14. Cost of Capital

SCWC requests a rate of return on rate base of 10.15% for each of the years 2003, 2004 and 2005. ORA recommends the Commission adopt a rate of return on rate base of 8.54% for 2003, 8.52% for 2004 and 8.55% for 2005. In determining the proposed rates of return, both parties recommend adoption of a capital structure composed of 50% long-term debt and 50% equity. SCWC recommends a cost of debt of 7.8% and a return on equity of 12.45% for 2003, 2004 and 2005. ORA recommends a cost of debt of 7.67% for 2003, 7.63% for 2004 and 7.68% for 2005 and a return on equity of 9.41% for each of the years.

As discussed below, we adopt an average cost of debt of 7.67% in 2003, 7.63% in 2004 and 7.68% in 2005 and a return on equity of 9.90% for each of the years. This equates to a rate of return on rate base of 8.79% for test year 2003, 8.77% for 2004 and 8.79% for attrition year 2005.

14.1. Cost of Debt

SCWC derived its cost of debt by calculating the embedded cost of debt currently outstanding and the projected cost of new debt issues. SCWC derived a coupon rate of 8.00% for new debt issues by adding a spread of 150 basis points to the Blue Chip Financial Forecast of long-term treasury yields. The 150 basis point spread is an amount consistent with current spreads for an "A+" rated utility issue. To the 8.00% coupon rate for new debt issues, SCWC added 13 basis points to approximate the annual cost factor for issuing new debt, to arrive at a cost of new debt equal to 8.13%.

Once both the embedded cost of debt and the cost of new debt were calculated, SCWC adjusted the embedded cost of debt yearly to reflect reductions in debt due to sinking fund payments, maturities, and projected new debt issues. Accordingly, SCWC derived an estimated, effective cost of long-term debt for 2003 through 2005 of 7.77%, 7.83%, and 7.84%, respectively.

ORA updated the forecast used to determine the coupon rates, using the most recent DRI forecast. Table 5-1 in ORA's testimony shows its resulting forecast of the new issue coupon rate to be 6.63% for 2003 and 8.47% for 2005. The associated effective interest rates are shown to be 6.76% for the 2003 new issue and 8.62% for the 2005 new issue. ORA's resulting forecast of the average cost of debt is 7.67% in 2003, 7.63 in 2004 and 7.68 in 2005.

We reject SCWC's contention that ORA did not account for the issuance costs of new debt. As described above, ORA increased its forecast of the 2003 new issue coupon rate of 6.63% to an effective rate of 6.76% (13 basis points) and increased its forecast of the 2005 new issue coupon rate of 8.47% to an effective rate of 8.62% (15 basis points). This methodology is similar to SCWC's where it took the forecasted coupon rate of 8.00% for 2003 and 2005 and increased it by 13 basis points to derive the effective rate of 8.13%, for both years. Since ORA's forecast incorporates more recent information, to which SCWC does not object, we will adopt ORA's forecast of average long-term debt costs for the period 2003-2005.

14.2. Return on Equity

SCWC requests a return on equity of 12.45 percent. The utility performed four common equity market cost analyses, followed by three book value return on equity analyses. SCWC derived a market cost estimate of 10.87% by averaging costs indicated by the Ibbotson CAPM Method (11.98%), the Fama-French Three Factor Model (10.80%), the Risk Premium (RP) Analysis (10.89%) and Discounted Cash Flow (DCF) Model (9.81%).

After deriving the market-required return rates on market price of SCWC's equity, SCWC then converted those rates to "book value" equivalent return rates. SCWC asserts that a return based strictly on market prices is applicable directly to book value only if the price to book value ratio is already 1.00. Since its current market to book ratio is about 1.9, SCWC used three methods to make the book value conversion. SCWC's conversion of the DCF market cost results in a book value return on equity of 11.10%. The use of the Modigliani & Miller Conversion of Market Leveraged Cost results in a return of 11.58% and the Brigham Leverage Curve-Based Conversion results in a return of 11.66%. The average of the three methods is 11.45%.

To its estimated book value return on equity of 11.45%, SCWC added 10 basis points to account for incremental business risk resulting from return variability, market capitalization, customer mix, capitalization size shortfall and other specific risks. SCWC asserts that water utilities, being the most capital intensive of the utilities as well as the smallest, are particularly susceptible to the risk phenomena of combined forms of leverage and small size. In addition, SCWC added 90 basis points to its required book value return on equity to account for the risks created by the Commission's issuance of Resolution W-4294 and D.03-06-072 in the balancing account OIR.

ORA's 9.41% return on equity is derived from a quantitative analysis using two financial models, DCF and RP, to estimate investors' expected return on equity for SCWC. ORA applied both models to a group of comparable water utilities. The DCF model projected returns on equity of 8.00% based on the 3-month dividend yield, 8.00% based on the 6-month dividend yield, and 7.99% based on the 12-month dividend yield, with an average of 8.00%. The RP model combined average equity risk premiums with average interest forecasts for the test period (years 2003 to 2005). Based on the average 10-year risk premiums, ORA calculated an expected return on equity of 10.59% for the 10-year Treasury bond yield and 10.90% for the 30-year Treasury bond yield. Using the 5-year average risk premium produced expected returns of 10.70% for the 10-year Treasury bond yield and 11.12% for the 30-year Treasury bond yield. Averaged together, ORA calculated an average ROE of 10.83% based on the RP model. Averaging the results of the two financial models produces ORA's expected return on equity of 9.41%.

As discussed below, we have developed a return on equity range and determined that the 9.9% midpoint value of that range will provide SCWC an appropriate return on equity for the years 2003 through 2005.

14.3. Return on Equity - Discussion of Models

The differences in return on equity recommendations between SCWC and ORA are caused by (1) SCWC's adjustment to the model results to convert from a market basis to a book value basis, (2) SCWC's use of two additional financial models, (3) differences in the financial model assumptions and inputs, and (4) SCWC's adjustments to reflect additional risk. This section discusses the first three items

Regarding SCWC's position that it is inappropriate to compare book value returns with market derived returns, it is not clear that the adjustment or the magnitude of the adjustment proposed by SCWC is reasonable or necessary. SCWC's use of the market to book conversion raises its equity return recommendation from 10.87% to 11.45%, or 58 basis points. ORA argues plausibly that the current high market to book ratios for regulated water companies indicate that authorized returns should actually be lowered rather than raised. More importantly, we must recognize that comparisons and averaging of DCF and RP results have been the basis for many of our decisions regarding equity return levels. These past Commission authorized returns have not included SCWC's proposed adjustment. SCWC has not demonstrated, based on its financing experience, or any other practical criteria why the adjustment is necessary at this time. We are reluctant to make a make a major change such as this without convincing evidence supporting the necessity for the change. We will therefore not include the market to book conversion in determining a reasonable equity return. This treatment and the return on equity that we are authorizing today are in line with recent authorizations for other Class A water companies.

SCWC argues that an additional premium is necessary in factoring risk, because small companies are at greater risk than large companies. Specifically, in its CAPM analysis, SCWC uses a small company premium, which is calculated using data from the 2001 Ibbotson Associates SBBI yearbook. The data include all stocks listed on the New York Stock Exchange, American Stock Exchange, and the NASDAQ. The vast majority of the companies on these exchanges are non-regulated and non-water. The Commission has stated that water utilities should not be compared to companies in other industries (D.01-04-034, mimeo. at p.13-14; D.90-02-042, mimeo. at p. 38.) As stated in D.92-01-025, "[d]ue to the revenue recovery mechanisms in place for water utilities, we find that water utilities do not face the same overall risks as energy and telecommunications utilities." Therefore, for the determination of a reasonable range for equity returns we will not rely on the CAPM, nor will we use the Fama-French Three Factor Model, which appears to be similar in many respects to the CAPM.

The financial models we will consider are the DCF and RP models, both of which have been used in the past by the Commission in determining equity returns for Class A water utilities. Also, both models were run by ORA and SCWC, although there were significant differences in the results.

In its prepared testimony analysis, ORA did not include Artesian Resources as one of the comparable companies. In response to SCWC's criticism for the omission, ORA reran the DCF and RP models the same way they were run for its report, with the inclusion of Artesian Resources. Based on numbers provided during evidentiary hearing,⁴³ if Artesian Resources were included in the comparable group, ORA's DCF analysis would yield a return of 8.35% and the RP analysis would yield a return of 10.52%. The average of the methods would be 9.43%, two basis points higher than ORA's recommended value of 9.41%.

In many of our decisions we have defined a reasonable range for equity returns. The authorized return would fall somewhere within that range. Model results for both ORA and SCWC show a wide range of equity returns. ORA's original analysis resulted in an 8.00% result from its DCF model, and a 10.83% result from its RP model - 283 basis points apart, or a recommended average that is 17.6% higher than the DCF result, and 13.1% lower than the RP result. SCWC asserts that this large disparity of 283 basis points between the DCF and RP results makes the ORA analysis unreliable. With the inclusion of Artesian Resources in the comparable group, ORA's spread between the DCF and RP model results would be reduced to 217 basis points. In its market cost of equity analysis, SCWC shows a range from 9.81% to 11.98%, or 217 basis points. The spread between SCWC's DCF (9.81%) and RP (10.89%) results, which amounts to 108 basis points, is somewhat less than the comparable ORA spread. However, it is clear that a fairly large spread in model results is common and not counter to our objective to determine a reasonable range. Once such a range is determined, we will exercise our judgment in determining the authorized return for SCWC.

Regarding the differences in the results, SCWC and ORA criticize each other's models. For instance, for the DCF, ORA criticizes SCWC's adjustment of the dividend yields to account for market pressure and issuance costs. ORA states that, in D.92-11-047, the Commission rejected the use of issuance costs and sinking fund effects in determination of rate of return. SCWC criticizes ORA's DCF formula on the grounds that ORA uses data from the most recent six and twelve- month periods, and not just the most recent three- month period, which SCWC characterizes as the only "current" data. ORA asserts that its approach takes advantage of a longer time period to average out any short-term aberrations, though in this case the results for the three, six and twelve months were all within one one-hundredth of a percent.

Regarding the RP analysis, SCWC criticizes ORA's model on the ground that it uses "too much history," which is wrong because "history does not repeat itself." ORA argues that using just the 5-year period, which SCWC proposes, would ignore the longer-term trends in the model. ORA asserts that its RP model is superior to SCWC's proposal because it balances historical trends in the risk premium with forecast interest rates to arrive at return on equity.

Based on the record, we do not see either SCWC's or ORA's analyses as clearly superior, and we will consider the determinations of both ORA and SCWC. We adopt an equity return range of 9.08% to 10.70% for SCWC. We derive the floor rate by taking the simple average of the parties' DCF results and the ceiling by taking the simple average of the RP results.⁴⁴

In order to determine where SCWC should fall in that ROE range, we next assess the risk factors.

14.4. Return on Equity - Discussion of Risk

We see no indication of high financial, business or regulatory risk for SCWC. After considering that along with evidence on the financial models, adjustments to the models, interest rate trends, the current economy and our informed judgment, we have determined that the midpoint of our range (9.90%) reflects an appropriate equity return for SCWC. We therefore authorize this return for SCWC's Region III for the years 2003, 2004, and 2005. This return along with our adopted capital structure and costs of debt, equates to a rate of return on rate base of 8.79% for test year 2003, 8.77% for test year 2004 and 8.79% for attrition year 2005.

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. Both SCWC and ORA utilize a capital structure consisting of 50% debt and 50% equity, and neither party asserts increased or decreased financial risks associated with that structure.

Business risks pertain to uncertainties resulting from competition and the economy. A utility that has the most variability in operating results has the most business risk. Regulatory risk pertains to new risks that investors face from future regulatory actions that we, and other regulatory agencies, might take. Assessments of these risks are conducted to determine whether there is a need to increase return to compensate investors for added risk.

In its application, SCWC identifies and quantifies two elements of risk. First, it asserts business risk associated with the capital intensive nature of water utilities and operating leverage concerns caused by California water utilities having higher than average quantities of purchased water (fixed costs). However, the Commission has provided various regulatory mechanisms that deal with this risk - balancing accounts for purchased water, purchased power and pump taxes; memorandum accounts for catastrophic events; memorandum accounts for SDWA compliance; 50% fixed cost recovery and construction work in progress in rate base. Also, in this decision, we have approved memorandum account treatment for the Calipatria treatment plant. All of these factors tend to reduce risk associated with the capital intensive and high fixed cost nature of SCWC's operations. We see no reason to add a premium to the ROE to compensate for these risks.

The second risk specifically identified by SCWC concerns potential adjustments to balancing account procedures raised in Resolution No. W-4294. At the time of the application filing, related issues were being considered in R.01-12-009 where, on June 19, 2003, the Commission issued a final decision. In summary, D.03-06-072 revised the existing procedures for recovery of under collections and over collections in balancing-type memorandum accounts (accounts) existing on or after November 29, 2001 as follows: (1) If a utility is within its rate case cycle and is not over earning, the utility shall recover its account subject to reasonableness review; and (2) If a utility is either within or outside of its rate case cycle and is over earning, the utility's recovery of expenses from the accounts will be reduced by the amount of the over earning, again subject to reasonableness review. The utility shall remove the amount of the over earning from the account and shall amortize it below the line. Utilities shall use the recorded rate of return means test to evaluate earnings for all years.

Conclusion of Law 7, states in relevant part:

"[T]he readjustment of a utility's specific rate of return is not within the scope of this industry-wide proceeding. The appropriate forum for such a utility-specific inquiry is a utility's general rate case or other appropriate proceeding the Commission may designate in the future."

The affect of D.03-06-072 is to limit recovery of costs subject to balancing accounts when the utility is over earning its authorized rate of return. The issue is whether this imposes additional risk on the utility to the extent that a premium should be added to the equity return. In the policy discussion of D.03-06-072, we state:

"Like the Edison case, we believe that a revision to our existing procedures is necessary here in order to effectively correct distorted results. The existing procedures for recovery of under and over collections in balancing accounts, which we suspended as of November 29, 2001, were originally established for the utilities to recover unanticipated increases in electricity costs between general rate cases, without the need to file an additional rate case application. The procedures also served the purpose of protecting shareholders from having to finance large unanticipated expenses until the next general rate case.

"These procedures served, in effect, as insurance to protect a utility against its failure to earn its authorized earnings due to unanticipated expenses beyond the utility's control. When a person obtains insurance, the insurance is paid or invoked when the event insured against occurs. Similarly, offset balancing account recovery should only occur when the utility fails to earn up to its authorized rate of return due to unanticipated expenses beyond its control and that are the subject of the balancing account. To the extent a utility is earning above its authorized rate of return, recovery of the balancing account should be reduced by the amount of over earning since the event insured against (i.e., the failure to earn its authorized earnings) has not occurred.

"Thus, the existing procedures become problematic when they have the effect of enhancing utilities' earnings above the Commission-authorized rates of return. It is unreasonable and unnecessary to permit the utilities to pass through to ratepayers the dollar-for-dollar costs accumulated in their balancing accounts when these same utilities are earning more than their authorized rate of return, particularly when their ratepayers are also experiencing the same increased electrical costs in their own homes. To permit such recovery would be to grant the utilities an unanticipated windfall at ratepayer expense." (D.03-062-072, mimeo. at pp. 15-16.)

As described above, D.03-06-072 corrects an imbalance in the risks associated with previous balancing account procedures. The opportunity for unanticipated windfalls caused by balancing account protection should not have existed from the beginning. It would be illogical for us to increase the ROE by 90 basis points, as requested by SCWC, to compensate for the loss of what has been determined to be an illegitimate opportunity.

Additional information in the record indicates that SCWC is financially healthy. In evaluating SCWC's risk, ORA considered the Standard & Poor's (S&P) credit rating of SCWC, since this rating factors in a company's total risk. S&P rates SCWC A+/Stable, Business Profile 3. ORA Table 3-1, contained in Exhibit 9, shows S&P benchmark financial ratios as compared to SCWC for the years 1997-2001. According to ORA, based on those ratios, SCWC's overall rating would be "A," a strong indication that it is a financially healthy company. Placing SCWC in the middle of our ROE range appears fair and puts the authorized return in line with that recently authorized for other large water utilities under our jurisdiction.⁴⁵