XII. Return on Common Equity

At issue is the appropriate ROE for PG&E's electric and gas distribution operations for 2001. 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 methods and risk factors prior to arriving at a fair ROE.

A. Analytical Methods

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

PG&E, FEA, Knecht and Czahar, and ORA, each of which used different assumptions and numbers to run their financial models, presented CAPM, DCF, and RPM financial models in this proceeding.

Aglet opted to adjust PG&E's last litigated ROE of 10.60% for test year 1999 to reflect PG&E's target debt rating and changes in interest rate forecasts and risk factors, as addressed in our subsequent risk factor discussion.

The financial models are used only to establish a range from which the parties apply their individual judgment to determine a fair ROE. Although the parties agree that the models are objective, the results are dependent on the subjective inputs. From these subjective inputs the parties advance arguments in support of their respective analyses and in criticism of the input assumption used by other parties. These arguments will not be addressed extensively in this opinion, since they do not alter the model results.

The following table summarizes the 7.16% - 14.70% broad ranges of results derived from the CAPM, DCF, and RPM quantitative financial models used by PG&E, FEA, Knecht & Czahar, and ORA to arrive at their respective recommended ROE for PG&E:

Party	CAPM	DCF	RPM
PG&E	11.70-12.70%	11.60-11.70%6	13.20-14.70%7
FEA	7.16-10.80	8.60-12.62	9.83-12.03
Knecht/Czahar	10.22	10.33-10.70	10.73
ORA8	9.38-10.24	10.98-12.07	9.43-10.35

B. Risk Factors

The parties fine-tuned the results of their respective financial models in arriving at their recommended ROEs based on informed judgment with respect to financial, business, and regulatory risks expected to occur in 2001.

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. In general, the lower the proportion of a utility's total capitalization consisting of common equity, the higher the financial risk. Therefore, as a utility's debt ratio increases, a higher ROE may be needed to compensate for that increased risk.

Business risk is defined to be the degree of variability in operating results. That is, a company 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 which include, but are not limited to, deregulation, poor management; a failed marketing campaign; fire in a factory; and, greater fixed costs in relationship to sales volume.

Regulatory risk pertains to new risks that may result from future regulatory action that this and other regulatory agencies might take. It also includes the potential disallowance of operating expenses and rate base additions.

An adjustment to reflect a change in financial risk is not appropriate in this proceeding because PG&E's capital structure being adopted in this proceeding is identical to its currently authorized capital structure. However, the parties differ on whether an adjustment is necessary to reflect changes in business and regulatory risks. The following table summarizes the ROE position of each party, which includes the effect of perceived business and regulatory risks.

Recommended Return
Party	Electric	Gas
PG&E	12.40%	12.40%
Aglet	10.50	10.50
FEA9	10.80	10.80
Knecht & Czahar	10.50	10.50
ORA	10.40	10.40

1. PG&E

a) Electric Distribution Operations

PG&E recommends a 12.50% to 13.50% ROE range for its electric distribution operations based on four benchmarks. These benchmarks include the last five years risk of integrated utilities, results of deregulated generation and regulated transmission and distribution in the United Kingdom, results of prior de-regulations such as the AT&T breakup, and results of its financial model study of gas distribution companies. However, PG&E rejects the results of its electric utility study on the basis that the electric industry restructuring makes the results of its electric utility DCF and CAPM models unreliable. It also concludes that its CAPM electric utility results are unreliable because "betas" are biased downward due to electric utility stocks being more sensitive to interest rate risks and because the U.S. financial market volatility in the summer of 1998 distorts the standard measurement of electric utility betas. Further, the standard form of CAPM underestimates the cost of capital for stocks with betas less than a 1.0 market average.

PG&E attempted to draw risk parallels between electric utilities recently divested from state ownership in Great Britain and the restructured telephone industry in the United States in a prior cost of capital proceeding to support its opinion that its investors demand more return from a distribution only utility company. We found, in a prior decision, no meaningful comparison between Great Britain electric utilities and United States electric utilities and no meaningful relationship between the deregulation of the telephone industry and the deregulation of the electric distribution industry.¹⁰

b) Gas Distribution Operations

To determine a reasonable ROE for its gas distribution operations, PG&E utilized CAPM, DCF, and RPM financial models. Its CAPM model was based on the three months ending March 2000 average yield to maturity on long-term U.S. Treasury bonds, average beta of the Value Line companies in PG&E's gas distribution companies, and the Ibbotson arithmetic mean risk premium on the S&P 500.

For its DCF financial model, PG&E utilized parallel models. One model was based on quarterly data with and without flotation cost and the other model was based on annual data with and without flotation cost. Its RPM model was based on a study of comparable returns received by bond and stock investors over the last 62 years.

(1) Flotation cost

PG&E included a 25 basis points allowance for flotation costs in each of its CAPM, DCF, and RPM models because it believes that the inclusion of flotation cost provides the best estimate of cost of capital. Flotation costs are the costs incurred in issuing new stock. Flotation costs commonly include underwriter costs for marketing, consulting, printing and distribution projects, legal costs and discounts that must be provided to place new stock. The CAPM, DCF, and RPM financial models conducted by PG&E for its gas distribution operations results in a 12.10% to 14.20% ROE range with flotation cost compared to its 11.70% to 13.90% ROE range without flotation cost.

The inclusion of flotation cost in the various financial models is not a new issue. We concluded in D.92-11-047 that any merit to a flotation adjustment would apply only to existing stock at the time of actual new issuances. We also concluded in that decision that a flotation adjustment is not applicable to sales in the secondary market, and that such an adjustment is inappropriate as long as utility stocks are trading significantly above their book value. We further concluded in D.92-11-047 that any reconsideration of a flotation adjustment in a future proceeding would require a showing of theoretical, practical, utility and market specific data, and a showing that a flotation cost adjustment does not shift the burden of the transaction costs from investors to ratepayers. We would consider referring the flotation adjustment to a workshop only upon such a showing.¹¹

PG&E does not anticipate issuing any new stock in the 2001 test year. PG&E also confirms that nothing with respect to flotation cost has changed since we denied the use of flotation cost. Further, PG&E excludes the impact of flotation cost from its requested ROE in this proceeding. Despite this, PG&E now reargues including flotation cost in a proceeding which PG&E does not seek an adjustment for its inclusion and does not attempt to satisfy the requirements set forth in D.92-11-047. This is a nonproductive use of shareholder, ratepayers and the Commission's time. Absent PG&E's intent to seek recovery of flotation cost through its ROE and absent evidence to support its claim that the inclusion of flotation cost provides the best estimate of cost of capital, PG&E should not relitigate this issue in subsequent cost of capital proceedings.

(2) Quarterly DCF Model

Although PG&E utilized parallel DCF models, it relies on the 11.60% result of its annual DCF model instead of the 11.70% result of its quarterly DCF model. PG&E does not recommend the use of its quarterly DCF financial model results for this proceeding because the Commission has previously rejected the use of quarterly results.¹²

Similar to the issue of flotation cost, PG&E submits informational testimony on a quarterly DCF model which has been rejected by the Commission and which PG&E does not seek to utilize in this proceeding. Absent PG&E's intent to utilized a quarterly DCF model in seeking a ROE and absent evidence to substantiate that a quarterly DCF model is substantially superior to an annual DCF model, PG&E should not utilize a quarterly DCF model in subsequent cost of capital proceedings.

c) Summary

PG&E seeks a 12.40% ROE for its gas distribution operations. Its recommendation is based on the simple average of its 12.70% CAPM, 11.60% DCF, and 13.90% RPM financial model results from its selected gas distribution companies based on annual data without flotation cost. PG&E seeks the same 12.40% ROE for its electric distribution operations based on the results of its gas distribution financial models because its DCF and CAPM electric utility results are not reliable. PG&E believes that the gas distribution companies used in its gas financial models are a reasonable proxy for the risk of their electric distribution operations. This proposal is consistent with its 1999 cost of capital decision where we found no difference in PG&E's cost of capital between its gas and electric operations¹³

2. Aglet

Aglet arrived at its recommended 10.40% ROE for PG&E by reducing PG&E's last litigated ROE of 10.60% for test year 1999 by 45 basis points to restore PG&E's debt rating to its targeted single-A debt rating. It also reduced PG&E last litigated ROE by 20 basis points to reflect PG&E's improved earnings trend newly implement policy to protect shareholder earnings and improved regulatory climate in California. At the same time, Aglet increased PG&E's last litigated ROE by 55 basis points to reflect 110 basis points increase in DRI's 30-year Treasury bond forecast. The 110 basis points change is the difference between the October 1998 DRI forecast of 4.71% for 1999, PG&E's last litigated ROE proceeding, to the July 2000 DRI forecast of 5.81% for 2001, one basis points lower than the October 2000 DRI forecast of 5.82% for 2001.

3. FEA

FEA's witness arrived at FEA's 10.80% ROE recommendation based on the midpoint of its 10.70% upper end company specific DCF and 10.90% midpoint of its RPM financial model results. Although FEA also conducted a CAPM model, it rejected its CAPM results because of internal inconsistencies in beta data. No adjustment was made for changes in business and regulatory risks other than those included in FEA's financial models.

Subsequent to the close of the evidentiary hearing, FEA changed its recommended ROE from 10.80% to 11.00%, as addressed in its opening brief. The revised ROE was derived from the simple average of its 10.00% CAPM based on adjusted Value Line betas and its 11.90% RPM based on the average expected returns based on the RPM using DRI forecasted interest rates.

4. Knecht and Czahar

Knecht and Czahar also utilized CAPM, DCF, and RPM financial models. However, because of systematic implementation problems from using historical data that disabled their CAPM and RPM models, they relied on the simple average results of two DCF methods in recommending a 10.50% ROE. These DCF methods were the all utilities Value Line Dividends method with a 10.33% result and the all utilities time weighted retention ratio method with a 10.70% result.

a) Flotation cost

Included in the results of Knecht and Czahar's DCF models is a 4% premium for partial flotation cost. However, Knecht and Czahar have not submitted the necessary flotation cost showing set forth in D.92-11-047 as addressed in our prior PG&E flotation cost discussion to enable us to assess the reasonableness of a partial flotation cost adjustment. Consistent with our flotation cost discussion of PG&E's gas distribution financial model results, we reject Knecht and Czahar's DCF results that include the effect of flotation cost. With the exclusion of their partial flotation cost premium, Knecht and Czahar represent that their recommended 10.50% ROE would be reduced by 36 basis points to 10.24%.

b) Time Weighted Retention Ratio

The all utilities time-weighting retention ratio method deviated from the traditional DCF Model. Knecht and Czahar utilized a ten-percent time-weighting factor to give the greatest weight to the most recent years because they contend that the recent future is more similar to and a better predictor of the future than is the distant past. No other party utilized a time-weighting factor.

There is no evidence to substantiate the reasonableness of a ten-percent time-weighting factor, let alone any weighting factor, in the DCF model. Further, there is no testimony to enable us to determine whether a time weighted DCF model would be preferable over a non-time weighted DCF model. We reject the use of a time-weighting factor in the DCF model at this time.

c) Daily Betas

Knecht and Czahar also deviated from the traditional CAPM model by utilizing daily betas. They believe this approach would decrease bias and improve the estimate of betas over the use of monthly betas. The issue of betas is the subject of a workshop agreed to by all of the parties in PG&E's 1999 cost of capital proceeding and required by D.00-06-040. Because Knecht and Czahar's recommended ROE is not based on their CAPM results, the use of daily betas need not be addressed further in this proceeding. If this issue arises in future proceedings, we would expect the parties to address whether the use of monthly betas adequately addresses the need to analyze betas in a cost of capital proceeding. Parties should also address whether it is necessary and productive to analyze betas down to weeks or days.

5. ORA

ORA arrived at its recommended 10.40% ROE by averaging the results of its CAPM, DCF, and RPM financial models for gas and electric utilities and weighting the resultant averages by PG&E's electric and gas distribution revenue requirement. No adjustment was made to the model results for changes in business and regulatory risks other than those included in the financial models.

6. Subsequent Events

Subsequent to the October 30, 2000 submittal date of this proceeding, significant changes began to occur in the California energy market that impacted PG&E's ability to attract capital. Because the ability to attract capital is an important issue in determining an appropriate ROE for PG&E, we reopen this proceeding to take official notice of the documents listed in Exhibit B. All parties have the opportunity to address the subsequent inclusion of official notice items summarized below as part of their comments to the proposed decision, as described in Section 16.

PG&E informed the Securities and Exchange Commission (SEC) in its January 2, 2001 8-K filing that its financial condition has deteriorated to an extent that it expects to run out of cash by late January or early February.¹⁴ By D.01-01-018, the Commission, acting on a PG&E emergency request, allowed PG&E to raise its revenues by increasing the electric bill of each customer by one cent per kilowatt-hour. The increase is a temporary surcharge intended to improve the ability of PG&E to cover its cost of procuring future energy in wholesale markets that it cannot produce itself to serve its energy load.

As PG&E reported in its January 5, 2001 8-K filing, its credit ratings were downgraded by its primary rating agencies. Standard & Poor's (S&P) reduced PG&E's corporate credit rating from A+/A-1 to BBB-/A-3 and senior secured debt from AA- to BBB. Moody's Investor Service (Moody's) reduced PG&E's corporate credit rating from A3 to Baa3 and its senior secured debt from A1 to Baa2. Downgrading makes it more difficult and costly for PG&E to attract new capital. Subsequently, PG&E reported to the SEC through its January 10, 2001 8-K filing that it is unable to borrow money and is foreclosed from the capital markets because of its current financial condition.

The January 10, 2001 8-K filing also reported that PG&E suspended its $110 million fourth quarter 2000 common stock dividend and decided not to declare its regular preferred stock dividends for the three-month period ending January 31, 2001. Because dividends on all utility preferred stock are cumulative, the utility is precluded from paying any dividends on its common stock until all cumulative preferred stock dividends are paid.

Finally, PG&E reported to the SEC, in its January 17, 2001 8-K filing, that the rating agencies further downgraded its credit ratings on January 16, 2001, to a level considered by the financial community to be below investment grade. S&P downgraded PG&E's corporate credit rating from BBB-/A-3 to CC/C and PG&E's senior secured debt from BBB to CCC, well below S&P's minimum investment grade rating of BBB-. Moody's downgraded PG&E's corporate credit rating from Baa3 to Caa3 and PG&E's senior secured debt from Baa2 to B3, well below Moody's minimum investment grade rating of Baa3.

C. Conclusion

It is reasonable to consider the current estimate and anomalous behavior of interest rates when making a final decision on authorizing a fair ROE. We must also consider the impact of the recent downgrades. 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."¹⁵ Consistent with this statement, we have had a practice of using one-half to two-thirds of the change in the benchmark interest rate.¹⁶ We see no reason to do otherwise in this proceeding. We will continue to moderate changes in ROE relative to the most recent changes in interest rates.

PG&E tested the reasonableness of its 12.40% ROE request with a comparison of its 1990 to 1999 authorized ROEs to the DRI forecast of AA utility bond for each of its test years. Based on that test, PG&E observed that its requested ROE is below the 385 basis points average spread of DRI AA utility bond forecast during the ten year period. It also observed that its requested ROE is below an updating of its 1999 ROE with 44% of the change in the DRI forecast of AA utility bond that has occurred from October 1998 to August 2000. Finally, it observed that its requested ROE is within the 10.60% to 12.60% range between its authorized ROEs and DRI forecast of AA utility bonds for the past ten years.

PG&E's reasonableness approach is similar to the method utilized by Aglet to establish its recommended ROE for PG&E. Aglet updated PG&E's 1999 authorized ROE to reflect an upward change in interest rates. However, Aglet utilized 50% of the upward change in the 30-year Treasury bond forecast instead of PG&E's 44% change in the AA utility bond forecast.

Although both PG&E and Aglet used PG&E's last litigated cost of capital proceeding as a base, forecasted interest rates apparently peaked in PG&E's 2000 cost of capital proceeding, which resulted in an all-party settlement. The reasonableness of PG&E's 2000 cost of capital all-party settlement was tested against the 10.90% to 11.30% ROE range found reasonable in that proceeding before being adopted. Although that range reflected no change in risk, it did reflect a substantial increase in forecasted interest rates.¹⁷ For example, the DRI forecast for the 30-year Treasury bond increased approximately 140 basis points and the AA utility bond increased approximately 185 basis points from PG&E's 1999 cost of capital proceeding to its 2000 cost of capital proceeding.

This trend of a substantial increase in forecasted interest rates has stopped and, in this proceeding, changed to a downward trend. The October 2000 DRI forecast being used in this proceeding show that interest rates are declining from the April 2000 DRI forecast used in PG&E's 2000 cost of capital proceeding. The 30-year Treasury bond forecast is approximately 20 basis points lower and AA utility bond forecast is approximately 20 basis points lower. Hence, a downward adjustment to the financial models found reasonable in this proceeding is justified to reflect lower forecasted interest rates and to continue making changes in the ROE relative to changes in interest rates.

Although PG&E testified that its risk has not change in the past few years, Aglet demonstrated that PG&E's risk has been reduced due to an improved regulatory climate and PG&E's stated intent to not have shareholders subsidize its ratepayers. This reduced risk may warrant a moderate reduction in PG&E's authorized ROE. However, we find no basis to accept Aglet's proposed 20 basis points reduction in ROE for such reduced risk, especially since taking official notice of the significant charges in the California energy market that began to occur after this proceeding was first submitted.

We need not be concerned that a 45 basis points reduction in ROE, as proposed by Aglet, to bring PG&E's debt rating down from a AA minus/negative to its targeted single-A debt rating would result in a downgrading of PG&E's debt rating. As PG&E informed the SEC in its recent 8-K filings, PG&E's credit ratings have since been downgraded twice and now considered to be below a minimum investment grade. These recent downgradings restrict PG&E's ability to attract capital. This is because bond ratings are developed using total equity and debt, including short-term debt. Hence, a 45 basis points reduction in PG&E's ROE as proposed by Aglet would only drive PG&E's credit ratings further below a minimum investment grade and result in greater risk to PG&E.

In the final analysis, it is the application of judgment, not the precision of these models, which is the key to selecting a specific ROE estimate within the range predicted by analysis. We affirmed this view in D.89-10-031,¹⁸ which established rates of return for GTE California, Inc. and Pacific Bell, noting that we continue to view these models with considerable skepticism.

As addressed in our discussion of analytical methods and risk factors, parties derived different results because of their use of subjective inputs. Those parties that used the CAPM, DCF, and RPM models each used different proxy groups, risk-free rates, betas, market risk premiums, growth rates, calculations of historical market returns, and time periods within their respective models. Although each party addressed the strengths of their respective financial modeling results, other parties addressed their defects.

After considering the evidence on the market conditions, trends, interest rate forecasts, quantitative financial models, risk factors, and interest coverage presented by the parties and applying our informed judgment, we conclude that a ROE range from 10.76% to 11.25% is just and reasonable for PG&E. The downward trend in interest rates warrants a ROE at the lower end of this range. However, with the recent turmoil in the California energy market, deterioration of PG&E's financial condition and downgrading of PG&E's credit ratings to below investment grade offsets the setting of PG&E's ROE at the lower end of the range. We conclude that PG&E should maintain its currently authorized ROE, which is at the upper end of this range. Hence, we adopt a 11.22% ROE as being just and reasonable for PG&E. The 11.22% ROE is within the 10.60% to 12.60% ROE range used by PG&E to test the reasonableness of its requested ROE. It is also within the 10.90% to 11.30% ROE range found reasonable for PG&E's 2000 ROE adjusted downward to reflect recent results from the financial models in this proceeding and a recent downward trend in the AA utility bond forecast.