V. Discussion
A. Water Quality Improvement Program
Valencia's water supply from groundwater wells contains concentrations of calcium and magnesium ranging between 300 and 700 parts per million (ppm).5 According to the Applicant, water is considered very hard if the combined concentrations of calcium and magnesium exceed 300 ppm. In response to numerous customer complaints about this problem, Valencia proposes to construct a demonstration project to evaluate a remedy for the hardness of its groundwater supplies.
Valencia's customers receive a blend of imported water from the State Water Project (SWP), purchased water from Castaic Lake Water Agency (CLWA) and local groundwater pumped from company-owned wells. Valencia blends the hard water from its groundwater wells with softer water from the SWP and the CLWA in order to deliver a more uniform and consistent water quality for its customers. This blending reduces the hardness of the water delivered to Valencia's customers but does not result in overall hardness levels generally acceptable to consumers.
Testimony at the evidentiary hearing established that all of Valencia's customers currently pay a high price for hard water. Some have absorbed the price in the form of clogged plumbing and damaged appliances, while others have installed self-regenerative water softeners in their homes and businesses. These devices extract the dissolved minerals in the water in the form of calcium and magnesium chloride. Valencia's Chief Executive Officer testified that 11,000 to 12,000 of Valencia's residential customers currently spend, on average, $50 per month or more for water softening in addition to paying their water bills.6 Other evidence including correspondence from Valencia customers supporting the water softening project suggests that the average Valencia customer incurs between $300 and $900 a year in direct and indirect costs as a result of the extreme hardness of the water.7
Although the in-home water softeners substantially reduce hardness levels in water used for drinking and washing, the water softening devices periodically discharge the captured calcium and magnesium chlorides into the sewer system. This additional chloride loading to local wastewater treatment plants causes high chloride discharges into the Santa Clara River. The Los Angeles Regional Water Quality Control Board currently is reviewing the total maximum daily load for chloride in the Santa Clara River and is in the process of setting new discharge limits for chloride that may require the Santa Clarita Valley Sanitation District of Los Angeles County (Sanitation District), owner of the existing wastewater treatment plants, to install very expensive treatment equipment8 and a brine line to remove the chlorides before effluent is discharged to the river. One possible alternative being considered by the District to reduce chloride loading to the Santa Clara River is Valencia's wellhead water softening project.9
Kennedy/Jenks Consultants conducted a Wellhead Softening Feasibility Study (Feasibility Study) for the Applicant in April 2006. The Feasibility Study describes the hardness problems characterizing Valencia's water supplies, evaluates several candidate softening technologies and recommends, based on cost and non-cost factors, the choice of pellet softening as the treatment process most appropriate for Valencia.10 The Feasibility Study also recommends, in order to confirm consumer acceptance of centralized softening and to refine project costs, that Valencia construct a demonstration plant. The Feasibility Study estimates the capital cost for the demonstration plant at $1.7 million, with an annual operation and maintenance cost of $170,000.11
The Feasibility Study includes an analysis of the projected economic benefits of a system-wide water softening project. This analysis concludes that the benefits to Valencia's customers from softening all the groundwater could be as high as $23.8 million per year - consisting of $2.6 million in savings from reduced water softener use and $21.2 million in avoided wastewater treatment costs.12 For reasons discussed below, we believe this estimate must be revised.
Valencia's request for authorization to build the demonstration water-softening project was supported by state, county and local government officials and customers of the water company. Most significant, no Valencia customer opposed the request. However, DRA did oppose the request and it is to its opposition that we now turn.
DRA opposes having ratepayers finance water quality improvement projects that do not result in direct health benefits and asserts that Valencia's water softening project fits that category. DRA sees a lack of customer support and argues that ratepayers should not be asked to pay for non-essential aesthetic enhancement projects.13 Interpreting survey results as indicating that most customers do not own water softeners and do not favor paying more for softer water, DRA sees no compelling argument for allowing the project.14
DRA denies that other water purveyors in the Santa Clarita area consider water softening a priority or that Valencia's project will provide environmental benefits. DRA also doubts Valencia's claim that the project could save $21 million in avoided wastewater treatment costs. DRA argues that Valencia ratepayers might end up having to pay for both Valencia's water softening project and an advanced wastewater treatment plant.15
DRA sees the environmental problem of self-regenerating water softeners as a regional problem, requiring a regional solution. DRA does not want Valencia's ratepayers to bear the cost of Valencia investing in an uncertain solution to the problem and would prefer to see other alternatives pursued.16
Based on Valencia's data showing that the demonstration softening project will require a 2.2% increase in customer bills, DRA calculates that a 45% increase in customer bills would be needed to fund full deployment of a water softening program. DRA concludes that the cost of full deployment outweighs the benefits. Having decided that the overall project is unreasonable, DRA considers the demonstration project unreasonable as well. DRA fears that allowing the demonstration project to go forward would increase pressure to approve the entire project, to which DRA is opposed.17
Hard water is clearly more than an aesthetic problem. The record establishes that the scale from hard water clogs pipes, hot water heaters, washing machines and dishwashers, necessitating frequent repairs and replacements. The damage done by hard water is a principle reason that nearly half of Valencia's customers incur the substantial extra monthly costs associated with in-home water softening devices. If softening the water at the source is cheaper than softening water in the home, all Valencia customers will benefit from switching to pre-softened water.
If Valencia pre-softens the water it delivers to its customers and those customers disconnect existing home water softeners (or don't buy additional home water softeners), Valencia's customers will be conferring a benefit on people and businesses downstream. The Santa Clara River will have fewer chlorides discharged into it than it would if Valencia didn't build the water-softening plant. Softer water in the river will reduce the need for brine removal downstream, to the benefit of downstream residents and businesses. But is this unintended benefit to third parties sufficient reason to deny the application for a permit to build the demonstration plant? We think not.
Rather than seeking immediate approval to build a full-scale plant, Applicant prudently proposes to test the recommended water softening technology on a pilot basis before deciding whether to propose building a full-scale plant. The demonstration project will establish how effective pellet softening is in removing dissolved minerals from the groundwater, how costly it is and whether a less costly alternative would produce satisfactory water quality. Valencia's project will also provide data to other upstream water companies and to the Sanitation District that can be used to develop a region-wide response to the water hardness problem.
To analyze the costs and benefits of wellhead water softening to Valencia ratepayers, it is necessary to make certain assumptions. The first of these concerns the proportion of the cost of in-home water softening that can be attributed to customer dissatisfaction with the aesthetic qualities of hard water. Among the 11 to 12 thousand customers estimated to be operating home water softeners, we assume that some are doing so primarily to eliminate unsightly deposits on glassware and automobiles, to get cleaner clothing from their washing machines or for other reasons. Others, we assume, are softening the water primarily to avoid clogged pipes and damaged appliances. Unfortunately, the company's customer survey did not attempt to distinguish between these reasons for installing home water softening equipment. Since those customers who have installed home water softeners are paying $50 a month or more to operate them, $50 a month appears to be a reasonable estimate of the total cost of hard water to the average Valencia customer. In the absence of record evidence regarding the allocation of that cost between aesthetic and economic problems, we will assume that half the cost ($25 a month per customer) is a reasonable estimate of the unavoidable economic burden imposed by hard water on the average Valencia customer. This equates to an annual cost of approximately $8.5 million18 to Valencia ratepayers which could be avoided by wellhead water softening.
To calculate the avoided costs of reduced wastewater treatment requires another assumption. The home water softeners operated by Valencia customers contribute about 45% of the chlorides flowing into the wastewater stream from the upper Santa Clarita Valley.19 The remaining 55% of the discharged chlorides come from water softeners operated by customers of other water companies. There is an inverse relationship between the amount of chloride discharge from upstream water company customers and the cost of building a new downstream wastewater treatment plant. If all upstream chloride discharges were eliminated, a new plant would be unnecessary. If none of the upstream discharges were eliminated, a new plant costing nearly half a billion dollars would be necessary.
If some, but not all, upstream discharges were eliminated, witnesses for the Applicant and DRA agree that the size of the downstream plant could be reduced. Unfortunately, the record is silent on the exact relationship between partial chloride discharge abatement and size reduction of the downstream plant. Recognizing that there is a minimum size for any such plant, we assume that if Valencia alone among the upstream water companies adopted wellhead water softening, thus eliminating 45% of the chloride discharge, the size of any future wastewater treatment plan would be reduced by 22.5%.
With these assumptions in place, it is possible to estimate the costs and benefits to Valencia ratepayers under different scenarios. The scenario of concern to us is one in which Valencia alone adopts wellhead water softening. On the assumptions we have made, would such an action be beneficial to Valencia ratepayers? The Kennedy-Jenks study estimated that eliminating the need for a new wastewater treatment plant would save Valencia ratepayers $21.2 million a year in avoided wastewater treatment costs. Reducing the size of the plant by 22.5% rather than eliminating it entirely would therefore save Valencia ratepayers $4.77 million a year (22.5% of $21.2 million). To these avoided costs, we add the previously calculated economic benefit from eliminating hard water at the wellhead of $8.5 million, for a total estimated annual benefit to Valencia ratepayers of $13.27 million from wellhead water softening, even if no other water company follows Valencia's lead.
On the cost side of the equation, the full-scale plant is estimated to cost $32.328 million to build and $6.6 million annually to operate. To the $6.6 million annual operating cost, we need to add approximately $3.3 million a year to amortize plant construction costs, for a total annual cost of wellhead water softening of $9.9 million. Thus, even if Valencia alone chooses to adopt wellhead water softening, the projected annual savings to Valencia ratepayers are approximately $3.4 million ($13.3 million - $9.9 million).
Not surprisingly, given the favorable economics of the proposed project and its beneficial environmental effects, it has received wide support from the public bodies responsible for water supply and water quality in the Santa Clarita area. These expressions of support include a letter from State Senator George Runner and resolutions of support adopted by the City of Santa Clarita and the Santa Clarita Valley Sanitation District. At the Public Participation Hearing, various speakers supported the project and no one opposed it. We also note that the matter of chloride pollution of the Santa Clara River is the subject of SB 475, authored by Senator Runner and signed into law last year by the Governor, which provides financial incentives for homeowners to remove their water softeners.
For the above reasons, we find it to be in the public interest to approve the construction of the demonstration plant.
B. Valencia's Capital Structure
Valencia's capital structure is set forth in Section 2.5 of its Application. Valencia's last adopted capital structure, for Test Year 2004, was agreed upon as part of a settlement between Valencia and DRA and consisted of 34.71% debt, 3.79% preferred stock, and 61.5% common equity. Valencia asks that we base its rates on its actual capital structure, which, for Test Year 2007-2008, will consist of 27.95% debt, 3.05% preferred stock, and 69% common equity.20
DRA witness Mehboob Aslam testified that DRA recommends a capital structure consisting of 46.40% long-term debt, 1.00% preferred equity, and 52.60% common equity.21 Since the cost of long-term debt is substantially lower than the cost of common equity, the DRA recommendation would significantly reduce the overall rate of return on rate base and the overall return allowance for Valencia, as compared with a calculation of return requirements based on Valencia's actual capital structure.
Noting Valencia's citation22 of a study, by Scott and Martin, that found statistically significant evidence that smaller firms seeking low-cost capital structures will have higher equity ratios than larger firms in the same industry, DRA witness Aslam denied that Valencia was a small company "in the true meaning of the term," because its parent company, Lennar Corporation, is publicly traded and reported annual revenue over $13 billion in the year 2005, and because for investors, "it is the size of Lennar not that of Valencia which matters."23 He also denied the relevance of the cited study, because it addressed "non-regulated" firms.24
According to Aslam, "DRA has two choices" - either to adjust Valencia's capital structure to one "more in tune with the other Class A water utilities in California" or to reduce Valencia's rate of return to "accommodate" a low financial risk due to increased reliance on equity. He chose the former approach as "more practical" and "less controversial." Based on data for four major Class A water companies in California, he calculated an average common equity ratio of 52.6% and imputed that ratio to Valencia.25
In choosing between Valencia's actual capital structure and DRA's recommended capital structure, we note that Valencia is a subsidiary of a publicly traded company, Lennar Corporation. Thus, its actual capital structure is to some degree arbitrary and could be changed by the parent at any time, for instance, by injecting additional equity into the Applicant or using parent funds to retire the subsidiary's indebtedness, although there are sound economic reasons to doubt that the parent would actually take these steps.26 On the other hand, there is much to be said for basing a rate of return calculation on an applicant's actual capital structure rather than imputing an artificial capital structure in order to back into a rate of return. As Applicant's witness Zepp pointed out, a goal of ratemaking is to approximate the economic returns that a regulated company would achieve in a competitive environment.27 In the absence of evidence that the Applicant is manipulating its capital structure in order to achieve an artificially high rate of return, basing a rate of return calculation on its actual capital structure is consistent with that goal.
In response to DRA's contention that Valencia should not be considered a small company because its parent Lennar is a large public company, Applicant's witness Milleman testified that "Valencia is not Lennar." He noted that "if Valencia truly were viewed by the investment world as a real estate company, then DRA should be recommending rates of return applicable to real estate companies."28 Instead, DRA compared Valencia to large publicly traded water utilities that are in a far less risky business than speculative real estate development. If Lennar sometimes earns a high rate of return, that is because "the speculative real estate industry demands a higher rate of return then the regulated water industry."29 DRA invokes Valencia's large parent company to deny recognition of Valencia's small company risks, but ignores the greater market risks facing the parent company.30
Milleman also testified that comparing Valencia's capital structure with the capital structure of major water companies such as California Water Service Company or Golden State Water Company is inappropriate. Because these companies are large and geographically dispersed, "they are able to spread their risks. They are publicly traded and have ready access to capital. They have finance departments, legal departments, regulatory staffs, general office support, and other large-company resources. Valencia has two associates to cover these issues."31 Milleman suggested that a more realistic approach would be to compare Valencia (28,300 connections) with Great Oaks Water Company (20,000 connections) or Class B water utilities (2,001 to 10,000 connections). We agree that a comparison with Great Oaks is appropriate but not with Class B utilities.
Our charge as regulators is to see to it that Valencia's customers continue to receive an adequate supply of good-quality water at a reasonable price. If, in an effort to keep water prices to consumers low, we provide Valencia with inadequate income by manipulating its capital structure or by any other means, we jeopardize the future adequacy and quality of the water supply. Accordingly, we choose to adopt Valencia's actual capital structure rather than DRA's imputed capital structure and we deem Valencia to have a capital structure for this rate-making period consisting of 27.95% debt, 3.05% preferred stock,32 and 69% common equity.33
C. Return on Common Equity (ROE)
Having established the proportion of the various components of the rate base, we turn now to the question of the rate of return that Valencia should be permitted to earn on the common equity portion of the rate base. DRA's expert Aslam submitted two studies of rates of return earned by a group of six "comparable" water companies that implied rates of return on Valencia's common equity of 8.40% and 10.53% respectively. Aslam took the average of those studies and recommended an ROE of 9.46%.34 Applicant's witness Zepp criticized the methodology used by Aslam and submitted a study proposing an ROE of 11.75%.35 In support of this recommendation, witness Zepp testified that small water companies like Valencia typically are capitalized with higher equity ratios than larger water companies and pay higher costs for their equity than do larger water companies.36 He explained that the higher costs of equity result from the higher risks typically incurred by small water companies. He identified two areas of "additional risk" facing Valencia: the additional regulatory risk of operating in California and litigation and other risks specific to Valencia.
Noting that DRA has recognized in other cases that "the business risk of a regulated utility consists primarily of regulatory risk," Zepp pointed out that "at least two institutions that evaluate regulatory risk indicate such risk is higher in California than in other states."37 He discussed a number of reasons why Valencia and other California water utilities face above-average regulatory risk, including a relatively short three-year GRC cycle, constraints on forecasting future sales, and the potential for water quality lawsuits. He also noted that three of the six utilities in Aslam's "comparable group" do not operate in California and therefore do not face comparable regulatory risk.38
Zepp testified that Valencia faced additional company-specific risks due to its small size, the possibility of catastrophic events affecting its service area, and the active and continuing intervention of "no growth" advocates in regulatory and court proceedings affecting Valencia.39 The constant contention over land development and water supply in the Santa Clarita Valley creates the risks that growth in future sales may be substantially below forecasts, that Valencia may incur unexpected legal costs and that litigation over perchlorate contamination may drain Valencia's resources.40 Each increment of risk faced by Valencia that is not faced by larger water companies or companies located outside of California increases Valencia's cost of capital relative to such other companies. Zepp testified that these added risks increase the equity return required by Valencia by at least 90 basis points (0.90%) above the average ROE required by companies that do not face these risks.
Zepp's analysis of the cost of equity capital for Valencia, was received into evidence as Exhibit 4. It includes a study of the current cost of equity capital for a group of benchmark water utilities (the "comparable group"), using both Discounted Cash Flow (DCF) analysis and several variations of Equity Risk Premium analysis.
Applying a DCF analysis, which computes the cost of equity as the sum of expected dividend yield and expected dividend growth, Zepp calculated the cost of equity for the comparable group as falling in a range of 10.7% to 10.9%.41 Removing one of the six companies, Connecticut Water Service, from the six-company benchmark group produced a slight increase in the benchmark equity range to one of 10.8% to 11.0%.42
Zepp then applied the Equity Risk Premium concept to calculate the cost of equity for the comparable group. The Equity Risk Premium is the difference between the utility's cost of equity and the interest rates paid on Treasury securities, which are essentially risk-free. He developed four different risk premium estimates, based on similar but distinct calculations. The first was an update and restatement of the risk premium analysis DRA presented in a November 2004 California-American Water Company (CalAm) GRC, producing an indicated cost of equity for the comparable group in a range of 10.6% to 10.9% (implying a cost of equity for Valencia of 11.5% to 11.8% when the 0.9% risk premium is added to the indicated ROE range for CalAm).43 Zepp then substituted authorized ROEs for the earned ROEs as proxies for the cost of equity, which produced a range for the comparable group of 10.8% to 11.4% (implying a cost of equity for Valencia of 11.7% to 12.3%).44
On the assumption that costs of equity for water utilities move in the same direction as interest rates but by smaller amounts, Zepp performed a third risk premium analysis that produced a 10.7% cost of equity for the benchmark water utilities and implied an 11.6% cost of equity for Valencia.45 A fourth risk premium analysis based on dividend yields and growth rates derived from published reports of the comparable group for the years 1999 through 2005 produced an equity cost range of 10.7% to 11.1% for the comparable group and implied an equity cost range of 11.6% to 12.0% for Valencia.46 The combination of these four Equity Risk Premium estimates indicated a benchmark equity cost range of 10.5% to 11.4% and implied an equity cost range of 11.4% to 12.3% for Valencia.47
Based on these analyses, Zepp recommended that the Commission authorize an ROE of 11.75% for Valencia.48
DRA's analysis of Valencia's cost of equity was provided by Aslam in Exhibit DRA-12. Like Zepp, he applied two financial models, DCF and Equity Risk Premium, to a "comparable group" of six49 publicly-traded water utilities, to estimate investors' expected ROE for Valencia.
The two experts differed slightly in their calculation of dividend yields for the companies in the comparable group. Zepp calculated average yields of 2.87% to 3.06% while Aslam calculated average yields of 2.78% to 2.84% for the same companies. For the other main element of DCF analysis, the earnings growth rate, the differences were far greater. Zepp relied solely on forecasts of earnings growth while Aslam combined historical growth rates with forecasted growth rates. The result of these different approaches is that while Zepp used an earnings growth rate at 7.59%, the average of two forecasts of earnings growth, Aslam used a growth rate of only 5.44%, the average of a 2.74% average historical growth rate and an 8.14% average forecast growth rate.50 The 2.74% average historical growth rate used by Aslam was an average of six calculated historical averages, ranging from a five-year earnings growth rate of 0.24% to a five-year earnings growth rate of 4.52%.51 While Valencia questions the accuracy of Aslam's calculation of historical growth rates, the company primarily argues that the combining of historical growth rates with forecasted growth rates over-weights past data because analysts' forecasts have already taken those data into account.52
Aslam's Equity Risk Premium analysis compared historical ROEs for the comparable group with yields on 10-year and 30-year Treasury bonds, with the differences between those annual averages assumed to be the annual risk premiums. He then calculated ten-year and five-year averages of those risk premiums and added those averages to forecasted future interest rates, thereby showing four projected ROEs ranging from 10.27% to 10.75%, from which he derived an average risk premium ROE calculation of 10.53%.53
Aslam contested several aspects of Zepp's cost of capital analysis. In the area of company-specific risk, as discussed in the context of the capital structure issue, he claimed that Valencia should not be considered a small company, because its parent company is large.54 He denied that regulatory risk is higher in California than in other states, noting favorable investor responses to changes at the Commission, and he downplayed the risks presented by "no-growth" opposition, water quality litigation, uncertainty about future growth, or other litigation.55
Aslam challenged Zepp's inclusion in his DCF analysis of "sv-growth" - sustainable growth that is expected from external sources, which can be expected when market price exceeds book value.56 Aslam asserted that if the market to book ratio exceeds one, the authorized return must be greater than investors require, and so he opposed any recognition of "sv growth."57
Aslam objected to any method of risk premium calculation other than the comparison of earned ROE to low-risk Treasury bonds, which both he and Zepp employed. He opposed Zepp's use of authorized ROE - as distinguished from earned ROE - in a variant of his risk premium model. He also objected to Zepp's consideration that costs of equity for water utilities may move in the same direction as interest rates but by less, as well as his use of a composite DCF/risk premium model.58 We address DRA's objections in detail in Section VII, below.
DRA's case for an ROE below that recommended by Valencia's expert was unfortunately marred by a series of factual and methodological errors that rendered its conclusions of dubious value. The errors that came to light during the evidentiary hearings included:
· Aslam failed to note that a number of the utilities in the comparable group of publicly-traded water utilities had stock splits during the last ten years. Failure to correct for a stock split causes earnings in years subsequent to the split to appear lower than they are. This failure to note stock splits caused the dividend yield estimates in Aslam's original Table 2-2, the historic growth estimates in his original Table 2-3, and his DCF equity cost estimates to be understated. The failure to correct for stock splits is particularly egregious in light of the fact that DRA had already made those corrections in prior water company rate cases including the recent Park Water Company case, A.06-01-004. Making these corrections to Aslam's calculations increased the average DCF equity cost estimate for the comparable group from 8.06% to 9.31%.59
· Aslam used six years of data to calculate what he called a five-year average, and eleven years of data for his ten-year average.60
· Aslam's correction of his data for Aqua America to reflect a stock split in 2005 somehow caused the indicated earnings per share in 2005 to go down while leaving the indicated earnings in 2004 unchanged. The witness was unable to provide an articulate explanation of that anomaly. Nor could he explain discrepancies between his presentation of Aqua America's earnings per share and those presented by the DRA witness in the Park Water case.61
· Aslam used erroneous data in applying his risk premium equity cost model. Some of the factual data were contradicted by SEC filings by companies in the comparable group while other data were conceptually wrong, e.g., based on year-end equity rather than average equity. These errors were also egregious in light of the fact that DRA had used correct data in the recent San Jose Water case.
Rerunning the DRA analysis with restated and corrected data, including a 60 basis point risk premium adjustment, produces a required ROE of 10.19%. See Section VII, below for detailed discussion of adjustments made to the discounted cash flow and risk premium models in response to comments.
D. Payroll Expense
Valencia's forecast of payroll expense for Test Year 2007-2008 is based on its recorded payroll expense for calendar year 2006, which includes the salaries for three new positions created and filled in 2005 and 2006. The 2006 amount is routinely escalated to the test year, but the test year amount also includes payroll amounts for three positions proposed to be added in 2007 and 2008. DRA's witness proposed to disallow the base year payroll expense to the extent of salary increases Valencia granted to certain employees in 2005 and 2006. He further proposed to disallow all payroll expense associated with the three new positions for which employees were hired in 2005 and 2006 and the three further positions for which Valencia plans to hire in 2007 and 2008.62
Valencia granted salary increases in 2005 to certain of its employees, based on a salary survey conducted in 2004. DRA calculates payroll expense by using the recorded expense for 2004 as a base and then applying standard labor Escalation Factors to estimate payroll expense for Test Year 2007-2008.63 By doing so, DRA excluded the salary increases Valencia granted in 2005 from the base on which DRA calculated payroll expense for the Test Year. DRA's RO Report provided three reasons for the proposed disallowance of payroll costs related to the 2005 salary increases. These were: (1) that Valencia "did not receive authorization from the Commission to implement the adjustment in payroll expenses"; (2) that "[t]he Commission was not given the opportunity to address this matter because [Valencia] failed to bring this matter before the Commission"; and (3) that Valencia's application "does not provide an explanation of the survey and the justification of implementing the adjustments to payroll expense."64
DRA witness Matsuoka testified that Valencia should have sought approval from the Commission or from the Water Division prior to this GRC to recover, or at least to track, the increased expense associated with the salary increases granted in 2005, even though Valencia did not intend to seek recovery of such costs for the period prior to Test Year 2007-2008.65 Absent such prior approval, Matsuoka testified that Valencia should not be allowed to request the recovery of salary increase costs in the future Test Year.66 Although Matsuoka acknowledged that it was not within the Water Division's authority to tell Valencia whether to increase its employees' salaries, he insisted that "as far as estimating for future test years for a GRC, that amount remains questionable as a basis," and he questioned it.67 He did not assert that it was an imprudent business decision for Valencia to grant the 2005 salary increases, but he emphasized that for ratemaking purposes it would be necessary to find the expenses reasonable. He agreed that the record developed in this GRC would provide the basis for that determination.68
It is not necessary for Valencia to obtain our approval before adjusting salaries even though, as noted by Matsuoka, Valencia needs to demonstrate that those salary increases were reasonable in order to recover them in rates. Valencia's salary survey compared Valencia's salaries to those offered by neighboring water companies. The survey showed that Valencia's salaries were below the low end of the salary range of the other companies surveyed. In response to this information, management adjusted Valencia's salaries to meet competition and prevent the loss of valuable employees.69 This was a prudent decision undertaken on the basis of relevant information and Valencia is entitled to seek recovery of those increased salary costs in this GRC.
DRA would disallow the salaries associated with three employee positions Valencia created in 2005 and 2006: a Conservation Coordinator and two Water Treatment Technicians. Valencia created these positions to ensure compliance with specific conservation and water treatment mandates.
5 Applicant Exhibit 8.
6 Exhibit 8 (DiPrimio), at 2.
7 Id. at 34.
8 Evidence at the hearing established that a new treatment plant that would reduce chloride discharges in the river to acceptable levels could cost as much as $447 million. Exhibit 8 at 34. The technology that would be employed in such a plant would be the so-called "reverse osmosis" process, in which hard water is forced through a membrane that filters out the chlorides and disposes of the resulting brine by piping it through a dedicated line to a disposal site.
9 Exhibit 8 (DiPrimio) 2. As discussed more fully below, DRA objects to the construction of the demonstration water softening project in part because Valencia customers would be charged for conferring a benefit on non-Valencia customers, i.e., the downstream users of water from the Santa Clara River.
10 Id. (Takiichi) 1-3 and 9-29. The Applicant's expert testified that pellet softening would not completely eliminate dissolved minerals in the treated water but would reduce them enough that the blended water received by Valencia's customers would be acceptably soft. To further soften the water would require the use of the reverse osmosis process and would not be cost-effective.
11 Id. at 4, 37-44.
12 Id. at 4, 37-44.
13 Exhibit DRA-7 (Gomberg), at 4-12.
14 Id. at 4-13 to 4-14.
15 Id. at 4-14 to 4-15. DRA introduced no evidence to rebut the cost savings estimated in the Kennedy-Jenks study.
16 Id. at 4-16 to 4-17.
17 Id. at 4-18 to 4-19.
18 $300 per customer times 28,300 customers = $8,490,000.
19 Tr. 221:19-222:21 (DiPrimio).
20 See Application, 9-11; see also Exhibit 1 (Conway/VWC), at 11-1 to -2.
21 Exhibit DRA-12 (Aslam), at 1-1; see also id. at 1-2.
22 Aslam refers to this citation as appearing in Valencia's application. In fact, it appears in Exhibit 4 (Zepp), at 29.
23 Exhibit DRA-12 (Aslam), at 3-3.
24 Id.
25 Id. at 3-3.
26 While the record is silent on the comparative costs of capital to parent and subsidiary, we note that Lennar is engaged in a risky and volatile business, speculative real estate development, while Valencia, as a regulated company, has a guaranteed rate of return on its capital and the legal ability to recover its costs through rates. It is not unusual for a regulated subsidiary to have a lower cost of capital than an unregulated parent in such a situation and it is unlikely that the parent would choose to replace lower-cost subsidiary capital with higher-cost parent capital.
27 Exhibit 33 (Zepp), at 5.
28 Exhibit 25 (Milleman), at 24.
29 Id. at 24-25.
30 In a recent Form 8-K report filed with the Securities and Exchange Commission (SEC), Lennar reported a net loss of $195.6 million, or $1.24 per diluted share, for its 4th quarter ended November 30, 2006, compared to net earnings of $581.2 million, or $3.54 per diluted share, in the 4th quarter of 2005. Lennar Corporation, Current Report Form 8-K, filed January 17, 2007, Exhibit 99.1.
31 Exhibit 25 (Milleman), at 27.
32 DRA recommended a reduction from 3.05% of preferred stock to an imputed 1%. However, there was insufficient justification for this recommendation and it ignores the fact that preferred stock functions effectively like debt in that preferred dividends are essentially equivalent to bond interest that is periodically paid to the holder. Thus little, if anything, is gained from substituting one form of fixed-payment obligation for another.
33 From time to time, to keep water rates reasonable, we have imputed a capital structure that includes approximately one-third debt to some small water companies that are actually capitalized with 100% equity. See, e.g., Great Oaks Water Company, Resolution W-4594 (May 11, 2006) (Exhibit DRA-8), at 8. However, Valencia's rates are already moderate relative to the rates of comparable companies in the region and we do not believe that an artificial capital structure is necessary to insure continued reasonable rates.
34 DRA 12 (Aslam) at 1-2.
35 Exhibit 33 (Zepp), at 4.
36 Id. at 5.
37 Exhibit 4 (Zepp), at 15-16.
38 Id. 16-19.
39 Id. at 19-26.
40 Id. at 26-28.
41 Id. at 6, 31-39.
42 Id. at 39. Zepp explained that estimates of forward-looking growth for Connecticut Water are not available, and so for purposes of DCF analysis, growth rates must be determined from other sources and DRA's approach of relying on past data but not including past growth in stock prices produces understated results.
43 Id. at 40-41.
44 Id. at 41-42.
45 Id. at 42-44.
46 Id. at 42-44.
47 Id. at 6, 44.
48 Id. at 54.
49 Aslam's testimony refers to seven publicly-traded water companies but his analysis used only the same 6 companies used by Zepp.
50 Compare Exhibit 4 (Zepp), Tables 8 through 12, with Exhibit DRA-12 (Aslam), at 2-4 to 2-5 and Table 2-5, and Exhibit DRA-15 (Aslam), Table 2-5.
51 Exhibit DRA-15, Table 2-3.
52 Exhibit 4 (Zepp), at 33-34. In support of this argument, Zepp cited a study finding that relying only on forecasts of earnings growth yielded better results in a DCF analysis.
53 See Exhibit DRA-12 (Aslam), at 2-5 to 2-6 and Tables 2-6 through 2-8.
54 Id. at 4-2.
55 Id. at 4-3 to 4-6.
56 Zepp testified that some utilities in the comparable group have sold stock at prices above book value in recent years, thus achieving "sv growth," and that knowledgeable investors would expect such "sv growth" in the future. Exhibit 4 (Zepp), at 36. He explained that failure to recognize this type of growth results in serious understatement of the overall earnings growth rate. Id. at 37-38.
57 Exhibit DRA-12 at 4-7 to 4-8.
58 Id. at 4-8 to4-9.
59 Exhibit 33 (Zepp), at 8; see also, Tr. 170:11-173:12, 176:24-177:19 (Zepp).
60 Tr, 255.15-257.7.
61 To help overcome this factual confusion, Valencia requested and was granted permission to submit a late-filed exhibit presenting earnings information for one or more of the "comparable group" companies. Valencia provided that information by late-filed Exhibit 40 to present a corrected calculation of earnings per share growth for Aqua America and San Jose Water based on their SEC Form 10-K filings for the years 2002, 2004, and 2005. The excerpts from SEC filings in Exhibit 40 reflect the effects of several stock splits for each of the two companies. Attachment 5 to Exhibit 40 extracts relevant earnings per share data from the SEC filings and displays those data, adjusts them to reflect recent stock splits, and thereby presents comparable earnings per share data for years 2000 through 2005 for the two companies. On this basis, Attachment 5 shows average annual growth in earnings per share over that five-year period as 7.8% for Aqua America and 15.5% for SJW Corp. (the parent company of San Jose Water). The Attachment 5 table shows that a single error in the most recent year for Aqua America resulted in a 5.6% understatement of its earnings growth rate, while a series of errors in SJW Corp. data produced an 11.0% understatement of the earnings growth rate for that company. These errors caused Aslam's calculation of five-year earnings growth for the six-company "comparable group" to be understated by nearly 3%: (11.0% +5.6%) / 6 ≈ 2.77%. As summarized in Exhibit 40, at 3:
"[T]hese substantial understatements of earnings growth for two of DRA's six comparable group companies contributed to a substantial error in Mr. Aslam's Discounted Cash Flow analysis and, consequently, a substantial underestimation of Valencia's cost of equity capital."
62 Exhibit DRA-7 (Matsuoka), at 3-4 to 3-9; Tr. 116:20-25 (Matsuoka); Exhibit 25 (Milleman), at 2.
63 Exhibit DRA-7 (Matsuoka), at 3-5 to 3-6.
64 Id.
65 Tr. 117:20-119:25 (Matsuoka).
66 Tr. 120:9-15.
67 Tr. 121:1-14.
68 Tr. 121:15-122:8.
69 Exhibit 28 (Data Request EYM-17), at 2 (Response to Request 3); see also Exhibit 25 (Milleman), at 2-3.
