for the company and for the benchmark4 companies so that the distribution company in analysis reaches the same percentage of the private reference distributed between 2023 and 2027. As the new Eletroacre base exceeded the reference value, it was not necessary to estimate investments for asset replacement in 2022.

    Given the assumptions and their differences, we find the following values for the two consultancies for Eletroacre.

    Table 29. Comparison of Services A and B

    2 Resolution that approved the review of the Assets control manual of the Electric Sector - MCPSE.

    3 Cemar, Coelce, Coelba and Celpe.

    4 Cemar, Energisa Sergipe and Energisa Paraíba

    Page 47


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Service A              R$ Million 
      1st Five-Year  2nd Five-Year  3rd Five-Year  4th Five-Year  5th Five-Year  6th Five-Year   
                  Sum for the 
    Investment  Period  Period  Period  Period  Period  Period   
                  Period 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
    AT Extension  217  72  60  60  60  60  529 
    AV/LV Extension  106  106  106  106  106  106  636 
    Improvement  114  65  65  65  65  65  439 
    Renewal (maintenance)  62  262  306  356  412  182  1,580 
    Luz para Todos (Light for Everyone)  7  -  -  -  -  -  7 
    Infrastructure and Support  21  -  -  -  -  -  21 
    Replacement  -  -  -  -  -  -  - 
    Total  527  505  537  587  643  413  3,212 
    Service B              R$ Million 
      1st Five-Year  2nd Five-Year  3rd Five-Year  4th Five-Year  5th Five-Year  6th Five-Year   
                  Sum for the 
    Investment  Period  Period  Period  Period  Period  Period   
                  Period 
      2018-2022  2023-2027  2028-2032  2033-2037  2038-2042  2043-2047   
    AT Extension  287  50  100  50  100  50  636 
    AV/LV Extension  107  79  79  79  79  79  503 
    Improvement  115  93  93  93  93  93  581 
    Renewal (maintenance)  63  24  29  24  29  24  195 
    Luz para Todos (Light for Everyone)  74  -  -  -  -  -  74 
    Infrastructure and Support  21  10  10  10  10  10  71 
    Replacement  -  -  -  -  -  -  - 
    Total  667  256  311  256  311  256  2,059 

     

    Note that the difference between the consortiums is R$ 1.2 billion, notably in item Renovation (maintenance) in which Service A estimates that from the 2nd five-year period the investments are equivalent to the value of the depreciation of the assets and Service B calculates them from their share in the 2018-2022 Five Year Plan. It should be noted that this difference in the volume of projected investments does not have a significant impact on valuation, since the investment increases the Company's Remuneration Base, and consequently, it is reflected in a higher tariff.

    By way of comparison, the value of the investment made from 2012 to 2016 to the values of December 2016 by Eletroacre is as follows:

    Table 30.  Comparison of investment carried out from 2012 to 2016, at values 
    of Dec./2016 Services A and B
                  R$ Million 
    Investment  2012  2013  2014  2015  2016  Five-Year Period 
                  2012-2016 
    Updated Values (Dec./16)  71  79  57  57  65  329 

     

    2.7 REGULATORY REMUNERATION BASE

    Both consortiums used the data contained in the valuation reports of the assets as reference to the date of February 28, 2017 to prepare the projection of the Regulatory Remuneration Base (BRR). In these reports, both closed BRR and the incremental BRR

    Page 48


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    from 3CRTP to 4CRTP were assessed. In the case of Eletroacre, the following reports were used:

    Both service A and B adopted different projection criteria for the asset base unitized through Feb/17 (closed BRR - starting point for 4CRTP) and for unitizations occurring after Feb/17 and new investments (incremental BRR).

    Regarding the starting point for 4CRTP, both consortiums use the full appraisal report (delivered on 07/27/2017), however, they present different criteria for the reversal of Fixed Assets in Use (AIS).

    For service A, the reversal is treated in two ways: i) reversal of the revised closed base (Closed Base of the Full Report - Closed Base of the Incremental Report) in the value set at 5.73% for the companies from the North and 5.01% for companies from the Northeast, based on the average adjustment (reversal) of the group of distribution companies appraised in the 3CRTP by region, i.e., this reversal is made only in the AIS of the difference found in the opening of the closed base compared to that approved by Aneel in 3CRTP adjusted for inflation; and ii) The incremental base between 3CRTP and 4CRTP -the lowest value was applied between the average adjustment of the group of distribution companies appraised in 3CRTP by region and that verified by the company, without evidencing in the Appraisal Report the invested amount.

    Unlike Service A, Service B used only a disallowance methodology to reach the Remuneration Base of Feb/17 of the Distribution Company. The disallowances were performed between the difference in AIS of the two assessment reports of the Closed Base and the Incremental Base. This difference would be the opening of the Closed Base. That is, the disallowance for Service B was performed only at the opening of the Closed Base and occurred in the AIS. The percentage of disallowance was calculated from the relation

    Page 49


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    between the final AIS (with disallowances) and the initial AIS (without disallowances) of the incremental BRR of the distribution companies 3RTP of Eletrobras group, obtaining a mean of 9.2%. If the company has a disallowance lower than this mean, it is considered the mean of disallowance percentage itself (9.2%). Otherwise, the mean between the percentage of the company and the maximum percentage of disallowances (20.1%) is used. In the case of Eletroacre, it falls within the second criterion, since it presented a 9.9% disallowance of the AIS in the 3CRTP, then, adopting a disallowance of 15% for the Closed Base. This disallowance percentage is applied only in the difference between the AIS of the two reports and added to the value of the base of the incremental report. That is, the disallowance of 15% is not applied in the total AIS of the Complete Base, it is calculated only for the increment of the closed base.

    Therefore, the following information was obtained for the Closed base, in the position of February 2017, for both services:

    Table 31.  Closed base, in the position of February 2017 - Services A and B 
     
          R$ Million 
          Armored Base 
      Description     
        Service A  Service B 
    (1) Property, Plant And Equipment in Service (New Replacement Value)  1,060  1,030 
    (2) Index of Full Utilization  1  1 
    (3) Gross Special Obligations  314  314 
    (4) Fully Depreciated Assets  57  57 
    (5) Gross Compensation Reference = (1) - (2) -(3) - (4)  689  659 

     

    Page 50


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    (6) Accumulated Depreciation  357  357 
    (7) Net AIS (Market Value In Use)  703  673 
    (8) Index of Depreciated Utilization  1  1 
    (9) Compensation Reference Value (VBR)  703  672 
    (10) Warehouse in Operation  1  1 
    (11) Deferred Assets  -  - 
    (12) Special Net Obligations  274  274 
    (13) Lands and Easements  2  2 
    (14) Base of Total Net Compensation = (1) - (6) - (8) + (10) + (11) - (12) +     
    (13)  432  402 
    (15) RGR PLPT Balance  -  - 
    (16) RGR Balance Other Investments  40  35 
    (17) Depreciation Rate  4%  3% 
    (18) Share of Regulatory Reintegration = (5) * (17)  26  23 
    (19) Actual WACC before Taxes  11%  ND 
    (20) RGR PLPT Rate  1%  ND 
    (21) RGR Rate of Other Investments  4%  ND 
     
    (22) Capital Compensation (RC) = (15) * (20) + (16) * (21) + [(14) - (15) -  46  ND 
    (16)] * (19)     

     

    For the years subsequent to the 4CRTP Starting Base, both Services apply adjustment for inflations for the IPCA, additions and write-offs of the assets that were considered in the valuation until the end of the concession period.

    However, it should be noted that both Services did not incorporate the positions of Fixed Assets in use into the Remuneration Base, since, according to the services, the assessors’ report of the Remuneration Base does not take into account the specific assessment of the condition and characteristic of these assets concerning (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the reassessment process (such as the calculation of VNR, depreciation, utilization rate and valuation of losses). In the case of Eletroacre, according to data from June 2017, it should be noted that the Fixed Assets in Use (AIC), obtained from balance sheet information, totaled R$ 139.2 million in the distribution and administrative use.

    In addition, Service B segments the assumptions for the projection of BRR transactions into two groups. The first group operated the closed BRR by the IPCA adjustment for inflation and the discount of the depreciation and write-off of the assets that went into operation until February 17 and in a second group operates the incremental BRR (new investments) through the additions of assets projected after Feb/2017, less the losses related to these assets.

    Thus, from 2018, service A monthly corrects, through the IPCA, the items Fixed Assets in Use, Warehousing in operation and land and easements and increases them according to

    Page 51


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    their projection of the investment. The gross special obligations item is also updated by the IPCA and by the increase in projected special obligations. On this incremental Base, the average adjustment of privatized benchmark distribution companies was applied as disallowance by region obtained through the 3CRTP Incremental Base. For the Northeast companies, Celpe, Energisa Paraíba and Energisa Sergipe were considered, giving a value disallowance of 2.09%. For the North companies, the disallowance value was 2.83%, using Celpa and Cemar as benchmark. In relation to the depreciation rate, the same verified in 3CRTP was maintained for each distribution company assessed. In relation to asset write-offs, the ratio of fully depreciated assets to fixed assets in use observed in the 4RTP BRR resulted in an equilibrium rate calculated as a simple mean between Eletrobras' companies. Thus, starting from the company's ratio, it varies at a constant monthly rate until reaching that equilibrium rate of 12.54% at the end of the concession.

    Service B, as mentioned earlier, operated BRR with different premises for the base that served as a starting point and the one that incorporates the new investments. The transaction of the closed BRR was made by monetary restatement by the IPCA and by the discount of depreciation and write-offs. The depreciation rate, unlike service A, is not constant, the rate falls from the average level calculated in Feb/17, as the assets are being written-off. With respect to the asset write-offs, the average write-off/AIS ratio recorded in the period between 3CRTP and Feb./2017 was applied in the AIS of each month. The incremental BRR transaction (new investments) was made based on the projection of the investments rejected in 5%, except for the unitizations from Mar./2017 to May/2017, which were reduced at 15%. The depreciation rate used was obtained by the weighted mean of the investments of the Five-Year Plan 2018 to 2022 per group of assets and their respective depreciation rates, obtaining 3.79% of the AIS. This rate was considered constant throughout the concession. Regarding the write-offs, until Feb/27, they were considered null in view of the reduced probability of assets being written off in the first 10 years. After this period, 1/3 of the BRR Closed/AIS write-off ratio was applied in a staggered manner every 5 year until reaching the value of the integral ratio as of Mar/42.

    Accordingly, given the differences in the investment projections, the criteria for disallowance, depreciation and write-offs, we obtained at the end of the concession (Feb / 48) the following values of Net Remuneration Basis for Eletroacre according to the projections of the two services:

    Table 32. Net Remuneration Base- Services A and B

    Page 52


     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    The table below shows X factor and its components calculated by Services A and B in years of tariff reviews5:

    Table 38.Table 33. X Factor and its components - Services A and B

    Service A         
      X Factor  Pd  T  Q 
    2019  -0.32%  0.00%  0.00%  -0.32% 
    2023  -0.66%  0.98%  -1.56%  -0.08% 
    2028  1.00%  0.98%  0.32%  -0.30% 
    2033  1.35%  0.98%  0.37%  0.00% 
    2038  0.99%  0.99%  0.00%  0.00% 
    2043  0.99%  0.99%  0.00%  0.00% 
    2048*  0.99%  0.99%  0.00%  0.00% 
    Service B         
      X Factor  Pd  T  Q 
    2019  0.38%  0.00%  0.00%  0.38% 
    2023  1.74%  1.42%  0.76%  -0.44% 
    2028  1.24%  1.48%  0.00%  -0.24% 
    2033  1.23%  1.42%  0.03%  -0.22% 
    2038  1.20%  1.40%  0.14%  -0.34% 
    2043  0.95%  1.38%  0.02%  -0.45% 
    2047*  0.30%  1.36%  0.02%  -1.08% 

     

    As can be seen in the table above, the estimated values for the X Factor are quite divergent, especially for the Q and T components, affecting, as a whole, the X-Factor as a whole.

    Table 39.Table 34. X factor - Differences between Services A and B

    Service A - Service B       
      X Factor  Pd  T  Q 
    2019  -0.70%  0.00%  -0.32%  -0.38% 
    2023  -2.40%  -0.44%  -0.84%  -1.12% 
    2028  -0.24%  -0.50%  -0.30%  0.56% 

     

    5 It should be noted that Service B provided the value of the component Q for all thirty years of concession, while Service A presented the values only in the years of tariff reviews.

    Page 54


     



     



     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Component T of the X-Factor is applied to Installment B so that the tariff increases or decreases so that the recognized regulatory operating costs reach the efficiency threshold defined by Aneel.

    In defining efficient operating costs, the distribution company’s costs, the efficient level of costs and the characteristics of the concession area are observed. Aneel uses a model whose objective is to define a feasible production/cost frontier as a way of identifying the level of efficiency of companies.

    In a simplified way, the identification of the efficient level of costs is obtained by comparing the distribution companies by means of a benchmarking method that takes into account the attributes of each utility. From these conditions, a target of efficient operational costs is established to be reached throughout the tariff cycle. At the time of the tariff review, the regulatory goal is compared to the operational costs coverage present in the utility´s tariff, called operating costs revenue. From the difference between the regulatory goal and the operating costs revenue, a regulatory trajectory will be calculated. Part of the difference will be incorporated at the time of the tariff review and the remaining installment will be considered for the purpose of calculating component T of the X Factor.

    Services A and B, following the provisions of Aneel's Technical Note No. 149/2017, used a zero value for component T in the period between the date of signature of the agreement and the first ordinary tariff review, that is, from 2018 to 2022. From 2023, Services A and B found very divergent values for Component T of X Factor.

    2.9 WACC

    Services A and B adopted the method of Capital Asset Pricing Model - CAPM to estimate the return that the shareholders expect to arrange with the capital employed in the business, for discounting the current value of the distribution company’s cash flow. It should be noted that this method is well-established and widely used, either by the market at large, and by Eletrobras.

    In the calculation of the Weighted Average Cost of Capital WACC, Services A and B applied different approaches of the CAPM method, thus allowing quite different results.

    Below are the main differences found in the calculation of WACC by both Services.

    Page 57


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Capital Structure

    Beta

    ·Service A: used comparable domestic companies. Unlevered Beta: 0.68;

    ·Service B: considered comparable domestic companies, with a larger sample. Unlevered Beta: 0.43.

     

    Ke

    ·Service A: applied the CAPM method and a “foreign exchange risk” of 4.17% at nominal Ke in Dollars to achieve a nominal Ke in Brazilian Reais. Therefore, this “foreign exchange risk,” which derives from the “Future DI” rate and “Fraction of exchange rate coupon,” works as a differential of inflation. To achieve such rate in real terms, it was deflated by IPCA (4.23%);

    ·Service B: applied the CAPM method.

     

     

    The table below shows the result for Ke under real terms:

    Table 41.Table 36. Profit or Loss of Real Ke Services A and B

    Actual Ke 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2030 Service A 7.95% 7.81% 7.53% 7.21% 7.02% 6.85% 6.70% 6.55% 6.43% 6.31% 6.31% Service B 11.73% 10.16% 10.18% 10.42% 10.42% 10.42% 10.42% 10.42% 10.84% 10.84% 10.79%

    Kd

    Page 58


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    the amount of debt of each company, achieving an average weighted cost, which was invested in a projection for CDI (source: Bradesco).

    · Service B: calculated Kd by the sum of SELIC and a spread of 0.64%.

    The table below shows the Kd result in real terms after taxes.

      Table 42.Table 37. Kd in real terms after taxes Services A and B   
    Actual                       
    Kd 2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  4.25%  4.09%  4.27%  4.19%  4.19%  4.19%  4.19%  4.19%  3.25%  3.25%  3.25% 
    Service B  4.44%  3,34%  3.36%  3.29%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25%  3.25% 

     

    WACC

    Based on the values of Beta, Ke and Kd above, Services A and B calculated WACC in real terms after taxes, according to the results included in the table below.

    Table 43.Table 38. WACC Services A and B
    WACC  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  2030 
    Service A  8.09%  7.20%  7.30%  7.39%  7.39%  7.39%  7.39%  7.39%  7.60%  7.60%  7.58% 
    Service B  5.29%  4.80%  4.89%  4.87%  4.93%  5.00%  5.06%  5.12%  5.19%  5.26%  5.26% 

     

    Out of the results stated above, it should be noted that the factor with a higher impact on the rate presented by Service A was the use of the foreign exchange risk (4.17%) in a heterogeneous manner to the market practices, as a differential of inflation, as this is applied on a Ke (nominal in USD), in a scenario where the CPI US inflation is 2.30% and IPCA, 4.23%. The result of this is WACC, in real terms, which is higher compared to the number obtained in calculation by Service B.

    Page 59


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    3. THE RESULT OF ECONOMIC AND FINANCIAL ASSESSMENTS AND BY MULTIPLES

    Service A's assessment calculated the enterprise value from R$ 921,169,419.22 to ELEROACRE, considering, as already mentioned, the granting of a new concession. After deducting the debts, liabilities with suppliers and contingencies, Eletroacre´s equity value calculated by Service A is negative R$ 159,107,674.07, as shown below. To obtain these values, the consultancy used the WACC of 7.57%.

    Table 44.Table 39. Enterprise value Service A

     
        Free Cash Flow to Firm (FCFF) 
    Base date of Enterprise Value    Feb./2018  Dec./2016 
    Adjustment on Enterprise Value    -  IPCA 
    Enterprise Value (BRL)  A  969,447,975.51  921,169,419.22 
    Cash (BRL)  B  16,006,000.00  16,006,000.00 
    Gross Debt (BRL)  C  (871,635,025.17)  (871,635,025.17) 
    Net Debt (BRL)  B+C  (855,629,025.17)  (855,629,025.17) 
    Tax, Labor and Social Security Contingencies (BRL)  D  (77,852,000.00)  (77,852,000.00) 
    Legal Contingencies (BRL)  AND  (138,379,219.28)  (138,379,219.28) 
    Environmental Adjustments (BRL)  F  (8,416,848.84)  (8,416,848.84) 
    Valuation (R$)  A+B+C+D+E+F  (110,829,117.78)  (159,107,674.07) 

     

    Source: Assessment Report of Eletrobras Distribuição Acre, Service A.

    Service B calculated the enterprise value of R$ 944,913,418.27 for Eletroacre, in view of the granting of a new concession. After deducting the debts, liabilities with suppliers and contingencies, Eletroacre equity value calculated by Service B is negative R$ 135,363,675.02, as shown below.

    Table 45.Table 40. Enterprise value - Service B 
    Result  Dec./16 
    Operation NPV  826,942,945.56 
    NPV of Terminal Values  117,970,922.70 
    Enterprise Value  944,913,419.27 
    (-) Net Revenue  (855,629,025.17) 
    (-) Contingencies  (216,231,219.28) 
    (-) Environmental Adjustments  (8,416,848.84) 
    Equity Value  (135,363,675.02) 

     

    Page 60


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Source: Product 10: Eletroacre Conclusive Economic and Financial Assessment Report, Service B.

    In order to obtain these values, Service B used the rolling WACC, which is considered a mobile (leveraged) capital structure, according to the table below.

    Table 46.Table 41. Rolling WACC - Mais Energia B Consortium
    Parameters  2017  2018  2019  2020  2021  2022  2023  2024  2025  2026  > 2026 
    Rate Free from Risk (Rf)  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78%  2.78% 
    Country Risk  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53%  2.53% 
    Market Premium  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90%  6.90% 
    Non-Leveraged Beta (²u)  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43  0.43 
    % Debt (Kd) - [D/(D+E)]  54.0%  50.7%  47.3%  44.0%  40.6%  37.3%  33.9%  30.6%  27.3%  23.9%  23.9% 
    % Own Capital (Ke) - [E/(D+E)]  46.0%  49.3%  52.7%  56.0%  59.4%  62.7%  66.1%  69.4%  72.7%  76.1%  76.1% 
    Taxes - IRPJ/CSLL  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0%  34.0% 
    Leveraged Beta (²L)  0.76  0.72  0.68  0.65  0.62  0.59  0.57  0.55  0.53  0.52  0.52 
    Premium Size  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00%  0.00% 
    Inflation - Brazil  4.33%  4.43%  4.36%  4.31%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23%  4.23% 
    Inflation - USA  2.40%  2.30%  2.30%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40%  2.40% 
    Cost of Own Capital  12.62%       12.55%  12.22%  11.83%  11.55%  11.37%  11.21%  11.06%     10.93%  10.81%  10.81% 
    Selic  10.71%  9.07%  9.03%  8.87%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73%  8.73% 
    Spread  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64%  0.64% 
    Cost of Debt  11.35%  9.71%  9.67%  9.51%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37%  9.37% 
    Debt Cost - Net of Taxes  7.49%  6.41%  6.38%  6.28%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18%  6.18% 
    WACC in current CRL (Post-Tax)  9.86%  9.44%  9.46%  9.39%  9.37%  9.44%  9.50%  9.57%  9.64%  9.71%  9.71% 

    Source: Product 10: Eletroacre Conclusive Economic and Financial Assessment Report, Service B.

    With respect to the multi-market valuation, Service A, in its Assessment Report of

    Eletrobras Distribuição Acre, attached to this IT, obtained Eletroacre’s enterprise value equal to R$ 645,646,000, which represents a difference of 30% vis-à-vis the value of the economic and financial assessment.

    Service B, in Product 10: Eletroacre Conclusive Economic and Financial Assessment Report, attached to this IT, did not obtain the enterprise value, but performed comparative analyses of different implicit multiples to the result of the Discounted Cash Flow assessment.

    Page 61


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    3.1 AFAC

    The study presented considers that all Advances for Future Capital Increase - AFAC’s made by Eletrobras in Eletroacre and which are accounted for will be converted into debt or loan. On October 17, 2017, Eletroacre had R$ 82,45 million in AFAC accounted for. The AFAC contracts already provide for the conditions to be considered in case of conversion into debt. Accordingly, Eletrobras must convert the AFACs into debt, according to a loan agreement executed under the following conditions:

    R$ 63.54 Million;

    4.      CONTINGENCIES
    4.1      LEGAL CONTINGENCIES

    Eletroacre´s legal due diligence, performed by Service B and listed in Product 4: Due Diligence Legal Report - Eletroacre, pointed out some critical points, although such points do not impede the process of privatization of the distribution company, the new investor should be aware of them so that it can take the applicable measures.

    Page 62


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    4.2 ACCOUNTING CONTINGENCIES

    ANEEL carried out inspections at the distribution companies of Eletrobras - AmD, Eletroacre, Ceron and Boa Vista Energia, regarding the onlendings of CCC's funds, from June 2009 to June 2016, which aims to cover generation costs in current regulations, namely: Law 12.111/2009, Decree No. 7.246/2010, ANEEL Normative Resolution 427/2011 and Normative Resolution ANEEL No. 597/2013.

    The inspections pointed out that the management of CCC, at the time in charge of Eletrobras, was not applying a series of rules set forth in ANEEL Normative Resolution 427/2011 for the monthly calculation of the financial criteria that involve the total cost of generation of the beneficiaries.

    In the case of Eletroacre, the preliminary decisions of ANEEL are unfavorable to the distribution company, since the Agency, by means of Technical Note No. 90/2017-SFF-SFG-SRG/ANEEL, dated 07/29/2017, understands that Eletrobras, as manager of CCC at the time, has to reimburse CCC in the amount of R$ 276.55 million, for June/2017, connected to the largest transfer to Eletroacre. In addition to this, considering that in Eletroacre’s Balance Sheet, 06/30/2017, records a receivable from CCC of R$ 296.17 million and that the position of the Agency is maintained, there is an accounting risk of approximately R$ 572.7 million.

    This issue is still under discussion between the interested parties and, therefore, ANEEL's decisions unfavorable to Eletroacre can be reversed. However, the potential reversal of an unfavorable decision should not occur in a timely manner from the completion of this document or even within the timeline of privatization planned for the distribution company. If Eletroacre does not succeed in administrative appeals with ANEEL, it will file legal appeals to try to succeed in this matter.

    This issue was pointed out as being of great risk to the eventual buyer of Eletroacre, and may even render the auction unfeasible, since the amount under discussion may cause an impact of R$ 572.7 million. In this sense, the proposed model for privatization seeks to neutralize potential effects on the valuation due to uncertainty in the amount effectively recognized of the accounting difference of CCC credits.

    The modeling proposal foresees that Eletroacre will transfer to the Eletrobras registered CCC credit in the amount of R$ 296.17 million and debts (which Eletroacre has with

    Page 63


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Petrobras) in the same amount, which are already backed by this credit. Accordingly, the accounting effect of such a transfer in Eletroacre would be zero, and the credit risk is transferred from Eletroacre to Eletrobras.

    Although Eletroacre has recognized this amount as a claim in its balance sheet, in the event that ANEEL maintains its position on the matter, even though Eletroacre has filed an appeal, it is probable, based on accounting rules, that the current claim becomes a provision. That being so, there is an operation that would originally be neutral for Eletrobras to become a deficit operation in large sums. In this case, Eletrobras, in addition to assuming a debt of R$ 296.17 million with Petrobras, would record in its result a provision of R$ 276.55 billion with CCC, thus reaching a possible negative impact of R$ 572.7 billion on its balance sheet.

    Table 47.Table 42. Accounting impact risk

        R$ thousand 
    Value  Balance Value  Possible 
    ANEEL  (Jun/2017)  Accounting 
        Impact 
    -276,550.90  296,167.00  -572,717.90 

     

    Notwithstanding the foregoing, it is worth mentioning, according to the legal opinion prepared by the law firm Décio Freire and Associados, dated September 18, 2017, referring to Administrative Proceeding no. 48500.004972/2016-34 of AmD, which, owing to its similarity, is being duplicated for Eletroacre, while there is no decision of the ANEEL’s

    Board wholly or partially ratifying the conclusions made, event in which there is no administrative act that implies an obligation to reimburse the fund. Additionally, the opinion considers that there is a solid legal basis for the review of ANEEL’s rulings and ultimately states that it is too early to provide for any potential loss connected to said Administrative Proceeding or to disclose the matter in questions in explanatory notes to the Financial Statements, as reproduced below:

    "Under all of the circumstances described above, we believe there is a solid legal basis for reviewing the provisions under the Technical Notes numbers 52/2017-SFF-SFG-SRG and 141/2017-SFF-SFG-SRG, as well as a manifest failure to comply with decisions rendered by Federal Regional Court of the

    Page 64


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    1st Region, in the records of Writ of Mandamus no. 0026107-81.2012.4.01.3400.

    This way, we understand that it is too early to provide for any potential loss connected to the Administrative Proceeding in question, or to disclose the matter in explanatory notes to the Financial Statements.”

    5. CAPITAL STRUCTURE

    The structure developed is aimed at enabling the sale of the Eletrobrasdistribution companies in the business model associated with the granting of a new concession. The following is the current Eletroacre corporate structure.

    Table 48.Table 43. Capital structure of Eletroacre on 12/31/2016 
    Total Shares      108,271,416,523 
    Capital  R$    475,788,771.91 
    Par Value of Shares  R$    0.00439441 
     
    Interest    Shares  (%) 
    Eletrobras    104,706,713,291  96,71% 
    Minority    3,564,703,232  3.29% 
     
    Classes of Shares    Shares  (%) 
    Common    78,582,126,234  72.58% 
    Preferred    29,698,290,289  27.42 % 
     
    Source: Product 11: Eletroacre Privatization Modeling Proposal, Service B. 

     

    The economic value of the company's shares was determined as of 12/31/2016. Then, the definition of values was performed, considering the mean values between Services A and B, as shown in the table below.

    Table 49.Table 44. Valuation of ELETROCRE Mean of Services A and B 
    Service A  R$  (159,107,674.07) 
    Service B  R$  (135,363,675.02) 
    Average of Services A and B  R$  (147,235,674.55) 
     
    Source: Product 11: Eletroacre Privatization Modeling Proposal, Service B.   

     

    Page 65


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    As mentioned in item 6.3 of the IT, adjustments in the equity value of Eletroacre were necessary. The adjustments made were intended to contemplate the update of the balance sheet up to June 2017, which totaled R$ 33,505,802.55. Thus, the new equity value is negative R$ 113,729,781.99.

    Table 50.Table 45. Equity Value Final

    Average of Services A and B  R$  (147,235,674.55) 
    Consolidated Adjustments  R$  33,505,802.55 
    Equity Value Adjusted  R$  (113,729,871.99) 

     

    Source: Product 11: Eletroacre Privatization Modeling Proposal, Service B.

    At the end of the privatization process, Eletroacre should present the capital structure described below, considering the total adhesion of employees and retirees to the stock offering:

    Table 51.Table 46. Eletroacre's capital structure after sale
    Total Shares      710,018,610,603,462 
    Capital    R$  854,128,324.96 
    Par Value of Shares    R$    0.00000120 
     
    Interest  Common Shares  Preferred Shares  Total Shares  (%) 
    Investor  469,037,986,807,138  171,859,251,826,389   640,897,238,633,527  90.2649% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retirees  50,584,239,078,573  18,533,568,188,129  69,117,807,266,702  9.7346% 
    Other Minority  1,795,985,579  1,768,717,653  3,564,703,232  0.0005% 
    Shares per Class      Total Shares  (%) 
    Common      519,624,021,871,291  73.1846% 
    Preferred      190,394,588,732,171  26.8154 % 
     
    Source: Product 11: Eletroacre Privatization Modeling Proposal, Service B.     

     

    5.1      CAPITALIZATION BY ELETROBRAS AND BY THE INVESTOR
    5.1.1      ELETROBRAS

    Considering the set amount of R$ 50,000.00 for Eletroacre, with the granting of a new concession, the conversion of debt into equity and the assumption of Eletroacre's debt by Eletrobras in exchange for actions in Stage 1 of the modeling of sales. Accordingly, there is a need to capitalize Eletrobras in the amount of R$ 113,779,871.99, as detailed below.

    Page 66


     
      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    If Eletrobras chooses to increase its stake in up to 30% in Stage 2, as detailed in item 6.5 of IT, it will have to capitalize the distribution company in the amount of R$ 102.3 million, which may occur through capital contribution or conversion debt.

    5.1.2 INVESTOR

    As provided in Stage 2 of the modeling, the investor should make a capital contribution with subscription of new shares in the company. Therefore, the estimated contribution for the investor is R$ 238,805,729.30.

    6. NECESSARY ADJUSTMENTS AT ELETROBRAS

    According to the proposed model to reach the minimum value of R $ 50,000.00, Eletrobras must capitalize R$ 113.8 million in Eletroacre, in addition to assuming a further R$ 284.25 million, which today are backed by a CCC's claim of the same amount, which will also be transferred to Eletrobras, as explained in item 4.2.

    Thus, Eletrobras should prioritize the cheapest debts to be capitalized/assumed, since other debts will be settled by the new Controller. The table below shows Eletroacre's debt amount by type of creditor and its associated costs:

    Table 52.Table 47. Debt amount per creditor and its costs
    RO + AFAC World
    Bank 
    RGR Obligation
    CCC: 
    Related Parties Petrobras Total
    Average
    Cost 
    12.00% 6.59% 5.00% 8.75% (Selic) 8.75%
    (Selic) 
     
     
     
    Eletroacre  357.89  49.70  7.36  81.61  -  461.80  958.36 

     

    As can be seen from the table above, Eletroacre's cheapest debts, which will be assumed by Eletrobras to be fully or partially capitalized, are as follows: i) RGR, which has a cost of 5% per year; ii) World Bank, with a cost of 6.59%; iii) Petrobras, at SELIC cost (8.75%); and RO, with a cost of 12.59%. It is important to note that ANEEL must authorize Eletrobras to assume RGR's debt without having to remove all of it at the time of the transfer.

    Page 67


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Debts with the World Bank refer to the raising of funds from this bank for the implementation of the "Projeto Energia +". The RGR debts refer to financing of investments obtained before the designation period as temporary public service provider and RGR resources received during the temporary provision. The Ordinary Resources are the transfers of Eletrobras' own funds to the distribution companies through loan, in order to cover operating deficits. Related parts debts were contracted with other distribution companies or Eletrobras companies, except the Holding Company. Among the main parties related to outstanding credits are, for example, Eletronorte, Furnas and Chesf.

    The table below shows the detailing of the capitalization by Eletrobras into Eletroacre, according to the source of funds, to achieve the set price of R$ 50,000.00. It should be underlined that, in this scenario, the preferred option was to convert the RO debts and before the World Bank in capital, although the common resources have a higher cost (12.00%), as the debts with Petrobras are posted as bonds under the CCD’s executed.

    Table 53.Table 48. Conversion of debts in capital, per distribution company R$million

    Debt  Capitalization 
       
    RO + AFAC  64.08 
    IBRD  49.70 
    Total  113.78 

     

    If the Shareholders’ Meeting of Eletrobras decides to assume the debt claims of CCC and

    CDE recorded in the balance sheets of the distribution companiess, under the terms of Paragraph 1, Article 3 of CCPI Resolution 20/2017, as amended by CPPI Resolution no. 28/2017, the table below sets forth the amounts of debts, by source of funds, to be capitalized to achieve the set price in Eletroacre, as well as the assumption of debts in the same amount of CCC and CDE debt claims. In this context, the debts with Petrobras are prioritized, since most of these debts are already backed up by receivables of CCC, substantiated in Contracts of Admission of Debt, wherein Eletrobras stands as a guarantor.

    Page 68


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Table 57.Table 49. Conversion of debts in capital and Assumption of Debt Claims of CCC and CDE, per distribution company R$ million

    Debt  Capitalization  Assumption  Total 
    RO + AFAC  56.72  -  56.72 
    IBRD  49.70  -  49.70 
    RGR  -  7.36  7.36 
    CCC Reimbursement  -  81.61  81.61 
    Petrobras  7.36  207.20  214.56 
    Total  113.78  296.17  409.95 

     

    It should be noted that the amount corresponding to the assumption of debts of R$ 296.17 million, has as counterparty the CCC's claims recorded in the Eletroacre's Financial Statements, which, as explained in item 4.2, are under discussion with ANEEL, with the risk of not realizing and becoming a debt of R$ 276.5 million.

    The table below summarizes the impact on Eletrobras in the event that the company succeeds in its administrative / judicial litigation and in case it cannot reverse the position of ANEEL. It is important to note that the result of the administrative / judicial funds may be in intermediate amounts to those presented in the table below.

    Table 58.Table 50. Impact Scenarios at Eletrobras - R$ million

    SCENARIOS Assumption of Debts  Assumption of Debt
    Claims of CCC 
    Impact on Eletrobras
    Current position  409.95  296.17  -113.80 
    ANEEL Inspection  409.95  -276.55  -686.50 

     

    6.26.1 FIXED ASSETS IN USE

    As referred to in Item 2.7 of this Appendix, it should be pointed out that the studies performed by Services A and B did not incorporate the positions of Fixed Assets in Use (AIC) into the Remuneration Base since, according to the Services, the assessors' report about the Remuneration Base does not take into account the specific assessment of the

    Page 69


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    condition and characteristic of these assets concerning: (i) the eligibility to provide the regulated electricity distribution service, (ii) the physical accounting reconciliation; (iii) its sources of financing and (iv) other critical parameters of the reassessment process. The table below shows the Fixed Assets in use, ascertained based on data of interim accounting balance sheets for each distribution company. The consideration of these assets, amounting to R$ 139.2 million at Jun./2017, would improve the results of the economic and financial assessments performed by the Services, and consequently, reduce the amount to be contributed by Eletrobras in the distribution company to reach the minimum amount of R$ 50,000.00, as defined in CPPI Resolution 20/2017.

    7. SETTLEMENT COSTS

    In the 165th EGM, held in July 2016, Eletrobras' shareholders decided, by majority vote, to approve: (i) the non-extension of the concessions of its distribution companies; (ii) the transfer of the ownership control of the distribution companies until December 31, 2017, pursuant to Paragraph 1-A of article 8, of Law 12.783/2013; and (iii) if there is no transfer of control under the terms of the previous item, the adoption of provisions for settlement thereof.

    Thus, the modeling study carried out by Service B calculated the cost with the settlement of Eletroacre, in the case of non-occurrence of item (ii) above or not accepted by Eletrobras of the adjustments made to reach the equity value equal to zero.

    For these cases, Service B developed a comparative analysis between the distribution company's sales value associated with the grant of a new concession and the eventual settlement value of the company. Thus, it was possible to evaluate if the sale of the said state-owned company would be more attractive than its liquidation.

    Since the sale value considers adjustments to the equity value (equity value equal to zero), the adjustments were also considered in the settlement value estimate. However, there were exceptions to the Tax and Negative Adjustments, since these do not impact on the settlement value, and for the AFACs, since these can be paid in the event that the distribution companies are not sold.

    Page 70


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    As shown below, in the case of Eletroacre, the best economic option is to sell the company, considering the value of the assets and the concession. The balance between selling and liquidating is favorable for the first option, given the following table:

    Table 59.Table 51. COMPARISON BETWEEN THE SALE AND SETTLEMENT OF

    ELETROACRE
    Final Equity Value  R$  (113,729,871.99) 
    Settlement Amount (Prior to Adjustments)  R$  (514,632,455.55) 
    (+) Indemnities  R$  419,227,871.57 
    (-) Net Debt  R$  (855,629,025.17) 
    (-) Contingencies  R$  (216,231,219.28) 
    (+/-) Payments/Receivables  R$  137,999,917.33 
    Settlementn Amount (After Adjustments)  R$  (402,348,970.72) 
    =Value of settlement (After Adjustments)  R$  (514,632,455.55) 
    (+/-) Balance Adjustments  R$  5,150,000.00 
    (+) RGR PPST Reincorporation  R$  107,133,484.83 
    Final Equity Value Settlement Amount (After Adjustments)  R$   288,619,098.73 

    Source: Product 11: Eletroacre Privatization Modeling Proposal, Service B.

    The results presented above corroborate that the sale of the distribution company, associated to the granting of a new concession, is the option that is less onerous to Eletrobras. If Eletrobras had to simply liquidate the distribution company under the scope of Eletrobras, the latter would have to pay a loss of R$ 402,348,970.72, an amount greater than R$ 288,619,098.73 when compared to the final equity value of negative R$ 113,729,871.99.

    In relation to the CCC credits in discussion with ANEEL, the effects shown in item 6 impact the company in both cases, of sale and liquidation. Thus, it has no direct impact on the comparison of values, so that the economic decision remains unchanged in the event of unsuccessful lawsuits.

    In addition, in the event of settlement of the distribution company, Eletrobras would not be entitled to the benefits arising from the relaxation of regulatory parameters, which generated a reduction in costs of R$ 150.0 million for Eletroacre.

    Page 71


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

     

    8.      ANALYSIS OF THE BENEFITS AND RISKS OF DISPOSAL
    8.1      BENEFITS
     

    (i) The sale of Eletroacre's shares will result in the adoption, by Eletroacre, of a legal regime similar to that of other non-state agents operating in the sector, allowing for
    significant operating gains;

     

     
     
     

    (ii) Improvement of Eletrobras' Ebitda and elimination of the need for disbursement of funds by Eletrobras to finance the cash deficit and investments of Eletroacre, contributing to the reduction strategy of the net debt / ebitda indicator;

     

     
     
     

    (iii) It prevents the need for capitalization by shareholders to meet the financial sustainability targets imposed by ANEEL for the extension of the concession, and for investments in the expansion and improvement of service quality;

     

     
     
     

    (iv) The disposal, along with the grant of concession allows the use of relaxation of the regulatory parameters approved in the period from 2018 to 2023, when the tariff
    review of the new concession contract will occur; and

     

     
     
     

    (v) This flexibility has led to an improvement in the equity value of the distribution company of around R$ 150.0 million.

     

     
    8.2      RISKS
    (i) Frustration of the disposal of the shares or non-performance of the auction, due to several factors, notably due to possible judicial, political, economic-financial, regulatory, governmental or administrative impediments;
    (ii)      Absence of investors and / or companies interested in purchasing;
    (iii)      Reduced competitiveness in the bidding procedure due to the current economic situation of the country; and

     

    (iv)      Amendment of the minimum conditions for sale by the TCU.

    Page 72


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    9. FINAL REMARKS

    Lastly, it should be noted that the economic and financial assessment methodology was found to be widely known and adopted by the market, which is the Free Cash Flow to Firm - FCFF, in nominal terms. In addition, the calculated discount rates were based on the CAPM - Capital Asset Pricing Model method, which estimates the return that the shareholders expect to obtain from the company, given this is a well-established method widely used either by the market at large or by Eletrobras.

    Additionally, some points need to be highlighted. Among them, there is the need for approval of the minimum amount and the adjustments necessary for achieving such amount by the Executive Board and Board of Directors of Eletrobras, in addition to the need to submit the item to the general shareholders' meeting.

    With regard to the need for adjustments, the study establishes the need for capitalization by Eletrobras of the debts in exchange of shares at Eletroacre in the value of R$ 113.78 million, in order for the company’s value to achieve the minimum value of R$ 50,000.00.

    Additionally, Eletrobras shall assume R$ 296.17 million in debts under the debt claim corresponding entry, with CCC/CDE, at the same amount.

    It should be noted that the studies do not consider the fixed assets in use, amounting to R$ 139.2 million, referenced back to values of Jun./2017. If, in the modeling, AIC’s were considered unitized, even if set aside, they would be able to reduce the value to be disbursed in the distribution company by Eletrobras.

    Furthermore, it should be noted that in the proposed sales model there is the possibility that Eletrobras will remain with one share owned by Eletroacre, with the purpose of increasing its shareholding in the distribution company by up to 30%. The increase in interest, if it is the option adopted by Eletrobras, must be carried out within six months after the date of the auction.

    It should also be reiterated that the modeling entertained the possibility that Eletroacre has values of CCC previously stated as amounts receivable in the balance sheet, which were currently reverted into obligations payable, owing to technical decisions of ANEEL. The difference of understanding between Eletroacre and ANEEL achieves amounts as large as R$ 572.1 million. Therefore, in view of the relevance of the value discussed for the valuation of Eletroacre and the unpredictability of the deadline for a final decision on this

    Page 73


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    matter, it was recommended by Service B that the responsibility for such debt claims -including possible benefits and burdens - is transferred from Eletroacre for Eletrobras’ Holding. Thus, the effect of the current questioning would be neutral on the assessment of the market value of the distribution company and, therefore, on its sales modeling. However, this recommendation requires the consent of the regulators and should be accompanied by considerations of the company's external auditors. Furthermore, the uncertainties existing within the scope of Eletroacre should also be taken into consideration, as this operating will make them be incurred by Eletrobras. Therefore, this point should be considered by the managers of the Holding.

    It is worth noting that Service B performed a comparative analysis between the option of divesting the shareholder control of the distribution company with the granting of a new concession and the simple liquidation option of the distribution conmpany, as deliberated by the shareholders at the 165th EGM of Eletrobras. With this analysis, one may conclude that the liquidation of the distribution company, under the scope of Eletrobras, as established under the modeling proposed by BNDES, is the most onerous option to Eletrobras, given that, for such purpose, it would have to bear an additional cost of R$ 288.7 million connected to the sale scenario.

    Lastly, it should be noted that in the case of the simple distribution company’s liquidation, Eletrobras would not benefit from the relaxation of regulatory parameters, which contributed to an improvement in the value of Eletroacre’s equity value at R$ 150 million.

    Page 74


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Appendix Rectification

    To Joint Technical Information DFP/DFF/DDE 001/2017

    AMAZONAS ENERGIA Analysis and Results (AmE D)

    BOA VISTA ENERGIA Analysis and Results

    CEAL Analysis and Results

    CEPISA Analysis and Results

    CERON Analysis and Results

    ELETROACRE Analysis and Results

    Página 1


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table of Contents   
    1.  OBJECTIVE  3 
    2.  Amazonas Energia APPENDIX RECTIFICATION  3 
    3.  Boa Vista Energia APPENDIX RECTIFICATION  4 
    4.  Ceal APPENDIX RECTIFICATION  5 
    5.  Cepisa APPENDIX RECTIFICATION  6 
    6.  Ceron APPENDIX RECTIFICATION  7 
    7.  Eletroacre APPENDIX RECFTIFICATION  8 

     

    Página 2


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    1. OBJECTIVE

    This document aims to rectify the Appendixes of AmE D, Boa Vista, Ceal, Cepisa, Ceron and Eletroacre, dated 11/30/2017, due to adjustments in the number of shares and capital stock of the distributors in the modeling studies of distributors forwarded by BNDES on 12/18/2017.

    2. AMAZONAS ENERGIA APPENDIX RECTIFICATION

    The tables "Table 43. AmE D's corporate structure as of 12/31/2016" and "Table 46. AmE D's corporate structure after sale" are replaced, in the Appendix Analysis and Results of Amazonas Energia (AmE D), by the following tables.

    Table 43. AmE D Corporate Structure on 12/31/2016

    Total of Shares      6,276,666,628 
    Capital Stock  R$    4,610,171,103.19 
    Shares Nominal Values  R$    0.73449354 
    Participation    Shares  (%) 
    Eletrobras    6,276,666,628  100.00% 
    Minorities    0  - % 
     
    Type of Shares    Shares  (%) 
    Common    6,276,666,628  100.00% 
    Preferred    0  - % 

    Source: Product 11: Proposed Modeling of Privatization of AmE D, Service B, Dec/2017.

     

      Table 46. AmE D Corporate Structure after sale   
     
     
    Total of Shares        201,043,849,012,596 
    Capital Stock      R$  14,068,005,253.69 
    Shares Nominal Value      R$  0.00006997 
     
    Participation  Common Shares  Preferred Shares  Shares  (%) 
    Investor  180,939,447,700,231  0  180,939,447,700,231  90.0000% 
    Eletrobras  1  0  1  0.0000% 
    Empluyees and Retired  20,104,401,312,364  0  20,104,401,312,364  10.0000% 
    Others Minorities  0  0  0  - % 
     
    Type of Shares      Total of Shares  (%) 
    Common      201,043,849,012,596  100.00% 
    Preferred      0  - % 

    Source: Product 11: Proposed Modeling of Privatization of AmE D, Service B, Dec/2017.

    Página 3


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    3. BOA VISTA ENERGIA APPENDIX RECTIFICATION

    The tables "Table 48. Boa Vista Corporate Structure on 12/31/2016" and "Table 51. Boa Vista corporate structure after sale" are replaced, in the Appendix Analysis and Results of Boa Vista Energia, by the following tables.

    Table 48. Boa Vista Energia Corporate Structure on 12/31/2016

    Total of Shares      320,742,889 
    Capital Stock  R$    684,204,245.96 
    Shares Nominal Value  R$    2.13318602 
    Participation    Shares  (%) 
    Eletrobras    320,742,889  100.00% 
    Minorities    0  - % 
    Type of Shares    Shares  (%) 
    Common    320,742,889  100.00% 
    Preferred    0  - % 

    Source: Product 11: Proposed Modeling of Privatization of Boa Vista Energia, Service B, Dec/2017.

    Table 51. Boa Vista Energia Corporate Structure after sale

    Total of Shares        1,882,197,882,831 
    Capital Stock      R$  1,221,879,401.83 
    Shares Nominal Value      R$  0.00064918 
    Participation  Common Shares  Preferred Shares  Shares  (%) 
    Investor  1,693,977,931,021  0  1,693,977,931,021  90.0000% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retired  188,219,951,809  0  188,219,951,809  10.0000% 
    Otherm Minorities  0  0  0  - % 
    Type of Shares      Total of Shares  (%) 
    Common      1,882,197,882,831  100.00% 
    Preferred      0  - % 

     

    Source: Product 11: Proposed Modeling of Privatization of Boa Vista Energia, Service B, Dec/2017.

    Página 4


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    4. CEAL APPENDIX RECTIFICATION

    The tables "Table 42. CEAL Corporate Structure on 12/31/2016", "Table 46. CEAL corporate structure after sale Scenario 1 and "Table 48. CEAL corporate structure after sale – Scenario 2” are replaced, in the Appendix Analysis and Results of CEAL, by the following tables.

    Table 42. CEAL Corporate Structure on 12/31/2016

    Total of Shares      689,524,384 
    Capital Stock  R$    734,753,835.00 
    Shares Nominal Value  R$    1.06559514 
    Participation    Shares  (%) 
    Eletrobras    689,524,384  100.00% 
    Minorities    0  - % 
    Type of Shares    Shares  (%) 
    Common    677,858,321  98.31% 
    Preferred    11,666,063  1.69% 

    Source: Product 11: Proposed Modeling of Privatization of CEAL, Service B, Dec/2017.

    Table 46. CEAL Corporate Structure after sale Scenario 1

    Total of Shares        8,370,010,415,171 
    Capital Stock      R$  1,341,653,392.10 
    Shares Nominal Value      R$  0.00016029 
    Participation  Common Shares  Preferred Shares  Shares  (%) 
    Investor  7,402,266,527,625  125,306,976,572  7,527,573,504,197  89.9351% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retired  826,131,680,439  16,305,230,534  842,436,910,973  10.0649% 
    Others Minorities  0  0  0  - % 
    Type of Shares      Total of Shares  (%) 
    Common      8,228,398,208,065  98.3081% 
    Preferred      141,612,207,106  1.6919% 

    Source: Product 11: Proposed Modeling of Privatization of CEAL, Service B, Dec/2017.

    Página 5


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017

    Table 48. CEAL Corporate Structure after sale Scenario 2

            8,770,108,413,078 
    Stock Capital      R$  1,915,352,157.18 
    Shares Nominal Value      R$  0.00021840 
    Participation  Common Shares  Preferred Shares  Shares  (%) 
    Investor  7,757,612,096,239  132,904,781,176  7,890,516,877,415  89.9706% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retired  864,114,851,751  15,476,683,911  879,591,535,662  10.0294% 
    Others Minorities  0  0  0  - % 
    Type of Shares      Total of Shares  (%) 
    Common      8,621,726,947,991  98.3081% 
    Preferred      148,381,465,087  1.6919% 

    Source: Product 11: Proposed Modeling of Privatization of CEAL, Service B, Dec/2017.

    5. CEPISA APPENDIX RECTIFICATION

    The tables "Table 42. Cepisa Corporate Structure on 12/31/2016" and "Table 46. Cepisa corporate structure after saleare replaced, in the Appendix Analysis and Results of Cepisa, by the following tables.

    Table 42. Cepisa Corporate Structure on 12/31/2016

    Total of Shares      779,223,552 
    Capital Stock  R$    1,272,747,054.06 
    Shares Nominal Value  R$    1.63335291 
    Participation    Shares  (%) 
    Eletrobras    779,223,552  100.00% 
    Minorities    0  - % 
    Type of Shares    Shares  (%) 
    Common    744,131,334  95.50% 
    Preferred    35,092,218  4.50% 

    Source: Product 11: Proposed Modeling of Privatization of Cepisa, Service B, Dec/2017.

    Página 6


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    Table 46. Cepisa Corporate Structure after sale
    total of Shares        12,492,680,402,577 
    Capital Stock      R$  2,074,326,370.52 
    Shares Nominal Value      R$  0.00016604 
    Participation  Common Shares  Preferred Shares  Shares  (%) 
    Investor  10,731,062,154,927  505,168,319,634  11,236,230,474,561  89.9425% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retired  1,199,012,217,021  57,437,710,994  1,256,449,928,015  10.0575% 
    Others Minorities  0  0  0  - % 
    Type of Shares      Total of Shares  (%) 
    Common      11,930,074,371,949  95.4965% 
    Preferred      562,606,030,628  4.5035% 

    Source: Product 11: Proposed Modeling of Privatization of Cepisa, Service B, Dec/2017.

    6. CERON APPENDIX RECTIFICATION

    The tables "Table 43. Ceron Corporate Structure on 12/31/2016" and "Table 46. Ceron corporate structure after saleare replaced, in the Appendix Analysis and Results of Ceron, by the following tables.

    Table 43. Ceron Corporate Structure on 12/31/2016

    total of Shares      1,212,156,653 
    Stock Capital  R$    1,325,368,848.76 
    Shares Nominal Value  R$    1.09339733 
    Participation    Shares  (%) 
    Eletrobras    1,212,156,653  100.00% 
    Minorities    0  - 
    ype of Shares    Ações  (%) 
    Common    1,212,156,653  100.00% 
    Preferred    0  - 

    Source: Product 11: Proposed Modeling of Privatization of Ceron, Service B, Dec/2017.

    Página 7


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
    Table 46. Ceron Corporate Structure after sale   
    Total of Shares        15,672,967,962,122 
    Capital Stock      R$  3,465,780,082.76 
    Shares Nominal Value      R$  0.00022113 
    Participation  Common Shares  Preferred Shares  Shares  (%) 
    Investor  14,105,668,965,596  0  14,105,668,965,596  90.0000% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retired  1,567,298,996,525  0  1,567,298,996,525  10.0000% 
    Others Minorities  0  0  0  - 
    Type of Shares      Total of Shares  (%) 
    Common      15,672,967,962,122  100.0000% 
    Preferred      0  0.0000% 

    Source: Product 11: Proposed Modeling of Privatization of Ceron, Service B, Dec/2017.

    7. ELETROACRE APPENDIX RECFTIFICATION

    The tables "Table 43. Eletroacre Corporate Structure on 12/31/2016" and "Table 46. Eletroacre corporate structure after saleare replaced, in the Appendix Analysis and Results of Eletroacre, by the following tables.

    Table 43. Eletroacre Corporate Structure on 12/31/2016

    Total of Shares      108,271,416,523 
    Capital Stock  R$    475,788,771.36 
    Shares Nominal Value  R$    0.00439441 
    Participation    Shares  (%) 
    Eletrobras    104,706,713,291  96.71% 
    Minorities    3,564,703,232  3.29% 
    Type of Shares    Shares  (%) 
    Common    78,582,126,234  72.58% 
    Preferred    29,689,290,289  27.42% 

    Source: Product 11: Proposed Modeling of Privatization of Eletroacre, Service B, Dec/2017.

    Página 8


     

      Date: 30/11/2017

    IT DFP/DFF/DDE
    001/2017
     
    Table 46. Eletroacre Corporate Structure after sale   
    Total of Shares        710,018,544,395,161 
    Stock Capital      R$  854,128,299.67 
    Shares Nominal Value      R$  0.00000120 
    Participation  Common Shares  Preferred Shares  Shares  (%) 
    Investor  469,039,554,055,764  171,857,684,732,396  640,897,238,788,160  90.2649% 
    Eletrobras  1  0  1  0.0000% 
    Employees and Retired  50,582,669,496,470  18,535,071,407,298  69,117,740,903,768  9.7346% 
    Others Minorities  1,795,985,579  1,768,717,653  3,564,703,232  0.0005% 
    Type of Shares      Total of Shares  (%) 
    Common      519,624,019,537,814  73.1846% 
    Preferred      190,394,524,857,347  26.8154% 

    Source: Product 11: Proposed Modeling of Privatization of Eletroacre, Service B, Dec/2017.

    Página 9


     

      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

     

     

    Addendum   
     
    DFP/DFF/DDE 001-B/2017 Joint Technical Information 
     
     
    Title  Analysis of the sale of shareholding from Eletrobras in the distributors Amazonas Distribuidora de Energia (AmE D), Boa Vista Energia, Ceal, Cepisa, Ceron and Eletroacre. 
     
     
     
    Subject
     
    Analysis of external legal opinions and technical report outsourced for grounding legal, economic, and financial aspects concerning the
     process of privatization and the possible dissolution and liquidation of Eletrobras’ distributors. 
     
     
     
     
    Authors
     
    Felipe Baptista da Silva DFP;
    Marcos Barreto de Faria Pinho DFPI;
    Flávia Xavier Cirilo de Sá DFPI;
    Pedro Paulo da Cunha DFF;
    Frederico Pinto Eccard DDEF;
    Luciana Pereira de Souza DDEF. 
     
     
     
     
     
     
    Company
     
    Eletrobras, Amazonas Distribuidora de Energia (AmE D), Boa Vista Energia, Ceal, Cepisa, Ceron, Eletroacre 
     
     
    Category
     

    Addendum to the Technical Information. 

     
    Keywords
    Amazonas Energia (AmE D), Boa Vista Energia, Ceal, Cepisa, Ceron, Eletroacre, Sale, Transfer, Liquidation. 
     
     
     
     
      Page 1 

     


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    Contents   
    1.  Disclaimer  3 
    2.  Purpose  3 
    3.  External Opinions  4 
    3.1  Legal Aspects RELATED to the privatization and liquidation of the DISTRIBUTION 
    COMPANIES  4 
    3.2  Legal Aspects RELATED to THE commutativity  6 
    3.3  Legal statement of the lawyers of the consortium/BNDES  8 
    4.  Economic and Financial Aspects  10 
    5.  Conclusion  12 

     

    Page 2

    Date: December 26, 2017 
     
    IT DFP/DFF/DDE
    001-B/2017

     


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    1. DISCLAIMER

    This Addendum IT DFP/DFF/DDE 001-B is an integral part of IT DFP/DFF/DDE 001/2017, dated November 30, 2017, and should be read jointly with it.

    This Addendum supersedes Addendum IT DFP/DFF/DDE 001-A/2017, dated December 21, 2017.

    2. PURPOSE

    This Addendum is targeted at supplementing the DFP/DFF/DDE 001/2017 Technical Information, dated November 30, 2017, especially with respect to the conclusion and the cost of liquidation of the distributors, as well as to replace the Addendum IT DFP/DFF/DDE 001-A/2017, dated December 21, 2017, incorporating the analysis made by the external consultants outsourced by Eletrobras to support the company’s senior management decision-making, especially with respect to the legal aspects connected to the privatization and/or liquidation of the distributors - Amazonas Energia Distribuidora S.A. Amazonas Distribuidora de Energia (AmE D), Boa Vista Energia S.A Boa Vista Energia, Companhia Energética de Alagoas S.A.Ceal, Companhia Energética do Piauí S.A. Cepisa, Centrais Elétricas de Rondônia S.A. Ceron, Companhia de Eletricidade do Acre S.A. Eletroacre, in light of the modeling approved by the Board of the Investment Partnership Program CPPI, by means of Resolution no. 20, dated November 8, 2017, amended by Resolution no. 28, dated November 22, 2017, which approve the assessments made by BNDES.

    Aiming this purpose, the following lawyers were engaged: (i) Arnold Wald, full professor of Civil Law of the Law School of the State University of Rio de Janeiro; (ii) Gustavo Binenbojm, full professor of the Law School of the State University of Rio de Janeiro; and (iii) Nelson Eizirik, LLM by PUC-RJ, founding member of the law firm Carvalhosa e Eizirik Advogados.

    The provision of the opinions by Arnold Wald, Gustavo Binenbojm and Nelson Eizirik took place after the conclusion of IT DFP/DFF/DDE 001/2017, dated November 30, 2017, which led to the need for an Addendum.

    Moreover, a request was made by the lawyers of the Consortium engaged by BNDES -Loeser and Portela Advogados - LPA about the modeling of the sale of the distributors

    Page 3


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    proposed by the Bank and approved by CPPI. The said positioning was forwarded by means of the Letter AJ/JUADE no. 20/2017, dated December 21, 2017, whose annex is the Technical Note of said firm.

    Additionally, an opinion was received from lawyer Anderson Schreiber, Professor of Civil Law at the State University of Rio de Janeiro, on the commutativity and contractual balance of the privatization of the six distributors of Eletrobras, under the terms of the modeling approved by CPPI, especially given the provisions under Article 245 of Law 6,404/1976 (LSA).

    It should be noted that the positionings of Loeser and Portela Advogados - LPA and Anderson Schreiber were not still provided to Eletrobras on the conclusion of the IT Addendum DFP/DFF/DDE 001-A/2017, dated December 21, 2017, which is the main reason for their review.

    Lastly, the internal cost of liquidation of the distributors was analyzed by means of Technical Information DFP/DFC/DDEF 001/2017, dated December 21, 2017, with Eletrobras as liquidator, i.e., if not applied to Law 8029/1990, from the economic and financial viewpoint, based on the Financial Statements of June 30, 2017 of the distributors, with the purpose of performing a test of adherence to the numbers presented by BNDES in its assessments.

    3. EXTERNAL OPINIONS

    3.1 LEGAL ASPECTS RELATED TO THE PRIVATIZATION AND LIQUIDATION OF THE DISTRIBUTION COMPANIES

    The external opinions outsourced (by Arnold Wald, Gustavo Binenbojm and Nelson Eizirik), who mainly reviewed the legal opinions which addressed: (i) the privatization of the distributors in light of the conditions approved by CPPI; (ii) the interpretation of the provisions under Law 6,404, dated December 15, 1976, which provides for corporations; and (iii) the interpretation of Law 8029, dated April 12, 1990, and Law 9,491, dated September 9, 1997, with respect to the hypothesis of dissolution of the distributors.

    Page 4


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    Initially, it should be noted that the three lawyers, in their respective opinions, conclude that the CPPI approved modeling is unfeasible, given that it imposes the assumption of a billion-figure indebtedness by Eletrobras, amounting to BRL 11.24 billion, without the appropriate offset, thus impacting the company’s finances as a whole.

    In regards to the latter issue, the Eletrobras’ managers draw attention, alongside the executive board, that the principles of diligence and loyalty should be met, as well as the standards of liability established by the applicable laws and Securities Commission CVM, given that they are personally liable for the efforts that lead to an actual loss to the company, and, hence, shall be subject to any penalties, administrative and civil sanctions, as the laws so establish.

    In this context, the three opinions also conclude that the liquidation is the only judicially feasible alternative, which should be carried out under the terms of Articles 21 and 23 of Law 8029/1990. In this case, the Federal Government shall be responsible for the liquidation of the distributors as successor “in its rights and obligations arising out of the legal standard, administrative act or contract, as well as in the other pecuniary obligations.” On account of this, Eletrobras shall not remain liable for the debts of the distributors it secures, and the indebtedness of Eletrobras, incurred as a result of the purchase of the distributors from the States, before the Global Reversion Reserve Fund RGR, shall be cancelled.

    Alternatively, the opinions call attention to the fact that the recommendation for liquidation by the Eletrobras’ senior management is supported by the decision made at the

    165th EGM and the laws, thus not indicating legal impacts on its managers, given that the applicable legal rules are met, as well as the respective fiduciary duties.

    Furthermore, they assert that, upon expiration of the distributors’ concessions, the next logical step is to liquidate them, with the full deprivation of their corporate purpose, pursuant to Article 206 of Law 6,404/1976, given that they were established by law to provide public utility. This scenario becomes more evident with the possibility that the Federal Government holds a pure bidding for the concessions, as provided for under Article 5 of Decree 9,192/2017.

    Furthermore, the opinions also remember that the purchase of direct controlling interest in the distributors has not happened owing to a corporate deliberation of Eletrobras’

    Page 5


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    shareholder, but under a mandatory ruling of the Federal Government, insofar as Eletrobras is a mere means or instrument to accomplish the public policy of federalization, in order to subsequently privatize them, over a term of 90 days.

    Lastly, attention should be drawn to the concept of commutativity, also addressed in the opinions at hand, as established under Article 245 of LSA. It points out that, in the transactions between related parties, a balance should be in place relating to the performances of each of them, without favoring to holding company or subsidiary.

    In this vein, managers have no power, in such a way that compromises the company, to favor an affiliated company, holding company or subsidiary, and its obligation is to care for the concept of commutativity established in LSA, namely: the balance between performances and counterperformances of each party.

    This way, the non-existence of commutativity can be evidenced in the following cases: (i) the contracting took place in non-abidance by the standards generally adopted in the market in similar business; (ii) the business is not performed in such a way that serves the best interests of the company; or (iii) the transaction would not have been completed had the parties been independent.

    3.2 LEGAL ASPECTS RELATED TO THE COMMUTATIVITY

    For the purposes of clarification about the legal concept of commutativity and its application, in light of the assessments prepared by BNDES and the sales modeling approved by CPPI, the opinion of the lawyer Anderson Schreiber was outsources.

    Said opinion initially draws attention to the fact that Article 245 of LSA should be read jointly with Article 154 of the same Law. Therefore, he underscores that the analysis of commutativity as balance may not be performed on an isolated basis, yet it should cover the entire complexity of the object (content) of the business transaction:

    “The object of contract should be considered not only its main performances. The person interpreting it should be heedful not only of the sum of the main and accessory performances, but the entire set of rights, faculties, obligations, burden and other legal situations deriving from the contract rights and obligations in their broad sense , with the purpose of spotting is real consequence on the contracting parties. The principle at

    Page 6


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    hand is interested not only in the legal situations per se, but else their economic consequences, i.e., the economic sacrifices supported and the economic benefits earned by the parties on account of the rights and obligations, always broadly speaking, in which they are part of the contract.”8

    He also stresses that the notion of commutativity is not only concerned with the mere possibility of estimating the value of the performance and counterperformance by the time of the execution of the contract (commutativity as a notion contrary to randomness), but else its relatedness to a relationship of proportionality between economic benefits and economic sacrifices deriving from the contract.

    However, in light of the modeling approved, the opinion underlines that the value of the debt to be assumed by Eletrobras and the value of the counterperformance seem, however, precisely set, so that, by the time of the engagement, the involved parties shall have immediate awareness of the value of its reciprocal performances. Thus, the contract shall technically be commutative, as there shall be no subsequent ascertainment of the values involved, and all of them are already previously set, on the occasion of the execution of the contract.

    The opinion still affirms that Article 245 of LSA does not preclude the execution of the non-commutative contracts, but else impedes the performance of managers in such a way that bring loss to the company. With respect to the distributors’ sales modeling, the opinion states that it was motivated by the intention to impede the worsening of the accrued loss by Eletrobras, so that the term “company’s loss” should be analyzed with due regard to the economic outcomes of the different alternatives available for Eletrobras (sale x liquidation).

    Accordingly, based on the information and documents provided and the sales modeling approved by CPPI, the opinion concludes that the commutativity provided for under LSA is also found in the sales modeling of CPPI, since the amounts of debt assumption and sales of the distributors are known prior to the contract execution, as well as there will be no loss to the company, since, among the available alternatives, according to BNDES’ assessments, the distributors’ sale is less onerous than the liquidation thereof.

    Page 7


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    Lastly, considering the alternatives to remain in the current situation of large losses and to liquidate the distributors, which, based on BNDES’ assessments, prove to be economically more onerous for the company than the sale of such distributors, the opinion concludes that, from the technical viewpoint, there are not such prerequisites necessary for the accountability of the managers, based on Article 245 of LSA.

    It should be noted that the aforementioned opinion did not evaluate the applicability of Law 8,029/1990, and therefore the comparison between the alternative of sale and liquidation was based on the figures presented by BNDES, considering Eletrobras as liquidator.

    3.3 LEGAL STATEMENT OF THE LAWYERS OF THE CONSORTIUM/BNDES

    Eletrobras, by means of its Board of Directors, requested that BNDES stated its opinion with respect to the lawfulness of the distributors’ modeling proposed by the Bank, and approved by CPPI. The aforementioned positioning was sent by means of Letter AJ/JUADE no. 20/2017 dated December 21, 2017, along with the Technical Note of the law firm Loeser and Portela Advogados - LPA, representative of the Consortium Mais B Energia, responsible for said modeling.

    The Technical Note of LPA was intended to provide brief legal considerations on the main points raised in the 3 (three) opinions issued by lawyers Nelson Eizirik, Arnold Wald and Gustavo Binenbojm, and Opinion PGFN/CAS/NO. 2052/2017, all relating to aspects of the privatization process of the Eletrobras’ distributors.

    Initially, the Technical Note states that the modeling proposal was designed to mitigate risk, to make feasible and optimize the privatization process and to evaluate synergies among distributors, in order to generate value to the current shareholder without prejudice to consumers and public power.

    Concerning the conversion and the assumption of debts foreseen in the modeling, the Technical Note states that there is a corresponding entry for Eletrobras, namely, the receipt of new shares of the distributors, which stands as an asset with potential valuation, and thus the corporate relationships are preserved, concluding that it is farfetched that penalties be imposed on Eletrobras’ managers owing to the conversion and assumption of

    Page 8


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    the debts referred to, and Article 245 of LSA would be thus fully met in keeping with the modeling prepared by Consortium Mais B Energia, and approved by CPPI.

    Related to the decision to liquidate or sell the distributors, the Technical Note states that the modeling assessments, considering the non-application of Article 23 of Law 8,029/1990, were performed with the purpose of checking which alternative was more advantageous for Eletrobras and the stakeholders at large, finding that the sale of the distributors was the best alternative for Eletrobras.

    Thus, the Technical Note concludes that the decision to approve the modeling, on the bases proposed, has legal and economic grounds, and includes the commutativity that is covered under Article 245 of LSA.

    Concerning the application of Law 8,029/1990, the Technical Note states that the provisions of the aforementioned Law do not apply to the context of privatization, since there was no deliberation of CPPI in order to approve the dissolution of the distributors as a modality of privatization. Therefore, in the event of liquidation of the distributors, Eletrobras shall be the sole and exclusive responsible for their obligations, incurring costs higher than those in the approved sales modeling.

    Lastly, it is important for Eletrobras to decide to continue the privatization of some distributors, as well as for the liquidation of others.

    As mentioned above, the Technical Note of LPA addresses the assumption of BRL 11.24 billion of the debt of the distributors by Eletrobras as commutative, owing to the fact that Eletrobras, by capitalizing BRL 11.24 billion of debt, would be increasing its shareholding base, and booking an asset with potential appreciation. However, taking into consideration the modeling proposed by BNDES and approved by CPPI Resolution no. 20/2017 and no. 28/2017, this fact did not happen, insofar as, while assuming a BRL 11.24 billion debt, despite increasing the shareholding base, the company shall be sold at a set price for Eletrobras, amounting to BRL 50 thousand, with no possibility of gains higher than this amount.

    The aforementioned valuation potential is only possible if Eletrobras exercises the option of diluting the new distributors' controller by up to 30%, and for that purpose, it will have to capitalize another R $ 1 billion approximately.

    Page 9


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    4. ECONOMIC AND FINANCIAL ASPECTS

    According to IT DFP/DFF/DDE 001/2017, dated November 30, 2017, and to the modeling assessments performed by BNDES, and admitting the withdrawal from the scenario of application of Law 8029/1990, a scenario which was defended by the lawyers outsourced by Eletrobras - Arnold Wald, Gustavo Binenbojm and Nelson Eizirik - the sale of the distributors would be the least onerous option for Eletrobras, considering the legal statements on the subject of commutativity, defended by the law firms of Loeser and Portela Advogados - LPA and that of Anderson Schreiber, as can be observed in the following table, based on the liquidation cost established in the BNDES assessments.

    Distributor  Liquidation Cost (BRL)  Sale Cost (BRL) 
    Amazonas Energia  12,060,245,607.63  8,911,866,558.94 
    Boa Vista Energia  465,696,747.78  342,120,486.20 
    Ceal  569,913,103.09  50,000.00 
    Cepisa  740,112,396.38  50,000.00 
    Ceron  2,398,260,429.38  1,872,522,463.43 
    Eletroacre  402,348,970.72  113,779,872.00 
    Total  16,636,577,254.98  11,240,389,380.57 

     

    In addition to the assessments performed BNDES, which already indicated that, in the case of non-application of Law 8,029/1990, the liquidation of the distributors would be more onerous to Eletrobras, an internal analysis of the adherence of the valued indicated by BNDES, by means of the Technical Information DFP/DFC/DDEF 001/2017, dated December 21, 2017, considering calculation of the cost of liquidation of the six distributors based on the deliberation made at the 165th AGE dated July 22, 2016. Thus, the Technical Information shows the cost of the liquidation from an economic and financial and accounting viewpoint, that is, not involving legal and regulatory aspects, starting from the Financial Statements of June 30, 2017 of the distributors.

    The Technical Information points out that, strictly from the economic and financial and accounting viewpoint, the cost of liquidation of all six distributors, with the liquidation happening within Eletrobras, would amount to about BRL 17.73 billion. It should be noted that, in the case of AmE D, in order to make the liquidation value comparable to the

    Page 10


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    selling value indicated by BNDES, AmE D was already unbundled from AmGT, that is, no value was attributed to AmGT as can be seen in the following table. This procedure was also adopted by BNDES to determine the valuation of AmE D.

    It should be noted that, in internal assessments, which still need to be confirmed by an independent consulting firm, which are under the engagement phase, the preliminary value of AmGT is approximately BRL 3.2 billion.


    Thus, it should be noted that the values obtained by the adherence test conducted by Eletrobras are in line with the liquidation values presented by BNDES, which is the most onerous option for Eletrobras, in the case of non-application of Law 8029/1990.

    It should also be noted that the test, represented by the table above, may contain inaccuracies and/or distortions, not necessarily representing the actual cost of liquidation of the distributors, which can only be determined in a formal process based on defined rules, which are currently unknown and, for this reason, should be taken as an indicative reference for the BNDES’ assessment.

    Page 11


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    5. CONCLUSION

    Given the legal reasoning exposed in the three external legal opinions engaged by Eletrobras (Arnold Wald, Gustavo Binenbojm and Nelson Eizirik), which indicate that the modeling proposed by BNDES and approved by means of Resolution no. 20/2017, as amended by Resolution no. 28/2017 is, under the aspects directly impacting Eletrobras, legally questionable, and impacts the finances of Eletrobras, if the latter may assume the distributors’ debts, without consideration or commutativity, which should be abided by according to the provisions under Article 245 of LSA.

    The assumption mentioned above, as foreseen in the said modeling of sale proposed by BNDES, shall amount to BRL 11.24 billion, which sums the cases of Amazonas Energia Distribuidora, Boa Vista Energia, Centrais Elétricas de Rondônia and of Companhia de Eletricidade do Acre, and given that the selling value of all 4 distributors shall be BRL 200 thousand. For the cases of Companhia Energética de Alagoas (within the scenario of the Bresser Plan’s agreement) and Companhia Energética do Piauí, the modeling indicates the capitalization of value exactly equivalent to the sales value to both companies, i.e., BRL 100 thousand as a whole or R$ 50 thousand for each of them, thus preserving the proper commutativity.

    Therefore, if Law 8,029/1990 is applied, the most legally and financially appropriate alternative, appropriate and in keeping with the Company’s and its shareholders’ interests, especially with respect to its perpetuity, is the dissolution and liquidation of the distributors Amazonas Energia Distribuidora, Boa Vista Energia, Centrais Elétricas de Rondônia and Companhia de Eletricidade do Acre, under the terms of said law, given that, in this hypothesis, the Federal Government shall be responsible for the procedure of liquidation, and the universal successor of the assets, rights and obligations of the distributors, and, in the cases of Companhia Energética de Alagoas (under the scenario of the Bresser Plan’s agreement) and Companhia Energética do Piauí, concerning the sale according to modeling proposed by CPPI.

    It should be also underlined that, in the scenario of the liquidation of the distributors, as provided for under Law 8,029/1990, Eletrobras shall not remain liable for the debts of the distributors it secures by such occasion, and the indebtedness of Eletrobras with the Global Reversion Reserve Fund RGR, arising for the purchase of such companies, shall be cancelled, in accordance with said legal provisions. Still, in the latter case, the aggregate

    Page 12


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    owners’ equity of all four distributors, amounting to negative BRL 12.8 billion, as of June

    30, 2017, would be reverted to the Balance Sheet of Eletrobras, thus recovering loss undergone by the Company’s shareholders with the time, through the full reversal of allowance for deficiency.

    However, in the event that the thesis of the application of Law 8,029/1990, as concluded in the Technical Note of the law firm Loeser and Portela Advogados, the liquidation cost, as shown in the BNDES modeling, confirmed by an internal assessment of Eletrobras, is higher than the cost of sale.

    Taking into account the decision of the Eletrobras Shareholders at the 165th EGM held in 2016, which deliberated for selling them until December 31, 2017 or to liquidate the distributors, only these two alternatives remain for Eletrobras - to sell or liquidate.

    In this point, attention should be drawn to the conclusion of the Opinion of Mr. Anderson Schreiber:

    "... Article 245 of the Corporations Act does not have the purpose of preventing the execution of non-commutative contracts, rather, it is targeted at preventing managers from acting in such a way to bring

    “loss to the company,” and instead favoring affiliated, subsidiaries or

    Holding companies. Hence, the conclusion of a non-commutative contract that does not cause any harm to the company (a pure donation, for instance, with the company being a mere beneficiary) is a hypothesis that does not bring on the levy of Article 245 of the Corporations Act, and for certain, the managers' responsibility may not even be raised in this situation. The same is true if there is a non-commutative contract that imposes loss onto the company, but a loss which is smaller than that represented by the other available alternatives.” (emphasis added)

    Therefore, analyzing the liquidation costs, in the case of non-applicability of Law 8029/1990, and those required to make the sale feasible, despite the latter representing an assumption of BRL 11.24 billion in debt, and imposing burden on Eletrobras, the latter option is less onerous than the liquidation, which, according to BNDES, amounts to BRL

    Page 13


     
      Date: December 26, 2017
    IT DFP/DFF/DDE
    001-B/2017 

    16.64 billion. It should the underlined that the thesis of non-application of Law 8,029/1990 has been dispelled.

    Accordingly, in view of the foregoing, if it is legally concluded that Law 8,029/1990 has not been applied, the senior management of Eletrobras is recommended to sell the controlling interest of the six distributors, according to the modeling approved by CPPI. taking into consideration the fact that the assessments prepared by BNDES and the adherence tests conducted by the Eletrobras’ staff, the cost to be incurred by the Company with the liquidation of the distributors would be higher than the value of assumption and conversion of debts, in the case of their sale.

     

     

    Felipe Baptista da Silva DFP    Pedro Paulo da Cunha - DFF 
       
      
    Marcos B. F. Pinho DFPI  Frederico Pinto Eccard - DDEF 
       
      
    Flávia Xavier Cirilo de Sá DFPI      Luciana Pereira de Souza - DDEF  
       
       

    Agreed: 

     

     
    Armando Casado de Araujo    Luiz Henrique Hamann 
    Chief Financial Officer and Investor  Distribution Officer 
    Relations Officer   

     

    Page 14


     

     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017

     
    DFP/DFC/DDEF 002/2017 Joint Technical Information 
     
     
    Title  Analysis of the cost of liquidation of distributors Amazonas distribuidora de Energia (AmE D), Boa Vista Energia, Ceal, Cepisa, eron and Eletroacre. 
     
     

    Subject
     
    Analysis of the cost of liquidation of the distributors of Eletrobras based on the deliberation made at 165th AGE to sell or liquidate tese companies until 12.31.2017. 
     
     

    Authors
     
    Felipe Baptista da Silva DFP;
    Marcos Barreto de Faria Pinho DFPI;
    Rodrigo Vilella Ruiz - DFC;
    Marcos José Lopes - DFCC;
    Frederico Pinto Eccard DDEF;
    Luciana Pereira de Souza DDEF. 
     
     
     
     
     
    Company  Eletrobras, Amazonas Distribuidora de Energia (AmE D), Boa Vista Energia, Ceal, Cepisa, Ceron, Eletroacre 
     

    Category
     
    Technical Information. 

    Keywords

    Amazonas Energia (AmE D), Boa Vista Energia, Ceal, Cepisa, Ceron, Eletroacre, Liquidation. 

     

    Page 1


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017

     
    Contents   
    1.  Purpose  3 
    2.  ANALYSIS OF FINANCIAL STATEMENTS  4 
    2.1  ASSETS  4 
    2.2  LIABILITIES  6 
    3.  Reclassification of the Balance Sheet  8 
    4.  ADJUSTMENTS OF THE LIQUIDATION COST  16 
    5.  Conclusion  18 

     

    Page 2


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    1. PURPOSE

    This DFP/DFC/DDEF 002/2017 Technical Information replaces the DFP/DFC/DDEF 001/2017 Technical Information and is intended to support the analysis of the cost of liquidation of distributors Amazonas Energia Distribuidora S.A. Amazonas Distribuidora de Energia (AmE D), Boa Vista Energia S.A Boa Vista Energia, Companhia Energética de Alagoas S.A. Ceal, Companhia Energética do Piauí S.A. Cepisa, Centrais Elétricas de Rondônia S.A. Ceron, Companhia de Eletricidade do Acre S.A. Eletroacre based on the deliberation made on the 165th EGM, dated July 22, 2016.

    On that occasion, Eletrobras' shareholders decided to approve that, should the transfer of control not occur until the date of December 31, 2017, all necessary steps should be taken to liquidate these electricity distribution utility companies, under the terms of item 11 of the minutes of the 165th EGM, as transcribed below:

    11. Approve, by majority vote, that the concession of distributors Companhia Energética de Piauí - CEPISA; Companhia Energética de Alagoas - CEAL; Companhia de Eletricidade do Acre - ELETROACRE; Centrais Elétricas de Rondônia S.A - CERON; Boa Vista Energia S.A; and Amazonas Distribuidora de Energia S.A be returned at any time and that the provisions of its liquidation are adopted, in the following events: (i) the transfer of controlling interest referred to in Item 9 above is not performed until December 31, 2017;"

    This Technical Information (IT) will show the liquidation cost from an economic and financial and accounting viewpoint, i.e., it does not involve judicial and legal aspects. The purpose shall be to compare the number of liquidation costs of each distributor presented by BNDES with the amount estimated by Eletrobras, based on the assumptions that shall be described in this paper. IT is divided in five topics, including purpose and conclusion. Firstly, the rationale used to break down the asset accounts into five groups and liabilities accounts into six groups shall be explained. The following topic will address the reclassification of the Balance Sheets of June 30, 2017 of Amazonas Energia, Boa Vista Energia, Ceal, Cepisa, Ceron, and Eletroacre according to the premise adopted for the breaking down of accounts in the large groups in the section above. In the topic on liquidation cost adjustments, the rationale and arrangement will be presented to allow that the distributor’s assets and liabilities be closer to a liquidation scenario.

    Page 3


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    All figures in the tables below are in thousands of Brazilian Reais, according to the standard used in the financial statements of the companies.

    2. ANALYSIS OF FINANCIAL STATEMENTS

    This work was developed based on the Financial Statements of June 30, 2017 for each one of the six Eletrobras’ distributors. The choice of this base date is targeted at making this analysis comparable in time with the option of modeling of the same of shares of the said distributors which BNDES had as responsible for the performance of the work, based on the economic and financial assessments prepared by the independent consultants1.

    All Asset accounts were grouped into five groups in order to associate accounts with similar degrees of realization. Accordingly, the assets were divided in: i) indemnifiable; ii) immediate liquidity; (iii) to realize; (iv) Reimbursement of CCC; and (v) non-realizable. In turn, the liabilities were broken down in six groups: (i) immediate payability; (ii) compensable; (iii) Eletrobras System; (iv) Petrobras System; (v) other liabilities; and (vi) shareholders’ equity.

    2.1 ASSETS

    · INDEMNITIABLE ASSETS

    For the indemnifiable group, accounts were included in which ANEEL/FEDERAL GOVERNMENT are expected not to reimburse the CNPJ of the distributor or its holding company with its liquidation. This assessment included all regulatory assets, such as CVA and financial components, and ANEEL recognizing that the distributor would be entitled to be reimbursed in the next tariff adjustment. As the company will be liquidated, the costs incurred and already paid by the distributor, for instance, the purchase of sale at a price higher than that established in the prior tariff adjustment, the company would have the right of reimbursement ensured.

    1 The economic and financial assessments performed by the independent consultants adopted the criteria of nominal Free Cash Flow to Firm - FCFF, having the December 31, 2016, as the base date. The discount rate used was WACC. Subsequently, adjustments were made to the valuation of the distributors considering the last financial statement available at the time of completion of the assessment, dated June 30, 2017.

    Page 4


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    In the same logic is the right of reimbursement - CDE, which includes the low-income social tariff and the tariff discount subsidies, both with values homologated by ANEEL, and right of reimbursement by the distributor. Another item classified as indemnifiable is the short-term electricity, an amount recognized by CCEE, but still not paid to companies owing to the delinquency with the Chamber.

    In this indemnifiable group, the item with the greatest impact is the financial asset - public utility concession, which includes the distribution assets that make up the distributor’s regulatory remuneration base (BRR) that have not yet been amortized/depreciated. Within this account in the Balance Sheet is the financial asset in use, comparable to the BRR of the company, and the financial asset in course, which would be connected to the investments made that have not yet been completed or that have already been completed and still unitized by the company’s accounting. This financial asset in course shall be part of BRR in the future. The distribution financial asset, because it is essential to the operation, is added to the concession at the time of the bidding process for a new utility company. Therefore, these non-depreciated assets shall be reimbursed by the Federal Government to the CNPJ of the liquidated company.

    · IMMEDIATE LIQUIDITY

    Within immediate liquidity, there is the cash and cash equivalents account, in addition to the securities account. Funds that companies can quickly turn into cash for payment of liabilities.

    · TO REALIZE

    Assets to realize are items according to which the receipt of funds is neither immediate nor indemnifiable, i.e., they are direct for the companies, although their realization is uncertain and with values that may be negotiated. This group includes: (i) customers, (ii) services in course, (iii) warehousing, (iv) investments, (v) fixed assets, (vi) securities and pegged deposits, and (vii) other assets entered under the Balance Sheet. Concerning item (vi), the securities and pegged deposits are assets of difficult liquidation, but can be used to write off values of liabilities for provisions for civil, labor and tax risks. Lastly, item (i), the customers’ account, is the most representative of the group “To Realize” under the assets. This item includes all pending invoices of consumers, which have not yet been

    Page 5


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    received by the distributor, already discounting the allowance for doubtful accounts (PCLD).

    · CCC REIMBURSEMENT

    In this group, the account of the Right of Reimbursement of CCC is included, where all the values booked by the companies to be received from the CDE/CCC Industry-Related Fund, as consideration for expenditure with fuel, machine rental and taxes on the generation of energy in isolated systems, given that the amounts thereof were not covered by the distributors’ final consumer tariff. This item is the most controversial asset of Amazonas Energia, Boa Vista, Ceron and Eletroacre, as it is the object of ANEEL’s inspection, still in progress, whose preliminary values calculated by the agency are quite different from those booked by these companies. Because it is a relevant and controversial item, it will be better addressed later on this IT.

    · NON-REALIZABLE

    Non-realizable items of assets in the event of liquidation of Eletrobras’ distributors, the accounts of (i) taxes and social contributions, and (ii) intangible assets, were considered. Item (i), owing to any tax credit being deposited with the liquidation of the company, may not be used or passed on to another company. Concerning item (ii), the intangible asset also ends with the completion of the company’s operations and their liquidation.

    2.2 LIABILITIES

    · IMMEDIATE PAYABILITY

    The liability group of immediate payability consisted of social and labor obligations and post-employment benefits. According to the criteria used, these two items would have preference for payments in a liquidation.

    · COMPENSABLE

    The compensatory liabilities include the regulatory liabilities accounts, the obligations of reimbursement and the industry-related charges. They are classified as compensable because they are subject to a decrease in the amount of reimbursement by ANEEL/FEDERAL GOVERNMENT to be received by the distributors at the time of their liquidation. The regulatory liability has the same explanation of the regulatory asset, with

    Page 6


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    the difference being a future disbursement of the company, instead of a right. This account includes the CVA and the negative financial components, that is, the distributor invoiced its consumers in this tariff cycle in excess of what had been set forth in the last adjustment, and should be compensated in the next adjustment. As a subsequent adjustment shall not happen, the company shall pay or discount the value to the received.

    The industry-related charges booked under liabilities refer to amounts owed by the companies to the industry-related fund CDE, and that, since it is an industry-related debt, it should also be subject to offset, for the purpose of decreasing the value to be received by the distributors. In turn, concerning the reimbursement obligations, the values received in excess are chiefly connected to the CDE/CCC industry-related fund, as an advance payment, which should be offset in the future.

    · ELETROBRAS SYSTEM

    The accounts included in this group cover all the liabilities with companies of Eletrobras System, and not only with the holding. They include loans of ordinary funds (RO) agreed with Eletrobras, former RGR loans and financing, prior to the appointment period (August 2016), loans with Eletrobras owing to the Project Energia Mais and the Program Luz para Todos (Light for Everyone). Additionally to the debts with Eletrobras, there are liabilities for the purchase of energy with Eletronorte, CHESF and the other distributors of Eletrobras System.

    · PETROBRAS SYSTEM

    For the four distributors of Eletrobras, namely Amazonas Energia, Ceron, Boa Vista and Eletroacre, this group has the heaviest weight in the liabilities of these companies. It refers to the fuel debt contracted with Petrobras, BR Distribuidora and Cigás to serve the isolated systems. Part of this debt was already negotiated with the execution of contracts of admission of debt (CCD) in December 2014, backed by the receipt of the CCC industry-related fund. As part of the scope of ANEEL’s inspection work, the guarantees of the CCC fund are under review and some companies (Amazonas Energia and Eletroacre) have stopped receiving the monthly installments of the CCD agreed between the distributor and the industry-related fund itself.

    Page 7


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    · OTHER LIABILITIES

    Other liabilities include further liabilities with suppliers, excluding the Eletrobras System and Petrobras System, taxes and social contributions, provision for civil, labor and tax risks, and other liabilities.

    · NON-CURRENT LIABILITIES

    Similarly to a part of the assets being classified as non-realizable, a part of the assets was also considered as shareholders’ equity. The items onerous contract and loans of the RGR during the period of appointment and provision of services were withdrawn from the liabilities enforceable from the companies at the time of liquidation. The onerous contract is based on the impairment test annually performed by the companies, and if the test proves negative, the intangible asset is written off and, if the necessary intangible asset is not enough, an onerous contract is still entered under the liabilities account. This would be an accounting record of a negative future result of the company. With its liquidation, this onerous contract is expired and the item is removed for the company’s liabilities.

    Concerning the RGR loans during the appointment period, it should be pointed out that, even though they are booked under liabilities, ANEEL and MME, in their ordinances and resolutions, have already defined that these loans would be under concession and would be paid by means of tariffs to the final consumers between the 6th and the 30th year of the new concession. Therefore, it is understood that, with the liquidation of these distributors, this RGR loan shall be incurred by the new utility company, which shall have the right to enter the amount equivalent to the future tariff in its payment.

    Concerning Amazonas, in addition to the accounts above, the account “Provision for deficiency of subsidiaries,” which corresponds to the negative shareholders’ equity of

    Amazonas GT AmE GT, was also withdrawn from liabilities. As the assessment considers that there shall be unbundling of Amazonas Energia, such deficiency of AmE GT should be disregarded.

    3. RECLASSIFICATION OF THE BALANCE SHEET

    In this section, the assets and liabilities per company with the proposed new reclassification will be presented. Assets divided in five groups: i) indemnifiable; ii) immediate liquidity; iii) to realize; (iv) Reimbursement of CCC; and (v) non-realizable. In

    Page 8


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    turn, the liabilities were broken down in six groups: (i) immediate payability; (ii) compensable; (iii) Eletrobras System; (iv) Petrobras System; (v) other liabilities; and (vi) shareholders’ equity.

    · BALANCE SHEET OF ELETROACRE

    On 06/30/2017    On 06/30/2017     
    Asset      1,060,054  Liability    1,467,425 
    Recoverable      551,906  Immediate enforceability    9,368 
    Regulatory Asset    60,641  Social and labor obligations    8,912 
    Financial Asset public service  430,497  Post-employment benefits    456 
    concessions             
    Reimbursement  right  CDE  60,768       
            Offsettable    130,744 
    Short-Term Electrical Energy         
            Regulatory Liability    32,302 
    Cash to current liabilities    18,437  Reimbursement obligations    81,609 
    Cash and equivalents Securities  18,437  Sectoral  charges  16,833 
            Eletrobrás System    402,263 
    Unpaid      175,147  Advances for future capital increase    69,462 
    Customers CP      85,456  Eletrobrás RGR    7,361 
    Customers LP      44,679  Eletrobrás RO    325,440 
    Warehouse      2,313  Eletronorte     
    Services in progress    10,473       
    Other Assets      12,723  Petrobrás System    443,643 
    Assets to disposal    86  BR Distribuidora    443,643 
    Guarantees and linked deposits  10,552       
            Other Liabilities    374,274 
    Investments             
    Fixed      8,865  Suppliers    55,288 
            Taxes and social contributions    124,161 
    CCC Reimbursement    284,250  Provision for civil, labor and tax risks    146,491 
    Reimbursement right CCC CP  43,607  Other Liabilities    22,939 
    Reimbursement right CCC LP  240,643  R&D    25,395 
    Non-current      30,314  Net equity    107,133 
    Taxes and social contributions  10,020  Onerous concession     
    Intangible      20,294  RGR Service provision    107,133 
    Total Asset      1,060,054  Total Liability    1,467,425 
    Non-current      30,314  Net equity    107,133 
    Total asset for liquidation    1,029,740  Total liabilities for liquidation    1,360,292 
    Recoverable      551,906  Immediate enforceability    9,368 
    Cash to current liabilities    18,437  Offsettable    130,744 
    Unpaid      175,147  Eletrobrás System    402,263 
    CCC      284,250  Petrobrás System    443,643 
            Other    374,274 

     

    Total Net Balance  (330,552) 
    Net balance linked to the concession  421,162 
    Corporate balance  (751,714) 

     

    Page 9


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    The total net balance between assets and liabilities for the liquidation of Eletroacre is negative at BRL 330.5 million, and the balance connected to the concession, indemnifiable assets and compensable liabilities, is positive at BRL 421.1 million. The corporate balance not connected to the concession is negative at BRL 751.1 million.

    · BALANCE SHEET OF AMAZONAS

    On 06/30/2017 On 06/30/2017
    Asset      10,837,074  Liability  20,940,418 
     
    Recoverable      3,369,550  Immediate enforceability  60,823 
    Regulatory Asset      37,202  Social and labor obligations  58,663 
    Financial Asset public service concessions  2,137,373  Post-employment benefits  2,160 
    Reimbursement right CDE    86,860     
    Commercial leasing      1,108,115  Offsettable  2,398,018 
            Regulatory Liability  115,906 
    Cash to current liabilities    318,086  Commercial leasing  1,124,850 
    Cash and equivalents    80,098  Reimbursement obligations  1,157,262 
    Securities      237,988  Sectoral charges   
    Unpaid      1,577,310  Eletrobrás System  1,871,850 
    Customers CP      656,331  Advances for future capital increase  117,446 
    Customers LP      97,988  Eletrobrás RGR  91,616 
    Warehouse      115,001  Eletrobrás RO  1,491,324 
    Services  in  progress    Ame GT  171,464 
    Other Assets      131,794     
    Guarantees and linked deposits    458,095  Petrobrás System  12,593,154 
    Investments      17,107  BR Distribuidora  7,350,034 
    Fixed      100,994  Cigás/Petrobrás  5,243,120 
    CCC      3,968,689  Other  2,411,721 
    Reimbursement right CCC CP    140,806  Suppliers  474,885 
    Reimbursement right CCC LP    3,827,883  Taxes and social contributions  48,306 
            Provision for civil, labor and tax risks  1,688,628 
    Non-current      1,603,439  Other Liabilities  199,902 
    Taxes and social contributions    1,472,176  R&D   
    Intangible      131,263     
            Net equity  1,604,852 
            Onerous concession  378,899 
            RGR Service provision  1,000,596 
    Total Asset      10,837,074  Provision for unsecured liability in subsidiaries  225,357 
    Non-current      1,603,439     
    Total asset for liquidation    9,233,635     
    Recoverable      3,369,550  Total Liability  20,940,418 
    Immediate liquidity      318,086  Net equity  1,604,852 
    Unpaid      1,577,310  Total liabilities for liquidation  19,335,566 
    CCC      3,968,689  Immediate enforceability  60,823 
            Compensable  2,398,018 
            Eletrobras System  1,871,850 
            Petrobras System  12,593,154 
            Other  2,411,721 

     

     

    Page 10


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


     

    Total Net Balance  (10,101,931) 
    Net balance linked to the concession  971,532 
    Corporate balance  (11,073,463) 

     

    The total net balance between assets and liabilities for the liquidation of Amazonas Energia is negative at BRL 10,101.9 million, and the balance connected to the concession, indemnifiable assets and compensable liabilities is positive at BRL 971.5 million. The corporate balance not connected to the concession is negative at BRL 11,073.4 million.

    Page 11


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    · BALANCE SHEET OF CEAL

    On 06/30/2017 On 06/30/2017
    Asset    1,907,905  Liability  2,603,737 
     
    Recoverable    1,015,507  Immediate enforceability  65,176 
     
    Regulatory Asset    103,188  Social and labor obligations  23,043 
    Financial Asset public service concessions  885,014  Post-employment benefits  42,133 
    Reimbursement right CDE    27,305     
    Short-Term Electrical Energy      Offsettable  148,596 
          Regulatory Liability  57,888 
    Cash to current liabilities    75,674  Reimbursement obligations CP  38,764 
    Cash and equivalents    72,667  Sectoral charges  51,944 
    Securities    3,007     
          Eletrobrás System  1,599,658 
    Unpaid    799,285  Advances for future capital increase  159,155 
    Customers CP    337,307  Eletrobrás RGR  54,108 
    Customers LP    307,947  Eletrobrás RO  1,356,407 
    Warehouse    9,737  Chesf (codevasf)  29,988 
    Services in progress    7,204  Distributors   
    Other Assets    32,426     
    Guarantees and linked deposits    73,960     
    Investments    168  Other  446,195 
    Fixed    30,536  Suppliers  142,001 
          Loans and Financings (Other)  6,401 
    CCC Reimbursement      Taxes and social contributions  144,114 
    Reimbursement right CCC CP      Provision for civil, labor and tax risks  125,725 
    Reimbursement right CCC LP      Other Liabilities  27,954 
          R&D   
    Non-current    17,439     
    Taxes and social contributions    17,439  Net equity  344,112 
    Intangible      Onerous concession  7,808 
          RGR Service provision  336,304 
     
     
    Total Asset  1,907,905       
    Non-current  17,439    Total Liability  2,603,737 
    Total asset for liquidation  1,890,466    Net equity  344,112 
    Recoverable  1,015,507    Total liabilities for liquidation  2,259,625 
    Cash to current liabilities  75,674    Immediate enforceability  65,176 
    Unpaid  799,285    Offsettable  148,596 
    CCC      Eletrobrás System  1,599,658 
          Other  446,195 

     

    Total Net Balance  (369,159) 
    Net balance linked to the concession  866,911 
    Corporate balance  (1,236,070) 

     

    The total net balance between assets and liabilities for the liquidation of Ceal is negative at BRL 369.1 million, and the balance connected to the concession, indemnifiable assets and

    Page 12


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    compensable liabilities, is positive at BRL 866.9 million. The corporate balance not connected to the concession is negative at BRL 1,236.0 million.

    · BALANCE SHEET OF BOA VISTA

    On 06/30/2017 On 06/30/2017
    Asset  695,319  Liability  1,405,414 
    Recoverable  227,568  Immediate enforceability  20,080 
    Regulatory Asset  14,853  Social and labor obligations  18,264 
    Financial Asset public service concessions  202,982  Post-employment benefits  1,816 
    Reimbursement right CDE  3,392     
    Short-Term Electrical Energy  6,341  Offsettable  81,095 
        Regulatory Liability  4,096 
    Cash to current liabilities  6,755  Reimbursement obligations  76,999 
    Cash and equivalents Securities  6,755  Sectoral charges   
           
        Eletrobras System  696,666 
    Unpaid  162,459  Advances for future capital increase  97,391 
    Customers CP  82,895  Eletrobrás RGR  7,974 
    Customers LP  8,851  Eletrobrás RO  57,373 
    Warehouse  16,048  Eletronorte  533,740 
    Services in progress Other Assets    Distributors  188 
      9,439     
    Guarantees and linked deposits  29,915  Petrobras System  327,572 
    Investments Fixed    BR Distribuidora  327,572 
      15,311     
        Other  149,400 
    CCC Reimbursement  278,360  Suppliers  62,526 
    Reimbursement right CCC CP  81,150  Taxes and social contributions  22,941 
    Reimbursement right CCC LP  197,210  Provision for civil, labor and tax risks  56,933 
        Other Liabilities  3,494 
    Non-current  20,177  R&D  3,506 
    Taxes and social contributions  15,657     
    Intangible  4,520  Net equity  130,601 
        Onerous concession  14,970 
        RGR Service provision  115,631 
    Total Asset  695,319     
    Non-current  20,177     
    Total asset for liquidation  675,142  Total Liability  1,405,414 
    Recoverable  227,568  Net equity  130,601 
    Cash to current liabilities  6,755  Total liabilities for liquidation  1,274,813 
    Unpaid  162,459  Immediate enforceability  20,080 
    CCC  278,360  Offsettable  81,095 
        Eletrobrás System  696,666 
        Petrobrás System  327,572 
        Other  149,400 

     

    Total Net Balance  (599,671) 
    Net balance linked to the concession  146,473 
    Corporate balance  (746,144) 

     

    Page 13


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    The total net balance between assets and liabilities for the liquidation of Boa Vista is negative at BRL 599.6 million, and the balance connected to the concession, indemnifiable assets and compensable liabilities is positive at BRL 146.4 million. The corporate balance not connected to the concession is negative at BRL 746.1 million.

    · BALANCE SHEET OF CEPISA

    On 06/30/2017 On 06/30/2017
    Asset  1,884,065  Liability  3,154,388 
     
    Recoverable  1,099,406  Immediate enforceability  38,102 
     
    Regulatory Asset  153,350  Social and labor obligations  33,806 
    Financial Asset public service concessions  923,040  Post-employment benefits  4,296 
    Reimbursement right CDE  23,016     
    Short-Term Electrical Energy    Offsettable  81,292 
        Regulatory Liability  70,789 
    Cash to current liabilities  24,063  Reimbursement obligations   
    Cash and equivalents  6,542  Sectoral charges  10,503 
    Securities  17,521     
        Eletrobrás System  1,843,625 
    Unpaid  717,232  Advances for future capital increase  308,978 
    Customers CP  349,472  Eletrobrás RGR  278,615 
    Customers LP  225,652  Eletrobrás RO  1,256,032 
    Warehouse  15,337  Eletronorte   
    Services in progress  46,362  Distributors   
    Other Assets  25,105     
    Guarantees and linked deposits  17,124  Other  675,055 
    Investments  146  Suppliers  134,079 
    Fixed  38,034  Loans and Financing  59,114 
        Taxes and social contributions  220,863 
    CCC Reimbursement    Provision for civil, labor and tax risks  142,231 
    Reimbursement right CCC CP    Other Liabilities  50,083 
    Reimbursement right CCC LP    R&D  68,685 
     
    Non-current  43,364  Net equity  516,314 
    Taxes and social contributions  24,020  Onerous concession  32,691 
    Intangible  19,344  RGR Service provision  483,623 
     
    Total Asset  1,884,065  Total Liability  3,154,388 
    Non-current  43,364  Net equity  516,314 
    Total asset for liquidation  1,840,701  Total liabilities for liquidation  2,638,074 
    Recoverable  1,099,406  Immediate enforceability  38,102 
    Cash to current liabilities  24,063  Offsettable  81,292 
    Unpaid  717,232  Eletrobrás System  1,843,625 
    CCC Reimbursement    Other  675,055 

     

    Total Net Balance  (797,373) 
    Net balance linked to the concession  1,018,114 
    Corporate balance  (1,815,487) 

     

    Page 14


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    The total net balance between assets and liabilities for the liquidation of Cepisa is negative at BRL 797.3 million, and the balance connected to the concession, indemnifiable assets and compensable liabilities, is positive at BRL 1,018.1 million. The corporate balance not connected to the concession is negative at BRL 1,815.4 million.

    · BALANCE SHEET OF CERON

    On 06/30/2017 On 06/30/2017
    Asset  5,700,135  1 Liabilities  7,299,359 
    Recoverable  1,222,975  Immediate enforceability  30,406 
    Regulatory Asset  40,371  Social and labor obligations  25,786 
    Financial Asset - public service concessions  1,165,919  Post-employment benefits  4,620 
    Reimbursement right - CDE  16,685     
    Short-Term Electricity  -  Offsettable  240,497 
        Regulatory Liability  62,255 
    Cash to current liabilities  90,106  Reimbursement obligations CP  13,203 
    Cash and equivalents  37,526  Reimbursement obligations LP  151,353 
        The adjustment of the related liability at   
    Bonds and Securities  52,580    13686 
        present value is recognized as finance costs.   
    Unpaid  483,071  Eletrobrás System  3,294,107 
    Customers CP  230,278  Advance for Future Capital Increase  - 
    Customers LP  16,677  Eletrobras - RGR  58,179 
    Warehouse  16,211  Eletrobras - LPT  18,793 
    Services in progress  14,316  Eletrobras - Energia +  106,013 
    Other Assets  40,778  Eletrobras - RO  701,593 
    Guarantees and linked deposits  135,089  Eletronorte  2,409,529 
    Investments  1,806     
    Fixed Assets  27,916  Petrobrás System  2,467,688 
        BR Distribuidora  2,467,688 
    CCC Reimbursement  3,830,608     
    Reimbursement right - CCC CP  239,539  Other Liabilities  869,054 
    Reimbursement right - CCC LP  3,591,069  Suppliers  473,520 
        Taxes and social contributions  79,361 
    Non-current  73,375  Provision for civil, labor and tax risks  220,745 
    Taxes and social contributions  39,214  Other Liabilities  95,428 
    Intangible Assets  34,161     
        Net equity  397,607 
        Onerous concession  102,958 
        RGR - Service provision  294,649 
    Total Asset  5,700,135     
    Non-current  73,375     
    Total asset for liquidation  5,626,760  Total Liability  7,299,359 
    Recoverable  1,222,975  Net equity  397,607 
    Cash to current liabilities  90,106  Total liabilities for liquidation  6,901,752 
    Unpaid  483,071  Immediate enforceability  30,406 
    CCC  3,830,608  Offsettable  240,497 
        Eletrobrás System  3,294,107 
        Petrobrás System  2,467,688 
        Others  869,054 

     

    Total Net Balance  (1,274,992) 
    Net balance linked to the concession  982,478 
    Corporate balance  (2,257,470) 

     

    Page 15


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    The total net balance between assets and liabilities for the liquidation of Ceron is negative at BRL 1274.9 million, and the balance connected to the concession, indemnifiable assets and compensable liabilities, is positive at BRL 982.4 million. The corporate balance not connected to the concession is negative at BRL 2,257.4 million.

    4. ADJUSTMENTS OF THE LIQUIDATION COST

    · COST OF REMOVAL OF PERSONNEL

    For the exercise of the liquidation cost, a 50% FGTS fine was estimated for each distributor, plus twice of the projected payroll sum established as notice.

    The values considered are summarized in the table below:

      Termination 
      Cost 
    AmE D  260,685 
    Ceron  215,165 
    Eletroacre  72,966 
    Boa Vista  126,598 
    Ceal  130,425 
    Cepisa  309,169 
    Total  1,115,009 

     

    · BRESSER PLAN

    Similarly to the Financial Statements of Ceal on June 30, 2017, only the value of BRL 11.79 million is allocated to Bresser Plan, and that made an adjustment necessary to such account given the high risk of loss in this proceeding. The last proposal sent by Eletrobras to the Union was BRL 178.0 million, which corresponds to 10% of the total adjusted value of the value. This way, the difference between what has already been allocated to the Bresser Plan and the last proposal of Eletrobras was the value of the adjustment of contingent liabilities of Ceal, amounting to BRL 166.20 million.

    · FINANCIAL ASSETS OF THE CONCESSION - IN USE

    The financial asset of the concession was treated as an indemnifiable asset in the liquidation calculation. However, an assumption of a disallowance of 10% of the value of the financial assets was adopted over the course of the position of June 2017. This

    Page 16


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    premise was established assuming that, on average, 5% of these assets are not unitized on the average of the Brazilian distributors. As the distributors of Eletrobras have a history of performance lower than the others, the double amount of such disallowance was adopted.

    The table below shows the values set aside, as well as those considered as indemnifiable.

    Financial
    Asset in
    progress
    Adjustment
    (10%)
    Financial
    Asset in
    progress
    considered
    recoverable 
    AmE D  559,580  (55,958)  503,622 
    Ceron  497,780  (49,778)  448,002 
    Eletroacre  110,541  (11,054)  99,487 
    Boa Vista  45,860  (4,586)  41,274 
    Ceal  148,645  (14,865)  133,781 
    Cepisa  97,628  (9,763)  87,865 
    Total  1,460,034  (146,003)  1,314,031 

     

    · CUSTOMERS

    For the purposes of the calculation of liquidation, the Customer account already net of Provision for Doubtful Debtors (LCPD) was treated as provision to realize. Nonetheless, adjustments were made in keeping with the time over which such invoices remained overdue, in addition to the smaller probability of realization of such assets. Accordingly, the following assumptions were adopted for the customer account, with PCLD already discounted:

    ü Unmatured adjustments were not made since invoices of current receipt of the distributor were considered, and they have not become overdue as of yet;ü Overdue up to 90 days - disallowance of 20%ü Overdue for more than 90 days - disallowance of 50%ü Debt claims renegotiated disallowance of 50%.

    Page 17


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    Below are the customers’ accounts, the value of the adjustments and the value considered as assets to realize for calculating the cost of liquidation:

    Customers Adjustment Customers
    considered
    unpaid 
    AmE D  754,319  (153,570)  600,749 
    Ceron  246,955  (35,850)  211,105 
    Eletroacre  130,135  (35,971)  94,165 
    Boa Vista  91,746  (17,863)  73,883 
    Ceal  645,254  (207,549)  437,705 
    Cepisa  575,124  (187,803)  387,321 
    Total  2,443,533  (638,606)  1,804,927 

     

    · POSSIBLE CONTINGENCIES

    In the liabilities of the Eletrobras’ distributors, a 100% provision is accounted for losses classified as probable, as a default procedure. This way, the provisions considered possible and remote are not recorded under liabilities, although they are broken down in the company’s explanatory notes. Thus, the premise adopted in this IT was that 30% of the provisions classified as possible shall be included in the calculation of the liquidation cost, in an attempt to quantify the increase in the legal issues that a liquidation process will bring to the companies.

    5. CONCLUSION

    In view of the foregoing in this Technical Information, it may be concluded that, from a strictly economic and financial and accounting viewpoint, the liquidation value of all six distributors for the base date of June 30, 2017 is negative at BRL 17,731.29 million, with the liquidation taking place within Eletrobras, as can be seen in the table below. In the last line of the table, the liquidation value ascertained by BNDES is shown.

    Page 18


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017



    It should be noted that, in the assessment for privatization of Amazonas Energia, performed by BNDES, Amazonas G&T was treated as neutral, i.e., there were neither positive impacts, nor negative impacts on the cash flow and the distributor’s valuation.

    Thus, to make these numbers comparable, the same assumption was applied for calculating the cost of liquidation.

    Page 19


     
     

    Date: December 26, 2017

    IT DFP/DFC/DDEF

    002/2017


    The costs of termination of current contracts of the companies, as well as all the costs associated with the liquidation process, where a structure shall have to be maintained by Eletrobras for a long period, were not considered in the aforementioned liquidation value.

    Lastly, it should also be noted that the liquidation cost, presented in the table above, may contain inaccuracies and/or distortions, not necessarily representing the actual cost of liquidation of the distributors, which can only be determined in a formal process based on defined rules, which are currently unknown and, for this reason, should be taken as an indicative reference for the BNDES’ assessment.

     

     

     

    Felipe Baptista da Silva DFP 

     

     

    Rodrigo Vilella Ruiz - DFC 

     

         

    Marcos B. F. Pinho DFPI 

     

     

    Frederico Pinto Eccard - DDEF 

     

         

    Marcos José Lopes - DFCC 

     

     

    Luciana Pereira de Souza - DDEF 

     

     

    Page 20

     

    SIGNATURE
     
     
    Pursuant to the requirements of the Securities Exchange Act of 1934, the registrant has duly caused this report to be signed on its behalf by the undersigned, thereunto duly authorized.

    Date: January 8, 2018
    CENTRAIS ELÉTRICAS BRASILEIRAS S.A. - ELETROBRÁS
    By:
    /SArmando Casado de Araujo
     
    Armando Casado de Araujo
    Chief Financial and Investor Relation Officer
     
     

     

     
    FORWARD-LOOKING STATEMENTS

    This press release may contain forward-looking statements. These statements are statements that are not historical facts, and are based on management's current view and estimates offuture economic circumstances, industry conditions, company performance and financial results. The words "anticipates", "believes", "estimates", "expects", "plans" and similar expressions, as they relate to the company, are intended to identify forward-looking statements. Statements regarding the declaration or payment of dividends, the implementation of principal operating and financing strategies and capital expenditure plans, the direction of future operations and the factors or trends affecting financial condition, liquidity or results of operations are examples of forward-looking statements. Such statements reflect the current views of management and are subject to a number of risks and uncertainties. There is no guarantee that the expected events, trends or results will actually occur. The statements are based on many assumptions and factors, including general economic and market conditions, industry conditions, and operating factors. Any changes in such assumptions or factors could cause actual results to differ materially from current expectations.