V. 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.14 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. 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 assure 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 and risk factors prior to exercising informed judgment to arrive at a fair ROE.

A. Financial Models

Historically, quantitative financial models are used as a starting point to estimate a fair ROE. The models commonly used in ROE proceedings are the Capital Asset Pricing Model, Discounted Cash Flow Analysis, and Market Risk Premium. Detailed descriptions of each financial model are contained in the record and are not repeated here.

Although the parties agree that the models are objective, the results are dependent on subjective inputs. For example each party uses different proxy groups, risk-free rates, beta, market risk premiums, growth rates, calculations of market returns, and time periods within their respective financial models. It is the application of these subjective inputs that result in a wide range of ROEs being recommended by the parties.

The overall ROE range for PG&E's electric operations is from 9.05% to 15.80%, PG&E's gas operations is 9.35% to 13.88%, SCE is 8.99% to 13.88%, Sierra is 8.89% to 13.85%, SDG&E's electric operations is 8.97% to 14.85%, and SDG&E's gas operations is 9.35% to 14.85%. PG&E has both the largest and smallest overall ROE range, with its electric operations having a 675 basis points spread and its gas operations 453 basis points.15 A tabulation summarizing the broad ROE range results derived from the various financial models used by the utilities, Aglet Consumer Alliance (Aglet), Federal Executive Agencies (FEA), and ORA is set forth in Appendix A.

From these broad ROE ranges the parties advance arguments in support for 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 model results. The financial models are used only to establish a range from which the parties apply their individual judgment to determine a fair ROE.

B. Risk Factors

Risk factors consist of financial, business, and regulatory 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.

None of the utilities propose a major change in their capital structure for the test year. We have adopted a moderate change for Sierra, which increases its long-term debt to 54.87% from 47.56%, resulting in Sierra's shareholders taking on additional risk. Absent a shift in the other utilities' capital structure there is no additional financial risk associated with their debt/equity ratios to consider.16

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.

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.

C. Discussion

We must set the ROE at the lowest level that meets the test of reasonableness.17 At the same time our adopted ROE should be sufficient to provide a margin of safety for payment of interest and preferred dividends, to pay a reasonable common dividend, and to allow for some money to be kept in the business as retained earnings.

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, which established ROEs for GTE California, Inc. and Pacific Bell, noting that we continue to view the financial models with considerable skepticism.

We consistently consider the current estimate and anomalous behavior of interest rates when making a final decision on authorizing a fair ROE. In PG&E's 1997 cost of capital proceeding we stated "Our consistent practice has been to moderate changes in ROE relative to changes in interest rates in order to increase the stability of ROE over time."18 That consistent practice has also resulted in the practice of only adjusting rate of return by one half to two thirds of the change in the benchmark interest rate.19

In past Commission decisions, 30-year treasury bonds and AA utility bond interest rates were referenced as benchmarks for evaluating cost of common equity for electric utilities. In the 1999 cost of capital order, the Commission discussed evidence on the historic spread between authorized ROE and 30-year Treasury bond rates from test years 1990 to 1998, before deciding that AA utility bonds provide a better benchmark for setting the ROE.20 In the recent ROE application of SDG&E and other major energy utilities, we continued to rely on AA utility bonds during a period of time that the DRI forecast for 30-year Treasury bonds dropped during a tremendous turmoil in the foreign market.21 At that time investors were fleeing to the safety of U.S. Government backed securities, resulting in Treasury rates falling to unusual low rates in comparison to AA utility bonds.

Consistent with our practice to moderate changes in ROE relative to changes in interest rates we compare the most recent trend of DRI interest rate forecasts between the date that testimony was prepared in April to the date this matter was submitted in September, 2002. The interest rate trend is going in a downward direction. The September 2002 AA utility bond interest rate forecast for test year 2003 is 7.16%, a 46 basis point decline in interest rate from the April 2002 forecast of 7.62%.

We also compare the most recent DRI interest rate forecast to the DRI forecast used in the utilities prior ROE proceedings. Except for SDG&E, interest rates have changed materially downward when the test year 2003 forecast for the utilities is compared to the test year forecast used in their prior rate proceeding. For example, the forecast AA utility bond interest rate used to moderate PG&E's test year 2000 ROE was 7.72%,22 SCE's test year 1997 ROE was 7.92%; and, Sierra's test year 2001 was 7.48%.23 The AA utility bond interest rate used to moderate SDG&E's test year 1999 ROE was 5.87%, 129 basis points lower than the forecast AA utility bond interest rate for test year 2003.24 This downward trend in interest rates, except for SDG&E, warrants a downward adjustment in ROE.

The utilities perceive the existence of increased diversifiable and non-diversifiable business and regulatory risks.25 These risks include the financing of large under-collections, municipalization, Performance-Based Ratemaking mechanisms, cogeneration, direct access, distributed generation, substations, photovoltaic systems,26 diesel fuel, propane, power procurement, and legislation pending in Sacramento regarding generation, procurement, and purchase of high cost renewable energy.

The utilities contend that these perceived risks require higher ROEs so their shareholders can be fairly compensated and the utilities may continue to attract capital. To accomplish this result, each utility applied above average risk factors in their financial models. Sierra and SDG&E, apparently perceiving even higher risks, further adjusted their financial model results upward; Sierra by 75 basis points and SDG&E by 100 basis points.

While FEA also reflects above average risks for the California utilities in its financial models, Aglet and ORA do not. Aglet and ORA contend that no upward ROE adjustment is necessary because the risks perceived by the utilities are diversifiable, requiring no adjustment. They cite D.94-11-076 as support for excluding a diversifiable risk adjustment.

In that decision, the Commission concluded that little or no weight should be given to diversifiable risk and that general economic factors such as interest rates and financial market trends carry more weight than risks associated with individual utilities or utility industries.27 By that same decision we found that distinctions between diversifiable and non-diversifiable risks are not clear-cut.28

We could analyze each of the risks identified by the utilities to determine which are diversifiable, which are not, and which may be a hybrid. However, such an exercise is not productive at this time.29 This is because, irrespective of the final result of any such exercise, in the eyes of the public and investors the electric utility industry is highly unstable at this time. The utilities are being increasingly driven by industry-specific 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, important components of the Hope and Bluefield decisions.

Therefore, we find that there is a net increase in utility risks that warrants the ROEs being adopted in this proceeding to be set at the upward end of an ROE range found just and reasonable for each utility.

Having discussed the generic factors used in setting a fair ROE we next address a fair and reasonable ROE for each utility.

D. Fair and Reasonable ROEs

The utilities, interested parties, and ORA used their individual financial model results, risk assessments, and informed judgment to recommend an appropriate ROE for the utilities. The following tabulation summarizes these recommendations.

UTILITY AGLET FEA ORA

Both PG&E and Aglet recommend a ROE differential between PG&E's electric and gas operations. While PG&E seeks a 25 basis points differential between its electric and gas operations, with its electric operations being assigned the higher ROE, Aglet recommends a 10 basis points differential, with PG&E's gas operations being assigned the higher ROE.

PG&E bases its electric recommendation on the results of four financial models, each of which are further divided into four separate levels of risk. Two of the initial financial models are variations of the Empirical Capital Asset Pricing method (ECAPM) rejected in D.99-06-057 because it artificially raises the ROE requirement.31 Half of these models are based on financial data from eight electric utilities in states not active in energy restructuring and the other half are based on 12 electric utilities with 65% or more of their revenue coming from regulated electric operations. Its gas recommendation is based on the results of three financial models that use information from 12 of 19 selected local natural gas distribution companies.32

Aglet's ROE differential is based on its Discounted Cash Flow model result that shows local natural gas distribution companies in its model have ROEs 12 basis points higher than electric and combination utility ROEs in its study. Aglet also places reliance on evidence in other proceedings demonstrating that gas operations have higher sales volatility.

This opposing view on whether PG&E's electric or gas operations should have a higher ROE demonstrates subjectivity in the financial models and explains why we view the financial model results with skepticism. Upon review of the financial models pertaining to PG&E's electric and gas operations we concur with PG&E that a differential in ROE between its electric and gas operations is warranted because of differences in electric and gas financial model results, with its electric operations requiring a higher ROE. Based on the financial model results and informed judgment, we conclude that the differential between PG&E's electric and gas ROE should be 20 basis points.

Aglet recommends a 10 basis points ROE differential between SDG&E's electric and gas operations, with the higher ROE going to the gas operations. Aglet's recommendation is based on its Discounted Cash Flow model result showing that gas utility ROEs are 12 basis points higher than electric and combination utility ROEs and reliance on evidence in other proceedings demonstrating that gas operations have higher sales volatility. This is insufficient evidence for us to conclude that SDG&E's gas operations warrant an ROE different from its electric operations. Even if we were satisfied that sufficient evidence exists, this differential, when combined with changes in interest rates and risks factors, would not warrant a 10 basis points differential between SDG&E's electric and gas operations. We reject Aglet's ROE differential for SDG&E's gas operations.

After considering the evidence on the market conditions, trends, creditworthiness, interest rate forecasts, quantitative financial models based on subjective inputs, risks factors, and interest coverage presented by the parties and applying our informed judgment, we conclude that the following subjective range of ROE for the utilities are fair and reasonable.

ROE RANGE

More specifically, the ROE ranges reflect judgmental tempering of the financial model results, increased business and regulatory risks, and a downward trend in the test year AA utility bond interest rate forecast. For PG&E, its range also reflects no change in its capital structure, suspension of dividend payments, and a speculative grade credit rating. For SCE, the range also reflects no change in its capital structure, suspension of dividend payments, embedded debt cost higher than the most recent test year AA utility bond interest rate forecast, and a speculative grade credit rating. For Sierra, its range also reflects a change in its capital structure and a speculative grade credit rating. For SDG&E, its range also reflects no change in its capital structure, embedded long-term debt cost lower than the most recent test year AA utility bond interest rate forecast, and an investment grade credit rating.

Given the increased business and regulatory risks, downward trend in the test year AA utility bond interest rate forecast, embedded long-term debt cost in relation to test year AA utility bond interest rate forecast, and ability to pay dividends of the individual utilities, we adopt the following ROEs as being fair and reasonable for the 2003 test year.

14 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). 15 One basis point equals 0.01%. 16 As discussed previously, this proceeding will remain open to allow PG&E to update its capital structure, long-term debt and preferred stock costs, and return on common equity, as a result of the bankruptcy proceedings. 17 46 CPUC2d at 369 (1992), 78 CPUC at 723 (1975). 18 77 CPUC2d 556 at 563 (1996). 19 57 CPUC2d 533 at 549 (1994). 20 196 P.U.R. 4th, 438. 21 D.99-06-057, mimeo, p. 49 (1999). 22 See D.00-06-040, mimeo, p. 16 (2000). Although this decision involves a settlement not constituting approval of, or precedent regarding, any principle or issue of the settlement in that proceeding or in any future proceeding, that decision adopted, independent of any settlement, a range of ROE reflective of interest rate trends to assess the reasonableness of the settlement. 23 See 69 CPUC2d 327 at 339 (1996) and, D.00-12-062, mimeo, p. 20 (2000). 24 D.99-06-057 mimeo, p. 49 (1999). 25 Diversifiable risk, or unique risk, is the firm-specific risk that can be eliminated by assembly of a balanced portfolio of investments. 26 Photovoltaic systems produce electricity for a specific facility eliminating the need for a customer to purchase energy from a utility. 27 57 CPUC2d 533 at 550 (1994). 28 Id. at 549. 29 Although we are not conducting a diversifiable risk exercise in this proceeding, we are not reversing our stated position of excluding diversifiable risk. 30 FEA makes no ROE recommendation for Sierra. 31 Id. pp. 45-46. 32 PG&E excluded seven local natural gas distribution companies from its financial models because those companies have not paid dividends in the past five years, have fewer than three analysts forecasts or have announced a merger.

Previous PageTop Of PageNext PageGo To First Page