Rents in the electricity generation sector

Rents in the electricity generation sector

Report within the scope of

Measure 5.15 of the 2nd update of the MoU onspecific economic policy conditionality

31-01-2012

CONTENTS

1. Introduction............................................................11.1Background................................................................11.2Program for Financial Assistance....................................31.3Goal and Structure of the Report....................................4

2. Rate of Return Benchmark......................................62.1CMEC........................................................................72.2PPA..........................................................................72.3PRE..........................................................................8

3. Effective Rates of Return.........................................93.1CMEC........................................................................93.2PPA..........................................................................93.3PRE........................................................................10

4. Excessive Rents of Generation Assets....................114.1CMEC......................................................................114.2PPA........................................................................124.3PRE........................................................................134.4Total.......................................................................14

5. Correction of Excessive Rents: Impacts on the Tariff Debt Reduction and on Electricity Prices........15

6. Conclusions...........................................................17

Appendices

Secretary of State for Energy‘s OfficeRua da Horta Sêca, nº 15, 1200-221 Lisbon, PORTUGAL

TEL + 351 21 324 54 00 FAX + 351 21 324 54 60 EMAIL [email protected] www.portugal.gov.pt

Appendix 1Benchmark on rates of return for generation assets in Portugal, Cambridge Economic Policy Associates, Ltd, January 2012

Appendix 2Benchmark on opportunity cost of capital for investments in the special regime,AT Kearney, January 2012

Appendix 3Effective rates of return of PRO generation assets (CMEC and PPA)

Appendix 4Effective rate of return of PRE generation assets (wind, small hydro, photovoltaic, biomass, CHP)

Appendix 5Comment on electricity pricing policy options

Appendix 6References



1. INTRODUCTION

1.1 BACKGROUND

Energy policy undertaken since the 90’s in the Portuguese National Electric System (SEN) has been continuously based on mechanisms devoted to support generation agents. Generation business is developed under an ordinary regime (PRO) and a special regime (PRE), both benefiting from guarantees and support schemes that considerably reduce, or even eliminate, market risks of the said agents, as follows:

• PRO- Power Purchase Agreements (PPAs): In force since the 90’s, the remaining two PPA’s (600 MW coal power plant and 1000 MW CCGT power plant) benefit from pre-established profitability, which is independent from the effective generation of the power plants (quantity risk-free) and from fuel costs (price risk-free); Additionally, there is an incentive (or a penalty) depending on the relation between real and contractual availability; Major shareholder of those power plants is International Power;- Costs of Maintenance of the Contractual Equilibrium (CMECs): In force since July 2007 and regulated by Decree-Law 240/2004, of December 27th, which sets the rules for the early termination of the PPAs, this mechanism is applied to a set of power plants (currently a1180 MW coal power plant, a 940 MW fuel-oil power plant and 26 large hydro power plants with a total installed capacity of 4100 MW) which sell generated electricity in the market and profit from a compensation that supposedly corresponds to the difference between the market obtained income and the one that would be obtained under PPA regime; Also in this case, asset remuneration is guaranteed and is independent from market income; All the CMEC power plants belong to EDP;

1Secretary of State for Energy‘s Office

Rua da Horta Sêca, nº 15, 1200-221 Lisbon, PORTUGALTEL + 351 21 324 54 00 FAX + 351 21 324 54 60 EMAIL [email protected] www.portugal.gov.pt

- Other: The remaining PRO generating capacity (currently, 3 CCGT power plants: 1175 MW + 830 MW + 830 MW; and a large hydro power plant: 240 MW) benefit from power guarantee mechanism, justified by a contribution that they provide to security of electricity supply.

• Renewable PREGeneration plants using renewable energy sources, benefiting from dispatch priority (grid reception of the generation is guaranteed) and from a feed-in tariff (remuneration scheme that guarantees a fixed purchasing price along the major part of the operating lifetime – typically first 15 years); There are around 500 renewable PRE power plants, corresponding to an installed capacity of 6000 MW;

• Combined heat and power (CHP) PRECHP benefits from a guaranteed remuneration for the total electricity generation (including generation for self consumption), which is based on the “avoided costs” methodology and is immune to inflation rate, oil price and currency variations; There are around 160 CHP power plants for an installed capacity of 1550 MW.

Energy policy options that have been adopted in the last two decades allowed that more than 95% of the total installed capacity in the Portuguese electric sector work in a mitigated risk (or even a risk-free) environment, where the costs associated to support schemes are transferred to the electric system and, therefore, to electricity consumers.Under this policy, such incentives and support mechanisms paid to electricity generators (included in the so-called CIEG’s - Economic General Interest Costs) should be totally transferred to consumers via access tariff. In the last five years, the growth of these system costs would have had strong impacts on the final price of electricity, if such transfer to consumers had actually occurred. To partially



avoid those politically and economically sensitive impacts on consumers’ bills, it was decided that the costs would be deferred, the corresponding debt being provisionally computed in the balance sheet of the supplier of last resort.At the end of 2011, the accumulated value of the tariff debt was 1750 M€. It is being passed to consumers through the grid access costs by an annual payment of 170 M€ average up to 2024, in accordance with Decree-Law 165/2008, of August 21st. In 2012, if total costs were to be reflected in the grid access tariff, electricity prices would increase 27% in average, in addition to the increase in VAT from 6% to 23%, in force as of October 2011. Due to this fact, in 2012 a new tariff deficit of 1080 M€ will be created (in addition to the existing debt) by the application of a smoothing mechanism of the renewable PRE overcosts (Article 73-A of the Decree-Law 78/2011, of June 20th) and of an extraordinary deferral of the payment of the CMEC adjustment to 2013.If no measures are promptly taken, the payment of all the tariff debt and the future costs of the electric system between 2012 and 2020 will lead to an annual growth rate of electricity prices of 4,7%. Taking into account that, in 2012, Portuguese residential consumers will already have the most expensive electricity in Europe (corrected by power purchase parity) and that electricity prices for industry will be higher than the European average (access tariffs increased up to 20%, even with the creation of the new debt), it appears clear that this is an unsustainable scenario for the Portuguese economy competitiveness and for the families and, therefore, that the Portuguese electric system is unsustainable from the economic point of view.In addition, recent years events in the financial markets have largely increased the cost of debt, thus aggravating this already difficult situation.



1.2 PROGRAM FOR FINANCIAL ASSISTANCE

The solution for the problem identified above should be achieved by progressively correcting the unreasonable and unbalanced responsibility of financing the electric system – currently impending on electricity consumers –, by way of (i) searching new sources of income; (ii) mobilizing the generation agents – major beneficiaries of the protection system in force – to cooperate in this balancing process; and (iii) engaging the elimination of excessive and unjustified support to certain agents, in order to stimulate electricity market.These solutions are in line with the Program for Financial Assistance signed by the Portuguese State, the European Union, the International Monetary Fund and the European Central Bank, which explicitly states that measures must be taken in order “to limit the additional cost associated with the production of electricity under the ordinary regime, in particular through renegotiation or downward revision of the guaranteed compensation mechanism (CMEC) paid to producers under the ordinary regime and the remaining long-term power-purchase agreements (PPAs)”, to review the efficiency of support schemes to CHP and renewables and, regarding new contracts, to “revise downward the feed-in tariffs and ensure that the tariffs do not over-compensate producers for their costs”.Within the framework of the second update of the Memorandum of Understanding (MoU) on specific economic policy conditionality, dated 9th of December 2011, a new measure was introduced with the aim of achieving electric sector sustainability, as follows:“5.15. Measures to set the national electricity system on a sustainable path leading to the elimination of the tariff debt (déficit tarifário) by 2020 and ensuring that it will stabilise by 2013 will be adopted. The latter deadline is subject to a review based on a government proposal which will also specify how excessive rents in the standard (CMECs, PPAs, and power guarantee mechanism) and



special regimes (co-generation and renewables) will be corrected. This proposal will consider the merits of a full range of measures which will cover all sources of rents”.

1.3 GOAL AND STRUCTURE OF THE REPORT

The timeframe to accomplish the above-mentioned measure is extremely compressed, since it must be fulfilled until the end of January 2012. The present report aims at identifying the excessive rents in electricity generation sector and the impacts of their correction on the tariff debt, in order to provide an adequate answer to the issues raised by measure 5.15 of the updated MoU, i.e., how excessive rents will be corrected.For the purpose of a clear understanding of the process, the structure of the report reflects the adopted approach:Chapter 2 provides the benchmark for the efficient rate of return on investments associated with the generation assets of the regimes in force (PPA, CMEC, PRE1). It is based on the correspondent cost of capital (WACC – Weighted Average Cost of Capital), calculated taking into account all the relevant aspects ( characterization of the different source of risks and the correspondent level, gearing level, remaining maturity).Chapter 3 is devoted to the evaluation of the effective rate of return of PPA, CMEC and PRE generation assets. In the case of PPA power plants, the rate of return was determined considering the initial investment and the fixed charged paid along the contracts duration. The rate of return of the generation assets under CMEC regime is determined by computing financial neutrality with the former PPA regime, taking into account the opportunity cost of capital in the 1 Power guarantee mechanism is not addressed in this report since it is currently under revision by the Government, in accordance with measure 5.13 of the MoU. The revision will (i) eliminate payments to power plants decided before the existence of the mechanism (some of them were already in operation for several years); and (ii) establish a new rational that provides incentive to build additional generating capacity, taking into account adequate security of supply levels and the existence of other mechanisms, such as interruptibility service.



starting date of CMEC and the two main aspects of the CMEC process: The calculation of the value of the former PPAs when they were terminated; and The calculation of the value of the extension of the use of water from the end of the hydro PPAs to the end of the concession of public hydrological domain. Regarding the PRE regime assets, the correspondent rates of return were calculated using historical data for CAPEX, OPEX and generation revenue.Chapter 4 shows the excessive rents for each of the generation regimes in force, which are derived from the difference between the effective rate of return and the correspondent WACC benchmark. They are expressed in relative terms, as well as in M€ per year. The expected future impacts associated with the correction of the excessive rents and the accumulated value of the excessive rents since the beginning of the operation of the generating assets are also presented.Chapter 5 evaluates the positive impacts on the reduction of tariff debt and on electricity prices that can be achieved with the correction of the excessive rents, including a welfare analysis based on the transferred amounts from generation agents to the consumers, the associated surplus and the results in terms of economic efficiency.Chapter 6 provides the main conclusions of the report: the amount of excessive rents for each of the generating regimes; and the impacts on the evolution of the electric system debt and the electricity prices in an excessive rents elimination scenario.The report is supported on the following set of appendices:Appendix 1 – Benchmark on rates of return for generation assets in Portugal, Cambridge Economic Policy Associates, Ltd, January 2012;Appendix 2 – Benchmark on rate of return for investments in the special regime, AT Kearney, January 2012;Appendix 3 – Effective rate of return of PRO generation assets (PPA and CMEC);



Appendix 4 – Effective rate of return of PRE generation assets (wind, small hydro, photovoltaic, biomass, CHP);Appendix 5 – Comment on electricity pricing policy options, Prof. João ConfrariaAppendix 6 – References



2. RATE OF RETURN BENCHMARK

The adequate rates of return on investments associated with the generation assets of the regimes in force (PPA, CMEC, PRE) must be based on patterns of economic efficiency. Those patterns are reflected by the opportunity cost of capital for regulated generators, at the time the investments (or the transactions/contracts) were made. Therefore, they are provided by the correspondent WACC (Weighted Average Cost of Capital).The determination of WACC requires estimates of:• The required rate of equity return/cost of equity (COE);• The required rate of debt return/cost of debt (COD); and• The mix of debt and equity in the capital structure – also referred to as the level of gearing.

The estimations must take into account relevant aspects, like the characterization of the different source of risks and the correspondent level, the gearing level and the (remaining) maturity of the assets.The WACC for the ordinary regime generation (CMEC and PPA) was determined by CEPA - Cambridge Economic Policy Associates Ltd (see Appendix 1), using a benchmark based on a set of representative generation companies (for the equity measure) and raising bond finance (for the debt measure), which were then adjusted to the PPA and CMEC Portuguese cases. Nevertheless, the adopted approach has led to WACC estimates on the upper side of the range applicable to that of PPA's and CMEC's, since the benchmarks used seem not to include any power generators facing PPA's as immune to systematic risk factors as those of the Portuguese generators2.For the special regime generation (PRE: Wind, small hydro, photovoltaic, biomass, CHP), the WACC was estimated by AT 2 Looking at substantial risk factors, in the case of the Portuguese PPA's there is no systematic risk on these contracts. In practice, generating companies are exposed to some volatility in capital markets (not due the substantial risk factors) which seems to have a small positive correlation with market volatility. This justifies an asset beta slightly above zero.



Kearney (see Appendix 2), which used a similar approach for each of the different technologies.Since it was adopted the perspective of the system costs, the relevant WACC should be provided by pre-tax values. That is the case of the WACC for PPA and CMEC regimes. Nevertheless, due to the high number of PRE power plants and the time constraints to perform this work, the effective rate of return of those power plants is provided by their Internal Rate of Return (IRR), which means that it must be compared with the opportunity cost of capital post-tax.The estimated WACC values that reflect the opportunity cost of capital are as follows.

2.1 CMEC

The opportunity cost of capital for the CMEC generating assets was assumed as the WACC in the starting date of this mechanism (July 2007). Since the substantial market risk factors are identical for the different generating technologies, it was assumed the same asset beta for all the contracts. The same assumption was taken for the gearing level, because EDP has no project finance associated with individual CMEC power plants and that the underlying risk of the different contracts is practically the same for all power plants, regardless of technology.In accordance with CEPA study, and taking into account the remaining maturity of the CMEC generating assets in 2007, the correspondent WACC should stand between 5,4% pre-tax real (remaining maturity < 10 years) and 5,8% pre-tax real (remaining maturity 10-30 years)3. Considering that the WACC used under the framework of Decree-Law 240/2004, of December 27th, was within the above range, it was adopted that same value:

WACC = 7,55% (pre-tax nominal)

3 Pre-tax nominal interval: (1+5,4%)*(1+2%)-1; (1+5,4%)*(1+2%)-1] = [7,51%;7,92%]



2.2 PPA

The guidelines for the structural changes in the organization of the electric system, required to develop a competitive and efficient electricity market, in accordance with the Directive 2003/54/EC, of June 26th, which established the common rules for the European internal electricity market, started to be settled by the Decree-Law 185/2003, of August 20th.The ability to implement the guidelines for the generating system, defined with the aim of the creation of the Iberian electricity market (MIBEL), was provided by the legal authorization created by the Law 52/2004, of October 29th, which has defined the conditions for the early termination of the PPAs and the correspondent adequate compensation measures.In the case of the current two remaining PPAs, the Decree-Law 240/2004, of December 27th, was not applied because their holders have decided for their non-termination, and the previous remuneration scheme was kept without any change. Taking into account that the evolution of regulatory and economic frameworks (namely due to the creation of regional energy markets and the Euro currency) have introduced relevant changes in the financial market conditions, the remaining two PPAs should have been, at least, renegotiated in order to incorporate those conditions.Under this context, it is assumed that the opportunity cost of capital for the two remaining PPAs is given by the WACC at the date of the termination of the other PPAs (July 2007), taking into account the correspondent remaining maturity (10-30 years):

WACC (Tejo Energia) = 7,92% (pre-tax nominal)WACC (Turbogás) = 7,92%(pre-tax nominal)

2.3 PRE

The opportunity cost of capital for special regime generation is given by the following values of WACC (post-tax nominal):



1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010Wind - - - 8,6% 8,3% 8,0% 7,9% 7,5% 7,3% 6,8% 7,2% 6,8% 6,7% 9,7%Photovoltaic - - - 8,7% 8,5% 8,2% 8,1% 7,8% 7,6% 7,0% 7,4% 7,0% 7,0% 10,1%Biomass - - - 8,9% 8,6% 8,3% 8,2% 7,9% 7,7% 7,1% 7,5% 7,1% 7,0% 10,2%Small Hydro - - - 8,3% 8,1% 7,8% 7,7% 7,4% 7,2% 6,8% 7,2% 6,9% 6,8% 9,8%CHP 8,8% 7,6% 8,6% 8,6% 8,4% 8,1% 8,1% 7,7% 7,6% 7,0% 7,5% 7,1% 7,0% 10,2%

For a certain PRE investment, the correspondent WACC is given by the estimated value for the year prior to the commissioning date of the power plant.



3. EFFECTIVE RATES OF RETURN

This chapter presents the effective rate of return of CMEC, PPA and PRE generation assets.

3.1 CMEC

The effective rate of return of the generation assets under CMEC regime is determined taking into account the former PPAs return on assets (10,67% pre-tax nominal) and assessing the two main financial operations of the CMEC process: The calculation of the value of the former PPAs when they were terminated; and The calculation of the value of the extension of the use of water from the end of the hydro PPAs to the end of the concession of public hydrological domain (see Appendix 3).

0%

2%

4%

6%

8%

10%

12%

14%

0%

2%

4%

6%

8%

10%

12%

14%

Former PPAs return Effect of CMEC initialvalue calculation

Effect of hydroconcession extension

Effective return

The effective rate of return obtained for the generating assets under CMEC mechanism is:

14,22% pre-tax nominal4.

4 This effective rate of return does not include the availability premium (30 M€/year average) and the correction factor applied to hydro and coal generation calculated by Valoragua model in the annual revision of the initial value considering real conditions (10 M€/year average).



3.2 PPA

In the case of PPA power plants, the rate of return was determined considering the value of assets and the cash-flows given by the “fixed charged”.

The effective rates of return estimated for the PPAs generating assets are:

Tejo Energia – 13,23% pre-tax nominal 5;Turbogás – 12,91% pre-tax nominal 6.

3.3 PRE

Profitability associated with the investments in the special regime is based in the internal rate of return (IRR) of the projects. Due to the large number of PRE power plants (around 500) and the time constraints to develop this work, the IRR for the different special regime technologies were based on the relevant characteristics of the projects, with the intent of representing the technology typical conditions (and not specific projects). If the evolution of the relevant variables is considered, typical characteristics are adequate to evaluate the global value of the excessive rents. Nonetheless, any solution to correct those rents will require a more detailed assessment on the economics of the different projects.For the calculation of IRR of the different PRE technologies, the following variables were carefully determined:

- Feed-in-tariff;- Licensing and grid connection date;- Net operating hours;- Plant expected life-time;

5 This effective rate of return does not include the availability premium (1-2 M€/year) and sale of ashes (2-3 M€/year).

6 This effective rate of return does not include the availability premium (4 M€/year average)



- CapEx (i.e. EPC costs, electrical and grid connection costs, promotion costs, other);- OpEx (i.e. O&M, Fuel, Insurances, Rents, other).

These variables were estimated on an annual basis since 2000 until 2010 (for CHP since 1997). For that purpose, several sources were used and industry experts were consulted to cross-check assumptions and results.The obtained IRR values, referred to the licensing date, are as follows (nominal terms):

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010Wind - - - 8,3% 9,3% 9,5% 9,8% 9,1% 7,9% 7,8% 7,2% 6,9% 8,1% 7,3%Photovoltaic - - - 0,9% 1,5% 2,9% 5,0% 5,8% 11,0% 5,0% 5,3% 6,6% 10,3% 9,5%Biomass - - - 6,0% 6,9% 7,7% 8,5% 9,5% 10,0% 9,8% 9,6% 9,8% 10,3% 10,1%Small Hydro - - - 7,9% 8,5% 9,0% 9,5% 10,0% 9,9% 10,1% 10,1% 10,5% 10,9% 11,0%CHP < 10 MW 15,2% 14,5% 14,6% 15,1% 16,0% 16,5% 16,6% 16,7% 16,0% 15,3% 15,5% 14,7% 15,5% 15,8%CHP > 10 MW - - - 15,6% 15,8% 16,4% 16,5% 16,5% 15,5% 14,0% 13,3% 12,6% 12,9% 13,0%



4. EXCESSIVE RENTS OF GENERATION ASSETS

The excessive rents are determined by evaluating eventual differences between the return on the resource initially contracted and the opportunity cost of that resource at the time the contracts were made. They are expressed in relative terms, as well as in M€ per year. The expected future impacts associated with the correction of the excessive rents and the current accumulated value of the excessive rents are also presented.

4.1 CMECAs shown in section 3.1, the excessive rents in the CMEC regime are due to three factors:

- Calculation of former PPAs value ate the date CMEC mechanism started (July 2007);- Calculation of the value for the extension of hydro concession;- Rate of return on CMEC assets (10,67% instead of 7,55% nominal pre-tax, which was the opportunity cost of capital of EDP in 2007).

The correction of the excessive rents associated with these factors has the following effects:

WACC nominal pre-tax (%): 7,55

Effective rent nominal pre-tax (%): 14,22

Excessive rent nominal pre-tax (%): 6,67

Annuity of excessive rent (M€2011/ ano): 165

Value of excessive rent correction 2012-2020 (M€2011): 1348

Value of excessive rent 2007-2011 (M€2011): 785

Value of excessive rent 2007-2020 (M€2011): 2133

The effect on the annuity due to the correction of the excessive rents in CMEC regime is disaggregated as follows:



Effi cient CMEC annuity

-100.000

-80.000

-60.000

-40.000

-20.000

0

20.000

40.000

60.000

80.000

100.000

-100.000

-80.000

-60.000

-40.000

-20.000

0

20.000

40.000

60.000

80.000

100.000

CurrentCMEC

annuity

Correctionof CMEC

initial value

Eq. annuityof former

PPAs return

Correctionformer

PPAs return

Correctionof hydro

concessionextension

value

EfficientCMEC

annuity

103

EUR

per y

ear

There are other sources of rents that may be addressed (and were not included in the rate of return or in the annuity calculations in the above figures), such as the availability premium (30 M€/year average) and the correction factor applied to hydro and coal generation in the annual revision (10 M€/year average). Regarding the former, it can be argued that it is a result of a good performance and adequate investments in the power plants, however the current level of this premiums suggests that there is potential to improve the allocation of benefits, risks and costs regarding this incentive.

4.2 PPAExcessive rents in current remaining PPAs are due to the high rates of return of the underlying generating assets. The correction of the excessive rents has the following effects:



Tejo Energia Turbogás

WACC nominal pre-tax (%): 7,92 7,92

Effective rent nominal pre-tax (%): 13,23 12,91

Excessive rent nominal pre-tax (%): 5,31 4,99

Annuity of excessive rent (M€2011/ ano): 13 8

Value of excessive rent correction 2012-2020 (M€2011): 106 65

Value of excessive rent 2007-2011 (M€2011): 62 38

Value of excessive rent 2007-2020 (M€2011): 168 103As it happens with CMEC power plants, in PPAs are other sources of rents that may be addressed, such as the availability premium (1-2 M€/year in Tejo Energia and 4 M€/ year average in Turbogás) and the sale of burnt coal ashes (2-3 M€/year in Tejo Energia). The rate of return presented above does not include these rents. As well as in the case of previous section, there is a potential for improving the allocation of these benefits and costs.

4.3 PRE

The excessive rents of PRE power plants correspond to the difference between IRR and WACC, determined on an annual basis for each technology. In order to assess the impact of the excessive rents from the point of view of the consumers (costs), the reduction of the feed-in-tariff required to eliminate the excess was calculated whenever IRR was higher that the WACC.The effects of the correction of the estimated excessive rents are as follows:



Wind Photovoltaic Biomass Small Hydro CHPAnnuity of excessive rent (M€2011/ ano): 54 9 5 3 42

Value of excessive rent correction 2012-2020 (M€2011): 439 73 42 25 339

Value of excessive rent 2000-2011 (M€2011): 284 32 14 9 264

Value of excessive rent 2000-2020 (M€2011): 723 105 56 35 602

A remark must be done regarding CHP and small hydro:i. Due to lack of information about CHP plants built prior to

1997 and the compressed timeframe to develop this work, they were not considered. If we assume that they have an IRR and an opportunity cost of capital equivalent to the 1997 CHP plants and taking into account that their installed capacity is over 500 MW, the associated yearly excessive rent is approximately 25 M€.

ii. Performing the same exercise to small hydro, the amount of the excessive rent associated with the plants that were commissioned before 2000 (250 MW) can reach 5 M€.

4.4 TOTAL

CMECs PPAs PRE TOTALAnnuity of excessive rent (M€2011/ ano): 165 21 113 299

Value of excessive rent correction 2012-2020 (M€2011): 1348 171 919 2439

Value of excessive rent up to 2011 (M€2011): 785 100 602 1486

Total value of excessive rent up to 2020 (M€2011): 2133 271 1521 3925



5. CORRECTION OF EXCESSIVE RENTS: IMPACTS ON THE TARIFF DEBT REDUCTION AND ON ELECTRICITY PRICES

The assessment of the impacts on the tariff debt reduction and on electricity prices is supported on the base case developed by the Ministry of Economy and Employment / Secretary of State for Energy along the second half of 2011, namely on the:

- Assumptions (electricity demand, inflation rate, fuel prices, electricity market prices); and

- Measures to be adopted (80% of the revenues of CO2 trading of electricity companies, the regulation of current CHP law and the revision of the power guarantee mechanism).

The expected evolution of the electric system costs is as follows.

2013E

7,1 0,4

1,1

2,6

2012E

7,3 0,4

1,4

1,9

2011

6,2

0,4

1,5

1,5

0,4

+18%

2020E

7,9

4,5

8,1

2018E

7,7

4,8

2019E

4,3

2017E

7,4

+2,3% pa

0

-1,9% pa

M €10

8

6

4

2

0,4

3,5

2014E

7,0

3,1

0,7

3,04,0 4,0

2015E

7,1

2016E

7,30,4

0,40,4 0,4

3,7 3,0 2,8 2,8 2,9 3,03,0

3,0 2,9

Energy & Comercialisation (liberalized)Excessive Rents + Pwr. Guar. + CHP lawEnergy & Comercialisation (regulated)Access Tariff

Previsional evolution of regulated and liberalized costs

The excessive rents of CMEC, PPAs and PRE (~300 M€/year), the power guarantee mechanism and the over compensation of older CHP power plants represent more than 5% of electric system total costs and 13% of the access tariff.



The elimination of these costs, associated to the revenues of 80% of the CO2 allowances of electricity companies, will provide the conditions required to fulfill the measure 5.15 of the MoU:

- Stabilize the tariff debt by 2013; and- Total elimination of the tariff debt by 2020.

Impacts on the evolution of the tariff debt (mM€)

6

5

4

3

2

1

0

CHP Law

Power Guarantee

Excessive rents

2020E

5,4

1,8

2,8

2019E

5,7

0,7

0,2

1,7

2,5

2018E

5,8

0,8

0,7

1,5

0,3

2,2

2017E

5,7

0,9

1,1

1,3

1,9

2016E

5,5

1,1

1,5

1,0

1,5

2015E

5,1

1,2

1,7

1,1

2014E

4,6

1,4

1,8

0,7

2013E

4,0

1,5

1,7

0,4

2012E

2,9

1,6

1,2

0,5

2011

1,8

1,8

2010

1,9

1,9

Existing debt

New debt

CO2

0,3 0,8 0,3

0,3

0,30,30,20,1

0,1 0,1 0,10,2

0,30,3

0,3

Evolution of the tariff debt with correction of excessive rents and other measures (mM €)

6

5

4

3

2

1

0

-1

2020E

0.0

2019E

0.9

0.70.2

2018E

1.5

0.80.6

2017E

2.0

0.90.8

0.3

2016E

2.6

1.10.9

0.6

2015E

2.9

1.2

0.6

1.2

2014E

3.2

1.4

0.6

1.3

2013E

3.2

1.50.4

1.3

2012E

2.9

1.6

0.30.9

2011

1.8

2010

1.9

-2

Existing debt

New debt - Additional

New debt – Renew. deferral

Furthermore, this evolution of the tariff debt has associated a set of important advantages for the economy and for the country:- Limits to 1,0% real the average annual growth of electricity

prices from 2012 to 2020;- Although it represents an increase in real terms, it fulfills

important targets already committed by the Government such



as the one in the framework of the social dialogue, limiting the growth of electricity prices to 1,5% real;

- Avoid the creation of debt over 3200 M€, setting free financial resources to the development of the economy;

- Transfers resources from generation agents to consumers, which promote the creation of surplus and improvement in the economic efficiency (see Appendix 5).



6. CONCLUSIONS

The present report aimed at providing a comprehensive response to measure 5.15 of the updated MoU, namely the identification of excessive rents in electricity generation sector and the impacts of their correction on the tariff debt and on electricity prices.The excessive rents have been determined by evaluating eventual differences between the return on the resource initially contracted and the opportunity cost of that resource at the time the contracts were made. Given this, the opportunity cost of capital for the different generation regimes (CMEC, CAE and PRE) were identified taking into account the different source of risks and the correspondent level, gearing level and the remaining maturity of the generating assets. The subsequent evaluation of the effective rates of return allowed to derive the excessive rents.

If the cumulative transfer of income from producers to consumers is smaller than the cumulative value of excessive rents, prices will always be enough to cover the opportunity cost of resources used in electricity production. In particular, shareholders will always receive a return at least as large as the opportunity cost of their capital. However, if the cumulative transfer of income from producers to consumers is larger than the value of excessive rents, the Government should use revenues from other sources, such as carbon trading of electricity companies, to compensate the difference, and ensuring that stakeholders always receive the opportunity cost of their equity.



CMECs PPAs PRE TOTALAnnuity of excessive rent (M€2011/ ano): 165 21 113 299

Value of excessive rent correction 2012-2020 (M€2011): 1348 171 919 2439

Value of excessive rent up to 2011 (M€2011): 785 100 602 1486

Total value of excessive rent up to 2020 (M€2011): 2133 271 1521 3925

In practice, as it was demonstrated in chapter 5, limiting the transfers from generators to consumers to the cumulative value of the excessive rents, associated with 80% of the revenues of CO2

trading of electricity companies and the regulation of current CHP law, will allow total elimination of tariff debt up to 2020 and keep annual average electricity prices growth under 1,0% real. Therefore, in accordance with the economic theory of regulation, the economical efficiency is enhanced.The first step towards the correction of the above identified excessive rents requires a negotiation process with the involved generators. Taking into account the large number of stakeholders involved, the far-reaching strength of their lobbies and information asymmetry, it is of paramount importance that the Portuguese State shall be advised by an independent and professional negotiation team. The proposed composition of such a team is as follows:

- Leader- Expert in negotiation- Expert in finance with experience in the electric sector

The negotiation team is to be supported and coordinated by the Ministry of Economy and Employment / Secretary of State for Energy. The approach to be used must be adequate to the specific issues of different generation regimes, namely the complexity (much higher in CMEC and PPAs than in PRE) and the number of stakeholders involved (hundreds in PRE; one in CMEC and PPAs).



APPENDIX 1

BENCHMARK ON RATES OF RETURN FOR GENERATION ASSETS IN PORTUGAL, CAMBRIDGE ECONOMIC POLICY ASSOCIATES, LTD, JANUARY 2012



BENCHMARK ON RATES OF RETURN FOR GENERATION ASSETS IN PORTUGAL

MAIN REPORT

January 2012

CONFIDENTIAL

Cambridge Economic Policy Associates Ltd



OUTLINEThis report has been prepared by CEPA for the Government of Portugal. It provides an analysis of rates of return for particular generating assets in Portugal. The analytical study was carried out in an extremely compressed timeframe, which heavily influenced the approach taken and the scope of the work.

The focus of the analysis has been on the PPAs signed in 1993/4 with private developers and in 1996 with EDP, and the subsequent conversion of the EDP PPAs to the CMEC regime in 2007. Our analysis of the risk profile of these assets suggests that they are low risk generating assets which can be compared to European regulated network utility risk. At the request of the Government, we have focused on returns over the period 2000 - 2010.

We have also commented on returns to European renewables, but due to time restrictions this review has been based on a limited review of published sources and the outputs of our comparative analysis for Portuguese generators.

Our approach to calculating the return has been to calculate a generic integrated electricity/generation WACC to provide a benchmark against which the PPA/CMEC returns can be compared. The risk-free rate can be computed from government bond data, whilst the calculation of the Equity Market Premium is more complex, as there is no agreed approach – we have therefore relied on published works and regulatory decisions. Asset betas can be derived from observed equity betas and gearing for our data set. For the cost of debt, we have used the debt premium from our data set and combined that with the risk-free rate.

The results of our analysis show that risk-free rates in Portugal fell steadily until around 2007. Since then they have risen sharply, as is well reported. The Equity Market Premium was considered by many to be relatively high in the early 1990s, before becoming more commonly treated as around 5%. Corporate tax rates were also relatively high in the 1990s. These factors drove a relatively high real pre-tax WACC in the early 1990s which fell sharply such that by 2001 it was heading below 6% and it remained below that level for an extended period. Although increasing in 2007, it was at that time likely to have been in the 5.5%- 6% range. Headline rates have risen again since then, but a review of the appropriate return for generating assets located in Portugal at a time of sovereign stress has been outside of the scope of this work.

3

IMPORTANT NOTICE

This report has been commissioned by the Ministry of Economy and Employment. However, the views expressed are those of CEPA alone. CEPA accepts no liability for use of this report or any information contained therein by any third party. © All rights reserved by Cambridge Economic Policy Associates Ltd.



CONTENTS

1. Introduction............................................................52. Structure of Generation in Portugal and Risk..........6

2.1. Background..............................................................................62.2. Summary of risk profile..........................................................102.3. Analyst views on risk..............................................................10

3. Approach to Building a Benchmark.......................113.1. Overall approach....................................................................113.2. The basic benchmark.............................................................123.3. Additional aspects of the approach........................................17

4. Our Benchmark.....................................................184.1. Inflation..................................................................................184.2. The Cost of Equity..................................................................194.3. The Cost of Debt.....................................................................264.4. Tax..........................................................................................274.5. The WACC..............................................................................28

5. Renewables............................................................32

5

1. INTRODUCTION

Cambridge Economic Policy Associates Ltd (CEPA) has been engaged by the Government of Portugal to advise on efficient levels of cost of capital for regulated generators, with a particular focus on the Power Purchase Agreement (PPA)-backed generators and the subsequent Costs for the Maintenance of Contractual Equilibrium (CMEC) scheme. At the request of the Government, this advice focuses the time period from 2000 to the end of 2010.

The timeframe for this analytical work has been extremely compressed – the project kicked-off on 6th January, a full draft report was produced for the 20th January and the final report on 27th January. Our understanding is that the timetable has in large part been driven by requests for information from the Troika (ECB, EU and IMF).

The compressed timetable has necessarily required an approach based on publicly available data, so for example there has been no real opportunity for discussions with the investment community (and those active in this community over the past decades) and limited opportunity to cross-check our approach with alternative approaches. Notwithstanding these limitations, we have, with the support of the Ministry of Economy and Employment and its advisers in our view been able to produce an analysis of returns that may be helpful to the Government in assessing the appropriate historic efficient returns for the generators in question.

The rest of this report is set out as follows:

Section 2 provides some background on the relevant generating assets in Portugal and assesses the level of risk in those assets and the relevant comparators.

Section 3 sets out our approach to building a benchmark for the Portuguese generating assets.

Section 4 provides the results of our analysis of benchmark returns.

Section 5 provides brief commentary on returns for renewables.

The report is supported by a number of Appendices, namely:

Appendix 1: Comparator Companies.

Appendix 2: WACC and Beta Estimates – Supporting Analysis.

Appendix 3: Alternative Views on Risk-free Rate and ERP.



7

2. STRUCTURE OF GENERATION IN PORTUGAL AND RISK

2.1. Background

Power Purchase Agreements (PPAs) were signed in 1993 and 1994 for one existing coal plant and one greenfield CCGT by private developers and then in 1996 for five thermal plants and 26 hydro plants belonging to EDP. We understand that most of the EDP plant was already operational in 1996. The offtaker was REN, at that time the wholly state owned transmission company belonging to EDP group, which was also wholly public at the time, but was on a path to floatation. This information can be seen in Figure 2.1 below.Figure 2.1: Timeline for significant share sales in EDP & REN



Redes EnergeticasNacionais (REN)

Power Purchase Agreements signed for 26 hydro & 5 thermal plantsEDP Public share: 100.0% REN Public share: 100.0%

2005/2006

1994

2012

2000

1997/1998

2004

Power Purchase Agreements terminated (July 2007). CMEC implemented.

EDP Public share: 51.5%




EDP Public share: 0%

First three phases of re-privatisation occur

Further share capital sold in 4th phase

State purchases large stake from EDP

REN Public share: 79.3%

State ownership further reduced

EDP sell half of their stake in the company

Two more stages of re-privatisation occur



Fully privatised company as Three Gorges Group buys

remaining state stake

State sells 40% of 51% REN stake

Energias de Portugal (EDP)

Following the publication of Decree-Law n.º 240/2004 on January 27, 2005, agreements were signed for the early termination of the PPAs held by EDP’s generation plants (the PPAs with the private developers remained in force). The Decree-Law established that in order to maintain the contractual equilibrium of the PPAs, the holders of such agreements, which included a significant portion of EDP's generation capacity in Portugal, had the right to receive compensation for the early termination of those agreements (CMEC). Termination depended on certain pre-conditions being fulfilled, including the launch of the spot electricity market at the Iberian level (MIBEL), which came into effect on July 1, 2007, although the relevant agreement for the EDP PPAs was signed on

9

15 June 2007. Generation plants operating under CMEC sell electricity directly on MIBEL i.e. are not dispatched by REN. However, plants have to match “Valoragua’s” simulations in terms of (quantity of) energy placed on the market. For the two remaining PPAs, R.EN Trading makes offers for these plants’ output.

We understand that the CMEC regulation set out the amount of initial compensation for the termination of the PPAs and at the same time it was established that EDP would pay an additional sum in relation to the extension of its rights to use public hydro resources for an average period of over 26 years.

Thus under the CMEC programme, each EDP generation plant that was under a PPA with REN was treated as follows:

Intention was to keep each plant ‘whole’ as per its PPA with REN.

The underlying remuneration of assets remained unchanged at 8.5% real pre-tax cost of capital and all operating costs (fuel, CO2, other variable and fixed) continue to be passed through to the consumer tariff.

The incentive scheme to promote power availability remained unchanged: we understand that this incentive regime was a source of additional profit for the generators.

We also understand that in order to calculate the termination payment, each plant was reviewed and the NPV of future cash streams computed, but that the NPV calculation and subsequent calculation of annuity payments gave a further and significant financial uplift to EDP.

We also understand that EDP may have acquiring rights to extend the use of hydrological resources at a price which created a further financial gain for EDP, and that the use of correction factors in hydro and coal generation also mitigates the risk of EDP

Table 2.1 below summarises how individual items are dealt with under the PPAs/ CMEC. This summary is based on a review of a private developer’s CCGT PPA and discussions with the Ministry.



Table 2.1: Risk and its treatment under the Portuguese PPAsRisk Treatment under

Portuguese PPAs/ CMEC

‘Standard’ international treatment

Comment

Offtaker credit-worthiness

Offtaker is REN, which is state-backed regulated network and, at the time of PPA signature, belonged to EDP group.

Depends on credit quality – enhancements are sometimes needed e.g. guarantee/ letter of credit.

Currently some comments from analysts about Iberian system credit risk and pressure on tariffs.

Fuel Water: notional charge but pass through. Pumping costs (for pump storage) passed through.Gas: 100% pass through of invoices for gas purchased under the Gas Agreement. REN/consumers carries take-or-pay risk.Coal: the generator is paid for 80% based on an annual contract for forecast consumption, which allows the generator to buy forward. 20% is required to match or beat the Platts API2 index. Train operator has take or pay agreement with the PPA coal plant, but for 50% of annual output.Some exchange risk as payments to generators based on average monthly currency values which might differ to spot price.

Index for thermal plants.

Coal: Portuguese generators import coal from the world market, so low fuel cost risk. Some notional take or pay risk with train operator.Coal typically runs baseload apart from very wet years.

Hydrological

No risk, as the CMEC ‘revisibility mechanism’ adjusts the generation of the power plants to the hydrological conditions.

11

Risk Treatment under Portuguese PPAs/ CMEC


Comment

O&M Fixed O&M: Based on pre-defined values, updated monthly using five indices (soft steel, diesel, labour cost, consumer prices, consumer prices other goods).Variable O&M: Based on pre-defined values, updated monthly by consumer price index excluding housing (CPI).

Often fixed ex ante with inflation index and / or periodic (3 – 5 year) re-openers for union contract renegotiation..

Some protection from changing operating regime as variable O&M based on unit values.

Project costs

Capacity payments were fixed for most plants when post construction.

Owner usually negotiates EPC before signing PPA.Recovery of capital costs typically linked to availability targets. Also, in a “pre-financed” plants, typical to reflect actual debt financing.

Debt costs For existing plant, no gearing at project level, but we understand that EDP did gear up at corporate level.For new plant, some pass though / link to interest rate movements.

Owners’ risk – owner would hedge to extent possible.

See above.

Availability For existing plant, additional payments for outperformance and targets appear low (unchanged from original targets).

Straight leverage onto recovery of fixed costs.



Risk Treatment under Portuguese PPAs/ CMEC


Comment

Currency Original contracts in Esc converted into Euro at prevailing rate.

Hard currency capacity payments, or only portion in a ‘soft’ currency.

Portion of currency risk removed through treatment of fuel as effective pass through.

Carbon/ emissions

Generators granted a ‘free’ allowance for five years and emissions costs treated as pass through. Excess allowances are sold and go to the system. Allowances will be paid for from 2013 but costs passed through. There is an incentive to reduce cost with swaps of 10% of EUA allowances with CEER allowances (50% of the cost reduction).Any investment to meet emissions standards would be treated as pass through.

Unclear. We understand that the additional income from these allowances is not significant.

Termination

In some of the thermal plants, REN is responsible for site remediation.

Other We understand that dispatch risk is effectively with REN.

2.2. Summary of risk profile

Based on the above review of risks, it would seem that the PPAs are a low risk asset class, closer to a regulated utility/ regulated network return. The main residual risk would be in a failure to operate e.g. due to fire, against which insurance is available.

This low risk is further evidenced by the fact that it is understood that the private developers geared their coal plant at high levels, whilst EDP over time increased its gearing at the group rather than project level as would be expected. This is illustrated in Figure 2.2 below.

13

Figure 2.2: EDP gearing

Comparators

EDP

0%

10%

20%

30%

40%

50%

60%

70%

1994 1996 1998 2000 2002 2004 2006 2008 2010

Debt

/(De

bt +

Equi

ty)

Source: Bloomberg. Chart also shows gearing for comparators, which will be examined further below.

2.3. Analyst views on risk

Much of the evidence we present below is quantitative evidence supporting the calculation of a reasonable cost of capital for the generation business enjoying the benefit of the long-term PPA agreements. However, it is also relevant to consider the market perceptions of EDP. As can be seen in Figure 2.1., EDP was not listed at the time of the signature of the PPAs in 1996 as the IPO was in 1997, so evidence of market perceptions relate to a later period, but we believe these are relevant. To obtain market views we have had conversations with a limited number of analysts covering the European utilities at that time.

European legislation forcing the opening of electricity markets came into force in February 1999, but by the time of the EDP flotation was already law. This placed threats to the earnings of the large liquid utility companies which were listed at that time. In Germany, the threat to prices appeared to be severe, with no recognition of the need to cut costs, putting pressure on the ratings of the utility conglomerates RWE, VEBA and VIAG. In Spain, the threat was less severe, but the government was overseeing a substantial reform of the sector with the introduction of a wholesale electricity market and new forms of regulation for the infrastructure segments. Valuation of generation assets in these markets required an estimate of a future wholesale market price in an unknown competitive environment.



The environment in Portugal was, and was seen to be, very different. EDP had not been restructured into smaller companies as was considered in the early 1990s, but was dominant in its market. The competitive threat would be low even without the PPAs. But the PPAs were seen to provide earnings security for many years for the generation assets. The attraction of EDP to investors was that the revenues were secure, and that value could be created for the shareholders from cost cutting. The continued role of the government in the shares was seen to provide added security.

As a result of these factors, EDP was seen as an integrated regulated utility, with correspondingly lower risks than the other countries where there were greater competitive threats.

The replacement of the PPA agreements with CMECs was intended to preserve the value of the generation assets (i.e. not change returns). However, issues about the passing on of the tariff deficit in Spain to consumers and potential risks of this in Portugal meant that the cost of capital for EDP was seen to be similar to those of Spanish companies.

3. APPROACH TO BUILDING A BENCHMARK

Establishing a benchmark required rate of return or weighted average cost of capital (WACC) for generation companies requires estimates of:

The required rate of equity return/cost of equity (COE);

The required rate of debt return/cost of debt (COD); and

The mix of debt and equity in the capital structure – also referred to as the level of gearing.

It is also important to specify whether a post- or pre-tax estimate is being derived.

This section of the report provides an overview of the approach we have adopted to establishing our estimate of the benchmark WACC – a series of appendices provide more detailed information on the approach adopted and the data used in deriving the estimate. Our estimate is reported in the following section and includes a high-level discussion of the results.

3.1. Overall approach

When determining a benchmark for the WACC of the PPAs in the power sector in Portugal there are two choices of approach that could be used, specifically:

15

a WACC based on the observed rates demanded for PPAs that had the same structure and risk as those in Portugal could be developed; or

a generic WACC could be developed for (i) pure generators, (ii) good comparators, (iii) integrated utilities, (iv) transmission utilities, and (v) renewable energy companies only.

While the former approach would provide a clear benchmark against which the Portuguese PPAs could be measured the following problems arise:

the level of information available on the WACC (both the headline rate and the true expected rate which incorporates additional returns) for each PPA is likely to be limited and also confidential;7

even if the information is available, the number of PPAs signed in any one year is unlikely to be significant, so meaning that the pool of observations for any one year is limited; and

within this pool, the PPAs which match the risk profile, maturity and overall structure of those against which the benchmark is being compared are likely to be even more limited.

Consequently, while this approach should yield a good benchmark against which the actual PPAs can be measured the reality is that implementing the approach would be, at best, difficult if not impossible.

As an alternative, it is possible to consider the development of a more generic measure of the WACC for generation companies, based on those companies that are listed (for the equity measure) and raising bond finance (for the debt measure). Obviously this approach also has its limitations, namely:

the comparability of listed companies which may be operating a portfolio of plant types, under a range of different contractual/ competitive arrangements in one or more countries;

the majority of generating companies are likely to be part of larger entities that tend to be integrated energy companies; and

a number of mergers and acquisitions have taken place in the energy sector since the late-1990s).

7 Additional returns refer to the types of payment that might be expected to arise from setting targets in the PPA at a level which any company would expect to beat – say a low availability target that should allow the company to earn abnormal returns without operating above normal levels of efficiency.16



Clearly these concerns are limitations to the approach, but the practical reality is that this approach is at least implementable. However, as discussed later in this report, the way in which the results are interpreted will be important owing to the differences in the average risk of the generation portfolio and the specific Portuguese PPAs.

3.2. The basic benchmark

Given the general approach outlined above, we have adopted a standard approach to estimating the benchmark and its constituent elements. Each of these are discussed in turn.

3.2.1.The cost of equity (COE)

Our approach to estimating the required COE is based on the Capital Asset Pricing Model (CAPM), a standard corporate finance approach to addressing this question. The CAPM links the required rate of return to three factors:

the general risk-free rate in the economy;

the level of additional return required for holding the entire “risky” portfolio of assets in the country (the equity or market risk premium – ERP or MRP); and

the relative risk exposure of the specific asset or asset class compared to the whole portfolio (the equity beta).

These elements are combined as:8

Re=R f+βe (Rm−Rf )

Where: Re is the return on equity;

Rf is the risk-free rate;

ße is the equity beta;

Rm is the market return.

Each of these elements is now discussed in turn.

Risk-free rateThe risk-free rate is an estimate of the underlying cost of riskless borrowing in an economy. It is normal to proxy this rate through the use of government borrowing rates. Of course, government borrowing rates are not riskless – apart from the general risk of the government there is

8 These should be presented as expected values but for simplicity the equation has been simplified.

17

risk of unexpected inflation affecting nominal bonds. Whilst index-linked bonds are a partial solution to this there use is limited across Europe. Government bonds are, however, normally the least risky asset in an economy and, as such, typically the relevant benchmark against which riskier assets are measured.

The choice of the appropriate government security to act as the risk-free rate depends on:

the availability of different maturities;

the frequency of trades and size of the bonds;

whether nominal or index-linked bonds are issued; and

the time horizon of the investment.

For countries with a significant number of government bonds there are indices of specific maturity developed – say a constant five year or 10 year maturity. With these indices individual bonds are included while their maturity is within the band of the constant maturity and then dropped once it is outside the band (so the five year bond may include bonds with remaining maturities between three and a half and six and a half years). We have focused on these sort of Bloomberg index but also considered some individual bonds.

As discussed later, we use several different maturities to provide a range of time horizons for the WACC calculation.

Equity/market risk premiumWhile the principle behind the ERP or MRP (henceforth referred to solely as the MRP) is simple – as noted above, it is the additional return demanded by investors to hold the whole “risky” portfolio in a country – the measurement has proven a subject of intense academic debate. The basic problem arises from the observed values for the MRP, which are measured by comparing the returns on the market with returns on risk-free assets.

Figure 3.1 below summarises selected regulatory decisions and our view of the broad consensus regarding the MRP over time. In the 1990s, worldwide evidence pointed to a MRP in the broad range of c4% to 7%, or even as high as 9%. Many regulatory decisions on the WACC in the late 1980s and early 1990s used these figures.9

9 While there are some differences between the situation faced by generators to standard regulated utilities, regulatory determinations are a useful public source of views on elements of the WACC.18



Figure 3.1: Consensus range and regulatory decisions on MRP

0%

2%

4%

6%

8%

10%

12%

1990-95 1996-2000 2001-05 2006-10

Mar

ket r

isk p

rem

ium Hong Kong (1990s)

Oftel (1994)

MMC (1993)

Maine Public Utilities Commission (1999)

Oftel (1997)

CC (2000)Ofgem (1999-2001)

Oftel (2001)

ACCC (2003)AER (2011)

Ofgem (2006)

AEEG (2005)

CC (2010)

Source: CEPA research. The shaded area represents the approximate range of the consensus view on the MRP. We recognise that alternative views have existed outside this range – as demonstrated by regulatory decisions in Hong Kong in the 1990s. The selected regulatory decisions are intended to indicate the range of views taken by worldwide regulators and are not intended to be comprehensive. Several academics posed the question of how an MRP as high as 9% could be justified from a theoretical perspective when other approaches justified much lower values. This led to an intense debate about why using observed values to estimate the MRP provided misleading values. This focused specifically on the:

use of out-turn values rather than expected values;

problems of survival bias in equity indices;

problems of unstable ratings periods such that like-for-like calculations are not being undertaken; and

relative impact of unanticipated inflation on equities and bonds.

19

This analysis has led to alternative values for the MRP being developed. The consensus view is now arguably lower (and narrower), as reflected in Figure 3.1.

Equity betaThe final element of the CAPM calculation is the relative risk of the asset/company, as captured by the equity beta. When measuring the equity beta the risk that matters is referred to as market or non-diversifiable risk. This is because investors are assumed to have an efficient portfolio of assets which allow some risk to be mitigated through the ownership of assets that benefit from the risk that is negatively affecting another asset. As such, the only risk that matters is that which cannot be diversified away through the creation of an efficient portfolio.

The market as a whole has an equity beta value of 1 and if a company has an equity beta of less than 1 then it is less risky than the market as a whole and if the value is greater than 1 then the company is more risky. Two factors drive the equity beta value for a company:

underlying business risk reflecting the non-diversifiable risks of the business relative to the market as a whole (measured through the asset beta); and

financial risk which magnifies the underlying business risk’s effects on equity caused by the increasing use of debt. In principle this should be measured against the average gearing (financial structure) of the market as a whole, however, this is normally assumed to be constant and it is just the actual level of gearing for the asset that is considered.10

3.2.2.The cost of debt (COD)

The COD is traditionally measured as two separate elements:

the risk-free rate (described earlier); and

the debt premium.

The debt premium is a measure of the additional risk faced by investors for lending money to a company rather than the government. It is estimated by comparing the yield to maturity of the corporate bond with the yield to maturity of the appropriate government bond. This spread is then the debt premium.10 Clearly this assumption is not appropriate when considering long time series of data. However, we have continued to use the standard assumption for this work given the limitations faced with respect to time, data etc.20



When choosing government comparator bonds it is important to consider:

the maturity of the bond; and

the coupon.

Maturity is the most important determinant of the yield and so ensuring comparator bonds have as close a maturity to the corporate bond as possible is key. However, if there is then a choice concerning coupon it is better to find a coupon similar to that of the corporate bond since tax clientele effects might lead to different levels of demand and so yield.

3.2.3. Gearing

When calculating the WACC it is important to establish the capital structure of the business, i.e. the mix of debt and equity (referred to as the gearing). This plays several roles in the WACC calculation as it is used to:

de-lever equity betas into asset betas (and then re-lever them); and

weight the COE and COD in the WACC calculation.

The standard approach to measuring the level of gearing is:

Gearing= DebtDebt+Equity

Whilst there are preferred ways of measuring the debt and equity values, given the scale of the analytical work and the compressed timetables for the project we have relied upon the value that Bloomberg provides in order to de-lever the equity betas. We have re-levered the resulting asset betas based on our assumption regarding the appropriate notional capital structure.

3.2.4.Taxation

Both corporate and individual taxes can have an influence on the WACC:

corporate as differential treatment of debt and equity returns often exist, with debt interest being tax deductible; and

personal since dividend and interest payments received by investors are treated as taxable income.

Consequently it is important to be clear as to the pre- or post-tax status of the estimate. It is normal to estimate values that are either pre- or post-corporate tax while assuming the values are pre-personal taxes. The standard approach to doing this is to:

21

estimate a post-corporate tax value from observable information; and

make an adjustment based on the corporate tax rate to determine the pre-tax rate.

Whether the tax adjustment should be made to just the COE or the whole WACC is a point of disagreement. Our approach allows for any of the options to be considered.

3.3. Additional aspects of the approach

In the sub-sections above we have considered the overall approach that we will adopt to estimating the benchmark. There are two further issues that need to be considered when determining an estimate.

3.3.1.Time

The time horizon for an estimate of the WACC is important. In principle the WACC should be calculated for the life of the asset under consideration. In the case of the PPAs in the Portuguese power sector there are different lives arising because of:

the commissioning date of the various plant which already existed; and

the differences in lives dependent on the technology/ fuel source employed by the plant.

Given this it is important to consider what elements of the estimate might change as the time horizon changes. There are two key elements:

the risk-free rate; and

the debt premium.

It is normally expected that there is an upward sloping yield curve, i.e. borrowing longer maturity funds is more expensive than short. This is because of the time value of money, risk etc. As such, we would expect the long maturity WACC to be higher than a short maturity one, ceteris paribus. Of course, there may be periods when short-term money is more expensive than long maturity money, owing to short-term risks – the current financial crisis is an example of this. That can change the slope of the yield curve.

The debt premium is also expected to increase the longer the maturity of the bond owing to the greater risk compared to the government bond. As such, a long maturity WACC should be higher than a short-maturity one, 22



ceteris paribus. Anecdotal evidence in the 1990s suggested that every additional year of maturity added 2 basis points (0.02%) to the debt premium. However, it is also possible that if the maturity impact on the risk-free rate is significant then the debt premium may adjust for this and be lower for long dated debt (but still having an overall long dated COD greater than a short dated one).

To take account of the possible impact of maturity we calculate three maturities of WACC:

Short-dated (under 10 years);

Medium-dated (between 10 and 30 years); and

Long-dated (above 30 years).

In our main conclusions we match the maturity of each PPA or CMEC agreement to the maturity categories above.

3.3.2.Fuel type

It is possible that the choice of fuel type will affect more than just the maturity of the plant, especially if the exposure to risk is different. The most obvious example of this would be the comparison of thermal plant and hydro-electric ones. The latter are unlikely to face fuel cost risk while this can be important for thermal plant. Of course, whether the thermal plant faces a different level of risk will depend on the structure of the contract under which the power is being sold. For the PPA-backed generators in Portugal, our analysis indicates that you would not expect to see significant differential pricing according to fuel type.

Unfortunately, given the time and data limitations no meaningful information on the impact of technology has been discernible from our analysis.

4. OUR BENCHMARK

Having outlined our approach, this section reports the results of the data analysis that was undertaken. The length of the time series of WACC estimates required has created some specific data issues that are discussed in the appendices that support the section.

The basic source of data for this analysis has been Bloomberg. However, this has been supplemented in many cases, and again the additional data is discussed in the appendices.

23

We have developed a long-list of potential comparator companies (see Appendix 2). However, we have had to exclude some companies from this list due to lack of comparability or data issues.

4.1. Inflation

As we are seeking a real WACC, we need to deflate our nominal data. We have done so using inflation rates for Portugal based on OECD data (see Figure 4.1 below). In calculating forward-looking measures (e.g. estimates of the risk-free rate) we have used OECD forecasts as reported in the World Economic Outlook for the preceding year. All calculations involving inflation have been carried out based on the Fisher equation.

Figure 4.1: Inflation data for Portugal



OECD-Outturn

OECD-Forecast

ECB target

-2%

-1%

0%

1%

2%

3%

4%

5%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Year

-on-

year

infla

tion

Source: OECD11

4.2. The Cost of Equity

As discussed previously, three elements need to be considered:

the risk-free rate;

the MRP; and

the equity beta.

4.2.1.Risk-free rate

Figure 4.2 reports the different maturities of the nominal risk-free rate for Portugal. This is based on the yield to maturity of the various maturity bonds.

11 OECD outturn inflation based on harmonised CPI data from stats.oecd.org. OECD forecast inflation based on estimates for year-ahead Portuguese CPI and harmonised CPI data contained in annual World Economic Outlook reports from 1992-2010. ECB target assumed to be 2.0% for 2000-10.

25

http://stats.oecd.org/

Figure 4.2: Nominal yields to maturity on 5-, 10- and 30-year Portuguese government bonds

10 year

5 year

30 year

0%

1%

2%

3%

4%

5%

6%

7%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Nom

inal

bon

d yi

eld

Source: Bloomberg, OECD, CEPA analysisUsing these nominal yields and the inflation information previously provided allows us to estimate the real risk-free rate. This estimate is not perfect since it does not use a market estimate of the inflation over the appropriate maturity range, but it is the best that can be done with the information available. Figure 4.3 illustrates the resulting real risk-free rates across the different maturities of WACC.Figure 4.3: Real yields to maturity on 5-, 10- and 30-year Portuguese government bonds12

12 Nominal yields deflated based on OECD reported (harmonised) CPI inflation.26



10 year

5 year

30 year

0%

1%

2%

3%

4%

5%

6%

7%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Real

bon

d yi

eld

Source: Bloomberg, OECD, CEPA analysisThe real risk-free rate from in the early 2000’s was relatively low before rising to close to 6% after the credit crunch in 2007. However, it should be noted for context that it was relatively high in the 1994/95 period at around 6%.

4.2.2.Equity/market risk premium

As discussed in Section 3, the MRP has been subject to significant uncertainty owing to the problems implicit in the way that the value used to be calculated. The starting point for our estimation period is the time when views about the MRP were starting to significantly change and so an unthinking adoption of high values would not be appropriate – significant academic and professional debate about the right number was underway. As such, we have set out a range that changes over time (see Figure 4.4 below).Figure 4.4: The MRP range

27

0%

2%

4%

6%

8%

10%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Equi

ty ri

sk p

rem

ium

ove

r gov

ernm

ent b

onds

Range Base case

The upper value (6.5%) is towards the values used by some regulators in the mid-1990s (and interestingly regulators in Portugal and Australia have been increasing rates again to similar levels). The lower bound (4%) reflects the values used consistently for most of the 2000s. As shown in Figure 3.1, arguably the range of commonly-used values has fallen and narrowed over time, and we reflect this in our assumptions. Our base case is 5.0% for the 2000s.

4.2.3.Equity betas

Our estimate of the equity beta has used several steps:

1. estimate the equity beta of generation companies (which has needed to include integrated energy companies to ensure a sufficient sample size);

2. estimate the asset beta by de-gearing the equity beta estimates; and

3. estimate a generation/integrated company benchmark using the asset betas and a notional level of gearing.

The raw equity betas for our sample of integrated utility companies are presented in Figure 4.5 below.Figure 4.5: Raw equity betas for integrated utility companies



0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Jan-2000 Jan-2001 Jan-2002 Jan-2003 Jan-2004 Jan-2005 Jan-2006 Jan-2007 Jan-2008 Jan-2009 Jan-2010 Jan-2011

Range Interquartile range Median Annual assumption

Source: Bloomberg, CEPA analysisThe following sections summarise our approach to generating asset betas.

4.2.4.Gearing

Figure 4.6 provides evidence from our sample of companies on the level of gearing (taken from Bloomberg and defined as long and short term financial debt as a proportion of debt and equity).Figure 4.6 Gearing for comparator companies

Comparators

EDP

0%

10%

20%

30%

40%

50%

60%

70%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Debt

/(De

bt +

Equi

ty)

Our analysis of gearing shows that:

29

For the integrated utility companies in our sample, average gearing rose in the early 2000s as high as 50%, but has since fallen back to around 40%; and

Our good comparators have sustained higher average gearing levels, at around 50-60% since the early 2000s, suggesting that they are viewed as slightly lower risk than the broader sample.

We note that there is a significant amount of variation in actual gearing rates chosen by companies in our sample (see Figure 4.7 below).Figure 4.7: Range of gearing for comparator companies

0%

20%

40%

60%

80%

100%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Debt

/(De

bt +

Equi

ty)

Range Interquartile range Average

Source: BloombergWe therefore consider it appropriate to show a range of values under low and high scenarios (see Figure 4.8 below). Figure 4.8: Notional gearing assumptions



0%

20%

40%

60%

80%

100%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Debt

/(De

bt +

Equi

ty)

Range Base case

In our base case we take the view that the PPAs would attract a relatively high level of gearing owing to their relatively low level of risk and show and increase over time as investors became more familiar with these assets. However, in all scenarios we assume that gearing begins to decline slightly in 2007.

4.2.5.Asset betas

By de-gearing the equity betas (using the information outlined above) we have been able to estimate asset betas – reflecting the underlying business risk rather than a mix of business and financial structure risk (as captured in the equity beta). Figures 4.9 and 4.10 illustrate the results of this analysis. In the WACC estimation that follows, we use an annual estimate for each year; this value is labelled ‘Annual’ in the figures below (and in appendix figures), and represents the value based on evidence up to 1 January of each year.Figure 4.9: Asset betas: integrated utilities

31

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4



Figure 4.10 Asset betas: good comparators



0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9



Our analysis of asset betas shows that:

Asset betas fell to very low levels of around 0.2 in the early 2000’s before returning to around 0.4 on average in the late 2000s.

Our good comparators also have relatively volatile asset betas, declining sharply to 0.1 in the early 2000’s and then more recently (post 2007) stabilising at around 0.3-0.4. The evidence suggests they were then seen as lower risk or on average no more risky than the wider sample.

The central figures shown above are median values for each group of companies. We also considered the mean across companies (either a simple mean or weighted by market capitalisations). The mean values were more volatile, with the peaks around 2007 being substantially more pronounced. They were also frequently outside the interquartile range – higher, therefore than at least 75% of our comparator companies.

There are strong arguments to suggest that the degree of systemic risk faced under the PPAs would be towards the low end of that faced by many of the comparator businesses. Since it has not been possible in the time available to gather evidence for purely PPA-backed businesses, we have been forced to use our judgement to select an appropriate range. We consider it likely that the (relatively high) mean betas would be at the upper end of what would have been required by private investors into the PPA-backed assets. As a result, we use the median values as our base case, and refer to high and low cases based on the interquartile range (see Figure 4.11).

33

Figure 4.11: Asset beta assumptions

0.0

0.2

0.4

0.6

0.8

1.0

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Range Base case

4.3. The Cost of Debt

4.3.1.Risk-free rate

The risk-free rate for the cost of debt is of course taken to be identical to the risk-free rate shown in Section 4.2.1 above.

4.3.2.Debt premium

There is relatively little publicly available information on historic debt premia. For the early 2000s we have therefore relied on several sources for spread at issue for electricity company bonds. From the early 2000s onwards we have used daily spread estimates for electricity bonds based on secondary trading data. These are reported by Bloomberg and summarised in Figure 4.12 below.Figure 4.12: Spreads on utility company bonds vs. benchmark government bonds



Short (<10y)

Medium (10-30y)

Long (>30y)

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Spre

ad o

ver b

ench

mar

k go

vt. b

ond

Source: Financial Times, Bloomberg, CEPA analysis. We include only European and US bond issuances. We exclude six issuances by AES in 2001/02 due to the specific difficulties faced by Drax during that period. Data were available for all years for medium term bonds. Where data were unavailable for short or long term, we interpolated values based on data for the closest available year.Whilst there has been significant movement in bond spreads over the period, the medium term spread has been much more stable. This is likely to be in part due to the higher quality of available data for medium term bonds. Our assumption regarding the debt premium is driven by the asset life under consideration; where appropriate, however, we focus on the medium term as a ‘base case’ given the higher data quality.

4.4. Tax

We use the headline tax rate to compute a pre-tax cost of equity and thus a pre-tax WACC (cost of debt data is already pre-tax). We are advised that both the corporation tax rate and the municipal tax rate were applied to profits and as such use a combined (and by definition higher) rate to gross up the cost of equity.Table 4.1: Tax rates

2000-01 2002-04 2005-08 2009-10Total tax rate 35.2% 33.0% 27.5% 26.5%

Note: The total tax rate is a combination of the corporation tax rate multiplied by municipal tax at 10%

35

4.5. The WACC

This section presents our conclusions for the overall WACC. We also present further detail on how our estimates change over time, as well as an illustration of the relevant parameters for a specific scenario (the medium term base case). Note that the yearly data points are the outputs of our modelling, and should not necessarily be interpreted as individual point estimates.

It is based on the ranges shown above for each of the building blocks. Our estimates cover each year from 2000 to 2010, but the table below focuses on the appropriate comparator cost of capital at the time of the CMEC agreements in 2007. More detailed tables showing the detailed figures and building blocks that underpin each of the scenarios are available in Appendix 3, along with sensitivity and robustness checks.

Our estimates in Table 4.2 below are specific to a point in time (i.e. year) and time horizon (i.e. remaining asset life – short, medium or long). To capture the latter we adjust the time horizon used for our evidence on the risk-free rate and the debt premium. We do not consider we have evidence to justify calculating a rate that is specific to the technology concerned. In our view these rates are likely to have been very similar. We acknowledge that a range of plausible values exist. As well as our base case, we present low and high estimates based on alternative assumptions for gearing and beta.Table 4.2: WACC estimates for PPA/CMEC agreementsContract

Year Remaining asset life

Base case Low High

CMEC 2007 Short (<10 yrs) 5.4% 4.4% 6.2%CMEC 2007 Medium (10-30 yrs) 5.8% 4.8% 6.5%CMEC 2007 Long (30+ yrs) 5.9% 4.9% 6.7%

We are aware that the Ministry is particularly interested in the results of our analytical work for the year when the CMEC regime came into operation (2007). Our focus has been on estimating the returns that would have been required by private investors at the time. An important cross-check for this is the viewpoint of regulators: typically, WACCs are set by regulators or within contracts based on a view of the recent past and likely changes to rates over the period, i.e. a fixed allowed return is informed by past events and the regulator’s expectations of any changes in market conditions, plus the need to ensure that the company can finance itself over the period.



In 2007, we show for the ‘medium asset lives’ base case a pre-tax real WACC of 5.8%, based on data for that year. At this point in time our yearly WACC data was on an upward trend, but we note the Regulator allowed a pre-tax nominal WACC of 8% for EDP or about 5.2% real. We therefore believe that our 2007 base case estimate of the WACC is not unreasonable.

WACC over timeFigures 4.13, 4.14 and 4.15 show our WACC estimates over time for each category of asset life.

Figure 4.13: Short term (real, pre-tax)

0%1%2%3%4%5%6%7%8%9%

10%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Range Base case

Notes:Comparator companies are European integrated utilitiesReal risk-free rate based on 5 year nominal Portuguese government bond yields to maturity deflated based on OECD forecast inflationDebt premium based on <10 year corporate bond spreads over national government bondsRange based on alternative estimates for asset betas and notional gearing

Figure 4.14: Medium term (real, pre-tax)

37

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%

10%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Range Base case

Notes:Comparator companies are European integrated utilitiesReal risk-free rate based on 10 year nominal Portuguese government bond yields to maturity deflated based on OECD forecast inflationDebt premium based on 10-30 year corporate bond spreads over national government bondsRange based on alternative estimates for asset betas and notional gearing

Figure 4.15: Long term (real, pre-tax)

0%

1%

2%3%

4%

5%

6%7%

8%

9%

10%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Range Base case

Notes:Comparator companies are European integrated utilitiesReal risk-free rate based on 30 year nominal Portuguese government bond yields to maturity deflated based on OECD forecast inflation



Debt premium based on 30+ year corporate bond spreads over national government bondsRange based on alternative estimates for asset betas and notional gearing

39

Illustration of building blocks (for medium term base case scenario only)Table 4.3: Cost of capital breakdown – Medium term base case

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010Notional gearing 50% 51% 53% 54% 56% 57% 59% 60% 58% 57% 55%

Tax 35.2% 35.2% 33.0% 33.0% 33.0% 27.5% 27.5% 27.5% 27.5% 26.5% 26.5%

Inflation 2.2% 2.5% 3.3% 2.8% 2.2% 1.7% 1.8% 2.1% 2.6% 2.2% 1.6%

Risk-free rate (real) 3.3% 2.6% 1.7% 1.4% 1.9% 1.7% 2.0% 2.3% 1.9% 2.0% 3.6%

Debt premium 0.8% 1.7% 1.3% 1.1% 0.9% 0.3% 0.6% 0.8% 1.8% 2.1% 1.6%

EMRP 5.3% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0%

Asset beta 0.25 0.20 0.15 0.21 0.22 0.28 0.30 0.39 0.35 0.32 0.35

Equity beta 0.51 0.40 0.32 0.45 0.49 0.65 0.72 0.97 0.85 0.74 0.78

Cost of debt (real, pre-tax) 4.1% 4.3% 3.0% 2.4% 2.8% 2.1% 2.6% 3.1% 3.7% 4.0% 5.3%

Cost of equity (real, pre-tax) 9.3% 7.1% 4.8% 5.4% 6.5% 6.9% 7.8% 9.8% 8.4% 7.7% 10.2%

Cost of equity (real, post-tax) 6.0% 4.6% 3.2% 3.6% 4.4% 5.0% 5.6% 7.1% 6.1% 5.6% 7.5%

Cost of capital (real, pre-tax) 6.7% 5.7% 3.8% 3.8% 4.5% 4.1% 4.8% 5.8% 5.7% 5.6% 7.5%

Cost of capital (real, vanilla) 5.1% 4.5% 3.1% 3.0% 3.5% 3.3% 3.9% 4.7% 4.7% 4.7% 6.3%



Note: ‘Vanilla’ WACC based on pre-tax cost of debt and post-tax cost of equity.

41

5. RENEWABLES

As part of our analysis, we have been asked to consider any evidence emerging on the returns for renewables, although this has not been a focus of the work.

As part of our data set, we have identified a number of quoted comparator stocks which are, broadly speaking, ‘renewable only’ and the relevant beta analysis and WACC data is presented for them below. We have also undertaken a limited review of published literature on returns to renewables.

The renewable-only companies in our data set are:

Acciona SA

EDP Renovaveis

EDF EN

Enel Green Power

Greentech Energy (note: relatively low weighting in sample)

The asset betas for these companies are set out in Figure 5.1 below. Estimates in the early 2000s are at a very low average level of below 0.2, before rising to a (still low) average of around 0.4 post 2007.

Figure 5.1: Asset beta estimates for renewable energy companies

0.0

0.2

0.4

0.6

0.8

1.0

1.2



Source: Bloomberg, CEPA analysis



Gearing levels for these companies have tended to be above the average for the fuller sample, as illustrated in Figure 5.2.

Figure 5.2: Weighted average gearing for renewable energy companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

All companies Current sample

Source: Bloomberg, CEPA analysisThese factors combine with our standard assumptions for other inputs to provide similar WACC numbers to our main ‘Integrated utilities’ sample. Figure 5.3 below compares the WACC estimate for renewable energy companies with our base case estimate presented in Section 4. Figure 5.3: Base case vs. renewable WACC (pre-tax, real, base case assumptions)

0%1%

2%

3%

4%5%

6%

7%

8%9%

10%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Renewable generators Integrated utilities

Notes: See notes to Figure 4.13.

43



BENCHMARK ON RATES OF RETURN FOR GENERATION ASSETS IN PORTUGAL APPENDICES

January 2012

CONFIDENTIAL

Submitted by:

Cambridge Economic Policy Associates Ltd

45

IMPORTANT NOTICE

This report has been commissioned by the Ministry of Economy and Employment. However, the views expressed are those of CEPA alone. CEPA accepts no liability for use of this report



or any information contained therein by any third party. © All rights reserved by Cambridge Economic Policy Associates Ltd.

47

CONTENTS

Appendix 1: Comparator Companies...........................37Appendix 2: WACC and Beta Estimates – Supporting analysis......................................................................40

WACC estimates.............................................................................40Beta estimates................................................................................61

Appendix 3: Alternative Views on Risk-free Rate and ERP65



APPENDIX 1: COMPARATOR COMPANIES

The starting point of the analysis was the Bloomberg European Electricity Generation Index in the beginning of 2012. To those companies, we have added electricity companies previously listed on the stock market and currently acquired or delisted, as well as other utilities currently listed. The table below gives a brief categorisation of these comparators. Table A3: Comparator companies Company name Description of business Sub-categoriesGAS NATURAL SDG SA

Vertically integrated utility company

N/A

ENERGIAS DE PORTUGAL (EDP)

Integrated electric utility with position in renewables

N/A

ENDESA SA Vertically integrated electric utility Good comparatorIntegrated utility

REN Portuguese TSO and owner of transmission grids

Transmission utility

RED ELECTRICA DE ESPANA (REE) SA

Spain’s TSO and regulated transmission company


A2A SPA Vertically integrated utility company

Integrated utility

ACEA SPA Vertically integrated utility company

Integrated utility

ALPIQ HOLDIN-REG


Integrated utility

ACCIONA ENERGIA SA

Renewable energy power generator

Pure generatorRenewables

BKW FMB ENERGIE

Vertically integrated electric utility Integrated utility

CEZ AS Vertically integrated electric utility Integrated utility

CENTRICA PLC Integrated energy company Integrated utilityDRAX GROUP PLC Coal-fired electricity generation

companyPure generator

EDF Vertically integrated utility company

Good comparatorIntegrated utility

49

Company name Description of business Sub-categoriesENERGIEDIENST AG


Integrated utility

EDISON SPA Vertically integrated utility company

Pure generator

EDP RENOVAVEIS Wind-energy based renewables company

Pure generatorRenewables

ENEL GREEN POWER

Renewable energy generator Pure generatorRenewables

ENEA Vertically integrated utility company

Integrated utility

ENEL SPA Vertically integrated utility companyMajority shareholder in Endesa SA

Integrated utility

E.ON AG Vertically integrated utility company

Integrated utility

EVN AG Vertically integrated utility company

Integrated utility

FORTUM OYJ Vertically integrated utility company

Integrated utility

GREENTECH ENERGY

Renewable energy producer Pure generatorRenewables

GDF SUEZ Vertically integrated utility company

N/A

HAFSLUND-B SHS Vertically integrated utility company

Integrated utility

ROMANDE ENE-REG

Electricity producer and distributor

Integrated utility

IBERDROLA SA Integrated electric utility with large wind power portfolio

Good comparatorIntegrated utility

INTL POWER PLC Independent power generation company

N/A

IREN SPA Vertically integrated utility company

Integrated utility

MVV ENERGIE AG Vertically integrated utility company

Integrated utility

PGE SA Vertically integrated utility company

Integrated utility

PUBLIC POWER CORP


Integrated utility



Company name Description of business Sub-categoriesRWE AG Vertically integrated utility

companyIntegrated utility

SECHILIENNE-SIDEC

Independent power generation company

Pure generator

SSE PLC Vertically integrated utility company

Integrated utility

TAURON Vertically integrated utility company

Integrated utility

VERBUND AG Vertically integrated utility company Integrated utilityHyrdo

UNION FENOSA Vertically integrated utility company Good comparatorIntegrated utility

HIDRO CANTABRICO (HC ENERGIA)

Vertically integrated utility company Good comparatorIntegrated utility

FESCA An Endesa subsidiary Integrated utilityCIA SEVILLANA DE ELECTRICIDAD

An Endesa subsidiary Pure generator

ENEL VIESGO An Endesa subsidiary Integrated utilityGESA An Endesa subsidiary Integrated utilityPOWERGEN Vertically integrated utility

companyIntegrated utility

BRITISH ENERGY Producer of electricity & gasPart of EDF since 2009

Pure generator

NATIONAL POWER Vertically integrated utility company

Integrated utility

NATIONAL GRID Power transmission network company


SCOTTISH POWER Vertically integrated utility company Integrated utilityENERGINET.DK Danish Transmission Systems

OperatorN/A

TERNA SPA Italian Transmission System Operator


EDF ENERGIES NOUVELLES

Renewable energy generatorAcquired by EDF in 2011

Pure generator Renewables

51

APPENDIX 2: WACC AND BETA ESTIMATES – SUPPORTING ANALYSIS

This appendix presents further scenario analysis for our beta and WACC estimates.

WACC estimatesFor the WACC estimates, we present:

tables of the results underlying Figures 4.13, 4.14 and 4.15 in our main report;

detailed tables showing the breakdown of our WACC estimates into their component parts and presenting estimates of different variants of the WACC; and

the following robustness checks of alternative estimation assumptions for our base case:

o using beta estimates generated using different groups of comparator companies;

o using beta estimates based on the Eurotop index as the benchmark for all companies, instead of the relevant local market;

o calculating 5-year rather than 2-year rolling betas; and

o basing the WACC calculation on raw equity betas rather than re-levered asset betas.



Table A4.1: Cost of capital over time for different remaining asset lives2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Short termBase case 6.4% 5.1% 3.1% 2.6% 3.3% 3.4% 4.4% 5.4% 5.3% 4.3% 6.3%Low 5.2% 4.9% 2.6% 2.0% 2.9% 2.4% 4.1% 4.4% 4.7% 4.2% 5.7%High 7.2% 5.4% 4.0% 3.4% 3.9% 3.8% 5.1% 6.2% 5.5% 4.6% 6.5%

Medium termBase case 6.7% 5.7% 3.8% 3.8% 4.5% 4.1% 4.8% 5.8% 5.7% 5.6% 7.5%Low 5.5% 5.5% 3.3% 3.2% 4.0% 3.1% 4.5% 4.8% 5.1% 5.5% 6.9%High 7.6% 6.0% 4.7% 4.6% 5.1% 4.5% 5.5% 6.5% 5.9% 6.0% 7.8%

Long termBase case 8.1% 6.1% 4.3% 4.2% 4.8% 4.5% 5.2% 5.9% 5.7% 5.8% 7.8%Low 7.0% 5.7% 3.6% 3.5% 4.3% 3.5% 4.8% 4.9% 5.0% 5.6% 7.1%High 9.0% 6.6% 5.3% 5.1% 5.6% 5.0% 5.9% 6.7% 6.0% 6.2% 8.0%

Note: These estimates relate to Figures 4.13, 4.14 and 4.15 in the main report.

53

Table A4.2: Cost of capital breakdown – Short term base case2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Notional gearing 50% 51% 53% 54% 56% 57% 59% 60% 58% 57% 55%

Tax 35.2% 35.2% 33.0% 33.0% 33.0% 27.5% 27.5% 27.5% 27.5% 26.5% 26.5%

Inflation 2.2% 2.5% 3.3% 2.8% 2.2% 1.7% 1.8% 2.1% 2.6% 2.2% 1.6%


Debt premium 1.0% 1.7% 1.3% 0.8% 0.7% 0.3% 0.4% 0.5% 1.9% 1.8% 1.6%

EMRP 5.3% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0%

Asset beta 0.25 0.20 0.15 0.21 0.22 0.28 0.30 0.39 0.35 0.32 0.35

Equity beta 0.51 0.40 0.32 0.45 0.49 0.65 0.72 0.97 0.85 0.74 0.78

Cost of debt (real, pre-tax)

4.0% 3.8% 2.4% 1.3% 1.8% 1.4% 2.2% 2.6% 3.4% 2.7% 4.2%

Cost of equity (real, pre-tax)

8.7% 6.4% 4.0% 4.1% 5.2% 6.0% 7.5% 9.6% 7.9% 6.3% 8.8%

Cost of equity (real, post-tax)

5.7% 4.1% 2.7% 2.8% 3.5% 4.4% 5.5% 6.9% 5.7% 4.6% 6.5%

Cost of capital

6.4% 5.1% 3.1% 2.6% 3.3% 3.4% 4.4% 5.4% 5.3% 4.3% 6.3%



(real, pre-tax)Cost of capital (real, vanilla)

4.8% 4.0% 2.5% 2.0% 2.5% 2.7% 3.6% 4.3% 4.3% 3.5% 5.2%


55

Table A4.3: Cost of capital breakdown – Short term low case (high gearing, low beta)2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010


Tax 35.2% 35.2% 33.0% 33.0% 33.0% 27.5% 27.5% 27.5% 27.5% 26.5% 26.5%

Inflation 2.2% 2.5% 3.3% 2.8% 2.2% 1.7% 1.8% 2.1% 2.6% 2.2% 1.6%


Debt premium 1.0% 1.7% 1.3% 0.8% 0.7% 0.3% 0.4% 0.5% 1.9% 1.8% 1.6%

EMRP 5.3% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0%

Asset beta 0.12 0.16 0.07 0.12 0.16 0.14 0.26 0.25 0.24 0.28 0.25

Equity beta 0.32 0.45 0.21 0.37 0.52 0.49 0.96 1.01 0.90 0.98 0.84


4.0% 3.8% 2.4% 1.3% 1.8% 1.4% 2.2% 2.6% 3.4% 2.7% 4.2%


7.2% 6.8% 3.2% 3.6% 5.4% 5.0% 9.2% 9.8% 8.3% 7.9% 9.3%


4.7% 4.4% 2.1% 2.4% 3.7% 3.6% 6.7% 7.1% 6.0% 5.8% 6.8%

Cost of capital

5.2% 4.9% 2.6% 2.0% 2.9% 2.4% 4.1% 4.4% 4.7% 4.2% 5.7%




4.3% 4.0% 2.3% 1.7% 2.3% 2.0% 3.4% 3.7% 4.1% 3.6% 5.0%


57

Table A4.4: Cost of capital breakdown – Short term high case (low gearing, high beta)2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010


Tax 35.2% 35.2% 33.0% 33.0% 33.0% 27.5% 27.5% 27.5% 27.5% 26.5% 26.5%

Inflation 2.2% 2.5% 3.3% 2.8% 2.2% 1.7% 1.8% 2.1% 2.6% 2.2% 1.6%


Debt premium 1.0% 1.7% 1.3% 0.8% 0.7% 0.3% 0.4% 0.5% 1.9% 1.8% 1.6%

EMRP 5.3% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0%

Asset beta 0.35 0.25 0.28 0.32 0.31 0.33 0.40 0.50 0.40 0.40 0.39

Equity beta 0.57 0.41 0.47 0.57 0.55 0.61 0.77 1.00 0.78 0.75 0.71


4.0% 3.8% 2.4% 1.3% 1.8% 1.4% 2.2% 2.6% 3.4% 2.7% 4.2%


9.2% 6.4% 5.1% 5.0% 5.7% 5.8% 7.9% 9.7% 7.5% 6.3% 8.4%


5.9% 4.2% 3.4% 3.4% 3.8% 4.2% 5.7% 7.1% 5.4% 4.6% 6.2%

Cost of capital

7.2% 5.4% 4.0% 3.4% 3.9% 3.8% 5.1% 6.2% 5.5% 4.6% 6.5%




5.2% 4.0% 3.0% 2.5% 2.9% 2.9% 4.0% 4.8% 4.4% 3.7% 5.3%


59

Table A4.5: Cost of capital breakdown – Medium term base case2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010


Tax 35.2% 35.2% 33.0% 33.0% 33.0% 27.5% 27.5% 27.5% 27.5% 26.5% 26.5%

Inflation 2.2% 2.5% 3.3% 2.8% 2.2% 1.7% 1.8% 2.1% 2.6% 2.2% 1.6%


Debt premium 0.8% 1.7% 1.3% 1.1% 0.9% 0.3% 0.6% 0.8% 1.8% 2.1% 1.6%

EMRP 5.3% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0%

Asset beta 0.25 0.20 0.15 0.21 0.22 0.28 0.30 0.39 0.35 0.32 0.35

Equity beta 0.51 0.40 0.32 0.45 0.49 0.65 0.72 0.97 0.85 0.74 0.78


4.1% 4.3% 3.0% 2.4% 2.8% 2.1% 2.6% 3.1% 3.7% 4.0% 5.3%


9.3% 7.1% 4.8% 5.4% 6.5% 6.9% 7.8% 9.8% 8.4% 7.7% 10.2%


6.0% 4.6% 3.2% 3.6% 4.4% 5.0% 5.6% 7.1% 6.1% 5.6% 7.5%

Cost of capital

6.7% 5.7% 3.8% 3.8% 4.5% 4.1% 4.8% 5.8% 5.7% 5.6% 7.5%




5.1% 4.5% 3.1% 3.0% 3.5% 3.3% 3.9% 4.7% 4.7% 4.7% 6.3%


61

Table A4.6: Cost of capital breakdown – Medium term low case (high gearing, low beta)2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010


Tax 35.2% 35.2% 33.0% 33.0% 33.0% 27.5% 27.5% 27.5% 27.5% 26.5% 26.5%

Inflation 2.2% 2.5% 3.3% 2.8% 2.2% 1.7% 1.8% 2.1% 2.6% 2.2% 1.6%


Debt premium 0.8% 1.7% 1.3% 1.1% 0.9% 0.3% 0.6% 0.8% 1.8% 2.1% 1.6%

EMRP 5.3% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0%

Asset beta 0.12 0.16 0.07 0.12 0.16 0.14 0.26 0.25 0.24 0.28 0.25

Equity beta 0.32 0.45 0.21 0.37 0.52 0.49 0.96 1.01 0.90 0.98 0.84


4.1% 4.3% 3.0% 2.4% 2.8% 2.1% 2.6% 3.1% 3.7% 4.0% 5.3%


7.8% 7.5% 4.0% 4.9% 6.8% 5.8% 9.5% 10.1% 8.8% 9.3% 10.7%


5.0% 4.9% 2.7% 3.3% 4.5% 4.2% 6.9% 7.3% 6.4% 6.9% 7.9%

Cost of capital

5.5% 5.5% 3.3% 3.2% 4.0% 3.1% 4.5% 4.8% 5.1% 5.5% 6.9%




4.5% 4.5% 2.9% 2.7% 3.3% 2.7% 3.8% 4.1% 4.4% 4.8% 6.0%


63

Table A4.7: Cost of capital breakdown – Medium term high case (low gearing, high beta)2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010


Tax 35.2% 35.2% 33.0% 33.0% 33.0% 27.5% 27.5% 27.5% 27.5% 26.5% 26.5%

Inflation 2.2% 2.5% 3.3% 2.8% 2.2% 1.7% 1.8% 2.1% 2.6% 2.2% 1.6%


Debt premium 0.8% 1.7% 1.3% 1.1% 0.9% 0.3% 0.6% 0.8% 1.8% 2.1% 1.6%

EMRP 5.3% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0%

Asset beta 0.35 0.25 0.28 0.32 0.31 0.33 0.40 0.50 0.40 0.40 0.39

Equity beta 0.57 0.41 0.47 0.57 0.55 0.61 0.77 1.00 0.78 0.75 0.71


4.1% 4.3% 3.0% 2.4% 2.8% 2.1% 2.6% 3.1% 3.7% 4.0% 5.3%


9.7% 7.2% 6.0% 6.3% 7.0% 6.6% 8.1% 10.0% 8.0% 7.7% 9.8%


6.3% 4.6% 4.0% 4.2% 4.7% 4.8% 5.9% 7.3% 5.8% 5.7% 7.2%

Cost of capital

7.6% 6.0% 4.7% 4.6% 5.1% 4.5% 5.5% 6.5% 5.9% 6.0% 7.8%




5.5% 4.5% 3.6% 3.5% 3.9% 3.5% 4.3% 5.2% 4.8% 4.9% 6.3%


65

Table A4.8: Cost of capital breakdown – Long term base case2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010


Tax 35.2% 35.2% 33.0% 33.0% 33.0% 27.5% 27.5% 27.5% 27.5% 26.5% 26.5%

Inflation 2.2% 2.5% 3.3% 2.8% 2.2% 1.7% 1.8% 2.1% 2.6% 2.2% 1.6%


Debt premium 2.0% 1.0% 0.7% 0.6% 0.5% 0.2% 0.4% 0.5% 1.3% 1.4% 1.5%

EMRP 5.3% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0%

Asset beta 0.25 0.20 0.15 0.21 0.22 0.28 0.30 0.39 0.35 0.32 0.35

Equity beta 0.51 0.40 0.32 0.45 0.49 0.65 0.72 0.97 0.85 0.74 0.78


6.0% 4.2% 3.0% 2.5% 2.9% 2.4% 2.9% 3.0% 3.4% 3.8% 5.4%


10.3% 8.1% 5.7% 6.2% 7.3% 7.4% 8.4% 10.1% 8.8% 8.3% 10.6%


6.7% 5.2% 3.8% 4.1% 4.9% 5.4% 6.1% 7.3% 6.4% 6.1% 7.8%

Cost of capital

8.1% 6.1% 4.3% 4.2% 4.8% 4.5% 5.2% 5.9% 5.7% 5.8% 7.8%




6.3% 4.7% 3.4% 3.2% 3.8% 3.7% 4.2% 4.7% 4.7% 4.8% 6.5%


67

Table A4.9: Cost of capital breakdown – Long term low case (high gearing, low beta)2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010


Tax 35.2% 35.2% 33.0% 33.0% 33.0% 27.5% 27.5% 27.5% 27.5% 26.5% 26.5%

Inflation 2.2% 2.5% 3.3% 2.8% 2.2% 1.7% 1.8% 2.1% 2.6% 2.2% 1.6%


Debt premium 2.0% 1.0% 0.7% 0.6% 0.5% 0.2% 0.4% 0.5% 1.3% 1.4% 1.5%

EMRP 5.3% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0%

Asset beta 0.12 0.16 0.07 0.12 0.16 0.14 0.26 0.25 0.24 0.28 0.25

Equity beta 0.32 0.45 0.21 0.37 0.52 0.49 0.96 1.01 0.90 0.98 0.84


6.0% 4.2% 3.0% 2.5% 2.9% 2.4% 2.9% 3.0% 3.4% 3.8% 5.4%


8.8% 8.5% 4.9% 5.6% 7.5% 6.4% 10.1% 10.4% 9.2% 9.9% 11.1%


5.7% 5.5% 3.3% 3.7% 5.0% 4.6% 7.3% 7.5% 6.7% 7.3% 8.2%

Cost of capital

7.0% 5.7% 3.6% 3.5% 4.3% 3.5% 4.8% 4.9% 5.0% 5.6% 7.1%




5.9% 4.6% 3.1% 2.9% 3.6% 3.0% 4.1% 4.1% 4.3% 4.8% 6.2%


69

Table A4.10: Cost of capital breakdown – Long term high case (low gearing, high beta)2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010


Tax 35.2% 35.2% 33.0% 33.0% 33.0% 27.5% 27.5% 27.5% 27.5% 26.5% 26.5%

Inflation 2.2% 2.5% 3.3% 2.8% 2.2% 1.7% 1.8% 2.1% 2.6% 2.2% 1.6%


Debt premium 2.0% 1.0% 0.7% 0.6% 0.5% 0.2% 0.4% 0.5% 1.3% 1.4% 1.5%

EMRP 5.3% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0%

Asset beta 0.35 0.25 0.28 0.32 0.31 0.33 0.40 0.50 0.40 0.40 0.39

Equity beta 0.57 0.41 0.47 0.57 0.55 0.61 0.77 1.00 0.78 0.75 0.71


6.0% 4.2% 3.0% 2.5% 2.9% 2.4% 2.9% 3.0% 3.4% 3.8% 5.4%


10.7% 8.1% 6.8% 7.0% 7.7% 7.2% 8.7% 10.3% 8.4% 8.3% 10.2%


7.0% 5.2% 4.6% 4.7% 5.2% 5.2% 6.3% 7.5% 6.1% 6.1% 7.5%

Cost of capital

9.0% 6.6% 5.3% 5.1% 5.6% 5.0% 5.9% 6.7% 6.0% 6.2% 8.0%




6.6% 4.8% 3.9% 3.8% 4.2% 3.9% 4.7% 5.2% 4.8% 5.1% 6.6%


71

Table A4.11: WACC robustness checks (for medium term base case)2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Baseline 6.7% 5.7% 3.8% 3.8% 4.5% 4.1% 4.8% 5.8% 5.7% 5.6% 7.5%All companies 6.6% 5.7% 3.8% 3.8% 4.4% 4.0% 4.6% 5.6% 5.6% 5.6% 7.4%

Good comps. 7.1% 5.6% 3.6% 3.8% 4.3% 4.4% 5.4% 6.3% 5.3% 5.6% 7.6%

Euro 5.9% 5.5% 3.4% 3.4% 4.1% 3.3% 4.4% 5.6% 5.4% 5.6% 7.5%5-year 7.1% 6.1% 4.0% 3.9% 4.4% 3.8% 4.7% 5.2% 5.5% 5.6% 7.6%Raw equity 6.5% 5.9% 3.9% 4.0% 4.8% 3.8% 4.5% 4.9% 5.2% 5.4% 7.1%

Notes: ‘Baseline’ estimates are those presented in the main report for the medium term base case, using our ‘Integrated utilities’ set of comparators.‘All companies’ estimates use all relevant companies for which data are available.‘Good comps.’ estimates use a subset of companies identified as the most relevant comparators.‘Euro’ estimates use beta estimates calculated using the Eurotop 100 index as the reference index for each company (instead of each company’s local reference index).‘5 –year’ estimates use beta estimates calculated using a 5-year rolling window (instead of a 2-year). ‘Raw equity’ estimates use companies’ raw equity betas (instead of de-levered asset betas that are then re-levered using our notional gearing assumption).



Beta estimatesFor the beta estimates, we summarise asset betas for the following sets of comparator companies:

all companies; good comparators; pure generators; integrated utilities (base case); transmission utilities; and renewable energy companies.

73

Figure A4.1: Asset betas – All companies

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4



Figure A4.3: Asset betas – Pure generators

0.0

0.2

0.4

0.6

0.8

1.0

1.2



Figure A4.2: Asset betas – Good comparators Figure A4.4: Asset betas – Integrated utilities (base case)



0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9



0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4



Figure A4.5: Asset betas – Transmission utilities

0.0

0.1

0.1

0.2

0.2

0.3

0.3

0.4



Figure A4.6: Asset betas – Renewable generators

0.0

0.2

0.4

0.6

0.8

1.0

1.2



75

APPENDIX 3: ALTERNATIVE VIEWS ON RISK-FREE RATE AND ERP

Alternative approaches to the risk-free rate

The appropriate country risk-free rate in the context of the current (especially 2010 onwards) financial crisis in Europe is the topic of considerable debate. Some analysts have considered the spreads between a utility’s CDS pricing and the country CDS as indicative of the perceived appropriate risk-free rate i.e. the risk-free rate could be adjusted downwards by the spread. We replicate such analysis in the chart below. The issue though is one of interpretation – in the case of EDP and Portuguese Government debt, there is such a spread and in recent times it is over 200bps, which might indicate that EDP is perceived as less risky than Government short-term debt. But care is needed, as other factors drive the analysis e.g. international diversification of EDP earnings. Unfortunately we have not found similar data for REN.

But nonetheless the data is illustrative and does point to a case that in times of extreme stress on Government debt, the appropriate risk-free rate for a national utility might be marginally lower. Further evidence would be provided by the nominal yields on actual debt issuances by the utility.Figure A6.1: EDP vs. Portuguese government bond CDS prices

EDP CDS EUR SR 5YR Corp

PORTUG CDS USD SR 5YR Corp

0

200

400

600

800

1,000

1,200

1,400

Jan-2002 Jan-2003 Jan-2004 Jan-2005 Jan-2006 Jan-2007 Jan-2008 Jan-2009 Jan-2010 Jan-2011 Jan-2012

Source: Bloomberg



Another alternative approach to estimating the risk-free rate is applying the Portuguese debt premium (measured by the CDS on Portuguese government bonds) to the Euro zone risk-free rate (the yield to maturity of the German bund). Even if we have not employed this approach, the figure below demonstrates that our estimation yields practically the same results.

Figure A6.2: Alternative approaches: CEPA baseline vs. German bund + Portuguese CDS

0%

2%

4%

6%

8%

10%

Risk

free

rate

leve

ls (%

)

CEPA Portugal risk free rate German bund YTM + Portuguese CDS

Alternative approaches to the ERP

We are aware that analysts sometimes attempt alternative approaches to calculating the Equity Risk Premium, as compared to the more conventional regulatory approaches. One such approach is to consider growth prospects and back out ERP’s from share price data. We have not focused on this approach due to its ‘black box’ nature and its dependency on forward-looking assumptions, but it is nonetheless informative.

The chart below is extracted from a research note entitled Strategy Calls prepared by Exane BNP Paribas from December 2011. It shows an ERP with a mean of 4%, but moving significantly over time. This is in line with the lower end of our range.

77

Figure A6.3: Nominal equity risk premium (Europe; %)

Source: Exane BNP ParibasWe are also aware that Credit Suisse’s Valuation Report of 16 February 2007 refers to a Portuguese ERP of 5.63% based on the ‘Damoraran Files. Geometric Returns 1928-2005’. They also provide an historic ERP based on Credit Suisse research for Spain, with a 10 year average of 5.27% and for a ‘Euro 5’ (Germany, France, Netherlands, Italy, Spain) of 4.70%. Again, these numbers are consistent with our ranges.



APPENDIX 2

BENCHMARK ON OPPORTUNITY COST OF CAPITAL FOR INVESTMENTS IN THE SPECIAL REGIME, AT KEARNEY, JANUARY 2012



Determining the necessary rate of return for investments in the Special Regime is based on the weighted average cost of capital (WACC post tax) proxy. WACC is commonly used as the internal discount rate of a project proxy or the overall rate of return required by the array of capital loaners (Debt and Equity providers).

2

WACC value tree

WACC (%)

Equity cost (%)

Debt cost (%)

EE + D

KE (%)

DE + D

KD (%)

1 - t

Risk-free return (%)

Risk premium (%)

Risk-free return (%)

Credit spread (%)

+

x

x +

+

The presented WACC post tax values were calculated for year prices; therefore inflation was deducted from monetary items.Portuguese, Spanish and German WACC in the years 2000, 2005 and 2010 for five technologies under the Special Regime were considered for a comparative analysis on profitability levels: wind onshore, utility-scale solar photovoltaic, cogeneration/CHP (Combined Heat and Power), biomass and small hydro.Debt and Equity conditions as resulting from specific financing structures between different countries were incorporated in the model-based assessment by applying differentiated debt-to-equity levels. Until 2005, the more significant investments in renewable energy were financed in-between 20% to 30%. Towards the end of the decade, the European financial crisis hindered the ability for companies to get debt financing (due to lower credit lines but also to a significant cut (~50%) in the average loan maturity), and the

1

investments that were effectively made were forced to take on more equity. These assumptions are aligned with recent publications on financing Renewable Energy, according to which “a level of 80% debt to 20% equity might have been fairly common pre-financial crisis. However, the cost of debt increased during the financial crisis, and 60 – 70% debt funding became more common“13

There might be specific cases where the proportion of debt is higher or lower than the one estimated, but the general trend on capital financing structure is towards taking on more equity, which will of course aggravate the minimum required IRR for the projects. “The lack of debt available in the market due to problems in the banking sector meant difficulty for investors completing acquisitions where debt is required to supplement the available equity, or to enhance equity returns to an acceptable hurdle level.”14

The risk-free rates of return used in the calculation are the implied rates on the 10-year sovereign bonds, which are somewhat lower for Germany throughout 2000-2010, with the deteriorating situation in both Portugal and Spain towards the end of the decade.15 For inflation deduction it was used the average inflation verified (between 2000 and 2011) and expected (beyond 2011) during the project lifetime (20 or 25 years).Within the model-based analysis, a range of settings was also applied to reflect each investor’s appropriate risk: the market risk premium and the technology risk.The market risk premium incorporates two key variables impacting the remuneration of the projects. One is the country’s economic setting and its ability to fulfill its financial commitments (measured by the country’s sovereign rating and overall confidence levels), given that remuneration for renewable energy (feed-in tariff) is still regulated; the second is the remuneration or policy risk, which assumes a particularly relevant role in this business. This risk was

13 Bloomberg New Energy Finance “Private financing of Renewable Energy”14 Bloomberg New Energy Finance “Private financing of Renewable Energy”15 European Central Bank: Statistical Data Warehouse



considered to be slightly higher for Portugal than for Germany and Spain, given the characteristics of each of the three remuneration systems. Portugal rewards the electricity feed-in for a smaller period of 15 years (for most of the technologies), after which the operator sells in the free market (high volatility)16.. In Spain, the operator may choose between a fixed feed-in tariff for the whole life of the operation (zero risk) or a premium variable tariff with a high ceiling and a minimum acceptable floor; the variable tariff is also given for the whole life of the operation, except for the more mature wind onshore and biomass technologies, but even for these technologies a tariff is assured for the first 25 years of operation, reducing significantly the risk of the project. The German feed-in tariff scheme rewards for a higher period of 20 years, and despite its degression rates and no adjustments to inflation, its tariff is higher than most in Europe and significantly higher than the generation cost of electricity. Moreover, and in the specific case of wind onshore, the tariff for the initial period (typically 5 years) is dependent on the load factor of each wind farm; lower load factors get the higher initial tariff for a longer period, up to 20 years.The technology risk assumes technological uncertainty in terms of future development and deployment due to, for example, (unexpected) additional operational and maintenance cost or substantial reinvestments which (after a phase out of operational guarantees) have to be borne by the investors themselves. Wind onshore has typically a smaller risk while biomass and CHP is slightly higher17, mainly due to its dependence on fuel/feedstock. This risk acts as a multiplier over the industry’s beta (a 1.6-1.7 is typically used18 for renewable energy investments).16 In 2005 was published a decree-law (DL nº 33-A/2005) that limits the feed-in tariff regime to 15 years from the date of entry into force of this statute for all renewable power plants (already in operation and in licensing process). Thus, there is a considerable renewable power taking advantage of more than 15 years of feed-in tariff and in this case the risk is substantial lower.17 ECOFYS “Financing Renewable Energy in the European Energy Market”18 Green X "Deriving optimal promotion strategies for increasing the share of RES-E in a dynamic European electricity market“; RedPoint Energy “Electricity Market Reform Analysis of policy options”

3

Finally, the minimum credit spread considered for Portugal was of 300 bps over the interbank interest rate, typically required for a solid investor (e.g. AA rating) in terms of credit rating and for projects involving mature technologies, according to Bloomberg19. Interviews with Project Finance experts in a major Portuguese bank recognized that credit spreads between Portugal and Spain vary significantly, currently around 600 bps for Portugal and 400 bps for Spain. This increase and differential were reflected in the 2010 credit spread assumed for both countries, although in a more conservative manner (+50 bps over the credit spread of 2005). For Germany, a lower spread of 200 bps was used, in accordance to the Green X study sponsored by the European Union.As a result, two tables were built for each of the five different technologies: one with the yearly evolution of the reference WACC for Portugal (2000-2010) and a second one with a comparative analysis with Germany and Spain for years 2000, 2005 and 2010.

1. Reference WACC for Onshore Wind-farm projects

19 Bloomberg New Energy Finance “Private financing of Renewable Energy”



Post-tax real WACC evolution for Wind-farm projects- Portugal -

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Equity cost (%) 4,0% 3,9% 3,8% 3,8% 3,5% 3,5% 2,4% 2,5% 2,4% 2,3% 4,6%

E / (D+E) 30,0% 30,0% 30,0% 30,0% 30,0% 30,0% 20,0% 20,0% 20,0% 20,0% 35,0%

KE (%) 13,3% 13,1% 12,7% 12,5% 11,8% 11,6% 12,0% 12,5% 11,8% 11,3% 13,1%

Risk free return (%)

2,5% 2,5% 2,2% 2,2% 1,6% 1,5% 2,1% 2,8% 2,2% 1,8% 3,7%

Risk premium (%)

10,8% 10,6% 10,5% 10,3% 10,2% 10,1% 9,9% 9,8% 9,6% 9,5% 9,4%

Debt cost (%) 2,5% 2,5% 2,4% 2,4% 2,3% 2,3% 3,0% 3,4% 3,0% 2,8% 3,4%

D / (D+E) 70,0% 70,0% 70,0% 70,0% 70,0% 70,0% 80,0% 80,0% 80,0% 80,0% 65,0%

KD (%) 5,5% 5,5% 5,2% 5,2% 4,6% 4,5% 5,1% 5,8% 5,2% 4,8% 7,2%


2,5% 2,5% 2,2% 2,2% 1,6% 1,5% 2,1% 2,8% 2,2% 1,8% 3,7%

Credit spread (%)

3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,5%

1 - t 64,8% 64,8% 67,0% 67,0% 72,5% 72,5% 72,5% 73,5% 73,5% 73,5% 73,5%

WACC (%) 6,5% 6,4% 6,2% 6,2% 5,9% 5,8% 5,4% 5,9% 5,4% 5,1% 8,0%

Post-tax real WACC evolution for Wind-farm projects- Benchmarking analysis -

Portugal Germany Spain

2000 2005 2010 2000 2005 2010 2000 2005 2010

Equity cost (%) 4,0% 3,5% 4,6% 3,1% 2,5% 2,3% 2,6% 2,2% 3,7%

E / (D+E) 30,0% 30,0% 35,0% 25,0% 25,0% 25,0% 20,0% 20,0% 30,0%

KE (%) 13,3% 11,6% 13,1% 12,4% 10,2% 9,1% 13,0% 10,8% 12,3%

Risk free return (%) 2,5% 1,5% 3,7% 3,2% 1,7% 1,3% 3,1% 1,6% 3,8%

Risk premium (%) 10,8% 10,1% 9,4% 9,2% 8,5% 7,8% 9,9% 9,2% 8,5%

Debt cost (%) 2,5% 2,3% 3,4% 1,9% 1,7% 2,0% 2,9% 2,1% 3,3%

D / (D+E) 70,0% 70,0% 65,0% 75,0% 75,0% 75,0% 80,0% 80,0% 70,0%

KD (%) 5,5% 4,5% 7,2% 5,2% 3,7% 3,8% 5,6% 4,1% 6,8%

Risk free return (%) 2,5% 1,5% 3,7% 3,2% 1,7% 1,3% 3,1% 1,6% 3,8%

Credit spread (%) 3,0% 3,0% 3,5% 2,0% 2,0% 2,5% 2,5% 2,5% 3,0%

1 - t 64,8% 72,5% 73,5% 48,0% 61,1% 69,8% 65,0% 65,0% 70,0%

WACC (%) 6,5% 5,8% 8,0% 5,0% 4,2% 4,3% 5,5% 4,3% 7,0%

2. Reference WACC for Utility-scale Solar PV projects

5

Post-tax real WACC evolution for Utility scale PV projects- Portugal -

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Equity cost (%) 4,2% 4,1% 4,0% 4,0% 3,8% 3,7% 2,6% 2,7% 2,6% 2,5% 5,0%

E / (D+E) 30,0% 30,0% 30,0% 30,0% 30,0% 30,0% 20,0% 20,0% 20,0% 20,0% 35,0%

KE (%) 13,8% 13,7% 13,3% 13,3% 12,6% 12,4% 12,9% 13,5% 12,8% 12,4% 14,2%


2,5% 2,5% 2,2% 2,2% 1,6% 1,5% 2,1% 2,8% 2,2% 1,8% 3,7%

Risk premium (%)

11,3% 11,2% 11,1% 11,1% 11,0% 10,9% 10,8% 10,7% 10,7% 10,6% 10,5%

Debt cost (%) 2,5% 2,5% 2,4% 2,4% 2,3% 2,3% 3,0% 3,4% 3,0% 2,8% 3,4%

D / (D+E) 70,0% 70,0% 70,0% 70,0% 70,0% 70,0% 80,0% 80,0% 80,0% 80,0% 65,0%

KD (%) 5,5% 5,5% 5,2% 5,2% 4,6% 4,5% 5,1% 5,8% 5,2% 4,8% 7,2%


2,5% 2,5% 2,2% 2,2% 1,6% 1,5% 2,1% 2,8% 2,2% 1,8% 3,7%

Credit spread (%)

3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,5%

1 - t 64,8% 64,8% 67,0% 67,0% 72,5% 72,5% 72,5% 73,5% 73,5% 73,5% 73,5%

WACC (%) 6,7% 6,6% 6,4% 6,4% 6,1% 6,0% 5,6% 6,1% 5,6% 5,3% 8,4%

Post-tax real WACC evolution for Utility scale PV projects- Benchmarking analysis -


2000 2005 2010 2000 2005 2010 2000 2005 2010

Equity cost (%) 4,2% 3,7% 5,0% 3,2% 2,7% 2,5% 2,7% 2,3% 4,0%

E / (D+E) 30,0% 30,0% 35,0% 25,0% 25,0% 25,0% 20,0% 20,0% 30,0%

KE (%) 13,8% 12,4% 14,2% 12,8% 10,8% 10,0% 13,4% 11,6% 13,3%

Risk free return (%) 2,5% 1,5% 3,7% 3,2% 1,7% 1,3% 3,1% 1,6% 3,8%

Risk premium (%) 11,3% 10,9% 10,5% 9,5% 9,1% 8,7% 10,3% 9,9% 9,5%

Debt cost (%) 2,5% 2,3% 3,4% 1,9% 1,7% 2,0% 2,9% 2,1% 3,3%

D / (D+E) 70,0% 70,0% 65,0% 75,0% 75,0% 75,0% 80,0% 80,0% 70,0%

KD (%) 5,5% 4,5% 7,2% 5,2% 3,7% 3,8% 5,6% 4,1% 6,8%

Risk free return (%) 2,5% 1,5% 3,7% 3,2% 1,7% 1,3% 3,1% 1,6% 3,8%

Credit spread (%) 3,0% 3,0% 3,5% 2,0% 2,0% 2,5% 2,5% 2,5% 3,0%

1 - t 64,8% 72,5% 73,5% 48,0% 61,1% 69,8% 65,0% 65,0% 70,0%

WACC (%) 6,7% 6,0% 8,4% 5,1% 4,4% 4,5% 5,6% 4,5% 7,3%

3. Reference WACC for CHP projects



Post-tax real WACC evolution for CHP projects - Portugal -

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Equity cost (%) 4,0% 4,0% 3,9% 3,9% 3,7% 3,7% 2,6% 2,7% 2,6% 2,6% 5,1%

E / (D+E) 30,0% 30,0% 30,0% 30,0% 30,0% 30,0% 20,0% 20,0% 20,0% 20,0% 35,0%

KE (%) 13,4% 13,4% 13,1% 13,1% 12,5% 12,5% 13,0% 13,7% 13,1% 12,8% 14,6%


2,5% 2,5% 2,2% 2,2% 1,6% 1,5% 2,1% 2,8% 2,2% 1,8% 3,7%

Risk premium (%)

10,9% 10,9% 10,9% 10,9% 10,9% 10,9% 10,9% 10,9% 10,9% 10,9% 10,9%

Debt cost (%) 2,5% 2,5% 2,4% 2,4% 2,3% 2,3% 3,0% 3,4% 3,0% 2,8% 3,4%

D / (D+E) 70,0% 70,0% 70,0% 70,0% 70,0% 70,0% 80,0% 80,0% 80,0% 80,0% 65,0%

KD (%) 5,5% 5,5% 5,2% 5,2% 4,6% 4,5% 5,1% 5,8% 5,2% 4,8% 7,2%


2,5% 2,5% 2,2% 2,2% 1,6% 1,5% 2,1% 2,8% 2,2% 1,8% 3,7%

Credit spread (%)

3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,5%

1 - t 64,8% 64,8% 67,0% 67,0% 72,5% 72,5% 72,5% 73,5% 73,5% 73,5% 73,5%

WACC (%) 6,5% 6,5% 6,4% 6,4% 6,1% 6,0% 5,6% 6,1% 5,7% 5,4% 8,6%

Post-tax real WACC evolution for CHP projects- Benchmarking analysis -


2000 2005 2010 2000 2005 2010 2000 2005 2010

Equity cost (%) 4,0% 3,7% 5,1% 3,1% 2,7% 2,6% 2,6% 2,3% 4,1%

E / (D+E) 30,0% 30,0% 35,0% 25,0% 25,0% 25,0% 20,0% 20,0% 30,0%

KE (%) 13,4% 12,5% 14,6% 12,3% 10,8% 10,4% 13,0% 11,5% 13,7%

Risk free return (%) 2,5% 1,5% 3,7% 3,2% 1,7% 1,3% 3,1% 1,6% 3,8%

Risk premium (%) 10,9% 10,9% 10,9% 9,1% 9,1% 9,1% 9,9% 9,9% 9,9%

Debt cost (%) 2,5% 2,3% 3,4% 1,9% 1,7% 2,0% 2,9% 2,1% 3,3%

D / (D+E) 70,0% 70,0% 65,0% 75,0% 75,0% 75,0% 80,0% 80,0% 70,0%

KD (%) 5,5% 4,5% 7,2% 5,2% 3,7% 3,8% 5,6% 4,1% 6,8%

Risk free return (%) 2,5% 1,5% 3,7% 3,2% 1,7% 1,3% 3,1% 1,6% 3,8%

Credit spread (%) 3,0% 3,0% 3,5% 2,0% 2,0% 2,5% 2,5% 2,5% 3,0%

1 - t 64,8% 72,5% 73,5% 48,0% 61,1% 69,8% 65,0% 65,0% 70,0%

WACC (%) 6,5% 6,0% 8,6% 5,0% 4,4% 4,6% 5,5% 4,5% 7,4%

4. Reference WACC for Biomass projects

7

Post-tax real WACC evolution for Biomass projects - Portugal -

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Equity cost (%) 4,3% 4,2% 4,1% 4,1% 3,9% 3,9% 2,7% 2,8% 2,7% 2,6% 5,1%

E / (D+E) 30,0% 30,0% 30,0% 30,0% 30,0% 30,0% 20,0% 20,0% 20,0% 20,0% 35,0%

KE (%) 14,3% 14,1% 13,8% 13,7% 13,0% 12,9% 13,4% 13,9% 13,3% 12,8% 14,6%


2,5% 2,5% 2,2% 2,2% 1,6% 1,5% 2,1% 2,8% 2,2% 1,8% 3,7%

Risk premium (%)

11,8% 11,7% 11,6% 11,5% 11,4% 11,3% 11,2% 11,2% 11,1% 11,0% 10,9%

Debt cost (%) 2,5% 2,5% 2,4% 2,4% 2,3% 2,3% 3,0% 3,4% 3,0% 2,8% 3,4%

D / (D+E) 70,0% 70,0% 70,0% 70,0% 70,0% 70,0% 80,0% 80,0% 80,0% 80,0% 65,0%

KD (%) 5,5% 5,5% 5,2% 5,2% 4,6% 4,5% 5,1% 5,8% 5,2% 4,8% 7,2%


2,5% 2,5% 2,2% 2,2% 1,6% 1,5% 2,1% 2,8% 2,2% 1,8% 3,7%

Credit spread (%)

3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,5%

1 - t 64,8% 64,8% 67,0% 67,0% 72,5% 72,5% 72,5% 73,5% 73,5% 73,5% 73,5%

WACC (%) 6,8% 6,7% 6,6% 6,5% 6,2% 6,2% 5,6% 6,2% 5,7% 5,4% 8,6%

Post-tax real WACC evolution for Biomass projects- Benchmarking analysis -


2000 2005 2010 2000 2005 2010 2000 2005 2010

Equity cost (%) 4,3% 3,9% 5,1% 3,3% 2,8% 2,6% 2,8% 2,4% 4,1%

E / (D+E) 30,0% 30,0% 35,0% 25,0% 25,0% 25,0% 20,0% 20,0% 30,0%

KE (%) 14,3% 12,9% 14,6% 13,1% 11,2% 10,4% 13,8% 11,9% 13,7%

Risk free return (%) 2,5% 1,5% 3,7% 3,2% 1,7% 1,3% 3,1% 1,6% 3,8%

Risk premium (%) 11,8% 11,3% 10,9% 9,9% 9,5% 9,1% 10,8% 10,3% 9,9%

Debt cost (%) 2,5% 2,3% 3,4% 1,9% 1,7% 2,0% 2,9% 2,1% 3,3%

D / (D+E) 70,0% 70,0% 65,0% 75,0% 75,0% 75,0% 80,0% 80,0% 70,0%

KD (%) 5,5% 4,5% 7,2% 5,2% 3,7% 3,8% 5,6% 4,1% 6,8%

Risk free return (%) 2,5% 1,5% 3,7% 3,2% 1,7% 1,3% 3,1% 1,6% 3,8%

Credit spread (%) 3,0% 3,0% 3,5% 2,0% 2,0% 2,5% 2,5% 2,5% 3,0%

1 - t 64,8% 72,5% 73,5% 48,0% 61,1% 69,8% 65,0% 65,0% 70,0%

WACC (%) 6,8% 6,2% 8,6% 5,2% 4,5% 4,6% 5,7% 4,5% 7,4%

5. Reference WACC for Small Hydro projects



Post-tax real WACC evolution for Small Hydro projects - Portugal -

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Equity cost (%) 3,7% 3,7% 3,6% 3,6% 3,4% 3,4% 2,4% 2,5% 2,4% 2,3% 4,7%

E / (D+E) 30,0% 30,0% 30,0% 30,0% 30,0% 30,0% 20,0% 20,0% 20,0% 20,0% 35,0%

KE (%) 12,3% 12,2% 12,0% 12,0% 11,4% 11,3% 11,9% 12,5% 12,0% 11,6% 13,5%


2,5% 2,5% 2,2% 2,2% 1,6% 1,5% 2,1% 2,8% 2,2% 1,8% 3,7%

Risk premium (%)

9,8% 9,8% 9,8% 9,8% 9,8% 9,8% 9,8% 9,8% 9,8% 9,8% 9,8%

Debt cost (%) 2,5% 2,5% 2,4% 2,4% 2,3% 2,3% 3,0% 3,4% 3,0% 2,8% 3,4%

D / (D+E) 70,0% 70,0% 70,0% 70,0% 70,0% 70,0% 80,0% 80,0% 80,0% 80,0% 65,0%

KD (%) 5,5% 5,5% 5,2% 5,2% 4,6% 4,5% 5,1% 5,8% 5,2% 4,8% 7,2%


2,5% 2,5% 2,2% 2,2% 1,6% 1,5% 2,1% 2,8% 2,2% 1,8% 3,7%

Credit spread (%)

3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,0% 3,5%

1 - t 64,8% 64,8% 67,0% 67,0% 72,5% 72,5% 72,5% 73,5% 73,5% 73,5% 73,5%

WACC (%) 6,2% 6,1% 6,0% 6,0% 5,7% 5,7% 5,4% 5,9% 5,4% 5,2% 8,2%

Post-tax real WACC evolution for Small Hydro projects- Benchmarking analysis -


2000 2005 2010 2000 2005 2010 2000 2005 2010

Equity cost (%) 3,7% 3,4% 4,7% 2,8% 2,5% 2,4% 2,4% 2,1% 3,8%

E / (D+E) 30,0% 30,0% 35,0% 25,0% 25,0% 25,0% 20,0% 20,0% 30,0%

KE (%) 12,3% 11,3% 13,5% 11,4% 9,8% 9,4% 11,9% 10,5% 12,6%

Risk free return (%) 2,5% 1,5% 3,7% 3,2% 1,7% 1,3% 3,1% 1,6% 3,8%

Risk premium (%) 9,8% 9,8% 9,8% 8,1% 8,1% 8,1% 8,9% 8,9% 8,9%

Debt cost (%) 2,5% 2,3% 3,4% 1,9% 1,7% 2,0% 2,9% 2,1% 3,3%

D / (D+E) 70,0% 70,0% 65,0% 75,0% 75,0% 75,0% 80,0% 80,0% 70,0%

KD (%) 5,5% 4,5% 7,2% 5,2% 3,7% 3,8% 5,6% 4,1% 6,8%

Risk free return (%) 2,5% 1,5% 3,7% 3,2% 1,7% 1,3% 3,1% 1,6% 3,8%

Credit spread (%) 3,0% 3,0% 3,5% 2,0% 2,0% 2,5% 2,5% 2,5% 3,0%

1 - t 64,8% 72,5% 73,5% 48,0% 61,1% 69,8% 65,0% 65,0% 70,0%

WACC (%) 6,2% 5,7% 8,2% 4,7% 4,2% 4,3% 5,3% 4,2% 7,1%

9

APPENDIX 3

EFFECTIVE RATES OF RETURN OF PRO GENERATION ASSETS (CMEC AND PPA)



1. CMEC

Following the publication of Decree-law n.º 240/2004, of December 27, on January 27, 2005, were signed the agreements for the early termination of the power purchase agreements (PPA's) of EDP’s binding electricity power plants. The referred Decree-Law established that in order to maintain the contractual equilibrium of the PPAs, the owners of such agreements, which include a significant portion of EDP's generation capacity in Portugal, have the right to receive a compensation for the early termination of those agreements (CMEC). The effects of the termination of these agreements depended on the verification of a set of conditions, which included the launch of the spot electricity market at the Iberian level (MIBEL), which came into effect on July 1, 2007.On February 16, 2007, the Portuguese Government confirmed the decision to early terminate the PPAs and implement the CMEC mechanism, closing the rules to calculate the compensations due to the power generators for such early termination. On June 15, 2007, EDP and REN agreed on the early termination of the PPAs, with effect as of July 1, 2007. The new CMEC regulation set the amount of the initial value at 833 M€ (corresponding to the NPV of the PPAs at that date), which can be subject to securitization. It was also established that EDP would pay 759 M€ for the extension of the use of hydro public domain, securing the right to operate 26 hydroelectric plants with a capacity of 4100 MW, under free market conditions for an average period of over 26 years.Under this context, the effective rate of return of the generation assets under CMEC regime is determined by the assessment of those two main financial operations of the CMEC process: The calculation of the value of the former PPAs when they were terminated; and The calculation of the value of the extension of the use of water from the

1

end of the hydro PPAs to the end of the concession of public hydrological domain.

1.1 Rent due to the calculation of the initial value of the CMEC

The initial value of the CMEC was calculated by the difference between the 2007 value of the power plants’ fixed annual charge, estimated for the remaining PPAs contracted period, and the expected income stemming from the foreseen sale of electricity in the wholesale market (at 50 €/MWh), deduced by the estimated annual variable production costs.According to the Decree-Law 240/2004, of December 27th, the NPV of the early termination of a PPA for power plant k (CPk) is given by the following expression:

The major part of the fixed charges corresponds to the compensation and the amortization of the power plants’ net assets. The former PPAs, signed with EDP in 1996, defined a rate of return for the power plants’ net assets of 8,5% real pre-tax. Since the net asset value is linked to the evolution of a set of deflators, the equivalent rate of return is 10,67% nominal pre-tax, considering 2% inflation rate used in CMEC framework.To discount the futures cash flows (calculated with 10,67% rate of return), the Decree-Law 240/2004, of December 27th, established a rate given by the yield at the date of termination of the PPAs of the



Expected income stemming from the foreseen sale of electricity in the wholesale market (at 50 €/MWh) deduced by the estimated annual variable production costs.

Fixed charge at nominal price (for a 2% estimated inflation). This charge includes mainly the compensation and depreciation of fixed assets.

Portuguese Government Bond with residual maturity closest to the average remaining life of the PPAs, added by 0,25%. That discount rate was 4,85% nominal, leading to a CMEC initial value of 833 M€ 20.Finally, that NPV was to be recovered along 20 years through an annuity payment, which will be paid by all consumers. The interest rate used for calculating the annuity should match the lower of the following rates:

- The Weighted Average Cost of Capital (WACC) of the power generation activity to be defined by the Minister of Economy;

- The annual interest rate associated with payments to the holders of securities, if CMEC value is securitized.

As the CMEC haven’t been securitized, the annuity (81 M€) is being calculated based on the opportunity cost of capital of EDP, which has been defined by the Minister of Economy in 2007 as 7,55%.The calculation of CMEC initial value has, then, three implicit nominal rates: i) to remunerate the former PPAs net assets (10,67%); ii) to calculate the initial value of the CMEC (4,85%) and, finally; iii) to calculate the annuity correspondent to that initial value (7,55%). This approach does not respect financial neutrality with the former PPAs and benefits the generator: It uses low discount rates to calculate NPV of future EDP rights; and considers high taxes to calculate the value of future consumer liabilities.The effect on the rate of return due to the calculation of the initial value of the CMEC is associated with the difference between the value calculated in 2007 (833 M€) and the value corresponding to the financial neutrality with the former PPAs regime, considering a rate of return of 10,67% pre-tax nominal (513 M€). This effect

20 The initial value is subject to an annual revision based on the outputs of a model (Valoragua) which simulates Iberian market, taking into account the real data (demand, fuel prices, exogenous constraints, ...). The value of the annual revision is given by the difference between the market margin implicit in the initial value (which was estimated) and the correspondent one calculated by Valoragua using the real data.

3

corresponds to an increase in the former PPAs rate of return of 1,26 pp (to 11,93%) and to a contribution to the CMEC annuity of 31 M€.Because the initial value of the CMEC is higher than it should be if the former PPAs rate of return had been used, the excessive rent associated with the CMEC initial value can be divided in two components:

i. The difference between the rate of return that has resulted from the CMEC initial value calculation (11,93%) and the former PPAs rate of return (10,67%); and

ii. The difference between the former PPAs rate of return (10,67% pre-tax nominal) and the opportunity cost of the generator (7,55% pre-tax nominal)

1.2 Rent due to the extension of the use of the public hydrological resource

After the termination of former PPAs and the beginning of CMEC mechanism, EDP renegotiated with the Portuguese Government the extension of the use of water from the end of the hydro PPAs to the end of the concession of public hydrological resource. These renegotiations were framed by the Decree-Law 226-A/2007, of May 31st, which also establishes that part of the amount to be paid by EDP would be used to reduce 2006 and 2007 tariff deficit. It must be referred that the formers PPAs stated that, one year before the termination of the hydro power plants PPAs, the TSO has to launch a public tender in order to concede a new concession for the hydro power plants.The calculation of the value of the extension of the use of the public hydro resources was done in two steps. First, it was calculated the NPV of the market residual value of the hydro power plants concessions, which EDP has the right to receive at the end of the PPAs. To calculate this NPV, it was used a low discount rate, linked to the long term Portuguese Government Bonds’ yields at that date.



Second, the NPV of the expected cash flows of the hydro plants after PPAs termination has to be determined. The discounted rate used to calculate this NPV was higher, in line with the WACC which was considered adequate for a European electricity generator. As a result, EDP had to pay 759 M€ for the extension of the use of hydro public domain.As in the case of the calculation of the CMEC initial value, the consideration of two different rates to calculate the value of the extension of the use of the public hydro resources benefited the generator. If the rate used to discount the residual value and to calculate the future cash-flows had been the opportunity cost of capital (7,55%), the generator should have paid 1340 M€, i.e., more 581 M€. This effect corresponds to an increase in the former PPAs rate of return of 2,29 pp and to an annuity of 56 M€.

1.3 CMEC rent

0%

2%

4%

6%

8%

10%

12%

14%

0%

2%

4%

6%

8%

10%

12%

14%

Efficient return Former PPAs return Effect of CMEC initialvalue calculation

Effect of hydroconcession extension

Effective return

The effective CMEC rent is 14,22% and is due to three effects:- CMEC initial value calculation;- Extension of use of the concession of the public hydro domain;- Difference between the former PPAs rate of return and the

opportunity cost of the generator.

5

Effi cient CMEC annuity

-100.000

-80.000

-60.000

-40.000

-20.000

0

20.000

40.000

60.000

80.000

100.000

-100.000

-80.000

-60.000

-40.000

-20.000

0

20.000

40.000

60.000

80.000

100.000

CurrentCMEC

annuity

Correctionof CMEC

initial value

Eq. annuityof former

PPAs return

Correctionformer

PPAs return

Correctionof hydro

concessionextension

value

EfficientCMEC

annuity

103

EUR

per y

ear

1.4 Revisibility

The revisibility process is the annual revision of the CMEC initial value. It is based on the outputs of a model (Valoragua) which simulates Iberian market, taking into account the real data (demand, fuel prices, exogenous constraints, ...). The value of the annual revision is given by the difference between the market margin implicit in the initial value (which was estimated) and the correspondent one calculated by Valoragua using the real data.CMEC mechanism is composed by a fixed and an adjustment parcel:

- Fixed parcel – corresponds to a fixed annual and smoothed income that represents the forecasted PPA costs for all of each year between 2007 and 2027;



- Adjustment parcel – corresponds to the amount that corrects, using the real data, the first parcel that was forecasted (this amount corresponds to the annual revisibility).

The adjustment parcel is calculated as follows:

Capacity charge is calculated taking into account the assets remuneration as established in PPA and considering the availability parameter.

The revisibility depends mainly on the following factors: - In terms of capacity charge: Inflation indexes update and

power station performance which is reflected in the availability;

- In terms of net energy costs: market premium (market income – production costs).

The revisibility mechanism is the way of CMEC ensures the risk free environment in relation to what was established in CAE. If market conditions change, payments are revised to ensure initial CAE conditions. The risk remains with the consumer.The CAE conditions are emulated in a simulation model (Valoragua) with conditions and restrictions established in the CMEC amendments. It works like an incentive mechanism to the generator, which has to dispatch the power plants emulating an optimization program.After the first 42 months the producer has been able to have an extra return from this incentive, around an average upside of 10 M€ per year in favour of EDP (equivalent to the effect of the correction factor which is applied to hydro and coal generation that come out from Valoragua model).

7

Adjustment parcel = [Capacity Charge + Net Energy Costs]real - [Capacity Charge + Net Energy Costs]forecasted

Another relevant aspect is the availability reward mechanism. Both in PPA and CMEC terminology, “availability” means the ability of a unit to generate energy in response to a dispatch instruction. The importance of the power station availability lies on the fact of the amount of capacity charge depends on an availability correction (through a calculated factor). This factor results from the comparison between theoretical targets and real availabilities. Thus, it works as a reward or a penalty for the generator.The impact of this availability correction and the inflation update on the capacity charge amount from 2008 to 2010 was as follows.

It can be concluded that this incentive has more upside than downside and it yields an additional average of 33 M€ of return per year to EDP.

2. PPAs

On 15 June 2007 only two PPAs remained in force: for Tejo Energia power plant and for Turbogás power plant. The operation of these power plants and the selling on the wholesale market of the power generated are handled by an enterprise called REN Trading, created as a subsidiary of the TSO.The difference between the costs of the energy produced and the gains of reselling this energy in the market is transferred to the



costumers through the access tariffs. Therefore, the existence of sunk costs or benefits has a neutral effect on the market.As shows in the Figure, the impact of those costs in tariffs is very important, namely in the two last years for which there’s audited financial data, highlighting the fact that nowadays those power plants are not competitive in the MIBEL context.

0

50

100

150

200

250

2008 2009 2010

106

EUR

Tejo Energia Turbogás

The PPAs of those power plants establish that, as a counterpart to the electricity purchased and to the availability performance of the power plants, the producers receive an amount which enable to remunerate fixed costs related to the investments, the Capacity Charge plus the amount of variable costs, the Energy Charge, which are almost integrally related to the fuel consumption costs. Those two agreements are based on financial investments. Therefore, the Capacity Charge reflects the loans conditions which support these investments. In Tejo Energia’s case, Capacity Charge is decreasing in the course of the agreement. In Turbogás’s case, the Capacity Charge is almost constant during the PPA’s duration. Despite the fact that these PPA are so called “project finance”, the implicit rate of return of these investments can be estimated considering the initial investment and the payment of the capacity charges along the PPA duration.

9

APPENDIX 4

EFFECTIVE RATE OF RETURN OF PRE GENERATION ASSETS (WIND, SMALL HYDRO, PHOTOVOLTAIC, BIOMASS, CHP)



Profitability associated with the investments in the special regime is based in the internal rate of return (IRR) of the projects. Due to the large number of PRE power plants (around 500) and the time constraints to develop this work, the IRR for the different special regime technologies were based on the relevant characteristics of the projects, with the intent of representing the technology typical conditions (and not specific projects). If the evolution of the relevant variables is considered, typical characteristics are adequate to evaluate the global value of the excessive rents. Although, any solution to correct those rents will require a more detailed assessment on the economics of the different projects.For the calculation of IRR of the different PRE technologies, the following variables were carefully determined:

- Feed-in-tariff;- Licensing and grid connection date;- Net operating hours;- Plant expected life-time;- CapEx (i.e. EPC costs, electrical and grid connection costs, promotion costs, other);- OpEx (i.e. O&M, Fuel, Insurances, Rents, other).

These variables were estimated on an annual basis since 2000 until 2010 (for CHP since 1997). For that purpose, several sources were used and industry experts were consulted to cross-check assumptions and results.IRRs are normally superior to WACC to ensure investments, still the value generated above WACC threshold is ~140 M€ (at 2011 prices) calculated yearly at average industry standard assumptions.The latter value is ~69 M€ related to CHP production, ~54 M€ related to Wind-farm production, ~9 M€ related to Photovoltaic production, ~5 M€ related to Biomass and ~3 M€ related to Small Hydro.

1. Onshore Wind-farm projects

1

Main variables for determining the IRR of Onshore Wind-farm projects are total investment costs (including turbine and structure acquisition and installation, land contracting and power-grid connection), yearly fixed and variable O&M costs (including insurance and predictable replacement needs), average load factor (depends on the site quality in terms of wind speed and volatility as well on the efficiency and availability of the turbine), the lifespan of the facility and the existing legislation (regulatory scheme, attributable remuneration length and amount and taxation); based on available data and sources, the following assumptions have been derived with the purpose of characterizing the average yearly wind-farm installation:

- The National Energy and Geology Laboratory (LNEG) yearly collects turbine sale prices from manufacturers, establishing a lower and an upper band that results in an historic price band variation; this band has been coupled with known historic wind-farm data to set an estimation of the historic price evolution of wind-farm installations over the last decade, contained between 1,2 and 1,4 M€/MW; National Energy Plans, ECOFYS and IEA Wind data have been used to derive similar curves for relevant foreign countries, for benchmarking purpose.

- LNEG and E.Value estimations for the O&M evolution have been considered; O&M tends to grow during the life of the wind-farm, with an average value at approximately 30 k€/MW.year as of 2010.

- Average load factor has been assumed from observed historic electricity production and sector and experts interviews to be 2.300 nominal power equivalent hours.

- Economic data, such as past and expected inflation and tax rates has been gathered from the National Statistics Institute (INE), OECD, and IMF.



- Average remuneration granted has been computed based on the historic national legislation for wind power production.

- For cross-checking purpose other sources were used or consulted such as IEA, EWEA, IDAE, DGEG, REN.

The observed profitability of a typical wind-farm project in Portugal evolved over the last decade and converged in the last years to levels near cost of capital. Albeit a current front-runner and the current high share of wind electric output, Portugal was a late comer and in 2002 had less than 400 MW vs. 4.500 MW installed late 2011. For that matter stimulus and European-wise competitive remuneration was settled in 2002, in order to attract investment and guarantee the increase of renewable share in electric production and accomplish European targets.The remuneration framework was a success, resulting in a fast pace of wind-farm promotion and quick growth of the installed wind-farm capacity. At the same time wind-farm technology stabilized and matured during the decade, resulting in higher performance WTG at stable prices (in 2007-2009 an equipment shortage resulted in +20 p.p. inflation in WTG prices). In line with these events remuneration and attribution schemes were reviewed resulting in lower remuneration for promoters and lately curtailing profitability and slowing down installation pace.Real IRR is computed by discounting the project cash-flows at constant prices and real WACC is derived by subtracting the implicit project life inflation from the risk-free rate.Benchmarking of Portuguese, German and Spanish curves for average wind-farm IRR evolution during 2000-2010 reveal a similar pattern, with an expansion phase of promotion of the technology sustained on higher returns, followed by a stabilization phase of lower returns.Therefore it is fair to conclude that the typical profitability observed in the Portuguese Wind-farm projects followed the same pattern and range of the European references such as Germany and Spain.

3

Wind-farm projects (real IRR-WACC)- Portugal -

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

IRR (%) 6% 7% 7% 8% 7% 6% 6% 6% 5% 6% 5%

IRR - WACC (%) -0,9% 0,4% 1,0% 1,4% 1,2% 0,3% 0,6% -0,3% -0,3% 1,0% -2,6%

Tariff reduction (€/MWh)

- -2 -6 -9 -8 -1 -4 - - -6 -

Production (GWh)

- ~370 ~500 ~2150 ~1700 ~700 ~1350 ~1110 -

Margin cost (M€)(@ 2011 prices)

- -0,8 -3,5 -21,1 -15,1 -1,0 -5,1 - - -7,1 -

20102009200820072006200520042003200220012000

SpainGermanyPortugal

Wind-farm projects (IRR-WACC)- International benchmark -

2. Utility-scale PV projects

Main variables for determining the IRR of Utility-scale PV projects are total investment costs (in addition to the module, structure and installation, it is assumed that the inverter has to be replaced after ten years of operation, with a cost of nearly 30% of the initial investment), yearly fixed and variable O&M costs (tend to converge to approximately 1% of the initial investment), average load factor (depends on sun irradiation, that varies widely between countries and regions, and also on the efficiency of the device, systems with



axis trackers should attain some 10% more output), the lifespan of the facility and the existing legislation (regulatory scheme, attributable remuneration length and amount and taxation); based on available data and sources, the following assumptions have been derived with the purpose of characterizing the average yearly PV installation:

- The LNEG provided an investment historical evolution curve for the module price, based on existing projects; the data to set an estimation of the historic price evolution of PV installations over the last decade, contained between 2,7 (Current) and 6,0 (Highest historical value) M€/MW; IEA and EPIA data has been used to derive similar curves for relevant foreign countries, for benchmarking purpose.

- Average load factor has been assumed from observed historic electricity production and sector and experts interviews to be 1.450 - 1.900 (this value for a specific PV installation) nominal power equivalent hours. Yearly decrease of 1% in production due to material degradation.


- Average remuneration granted has been computed based on the historic national legislation for PV production.

- For cross-checking purpose other sources used or consulted such as E.value, Kema and Ecofys.

The observed profitability of a typical PV project in Portugal evolved over the last decade, starting in a non-supportive remuneration stance since technology was considered expensive when compared to substitute renewable. In 2005, remuneration changes fostered the installation of PV facilities and late 2011 already ~150 MW are installed; likewise, fostered utility-scale facilities such as Amareleja with ~50 MW and high load factors. More recently (2007 and 2010)

5

the remuneration scheme was adjusted to avoid over-paying due to PV stepped LCoE curve.Benchmarking of Portuguese, German and Spanish curves for average PV IRR evolution during 2005 (first year with licensed PV facilities in Portugal) -2010 revealed that the Portuguese remuneration scheme followed the German reference and was generally below Spanish target.Therefore it is fair to conclude that the typical profitability observed in the Portuguese PV projects converged to ranges of the European references such as Germany and Spain.

Utility-scale PV projects (real IRR-WACC)- Portugal -

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

IRR (%) -1% -1% 1% 3% 4% 9% 3% 4% 5% 8% 7%

IRR - WACC (%) -8,1% -7,2% -5,5% -3,4% -2,2% 3,3% -2,3% -2,5% -0,8% 2,8% -0,9%


- - - - - -73 - - - -63 -

Production (GWh)

- - - - - ~93 - - - ~26 -


- - - - - -7,3 - - - -1,7 -

200520042003200220012000 20102009200820072006

SpainGermanyPortugal

Utility-scale PV projects (IRR-WACC)- International benchmark -



3. Biomass projects

Main variables for determining the IRR of Biomass projects are total investment costs (rather similar to common thermal power plants), yearly fixed and variable O&M costs (with added cleaning and storage needs), average load factor (mainly depends on the availability of the resource; forestry residues and industrial by-products have been considered within this analysis), the lifespan of the facility and the existing legislation (regulatory scheme, attributable remuneration length and amount and taxation); based on available data and sources, the following assumptions have been derived with the purpose of characterizing the average yearly biomass installation:

- The technology is considered to be at a mature stage of development, with a moderate slope of investment costs decrease; as per the data for the capacity deployed over the last decade provided by LNEG, the initial investment would be close to 3 M€/MW; as well investment data was cross-checked with IDAE, Ecofys and DGEG (database of specific projects deployed in Portugal).

- Average load factor has been assumed from observed historic electricity production and sector and experts interviews to be ~6.000 nominal power equivalent hours. This is the average verified load factor in Portugal.

- The technology bears a relatively higher risk profile in comparison with other renewable sources, as the fuel adds uncertainty over the variable O&M component (~50% of the considered O&M), because of both the scarcity of the fuel

7

resource and its dependence on the oil price, as transportation is the single most relevant cost; economical sustainability of the facility implies proximity to the resource (~30 km), either forests and woods or industrial clusters (e.g. pulp and paper).


- A PCI of 14 GJ/ton and a heat conversion ratio of 13 MJ/kWh have been considered as energy intensity, and fuel cost of prepared biomass delivered at plant site in excess of 30€/ton. For industry biomass by-product it was considered a 20€/ton value assumption.

- For cross-checking purpose other sources used or consulted such as IDAE, Ecofys, Univ Coimbra and Centro da Biomassa para a Energia (CBE).

The observed profitability of a typical Biomass project in Portugal only started to be fairly attractive after the 2005 tariff increase which allowed some coupling with fuel prices; and simultaneously attract new investor and deploy capacity. Still the tariff revision was too rigid and did not fully followed fuel increase in the last years hampering new capacity development and tightening existing plants economics.

Biomass projects (real IRR-WACC)- Portugal -

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

IRR (%) 4% 5% 6% 6% 8% 8% 8% 8% 8% 8% 8%

IRR - WACC (%) -3,2% -2,1% -1,0% -0,1% 1,3% 1,9% 2,4% 1,7% 2,2% 2,8% -0,4%


- - - - - - -11 - -9 - -

Production (GWh)

- - - - - - ~90 - ~420 - -


- - - - - - -1,0 - -4,2 - -



4. Small Hydro projects

Main variables for determining the IRR of Small Hydro projects are total investment costs (albeit with less environmental impact, less scalable than regular large scale hydro plants; depends on the grid connection and fall height), yearly fixed and variable O&M costs, average load factor (depending if reservoir or run-of-river), the lifespan of the facility and the existing legislation (regulatory scheme, attributable remuneration length and amount and taxation); based on available data and sources, the following assumptions have been derived with the purpose of characterizing the average yearly Small Hydro installation:

- The technology is considered to be at a mature stage of development, with a moderate slope of investment costs decrease; as per the data for the capacity deployed over the last decade provided by DGEG (compiled data from promoters), the initial investment would be in-between 1,6 – 2,0 M€/MW; as well investment data was cross-checked with IST’s range of values.

- Average load factor has been assumed from observed historic electricity production and sector and experts interviews to be ~2.800 nominal power equivalent hours. This is the average verified load factor in Portugal.


- IDEA and Ecofys estimations for the O&M evolution have been considered to be ~40 k€/MW.year in 2010.

For cross-checking purpose other sources used or consulted such as IDAE, Ecofys, IST and DGEG.

9

The observed profitability of a typical Small Hydro project in Portugal grew significantly with the 2002 and 2005 tariff uplifts, enabling attractive remuneration and fostering capacity deployment.Due to lack of information from Small Hydro projects with license attributed before the year 2000 in what concerns their historic remunerations, average capital investment and operations costs, their profitability is outside the scope of this analysis; if we were to consider the hypothesis that those installed 1997 onwards (currently 15 years old or less) would be applied the same tariff reduction that matches the 2002 Small Hydro producers IRR with their cost of capital (~7 €2002), with a rough estimation of target installed capacity of approximately 60 MW and considering an average of 2.300 nominal power equivalent hours (reported average of small hydro < 10 MW production 1997-2000), that would translate into total yearly of approximately 1 M€ (at 2011 prices). Special caution should be taking when drawing conclusions from this result, as it derives from a high-level hypothesis. Even when taking into account those facilities, still some 60% of the total small hydro currently operating were not assessed in this work.

Small Hydro projects (real IRR-WACC)- Portugal -

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

IRR (%) 5% 6% 7% 7% 8% 8% 8% 8% 9% 9% 9%

IRR - WACC (%) -0,8% 0,0% 0,9% 1,4% 2,3% 2,4% 2,9% 2,5% 3,2% 3,7% 0,9%


- - -7 -10 -16 -16 -20 -17 -22 -24 -

Production (GWh)

- - ~20 ~40 ~5 ~5 ~15 ~100 ~5 ~5 -


- - -0,1 -0,5 -0,1 -0,1 -0,3 -1,7 -0,1 -0,1 -



5. CHP projects

Main variables for determining the IRR of CHP projects are total investment costs (although there is a broad range of available CHP devices, as a function of installed capacity, technology and fuel type, there is negligible variation in terms of investment costs), yearly fixed and variable O&M costs , average load factor (trade-off between thermal and electric efficiencies and overall global efficiency should tailor the site specific heat needs), the lifespan of the facility and the existing legislation (regulatory scheme, attributable remuneration length and amount and taxation); based on available data and sources, the following assumptions have been derived with the purpose of characterizing the average yearly CHP installation:

- For initial capital investment assessment over time, current COGEN CapEx estimations have been crossed with historic CHP business plans stored at DGEG in previous licensing processes, yielding values circa 1 M€/MW.

- Fixed O&M costs vary between 40 to 70 k€/MW.year and variable O&M range in a 5 p.p. band around the median value of 10€/MWh, depending on the plant and its dimension, according to COGEN and IEA.

- The most common fuels are diesel, biomass and natural gas, with the latter costing approximately 30 €/MWh to CHP producers as of 2011 year end, according to DGEG’s statistics.

- Average historically observed load factor is of around 6.000 nominal power equivalent hours for large scale (> 10 MW) CHP facilities and approximately 4.000 hours for smaller producer (< 10 MW).

The observed profitability of a typical CHP plant was strongly influenced by current legislation framework, released in 2002, allowing the producers to sell the totality of their outputs at an attractive remuneration, impelling further and bigger capacity

11

installations and increasing existing facilities’ margins; a new framework is expected to be released in the short term to foster heat optimal usage and reduce externalities over electrical production (Law is already published, still pending regulation issuing).Due to lack of information from CHP facilities built prior to 1997 (those that should be obliged to endorse in the new regulatory framework immediately), especially in what concerns their historic remunerations, capital investment and operations costs, and given the considerable change in the technological profiles of those devices, their profitability is outside the scope of this analysis; if we were to consider the hypothesis that they would be applied the same tariff reduction that matches the 1997 CHP producers IRR with their cost of capital (7,7 €1997), with a rough estimation of the installed capacity in 1998 of approximately 500 MW and considering an average of 5.000 nominal power equivalent hours, that would translate into total yearly of approximately 27 M€ (at 2011 prices). Special caution should be taking when drawing conclusions from this result, as it derives from a high-level hypothesis.

CHP <10MW projects (real IRR-WACC)- Portugal -

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

IRR (%) 12% 12% 11% 12% 13% 14% 14% 14% 14% 13% 14% 13% 13% 14%

IRR - WACC (%) 5,7% 6,1% 5,1% 5,6% 6,8% 7,7% 7,9% 8,4% 7,9% 7,8% 7,6% 7,1% 7,9% 5,1%


-8 -8 -7 -8 -10 -12 -14 -15 -16 -18 -17 -20 -19 -

Production (GWh)

~300 ~100 ~25 ~170 ~40 ~170 ~50 ~15 ~5 ~30 ~60 ~50 ~20 -


-3,2 -1,1 -0,2 -1,6 -0,5 -2,3 -0,8 -0,2 -0,1 -0,5 -1,1 -1,0 -0,4 -



CHP >10MW projects (real IRR-WACC)- Portugal -

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

IRR (%) 13% 13% 14% 14% 14% 13% 12% 12% 11% 11% 11%

IRR - WACC (%) 6,1% 6,6% 7,6% 7,7% 8,2% 7,4% 6,5% 5,4% 5,0% 5,3% 2,4%


- - -8 -9 -8,9 - - -7,8 -9,2 - -

Production (GWh)

- - ~75 ~50 ~700 - - ~780 ~460 - -


- - -0,7 -0,4 -6,3 - - -6,5 -4,4 - -

13

APPENDIX 5

COMMENT ON ELECTRICITY PRICING POLICY OPTIONS



Comment on electricity pricing policy options

João Confraria

1. The problem

This paper addresses the economic effects in the electricity market of two policy options for electricity prices from 2012 to 2020. In both cases the tariff debt is eliminated in 2020. Tariff debt is equal to the sum of annual tariff deficits adjusted by the appropriate interest rate, as defined by law. The policy options differ basically on the rate of price change and on eventual complementary policies to reduce the growth rate of electricity prices.

i. Option 1 implies nominal electricity prices growing at the average annual growth rate of 4.7%. It supposes the elimination of tariff debt by 2020 on a business-as-usual scenario, with strong impacts on the competitiveness and burden on families.

ii. Option 2 is characterized by nominal electricity prices growing at the annual rate of 3.4% and also aims no tariff debt in 2020, which means that it may be complemented by additional electricity policy measures, as set in SEE (2011), in order to mitigate the above mentioned impacts.

Option 1 is the initial scenario considered for electricity prices. It does not require any policy change. In both cases it is assumed that prices are always above short run marginal costs and short run variable costs. This is a reasonable assumption given the share of fixed costs in total electricity costs.The need for a policy change, from option 1 to an alternative along the lines of option 2 has been set in the Memorandum of Understanding of May 2011, signed by the Government of Portugal, the IMF, the European Commission and the European Central Bank,



as well as on additional commitments of the Portuguese Government in the Letters of Intent of December 2011. The analytical framework of this paper is the economic theory of regulation. The effects of different policies on market surpluses and on investment incentives are considered in section 2, and political and regulatory dynamics arising from both scenarios are the subject of section 3. The current legal framework, the legal foundations and the legal consequences of eventual policy changes are not considered here.2. Changes in market surpluses and investment incentives

The evaluation of the effects of both policy options on electricity markets is based on the following assumptions:

i. Initial prices, for 2011, were estimated as the weighted average of prices paid by household and business end users in each one of the bands defined in Directive 2008/92/EC, on price transparency, the weights being the share of each band in total electricity consumption of each group (household and business).

ii. For household end users, the own price elasticity of demand was assumed to be -0.2 and the income elasticity of demand 0.3, in line with the average of most estimates of short run electricity household demand (Kriström, 2008, Espey and Espey, 2004). For industrial end users available results from currently available estimates seem less consistent, and of course dependent on the energy intensity of each industry and on each country economic structure. Given this, it was assumed that the market elasticities of demand were identical for business end users and for household end users. In both cases eventual effects of electricity price changes on demand for other energy sources were not considered.

iii. All values are at nominal current prices.iv. Assumptions on GDP growth and inflation rates are as in SEE

(2011).



Main results are presented in table 1.In 2012 option 2 will lead to electricity prices 1.2% lower than option 1. The difference increases over time and in 2020 under option 2 electricity prices will be 10.6% lower. End users gains increase from €84.6 million in 2011 to 1047.6 million in 2020. As a percentage of annual electricity bills, they increase from 1.3% in 2012 to 10.8 % in 2020 (in this case, given the assumptions on elasticities, per cent values are the same for household and business end users). As usual, given the own price elasticity of demand, almost all of these end users gains are reductions in the electricity bill. Net welfare changes amount to just 1% of total end users gains in 2020. In 2011 the value is even lower, 0.1%, as the difference in prices is smaller than in 2020. In the long run net welfare change will be larger and net end users gains will be smaller, as they are able to adjust their consumption patterns to price increases under option 1. The decrease in consumption will be larger than in the short run. The fall in prices under option 2 will be followed by a larger increase in consumption than in the short run. For instance assuming long run own price and income elasticities of demand of -0.9 and 1 (Espey and Espey, 2004) net welfare change increases from 0,5% to 4.5% of total end users gains. Total end user gains increase from 85 million euros in 2011 to 967 million in 2020. The reduction of the electricity bill of end users will be lower than in the short run over the same years (table 2).



It should be stressed that household elasticities may depend on variables as age groups of household members, income, appliances in the household. If so, option 2 and option 1 may have an impact on income distribution. It is not discussed for lack of data.



Business demand is also subject to some variation among industries and according to firms’ technologies. Evidence suggests that even after controlling for differences in industries, an increase in the electricity intensity at the firm level implies a substantial increase in the (absolute value of) own price elasticity of demand (Bjorner et al, 2001). In the Portuguese industry, electricity intensive sectors include chemicals, steel or paper. We should expect, of course, these industries were more negatively than the average firm under option 1 vs. option 2. It may also happen that the companies do not manage to remain competitive at higher levels of electricity prices, and they may close and the industries disappear. This might not be policy problem. Many firms are closing at any moment, simply because they are inefficient. However, this is not necessarily the case here. As argued below, current electricity prices are higher than their economic cost and some producers are receiving what might be called excessive rents. If so, electricity intensive firms may lack competitiveness not because of any problem they could solve, simply because they compete in international markets and cannot transfer to their customers the excess value they must pay for domestic electricity. In this case the economic distortion implied by higher electricity prices may be larger than suggested in tables 1 and 2, for the sectors involved. From 2012 to 2020 end users will of course be able to invest to adjust to price changes. Therefore the estimates based on short run price elasticities can be considered as an upper bound on the value of reductions of the electricity bill and a lower bound on net welfare gains, given option 2 versus option 1. In 2020, assuming an interest rate of 6%, the cumulative value of the reduction of electricity bill, under option 2 versus option 1, will be in the range 4943 million euros (assuming long run adjustments in electricity use) to 5424 million euros (assuming short run adjustments).Policy measures under option 2 will have negative effects on producers incomes if there are no offsetting measures, as they will



bear the burden of rebalancing the domestic electricity system eliminating the tariff debt in option 2. Let´s assume for the moment that there are no offsetting measures. In a short run framework there are no major efficiency shortcomings as long as prices under option 2 are above short run marginal costs21.In this case, producers’ losses will always be less than consumers’ gains. However, it is clear that the return on assets will be lower, unless companies manage to transfer most of the losses to workers and suppliers. This transfer is not likely in the short run, although in the long run some adjustments are certainly possible, and desirable, if current contractual arrangements with workers and suppliers for the provision of the current level of services are less demanding for them than alternative contractual arrangements for the provision of the same services under a more competitive electricity industry structure.However, a large part of the reduction of the electricity bill to be paid by end users from 2012 to 2020 will not be a burden on producers. Under policy options considered in SEE (2011), additional resources will be available reallocating to the consumer the revenues generated by electricity companies in the CO2 market. Additional revenues will come from the regulation of the current legal framework for co-generation. By 2020, the cumulative revenues from these sources will amount to 2876.3 million euros. As they will be used to finance lower prices, the total burden on producers is estimated in the range 2066.8 million to 2547.8 million euros. This will be the cumulative value of the transfer of surplus from producers to consumers under option 2.In any case, under option 2, the return on equity falls. This loss being the result of policy changes a policy and regulatory credibility problem may be created.

21 To keep things simple, in the following discussion short run marginal costs are assumed constant and equal to short run average variable costs and long run marginal costs equal to long run average costs and constant. Fixed costs are assumed to be sunk.



This credibility problem will be obvious if there are sunk costs and if prices are above marginal costs but below total average costs. In this case it would still make sense to continue operations, from the producers’ point of view. However, investors would doubt the commitment of the State on making sure that they get an adequate return on capital invested in production and network facilities. The value of the assets would depreciate fast, as firms would avoid maintenance and replacement investments, because prices were not enough to pay for that. Consumers would also lose in the long run, because of lower quality of service and, in the end, because electricity production capacity and production would fall (Newbery, 1999).However, it seems that in the case of option 2, the Portuguese Government does not have this type of credibility problem. In the Memorandum of Understanding on Specific Economic Policy Conditionality, of 17 May 2011 (MoU), the Government committed to take measures to limit the additional cost associated with the production of electricity, in particular through negotiation or revision of the guaranteed compensation mechanism (CMEC) paid to producers under the ordinary regime and the remaining long term power purchase agreements (PPA). The Government also committed to review the efficiency of support schemes for renewable and co-generation, aiming at achieving lower tariffs.Recently, in the Memorandum of Economic and Financial Policies, attachment to the Letter of Intent to the IMF, of December 9, 2011, the Government committed to take bold steps to reduce excessive rents in energy markets, and to propose measures covering all sources of rents. The Government also recognized that current arrangements were not consistent with the sustainability of the current national electricity system. In the Memorandum of Understanding on Specific Economic Policy Conditionality, attachment to the letter to the Eurogroup, the European Commission and European Central Bank, of December 9, 2011, the Government committed to limit the additional costs associated with the



production of electricity under the ordinary and special regimes, as well as reviewing the efficiency of support schemes for generation and for renewable.So, according to these (strong) commitments, the Portuguese Government states that there are problems in the electricity markets associated with excessive rents, additional costs, lack of efficiency of some support schemes and, ultimately, lack of sustainability of the national electricity system. The problem is to know the meaning of concepts as excessive rents (the concept of additional costs is even more ambiguous, in this context).Rent may be a confusing concept. A rent can be defined as a payment for use of a resource, whether it be land, labor, equipment, ideas, even money (Alchian, 1987). Accepting this, excessive rents in the domestic electricity industry might be considered as payments to the owners of resources in excess of the minimum demanded by these owners to keep the resources employed, that is in excess of their opportunity costs. In this case the problem is to know if we are considering short run or long run opportunity costs. Considering short run costs would create the credibility problem outlined above. Excessive rents might include the returns on capital invested in sunk costs in electricity production, as in the short run they exceed strictly the payments necessary for continuing production22. However, investment would stop and capacity and production would decrease. That would not be efficient and of course the electricity system would not be sustainable, for lack of investment. So, we should assume that the concept of excessive rent implicit in the Portuguese Government commitments relates to payments to the producers exceeding the payments

22 Confusingly, a rent is also defined by others as a return above the amount necessary to keep a resource in its current employment at least in the short run (McChesney, 1998). This concept of rent is close to the concept of “excessive rent” outlined above. It is not used here. Because it creates a huge ambiguity in the concept of “excessive rents” it was not certainly the relevant concept of rent as understood by the Government.



necessary to cover the opportunity costs of the resources employed, including an adequate return on capital invested.

The evaluation of opportunity cost of resources presents further problems. What are the resources to be considered? Does it mean the opportunity cost of resources as they are today or the opportunity costs of resources at the time the decisions were taken? In general, excessive rents may be created in anyone of the markets where resources are bought: labor markets, equipment and intermediate consumption markets as well as in capital markets. This view seems to have been adopted by the Portuguese Government, that committed to review all sources of rents in the Memorandum sent to the IMF the Government. However, at the same time, the Government considered that there were high rates of return granted to generators in the standard and special regimes representing a significant share of electricity bills, suggesting that at least one source of excessive rents were presumably the returns on capital invested. If so, the current return on the capital invested in some electricity producing units is above the opportunity cost of capital.23

Concerning the evaluation of opportunity costs, if they are evaluated according to their current value, a new source of regulatory risk may be created. If the current value of the opportunity costs are used to consider that there are excessive rents and if this is followed by a price review, the Government will be changing the nature of the contractual regulatory process, at least implicitly, increasing regulatory risk. To avoid this, the evaluation of excessive rents should be done evaluating eventual differences between the return on a resource initially contracted and the opportunity cost of that resource at the time the contracts were made, evaluated according to the information available at that time. For instance, if the resource is equity, rents should be given by the difference between actual returns on equity (rate of return x value of equity) and the returns on 23 In this case of course, previous Government decisions contributed to the creation of excessive rents.



(the same level of) equity calculated using the opportunity cost of equity found in the review. If we are considering the opportunity cost of capital rents should be given by the difference between actual returns on capital and the returns on the same level of capital using the opportunity cost of capital found in the review.The same reasoning should apply to other resources, following the eventual evaluation of all sources of rents by the Government.24

To keep things simple in this reasoning let’s assume the definition of excessive rents suggested above, that excessive rents are earned on capital invested and that capital costs are sunk costs. In this case under option 2, the first problem for the government is to see if the cumulative value of surplus transferred from producers to consumers, relative to option 1, estimated in 2020 in the range 2066.8 to 2547.8 million euros is equal to or smaller than the value of excessive rents. If so, there is not, in principle, a credibility problem, in the sense outlined above, because the prices are always large enough to pay for the opportunity cost of capital. However, there may be some unintended redistribution of income if new shareholders bought stakes in the projects involved. The shareholders that sold their stake in the past received their share of the value of excessive rents and, under option 2, current shareholders will lose.In the opposite case, the Government creates a credibility problem. Shareholders do not receive at least the opportunity cost of equity. Eventually bondholders will also lose if the market value of the firms’ debt decreases. Given this, the Government should define additional policy measures offsetting these losses.

24 This discussion has been focused on prices of resources, not on the quantity employed of each resource. Including quantities in the analysis would imply a more general appraisal of efficiency in electricity production that seems to be beyond the scope of the Government commitments. Although the Government intends to review the efficiency of support schemes there is no commitment to make a detailed review of all the factors related to the efficiency of investments made and of production facilities in place.



In anyone of these two cases, further negative effects might follow. Firms affected by this policy have business in other markets, not subject to domestic regulation. Probably they have been using the money earned in the domestic regulated businesses to finance investments in non regulated markets or in regulated markets in other countries. They may face additional costs of financing the other activities as the returns on domestic regulated activities are diminished. To reduce this problem, the Government should allocate the maximum available resources generated by carbon trading to the electricity industry to reduce the transfer of surplus from producers to end users of electricity, as policy shifts from option 1 to option 2.Anyway the transfers of surplus to end users under option 2 should be considered as exceptional, and a result of current economic conditions, notwithstanding past policy errors at the source of excessive rents. However there is an additional argument for option 2. It involves appropriate risk management by the State. Under option 2, of course, there is always a risk that the reduction or elimination of excessive rents has a negative impact on the cost of debt as well as on future investment. But under option 1 there is the risk that the tariff debt will not be paid, as it is not sustainable for instance under some scenarios on oil prices and droughts in Portugal. Additional political risks should also be considered under this scenario. They will create burdens for the electricity producers. That is discussed below.3. Risk regulation and regulatory risk

The contractual arrangements with renewable energy producers and with the former monopolists that are currently considered by the Government as the source of the excessive rents in electricity markets, were the result of previous government intervention in electricity markets that was beyond the correction of market failures namely natural monopoly, negative externalities and asymmetric information, and of some equity objectives relative to improving the availability of electricity to everyone in the country. They were part



of policy framework with a strong interventionist flavor, aiming at the promotion of specific technologies in renewable energy production, of maintaining an adequate return on assets that became underused or obsolete because of renewable electricity growth, regulation or liberalization, of developing domestic industry and, eventually, supporting the internationalization of domestic companies. The costs and most of the risks associated with these policies would fall on end users. Ideally the impact would be felt over time using whenever necessary the tariff deficit system. Tariff deficit is basically defined by a set of rules allowing each year regulated prices less than regulated costs, the difference being consumer debt to be paid in the future, and earning an interest rate set by the Government. Consumer debt was to be created in case of some contingencies. Examples are high oil prices, or a combination of high oil prices with additional costs of renewable production, arising from unfavorable weather. Given the current levels of tariff debt, contingencies as higher oil prices or severe droughts may lead to unsustainable levels of tariff debt (SEE, 2011).This means that Government intervention created additional risks. Rules shifting risks to end users actually create the risk of a political backlash against them. That happens in a democratic society because these risks may well be the source of voters’ discontent. And of course winning political platforms are built on voters’ discontent. The interesting points to be highlighted in this case is that all voters are electricity consumers and that at least part of the source of their discontent are high prices caused by excessive rents recognized as such by the Government.25 To get their support political parties may well propose policies that impose increasing and excessive burdens on electricity firms.It should be noted that in the domestic market, the problem is not only price growth resulting from option 1. Currently end users have a perception that the prices they will be from now on will be above 25 EDP may particularly affected in such cases. Almost all voters are EDP customers.



EU average, if evaluated at new VAT rates as well as in PPP.

Comparisons based on purchasing power are certainly not appropriate to evaluate and compare efficiency in production, but they are good enough to evaluate the burden on families and businesses. Until now electricity prices paid by end users have been relatively low, not only because of the tariff debt, but also because of low VAT rates, relative to current VAT rates in most EU countries. In a sense part of the costs of the system were shifted from energy users to taxpayers, at least relative to most EU countries. In the end prices paid by domestic end users were until now below EU average. However this is not sustainable. With the increase in VAT rates, now in line with most EU countries, prices paid by domestic households become higher than EU average. If the tariff deficit created in 2012 is included they become even higher. And if the purchasing power of the population is considered prices paid by domestic end users are among the highest in the EU (tables 3 and 4).Moreover it is not clear that it will be feasible to maintain financing conditions for EDP if there are further increases in tariff debt, even if that is possible given current commitments of the Government, Until now, tariff debt has been sold by EDP in the market, avoiding short run liquidity problems. In the event of large increases and given the current situation of financial markets, it is not obvious that EDP will be able to sell it. In this case it is likely that the firm will have increasing difficulties to borrow in international markets. Borrowing costs will increase with negative impacts on prices. Even if EDP manages to sell tariff debt, the buyers are likely to be domestic banks. To do that, they will reduce credit available to economy. It is possible that given the amount involved this will be the source of significant recessionary effects. In these events it may well happen that the tariff debt will become politically unsustainable, basically because end users are strongly against it and the political system may well respond accordingly.



An additional negative effect may be the erosion of goodwill accumulated in the past by electricity producers. It may be difficult to explain to consumers prices that must be considered relatively high. So much that the Government has already considered they were the source of excessive rents that should be eliminated according to the commitments with the IMF.So, the policies on electricity production have basically increased the likelihood of adverse reactions in regulatory policies imposing potentially huge losses on producers. Government intervention in these cases involved options on the allocation of risks between end users and producers as well as the management of societal risks, related to the environment, self sufficiency and regulatory policies. It is not obvious that the Government was well prepared to do it. Risk regulation involves an appropriate methodology on the evaluation and management of risk. The point is to avoid the absence of necessary regulation, as well as regulations that do not improve welfare. The benefits, costs and risks created by Government intervention should be evaluated (Bounds, 2010). According to available evidence the Government was not involved in this type of analysis when launching policies as those mentioned above.Liberalization and competition might reduce some of these regulatory risks, as long as they have the potential to decrease prices. It does not seem likely in the short run. It is well known that moving from a regulated monopoly with prices oriented to costs to a liberalized oligopoly may well lead to higher prices. Adaptive behavior of oligopolists, as in a Cournot oligopoly leads to this outcome (Friedman, 2002; Joskow e Kahn, 2001). It should be noted that significant improvements in production efficiency relative to the monopoly situation may offset the consequences of such adaptive behavior. Anyway such improvements are not likely in the short run. In the long run, entry of low cost producers should be an important driver of price decreases. Again, such outcomes may not happen in the immediate future. Most entrants were renewable electricity



producers that were not the lowest cost producers. Eventually they will become more competitive in the event of higher carbon prices or higher oil prices. In any case all electricity sold to end users includes the cost of the tariff debt, as it is included in access tariffs. So, even if the entrants become low cost producers, retail prices may be so high that they cannot avoid an adverse political reaction against paying the tariff debt, or against other features of the current policy framework.







4. Concluding remarks

Under electricity pricing of option 2 there is a transfer of surplus from producers to consumers, as well as relatively small increase in allocative efficiency. This transfer of surplus should be evaluated first in the framework of the commitments made by the Portuguese Government in December 2011 to eliminate excessive rents in the electricity industry. If the cumulative transfer of income from producers to consumers is smaller than the cumulative value of excessive rents, prices under option 2 will always be enough to cover the opportunity cost of resources used in electricity production. In particular, shareholders will always receive a return at least as large as the opportunity cost of their capital. If the cumulative transfer of income from producers to consumers, under option 2, would be larger than the value of excessive rents, the Government should use revenues from other sources, as carbon trading of electricity companies, to compensate the difference, and making sure stakeholders always receive the opportunity cost of their equity. Even if the Government makes sure that prices will always cover the opportunity cost of resources employed in electricity production, it can be argued that option 2 may increase the cost of capital of electricity firms. Currently producers may be using some of the excessive rents to finance other operations in other markets, and so it may difficult for them to get additional loans, or to avoid increases in the cost of debt, in particular in the current market environment. Investors may lose some confidence in the system, in particular if they see excessive rents as an inalienable right. Unintended transfers of income may also happen, when the current holders of stock are different from the owners at the time the excessive rents were built into the system. To minimize these effects the participation of shareholders (and eventually other stakeholders, if it is found they are also the source of excessive rents) should be



reduced using other sources of funding. Again, revenues from carbon trading of electricity companies may be a source of a significant amount of resources to be allocated with that purpose. However these risks should be compared with risks under the alternative policy option. Option 1 is a source of political and regulatory risk. Under this option, end users will bear the burden of eliminating the tariff debt and they will become increasingly aware of the existence of excessive rents, as already recognized by the Government. At the same time the prices they will be paying will be among the highest in the EU, considering PPPs. In case of higher oil prices or severe droughts the increase in tariff debt may unsustainable. As all voters are electricity consumers it is likely that policies adverse to electricity companies will become increasingly popular. Sooner or later some Government may adopt them as a response to voters demand. So, additional risks under option 2, created by the reduction or elimination of excessive rents, may well be offset by additional risks of adverse energy policy developments under option 1, as well as by the negative impact of “badwill” thus created by electricity producers.



References

Alchian, A., 1987, “Rent”, in The New Palgrave Dictionary of Economics, Londres, The MacMillan Press

BjØrner, T., Togeby, M., Jensen, H., 2001, Industrial companies’ demand for electricity: evidence from micropanel, Energy Economics 23 595-617)

Bounds, G., 2010, “Challenges to Designing Regulatory Policy Frameworks to Manage Risks”, in OECD, Risk and Regulatory Policy, Paris, OECD

Espey, J., e Espey, M., 2004, Turning on the lights: A meta analysis of residential electricity demand elasticities), Journal of Agricultural and Applied Economics, 36:65-81.

Friedman, L, 2010, The Microeconomics of Public Policy Analysis, Princeton, Princeton University Press.

Joskow, P, Kahn, E., 2001, “A Quantitative Analysis of Pricing Behaviour in California’s Wholesale Electricity Market during Summer 2000” NBER WP 8157.

Kriström, B., “Residential Energy Demand” in OECD, 2008, Household Behaviour and the Environment Reviewing the evidence, Paris, OECD.

McChesney, F., 1998, “Rent from regulation”, in The New Palgrave Dictionary of Economics and the Law, Londres, MacMillan Reference Ltd.

Newbery, D., 1999, Privatization, Restructuring and Regulation of Network Utilities, Cambridge, The MIT Press.

SEE 2011, “Sustentabilidade do Sistema Eléctrico Nacional”.



APPENDIX 6

REFERENCES



CEPA - Cambridge Economic Policy Associates Ltd

Ian Alexander, CEPA Director

SummaryIan is a regulatory economist with over 18 years of experience in assisting Governments, regulators and companies address regulatory and private sector participation issues.He is a specialist in:

financial aspects of regulation (cost of capital, asset valuation); creating an environment for private sector participation (regime

design, options for private involvement); and regime design (treatment of investment, creation of incentives).

Ian has also been heavily involved in the development and implementation of regulation in the UK and other developed countries (Australia, Canada, Holland, Ireland, New Zealand and the USA). He has worked with regulators and companies on price reviews, competition issues and the development of regime issues. Further, he has worked across the whole range of regulated utility and infrastructure sectors including electricity, gas, telecommunications, transport (airports, rail and road) and water and is able to bring the experience from those sectors.In Northern Ireland Ian has worked with OFGEM and its predecessor on several occasions. This started with support to Ofreg during the MMC inquiry into NIE in 1997 and has recently focused on water charging issues in PC10. Ian also worked on water reform and charging issues for SIB. More generally in Ireland, Ian has worked on energy issues with CER (gas), the joint Regulatory Authorities (on transmission charging issues within the common arrangements for gas), EirGrid (electricity tariffs) and the Commission for Aviation Regulation (on an interim review for Dublin airport as well as the last air traffic control price review).Recently much of Ian’s work has focused on regulatory reviews. He led the CEPA team supporting Ofgem on examples of alternative regimes in the UK and elsewhere as an initial input to the RPI-X@20 review and has subsequently supported Centrica on related issues. With UKWIR, the UK water industry research body, Ian has been leading a project on whether the existing regulatory regime is fit for purpose. Work over the past couple of years with British Airways has focused on ways in which the existing airport regulation regime could be improved – linked to both the Department for Transport and Competition Commission reviews and Ian led support to IPART, the multi-sector regulator in New South Wales, Australia, on whether its regimes were best practice. Finally, Ian is a member of the academic/review panel that Ofwat has established to consider Future Regulation.Selected Experience

1

Ian is leading a CEPA team advising EirGrid, the Irish electricity transmission company, on preparation for its next five year price control. This addresses key areas of finance and operations, especially important for an asset light transmission system operator.

Ian led a small team from CEPA working with Ofgem on establishing an overview of where incentive based regulation has been undertaken and the precise form of the regime. This work was extended to include an analysis of further institutional issues linked to some of the sectors under consideration to determine whether other factors also affect the ability of implementing lessons from those sectors.

Ian led the CEPA team providing support to Ofgem on input prices and the possible options for addressing uncertainty. The workstream Ian led is looking at the design of a mechanism to address uncertainty should it exist.

Ian is leading a small team working with the UK water industry research body to consider whether the existing regulatory regime is fit for purpose and what changes may be appropriate as the industry structure changes under the proposed increase in competition.

Ian lead the CEPA team advising IPART in New South Wales (Australia) on how its approaches to incentive based regulation compared with best practice. The project assessed how the regulatory regimes employed by IPART for water, transport and electricity compared against other Australian regulators and international best-practice. Separately Ian is working with IPART providing specialist input on aspects of the allowed rate of return. This is focusing on the question of how to address inflation when estimating the real cost of capital and also special issues raised by assets where the operational life is likely to be less than the technical life.

Ian led a small team from CEPA working with Ofwat on bulk pricing issues. This review has arisen owing to the various forms of bulk water sales that occur in England & Wales and the possibility that different pricing principles are being followed when considering each legal form of bulk transfer. The work is being undertaken to inform Ofwat of the possible need for reform and the options available should reform be needed.

Ian has worked extensively with CER in Ireland. He was director responsible for the consortium providing support on the gas transmission and distribution price review. He lead two other gas projects underway with CER – on the Bellanaboy entry point for Corrib and the transmission tariff methodology under the Common Arrangements for Gas.

For nine months Ian was seconded to the Irish Commission for Aviation Regulation to provide support on the interim determination requested by Dublin Airport arising from the revised plans for Terminal 2. This work focused on the need for the project and its design – including a cost-benefit analysis – as well as more general regulatory advice on investment incentives, options etc.

Working with the QCA, Ian provided advice on the process to be followed to determine whether a provider should be price-capped.

Ian led a small team advising Water UK, the water industry body, on menu regulation and how it can be implemented as part of the next price review. This built on previous work CEPA undertook for Severn Trent on options for developing the regulatory regime in the UK water



sector. This included an analysis of the likely impact of different reform options on consumers, companies and other stakeholders – effectively an ex-ante impact analysis.

For the last airport price review Ian was part of the CEPA team advising British Airways on the regulatory and competition aspects arising from the London airports review. Advice on the need for regulation and specific elements of the regulatory control, such as financial aspects, were addressed through this work. Ian led a follow-up piece of work on options for reforming the regulatory regime as part of British Airway’s inputs to the various Government and Competition Commission reviews. Now Ian is working with BA on responding to the CAA consultation on the assessment of market power and possible alternative regulatory arrangements.

Ian led a small team providing specialist advice to the New Zealand Commerce Commission on asset valuation issues as part of a gas distribution determination.

As part of the 2007 gas distribution review CEPA has been advising, Centrica, a major user of the network, on aspects of the review. This has included analysis of cost of capital/corporate finance issues, efficiency, incentives and regulatory risk. Numerous submissions on behalf of Centrica have been published on the Ofgem website.

Working for two of the UK regulators (rail and water) Ian led a CEPA team considering the options for “indexing” some or all the allowed rate of return. This work includes an ex-ante assessment of the implications of any reforms including impacts on prices for consumers and the financial viability of the companies. Issues related to regulatory burden are also under consideration.

In preparation for the possible privatization of Schipol Airport the Dutch competition agency needed to assess whether the airport required a regulatory framework beyond that provided by general competition policy. Further, if explicit economic regulation was necessary what options existed? Ian led a team that assessed the passenger usage of the airport – own origination versus transfer etc – as well as the other transport options. Options such as dual or single-till regulation were reviewed.

Ian was part of a small team providing specialist advice on cost of capital issues during a gas interconnection dispute in Victoria (Australia) in 1998. This involved review of documentation, preparation of brief reports etc as an input to a broader consultancy support to the Division of the Treasury responsible for selling the gas distribution businesses.

During the contested takeover of South West Water in 1997, Ian was part of a team advising the company on regulatory implications of any takeover, especially in terms of the impact on the precision of the regulator to determine price limits.

During the period 1991-1999 Ian provided advice to companies in the water, electricity, gas, transport and telecoms sectors as part of their price reviews. This included specialist advice on the financial aspects of regulation (cost of capital, depreciation/economic life and asset valuation) as well as more general regulatory support.

Qualifications1988 BA (Hons) Economics, Churchill College, Cambridge

University

3

Employment History2005 – present

Director, CEPA

2002 – 2005

Senior Economist, South Asia Energy and Infrastructure Unit, The World Bank

2001 – 2002

Regional Coordinator, East and South Asia (based in Singapore)

2000 – 2001

Public-Private Infrastructure Advisory Facility, The World Bank

1999 – 2000

Managing Consultant, Frontier Economics, UK

1996 – 1999

Private Sector Development Specialist, Public Private Infrastructure Unit, World Bank.

1991 – 1996

Senior Consultant then Managing Consultant, London Economics, UKConsultant then Senior Consultant, OXERA, UK

Selected PublicationsIan has also published widely on regulatory issues. Over 15 papers have been published in various journals, conference proceedings and working paper series since 1996. These have included:The Regulation of Investment in Utilities: concepts and applications, Alexander and Harris, World Bank Discussion Paper 52, 2005.The Delhi electricity discom privatizations: Some observations and recommendations for future privatizations in India and elsewhere, Agarwal, Alexander and Tenenbaum, Energy and Mining Board Discussion Paper series, The World Bank, 2003.UK model on Developing and Transitional Economies: Common issues and Misconceptions, in The UK model of Utility Regulation: A 20th anniversary collection to mark the ‘Littlechild Report’ – retrospect and prospect, CRI, University of Bath, 2003.



Charles Groom, CEPA Director

SummaryCharles Groom joined CEPA in January 2006, prior to which he held the position of Director - Business Development for Globeleq, an emerging markets generation power company. With CEPA, Charles has continued to focus on economic and financial issues in the energy sector, and is currently advising Centrica on the appropriate allowed cost of capital for electricity and gas transmission and distribution. He recently advised DECC on cost of capital inputs for generators under the EMR proposals.Charles led CEPA’s advice to CER/ NIAUR on the costs of the Best New Entrant for the SEM for 2010, 2011 and 2012 and the CEPA’s advice on market power and liquidity in the SEM. He is Project Director of the consortium advising CER on the price review of Bord Gais Networks and NIAUR on its ongoing review of allowed revenues for NIE. Previously, Charles advised CER both on the production of the 2008 Gas Capacity Statement and on the financial regulation of electrical and gas contractors with respect to safety.With CEPA and previously with Globeleq, Charles managed several bids for the acquisition or development of independent power projects, including a pioneering gas-to-electricity project. As a former board director of several Independent Power Producers (IPPs), Charles is very familiar with project development and financing issues for IPPs. Charles has specialized experience in financial economics, corporate and project finance, business development, and investment management, skills which he has brought to bear in previous BNE calculations.Selected Experience

5

Fixed cost of a Best New Entrant Peaking plant (CER/NIAUR): project to determine the value of the Annual Capacity Payment Mechanism in the Irish single electricity wholesale market (determined by the fixed costs of a Best New Entrant (BNE) peaking plant). Charles led the project in 2010, 2011 and 2012 and provided specialist input on the economic and financial parameters applying to the notional BNE plant.

Centrica – RIIO-T1 and GD1: Charles is leading CEPA’s ongoing advice to Centrica’s gas distribution and transmission teams on their response to the RIIO strategy and implementation consultation documents. CEPA has drafted independent papers on financial issues and mechanisms for dealing with uncertainty, and advises Centrica on its own responses to Ofgem. CEPA works closely with the Centrica team, attending industry working groups and Ofgem meetings.

CER/NIAUR – Market power and liquidity in the Single Electricity Market (SEM): Charles led this project, which considered the sources of market power in the SEM, the need for mitigants and measures to promote liquidity. The project also involved a consideration of proposals for vertical and horizontal integration from ESB, and associated liquidity undertakings.

Department of Energy and Climate Change - Electricity Market Reform: CEPA is advising DECC on the detailed design and financial implications of its preferred low carbon generation support packages. Charles is leading the assessment of the cost of capital implications and availability of finance for different packages, especially the ‘classic’ Feed-in-Tariff and Contract for Difference, including ongoing consultations with market participants (suppliers, generators, banks, PE funds and investment banks).

Centrica – regulatory strategy: Charles was the Project Director for the CEPA team that provided strategic advice to Centrica on Ofgem’s most recent price control review of electricity distribution networks. Key areas of analysis have been: a review of performance against incentives; recommendations on the structure of the control, including re-openers; pension allowances and related incentives; and financial issues, including the appropriate allowed cost of capital. Charles led CEPA’s input to both Ofgem Authority meetings and working level meetings, including those managed by the Energy Networks Association.

Centrica - NTS Entry: Charles leads CEPA’s ongoing advice to review the performance of the existing NTS entry charging regime and develop options for reform. The issues considered to date include (i) the impact of European trading hubs and the NTS entry capacity regime on GB National Balancing Point prices (ii) impacts of TO commodity charges on current and future “discretionary” gas supplies (e.g. LNG) and implications for UK winter gas supply security (iii) historical and future gas demand and supply trends to the GB market and implications for future NTS entry capacity demand.

Commission for Energy Regulation, Ireland (CER) – gas transmission and distribution price control review: Charles is leading the CEPA-consortium undertaking the current review of the gas transmission and distribution revenue control. This follows the successful review by the CEPA-led consortium (which Charles project managed) which provided economic, technical and financial advice to



the CER in its 2006-7 review of the revenue controls. Key tasks have included: evaluating the theoretical, financial and technical assumptions underpinning the current revenue controls; reviewing the capital and operational expenditure of BGN during the current control period against allowed revenues; benchmarking BGN service provision, operating and capital investment activities; advising on cost of capital issues; designing BGN transmission and distribution revenue controls; and advising on maximum allowed revenues throughout the review period.

UK Electricity Distribution Network Owner – trigger mechanism for the cost of capital: as part of Ofgem’s recent electricity distribution price control review, Charles led CEPA’s advice to a DNO on the rationale for a trigger mechanism on the cost of capital. CEPA’s work included the modelling of a range of specific trigger mechanisms.

Office of the Rail Regulator (ORR): Charles was Project Director for the CEPA team advising ORR on cost of capital issues in the rail sector.

Commission for Energy Regulation, Ireland - financial and procurement advice: Charles directed the CEPA team that provided financial advice to CER in developing new systems for the regulation of the activities of electrical contractors and gas installers with respect to safety. CEPA was then appointed by CER to provide financial advice in developing the methodology for the designation process and in assisting in the operation of a competitive tender process to appoint the bodies, including designing a methodology for the submission/ evaluation of the financial aspects of designation, including advising on the procurement process, drafting financial sections of the tender documentation and supporting the CER in operating the tender evaluation process.

Department for Communities and Local Government – tenancy deposit scheme: As part of a CEPA team, Charles provided specialist advice and support to DCLG on the financial aspects of the procurement of private concessionaires to operate the tenancy deposit schemes in England and Wales.

PFI review: Charles managed CEPA’s advice to a NHS Trust on the causes of its deficit. The project included a review of the impact of the Trust’s PFI contract on its ongoing sustainability and financial statements.

Feasibility, design and implementation of InfraCo Asia: Charles led the work on the formal establishment of a new greenfield infrastructure project development facility and investment fund. Work included feasibility, design, founding investor sign-off, and advice on the appointment of a private sector management team (which was successfully achieved).

World Bank and Kenya Airports Authority (KAA): Charles is leading CEPA’s advice to KAA on its debt financing options for its ongoing investment programme. Charles is also managing Barclays input to the project (Barclays are CEPA’s sub-contractor for the project), especially on financing options and credit ratings.

World Bank and Government of Montenegro: CEPA provided advice to the government of Montenegro to develop an effective

7

Legal, Regulatory and Institutional framework to promote private investment in power generation. The project had a particular focus on promoting private investment into new generation. Charles was the project director for this assignment and provided specific expertise in PPP transaction and finance.

Globeleq - bid for an independent power plant, Africa: Charles advised on bid strategy, leading the funding strategy for the debt package and advising on construction, fuel supply, environmental, tax and accounting aspects of the bid.

PPP options for irrigation projects in Zambia: Charles is leading the multi-disciplinary team currently developing PPP options for the World Bank and Government of Zambia in relation to three irrigation infrastructure projects.

Strategic advice for a leading IPP. Charles led CEPA’s advice to Tsavo Power, an Independent Power Producer based in Kenya and part owned by Globeleq, on its expansion strategy. The work included advice on construction and operating costs, debt and equity financing packages and tariff.

Gas to electricity project acquisition, Tanzania. Led the in-country investment appraisal, including assessment of sector and regulatory risk for Globeleq’s. Project managed subsequent sale of promissory notes and shareholder-funded 50% expansion of plant, including negotiating and completing all related legal documentation.

Greenfield power projects, East Africa. Lead responsibility for assessing project economics, funding and risk for Globeleq’s assessment of the Bujagali hydroelectric project in Uganda and the Embakasi diesel plant in Kenya.

Infrastructure project appraisal. As CDC Country manager, assessed infrastructure projects in water (privatisation), airports (management), ports (concession), electricity transmission (export opportunities for a captive power project and cross-border) and integrated agribusinesses.

Project Finance training: Charles conducted a project finance training course for bankers, developers, and government officials in Nigeria.

Qualifications and HonoursChartered Management Accountant (ACMA) 1996University of Bristol, BSc (Hons), Economics and Politics II (i) (1990)Employment History2006 – present1997 – 20051990 – 1997

Director, CEPADirector, Globeleq, a CDC Group Company, London and Dar es SalaamPricewaterhouseCoopers – Management Consulting Division, London



Prof. David Newbery, CEPA Vice-Chairman

Summary

Professor David Newbery is a world-leading economist in his field. He has extensive consulting experience on the regulation of network utilities in a range of OECD and developing economies and has written and presented on issues of regulatory policy. David is CEPA’s Vice Chairman and a Professor of Applied Economics at Cambridge University, where he was the Director of the Department of Applied Economics from 1988 - 2003. He spent two years as a Division Chief in the World Bank and has been a visiting Professor at Berkeley, Princeton, Stanford and Yale. He is a fellow of both the Econometric Society and the British Academy. David is an academic member of DEFRA’s Environmental Economics Panel and was Chairman of the Dutch electricity Market Surveillance Committee from April, 2001- Mar 2005. He was a member of the Competition Commission from 1996 - 2002 and has been an advisor to Ofgas, Ofwat, Ofgem and the Office of Rail Regulation. His recent work has concentrated on the privatisation, restructuring and regulation of utilities, particularly in the energy sector, and market design and performance in liberalised electricity and gas markets.

Selected Experience Consultant to the National Treasury of South Africa on the reform of the

electricity industry 2007-8 Expert Adviser to CEPA led consortium advising the Commission on the

recent gas distribution and transmission price review. Expert Advisor, detailed study for DEFRA determining the direct and

indirect costs and benefits to the Russian Federation from ratifying the Kyoto Protocol.

Co-Project Director, series of studies for Portugal’s Competition Authority examining the gas and electricity markets, proposed mergers, and remedies that might be applied to mitigate any effects on competition.

Economic adviser to Ofgas and then Ofgem. Advised Ofgas on the network code and the regulation of TransCo; Ofgem

(electricity and gas); and Offer on use-of-system pricing and reforms of the pool.

Directed a sequence of large research projects on the British energy markets under contracts with the ESRC (1989-2010), EPSRC (2005-10) and EU (2005-8), several projects studying tax reforms and the transition of Hungary to the market economy, financed by the ESRC, PHARE, & ACE, and several projects on road pricing financed by the ESRC and DfT.

Worked with CET on preparing the privatisation of Poland’s 33 electricity distribution companies.

Member of World Bank teams advising the governments of Hungary, the Czech Republic, Romania Bulgaria, Bangladesh, Kenya, Uganda and South Africa on regulatory reforms and restructuring of the electricity, gas and oil sectors needed to improve sector performance.

Chairman, Dutch Electricity Market Surveillance Committee 2001-5.

9

Preparation for first electricity price control in the Netherlands

Qualifications1996 President – European Economic Association1993 Harry Johnson Price of Canadian Economic Association1991 Fellow of British Academy1990 Frisch Medal of the Econometric Society1976 PhD Economics – University of Cambridge1965 BA (Hons) Economics – University of CambridgeEmployment History2001 – present

Vice-Chairman, CEPA

1988 – present

Professor of Applied Economics

1988 – 2003 Director of Department of Applied Economics, Cambridge1987 – 1988 Ford Visiting Professor at University of California, Berkley1985 and 1987

Visiting Professor, Princeton University; Visiting Scholar, IMF

1981 – 1983 Division Chief, World Bank – Washington, D.C.1966 – 1988 Lecturer then Reader in Economics, Cambridge; Fellow and

Director of Studies in Economics, Churchill College, Cambridge

Selected PublicationsNewbery, D. (2000) Privatization, Restructuring & Regulation of Network Utilities, MIT PressNewbery, D. et al (1999) A European Market for Electricity?, London: CEPRNewbery, D. & Green, R. (1993) The Regulation of the Gas Industry: Lessons from Electricity, Fiscal Studies, 14 (2)



AT Kearney

11



13

CURRICULUM VITAE PROFESSOR JOÃO CONFRARIA

Name: João Confraria SilvaBorn: 1960Citizenship: Portuguese

1. Academic background

Doctor in Economics, 1991, Catholic University of Portugal.Master in Economics, 1988, New University at Lisboa.Undergraduate studies, 1977-1982, Catholic University of Portugal

Research areas: Economics of Regulation, Industrial Economics and Historical Economics.

2. Current professional activity

Assistant Professor of the School of Economics and Business, teaching courses on Public Policy, Regulation & Competition and Telecommunications.

Co-director of the Masters on Law and Business (School of Economics and Business and School of Law)

Coordinatior of Executive Programs on Regulation and Competition.

3. Previous Professional Activity

Vice-President of Instituto Nacional da Aviação Civil (Portuguese Civil Aviation Authority). (2007-2011)

Member of the Board of ICP - Anacom (the Portuguese Regulatory Agency for Telecommunications and Postal Markets) (1996 – 2002).

Other positions include Member of the Advisory Board of Portugal Telecom SGPS (2004-2009) and Assistant to the Deputy Minister of Trade (1991-1992)

Director of the Centre for Applied Studies at the Catholic University of Portugal (since 2005-2007)



Coordinator of the Executive Program on Economics of Regulation and Competition (since 2003-2007)

Representative of the Catholic University at the Professional Council of Ordem dos Economistas (since 1995).

Member of the Board of the MBA Program at the Catholic University of Portugal 1991-1993.4. Consulting work

Economic analysis of costs in water and waste industries, 2007, Instituto Regulador de Águas e Resíduos (regulatory authority)

Regulatory framework of water and waste industries, 2007, Instituto Regulador de Águas e Resíduos (regulatory authority)

Price regulation of postal services, 2006-2007, Correios de Portugal (postal operator)

Regulatory framework for electricity in Angola, 2007 (electricity regulator)

Competition in the banking sector in Angola, 2006 (private company)

Economic Impact of the High Speed Train Network, 2005, Rede de Alta Velocidade EP. (High Speed Network, Public Enterprise)

Regulation of Access Prices to Railway Networks, 2003-2006, Instituto Nacional dos Transportes Ferroviários (Regulatory Agency for Railways)

Price Regulation in Retail and Wholesale Markets, 2005, Portugal Telecom SGPS and PTC SA

Regulation of Access to Mobile Networks, 2005, CTT - Correios de Portugal, SA (National Post Office)

Quantitative Estimates of the Economic Impact of the Ocean on Portuguese GDP and Employment, 2004, Comissão dos Oceanos.( Ocean Commission)

Economic Trends in Portuguese Regions, 2003-2004, Banco Comercial Português SA.

Economic Integration through Foreign Direct Investment in (the Less Favoured Countries of) Central and Eastern Europe and Impact of the (Less Favoured Countries of the) European Union, 1996, European Commission.

15

Trade, Labour and Capital Flows: the Less Developed Regions, European Commission, 1996

Competitiveness Report, 1995, Forum para a Competitividade.

Regulation of Water Industry – Institutional Framework, 1995, Águas de Portugal SA. (Holding for main water regional and local companies)

Price Regulation of Natural Gas Markets, 1991-1996, Gás de Portugal SA.(Holding for main natural gas regional distributors)

Analysis of Portuguese Industrial Policy, 1991-1995, Ministério da Indústria e Energia.

Industrial Policy in Angola, 1994, Ministério da Indústria de Angola.

Enterprise behavior and economic adjustment in Eastern Europe, 1991-1994, World Bank

Official Stock Market Index, 1990, Bolsa de Valores de Lisboa (Lisbon Stock Exchange)

Analysis of market structure, Industrial and Services markets, 1988-1991, Citibank, Banco Fonsecas & Burnay and Bolsa de Valores de Lisboa.

5. Published work

5.1. Books

Regulação e Concorrência. Desafios do século XXI, Lisboa, Edições Universidade Católica, 2010, 2ª ed.

O Interesse Público na Política de Comunicações. 1910-2010, Lisboa, Fundação Portuguesa das Comunicações, 2010.

Comunicações na Idade Contemporânea. Cartas, Telégrafos e Telefones, (coord.), Lisboa, Fundação Portuguesa das Comunicações, 2009.

Comunicações na Idade Contemporânea. Tempos Digitais, (coord.), Lisboa, Fundação Portuguesa das Comunicações, 2008.

Regulação e Concorrência. Desafios do século XXI, Lisboa, Edições Universidade Católica, 2005.



Desenvolvimento económico e política industrial. A economia portuguesa no processo de integração europeia, Lisboa, Edições Universidade Católica, 2nd edition, 2001 (1st edition, 1995)

Condicionamento industrial. Uma análise económica, Lisboa, Direcção-Geral da Indústria, Estudos de Análise Industrial, 1992.

5.2 Chapters in books

“Regulação de preços na concessão da Companhia Anglo Portuguesa de Telefones” in Comunicações na Idade Contemporâne. Cartas, Telégrafos e Telefones, Lisboa, Fundação Portuguesa das Comunicações, 2009.

“Serviço universal e liberalização” in Comunicações na Idade Contemporâne. Cartas, Telégrafos e Telefones, Lisboa, Fundação Portuguesa das Comunicações, 2009.

“Os mercados de comunicações em números” in Comunicações na Idade Contemporâne. Cartas, Telégrafos e Telefones, Lisboa, Fundação Portuguesa das Comunicações, 2009.

“A eficiência na regulação de redes e serviços telefónicos: o início e o fim do século XX” (co-autor) in Comunicações na Idade Contemporâne. Tempos Digitais, Lisboa, Fundação Portuguesa das Comunicações, 2008.

“Mudanças institucionais. Regulação independente” in Comunicações na Idade Contemporânea. Tempos Digitais, Lisboa, Fundação Portuguesa das Comunicações, 2008.

“O Estado Regulador e a Liberalização”, 2007, in AAVV O Estado e a Economia. O Modelo Económico e Social Europeu no Século XXI, Lisboa, Fundação Friedrich Ebert / Instituto de Estudos para o Desenvolvimento.

“O Património” in A. Barreto (coord.), Fundação Calouste Gulbenkian. Cinquenta anos, em publicação, 2007.

“Consequências da Integração Europeia nos Objectivos da Regulação Económica Nacional” in M. C. Lobo e P. Lains, (eds.) Em Nome da Europa. Portugal em Mudança 1986-2006. Lisboa, Principia, 2007.

“Política Económica” in P. Lains e A. Ferreira da Silva, História Económica de Portugal, Séculos XVIII – XX, Imprensa de Ciências Sociais, Lisboa, 2005.

“Power and Discretion in Independent Regulation. The Case of Portuguese Communications Markets”, in M. Crew e P. Kleindorfer,

17

Regulatory and Economic Changes in the Postal and Delivery Sector, Kluwer Academic Publishers, Boston/Dordrecht/London, 2004. “Postal Services Cost Modeling” (co-author), in M. Crew and P. Kleindorfer, Postal and Delivery Services. Pricing, Productivity, Regulation and Strategy, Kluwer Academic Publishers, Boston/Dordrecht/London, 2002.

"Portugal: Industrialization and backwardness" in J.Foreman-Peck and G. Federico, European Industrial Policy. The twentieth century experience, Oxford University Press, Oxford, 1998.

“Economia Digital e Investigação Tecnológica”, in A.A.V.V. Jornadas Empresariais Portuguesas. Encontro de Vidago, (vol. III) Associação Industrial Portuense, Câmara de Comércio e Indústria, Porto, 1998

“Economia Digital. Interrogações sobre as transformações tecnológicas em finais do século XX”, in A.A.V.V. Jornadas Empresariais Portuguesas. Encontro de Vidago, (vol. II) Associação Industrial Portuense, Câmara de Comércio e Indústria, Porto, 1998.

"Problemas estruturais da economia portuguesa e a situação económica em 1917" in A.A.V.V. Fenomenologia e Teologia das Aparições, Actas do Congresso Internacional de Fátima, Santuário de Fátima, 1998.

"Competitiveness and industrial policy in Portugal", in R.Sugden, Y.Katsoulacos e P.Devine, Competitiveness, subsidiarity and objectives. Issues for European industrial strategy, London, Routledge, 1995.

"Conduct and performance in Hungarian manufacturing", in S. Estrin, A. Gelb, I. Singh, East European Enterprises in Transition (II), Eastern European Economies, vol. 31, nº.6, 1994.

"Continuity and change in Portuguese industrial policy: mobility regulation in the clothing and textile markets", in P.Bianchi, K.Cowling e R.Sugden (eds), Europe´s economic challenge, London, Routledge, 1994.

5.3. Journals

"Regulation and turbulence in the Portuguese wool textile markets", Small Business Economics, 11: 283-292, 1998.

"Az iparpolitika megváltozott szerepe - Portugália és Magyarország egybevetése", Közgazdasági Szemle, XLIII. évf., 1996.



"O programa 3 do PEDIP e a reestruturação dos mercados industriais", Economia, vol. XVIII, 1994.

"O Prémio Nobel da Economia de 1993", Economia, vol. XVII, nº1, 1993.

"Política industrial do Estado Novo. A regulação de oligopólios no curto prazo", Análise Social,n.112-113, vol.XXVI, 1991.

"Metodologia dos índices oficiais de valores mobiliários de rendimento variável" (co-autor ), Economia, vol.XIV, n.2., 1990 e Bolsa de Valores de Lisboa, 1991.

"Actividades económicas realizadas ilegalmente, em Portugal: problemas e perspectivas de análise económica", 1990, Economia, vol. XIII, nº3, Outubro de 1989.

"Aspectos da abordagem económica do comportamento criminal", Scientia Juridica, vol.XXXVIII, 1989.

"Nota sobre a estimação da evasão fiscal através de uma função procura de notas e moedas em circulação", Economia, Vol.XII, nº3, Setembro 1988.

"A fuga ao imposto sobre o rendimento - um survey da literatura", 1988, Economia, vol.XII, nº2, Maio 1988.

5.4. Compact Disk (CD-ROM)

“On the use of LRIC models in price regulation”, (co-author) International Training Programme on Utility Regulation and Strategy, Public Utility Research Centre Reference CD, 2001.

5.5. Consulting reports

“Public interest and regulation of Postal services”, 2006, Conselho Económico e Social.

Trade, Labour and Capital Flows: The Less Developed Regions, The Single Market Review, (co-author, project coordinated by the Centre for Economic Research and Environmental Strategy), Sub series VI: volume 3, Office for Official Publications of the European Communities, 1998.

O PEDIP e a evolução da indústria portuguesa, (co-author), Gabinete de Estudos e Planeamento do Ministério da Indústria e Energia e Centro de Estudos Aplicados, Lisboa, 1994.

19

O Leste Europeu. Ameaças e Oportunidades para a Indústria Portuguesa, (co-autor), Ministério da Indústria e Energia, Gabinete de Estudos e Planeamento do Ministério da Indústria e Energia e Centro de Estudos Aplicados, 1992.

5.6. Dictionaries and Encyclopedias

"Condicionamento industrial", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VII, Porto, Livraria Figueirinhas, 1999.

"Cortiça, indústria corticeira", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VII, Porto, Livraria Figueirinhas, 1999.

"Electrificação", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VII, Porto, Livraria Figueirinhas, 1999.

"Empresas e estruturas empresariais in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VII, Porto, Livraria Figueirinhas, 1999.

"Indústria", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VIII, Porto, Livraria Figueirinhas, 1999.

“Indústria de cimentos”, in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VIII, Porto, Livraria Figueirinhas, 1999.

"Indústria de conservas", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VIII, Porto, Livraria Figueirinhas, 1999.

"Indústria de lanifícios", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VIII, Porto, Livraria Figueirinhas, 1999.

"Indústria de moagem e panifícios", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VIII, Porto, Livraria Figueirinhas, 1999.

"José Nascimento Ferreira Dias", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VIII, Porto, Livraria Figueirinhas, 1999.

"Lei do Fomento e Reorganização Industrial", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VIII, Porto, Livraria Figueirinhas, 1999.



“Lei da Reconstituição Económica", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VIII, Porto, Livraria Figueirinhas, 1999.

"Metalurgia e metalomecânica", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. VIII, Porto, Livraria Figueirinhas, 1999.

"Política industrial", in A. Barreto e M. Filomena Mónica, eds., Dicionário da História de Portugal, Suplemento, vol. IX, Porto, Livraria Figueirinhas, 2000.

“Investimento", 1997, in F. Rosas e J. Brandão de Brito, eds., Dicionário de História do Estado Novo, vol. I, Círculo de Leitores, 1997.

"Zonas industriais e parques industriais", 1987, Enciclopédia Polis da Sociedade e do Estado, Verbo.

21