Oxera's April 2011 Report for the Energiekamer on the WACC of Dutch Energy Networks A Review for Gas Transport Services

(1)

18 April 2011

Oxera's April 2011 Report for

the Energiekamer on the WACC

of Dutch Energy Networks

A Review for Gas Transport Services

(2)

Project Team Graham Shuttleworth Richard Hern

REPORT QUALIFICATIONS/ASSUMPTIONS & LIMITING CONDITIONS

This report sets forth the information required by the terms of NERA’s engagement by Gas Transport Services and is prepared in the form expressly required thereby. This report is intended to be read and used as a whole and not in parts. Separation or alteration of any section or page from the main body of this report is expressly forbidden and invalidates this report.

This report is not intended for general circulation or publication, nor is it to be used, reproduced, quoted or distributed for any purpose other than those that may be set forth herein without the prior written permission of NERA. Neither all nor any part of the contents of this report, any opinions expressed herein, or the firm with which this report is connected, shall be disseminated to the public through advertising media, public relations, news media, sales media, mail, direct transmittal, or any other public means of communications, without the prior written consent of NERA.

Information furnished by others, upon which all or portions of this report are based, is believed to be reliable but has not been verified. No warranty is given as to the accuracy of such information. Public information and industry and statistical data, including without limitation the data included in other NERA reports referred to in this report are from sources we deem to be reliable; however, we make no representation as to the accuracy or completeness of such information and have accepted the information without further verification.

The findings contained in this report may contain predictions based on current data and historical trends. Any such predictions are subject to inherent risks and uncertainties. In particular, actual results could be impacted by future events which cannot be predicted or controlled, including, without limitation, changes in business strategies, the development of future products and services, changes in market and industry conditions, the outcome of contingencies, changes in management, changes in law or regulations. NERA accepts no responsibility for actual results or future events.

The opinions expressed in this report are valid only for the purpose stated herein and as of the date of this report. No obligation is assumed to revise this report to reflect changes, events or conditions, which occur subsequent to the date hereof.

All decisions in connection with the implementation or use of advice or recommendations contained in this report are the sole responsibility of Gas Transport Services. This report does not represent investment advice nor does it provide an opinion regarding the fairness of any transaction to any and all parties.

This report is for the exclusive use of Gas Transport Services. There are no third party beneficiaries with respect to this report, and NERA does not accept any liability to any third party. In particular, NERA shall not have any liability to any third party in respect of the contents of this report or any actions taken or decisions made as a consequence of the results, advice or recommendations set forth herein.

(3)

Oxera's April 2011 Report on WACC: A Review Contents

Executive Summary

i

1. Overview

1

2. Risk Free Rate

3

2.1. Estimation Method 3

2.2. NERA Comments 3

2.3. Updated Calculation vs Expert Opinion 3

2.4. Summary 4

3. The Equity Risk Premium (ERP)

5

3.4. Summary 10

4. Asset Beta

12

4.4. Summary 15

5. Gearing

16

5.4. Summary 18

6. Debt Premium & Issuance Costs

19

6.4. Summary 20

7. Inflation

21

7.4. Summary 22

(4)

Oxera's April 2011 Report on WACC: A Review

Executive Summary

This report forms the second of two reports examining Oxera’s calculation of the cost of capital for Dutch energy networks. In this report, we consider each parameter of the WACC estimated by Oxera. For each component of the WACC, we:

§ compare Oxera’s methodology in the April 2011 paper and its earlier paper released in February 2010;

§ identify where Oxera has changed its methodology to take account of the comments in NERA’s report of December 2010 and identify any remaining methodological errors; and § identify where methodological choices are Oxera’s own or merely calculations following

Frontier Economics’ earlier method or the directions from the Energiekamer (EK) or the Nederlandse Mededingingsautoriteit (NMa).

Oxera’s methodology is broadly unchanged from its previous report in February 2010. Many of the criticisms raised in our December 2010 report therefore still apply. In some areas, Oxera has changed its methodology to reflect our criticisms, for example by raising the ERP in 2009 to reflect the effects of the financial crisis. However, Oxera’s report remains an unreliable basis for setting the cost of capital, with likely understatement of the cost of capital in 2010 at least.

We note that on many occasions Oxera adopts a methodology which seems to be drawn from earlier work by Frontier Economics, or based on advice from the EK, but which differs from or even contradicts decisions that Oxera has taken in its work for other clients in recent months. The results shown in Oxera’s report cannot therefore be regarded as representing Oxera’s own expert opinion of the cost of capital.

(5)

Executive Summary

Table 1

Comparison of Oxera's WACC methodology

WACC components Oxera Methodology Feb 2010 NERA Critique: Source of error Effect of NERA Critique of Oxera Feb 2010 (%)

Oxera Methodological Changes Apr 2011

ERP 4-6% precedent Revised ERP estimate, inc. Financial Crisis 0.58 uplift of 0.3-0.6% in 2009 only

ERP 4-6% precedent (long-term) unchanged, except 2009

Risk Free Rate Av. 2 and 5 years (short-term) unchanged

Gearing 50-60% unchanged

Debt Rating A grade 2006-2008: A/BBB; 2009-2011: A

Asset Beta Adjustment Vasicek Unexplained assumptions 0.41 unchanged

Asset Beta European & US companies US Companies unknown unchanged

Asset Beta Includes liquid and illiquid stocks Illiquid companies unknown unchanged

Real WACC Historic outturns Inflation -0.16 unchanged

Selection from Range Mid-Point of WACC range Risks mean mid-point too low unknown not specified

Beta & ERP Errors in both compound Cross-Product Effect 0.11 unchanged

Total Calculated Adjustment 1.09

Inconsistent Data Sets (see above)

Gearing & CoD 0.15

(6)

Overview

1. Overview

This report forms part of a series examining the cost of capital for Dutch energy networks prepared by NERA and other consultants. The principal events in this workstream are: § In 2005, Frontier Economics produced for the DTe an estimate of the weighted average

cost of capital (the “cost of capital” or “WACC”) for energy networks in the Netherlands;1

§ In February 2010 (the “2010 report”), Oxera produced an update on this cost of capital that built on Frontier’s earlier methodology and which estimated the cost of capital for a

single year - 2010;2

§ In December 2010, NERA produced a paper for Gas Transport Services (GTS) which provided a critique of Oxera’s analysis.3 We identified a range of (1) theoretical and methodological errors, (2) arbitrary and indefensible choices and (3) biased outcomes. We also identified the lack of clarity over whether Oxera’s update was Oxera’s estimate of the cost of capital or merely a recalculation of Frontier’s own estimate.

§ In April 2011 (the “2011 report”), Oxera produced another update on the cost of capital,4 which estimated annual, forward-looking estimates of the cost of capital for every year from 2006 to 2011. The estimates of the cost of capital in each year allegedly take into account all information up until the previous December. In this paper, Oxera mentions some of the comments we have made, but broadly adopts the same approach as in its earlier paper.

In this paper, we consider each parameter of the WACC estimated by Oxera. For each component of the WACC, we:

§ compare Oxera’s methodology in the April 2011 paper with the methodology in its earlier report released in February 2010;

§ identify where Oxera has changed its methodology to take account of NERA’s comments and identify any remaining methodological errors; and

§ identify where methodological choices are Oxera’s own or merely calculations following Frontier Economics’ earlier method or directions from the Energiekamer (EK) or the Nederlandse Mededingingsautoriteit (NMa).

Oxera’s methodology is broadly unchanged from its previous report in February 2010. Many of the criticisms raised in our December 2010 report therefore still apply. In some areas, Oxera has changed its methodology to reflect our criticisms, for example by raising the

1

Frontier Economics, The Cost of Capital for Regional Distribution Networks: A Report For DTe, December 2005

2

Oxera, Updating the WACC for Energy Networks: Quantitative Analysis, 5 February 2010 and Oxera, Updating the

WACC for Energy Networks: Methodology Paper, 2 February 2010.

3

NERA, Critique of Oxera’s Report for the NMa on the Cost of Capital: A Report for Gas Transport Services, 22 December 2010

4

(7)

Overview

equity risk premium (“ERP”) in 2009 as a result of the financial crisis. However, Oxera’s report remains an unreliable basis for setting the cost of capital, with likely understatement of the cost of capital in 2010 at least.

Figure 1.1

WACC Estimates (Oxera 2011)

Year 2006 2007 2008 2009 2010 2011

Cutoff Dec-05 Dec-06 Dec-07 Dec-08 Dec-09 Dec-10

Risk-free rate (nominal) (%) 3.7 - 4.3 3.6 - 4.1 3.9 - 4 4 - 4.3 3.9 - 4 3.3 - 3.8

ERP (%) 4 - 6 4 - 6 4 - 6 4 - 6 4.3 - 6.6 4 - 6 Asset beta 0.33 - 0.41 0.36 - 0.44 0.37 - 0.46 0.43 - 0.49 0.38 - 0.47 0.35 - 0.45 Equity beta 0.56 - 0.85 0.63 - 0.93 0.65 - 0.97 0.75 - 1.04 0.66 - 1 0.61 - 0.95 Cost of equity (%) 6 - 9.4 6.1 - 9.7 6.5 - 9.8 7.2 - 11.1 6.6 - 10 5.7 - 9.5 Debt premium (%) 0.9 - 1.1 0.9 - 1.1 0.9 - 1.1 0.8 - 1.5 1.1 - 1.9 1.3 - 1.6 Cost of Debt (%) 4.6 - 5.4 4.5 - 5.2 4.8 - 5.1 4.8 - 5.8 5 - 5.9 4.6 - 5.4 Gearing (%) 50 - 60 50 - 60 50 - 60 50 - 60 50 - 60 50 - 60 Tax rate (%) 29.1 - 29.1 25.5 - 25.5 25.5 - 25.5 25.5 - 25.5 25.5 - 25.5 25.5 - 25.5 Pre-tax WACC (nominal) (%) 6.3 - 8.3 6.4 - 8.3 6.8 - 8.3 7.3 - 9.5 6.9 - 8.9 6.2 - 8.3

Inflation (%) 1.3 - 2.3 1.5 - 1.9 1.6 - 1.6 1.9 - 2.4 1.5 - 1.6 1.3 - 1.5

Pre-tax WACC (real) (%) 5.1 - 6.1 4.8 - 6.3 5.1 - 6.6 5.3 - 6.9 5.3 - 7.2 4.8 - 6.7

Previous estimates for NMa 2005 Det. 2008 Det. Jul-09 Dec-09

Pre-tax WACC (real) % 4.7-6.9 4.7-6.3 5.2-6.9 5.3-6.9

Source: Oxera,5 Det. = Determination, decisions taken by the EK/NMa

5

Oxera, Updating the WACC for Energy Networks: Quantitative Analysis, 5 February 2010, page 2, and Oxera, Cost of

(8)

Risk Free Rate

2. Risk Free Rate

2.1. Estimation Method

Oxera estimates the risk free rate in its April 2011 report using average yields on nominal Dutch sovereign bonds with a 10 year maturity. Oxera uses two and five year average yields to obtain a range for the risk free rate, with the higher of these historic measures forming the higher bound and the lower measure forming the lower bound of its range. This method is the same as the method Oxera used in its February 2010 report.

2.2. NERA Comments

In our December 2010 report, NERA made two key criticisms of Oxera’s calculation of the risk free rate:

1. that Oxera put undue emphasis on the last two years in its calculation of the risk free rate since the last two years feature twice (in both averages) in the calculation; 6 and

2. that the data used for estimating the ERP and risk free rate were not consistent; Oxera justifies its estimate of the ERP with reference to long run data (since 1900) but estimates the risk free rate over a relatively short period of two or five years.7

As we pointed out in our December report, Oxera has itself argued that a mismatch between the calculation periods for the ERP and risk free rate is inappropriate in previous work:

“There is clear evidence, supported by theoretical and empirical research in corporate finance, that higher forward-looking uncertainty is likely to be associated with a higher ERP. […] It should be noted that the determination of the ERP should be consistent with the risk-free rate.”8

Oxera’s calculation of the risk free rate still does not correspond with best practice due to the mismatch between the calculation periods for the risk-free rate and ERP.

2.3. Updated Calculation vs Expert Opinion

It may not be surprising that Oxera uses a method for calculating the risk free rate that is inconsistent with its previously published views because Oxera explains that it is following a previously established methodology (based on reports by Frontier Economics for the EK):

“In previous determinations for energy distribution and electricity

transmission networks, EK estimated the risk free rate based on the two- and five-year average of yields on Dutch sovereign debt with a maturity of ten

6

NERA, Critique of Oxera’s Report for the NMa on the Cost of Capital: A Report for Gas Transport Services, 22 December 2010, page 4

7

NERA, Critique of Oxera’s Report for the NMa on the Cost of Capital: A Report for Gas Transport Services, 22 December 2010, page 4

8

(9)

Risk Free Rate

years… The methodology adopted previously involved calculating two- and five-year average of the yields on ten-year Dutch bonds. The risk-free rate is estimated to be within a range. The upper and lower bounds of the range are provided by these two- and five-year averages.”9

At no point does Oxera explain that the EK’s methodology (based on

recommendations by Frontier) will lead to a biased result since the risk free rate is inconsistent with the ERP.

2.4. Summary

Oxera has not changed its methodology for calculating the ERP in its April 2011 report. It continues to place too much weight on the two years preceding the estimate. Oxera’s methodology remains inconsistent because Oxera calculates the ERP and risk-free rate over different time periods. Oxera has inherited the methodology from Frontier, but does not point out the sources of bias in its estimate.

9

(10)

The Equity Risk Premium (ERP)

3. The Equity Risk Premium (ERP)

Oxera bases its estimate of the ERP on its survey of the finance literature and regulatory precedent. In both its 2010 and 2011 reports, Oxera considers evidence from (1) long run equity and government bond returns in the Netherlands and the world, (2) surveys of ERP expectations and (3) regulatory precedent from Great Britain.

Oxera’s estimation method does not form an independent view of the ERP, but determines whether the ERP assumption used by the EK in previous determinations (4-6%) is consistent with available evidence. In its 2010 report, Oxera concluded that “there does not seem to be any sufficient basis for departing from the range [for the ERP] used at the last

determinations” for the ERP in 2010. 10 In its 2011 report Oxera again stated that 4-6% was appropriate for the ERP in 2010, but it did recognise the case for an increase as a result of the financial crisis for the calendar year 2009.

“A range of 4.0-6.0% for the ERP is consistent with the evidence presented for all periods with the exception of 2009…

Given the unprecedented market conditions at the end of 2008, a higher and wider range for the ERP is appropriate. A wider range is also consistent with the approach taken by Ofcom. On that basis, Oxera considers that it is

appropriate to increase the ERP range to take into account the unprecedented market conditions that prevailed at the end of 2008. Because it is difficult to quantify the precise amount by which the range should be increased, it is necessary to rely on proxies. One such proxy can be obtained from survey evidence, which suggests an increase of between 30 and 60 bp in the ERP following the start of the crisis (45 bp on average). Hence the ERP range for 2009 is increased to 4.3% to 6.6%.”11 [footnotes omitted]

In our December 2010 report, we demonstrated that Oxera had used the data available on the ERP in a selective manner that was not consistent with its own previous

estimates of the ERP. Oxera consistently argued in contemporaneous reports that the financial crisis had an impact on the ERP. For example:

§ on behalf of BT Openreach in August 2008, Oxera argued that “there have been clear and significant negative developments in capital markets over the past year, which have resulted in […] an increase in the cost of raising capital and a re-pricing of risks.”12

10

Oxera, Updating the WACC for Energy Networks: Quantitative Analysis, 5 February 2010

11

Oxera, Cost of Capital for GTS: Annual Estimates from 2006 Onwards – prepared for the NMA, April 2011, pages 14-15

12

(11)

§ on behalf of KPN in December 2008, Oxera included a 25bp increase in the ERP (to 6.25%) as well as estimating a higher long-run ERP estimate of 6% for the Netherlands.13 § On behalf of TRA (Bahrain) in November 2009, Oxera used a 50bp increase in the ERP

to account for the effect of the financial crisis, whereby the stock market was represented by an index of world equities. 14 Since Oxera assessed the cost of capital for 2010 in its report for the EK on the basis of data available up to December 2009, Oxera’s decision not to increase the ERP for the 2010 cost of capital in the Netherlands is not consistent with the 50bp increase that Oxera applied in Bahrain just one month earlier.

In its April 2011 report, Oxera increased the ERP for 2009 by 0.3% to 0.6% to account for the impact of the financial crisis. However, Oxera does not recognise the impact of the financial crisis on the ERP in 2008, 2010 or 2011, despite having made such an adjustment in previous reports for other clients.

3.2.1. Oxera does not outline criteria for increasing the ERP

In addition to being inconsistent with its previous estimates of the ERP, Oxera does not explain in detail how it made the decision not to raise the ERP in 2008, 2010 and 2011. Oxera merely stated that “evidence from equity markets suggests that implied volatility has reduced significantly and stabilised at more normal levels” without specifying what “normal” is.15 Moreover Oxera presents contradictory evidence which does not support this position, for example:

§ quoting Ofcom increasing the top end of the range for the ERP by 0.25% in 2008 as a result of the financial crisis.16 In the footnote accompanying this quotation, Oxera admits that Ofcom included this 0.25% uplift in a consultation for wholesale broadband access in January 2011 as well. 17 If Oxera uses Ofcom’s decision as regulatory precedent for estimating the ERP in 2008, a consistent approach would also use Ofcom as a regulatory precedent in 2011;

§ presenting historic evidence which shows a higher ERP in 2010 and 2011 than in 2009.18 If Oxera is right to include an increase in the ERP in 2009, it should add a similar increase in 2010 and 2011 as well;

§ presenting data showing that the historic volatility on European indices remained high until February 2010, i.e. that volatility did not fall until after the December 2009

13

Oxera, Cost of Capital Prepared for KPN, 1 December 2008, page iv

14

TRA of Bahrain, Cost of Capital Determination, November 2009, p. 52. Note that TRA states explicitly that its decision is based on Oxera analysis – see footnote 1

15

Oxera, Cost of Capital for GTS: Annual Estimates from 2006 Onwards – prepared for the NMA, April 2011, page 15

16

Oxera, Cost of Capital for GTS: Annual Estimates from 2006 Onwards – prepared for the NMA, April 2011, page 11, footnote 19

17

Oxera, Cost of Capital for GTS: Annual Estimates from 2006 Onwards – prepared for the NMA, April 2011, page 11, footnote 19

18

(12)

information date that Oxera claims to use when estimating the ERP for 2010.19 Since Oxera claims that the ERP increases with volatility, it is inconsistent to argue that the ERP fell from 2009 to 2010; and

§ Both of the studies on which Oxera claims to be basing its adjustment to the 2009 ERP were published after Oxera’s claimed “cut-off date” of December 2008.20

3.2.2. Oxera uses unreliable and contradictory survey evidence

In addition, the survey evidence that Oxera uses to estimate the uplift to the ERP for 2009 does not form a reliable basis for adjustment, for the following reasons:

§ Survey evidence is heavily dependent on the parties asked. In these cases the surveys on which Oxera relied were conducted among CFOs (Graham and Harvey) and academics (Welch). CFOs may have an interest in reporting equity risk premia that are too low, in order justify project financing decisions. On the other academic opinion may not reflect market sentiment. The potential for such biases in reporting by respondents are among the reasons UK regulators have excluded survey evidence on the ERP in the past;21

§ The response rate for the Graham and Harvey survey was just 5-8 per cent of those surveyed.22 This indicates a risk of sampling bias between those who participated and those who did not;

§ Oxera’s assumed ERP of 4-6% is not consistent with the estimates in the surveys by Welch and by Graham and Harvey:

– Welch’s survey provides a 95% confidence interval for the ERP of either 4.7%-5.3% or 5.6%-6.4% for the next 30 years using the geometric and arithmetic methods respectively. These estimates cover higher and narrower ranges of the ERP than that used by Oxera, both individually and taken together. Welch also estimates a 95% confidence interval of 4.9%-7.1% for 2009, with a median of 6%, i.e. 55 bp higher than the midpoint of Oxera’s range for 2009;23 and

19_{Oxera, Cost of Capital for GTS: Annual Estimates from 2006 Onwards – prepared for the NMA, April 2011, page 13,}

Figure 3.2

20_{Graham and Harvey, The Equity Risk Premium amid a Global Financial Crisis, May 2009, page 6 provides estimate of}

0.6% uplift to the ERP during recessions. The paper covers data from June 2000 to February 2009

Welch, I., "Views of Financial Economists On The Equity Premium And Other Issues," The Journal of Business 73-4, October 2000, 501-537, with 2009 update, no pages – retrieved from http://research.ivo-welch.info/equpdate-results2009.html. Published January 2009.

21

See, for example, Oxera’s own reference to Ofcom’s decision in which Ofcom states that it “affords this analysis relatively little weight”: Oxera, What is the cost of equity for RIIO-T1 and RIIO-GD1?, Prepared for the Energy

Networks Association, 4 February 2011, page 16, quoting from Ofcom, Proposals for WBA charge control – Consultation document and draft notification of decisions on charge control in the WBA market, 20 January 2011.

22

Graham and Harvey, The Equity Risk Premium amid a Global Financial Crisis, May 2009, page 1

23

Estimate derived by adding/subtracting the standard deviation divided by the square root of the sample size to/from the mean estimate provided by Welch, I., "Views of Financial Economists On The Equity Premium And Other Issues," The

Journal of Business 73-4, October 2000, 501-537, with 2009 update, no pages – retrieved from

(13)

– The median ERP since 2001 in Graham and Harvey’s survey is 3.46%, around 154 bp

lower than the midpoint of Oxera’s suggested range for the long run ERP.

Oxera provides no explanation as to why it believes that the estimated uplifts in the papers by Welch and by Graham and Harvey are accurate, when the estimate ranges of the ERP are substantially different from its own and from each other. Welch provides a point estimate equal to the upper bound and Graham and Harvey provide an estimate 54 bp lower than the lower bound of Oxera’s range.

§ Oxera does not explain where it obtains the estimated uplift of 0.3% and 0.6% in the Welch and Graham and Harvey papers:

– Welch’s estimates show no difference between the median estimate of theERP for 2009 and the median estimate of the arithmetic mean ERP for the next 30 years (both 6%).. However, he shows a 1% difference between both of these estimates and the median estimate of the geometric mean ERP for the next 30 years (5%). Welch also calculates the mean estimates of the ERP for 2009 (6.2%), the arithmetic mean ERP for the next 30 years (6%), and the geometric mean ERP for the next 30 years (5.1%). It is not clear how Oxera estimated the 0.3% uplift quoted, as 0.3% does not represent the difference between any of these averages.

– Oxera claims to source its 0.6% estimate from Graham and Harvey (2009). It seems likely that Oxera’s estimate of 0.6% originates from Graham and Harvey’s claim that “[d]uring recessions, the risk premium is 3.97% and during non-recessions, the premium falls to 3.37% %”24 i.e. a difference of 0.6%. This estimate does not relate specifically to the current recession, but rather to the average difference between the estimated ERP in 7 recessionary quarters and the 31 non-recessionary quarters

between 2001 and 2009. In other words, the estimate of the change in the ERP is not a forward-looking estimate which relates to changes in the ERP due to the impact of the current financial crisis, but is based on a very small number of historical observations. Oxera relies on this survey evidence, despite having dismissed it in recent work on the cost of capital for energy networks in Great Britain. In a report dated 4 February 2011,25 Oxera criticised survey data of the type used in its April 2011 report for the EK. Specifically, Oxera wrote that “respondents’ answers may be influenced by the way the questions are phrased”, that “there is a tendency for respondents to extrapolate from recent realised returns, making the estimates not entirely forward-looking” and that “the results are based purely on

judgement and are less reliable than estimates based on direct market evidence on pricing” (emphasis added).26 Oxera quoted a statement by British telecoms regulator, Ofcom, that “we afford this [survey] analysis relatively little weight, since participant surveys do not

24

Graham and Harvey, The Equity Risk Premium amid a Global Financial Crisis, May 2009, page 6 provides estimate of 0.6% uplift to the ERP during recessions. The paper covers data from June 2000 to February 2009

25

Oxera, What is the cost of equity for RIIO-T1 and RIIO-GD1?, Prepared for the Energy Networks Association, 4 February 2011.

26

(14)

provide the same quality of evidence as market-based measures.”27 In its work for British energy networks, therefore, Oxera concluded that “The survey evidence is mixed and it therefore seems inappropriate to place significant weight on this evidence.”28 Oxera provides no justification for taking a contrary view in its work for the EK.

3.2.3. DGM is a robust, forward-looking alternative to Survey Data The Dividend Growth Model (DGM) would have provided a robust alternative to the combination of ad-hoc survey evidence and the assumption of an ERP of 4-6% on the basis of regulatory precedent. The DGM estimates the ERP according to current forecast dividend rates divided by the price level (‘dividend yield’) plus an assumed long-run growth in dividends. A DGM approach has key advantages over survey data when estimating the ERP such as:

§ using directly-observable publically available data, which lessens the potential for selection biases; and

§ relying on the views of actors actually participating in the market. For example, a DGM approach would take dividend forecasts from analysts’ reports, which have proven to be a reliable basis for estimating the figures used by buyers and sellers participating in the market.

In its recent work for energy networks in Great Britain, Oxera used the DGM as a “cross-check on historical estimates”,29 but has chosen not to do so in its work for the EK. 3.2.4. Summary of Oxera’s approach

Oxera’s methodology has taken some account of NERA’s critique in that it recognises the impact of the financial crisis on the ERP, in a manner more consistent with other Oxera reports. However, Oxera’s analysis understates the full impact of the financial crisis by accounting for its impact in only one year. Oxera’s analysis remains subjective and non-transparent as it does not specify the criteria for choosing when to adjust the ERP for changes in market conditions and relies on survey evidence, which is not ‘market-tested’.

Oxera’s report suggests that its estimates of the ERP are not independent estimates, but grounded in regulatory precedent. For example, in its April 2011 paper, Oxera highlights the following regulatory precedent, which set the ERP at the same level as Oxera does for 2006-2008 and 2010-2011:

27

Oxera, What is the cost of equity for RIIO-T1 and RIIO-GD1?, Prepared for the Energy Networks Association, 4 February 2011, page 16, quoting from Ofcom (2011), Proposals for WBA charge control – Consultation document and

draft notification of decisions on charge control in the WBA market, 20 January 2011.

28

Oxera, “What is the cost of equity for RIIO-T1 and RIIO-GD1?, Prepared for the Energy Networks Association, 4 February 2011, page 16.

29

(15)

§ “In 2006, [the EK] estimated the ERP for energy networks to be in a range of 4-6%. § In 2008, the EK’s advisers judged that the likely increase in ERP resulting from the start

of the subprime crisis did not invalidate the prevailing range. The transport regulator in the Netherlands, Vervoerkamer, set the ERP at 4.0-6.0% in 2008, in its determination on pilotage services provided by harbour tugs.

§ Two years later, in 2010, the EK adopted the same estimate for the ERP, 4-6% at a time when equity markets had largely stabilised from the crisis.” 30

Oxera appears to work on the hypothesis that this regulatory precedent for establishing the ERP in the Netherlands as 4-6% is correct. Oxera then evaluates whether this position is unsustainable given the market data available:

“A range of 4.0%-6.0% for the ERP is consistent with the evidence presented for all periods with the exception of 2009 (see Table 3.4 below)…For all years leading up to the start of the crisis, the evidence does not justify a departure

from EK’s 4-6% range.” 31 [emphasis added]

In other words, Oxera does not undertake an unbiased assessment of the ERP. Instead, Oxera starts with the prior assumption that the ERP is 4-6% and will only change it in the face of evidence demonstrating that this assumption is incorrect.

By contrast, where Oxera increases the ERP to reflect the financial crisis (i.e. for calendar year 2009), Oxera relies on subjective, unreliable and ad-hoc survey evidence:

“Oxera considers that it is appropriate to increase the ERP range to take into account the unprecedented market conditions that prevailed at the end of 2008. Because it is difficult to quantify the precise amount by which the range should be increased, it is necessary to rely on proxies. One such proxy can be obtained from survey evidence, which suggests an increase of between 30 and 60bp in the ERP following the start of the crisis,” 32

Thus, Oxera applies its own estimate of the change in the ERP as a result of the financial crisis to the estimate of the ERP previously established by the EK, without checking whether those estimating the change in the ERP adopted the same starting point. As a result, Oxera does not estimate the ERP independently for any year.

3.4. Summary

In contrast to its February 2010 report, Oxera makes some adjustment to the estimated ERP in the light of the financial crisis, but only for 2009. This approach contrasts with the approach Oxera has applied in work for other regulators and companies. Oxera does not explain why it applies no uplift to the ERP in other years nor why it adopted this selective

30

Oxera, Cost of Capital for GTS: Annual Estimates from 2006 Onwards – prepared for the NMA, April 2011, page 9.

31

32

(16)

approach. The survey evidence that Oxera uses is unreliable and contradictory. A more objective and transparent approach is to use the DGM.

Overall, Oxera has not formed an independent estimate of the ERP in any year, but relies instead on regulatory precedent. In 2009, Oxera adopted its own methodology for raising the ERP, but the estimate of this increase is itself unreliable and not comparable with the

(17)

Asset Beta

4. Asset Beta

Oxera made only minor changes to its methodology for estimating the asset beta between February 2010 and April 2011. In both cases, to estimate a range for the asset beta, Oxera uses the median of two- and five-year correlations between a peer group of international utilities and the market portfolio. (Oxera claims to use a “National Index” in its 2010 and 2011 reports, without specifying which.) Oxera selects members of the peer group according to “whether they continued to satisfy EK’s criteria”.33 In both February 2010 and April 2011, the EK’s criteria included the following (Quotations are from Oxera’s April 2011 report, but analogous quotations may be found in its February 2010 report.):34

§ Changes in the business mix – “comparators must generate the majority of their earnings from energy network businesses”;35

§ Regulatory Framework - “consistent with previous EK methodology, a mixture of companies under price cap, revenue cap and cost of service regulation were included. Regulatory risk was also considered so that a regulated company could be excluded if it was exposed to excessive regulatory risk, or if there was a lack of a clearly defined framework”;36

§ Gearing – “Companies with high gearing were excluded”. 37

In addition, in April 2011 Oxera introduced a new criterion, responding to NERA’s comments about Oxera’s inclusion of illiquid stocks:

§ Liquidity – “A necessary condition for beta estimates is that markets for their securities are sufficiently liquid. Therefore, using EK’s methodology, only those companies with non-zero trading volume on at least 90% of all trading days were included in the sample. Furthermore, as the shares of smaller companies tend to be less actively traded a criterion on minimum size is included (i.e. annual revenue in excess of €100m).” 38

Oxera does not appear to have excluded any companies from consideration as a result of introducing this new liquidity criterion, so it may merely be a description of the

characteristics of the companies that were already in the sample. Although the sample for 2010 does not appear to have changed by comparison with Oxera’s February 2010 report, the median asset betas do appear to have changed slightly (see Table 4.1). Oxera offers no explanation for this change.

33

34

35

36

37

38

(18)

Asset Beta

Table 4.1

Median Beta Estimates 2010

Report Beta range

Oxera (2010) 0.39-0.45 Oxera (2011) 0.38-0.47

Source: Oxera.39

In our December 2010 report, NERA made four key comments on Oxera’s approach to calculating asset betas:

1. Illiquidity in some of the traded stocks distorts the betas calculated using five years of sample data (which could not be calculated for those stocks in the sample which were not publically traded in 2005);

2. The peer group chosen was not comparable as it included a large number of American firms subject to lower-risk cost-of-service regulation;

3. The impact of cash-calls and the review of the price control distort the betas estimated for some companies, particularly when using two years of sample data (e.g. Snam Rete Gas); and

4. Oxera adjusted the raw asset betas by the Vasicek method, which requires subjective input parameters.

4.2.1. Illiquidity

Oxera appears to have dealt with the first of these concerns by introducing a new criterion, “liquidity”, into the sample group. However, this criterion as defined by Oxera does not appear to bind and Oxera has not excluded the firms that NERA identified as being illiquid stocks over the period (e.g. Northwest Natural Gas which had an average 5-year bid-ask spread of over 10%).40

4.2.2. Comparability

In response to NERA’s criticism of the peer group, Oxera argues that including North American comparators is appropriate on the basis of “regulatory consistency with previous EK determinations”.41 In other words, Oxera’s method is not designed to provide the best estimate of the asset beta but to accord with previous estimates set out by the EK. Oxera also argues that “the lower risks resulting from cost of service regulation are balanced by the

39

Oxera, Cost of Capital for GTS: Annual Estimates from 2006 Onwards – prepared for the NMA, April 2011, page 16, and Oxera, Updating the WACC for Energy Networks: Quantitative Analysis, 5 February 2010

40

NERA, Critique of Oxera’s Report for the NMa on the Cost of Capital: A Report for Gas Transport Services, 22 December 2010

41

(19)

Asset Beta

higher risks resulting from non-regulated business segments”.42 If the lower risk of cost of service regulation happened to be exactly balanced by the non-regulated business segments of the North American comparators it would be entirely fortuitous. Moreover, this assertion that North American comparators are a suitable benchmark for European networks directly contradicts Oxera’s claims in work for other clients. For example, for the GB transmission and gas distribution networks, Oxera writes:

“Energy networks in North America provide the third cross-check [of the asset beta]. These are regulated under a ‘cost of service’-based regulatory

framework. This type of regulation is generally considered to expose networks to less risk than if they were regulated under price or revenue cap systems. For this reason the betas of the North American comparators would be expected to provide a lower bound on the asset betas for the UK networks.”43

In other words, in work for the Dutch regulator, Oxera has argued that North American utilities are a relevant comparator for Dutch energy neutrals because they face higher risks in the unregulated segment that offset the lower risks of the regulatory framework; in work for the British transmission and gas distribution networks Oxera has argued that North American comparators face lower risks and can only “provide a lower bound on the asset betas”. 44 The basket of comparator companies chosen by Oxera faces the following further problems: § Of the 19 firms in Oxera’s basket of comparators, 10 are from the USA or Canada, i.e. the

majority are not directly comparable with European networks; and

§ The range of the asset betas Oxera estimates is wider for the North American firms than is the case for the European firms.

4.2.3. Other concerns

In response to the other NERA comments, Oxera:

§ ignored the impact of cash-calls and the introduction of a new price control on its betas; § continued to use the Vasicek method of correcting betas, despite NERA’s alternative

suggestion that Oxera use the Blume method to increase the transparency of the results and because the Vasicek method relies on a number of subjective parameters. This

approach is inconsistent with Oxera’s work for the Energy Network Association (ENA) in the UK in February 2011, where Oxera calculates the Blume adjustment rather than the Vasicek adjustment.45 Moreover, in that same report for the ENA, Oxera discounts the Vasicek adjustment because “it is not straightforward to implement due to the need to make a prior assumption about the distribution of the beta.”46 In other words, Oxera has discounted the Vasicek adjustment in work for other clients for the same reasons that we suggested that Oxera use the Blume adjustment in our December 2010 report. Oxera

42

43

Oxera, What is the cost of equity for RIIO-T1 and RIIO-GD1, February 2011, page 23

44

45

46

(20)

Asset Beta

gives no reason for applying the Vasicek adjustment in this case, or any explanation as to how it overcame the need for “prior assumptions”.

As well as stating that it uses the previous comparator group to estimate the asset beta for “regulatory consistency”, 47 Oxera frequently uses terminology that shows that its calculation of the asset beta represents the EK’s instructions rather than its own best estimate of the asset beta prevailing in the market. For example, Oxera writes of its calculation of the asset beta: § “For the current assessment, Oxera began by reviewing the set of energy network

comparators considered in previous EK determinations. Comparators were added or removed from the sample across the different control periods depending on whether they continued to satisfy EK’s criteria” 48 [emphasis added];

§ “EK’s method is to form a range using the median of daily betas (measured over a two-year period) and the median of weekly betas (measured over a five-two-year period)” 49 [emphasis added]; and

§ “In previous determinations EK set its range for the asset beta on the basis of the median estimate for weekly data and the median estimate for daily data. This approach would yield ranges as shown in Table 4.3 for each annual estimate.” 50 [emphasis added]

4.4. Summary

Oxera’s estimates of the asset beta differ between its 2010 and 2011 reports, even for the same year (2010), for reasons that remain unexplained. That alone makes it difficult to have confidence in Oxera’s numbers. Oxera has failed to address NERA’s comments about the inclusion of North American companies in its estimation of the asset beta. Oxera does not believe that North American firms are appropriate comparators for European networks, according to Oxera’s own arguments in reports for other clients. Throughout its report,

Oxera’s language shows that Oxera is merely recalculating asset betas using the EK’s method, rather than applying a method it believes will produce accurate results. The methodology Oxera uses for calculating the asset beta does not therefore appear to represent Oxera’s own best estimate, but rather the view of the EK.

47

48

49

50

(21)

Gearing

5. Gearing

In its April 2011 paper, Oxera discusses a range of market evidence to determine a gearing assumption for energy networks in the Netherlands. Oxera’s work takes the same approach as it took in its February 2010 report, by first examining what credit rating network companies should seek to achieve and then considering what level of gearing would lead to that credit rating. In both cases, the analysis is qualitative, considering different categories of market and regulatory evidence and then summarising the content of the evidence.

Although the approach to setting the gearing assumption is the same in both the April 2011 and February 2010 papers, Oxera comes to slightly different conclusions about the

appropriate credit rating. In its February 2010 report, Oxera writes:

“a BBB rating may not meet the required standard of efficiency and/or constitute a reasonable economic assumption… These considerations broadly point to an appropriate rate reference towards the low end of the A range.” 51 In its 2011 paper, Oxera relaxed the target credit rating for the earlier periods it studied:

“For the earlier years, most regulatory precedents relied on a target of BBB+ and the difference in spreads for A and BBB bonds was not large, suggesting a minimal cost of downgrade from A to BBB. On balance, a target credit rating range of BBB/A is appropriate for 2006-08. For 2009 onwards, the evidence tilts towards A as a target credit rating. This is consistent with regulatory precedents and with the increased cost (in the form of higher bond spreads) associated with a downgrade to BBB during and after the financial crisis.” 52 The difference in credit rating for the different years has no impact on Oxera’s proposed gearing assumption, which remains at 50-60% as in Oxera’s February 2010 report. 53

In our December 2010 report, we argued that Oxera’s assumption of 50-60% was inconsistent with its assumption of a credit rating in the A band for 2010 as well as with Oxera’s earlier work. For example:

§ Moody’s indicates that a gearing level of 60% (when measured as net debt to RAB) is on the borderline between broad rating classes A and BBB;54

51

Oxera, Updating the WACC for Energy Networks: Methodology Paper, 2 February 2010, page 21

52

53

54

(22)

Gearing

§ Oxera’s approach is inconsistent with its own past practice. In 2007, Oxera estimated the cost of debt for the UK gas distribution networks at a level of gearing of 60% based on a mix of debt rated A and BBB;55 and

§ Oxera’s own analysis of actual debt-to-RAB ratios and associated credit ratings for mostly UK companies does not easily transfer to the Dutch framework as another key rating factor considered by Moody’s is the stability of the regulatory system (assigned a weight of 15% compared to 15% for net debt to RAV). The UK system has to be considered far more stable than the Dutch system, which was recently overturned by a court decision.

In its April report, Oxera has changed its target credit rating to A/BBB for 2006-2008. This approach is consistent with its own past practice and evidence from credit rating agencies that a 50%-60% gearing is consistent with A and BBB-rated debt. However, Oxera retains the same gearing assumption and increased the target credit rating to A for 2009-2011. The higher credit rating is inconsistent with assuming the same level of gearing (of 50%-60%), especially given the impact of the financial crisis during this later period. Oxera’s target credit rating is therefore still inconsistent with its gearing assumption for 2009-2011.Oxera’s approach biases the cost of capital downwards because it assumes that network companies can raise their credit rating to A-grade at a time of financial crisis with levels gearing as high as 60%, in contrast to the evidence.

Oxera takes clear ownership of its methodology and its gearing assumption. In its February 2010 report, Oxera publishes its gearing assumption as part of a separate “Methodology” paper rather than with its calculation of the risk free rate, ERP and other parameters. Moreover, the text of the April 2011 report clarifies that Oxera’s approach to setting the gearing assumption is its own:

“In the previous determination for energy distribution and electricity

transmission, EK had asked Oxera to review the estimation methodology for the gearing parameter. The methodology proposed by Oxera addressed the following questions in light of market evidence and regulatory precedent: - what is the appropriate target credit rating?

- what gearing assumption is consistent with this target credit rating.”56

55

See Oxera, Recent market evidence on the common WACC/CAPM parameters, Note prepared for Gas DNs, August 2007, available at:

http://www.ofgem.gov.uk/Networks/GasDistr/GDPCR7-13/Documents1/Oxera%20WACC_August%2031st.pdf

56

(23)

Gearing

5.4. Summary

(24)

Debt Premium & Issuance Costs

6. Debt Premium & Issuance Costs

Oxera uses the same methodology for assessing the debt premium in its April 2011 report as in its February 2010 report. Oxera defines a range using two measures. The first is the mean spread on corporate bond indices over the preceding five years. The second is the median spread on a basket of comparator companies’ debt over the preceding two years. In addition to the debt premium, Oxera includes an uplift on the cost of debt of 10-20 bp to cover debt issuance costs.

In our December 2010 report, our main comment on Oxera’s calculation of the debt premium was that Oxera uses only companies with a single “A” credit rating. This policy understates the size of the debt premium, because utilities with the level of gearing assumed by Oxera are unlikely to have a single “A” credit rating and in any case Oxera identifies A/BBB as the target credit rating (in the years 2006-2008 at least). Oxera has adopted the same position as NERA in its own work for other clients. For example, when estimating the cost of debt for the UK gas distribution networks at 60% gearing in 2007 i.e. before the economic crisis, Oxera based its estimate on a mix of “A”- and “BBB”-rated debt. 57

In its 2011 report, Oxera calculates the debt premium for each year using a different basket of companies. For 2006-2008, Oxera includes some companies with BBB-, BBB or BBB+ credit ratings. For 2009-2011, all of the comparator companies Oxera uses have A-, A or A+ credit ratings.

Oxera’s approach to estimating the debt premium does not represent a time-consistent pattern of the credit rating that energy networks can achieve. It is not clear how energy networks could increase their credit rating to A in the midst of a financial crisis while maintaining an unchanged level of gearing. Energy networks could not have foreseen the credit crisis in order to anticipate the need to increase their credit rating. Moreover, Oxera’s assumption of a constant gearing range of 50%-60% is not consistent with Oxera’s assumption of an

improving credit rating during the financial crisis.

In its April 2011 report, Oxera appears to suggest that it bases its methodology for calculating the debt premium on previous work by the EK:

“Consistent with EK’s earlier approach, and with the approach adopted for the risk-free rate, the range for the debt premium has been estimated on the basis of the five-year trailing average spread on the bond index for A rated debt and the median two-year trailing average spread on the sample of comparators.

57

See Oxera, Recent market evidence on the common WACC/CAPM parameters, Note prepared for Gas DNs, August 2007, available at:

(25)

Debt Premium & Issuance Costs

Adding 10-20 bp for debt issuance costs gives the following conclusions (see Tables 6.3 and 6.4 below)” 58

Whilst Oxera appears to have inherited the broad methodology for calculating the debt premium, the precise method for calculating the uplift for debt issuance costs is Oxera’s own. For its earlier February 2010 report, the EK had asked Oxera to review the treatment of debt issuance costs. Oxera concluded that “the evidence above suggests an uplift of 10-20 bp above market yields to account for debt issuance fees and debt-related overhead costs.”59

6.4. Summary

Oxera adopted broadly the same approach in April 2011 as in its February 2010 report. For the 2011 report, Oxera has calculated the debt premium using inconsistent credit ratings over time, which significantly understates the debt premium. The shift from A/BBB to A ratings between 2006-08 and 2009-11 is implausible as a financing strategy, inconsistent with the assumed level of gearing and leads to a downwardly biased estimate of costs.

Oxera has adopted the EK’s broad methodology for calculating the debt premium but the 10-20 bp uplift to cover debt issuance costs appears to be its own assumption.

58

59

(26)

Inflation

7. Inflation

Oxera uses the same methodology in its 2010 and 2011 reports for calculating the level of inflation used to deflate a nominal WACC to a real WACC. In both cases, Oxera defines a range between the higher and lower end of the two measures. The first measure is the average of five years of historic inflation and one year of forecast inflation. The second measure is the average of two years of historic inflation and one year of forecast inflation. Despite the fact that the methodology has not changed, the historic inflation numbers Oxera presents in its February 2010 and April 2011 reports differ as the February 2010 report uses inflation over the years June-May and the April 2011 report uses inflation over the years November-October.60

In our December 2010 report, NERA criticised Oxera’s method for using (mostly) historic inflation as theory and evidence suggest it does not accurately reflect the forecast inflation that investors would have priced into the risk free rate that Oxera has estimated. A more accurate method would have used the historic average over the past two and five years of ten year forecasts for inflation, in order to match the maturities and time frames Oxera used to estimate the risk free rate.

In its April 2011 report, Oxera continues using historic outturn inflation to define future inflation assumptions. Oxera does recognise that longer term forecasts are available, but states only that the longer term forecasts are “broadly consistent” with its own approach:

“Longer term inflation forecasts by the CPB can be used to assess the

estimates based on realisations. For instance, a four-year forecast of inflation is available from CPB for the period 2008-11, of 1.75%. An alternative forecast can be obtained from Consensus Economics. Ten-year forecasts for Dutch inflation are 1.8-1.9% in 2005 and 2007 and 2.1% in 2008. These cross-checks are broadly consistent with the range estimates reported above.” 61

The consistency of forecast inflation with the ranges that Oxera estimates is not surprising as Oxera’s methodology produces wide ranges for certain years. For example, Oxera estimates a range of 1.3% to 2.3% in 2006. In any case, a transparent and objective method would use the best available estimate for inflation expectations, rather than measures of historic inflation which happen to produce wide ranges, similar to expected inflation in some specific years.

60

See Footnotes to: Oxera, Updating the WACC for Energy Networks: Methodology Paper, 2 February 2010, page 14, Table 2.5 and Oxera, Updating the WACC for Energy Networks: Quantitative Analysis, 5 February 2010, Table 7.1

61

(27)

Inflation

Oxera bases its estimate of inflation in both its February 2010 and April 2011 reports on its own methodology rather than a methodology specified by the EK. In its February 2010 methodology report, Oxera notes that the EK previously used forecast rates but for reasons not explained asked Oxera to review its methodology:

“In previous decisions, EK’s inflation assumption was based on forecasts of CPI growth published by the CPB. This assumption was chosen owing to its availability over a suitable timeframe and the reputation of the CPB. EK also verified that the resulting estimate of the deflated risk-free rate was consistent with alternative estimates of the real risk-free rate.

EK has asked Oxera to review its treatment of inflation in the estimation of the WACC.” 62

Oxera’s choice of methodology does appear to have been influenced by specific objectives set by the EK. Among these objectives is compatibility with the risk free rate calculation, necessitating averages over two to five years. 63 In addition, Oxera explains that “EK is keen to base its inflation assumptions on economic data that is transparent and objective, and leaves minimum scope for arbitrary choice”.64 Nonetheless, the ultimate choice of method appears to have been selected by Oxera itself. Oxera’s methodology is not transparent or objective and it remains inconsistent with the method Oxera used to estimate the risk-free rate.

7.4. Summary

In the 2011 report Oxera continues to use its previous (2010) methodology for calculating the rate of inflation used to deflate a nominal WACC into a real WACC, despite the rationale for using long term forecasts to be consistent with market price information on the risk free rate. Oxera appears to have chosen the approach itself, after the EK asked it to review the

inflation-methodology as part of its February 2010 report, so the inconsistency is Oxera’s.

62

Oxera, Updating the WACC for Energy Networks: Methodology Paper, 2 February 2010, page 4

63

Full quotation: “In particular, EK is cognisant that its methodology for estimating the parameters of the WACC (notably, the risk-free rate and the debt premium) relies on two- to five-year historical averages. There is, therefore, a presumption that the inflation assumption used to deflate the WACC should be consistent with this aspect of the methodology.” Oxera, Updating the WACC for Energy Networks: Methodology Paper, 2 February 2010, page 4

64

(28)

Conclusion

8. Conclusion

Oxera’s methodology has remained broadly unchanged in its latest report of April 2011. It remains open therefore, to many of the criticisms that we made of it in our 2010 report. In some areas, Oxera has changed its methodology in response to our criticisms, for example by including:

§ an uplift to the ERP (albeit only in 2009) to reflect the costs of the financial crisis; and § BBB-rated debt in the calculation of the debt premium (albeit only in 2006-2008). However, these changes are not sufficient or comprehensive. Oxera has made no attempt to address the concerns that NERA raised regarding:

§ the consistency of the datasets for calculating the ERP and the Risk Free Rate; § the comparability and illiquidity of the peer group used to estimate the asset beta; § the compatibility of the gearing and cost of debt assumption;

§ the use of the Vasicek adjustment; and § use of historic instead of forecast inflation.

(29)

Conclusion

Table 8.1

Comparison of Oxera's WACC methodology

WACC components Oxera Methodology Feb 2010 NERA Critique: Source of error Effect of NERA Critique of Oxera Feb 2010 (%)

Oxera Methodological Changes Apr 2011

ERP 4-6% precedent Revised ERP estimate, inc. Financial Crisis 0.58 uplift of 0.3-0.6% in 2009 only

ERP 4-6% precedent (long-term) unchanged, except 2009

Risk Free Rate Av. 2 and 5 years (short-term) unchanged

Gearing 50-60% unchanged

Debt Rating A grade 2006-2008: A/BBB; 2009-2011: A

Asset Beta Adjustment Vasicek Unexplained assumptions 0.41 unchanged

Asset Beta European & US companies US Companies unknown unchanged

Asset Beta Includes liquid and illiquid stocks Illiquid companies unknown unchanged

Real WACC Historic outturns Inflation -0.16 unchanged

Selection from Range Mid-Point of WACC range Risks mean mid-point too low unknown not specified

Beta & ERP Errors in both compound Cross-Product Effect 0.11 unchanged

Total Calculated Adjustment 1.09

Inconsistent Data Sets (see above)

Gearing & CoD 0.15

(30)

NERA Economic Consulting 15 Stratford Place London W1C 1BE United Kingdom Tel: +44 20 7659 8500 Fax: +44 20 7659 8501 www.nera.com