5. Return on Common Equity

The legal standard for setting the fair rate of return has been established by the United States Supreme Court in the Bluefield and Hope cases.8 The Bluefield decision states that a public utility is entitled to earn a return upon the value of its property employed for the convenience of the public and sets forth parameters to assess a reasonable return.9 Such return should be equal to that generally being made at the same time and in the same general part of the country on investments in other business undertakings attended by corresponding risks and uncertainties. That return should also be reasonably sufficient to ensure confidence in the financial soundness of the utility, and adequate, under efficient management, to maintain and support its credit and to enable it to raise the money necessary for the proper discharge of its public duties.

The Hope decision reinforces the Bluefield decision and emphasizes that such returns should be sufficient to cover operating expenses and capital costs of the business. The capital cost of business includes debt service and stock dividends. The return should also be commensurate with returns available on alternative investments of comparable risks. However, in applying these parameters, we must not lose sight of our duty to utility ratepayers to protect them from unreasonable risks including risks of imprudent management.

We attempt to set the ROE at a level of return commensurate with market returns on investments having corresponding risks, and adequate to enable a utility to attract investors to finance the replacement and expansion of a utility's facilities to fulfill its public utility service obligation. To accomplish this objective, we have consistently evaluated analytical financial models as a starting point to arrive at a fair ROE.

5.1. Financial Models

The financial models commonly used in ROE proceedings are the Capital Asset Pricing Model (CAPM), Discounted Cash Flow (DCF) Analysis, and Historical Risk Premium Model (HRP). In addition to those financial models, SCE and SDG&E applied the Fama French Model to support their respective proposals. Detailed descriptions of these financial models are contained in the record and are not repeated here.

5.1.1. Proxy Groups

SCE, SDG&E and PG&E started with the same Value Line Investment Survey (Value Line) electric utility industry group list to establish proxy groups for their financial models. The utilities then screened those companies to ensure that the companies in their proxy groups were comparable to their respective utility. PG&E also selected a second proxy group consisting of non-utility companies.

The following tabulation summarizes the primary screens used by the utilities to verify the comparability of their utility proxy groups of companies:

The screening process resulted in a comparable group of 24 companies for SCE, 43 for SDG&E and 28 for PG&E. SDG&E's proxy group is larger than the other utilities because SDG&E employed fewer screens than SCE and PG&E.

The Division of Ratepayer Advocates (DRA) accepted the utility proxy groups and screens of SCE, SDG&E, and PG&E for use in DRA's financial models.

The Federal Executive Agencies (FEA) accepted the utility proxy groups of SCE, SDG&E and PG&E as base proxy groups. However, FEA applied an additional screen that excluded companies having less than 70% of revenues from electric utility operations. That additional screen resulted in FEA using smaller proxy groups in its financial models than SCE, SDG&E and PG&E. SCE and PG&E's proxy groups consisting of 24 and 28 companies, respectively, were each reduced to 22 companies. SDG&E's proxy group of 43 companies was reduced to 29 companies.

ATU declined to use the proxy groups selected by the utilities on the basis that ATU had no empirical basis to screen companies out of the proxy groups.11 Instead, ATU used a single proxy group consisting of 82 companies from Value Line's electric, combination and natural gas distribution utilities lists.

PG&E also used screens in selecting its non-utility proxy group of 54 companies. Those screens consisted of an S&P credit rating of A+ and three Value Line benchmarks. Those benchmarks were the payment of common dividends, a safety rank of 1,12 and a beta value of 0.94 or less.13

All parties used Value Line electric industry classifications for their proxy groups. A proxy group of non-utility companies was used solely by PG&E as a second proxy group. With each party using different companies in their proxy groups it is difficult to evaluate and assess the reasonableness of the financial model results and difficult to determine which results are comparable to SCE, SDG&E and PG&E.

Those difficulties are highlighted by ATU's lack of screening their proxy group and, unlike other parties, use of different companies from their proxy group for each of their financial models. For example, ATU used 75 of the 82 companies in their proxy group for their CAPM financial model and 65 companies in their DCF financial model.14 Approximately seven of the companies included in their proxy group did not have investment grade credit ratings.15 ATU used different companies from their proxy group in each of their financial models because they did not have sufficient data to run the individual models with their entire proxy group.16 In other instances, ATU deemed data not meaningful for inclusion in their financial models.17

A proxy, by common definition, is a substitute. Hence, companies selected for a proxy of a utility should have characteristics similar to the utility that the companies are selected to proxy. In order to assess comparability and reasonableness of financial model results, there should be no deviation from financial model to financial model of the companies selected for a proxy model. ATU's proxy group fails these basic principles. With ATU's use of different companies from its proxy group for each of its financial models, inclusion of non-investment grade companies, and exclusion of data for which no criteria or reasons have been given, comparability of their proxy group and financial model results can not be reasonably determined. Hence, the financial model results of ATU are given minimal weight for consideration in this proceeding.

PG&E utilized its non-utility proxy group on the basis that relative risk, not a particular business activity or degree of regulation, should be the salient criteria in establishing a meaningful benchmark to evaluate a fair rate of return.18 However, a 70 basis point CAPM and 250 basis point DCF model result differential between its utility and non-utility proxy groups leads us to question whether the non-utility proxy group is comparable to its utility proxy group and to PG&E.19

PG&E contends that the salient difference between the proxy groups, as compared with the utility companies, is that they have average investment risks associated with the non-utility companies having a higher credit rating of A+ versus BBB-, a safety range of 1 versus 3 or better, and financial strength of A+ versus B+. Accordingly, PG&E suggests that ROE estimates from its non-utility group should provide a conservative estimate of investors' required ROE for the utility operations of PG&E.20

PG&E has not substantiated that investment risks of its non-utility proxy group is comparable to the investment risks of its utility proxy group or to PG&E. The only explanation of comparability provided by PG&E was that its non-utility proxy group, having four indicators higher than its utility proxy group, demonstrates that PG&E needs an ROE that is higher than its utility proxy group results. However, risks between its non-utility and utility proxy groups associated with, among other matters, pricing and earnings do not appear to be comparable. Earnings for non-utility companies are dependent on the extent of competition and ability to price products or services at rates a buyer is willing to pay for a product or service while maintaining a competitive edge. Earnings for utility companies are dependent on a fair return on investments with reasonable pricing of utility services, irrespective of what a buyer is willing to pay for a product or service for which they may have no alternative.

Absent evidence that supports a comparison of the non-utility proxy group to its utility proxy group and to PG&E, we decline to consider the financial model results from PG&E's non-utility proxy group.

We acknowledge that each party should be able to choose a representation of companies that they deem appropriate for their proxy group. However, those proxy groups should be subject to basic criteria to enable us to determine comparability of the proxy groups to the utilities. As such, Value Line electric industry classifications should continue to be used in ROE proceedings where financial models require the use of a proxy group. Three basic screens should be used in selecting a comparable proxy group. Those screens are: (1) to exclude companies that do not have investment grade credit ratings, (2) exclude companies that do not have a history of paying dividends and (3) exclude companies undergoing a restructure or merger. Additional screens may be used to the extent that justification is provided. Also, to the extent feasible, parties should use comparable factors in their respective financial models.

We next review the financial models used by the parties to assess the comparability and reasonableness of their results.

5.1.2. CAPM

The CAPM is a risk premium approach that gauges an entity's cost of equity based on the sum of an interest rate on a risk-free bond and a risk premium. CAPM results for SCE ranged from a low of 8.80% by DRA to a high of 11.59% by SCE, results for SDG&E ranged from a low of 8.90% by DRA to a high of 11.73% by SDG&E, and results for PG&E ranged from a low of 8.90% by DRA to a high of 12.10% by PG&E. Those differences resulted in a 279 basis point spread for SCE, 283 basis points for SDG&E and 320 basis points for PG&E.

5.1.3. Risk-Free Rate

The risk-free rate is based on long-term treasuries. SCE, SDG&E and PG&E used a 2008 forecasted 20-year treasury rate for their risk-free rates.21 However, those rates were not comparable because they each used different time periods and sources. SCE used Global Insight's April 2007 forecast of an average 30-year treasury rate plus 12 basis points to project a 20-year rate.22 SDG&E used an average of Global Insight's 2008 forecasted 10-year and 30-year treasury rates. PG&E used a 2008 projected yield for a 20-year treasury from Global Insight, EIA and Blue Chip.

FEA and DRA also used long-term treasuries for their risk-free rates. However, both of those parties based their risk-free rate on historical data. FEA relied on a recent six-week time period for a 4.90% risk-free rate and DRA relied on actual July 26, 2007 rate of 5.00% based on 10-year and 30-year treasury rates.23

5.1.4. Risk Premium

A beta coefficient (BETA), a company-specific multiplier of general market risk, is the first of two components used to derive a risk premium. Each party used a different composite beta due to the use of different companies in their proxy groups and due to the use of different data sources and time periods. Irrespective of these differences, there was only a ten basis point difference in composite betas between the parties, which ranged from .85 to .95. That difference in betas is minor and does not significantly affect the CAPM equity cost rate results.24

Equity risk is the other component used to derive a risk premium. A risk premium is derived by multiplying a beta to an equity risk. Two methods, arithmetic and geometric, were used to calculate equity risks.25 SCE, SDG&E and PG&E used a 7.10% arithmetic equity risk based on Morningstar's Stocks, Bonds, Bills, and Inflation Valuation Edition Yearbook.26 FEA and DRA used arithmetic and geometric averages to develop their equity risk. FEA used a 5.30% mid-point long-term equity risk from a financial study while DRA used a 4.14% equity risk based on the results of studies performed by over a dozen academic scholars, the individual results of which ranged from a low of 2.00% to a high of 7.14%.

5.1.5. Discussion

Financial models, such as the CAPM, are dependent on subjective inputs. However, to the extent that the criteria for those subjective inputs can be standardized, the spread between the parties' financial model results will diminish, avoid controversy, and provide us with comparable results.

The CAPM risk-free rate is a good candidate for criteria standardization given that the parties all used long-term treasuries for their risk-free rates. Although the risk-free rate differed among the parties, those differences occurred from the use of different time periods. This is a test year ROE proceeding and, as such, risk-free rates should be based on what the interest rate is forecasted to be in the test year.

In prior ROE proceedings, the utilities have been allowed to update their test year debt and preferred stock costs based on test year forecasted interest rates from Global Insight. This ROE proceeding is no exception. With prior Commission adopted reliance on the forecasted interest rates of Global Insight, the risk-free rate should be based on Global Insight's long-term treasury rate for the test year. Hence, the individual parties' risk-free rates should be adjusted to a uniform rate. Although each party used a different method to derive their risk-free rates, we will apply a uniform 4.78% rate, based on a simple average of Global Insight's September updated forecast of 4.65% for a 10-year treasury and 4.91% for a 30-year treasury for this proceeding. 27 This is the same method that SDG&E used to arrive at its risk-free rate.28

The risk premium components of a CAPM are not viable candidates for standardization. That is because they are derived from subjective financial literature on historical and geometric data of which there is no evidence of a consensus on a preferred or acceptable method to mitigate perceived equity risks by investors. That lack of consensus is the primary cause of the approximate 300 basis point difference between the low and high CAPM estimates for SCE, SDG&E and PG&E.

The following tabulation compares the individual CAPM results based on risk-free rates derived by the individual parties to a uniform 4.78% risk-free rate.29 The adjusted model results are incorporated into our overall evaluation of the financial models results for each utility prior to applying informed judgment to arrive at a fair return on equity for SCE, SDG&E and PG&E.

 

SCE

SDG&E

PG&E

 

PRE

POST

PRE

POST

PRE

POST

Utility

11.59%

11.03%

11.73%

11.38%

12.10%

11.45%

FEA

9.41

9.29

9.41

9.29

9.56

9.44

DRA

8.80

8.59

8.90

8.63

8.90

8.68

5.2. DCF

The DCF model is used to estimate an equity return from a proxy group by adding estimated dividend yields to investors' expected long-term dividend growth rate.

DCF results for SCE ranged from a low of 9.11% by FEA to a high of 10.96% by SCE, results for SDG&E ranged from a low of 9.14% by FEA to a high of 11.11% by SDG&E, and results for PG&E ranged from a low of 8.99% by FEA to a high of 10.30% by PG&E.30 Those differences resulted in a spread of 185 basis points for SCE, 197 basis points for SDG&E and 131 basis points for PG&E.

5.2.1. Dividend Yield

Although the utilities and interested parties calculated their dividend yields differently they each derived a 3.50% rate, when rounded to a tenth of a percent.31 For example, SDG&E used a full year 2006 dividend on its proxy group as reported by Value Line, DRA used an average six months and July 2007 dividend yield for its proxy groups and FEA used Value Line year-ahead dividend per share projection for its proxy groups.

5.2.2. Expected Growth Rate

The expected growth rate amongst the parties varied from a low of 5.00% to a high of 7.00%. This variation was due to the use of different time periods and different financial services that provide historic and forecasted growth rates. SCE, SDG&E and PG&E based their expected growth rates on five-year growth forecasts applicable to their proxy groups from various financial services. DRA based its expected growth rates on an average of five-year and ten-year growth forecasts from three financial services. FEA relied on both a five-year historical and five-year growth forecast from Value Line, Reuters and Zacks.

5.2.3. Discussion

Differences in the DCF results are attributed to subjective growth forecasts for which there is no consensus on an acceptable method to derive them. The following DCF results are incorporated into our overall evaluation of the financial models for each utility prior to applying informed judgment to arrive at a fair return on equity for SCE, SDG&E and PG&E.

5.2.4. Historical Risk Premium (HRP)

Similar to the CAPM, the HRP is a financial model that gauges a company's cost of equity capital. The cost of equity capital is measured by adding a risk premium to a risk-free long-term treasury or utility bond yield. A risk premium is derived by an assessment of historic utility stock and bond returns.

SCE, SDG&E and PG&E used the April 2007 Global Insight 2008 forecasted bond rate for their base yields. SCE and SDG&E used a 6.23% AA utility bond forecast and PG&E used a 7.00% BBB forecast. SDG&E conducted a second HRP analysis using a projected 20-year Treasury yield by averaging Global Insight's 10-year and 30-year forecasts.

Unlike the utilities' use of forecasted rates for their base yields, FEA and DRA used historical rates. FEA used a current 4.90% long-term treasury yield for each of the utilities. DRA used a current 6.00% Aa bond rate for SCE and SDG&E and a 6.25% current BBB bond rate for PG&E. Similar to SDG&E, DRA conducted a second HRP analysis using a current 4.88% long-term treasury rate.

SCE, SDG&E and PG&E derived arithmetical risk premiums by comparing S&P electric utility group historical returns to public utility bonds. SCE and PG&E used a 1946 to 2006 historical period and SDG&E used a 1937 to 2006 historical period. From those comparisons, SCE derived a risk premium of 4.64%, SDG&E 4.63% and PG&E 4.34%. SDG&E also derived a 5.23% risk premium from an alternate HRP based on its comparison of S&P's electric utility group historical returns to long-term treasuries.

FEA modified SDG&E's HRP to reflect a total return yield instead of an income return yield. FEA then averaged its 3.83% geometric mean result with its 4.85% arithmetic mean result for an overall 4.34% risk premium to be applied to SCE, SDG&E, and PG&E. DRA converted each of the utilities HRP arithmetic means to geometric means and derived a 3.88% risk premium for SCE, 3.77% for SDG&E based on long-term bonds and 3.84% based on 20-year treasuries, and 3.54% for PG&E.

The parties employed similar methodologies in arriving at their HRP base yields as they did in arriving at their CAPM risk-free rates. For the reasons addressed in our prior CAPM discussion, the HRP base yields should be uniform and consistent. Hence, the criterion to be used in deriving base yields for the HRP model should be based on Global Insight's test year forecast of utility bond rates. The utility bond yield for those utilities having less than an A bond rating, such as PG&E, should be estimated by adding a premium to the Global Insight forecasted test year AA bond rate based on the average premium for utility bonds rated Baa over bonds rated AA during the most recent 24 months, as published by Moody's.

Differences in the individual party's HRP risk premium are due to the same reasons addressed in our prior CAPM discussion. Although there is a lack of consensus on the use of subjective inputs for calculating risk premiums, spreads between the parties' individual HRP results adjusted for an uniform base yield are less than 80 basis points after reflecting uniform long-term treasury and utility long-term bond rates.

It is general knowledge that public utility bonds are riskier than treasury bonds. Absent specific evidence justifying the use of either treasury or utility bonds over the other in the HRP, adjusted utility bond results should be used in considering a reasonable ROE range for SCE, SDG&E and PG&E. Those adjusted model results are incorporated into our overall evaluation of the financial models results for each utility prior to applying informed judgment to arrive at a fair return on equity for SCE, SDG&E and PG&E. The following tabulation compares the individual HRP results based on their respective risk-free long-term and utility bond yields rates and the uniform utility long-term bond yield criterion.

5.2.5. Fama French

SCE and SDG&E also used a Fama French model to estimate their cost of equity capital. The model provides us with an additional methodology to consider in setting a fair and reasonable ROE. Fama French is a three factor model designed to reward investors for a market factor (CAPM), exposure to a size risk and exposure to a value risk (high book-to-market vs. low book-to-market). SCE and SDG&E used the same time periods and nearly identical annualized arithmetic averages and risk-free rate on their proxy groups. However, SCE derived an average 15.04% equity cost from its proxy group and SDG&E 13.89%.

The Fama French model was introduced in a California ROE proceeding in 2005 by SCE. In that proceeding, SCE derived a 14.00% average and 13.90% midpoint result from its use of the model, which were significantly higher than its requested 11.80% ROE. Unable to explain that difference, SCE acknowledged that it would take a few years working with the model to see if that phenomenon would persist and if the estimates derived from the model are highly unstable from period to period.34 That stability concern has yet to be answered.

Proponents of the model assert that Fama French is an improvement on the CAPM because it adds two variables for consideration, a difference between market values and book value and company size compared to a norm. Opponents of the model assert that insufficient information about the model requires additional subjective inputs based on information not readily available and provides excessive results, as summarized in the following tabulation:

Although the Fama French model has been used in other jurisdictions, it is not clear that the methodology has been accepted for regulated utilities. It was used in a Massachusetts case which resulted in a settlement and in a Nevada case which may have accepted it as a data point.37 Irrespective of its use in other jurisdictions, Fama French results continue to appear unrealistically high in comparison to the results of SCE and SDG&E's other financial models, as summarized in the following tabulation:

There is insufficient evidence to substantiate that the additional subjective risk factors, size and exposure, are relevant to companies in a regulated industry in a state in which over 50% of the energy utilities revenue requirements are protected by balancing account recovery.38 There is also insufficient evidence to validate that the Fama French results are reasonable compared to the CAPM, DCF and HRP model results.

We conclude there is insufficient evidence to assess the applicability of the Fama French model to California regulated utilities and decline to incorporate the Fama French results into our ROE analyses.

5.2.6. Summary

Although the parties agree that the financial models are objective, the results are dependent on subjective inputs, as we have addressed in our prior financial models discussion. From these broad financial models results the parties advance arguments in support of their respective analyses and in criticism of the input assumptions used by other parties. These arguments will not be addressed extensively in this opinion, since they do not materially alter the model results. However, it should be noted that none of the parties agreed with the financial model results of the others.

In the final analysis, it is the application of informed judgment, not the precision of financial models, which is the key to selecting a specific ROE estimate. We affirmed this view in D.89-10-031, noting that it is apparent that all these models have their flaws and, as we have routinely stated in past decisions, the models should not be used rigidly or as definitive proxies for the determination of the investor-required ROE. Consistent with that skepticism, we found no reason to adopt the financial modeling of any one party. The models are only helpful as rough gauges of the realm of reasonableness.

5.3. Additional Risk Factors

We also consider additional risk factors not specifically included in the financial models. Those additional risk factors fall into three categories: financial, business and regulatory.

5.3.1. Financial Risk

Financial risk is tied to the utility's capital structure. The proportion of its debt to permanent capital determines the level of financial risk that a utility faces. As a utility's debt ratio increases, a higher return on equity may be needed to compensate for that increased risk.

Debt equivalence has an impact on the financial risk of SCE, SDG&E, and PG&E. As recognized in D.04-12-047, debt equivalence has been reflected in the utilities' credit ratings since at least 1990. In D.05-12-043, we affirmed that debt equivalence would be assessed on a case-by-case basis along with other financial, regulatory and operational risks in setting a balanced capital structure and fair ROE. Our goal in so doing was, and continues to be, to provide reasonable confidence in the utilities' financial soundness, maintain and support investment-grade credit ratings, and provide utilities the ability to raise money necessary for the proper discharge of their public duty. We have no reason to change that goal. Debt equivalence is considered in arriving at an overall ROE.

Although SDG&E has proposed an equity rebalancing mechanism for its capital structure, as addressed in our subsequent SDG&E's ROE discussion, none of the utilities have proposed a major change in their capital structure for the test year. Hence, we find no additional financial risk to address associated with their debt/equity ratios.

5.3.2. Business Risk

Business risk pertains to uncertainties resulting from competition and the economy. That is, a utility that has the most variability in operating results has the most business risk. An increase in business risk can be caused by a variety of events that include deregulation, poor management, a failed marketing campaign and greater fixed costs in relationship to sales volume.

Those risks result from operating in a hybrid generation industry composed of unregulated generators and regulated utility generation leading to an uncertainty with respect to how the California electric industry will be structured in the future. The primary business risk identified by the parties in this proceeding is electric procurement.

The primary components of electric procurement risks are purchase power agreements, Direct Access and renewable energy. However, recent actions by the Commission are leading toward a more defined California electric industry. Those actions include the issuance of a proposed decision in R.07-05-025 initiating a rulemaking to determine whether, when, or how Direct Access should be restored in California and establishment of a 20% renewable goal by 2010, in comparison to other states looking at renewable standards in a 2015 to 2020 time period.

California energy utilities are facing electric procurement risks today that are similar to the procurement risks that were reflected in the utilities' currently authorized ROEs approved in D.05-12-043. In that proceeding, the base range of ROE for SCE was increased 70 basis points to account for electric procurement risk and debt equivalence, SDG&E was increased 50 basis points for debt leverage, debt equivalence and electric procurement risk, and PG&E was increased 70 basis points for debt leverage, debt equivalence and electric procurement risks. Included in those increased basis points was a premium for perceived regulatory risk in California.39 Although regulatory risk was not specifically broken out from those premium basis points, we opt to do so in this proceeding and will address regulatory risk next.

The Commission has a history of protecting ratepayers while making the utilities whole. As recognized by SCE, the Commission has done much to mitigate investor risks since the height of the California energy crisis.40 There is no basis to conclude that we will do otherwise in the future. With recent Commission action leading toward a more defined California electric industry, the procurement risks reflected in D.05-12-043 should subside. Based on informed judgment, a 50 basis point premium for debt leverage, debt equivalence and procurement risk should be added to the ROE base range for SCE and PG&E, a 40 basis point premium for debt leverage, debt equivalence and procurement risk should be added to the ROE base range for SDG&E. We next address whether a basis point premium is needed for regulatory risk.

5.3.3. Regulatory Risk

Regulatory risk pertains to new risks that investors may face from future regulatory actions that we, and other regulatory agencies, might take. Examples include the potential disallowance of operating expenses and rate base additions, comparability of utility ROEs throughout the United States and rating agencies' outlooks for the California regulatory environment.

ATU considered the California regulatory climate from an investor's standpoint to be above average.41 In support of that position they and FEA identified the favorable balancing and memorandum account treatment that California utilities receive. That evidence shows that the potential for disallowance of operating expenses and rate base additions are low given the utilities' ability to recover a substantial portion of their revenue requirements through balancing and memorandum accounts. For example, SCE recovers 55% of its revenue requirements through these mechanisms, SDG&E 60%, and PG&E 48% overall, 41% of its electric and 60% of its gas revenue requirements.42 Disallowances from these balancing and memorandum accounts have not been material.43 Although comparisons between balancing and memorandum accounts of California utilities to utilities in other states were not provided, the extent and recovery of California balancing and memorandum accounts should be considered a positive aspect of the California regulatory climate.

Also in support of a favorable regulatory climate in California, ATU identified 17 energy companies throughout the United States that were authorized an average 10.28% ROE in the first half of 2007 in comparison to authorized ROEs of California energy companies.44 However, it is difficult from this to conclude that California energy companies experienced a favorable regulatory climate or that they should earn ROEs comparable to those ROEs authorized for energy companies throughout the United States. That is because those 17 energy companies, some (or all) of which may be part of a holding company, were not specifically identified in any of the proxy groups used in this proceeding. More importantly, there is no evidence on whether the capital structures of those utilities are comparable to California utilities. Finally, there is no evidence as to whether those authorized ROEs were approved with conditions or resulted from settlements.45

We do observe that the outlook for California's regulatory climate from rating agencies and investment companies is mixed. Comments from these entities vary, including that the regulatory environment in California for electric utilities has become more predictable and reliable, remains reasonable and balanced, has vastly improved, and is below average.46 Most of these comments are from sources with which investors are familiar.

The below average rating comes from Value Line, a source which all parties to this proceeding relied on for inputs to their financial models and on which investors rely. We understand that investors perceive California to have regulatory risk. We also acknowledge that investors are awaiting California regulatory action on, among other matters, whether, when and how direct access would be restored in California and how our current hybrid energy market structure between regulation and competition will end up. There is, however, no evidence that this Commission would change its past practice of making utilities whole should direct access be reintroduced. Hence, investors' perceived California regulatory risks warrant a 10 basis point upward adjustment to the base ROE ranges being adopted in this proceeding.

5.3.4. Summary

In addition to addressing these three risk factors, we could analyze each of the risks identified by the utilities to determine an appropriate risk adjustment to the financial model results. However, irrespective of the final result of any such exercise, the utilities are being increasingly driven by business and regulatory factors that include energy availability, ability to attract capital to raise money for the proper discharge of their public utility duties and to maintain investment-grade creditworthiness, all of which are important components of the Hope and Bluefield decisions. Based on the above financial, business and regulatory risks discussion; our duty to utility shareholders to protect them from unreasonable risks; and, application of informed judgment we conclude that the ROE ranges being adopted in this proceeding warrant a cumulative 60 basis point upward adjustment for SCE and PG&E, and a cumulative 50 basis point upward adjustment for SDG&E.47 Based on that same analysis, we conclude that the adopted ROEs should be set at the upper end of an ROE range found just and reasonable for each utility.

Having addressed the generic factors used in setting an ROE we now address a fair and reasonable return for the individual utilities.

5.4. SCE's Return on Equity

We will establish a base ROE range from the financial model results which reflects our above analysis of additional financial, business and regulatory risks.

The following tabulation summarizes the results of the individual financial models used by the parties, excluding Fama French, and adjusted to reflect utility bond yields, including the simple weighted average of the financial model results and proposed test year ROE for SCE:

After considering the evidence on market conditions, trends, creditworthiness, interest rate forecasts, quantitative financial models, additional risk factors, and interest coverage presented by the parties and applying our informed judgment, we arrive at a base ROE range of 9.60% (simple average of FEA and DRA's adjusted financial model results) to 10.90% (simple average of SCE's adjusted financial model results).49 To that range we apply a 60 basis point upward adjustment for business and regulatory risks resulting in a fair and reasonable ROE range from 10.20% to 11.50%. Consistent with our conclusion that the adopted ROE should be set at the upper end of an ROE range found just and reasonable, the proposed decision recommended a 11.40% ROE for SCE's test year 2008. However, upon consideration of and agreement with of SCE's comments to the proposed decision and oral argument (November 26, 2007) that its ROE should be set higher because its credit statistics under a 11.40% ROE is weaker than SDG&E's and PG&E's credit statistics, we find that SCE's ROE should be set at 11.50%, the top of the ROE range found reasonable in this proceeding.

Having arrived at an authorized ROE for SCE, we must assess whether that ROE is sufficient to maintain and support its credit ratings. Last year, SCE regained a business profile score of five that it held before SCE's financial crisis. Hence, its credit rating has improved from its previous business profile score of six. Because SCE's debt equivalence is much greater than it was before SCE's financial crisis, it is necessary to consider what impact its debt equivalence will have on its test year credit rating.50 A comparison of SCE's currently authorized 11.60% ROE and requested 11.80% ROE set forth in Appendix A, which reflects a worst case scenario that the impact of debt equivalence and three common credit ratios are the only factors used to set a utility's rating, demonstrates that the adopted ROE, which falls 10 basis points below the bottom of that ROE range, would not materially change SCE's BBB+ credit rating position within the S&P benchmarks. SCE's cash flow interest coverage would remain within the S&P mid-A credit rating range, its debt to capital ratio would remain within the mid-BBB range, and its cash flow to debt ratio remains in the lower BBB range. A test year 2008 ROE of 11.50% for SCE is fair, reasonable and adequate for SCE to maintain and support its credit ratings. This ROE is reasonably sufficient to ensure confidence in the financial soundness of the utility while, at the same time, balance the interests between shareholders and ratepayers.

5.5. SDG&E's Return on Equity

In addition to seeking a company-wide ROE for its test year 2008, SDG&E seeks to bifurcate from its company-wide investments its investment in San Onofre Nuclear Generation Station (SONGS). SDG&E also seeks approval of an equity rebalancing mechanism. Both of these issues must be resolved prior to addressing a company-wide ROE for SDG&E.

5.5.1. San Onofre Nuclear Generation Station

SDG&E seeks an ROE on its minority investment in the SONGS that is the equal to the company-wide ROE of SCE, a majority partner in SONGS. The basis for this request is that SDG&E perceives significant risk in a steam generator replacement project (SGRP) involving SONGS 2 and SONGS 3 due to cost caps adopted by the Commission and due to system upgrades that require moving 600-ton steam generators by barge from Long Beach, up the beach, and into containment structures. In addition, SDG&E contends that the prospect of being required to participate in future capital projects represents significant cost management risks for SDG&E over which it has little control.51

SDG&E's current investment in SONGS is less than $50 million and almost fully depreciated. Hence SDG&E is no longer earning a return on that investment. The SGRP involving SONGS 2 is not expected to be completed and placed into rate base until 2010. The SONGS 3 portion of the SGRP is not expected to be completed and placed into service until 2010-2011. Until those projects are completed and placed in service, SDG&E will earn an interest allowance for its investment used during construction of that project. That interest allowance will be capitalized and recoverable through an authorized ROE when it is placed in service. SDG&E expects its SONGS investment, excluding the interest allowance, to reach approximately $180 million by 2012.

Although SDG&E perceives substantial risk in its minority partnership in the SONGS SGRP, it has not substantiated why its investment in the project should earn the same ROE as its majority partner.52 This proceeding pertains to a 2008 test year ROE for SDG&E. Given that SDG&E's current investment in SONGS is almost depreciated, there is no immediate impact on SDG&E. Company-wide risk is considered in arriving at a fair ROE along with the other factors addressed in this order. Although there is an ROE difference between SDG&E and its majority partner SCE, there are reasons for that difference. The ROEs are based on company-wide factors including different capital structures, debt costs and credit ratings.

SDG&E has not justified why its SONGS investment should earn the company-wide ROE of SCE. The Commission has already approved upfront ratemaking treatment for the SONGS SGRP.53 To the extent that SDG&E perceives substantial risk above that considered in its overall ROE due to its participation in the SGRP, consideration of reduced risk exposure of SCE should be considered as a result of SCE taking on a partner.

To the extent that SDG&E should be authorized to earn an ROE equal to that of SCE, that ROE should be based on the impact of consolidating the investments, risks, capital structures, debt costs and credit ratings of both utilities for that specific project. If those differences were consolidated, SDG&E may be authorized a higher ROE on its SONGS investment than its company-wide investments if its SONGS investment is found riskier than its company-wide risk. The opposite would result for SCE if its SONGS investment was found to be less risky than its company-wide risk. Approval of this request would also require the company-wide ROEs of SDG&E and SCE to be re-calibrated based on the exclusion of SONGS investments from their company-wide investments. SDG&E's request for an ROE on its SONGS investment equal to that of SCE is denied.

5.5.2. Equity Rebalancing Mechanism

SDG&E seeks an automatic equity rebalancing mechanism to mitigate its perceived adverse effect of debt equivalence54 and to mitigate any credit impacts of Variable Interest Entities (VIE), the accounting of which impacts its balance sheet as determined by Financial Accounting Standards Board Interpretation No. 46 (FIN 46). SDG&E proposes to use its rebalancing mechanism each time it enters into a new purchase power agreement (PPA) or contract that results in debt equivalence or FIN 46 treatment. Although SDG&E seeks approval of its equity rebalancing mechanism back to May 8, 2007, the date of its application, we decline that request on the basis that any approval back to May 8, 2007 would conflict with our practice of authorizing rates on a prospective basis. However, we will consider it on a prospective basis.

SDG&E proposed its rebalancing mechanism on the basis that debt equivalence and VIEs adversely impact its credit ratings and, unless mitigated, will negatively impact its credit profile. To mitigate that negative credit profile, SDGE proposes to add equity and to reduce its debt in an amount equal to its authorized equity factor of the imputed debt equivalent and VIEs. SDG&E would then calculate the revenue requirement effect of that rebalancing by using its authorized common equity return and factoring in a gross-up for income tax expense and authorized debt costs.55

There is no dispute that debt equivalence and FIN 46 costs are economic costs and not accounting costs.56 Although FIN 46 is a new issue, debt equivalence is not.57 SDG&E is proposing an automatic adjustment to its capital structure each time it negotiates a new PPA or adjusts its balance sheet to comply with FIN 46. Those automatic adjustments would result in incremental revenue requirement increases caused by SDG&E increasing its equity ratio.

SDG&E's current S&P credit rating of A is based on a number of factors. Three of the many factors considered by S&P are: funds from operations to adjusted debt; adjusted total debt to total capitalization; and funds from operations to interest coverage. S&P provides a range of acceptability for each of those factors. As shown in Appendix A, a S&P credit rating of A reflects, among other matters not specifically identified, a 30% to 22% range for funds from operations to adjusted debt, a 42% to 50% range for debt to capital, and a 4.5 times to 3.8 times range for cash flow to interest coverage.

As acknowledged by SDG&E "you don't know what is going to degrade and what condition is ultimately going to put the company in a position where its credit gets downgraded."58 Hence, an adjustment due to debt equivalence or FIN 46 treatment would not automatically result in a credit downgrade. Absent compelling evidence, we decline to begin unbundling ROE on a project by project basis. SDG&E's rebalancing mechanism is not adopted.

5.5.3. Summary

The following tabulation summarizes the results of the individual financial models used by the parties, excluding Fama French, adjusted to reflect uniform utility bond rates, including the simple weighted average of the financial model results and proposed test year ROE for SDG&E:

After considering the evidence on market conditions, trends, creditworthiness, interest rate forecasts, quantitative financial models, additional risk factors, and interest coverage presented by the parties and applying our informed judgment, we arrive at a base ROE range of 9.50% (simple average of FEA and DRA's adjusted financial model results) to 11.10% (simple average of SDG&E's adjusted financial model results). To that range, we make two adjustments.

The first adjustment results from a disparity of credit ratings among the utilities included in the proxy groups of SDG&E, FEA, and DRA. Approximately 60% of those utilities have a lower medium grade credit rating of BBB in comparison to SDG&E's upper medium grade credit rating of A.60 Ten percent of the utilities in the proxy group have an investment grade credit rating of BBB-, only one notch above the lowest investment grade credit rating. With BBB utilities being more risky than SDG&E, the financial models results are skewed toward a riskier side. Therefore, it is necessary to counter-balance the skewed financial models results that include more risky utilities. We adopt a 30 basis point downward adjustment to the base ROE range being adopted for SDG&E. In making this downward adjustment, we recognize that the BBB credit rating of SDG&E's parent company is comparable to the majority of the proxy group utilities. However, the appropriate comparison in this instance is SDG&E, a regulated entity under our jurisdiction, to a proxy group of utilities.

The second adjustment results from our discussion of additional risk factors. That adjustment is a 50 basis points upward adjustment for business and regulatory risks. Those adjustments result in a net 20 basis point increase (minus 30 basis points plus 50 basis points) in base financial model range of 9.50% to 11.10% resulting in a fair and reasonable ROE range from 9.70% to 11.30%. Consistent with our conclusion that the adopted ROE should be set at the upper end of an ROE range found just and reasonable, we find that an 11.10% ROE is fair and reasonable for SDG&E's test year 2008.

Having arrived at an authorized ROE for SDG&E, we must assess whether that ROE is sufficient to maintain and support its credit ratings. A comparison of SDG&E's currently authorized 10.70 ROE and requested 11.60% ROE set forth in Appendix A, which reflects a worst case scenario that the impact of debt equivalence and the three common credit ratios are the only factors used to set a utility's credit rating, demonstrates that the adopted ROE, which falls within that ROE range, would not materially change SDG&E's A credit rating position within the S&P benchmarks. SDG&E's cash flow interest coverage would remain above the S&P top-A credit rating range, its debt to capital ratio would remain within the mid-BBB range and its cash flow to debt ratio in the upper BBB range. A test year 2008 ROE of 11.10% for SDG&E is fair, reasonable and adequate for SDG&E to maintain and support its credit ratings. This ROE is reasonably sufficient to assure confidence in the financial soundness of the utility while at the same time balance the interests between shareholders and ratepayers.

5.6. PG&E's Return on Equity

In D.06-06-014, the Commission ordered PG&E to identify, explain, and compare its assumed and actual Pension fund ROE to its requested ROE in future cost-of-capital (ROE) proceedings. The purpose of this comparison is to assess whether pension return assumptions are comparable to the ROE used in utility ratemaking. Hence, this issue must be resolved prior to setting PG&E's test year 2008 ROE.

5.6.1. Pension ROE

PG&E reported an assumed 8.0% long-run projected pension return for its year-ended December 31, 2006, which included returns to equity, debt, and other assets. Its projected ROE on investments for the same period was 9.0%. That 9.0% pension return is 270 basis points lower than the 11.70% ROE that PG&E is requesting for its company-wide test year 2008.

PG&E disputes the comparability of its pension ROE to its company-wide ROE and provides three reasons for the non-comparability. First, the pension ROE is not equivalent to a specific benchmark for investors' forward-looking required ROE. Second, the pension projection applies to equity investments made in the pension portfolio selected by pension managers. Third, the pension plan projection used some, but not all, of the same historical information referenced in its ROE analysis.

The objectives of a pension fund are fundamentally different from that of an equity investor in a single utility and the risk profiles are not comparable. The Employee Retirement Income Security Act dictates that pension funds must be diversified whereas a utility's ROE is based on risks specific to that utility's operations.

More importantly, pension fund returns are related to market value of assets held in the pension fund while a utility's ROE is applied to a book value rate base. This difference can best be illustrated by dividing an average pension fund return by PG&E's market-to-book ratio. Based on ATU's 9.62% calculated average pension fund return and DRA's market-to-book ratio of 1.9 for PG&E, PG&E would only need to earn a 5.06% ROE on its rate base to equal the 9.62% average pension fund return.61 However, a 5.06% ROE is 116 basis points below its long-term debt cost, effectively eliminating PG&E's ability to support its credit and to raise the equity necessary to fulfill its public utility responsibilities as required by Bluefield and Hope. Pension return assumptions are not comparable to the ROE used in utility ratemaking. Having resolved this issue, PG&E should not be required to continue comparing its pension return assumptions to its ratemaking ROE in future ROE proceedings.62

5.6.2. Summary

The following tabulation summarizes the results of the individual financial models used by the parties, excluding Fama French, and adjusted to reflect utility bond yields, including the simple weighted average of the financial model results and proposed test year ROE for PG&E:

After considering the evidence on market conditions, trends, creditworthiness, interest rate forecasts, quantitative financial models, additional risk factors, and interest coverage presented by the parties and applying our informed judgment, we arrive at a base ROE range of 9.70% (simple average of FEA and DRA's adjusted financial model results) to 10.90% (simple average of PG&E's adjusted financial model results). To that range, we add a 60 basis point upward adjustment for business and regulatory risks resulting in a fair and reasonable ROE range from 10.30% to 11.50%. In setting PG&E's test year 2008 ROE, we recognize that in April of 2007, Moody's placed PG&E on review for possible upgrade and that S&P upgraded its rating for PG&E to BBB+ from BBB in May of 2007.64 Consistent with the recently favorable credit ratings of PG&E, our duty to utility ratepayers and our conclusion that the adopted ROE should be set at the upper end of an ROE range found just and reasonable, we find that PG&E's authorized ROE for test year 2008 should remain at 11.35% ROE.

Having arrived at an authorized ROE for PG&E, we must assess whether that ROE is sufficient to maintain and support its credit ratings. A comparison of PG&E's authorized 11.35% ROE and requested 11.70% ROE set forth in Appendix A, which reflects a worst case scenario that debt equivalence and the three common credit ratios are the only factors use to set a utility's credit rating, demonstrates that the adopted ROE would not materially change PG&E's BBB+ credit rating position within the S&P benchmarks. PG&E's cash flow interest coverage would remain above the S&P top-A credit rating range, its debt to capital ratio would remain within the top-BBB range, and its cash flow to debt ratio moves to the lower A range. A test year 2008 ROE of 11.35% for PG&E is fair, reasonable and adequate for PG&E to maintain and support its credit ratings. This ROE is reasonably sufficient to assure confidence in the financial soundness of the utility while balancing the interests between shareholders and ratepayers.

8 The Federal Power Commission v. Hope Natural Gas Company, 320 U.S. 591 (1944) and Bluefield Water Works & Improvement Company v. Public Service Commission of the State of Virginia, 262 U.S. 679 (1923).

9 Hope held that the value of a utility's property could be calculated based on the amount of prudent investment minus depreciation.

10 Investment research company long-term growth rates.

11 Exhibit 55, p. 19.

12 A safety rank of 1 is the highest rank from a range of 1 to 5.

13 This threshold is equal to the average beta value for PG&E's utility proxy group.

14 Reporter's Transcript, Volume 5, p. 626 and p. 628.

15 Id., p. 629.

16 Id., p. 626.

17 Id., p. 634.

18 Exhibit 23, pp. 2-21 and 2-22.

19 PG&E derived a 12.10% CAPM result from its utility proxy group and a 12.80% result from its non-utility proxy group. It derived a 10.30% DCF result from its utility proxy group and a 12.80% DCF result from its non-utility proxy group.

20 Exhibit 23, pp. 2-22 to 2-23.

21 SCE used a 5.33% risk free rate, SDG&E 5.13% and PG&E 5.40%.

22 Although Global Insight forecasts 30-year and 10-year treasuries it does not forecast an interest rate for 20-year treasuries.

23 Exhibit 44, p. 44 and 45 and Exhibit 34, p. 4-27.

24 Exhibit 34, p. 5-14.

25 An arithmetic average accumulates changes year by year. A geometric average measures changes over more than one period on a buy and hold strategy.

26 SCE and SDG&E relied on a 2007 edition as shown in Exhibit 4, p. D-1 and Exhibit 45, Attachment 5, respectively. PG&E relied on a 2006 edition, as shown in Exhibit 23, p. 2-46.

27 See, for example, Late-Filed Exhibits, 64, 65, and 66.

28 Exhibit 45, p. 11.

29 The pre CAPM equity result is the individual party's CAPM result based on that individual party's risk-free rate. The post CAPM equity result is the result of changing only the individual party's risk-free rate to Global Insight's September 2007 forecast of 2008 average 30-year and 10-year treasury rates.

30 The DCF results of FEA are the simple average of its standard, mechanical and two-stage DCF results.

31 There is a tenth of a percent minor deviation from FEA and DRA's calculated 3.4% dividend yields for SCE.

32 FEA DCF results reflect a simple average of its standard DCF, mechanical DCF and two-stage DCF results for SCE, SDG&E and PG&E.

33 Blanks exist in the pre-adjusted columns because not all parties reported RPM results based on long-term treasury rates.

34 D.05-12-043 (2005), mimeo., p. 31.

35 Exhibit 4, p. F-1.

36 Exhibit 45, Attachment 6.

37 Reporter's Transcript Vol. 4, p. 531.

38 Exhibit 6, p. A.1; Exhibit 2, p. 11; and, Exhibit 26, p. 3.

39 See for example D.05-12-043, mimeo., p. 23.

40 Exhibit 4, p. 7.

41 Reporter's Transcript Vol. 5, p. 617.

42 Exhibits 6, 2, and 26, respectively.

43 Reporter's Transcript Vol. 4, p. 429.

44 Exhibit 27, p. 2.

45 Under Rule 12.5 of this Commission's Rules of Practice and Procedure, settlements do not constitute approval of, or precedent for any future proceeding.

46 Exhibit 57, pp. 42, 44, 46, and 18 thought 20, respectively.

47 The 60 basis point upward adjustment consists of 50 basis points for business risk and 10 basis points for regulatory risks. The 50 basis point upward adjustment consists of 40 basis points for business risk and 10 basis points for regulatory risks.

48 Averages are rounded to a tenth of a percent.

49 FEA's argument that the equal weighting of one DCF result with two risk premium results inappropriately gives more weight to risk premium is moot. The same 9.60% to 10.90% range is derived by equally weighting the two risk premium results and equally weighing that result with the DCF result.

50 Exhibit 4, p. 5.

51 Exhibit 12, p. MMS-7.

52 SDG&E clarified that the ROE request pertains only to the equity portion of its SONGS investment.

53 Exhibit 55, p. 8.

54 Debt equivalence is a term used by credit analysts for treating long-term obligations which are perceived as analogous to the long-term debt, such as purchased power agreements, leases, or other contracts, as if they were debt in assessing an entity's credit rating.

55 Exhibit 12, p. MMS-22.

56 Reporter's Transcript Vol. 3, p. 209.

57 Neither SCE nor PG&E has raised a FIN 46 concern or presented any testimony on the impact of FIN 46 on their respective financial statement and earnings.

58 Id., p. 237.

59 Averages are rounded to a tenth of a percent.

60 Exhibit 2, p.2 and Exhibit 34, pp. 3-11 and 3-12.

61 Exhibit 52, p. 3, and Exhibit 34, p. 5-52.

62 This issue was raised in a PG&E general rate case proceeding. See D.06-06-014.

63 Averages are rounded to a tenth of a percent.

64 Exhibit 55, p. 3.

Previous PageTop Of PageNext PageGo To First Page