6. Cost of Capital

A major issue in this proceeding is the cost of capital. Cost of capital is expressed as an overall rate of return, the result of which is the rate of return on rate base used to determine the revenue requirement in the summary of earnings. Rate of return is based on an adopted capital structure, long-term debt costs and a common equity return. The following tabulation summarizes the adopted capital structure and costs for the test and escalation years, capital structure, long-term debt costs, and common equity return.

We discuss the development of these figures in the following sections.

6.1. Capital Structure

Capital structure of SGV consists of long-term debt and common equity.4

6.1.1. SGV

SGV projected a test year capital structure of 26.10% long-term debt and 73.90% common equity. For escalation year 2009, it projected a 26.81% long-term debt and 73.19% common equity ratio, and for escalation year 2010, a 25.63% long-term debt and 74.37% common equity ratio. The overall weighted average capital structure for the test and escalation years was 26.18% long-term debt and 73.82% common equity.

6.1.2. DRA

Although DRA did not dispute SGV's projected weighted average capital ratio, it recommended adoption of an imputed capital structure consisting of 40.00% long-term debt and 60.00% common equity for the test and escalation years. That capital structure is approximately midway between the average equity ratio of a group of comparable water utilities and SGV's actual equity ratio. It is also the same capital structure adopted in SGV's prior LA Division GRC (D.05-07-044) and prior two Fontana Division GRCs (D.07-04-046 and D.04-07-034).

6.1.3. Discussion

SGV proposed to replace its currently authorized imputed capital structure of 40.0% long-term debt and 60.0% common equity with its projected actual capital structure.5 However, if the Commission found that SGV's adopted equity ratio should be less than its projected actual 73.82% common equity ratio, SGV would agree to an imputed 69% common equity ratio.6 Given that the burden of proof in GRC applications, such as this proceeding, rests upon the applicant to demonstrate the reasonableness of its request, we first address the method SGV used to arrive at its projected capital structure.

SGV used different time periods to forecast the long-term debt and common equity components of its capital structure.7 Use of those different time periods resulted in an understatement of its projected capital structure. That is because SGV included only half of its new $10.0 million debt issuance in its test year capital structure consisting of $49.0 million long-term debt (simple average of December 31, 2007 and December 31, 2008 data) and $138.7 million common equity (half of year end 2008 and 2009 data).8 Its method of forecasting test year long-term debt was also inconsistent with the method it used to forecast its test year rate base.9

If SGV had derived a capital structure consistent with its June 30th fiscal year-end test year and its June 30th fiscal year-end forecast of common equity and rate base, its test year long-term debt would have been increased by $5.0 million to $54.0 million from $49.0 million. That change to long-term debt would result in a test year capital structure of 28.02% long-term debt and 71.98% common equity.10 Consistent with that change, its weighted average test and escalation years' capital structure would then consists of 26.82% long-term debt and 73.18% common equity.11 Therefore, SGV's projected capital structures do not reasonably reflect its test and escalation test year's capital structures.

We next address the merits of adopting SGV's projected capital structure over an imputed capital structure. SGV's arguments for adopting its projected capital structure are similar to the arguments it testified to in 2003 and resulted in the adoption of its currently authorized imputed capital structure.12 Those arguments are that SGV, as a relatively small, privately held water utility facing substantial capital needs, requires a strong equity position to be able to sell bonds on reasonable terms to finance expected and unexpected investments in a timely manner. An imputed capital structure should only be adopted when there is compelling evidence that the company's actual capital structure is inefficient, not least cost.

In regards to that latter argument, SGV cited a recent Valencia Water Company (Valencia) GRC decision which found that in the absence of evidence of a utility manipulating its capital structure to achieve an artificially high rate of return, basing a rate of return calculation on its actual capital structure is consistent with a goal of ratemaking to approximate the economic returns that a regulated company would achieve in a competitive environment.13

Although there was no comparison of the Valencia decision to SGV in this proceeding, we did review the Valencia decision. We found that there were more dissimilarities than similarities between the two water companies. While Valencia's long-term debt ratio was higher, 31.05% versus 25.63%, Valencia was substantially smaller in size. A rate base comparison of Valencia to only the LA Division showed that Valencia was 30% of the size of that division in rate base, and a customer comparison of Valencia to both of SGV's operating division showed that Valencia was also only 30% of size of SGV.14

The Valencia decision was applicable only to Valencia and based on the facts presented in that proceeding. It should not be considered precedent-setting for all regulated water companies. Absent a meaningful comparison between SGV and Valencia, which was not provided, we have no basis to apply the results of the Valencia decision to SGV.

In regards to SGV being a relatively small, privately held water utility, the company relied on the same 1975 academic study used in the Valencia proceeding to support a higher equity ratio. The study which concluded that among unregulated firms, lower equity ratios are generally associated with larger companies. By inference, SGV would need higher equity ratios to offset higher financial and business risks.15 However, the results of that study are not applicable in this proceeding because it studied only unregulated entities and was completed more than 30 years ago. Even if that study could be deemed pertinent to regulated entities, its conclusion does not hold when the substantially smaller Valencia with a 69% equity ratio is compared to a larger SGV with a projected 74%.

SGV further cited two publications from financial experts in support for adoption of its actual capital structure. Both of these publications recommended that water utilities that strengthen the shareholders' equity stake by increasing their equity ratios. These publications were not based on current information as they were both issued well before our 2004 adoption of SGV's imputed capital structure. One was published in the summer of 1994 and the other June 2000.16 Further, strengthening shareholders' equity does not necessarily result in strengthening of ratepayers interests. Absent a balance between the interests of shareholders and ratepayers, a high ratio of equity may be both inefficient and burdensome to ratepayers through excess rates.

Finally, SGV contends that it has more financial risks than publicly traded companies because it is a private company whose borrowings through private placements with insurance companies are not rated by rating agencies. As a result, institutional lenders consider its long-term debt financing to be more speculative. Again, its testimony was similar to the proceeding in which we first adopted an imputed capital structure of 60.0% equity.17

Common equity financing is more costly than debt financing because of its higher risk. Because debt financing is less costly and is tax-deductible, ratepayers benefit from the use of debt financing, or leverage. As debt ratios are increased and equity ratios are correspondingly lowered, financial risk for equity investors increases, requiring greater returns on equity. Additional leverage is advantageous to ratepayers up to the point that overall capital costs begin to increase as a result of increased cost of equity caused by greater financial risk and increased cost of debt. Therefore, the sole reason for adopting a utility's capital structure should not be based on an absence of evidence that a utility is manipulating its capital structure.

As previously stated, the burden of proof in a GRC rests with the utility. The utility has the duty to substantiate that its capital structure balances the interest of both shareholders and ratepayers. SGV's reliance on the Valencia decision, dated studies, and a proposal to accept an imputed capital structure equal to Valencia's does not satisfy that duty.

As proposed by SGV, we could reflect impacts of financial and business risks through the adoption of both SGV's projected actual capital structure and a 90 basis point risk premium on its return on equity.18 However, the impact of those risks has traditionally been reflected in the authorized common equity return and we should continue to do so. Absent evidence that SGV's projected capital structure is an appropriate balance between shareholder and ratepayer interest, we again adopt an imputed capital structure. We recognize that SGV, as a privately held company, does not have access to the financial markets that public companies do. As such, we adopt a 37.0% long-term debt and 63.0% common equity capital structure, which approximates midway between the average equity ratio of the comparable water utilities used by the parties in this proceeding and SGV's actual equity ratio.19

6.2. Long-Term Debt Costs

SGV projected its test year long-term debt costs at 7.76% based on a simple average of actual 7.70% year end 2007 and forecasted 7.81% year end 2008 long-term debt costs.20 That forecasted long-term debt included a $10 million issuance of new long-term debt in 2008 at a 7.72% interest rate. The 7.72% interest rate was based on an average of long-term Treasury bond rates plus a spread of 246 basis points.21 22 The overall weighted cost of debt for test and escalation years was 7.79%. DRA accepted SGV's 7.79% overall cost of debt.23

In this instance, an agreement on an appropriate interest rate by the parties does not necessarily constitute reasonableness of SGV's forecasted long-term debt costs. Here, SGV used a calendar year end date to forecast its long-term debt costs while it used its test year fiscal year to forecast rate base. This difference is compounded by SGV's use of a simple average of different time periods and sources to forecast new long-term debt costs. The time periods and sources consisted of a December 2006 Blue Chip consensus, February 23, 2007 Value Line Quarterly, and March 2007 Global Insight and U.S. Economic Outlook.24 That simple average of different time periods and sources further compounds a difference between the time periods used for forecasting long-term debt and use of October 2007 escalation rates adopted for the test year's results of operations accounts.

This inconsistent use of time periods for estimating long-term debt costs does not substantiate the reasonableness of SGV's and DRA's agreed-upon weighted average test and escalation years' 7.79% long-term debt costs. Other than a statement that SGV had experienced a historical 246 basis point premium above long-term Treasure rates for its new issuances, there was neither evidence nor support for a continued use of that premium in this proceeding.25

Absent evidence to justify adoption of SGV's long-term debt costs, we must turn to informed judgment to arrive at a weighted average test and escalation years' long-term debt costs. SGV had year-end 2007 embedded long-term debt costs of 7.70% and that $10,000,000 of new debt would be issued in the test year. It is also general knowledge that long-term treasury rates have dropped since SGV's use of forecasted long-term treasury rates. For example, one of the long-term treasury forecast sources relied on by SGV was Global Insight.26 The March 2007 Global Insight rate of 5.27% used by SGV dropped 36 basis points to 4.91% in its September 2007 forecast.27 Based on these facts and application of informed judgment, we arrive at a 7.70% weighted average long-term debt costs.

6.3. Common Equity Return

The legal standard for setting a fair common equity return has been established by the United States Supreme Court in the Bluefield and Hope cases.28 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.29 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. The 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 must set the equity return at the lowest level that meets the test of reasonableness.30 At the same time, our adopted equity return should be sufficient to provide a margin of safety to pay interest, pay reasonable common dividends, and allow for some money to be kept in the business as retained earnings. To accomplish this objective, we have consistently evaluated analytical financial models as a starting point to arrive at a fair equity return.

6.3.1. Financial Models

The financial models commonly used in equity return proceedings are the Discounted Cash Flow (DCF) and Risk Premium Model (RPM). Detailed descriptions of these financial models are contained in the record and are not repeated here. The results of these financial models are used to establish a range from which parties apply risk factors and individual judgment to determine a fair equity return.

Although the parties agree that the models are objective, the results are dependent on subjective inputs. The parties used the same proxy group of six Class A water companies. However, they used different subjective inputs such as time periods, risk-free rates, market risk premiums, and growth rates. The application of these subjective inputs results in a difference of common equity returns being recommended by the parties, as shown by the results of their DCF and RPM models. From these subjective inputs, the parties advanced arguments in support of their respective analyses and in criticism of the input assumptions used by the other parties. SGV even went so far as to recalculate DRA's financial modeling based on selective changes.31 The following tabulation summarized the results of the individual financial models used by SGV and DRA, including the simple range of the DCF and RPM model results, overall average of the models and recommended equity return.

The subjective nature of the financial model results was shown by SGV's recalculation of DRA's financial modeling results based on selective changes. These selected changes resulted in a 40 basis point increase in DRA's financial models overall average of 10.3% to 10.7%, still 50 basis points lower than SGV's overall average. In the final analysis, it is the application of judgment, not the precision of these models, which is the key to selecting a specific equity return within the range predicted by analysis. We affirmed this view in D.89-10-031, which established equity returns for GTE California, Inc. and Pacific Bell, noting that we continue to view the financial models with considerable skepticism.

We find no reason to exclude or adopt the financial modeling results of any one party. Therefore, we establish an equity range based on the model results and informed judgment. After considering the evidence on interest rate forecasts, quantitative financial models based on subjective inputs, risk factors, recently authorized equity returns on California Class A water utilities, and applying our informed judgment, we conclude that a subjective equity return range deemed fair and reasonable for the test and escalation years is 10.5% to 11.0%.35

6.3.2. Additional Risk Factors

SGV added a 90 basis point premium to the results of its financial models on the basis that company-specific risks make SGV a riskier investment than the benchmark group of publicly traded water utilities used in the financial models. SGV did not specify how much of this additional premium resulted from each of its individual company-specific risks. These individual risks fell into three categories: financial, business, and regulatory risks.

Financial risk is tied to a utility's capital structure. The proportion of its debt to permanent capital determines the level of financial risk that a utility faces. SGV contends that it has less financing flexibility as a private company than publicly traded companies. However, as a utility's debt ratio increases, a higher equity return may be needed to compensate for that increased risk. SGV's low debt ratio (actual and imputed) equates to lower financial risk. A low debt ratio enables it to access the financing markets through private placements and mitigates its financial risks in comparison to publicly traded companies having higher debt ratios.

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 poor management, and greater fixed costs in relationship to sales volume.

Risks related to contaminated water supplies and a 1994 Commission finding that SGV faces a higher risk than other water utilities with respect to its source of water fall into this category. However, contaminated water supplies are not unique to SGV. As stated by SGV, risk results from uncertainty related to anticipated stringent requirements to treat contaminated water, such as new state arsenic requirements.36 However, these new requirements will be imposed on all California water utilities, including the two California water utilities included in the financial models' proxy group. To an extent, increased risks due to contaminated water supplies have already been factored in via the financial models. The Commission finding, approximately 14 years ago, that SGV faces higher risk than other water utilities was actually limited to its Fontana Division. SGV did not provide evidence in this proceeding to enable us to affirm that the finding is still valid or that it is applicable to the LA Division.

Regulatory risk pertains to new risks that investors may face from future regulatory actions that we and other regulatory agencies might take. SGV has rated its regulatory risk to be high on the basis that Value Line advises its investors that regulatory risk in California is higher than other western states and changes in the Rate Case Plan (RCP) related to the three-year GRC cycle and use of a standard sales forecast model have increased the risk that SGV will be able to earn a fair rate of return.

However, the below average regulatory rating from Value Line does not pertain to water companies, it pertains to energy companies.37 DRA identified a host of regulatory mechanisms that reduce regulatory risks. These mechanisms include the inclusion of construction work in progress in rate base, a rate design that enables SGV to recover 50% of its fixed costs irrespective of sales, balancing accounts, and memorandum accounts. Irrespective of SGV's perceived regulatory risks, such risks were already reflected in the financial models. Two of the six water companies, California Water Service and San Jose Water Corporation, both used in the financial models' proxy group, are California utilities experiencing the same regulatory environment as SGV.

These additional risks mitigated by regulatory mechanisms do not warrant a premium to the overall 10.50% to 11.00% equity range found reasonable in this proceeding. We adopt a 10.50% ROE, the lower end of the equity range found reasonable in this proceeding. This ROE is sufficient to provide a margin of safety to pay interest on long-term debt, pay reasonable dividends to the equity holders, and allow a reasonable amount of funds to be kept in the business as retained earnings.

While this authorized ROE is higher than the ROEs granted other Class A water utilities that litigated ROEs since 2000, it reflects DRA and SGV's recognition that SGV faces increased equity risk in comparison to other Class A water utilities. For example, as shown in Exhibit 25, DRA's 10.30% recommended ROE for SGV was 70 basis points higher than its 9.55% average ROE recommendation for other Class A water utilities that litigated an ROE. SGV's 12.10% recommended ROE was 68 basis points higher than the 11.42% average ROE recommended by other Class A water utilities that litigated an ROE.38

3 Return on rate base.

4 Debt due within one year, i.e., short-term debt, is excluded.

5 That imputed capital structure had not changed since first authorized for its Fontana Division on July 8, 2004, pursuant to D.04-07-034.

6 Exhibit 23, p. 6.

7 Exhibit 3, pp. 1-3 through 1-8.

8 The $10 million new issuance is due to expire on July 1, 2008, as shown in Exhibit 3, p. 1-7.

9 In order to determine the average fiscal year balances of rate base components, SGV added half of the two consecutive year end average balances, as detailed in Exhibit 8, p. 56.

10 As a result of that $5.0 million addition to long-term debt, its total capital structure of $187.7 million would also increase by $5 million to $192.7 million. Hence, its long-term debt ratio would be 28.02% ($54.0 million divided by $192.7 million total capital). That 28.02% long-term debt ratio would result in a 71.98% common equity ratio (100.00% minus 28.02%).

11 A simple average of SGV's test year capital structure adjusted to reflect the additional $5.0 million of long-term debt and both escalation years' forecasted capital structure.

12 D.04-07-034 (2004) mimeo., p. 52.

13 D.07-06-024 (2007) mimeo., p. 17.

14 Valencia's rate base was $38.4 million in comparison to SGV's LA Division $128.0 million. Valencia had approximately 28,000 customers in comparison to SGV's 92,000 customers, of which LA Division had 48,000 and Fontana Division 44,000. We did not compare Valencia's and the total SGV rate base because Fontana Division's rate base was not identified in this record.

15 D.07-04-024 (2007) mimeo., p. 16 and Exhibit 14, p. 48.

16 Exhibit 3, pp. 2-17 and 2-18.

17 Id., pp. 2-14 through 2-18 and Application (A.) 02-11-044, Exhibit 3, pp. 2-12 through 2-17.

18 Exhibit 14, p. 47.

19 Exhibit 27, p. 2-8.

20 Exhibit 3, Table B.

21 Although SGV testified that the $10 million new long-term debt issuance was issued in 2007, p. 1-1 of Exhibit 3, Table B to that exhibit shows that its debt was actually projected to be issued in 2008 and is reflected in its 2008 average long-term debt costs.

22 One basis point equals 0.01%.

23 Exhibit 27, p. 1-1.

24 Exhibit 3, p. 1-1.

25 Id.

26 Exhibit 14, Table 11.

27 Official notice is taken of Global Insight's September 2007 forecast. See, e.g., Exhibit 65 of A.07-05-003 et al.

28 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).

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

30 46 CPUC2d 319 at 369 (1992), 78 CPUC at 723 (1975).

31 Exhibit 23, pp. 6 - 10.

32 Exhibit 14, Table 19.

33 SGV added a 90 basis point premium to its average to reflect additional risks not reflected in the financial model results.

34 Exhibit 27, Table 2-8.

35 This range of equity return is 20 basis points higher than DRA's financial models' overall average and 20 basis points lower than SGV's.

36 Exhibit 8, p. 18.

37 Reporter's Transcript vol. 2, p. 27.

38 SGV's 12.10% recommended ROE was based on the results of its financial models plus its premium for additional risks.

Previous PageTop Of PageNext PageGo To First Page