RÉGIE DE L ÉNERGIE WRITTEN EVIDENCE OF PAUL R. CARPENTER FOR GAZ MÉTRO. The Brattle Group 44 Brattle Street Cambridge, MA (617)

Transcription

1 RÉGIE DE L ÉNERGIE WRITTEN EVIDENCE OF FOR GAZ MÉTRO The Brattle Group Brattle Street Cambridge, MA 0 () -00 May 00

2 TABLE OF CONTENTS I. OVERVIEW/SUMMARY... II. INVESTMENT RISK AND TRENDS IN THE NATURAL GAS MARKET ENVIRONMENT SINCE... III. EVALUATION OF GAZ MÉTRO S BUSINESS RISK SINCE AND IV. A. Determinants Of Business Risk And The Cost Of Capital... B. Gaz Métro s Business Risk Environment... C. Evidence of Changes in Gaz Métro s Business Risk Since and Increases in Uncertainty in Gas Use Per Customer Increased Risk due to Gaz Métro s Performance Incentive Mechanism... D. Conclusion... GAZ MÉTRO S BUSINESS RISK COMPARED TO OTHER LDCS IN CANADA AND THE UNITED STATES... A. Gaz Métro s Unique Business Environment.... Natural Gas Penetration in Québec Competition from Electricity.... Gaz Métro s Industrial Customer Load.... Conclusion... B. Gaz Métro s Return on Equity Compared to Allowed Returns on Equity for U.S. LDCs...

3 I. OVERVIEW/SUMMARY 0 0 Q. Please state your name, address and position. A. My name is Paul R. Carpenter. I am a Principal of The Brattle Group, an economic and management consulting firm with offices in Cambridge, Massachusetts, Washington D.C., San Francisco, California, London, England, and Brussels, Belgium. My office is located at Brattle Street, Cambridge, Massachusetts 0. Q. Will you briefly describe your educational background and professional qualifications? A. Yes. I am an economist specializing in the fields of industrial organization, finance and energy and regulatory economics. I received a Ph.D. in Applied Economics and an M.S. in Management from the Massachusetts Institute of Technology, and a B.A. in Economics from Stanford University. I have been involved in research and consulting on the economics and regulation of the natural gas, oil and electric utility industries in North America and abroad for over twenty years. I frequently have testified before federal, state and Canadian regulatory commissions, in federal court and before the U.S. Congress, on issues of pricing, competition and regulatory policy in these industries. Outside of North America, I have advised governments and regulatory bodies on the structure of their natural gas markets and the pricing of gas transmission services. These assignments have included testimony before the U.K. Monopolies and Mergers Commission and the Australian Competition Tribunal, and advice to the governments of, and regulators in, Greece, Ireland, the Netherlands, New Zealand and Australia. I have been extensively involved in the evaluation of the economics and regulation of the natural gas industry in North America. In Canada, I have advised pipeline companies and have previously testified before the National Energy Board ( NEB ) and the Alberta Energy and Utilities Board ( EUB ) on matters relating to pipeline competition and capacity expansion, including the Alliance Pipeline Ltd. certification proceeding. I gave evidence on business risk previously before the NEB in the multi-pipeline cost of capital case, on behalf of Foothills Pipe Lines, and in more recent NEB proceedings on behalf of

4 0 0 TransCanada PipeLines Limited. I provided written evidence on business risk before the Ontario Energy Board ( OEB ) on behalf of Union Gas Limited as part of its 00 rate application. I provided written testimony and appeared before the OEB on behalf of Enbridge Gas Distribution Inc. as part of its 00 rate application. Further details of my educational and professional background, as well as a listing of my publications, are provided in my curriculum vitae, which is appended to this evidence as Attachment A. Q. What assignment were you given in this proceeding? A. I have been asked by Gaz Métro to provide evidence concerning its business and regulatory risks as they affect the return on equity required by investors in Gaz Métro securities, and to render an opinion as to the adequacy of the current formula return to compensate investors for bearing those risks. In particular, I have been asked to evaluate whether, and the extent to which, there have been changes in the business risk and regulatory environment faced by Gaz Métro since, when the formula approach to return on equity was approved by the Régie, and since 00, when the formula was reevaluated. I have also been asked to provide evidence comparing Gaz Métro s business risk and current allowed return with the risk and allowed returns of other Canadian and U.S. natural gas local distribution companies ( LDCs ). Q. Could you summarize your conclusions? A. Yes. It is important for the Régie to undertake an updated, quantitative evaluation of Gaz Métro s allowed return. Changes in natural gas markets since and 00 suggest natural gas distribution investment risks have increased since the formula used to determine Gaz Métro s allowed return on equity was previously evaluated. Recent increases in equity thickness for NEB-regulated pipelines and Canadian distributors also suggest that investment risks have increased. And, the returns anticipated for new and proposed NEB-regulated pipelines suggest that the returns allowed under the NEB s formula are too low.

5 0 0 Gaz Métro s business and regulatory risk have increased since and 00, which implies that its allowed return should be increased relative to the return produced by the current operation of the ROE formula. Other comparative evidence suggests that Gaz Métro is riskier than average LDCs. In particular, it is riskier than the LDCs in Professor Chrétien s U.S. LDC portfolio, and riskier than Canadian LDCs that have recently received increases in their allowed returns via increases in their deemed equity thickness. Despite the fact that Gaz Métro is riskier than average LDCs, its allowed return is significantly lower than allowed ROEs for LDCs in the U.S. I understand that Gaz Métro is applying for a 0.% return on equity at its deemed equity thickness of.%. This request is at the low end of the range of required ROEs estimated by Professor Chrétien of.% to.%. In my opinion, the ROE that Gaz Métro is requesting is conservatively low, given the evidence that Gaz Métro has higher business risks and somewhat lower regulatory risks than typical gas LDCs in Canada and the U.S., and given that it has higher financial risks (at the.% deemed equity thickness) than the LDCs in Professor Chrétien s U.S. LDC sample. Finally, I note that an increase in ROE, represented by Gaz Métro s request over the current formula result, is appropriate given the increased business risks faced by equity investors in Gaz Métro since the ROE formula was introduced in and reviewed in 00. Q. How is the rest of your evidence organized? A. In Section II, I discuss standards for determining a fair return, and evidence that required returns for energy infrastructure have increased due to changes in natural gas markets since and 00. In Section III, I discuss the determinants of business risk, Gaz Métro s business risk environment, and how that environment has changed since and 00. Finally, in Section IV, I compare Gaz Métro s business environment to that of other LDCs in Canada and the U.S., and show how Gaz Métro s allowed return on equity compares to that of LDCs in the U.S.

6 II. WRITTEN EVIDENCE OF INVESTMENT RISK AND TRENDS IN THE NATURAL GAS MARKET ENVIRONMENT SINCE 0 Q. What is investment risk and how does it relate to a regulator s decision to establish an appropriate allowed return for a natural gas utility? A. One of the bedrock principles for establishing a fair return on equity for a regulated utility that has been long recognized by economists, regulators, and the courts in Canada and the U.S. is that a reasonable return should be comparable with the return available from the application of the capital to other enterprises of like risk. This principle implies that any evaluation of the reasonableness of a particular return should consider the level of risk faced by investors in the enterprise, and how those risks and returns compare with other companies with similar risks. Investment risk for a natural gas utility has been defined to include financial and business risks. Financial risk involves the extent to which debt is employed in the company s capital structure. Business risk is a somewhat more subjective concept and is more difficult to quantify precisely, but it is sometimes categorized to include the supply, demand (or market), competitive, operating and regulatory risks that might be faced by particular utilities. Both business and financial risks are accounted for in the methodology employed by Professor Chrétien to estimate the cost of capital for Gaz Métro. For example, as discussed in his evidence, the Fama-French and Adjusted CAPM methods employed to estimate the cost of equity for a National Energy Board, RH--0, p.-. See also National Energy Board, RH--00, p., where the NEB stated that a fair or reasonable return on capital should meet three requirements. It should: () be comparable to the return available from the application of the invested capital to other enterprises of like risk (the comparable investment standard), () enable the financial integrity of the regulated enterprise to be maintained (the financial integrity standard), and () permit incremental capital to be attracted to the enterprise on reasonable terms and conditions (the capital attraction standard). The NEB has recognized that the assessment of business risk is an inherently qualitative exercise. See National Energy Board, RH--, p. : The Board has systematically assessed the various risk factors for each of the pipelines but has not found it possible to express, in any quantitative fashion, specific scores or weights to be given to risk factors. The determination of business risk, in our view, must necessarily involve a high degree of judgment, and the analysis is best expressed qualitatively. See, for example, National Energy Board, RH--00, p.. I understand that in past decisions the Régie has treated regulatory risk as an element separate from business risk. In this evidence I include regulatory risk as a part of business risk conceptually, but I also evaluate its implications separately for Gaz Métro s return on equity.

7 0 0 benchmark gas LDC employ stock market data that reflect the historical business risks perceived by the equity investors in Professor Chrétien s sample portfolios. These data also reflect the market-value capital structures of the companies in the sample portfolios and thus their financial risks. The evidence I present compares Gaz Métro s particular business and financial risks to the risks that one would expect for a benchmark gas LDC, and thus assists the Régie in positioning Gaz Métro s required return on equity within the benchmark results provided by Professor Chrétien in his evidence. Q. You have mentioned the distinction between business and financial risk. How do those concepts relate to the allowed return on equity and the deemed equity ratio? A. Return on equity and the deemed equity ratio or thickness (when combined with the cost of debt) are the components which make up the total return on assets. It is, of course, the total return (the weighted average return on debt and equity) that matters to investors. But equity investors, as residual claimants on the firm s assets behind debt holders, are particularly concerned about the relationship between the equity ratio and the allowed return on equity. As residual claimants, equity investors bear added financial risk or leverage the lower the deemed equity thickness, all else equal. Q. How have changes in business risk under a formula approach to return on equity been traditionally accounted for by Canadian regulators? A. Ever since various Canadian regulators began to employ formulas for the determination of allowed ROEs, the practice has been to adjust the deemed equity thickness for perceived changes in business risk. Higher business risk would imply that greater equity thickness would be required. This procedure was fine as long the resulting total return on assets continued to be fair and reasonable. But because the evaluation of business risk is largely subjective, this procedure is not a substitute for the more rigorous determination as to whether the return on equity produced by the formula in combination with the deemed equity thickness is fair and reasonable. This determination requires the kind of analysis performed by Professor Chrétien in his evidence. Q. What has been the trend in allowed ROE under the NEB formula as compared to the formula applied by the Régie to Gaz Métro since?

8 A. The trend in ROE has been significantly downward under both formulas. In Figure I have plotted the authorized returns on equity produced by the NEB formula and those produced by the formula authorized by the Régie for Gaz Métro from to the present. Figure Gaz Métro's Authorized ROE vs. Authorized ROE under the NEB Formula 0.0% 0.00% Gaz Métro Authorized ROE Authorized ROE.0%.00%.0% NEB Authorized ROE.00% Source: NEB & Gaz Métro. Notes: Gaz Métro's Authorized ROE is its "base ROE," excluding incentives. The NEB Authorized ROE is effective for a calendar year. Therefore, the ROE authorized for is effective //. Gaz Métro's ROE is effective for a fiscal year. Therefore, the ROE authorized for is effective 0//. Note that even though the formulas are slightly different in the two jurisdictions, the trend over time in the resulting returns under the two formulas has been very similar. This is due, of course, to the fact that the two formula approaches are designed to largely track the trend in the yield on the Canadian Long Bond. As Figure demonstrates, the most significant declines in the formula returns have occurred since 00. The NEB Authorized ROE differs from Gaz Métro's Authorized ROE because the NEB and Gaz Métro adjustment formulas rely on different interest rates. The NEB adjustment formula relies on -months-out and -months-out forecasts of 0-year Government of Canada bond yields published in the November issue of Consensus Forecasts, and the spread between 0-year and 0-year Government of Canada bonds during the previous October. Gaz Métro s adjustment formula relies on -months-out and -months-out forecasts of 0-year Government of Canada bond yields published in the August issue of Consensus Forecasts, and the spread between 0-year and 0-year Government of Canada bonds during the previous July. Both the NEB and Gaz Métro adjustment formula use a 0. elasticity factor.

9 Q0. How does Gaz Métro s return on equity and equity thickness compare to that of other Canadian gas distributors? A0. I present recent approved common and preferred equity ratios and returns on equity for Gaz Métro and other Canadian gas distributors whose returns on equity are set using formula-based methodologies in Figure below. Figure Approved Capital Structures for Canadian Gas Distributors Utility Regulator Common Equity Ratio Preferred Equity Ratio 00/00 Return on Common Equity Pacific Northern Gas [] (West Division) BCUC 0.00%.%.0% Pacific Northern Gas N.E. [] (Ft. St. John/Dawson Creek) BCUC.00% 0.00%.% Pacific Northern Gas N.E. [] (Tumbler Ridge) BCUC.00% 0.00%.0% Terasen Gas BCUC.0% 0.00%.% Terasen Gas (Vancouver Island) BCUC 0.00% 0.00%.0% AltaGas [] EUB.00% 0.00%.% ATCO Gas EUB.00%.00%.% Enbridge Gas Distribution Inc. [] OEB.00%.%.% Union Gas OEB.00%.%.% 0 Centra Gas Manitoba PUB.0% 0.00%.% Gaz Métro Régie.0%.0%.% Gazifère Régie 0.00% 0.00%.% Source: Company annual reports, regulatory filings and decisions. Notes: [] 00 rates have not yet been approved by the BCUC. Reflects requested capital structure and formula-based ROE. [] 00 rates have not yet been approved by the EUB. Reflects requested capital structure and formula-based ROE. [] 00 rates have not yet been approved by the OEB. Reflects requested capital structure and formula-based ROE. 0 Q. What has been the recent experience in Canada with respect to changes in equity thickness to account for changes in business risk? A. In Figure I report the changes in deemed equity thickness that have been authorized by various Canadian regulators since the formula approach to ROE was adopted in their jurisdictions.

10 Figure Changes in Deemed Equity Thickness Utility Regulator Docket / Decision Date Approved Equity Thickness Prior Equity Thickness Change in Equity Thickness Pacific Northern Gas (West Division) BCUC G--0 Aug-0 0% % % Terasen Gas [] BCUC G--0 Mar-0 % % % Terasen Gas [] (Vancouver Island) BCUC G--0 Mar-0 0% % % ATCO Gas EUB 00-0 [ ] Jul-0 % % % ATCO Pipelines EUB 00-0 [ ] Jul-0.0%.% -0.% NGTL EUB 00-0 [ ] Jul-0 % % % Union Gas OEB EB Jun-0 % % % Foothills NEB TG-0-00 Dec-0 % 0% % TransCanada BC System NEB TG-0-00 Feb-0 % 0% % 0 TransCanada Mainline NEB Mar-0 0% [ ] % % TransCanada Mainline NEB RH--00 Apr-0 % % % TransCanada Mainline NEB RH--00 Jun-0 % 0% % Westcoast [ ] NEB TG-0-00 Aug-0 % (for 00); % (for 00) % % (for 00); % (for 00) Westcoast [ ] NEB TG--00 Aug-0 % 0% % Westcoast [ ] NEB RH-- Aug- 0% % -% Notes: [] BCUC Decision 00-0 also increased Terasen Gas and Terasen Gas (Vancouver Island) 's allowed formula-based ROEs. Terasen Gas' 00 ROE increased by 0.%, and Terasen Gas (Vancouver Island) 's 00 ROE increased by 0.%. [] In Decision 00-0, the EUB also considered AltaGas' equity thickness, and left it unchanged at %. [] In March 00, TransCanada submitted a settlement that would increase its deemed equity thickness to 0%. This settlement has not yet been approved by the NEB. [] Results for Westcoast are for Westcoast's Zones &. Tolls for Westcoast's Zones & are set differently pursuant to the NEB's Framework for Light-handed Regulation. [] Centra Gas Manitoba's rates are set using an equity thickness that is determined by formula, instead of a deemed equity thickness. Its approved equity thickness has changed over time as follows 0% ( test year),.% ( test year),.0% ( test year),.% (00/0 test year),.% (00/0 test year), and.% (00/0 test year). 0 Nearly all of the changes in equity thickness have been in the positive direction, and nearly all were associated with determinations by the regulators that business risks had increased for the utilities involved, or they were the results of settlements between the utilities and their stakeholders. Q. Could you describe some of the changes in the natural gas market environment in North America that have occurred since and their implications for natural gas utility infrastructure investment? A. Yes. The market environment in which gas utilities operate in North America has changed significantly since reflecting greater uncertainty in the supply of the gas commodity and greater uncertainty in the extent and timing of growth in demand. This uncertainty is partly reflected in significantly higher gas commodity price levels and

11 0 0 volatility. These changes have significant implications for the need for, and investment risk of, gas utility infrastructure. These changes have been recognized in Canada by the NEB in its summary of the feedback it received in 00 during its Energy Futures Project consultation sessions. In its report on these sessions under Key Messages, the NEB writes:. New Energy Paradigm There is consensus that the energy system in Canada has transitioned into a new paradigm characterized by tighter energy markets, high and volatile energy prices, high value currency, higher inflationary pressures, ageing infrastructure at all levels, and an uncertain demand response. Additional gas market uncertainty has also been created by the increased environmental concerns associated with greenhouse gas emissions and their role in long-term climate change. While some of the policy responses to these concerns, such as the pricing of carbon emissions through so-called cap-and-trade or taxation schemes, may ultimately favor natural gas relative to coal or oil, other policies may not (e.g., those that favor renewable energy sources or nuclear power.) No matter what the specific policies are that result from these heightened environmental concerns, it seems clear that higher and more uncertain prices for natural gas are more likely than not to be experienced in a carbon-constrained world. Q. What is the evidence for changes in natural gas commodity prices and price volatility since? A. Tighter supply/demand balances for natural gas have led to substantially increased prices and price volatility in North America since. To show this phenomenon, in Figure I have plotted the -month forward strip price of natural gas on the New York Mercantile Exchange (NYMEX) from April 0 to April 00. This is a useful index to National Energy Board, Energy Futures Project: Consultation Sessions Feedback, July 00, p..

12 reference because it reflects the broad market expectation of the level and volatility of future prices in a way that is normalized somewhat for seasonal effects. Figure Nymex Natural Gas -Month Average Future Prices (April 0 - April 00) Future Price (US$/GJ) Pre- -Month Standard Average Deviation. 0. Post- -Month Standard Average Deviation.. 0 //0 // // // // // // // // // //000 //00 //00 //00 //00 //00 //00 //00 Note: Null prices are omitted to adjust for market inactivity. Prior to the average of these forward prices was US $. per GJ with a standard deviation of US $0., a very low and stable price environment. As the figure indicates, since the average forward price more than doubled to US $. per GJ and the standard deviation of those prices ballooned to US $., an eight-fold increase. The same pattern can be seen for -month NYMEX forward prices shown in Figure. 0

13 Figure Nymex Natural Gas -Month Average Future Prices (April - April 00) Future Price (US$/GJ) Pre- -Month Standard Average Deviation. 0.0 Post- -Month Standard Average Deviation.. 0 // // // //000 //00 //00 //00 //00 //00 //00 //00 Note: Null prices are omitted to adjust for market inactivity. 0 While these changes in the commodity market pre and post- are dramatic, even those averages mask to some extent the market changes experienced in just the last months, as NYMEX forward prices on both a and -month basis have risen to the US $ to $ per GJ range. This reflects the market s current perception that we have entered a sustained period of high natural gas prices and high price volatility that has not been seen before in North America. Whether or not these changes in the gas commodity market are permanent or temporary is a subject of debate (the market, at least, is forecasting a continuation of the pattern for at least the next three years.) One thing that we can have some confidence in, however, is that there will be continued uncertainty in future prices and increased price volatility, uncertainty which contributes to greater demand growth and use-per-customer risk for gas utilities distributing the commodity. Q. Is there direct evidence that since the market has required higher returns for utility investment in Canada than the returns implied by the NEB formula?

14 A. Yes, there is. Figure provides the details of the returns agreed by the parties in recently completed and proposed Canadian pipeline projects. According to these data, these projects earn, or propose to earn, returns on equity in the range of.0 percent to.0 percent on equity thickness of at least percent. I will return to other evidence of returns for comparable risk LDC investments below. But these pipeline investments are some of the best examples we have in Canada of returns required by market participants making new capital investments in natural gas infrastructure outside of the NEB or provincial ROE formulas. Figure Most Recently Completed and Proposed Canadian Pipeline Projects Project Participants Year In-Service Tolling Methodology Capital Structure ROE Completed Pipelines Alliance Enbridge Fort Chicago 000 Negotiated cost of service arrangement, year contracts with renewal options starting at year 0, accelerated depreciation for contracts that are not renewed in year 0 0/0.% M&NE Proposed Pipelines Brunswick Duke Emera ExxonMobil Emera 00 ROE of % for Canadian portion (for first five years) and % for U.S. portion. Single postage stamp rates for services offered Seeking approval for a confidential negotiated toll agreement with a shipper holding almost all of the project's capacity, as well as designation as a Group pipeline / 0/0 [].00%.00% to.00% Mackenzie Delta Imperial ConocoPhillips Shell ExxonMobil APG 00 Negotiated cost of service agreement with ROE equal to the NEB multi-pipeline 0/0 NEB ROE +.% ROE plus basis points for the first 0 years; after 0 years ROE will be set by negotiated settlement or by an application by the transporter to the NEB Source: NEB & Company Websites. Notes: [] 0% represents Brunswick's minimum equity component 0 III. EVALUATION OF GAZ MÉTRO S BUSINESS RISK SINCE AND 00 A. DETERMINANTS OF BUSINESS RISK AND THE COST OF CAPITAL Q. How does one define business risk in the context of a company s cost of capital? A. Business risk in the context of a company s cost of capital is the uncertainty in the income earning capability of the firm s assets that investors in the company s equity are

15 0 0 exposed to over the economic lifetime of the assets. Business risk has been traditionally subdivided into market demand, supply, competitive, regulatory and operating risks. Q. What kinds of risks matter the most in evaluating a company s business risk from a cost of capital perspective? A. The risks that matter the most from an equity investor s perspective are those that cannot be diversified away through the holding of a broad portfolio of securities. Risks that are hard to diversify are those that are generally correlated with the level of (and changes in) general economic activity. Such risks are referred to as systematic. Broadly speaking, systematic risks associated with the gas distribution business include uncertainties in the demand for, and supply of, distribution services that are affected by changes in economic activity, including incomes and prices. Q. How does the fact that a company itself may be comprised of a diversified portfolio of businesses affect the evaluation of a company s business risk? A. Investors in a company that contains a diversified set of businesses will evaluate the total business risk of the company based on the weighted average risk of the portfolio of assets that makes up the company, where the weights are the market value shares of the individual businesses in the overall company portfolio. Investors will not view a company as less risky simply because the company has chosen to own a diversified set of assets. This is because investors themselves can always choose to diversify within their own investment portfolios, and therefore no additional risk-reduction value is assigned to a company that chooses to diversify. Q. How does rate regulation affect a company s business risk? A. On the one hand, rate regulation reduces a company s business risk if it provides equity investors some assurance that a fair return on and of capital will be earned over the lifetime of the firm s assets. On the other hand, regulation may enhance a company s business risk if investors perceive that there is uncertainty in the future regulatory National Energy Board, Reasons for Decision, TransCanada Pipelines Limited, RH--00, June 00, page.

16 treatment of the firm s businesses. That is why regulatory risk is sometimes evaluated as a separate component of business risk. While the equity securities of rate regulated firms are generally perceived as relatively stable, low-risk investments, the greater exposure of such firms to competition from other regulated and unregulated businesses, and alternative fuels, has changed that perception somewhat in recent years, particularly in the energy utility sector. B. GAZ MÉTRO S BUSINESS RISK ENVIRONMENT 0 0 Q. How would you characterize Gaz Métro s assets from a business risk perspective? A. Gaz Métro s assets are devoted to the provision of traditional gas distribution services to residential, commercial and industrial customers. Q0. How does Gaz Métro s gas distribution business subdivide in terms of customer classes? A0. In fiscal year 00, Gaz Métro served roughly,000 customers in five tariff classes, D (general service), D M (modular service), D (stable service), D (stable service for large volumes), and D (interruptible service). In terms of customer class, residential customers made up % of Gaz Métro s customers and % of its throughput, commercial customers made up % of customers and % of throughput, and industrial customers made up % of customers and % of throughput. Gaz Métro is forecasting total throughput of, 0 m for fiscal year 00, which represents a.% increase from the prior year estimate (the fiscal year 00 estimate) on a comparable weathernormalized basis. This.% increase in throughput is attributed to the increase in consumption of one large industrial customer, partially offset by a decrease in the consumption of small and medium load customers. Q. How does Gaz Métro earn income on its distribution services? A. Since 000, Gaz Métro s rates have been set using a Performance Incentive Mechanism. Under this mechanism, Gaz Métro s rates are set prior to the start of each fiscal year by comparing its projected cost of service to a revenue cap. Gaz Métro s projected cost is computed using traditional cost of service ratemaking principles, with an allowed return

17 0 0 on equity set using the Régie-approved formula-based methodology at a deemed equity thickness. Its revenue cap is computed by escalating the prior years cap based on the Québec Consumer Price Index minus an X-factor. The revenue cap is adjusted to account for certain exogenous factors (for example, the impact of weather on revenues and the impact of changes in interest rates on Gaz Métro s allowed return) and exclusions (for example, the overall cost of energy efficiency programs and the impact of prior year overearnings). If the revenue cap is greater than Gaz Métro s projected cost of service, then rates are set so that the projected productivity gain is shared between customers and Gaz Métro, with the portion of the productivity gain retained by Gaz Métro returned to customers after five years. If the revenue cap is less than Gaz Métro s projected cost of service, then rates are set to recover Gaz Métro s projected cost of service. In this case, the difference between Gaz Métro s projected cost of service and the revenue cap must be repaid to customers before Gaz Métro can share in productivity gains or collect overearnings in later years. At the end of each fiscal year, Gaz Métro s actual productivity gain is compared to its expected productivity gain. If its actual productivity gain is greater than the expected productivity gain, then the resulting overearnings are shared between customers and Gaz Métro (in the form of a contribution to an energy efficiency fund and a reduction in rates in the second subsequent fiscal year). If the actual productivity gain is positive but less than the expected productivity gain, then Gaz Métro retains all of the actual productivity gain (since in this case customers received a full share of the expected gain prior to the start of the fiscal year). If the actual productivity gain is negative, then customers and Gaz Métro share the deficit (with customers share in the form of an increase in rates in the second subsequent fiscal year). In this case, Gaz Métro must repay customers for their share of the deficit with interest before Gaz Métro can share in productivity gains or overearnings in later years. Since 00, Gaz Métro s authorized incentive return has been capped at basis points (after-tax) above its allowed return. Gaz Métro can only receive an authorized incentive return if certain performance indicators are met.

18 0 0 Q. How has the Performance Incentive Mechanism changed since it was put in place in 000? A. Gaz Métro s Performance Incentive Mechanism has been modified twice since it took effect in October 000. The first modifications was effective October 00, and the second will take effect October 00. The modifications in October 00 involved changes to several parameters that govern the Performance Incentive Mechanism. When the Performance Incentive Mechanism took effect in October 000, the X-factor was 0.%, customers received.% of projected productivity gains and customers received % of overearnings. Starting in October 00, the X-factor was increased to 0.%, customers share of projected productivity gains was increased to 0%, and customers share of overearnings was increased to %. The modifications that will take effect in October 00 are more fundamental. Most significantly, in recognition of the fact that Gaz Métro s business reality has evolved since the performance incentive mechanism was introduced, and in particular because sales volumes are more volatile and per customer volumes are down, primarily because of energy conservation initiatives, a new exogenous factor will be introduced that partially accounts for declining average use by small and medium load (SML) customers. This exogenous factor will be based on revenue variation due to declines in SML customer weather normalized volumes two years prior to the fiscal year. More precisely, it is calculated as: Performance Incentive Mechanism Agreed in Negotiated Settlement Process (NSP) R--00, [Translation Not approved by Participants], April, 00, page 0.

19 0 -[(ΔV t- % + 0.%) * 0% * VSML * Rate t- ], where ΔV t- % = total normalized consumption of a certain group of SML customers for year t-, VSML = total volume consumed by SML customers, and Rate t- = rate based on rate brackets used to convert volumes into revenues for year t-. The X-factor will decrease to its original level, 0.%. This decrease in the X-factor is meant to account, in part, for the fact that there is no explicit adjustment for volume variation by large load customers. Several other modifications in the Performance Incentive Mechanism will take effect in October 00. The return of past productivity gains will be smoothed by basing them on a five-year moving average. Changes will be made in service quality requirements in order to give Gaz Métro a greater incentive to make improvements in service quality indicators. If Gaz Métro achieves between % and 00% on specified indicators, then it will retain a corresponding percentage of projected productivity gains and overearnings. 0 The Performance Incentive Mechanism will include a target return of $ million per year for meeting a cumulative Global Energy Efficiency Plan (GEEP) volume reduction of,000,000 m per year. Finally, transmission and load-balancing revenues will be treated differently in order to decrease overearnings associated with these services, and 0 SML customers served continuously since that have not required any sales intervention by Gaz Métro. If Gaz Métro achieves less than % on specified service quality indicators, then it retains no projected productivity gains or overearnings. This target return will be prorated if Gaz Métro only partially achieves the GEEP volume reduction. Revenue from financial transactions related to transmission and load-balancing services (which are transactions such as loans of unused storage space that do not have an impact on the total volume of transmission and load-balancing tools used during a fiscal year) will be forecast at the start of a fiscal year. If actual revenues are lower than projected revenues, then the deficit will be placed in a deferral account and recovered from customers. If actual revenues are greater than projected revenues, then the difference will be shared %/% between Gaz Métro and its customers, regardless of whether Gaz Métro attains an incentive award or overearnings. Operational transactions related to transmission and load-balancing services will also be forecast prior to the start of the fiscal year. At the end of the fiscal year, customers will receive % of earnings from operating transactions, calculated as (actual average operating (continued )

20 0 0 a new exogenous factor will be introduced to adjust for the impact of changes in income and capital tax rates on the cost of service. Q. Is Gaz Métro continuing to make capital investments in its distribution business? A. Yes. Gaz Métro is proposing capital expenditure of $. million for fiscal year 00. This includes $. million in network expansion expenditures, $. million in deferred charges expenditures (mainly commercial programs and IT development), and $. million in network improvement expenditures. Gaz Métro made capital expenditures of $. million in fiscal year 00, including $ million in expenditures actually incurred in 00 and 00 but included in rate base in 00 associated with serving TransCanada s Bécancour cogeneration project, which is now Gaz Métro s largest customer. Gaz Métro continues to put new capital at risk. Q. Does the fact that the markets Gaz Métro serves are expected to grow mean that its assets are likely to be exposed to less business risk? A. No. Risk is not about expectations alone. Risk involves the uncertainty associated with the expected outcomes. Some of the riskiest firms that one can evaluate from an investment perspective are those that serve high growth but highly uncertain markets such as telecommunications or technology. A high growth market is certainly a positive factor from an equity investor s perspective all else equal. However, that same investor will demand a higher rate of return if the expected growth is more uncertain. Q. What are the principal classes of business risk to which Gaz Métro is exposed? A. Gaz Métro is exposed to market risk in its gas distribution business. Gaz Métro is also exposed to regulatory risk related to its Performance Incentive Mechanism. Incentive ratemaking regimes create greater expected short-term variations in earnings than does traditional cost-of-service regulation, particularly where deferral accounts are used to adjust for cost under- or over-recovery. I will return to this topic in more detail below. transactions price projected average operating transactions price) * minimum [actual operating transactions volume, projected operating transactions volume].

21 0 0 Q. How does market risk manifest itself in Gaz Métro s gas distribution business? A. The market risk to which Gaz Métro is exposed in its distribution business manifests itself in uncertainty over the future utilization of its distribution assets. Because Gaz Métro s gas distribution assets are sunk investments, and cannot be redeployed easily to another use should market conditions change, Gaz Métro s future income earning capability depends critically on the maximum utilization of its assets. While Gaz Métro is compensated to some extent for declining volumes under its Performance Incentive Mechanism, this compensation does not provide Gaz Métro with assured cost recovery protection should its asset utilization differ from forecasts. In this way, Gaz Métro bears some market risk that depends on asset utilization. Q. What factors could affect the utilization of Gaz Métro s distribution assets? A. Distribution asset utilization is a function of the wholesale and retail price of the gas commodity itself, of the price of competing fuels, of general economic activity in its service area, and of weather deviations from normal forecast conditions. Of these risk factors, the ones most important to equity investors (i.e., those that are systematic) are the level of prices and economic activity. Weather deviations from normal are less important to equity investors because they are not likely to be correlated with the market and hence they are a diversifiable risk. Again, this is because investors themselves can cheaply diversify away risks that are not correlated with movements in the general economy by holding a portfolio of equities, such as broadly-based mutual funds. Moreover, Gaz Métro is compensated for weather deviations via a deferral account that is included in rate base and amortized over a five year period. Q. To this point you have not mentioned supply risk. Does Gaz Métro face supply risk in its gas distribution business? A. Not to a significant degree, in my opinion. This is partly because Gaz Métro s gas supply costs are a pass-through item in its customers bills. Of course, to the extent these supply costs rise, the market risk to which Gaz Métro is exposed increases, as I describe below. But that is not the same as supply risk in that Gaz Métro does not face a significant risk that the utilization of its facilities will be reduced due to the unavailability of supply.

22 C. EVIDENCE OF CHANGES IN GAZ MÉTRO S BUSINESS RISK SINCE AND 00 0 Q. What changes in the business risk to which Gaz Métro is exposed since and 00 have you identified? A. I have identified two areas in which there has been a measurable increase in Gaz Métro s business risk that would matter to investors in its equity securities. These are: ) Increases in the level and volatility of gas commodity prices and thus increased uncertainty in gas use per customer; and ) Increased uncertainty associated with Gaz Métro s earnings due to its Performance Incentive Mechanism.. Increases in Uncertainty in Gas Use Per Customer 0 Q0. In Section II, you discussed changes in the level and volatility of gas prices since, and particularly since 00. How do these changes in the natural gas commodity price environment translate to Gaz Métro and its customers? A0. Ultimately these price level and volatility changes in the wholesale market are reflected in the retail market. To see this, I have plotted Gaz Métro s total billing rate for several of its tariff classes in Figures through. Using Gaz Métro s rate D as an example, Figure shows that during the period October to September ( Pre- ) the average billing rate was cents/m with a standard deviation of. cents/m, and that this amount had increased by % over the period. During October to September 00 ( Post- ), the average rate D monthly bill was cents/m (a percent increase) while the standard deviation of those amounts increased to. cents/m. Starting in October 00 ( Post-00 ), the average rate D monthly bill increased further to cents/m (an percent increase over pre- levels) while the standard deviation of those amounts increased to. cents/m. 0

23 Figure Tariff D 0 0 Monthly Total Billing Rate ( /m³) Rate D Average % Change Standard Deviation Pre-.0.. Post-..0. Post Oct- Oct- Oct- Oct- Oct- Oct-00 Oct-0 Oct-0 Oct-0 Oct-0 Oct-0 Notes: "Pre-" is 0/ - /. "Post-" is 0/ - /0. "Post-00" is 0/0 - /0.

24 Figure Tariff DM 0 0 Monthly Total Billing Rate ( /m³) Rate DM Average % Change Standard Deviation Pre-...0 Post-... Post Oct- Oct- Oct- Oct- Oct- Oct-00 Oct-0 Oct-0 Oct-0 Oct-0 Oct-0 Notes: "Pre-" is 0/ - /. "Post-" is 0/ - /0. "Post-00" is 0/0 - /0.

25 Figure Tariff D 0 0 Monthly Total Billing Rate ( /m³) Rate D Average % Change Standard Deviation Pre Post-... Post Oct- Oct- Oct- Oct- Oct- Oct-00 Oct-0 Oct-0 Oct-0 Oct-0 Oct-0 Notes: "Pre-" is 0/ - /. "Post-" is 0/ - /0. "Post-00" is 0/0 - /0. These fundamental changes in the commodity price environment have begun to induce changes in customer use of Gaz Métro s network. This can be seen in decreases in Gaz Métro s historical and forecast usage per customer. Q. What do the usage per customer statistics show? A. The statistics show significant declines in use per customer over time. Gaz Métro has provided me with data on weather normalized average use per customer for several of its tariff classes. These data are presented in Figures 0 and below. For example, for tariff class D estimated weather normalized average use in was. 0 m. By In forecasting weather normalized usage, Gaz Métro considers: diminishing demand due to extreme gas prices, energy efficiency induced by Gaz Métro programs and undertaken independently by customers, relative fuel prices (oil and electricity vs. natural gas), Hydro-Québec special rates, macroeconomic indicators like Québec GDP, and decreases in 0-year average heating degree days for weather normalization.

26 00, estimated weather normalized average use declines to. 0 m, a.% decline. The weather normalized average use estimate for 00 based on the current price forecasts is. 0 m, a decline of an additional.%. Figure 0 Estimated Normalized Average Usage for Tariff Class D (Annual 0³m³ per Customer).0. Normalized Average Usage (0³m³)

27 Figure Estimated Normalized Average Usage for Tariff Classes D M and D (Annual 0³m³ per Customer) 00 0 Normalized Average Usage (0³m³) Q. Do the recent changes in Gaz Métro s Performance Incentive Mechanism, namely the new exogenous factor meant to account for declining usage, compensate Gaz Métro for these decreases in average normalized usage? A. No. The new exogenous factor meant to adjust for declining usage only accounts for declines in usage for Gaz Métro s SML (small and medium load) customers. Moreover, it is incorporated with a two year lag. Therefore, the fiscal year 00 revenue cap will be adjusted for annual variation in 00 SML customer usage. Q. Can you discuss how changes in natural gas prices and the prices of competing fuels since and 00 have affected Gaz Métro s industrial (or large load) customers? A. Yes. Figure below shows the relative cost of natural gas and fuel oil for large firm and interruptible customers (bars lower than 00% indicate that natural gas is more expensive than fuel oil). As Figure indicates, the competitive position of natural gas has eroded relative to fuel oil since.

28 Figure Relative Cost of Fuel Oil Vs. Natural Gas for Large Industrial Customers 0% 00% Firm Industrial Customers; # Fuel Oil,.% Sulphur Interruptible Industrial Customers; # Fuel Oil,.% Sulphur Over 00%, gas is cheaper 0% 0% 0% 0% 0% Source: Gaz Métro Analysis. Figure shows the number of customers and load for Gaz Métro s large industrial (tariffs D and D ) customers for selected years between and 00. Gaz Métro has lost many large industrial customers since. Its large customer load has decreased significantly since, although this decrease has been partially offset by the addition of the TransCanada Bécancour cogeneration plant in September 00. Figure Gaz Métro's Large Industrial Customer Load ( Tariff Classes D and D ) Number of Customers Volume (0 m ),0,0,0,0 Q. What do you conclude regarding the effects of the changed gas commodity environment on the risk associated with Gaz Métro s gas distribution business?

29 A. Equity investors looking at Gaz Métro s gas distribution business will be primarily concerned with uncertainty surrounding the short and long-term utilization of Gaz Métro s distribution assets. Since and particularly since 00 there has been a fundamental change in the natural gas commodity price environment in which Gaz Métro operates. This change has already begun to affect negatively the utilization of Gaz Métro s network across its rate classes, and there is substantial future uncertainty in this utilization. This represents a significant change in Gaz Métro s business risk.. Increased Risk due to Gaz Métro s Performance Incentive Mechanism 0 0 Q. How does Gaz Métro s Performance Incentive Mechanism differ from traditional cost of service regulation in terms of regulatory risk? A. In general, Gaz Métro s Performance Incentive Mechanism exposes equity investors to greater business risks than would traditional cost of service regulation. This is primarily due to the productivity gains and overearnings sharing and service performance threshold features of the Performance Incentive Mechanism. How much greater the risk is depends on the particular form of cost of service regulation to which it is being compared. In Canada, most traditional cost of service regimes employ deferral accounts to true up the incurrence of costs with the recovery of costs in future periods. This form of cost of service regulation exposes utilities to the least long-term cost recovery risk, and depending on the extent to which deferral accounts are employed, may reduce short term earnings volatility as well. In the U.S., most cost of service regimes do not make extensive use of deferral accounts, and many do not require utilities to file frequent rate cases. Under this approach the regulator is making use of regulatory lag (the time between cost-of-service rate cases) as an incentive mechanism. I.e., to the extent the utility is able to reduce costs or increase productivity between rate cases, its earned returns will vary between rate cases. As a result, this approach to cost of service regulation exposes utilities to somewhat greater business risks. Gaz Métro s Performance Incentive Mechanism exhibits some of the characteristics of cost of service regulation with regulatory lag, but it is a riskier regime because the

30 0 0 revenue cap includes the effect of projected productivity gains that must be shared with Gaz Métro s customers before any gains or overearnings accrue to Gaz Métro s shareholders. Q. Previously you mentioned some of the changes that have been made to Gaz Métro s recently approved Performance Incentive Mechanism that will take effect in October 00. What is the likely effect of those changes on Gaz Métro s business risk? A. The three modifications I discussed previously that would seem to have the greatest potential to affect Gaz Métro s business risk are ) the new exogenous factor for declining average use by SML customers; ) the smoothing of past productivity gains by basing them on a five-year moving average, and ) the modification of the service performance indicators to between % and 00% to increase Gaz Métro s incentives to improve service quality. The first two of these changes would be expected to reduce Gaz Métro s business risk, all else equal, because they are likely to reduce Gaz Métro s exposure to short term earnings volatility. Of course, all else is not equal in this case because declines in use per customer are continuing to be experienced, as I have shown, and the SML exogenous factor does not fully account for these effects. There is also no explicit adjustment being made for volume variation by large load customers, but the X- factor has also been reduced to 0.% to account, in part, for this fact. Finally, the modification of the service performance indicators can be expected to increase Gaz Métro s business risk because it raises the hurdle which must be overcome before Gaz Métro can retain the relevant percentages of any productivity gains or overearnings. Q. What do you conclude with respect to the effect of Gaz Métro s Performance Incentive Mechanism on its business risk? A. On balance, Gaz Métro s Performance Incentive Mechanism can be expected to increase its business risk relative to the traditional models of cost of service regulation employed for other utilities in North America. The recent changes to its Mechanism can be expected to reduce its risks directionally, and only partially in response to the more risky market environment to which Gaz Métro is now exposed.