In June, the Federal Energy Regulatory Commission ("FERC" or "Commission") issued Opinion No. 531, which details three significant changes to the way FERC determines the rate of return on equity ("ROE") in public utility rate cases.[1] First, FERC modified its longstanding discounted cash flow ("DCF") model for calculating ROE.[2]Second, FERC ended its practice of applying a post-hearing adjustment to ROE based on changes in United States Treasury bond yields.[3]Third, FERC decided that the ROE of the group of public utilities in question should not be set at the "point of central tendency" established by the range of reasonable ROEs, but instead should be set at the point halfway between the range's point of central tendency and the range's highest point.[4] Each of these changes is explained below. 

In conjunction with Opinion No. 531, FERC set for hearing a backlog of cases involving disputes over public utility rates.[5] In each case, the Commission stated that "we expect the evidence and any DCF analyses presented by the participants in this proceeding to be guided by our decision in Opinion No. 531."[6] As compared to FERC's preexisting approach to ROE, Opinion No. 531 appears likely to increase the ROE component in public utility cost-based rates, thereby increasing a public utility's overall return. However, the long-term reach of FERC's new ROE analysis remains unclear.

How Regulators Use the DCF Model and Establish a Range of Reasonable ROEs

Under cost-of-service ratemaking, certain costs are considered operating expenses and are recovered dollar-for-dollar in the utility's annual revenue requirements, while other costs are capitalized, "thus entering the cost of service in the form of annual allowances for depreciation and return on the undepreciated portion of the investment."[7] In order to attract necessary capital, the utility must offer "a risk-adjusted expected rate of return sufficient to attract investors."[8] This "return" on the utility's investments represents the cost expended by the utility to raise capital.[9] 

For more than 30 years, the dominant method used by FERC to estimate investors' required rate of return has been the DCF model. The premise of the DCF methodology is that "an investment in common stock is worth the present value of the infinite stream of dividends discounted at a market rate commensurate with the investment's risk."[10] This "constant growth" DCF model can be expressed as a formula:

k = D/P (1+0.5g) + g.

In this formula, D is the current dividend, P is the price of the company's common stock, and g is the expected growth rate in the company's dividends. The formula solves for k, which represents the rate of return investors require to invest in a company's common stock.[11] 

In a rate case, FERC applies the DCF formula to each member of a group of comparable utilities, known as a proxy group. This generates a range of ROEs. Screening criteria, which can result in the exclusion of particular companies from the analysis, are applied to establish a range of reasonable ROEs. The ROE of the public utility that is the subject of FERC's review is selected from within this range. FERC's past practice has been to set the subject public utility's ROE at the point of central tendency of the proxy group's range of ROEs. 

Applying the Two-Step DCF Methodology to Public Utilities 

Since the mid-1990s, FERC has used a "one-step" DCF methodology in public utility rate cases (i.e., in the electric industry) while using a "two-step" DCF methodology in natural gas pipeline and oil pipeline rate cases. Under the one-step DCF methodology, FERC calculates two dividend yields for each proxy group company: one based on the proxy group company's highest stock price from a six-month study period, and one based on the proxy group company's lowest stock price from the same study period. Next, FERC develops two estimates of short-term dividend growth rates. A low cost of equity for each proxy group company is developed using the lowest dividend yield plus the lowest dividend growth rate projection. A high cost of equity for each proxy group company is developed using the highest dividend yield plus the highest dividend growth rate projection.[12]

In Opinion No. 531, FERC decided that henceforth it will apply the two-step DCF methodology in public utility rate cases. The result is a single average dividend yield calculated for each company in the proxy group. The dividend growth rate estimate for each proxy group company will take into account both projected short-term growth rates (constituting two-thirds of the total growth rate estimate) and projected long-term growth (one-third of the total).[13] The short-term growth rate estimate will be based on the five-year forecast for each proxy group company, as published in the Institutional Brokers Estimate System ("IBES"). The long-term growth rate estimate will be based on forecasts of the long-term growth of the economy as a whole, stated in terms of gross domestic product.[14] In the case before it, FERC reopened the record and established a "paper hearing" to give the participants in that case "an opportunity to present evidence concerning the appropriate long-term growth projection to be used for public utilities under the two-step DCF methodology."[15] 

Eliminating the Post-Hearing ROE Adjustment Based on U.S. Treasury Bond Yields

FERC's cost-of-service ratemaking for public utilities relies predominantly on "test-period" evidence, which is evidence about the subject utility's costs limited to a specific time period that ends before the rate case goes to hearing. Use of test-period evidence gives the parties a known universe of facts to dispute. One exception is FERC's use of post-test-period data regarding ROE. FERC's practice has been to adjust the subject utility's ROE based on U.S. Treasury bond yields. FERC determines the change in U.S. Treasury bond yields as of the date of its order as compared to such yields as of the end of the hearing in the case.[16] FERC then adjusts the final ROE by the amount of the change in U.S. Treasury bond yields. For example, a 1 percent drop in bond yields between the end of a hearing and FERC's order would result in a 1 percent downward adjustment to the utility's ROE.[17]

In Opinion No. 531, FERC decided that U.S. Treasury bond yields no longer "provide a reliable and consistent metric for tracking changes in ROE after the close of the record in a case."[18] Instead, FERC will allow participants in a rate case "to present the most recent financial data available at the time of the hearing, including post-test period financial data then available."[19] FERC already uses this approach in natural gas pipeline and oil pipeline rate cases.[20]

Selecting an ROE From Within the Range of Reasonable ROEs 

Once a range of reasonable ROEs is developed by applying the DCF model to each member of a proxy group, FERC must select one ROE from within that range. Traditionally, FERC has set the subject public utility's ROE at the "point of central tendency" within the range of ROEs. For a diverse group of public utilities, the point of central tendency is the midpoint within the range, which, as FERC uses the term, is the arithmetic mean of the single lowest and the single highest ROE. In contrast, for an individual public utility, the point of central tendency is the median.[21] The median is the middle number in a series, such that half of the numbers are higher and half are lower than the median. 

In Opinion No. 531, the issue in dispute was the appropriate ROE for the New England Transmission Owners, a group of utilities that had transferred operational control of their transmission facilities to ISO-New England. FERC decided that it would not set the ROE at the midpoint of the range of ROEs. Instead, FERC selected the point halfway between the midpoint and the highest point in the zone of reasonableness (the 75th percentile), which FERC described as the "central tendency for the top half of the zone of easonableness."[22]In Opinion No. 531, the resulting midpoint was 9.39 percent, but the point at the 75th percentile of the range was 10.57 percent. 

In rejecting the midpoint ROE, FERC explained its concern that the "capital market conditions in the record are anomalous, thereby making it more difficult to determine the return necessary to attract capital."[23] The New England Transmission Owners had argued that five other "benchmark methodologies" showed that the DCF-based midpoint in that case was too low to attract capital: (i) a risk premium analysis, which examines the premium that investors require to invest in equities; (ii) a capital asset pricing model ("CAPM"), which is a model that examines investor expectations about the future by taking into consideration the tendency of a stock's price to follow changes in the market as a whole; (iii) an analysis of natural gas pipeline ROEs; (iv) a DCF analysis applied to non-utilities; and (v) an expected earnings analysis, which involves a comparison of the earnings investors can expect to receive from investing in public utilities, as compared to investing in other opportunities of comparable risk.[24] 

FERC found that the risk premium analysis, the CAPM, and the expected earnings analysis were "informative" and supported the conclusion that the midpoint ROE was too low to attract capital. FERC did not consider the DCF analysis of non-utilities or the natural gas pipeline ROE analysis to be probative because they did not analyze the returns of public utilities.[25] 

At several points, FERC's analysis focused on the unique characteristics of companies in the business of building and owning electric transmission assets. For example, FERC found persuasive the fact that state regulatory commissions have approved public utility ROEs above the DCF midpoint ROE. According to FERC, transmission investment "entails unique risks that state-regulated electric distribution does not."[26] Investors in electric transmission infrastructure face risks such as "long delays in transmission siting, greater project complexity, environmental impact proceedings, requiring regulatory approval from multiple jurisdictions overseeing permits and rights of way, liquidity risk from financing projects that are large relative to the size of a balance sheet, and shorter investment history."[27] FERC emphasized that it has an obligation to set an ROE in this case "at a level sufficient to attract investment in interstate electric transmission," explaining that such investment "helps promote efficient and competitive electricity markets, reduce costly congestion, enhance reliability, and allow access to new energy resources, including renewables."[28] 

Opinion No. 531 may allow a utility to justify an ROE greater than the median without making a company-specific showing of relative risk. In prior decisions, FERC has required a showing that the risk affecting the subject company be higher than the risk faced by the other members of the proxy group. As recently as 2013, FERC explained that any analysis attempting to "demonstrate that a deviation from the median ROE is justified" must present a comparison between "the risk level of the subject company and the risk level of each of the proxy group companies. This is the crux of the analysis, and if it is lacking, the analysis is incomplete."[29] In light of the general evidence relied on in Opinion No. 531, a company-specific showing of relative risk may no longer be the "crux" of the analysis. 

Finally, FERC explained that its decision to set the New England Transmission Owners' base ROE above the range's point of central tendency involved considerations that are distinct from its analysis of "incentive adders" pursuant to Section 219 of the Federal Power Act.[30] FERC's task when evaluating a base ROE is to set the ROE at a level that "enables the utility to attract investment." In contrast, FPA Section 219 authorized FERC to establish incentive above that base ROE. FERC cautioned that it will not permit its new analysis of base ROE and its analysis of FPA Section 219 incentives to be combined in a way that results in an ROE that exceeds the top of the zone of reasonableness established by its new two-step DCF methodology.[31] 

Likely Effect on Public Utility Rates 

FERC's departure in Opinion No. 531 from three aspects of its existing policy on ROE will result in a higher ROE for the New England Transmission Owners. In the Initial Decision under review in Opinion No. 531, the Administrative Law Judge found that the prospective ROE for the New England Transmission Owners should be set at 9.7 percent.[32] In contrast, FERC's analysis in Opinion No. 531 resulted in a tentative finding that the appropriate ROE was 10.57 percent.[33]  

Opinion No. 531's higher ROE did not result from FERC's switch to the two-step DCF model. The two-step DCF model alone resulted in a midpoint ROE of 9.39 percent[34]—lower than the prospective 9.7 percent ROE approved in the Initial Decision. Rather, the higher ROE resulted from FERC's decision to select the midpoint of the "upper half" of the zone of reasonableness rather than selecting the midpoint of the full zone. Moreover, FERC's focus on the midpoint as the point of central tendency (because the ROE of a group of utilities was at issue) raises questions about how FERC's analysis from Opinion No. 531 will be applied in the context of setting a single public utility's ROE, where the Commission uses the median of the range of reasonable ROEs as opposed to the midpoint.[35]Notwithstanding the differences in the ROE analysis for a single utility and grous of utilities, the Commission's order in Seminole Electric Cooperative, Inc. v. Florida Power Corp. instructs the parties to apply Opinion No. 531 in establishing a single utility's ROE.[36]

FERC also emphasized that the new two-step DCF model produces "a narrower zone of reasonableness, consistent with the fact [that] different firms in a regulated industry would not ordinarily be expected to have widely varying levels of profitability."[37] This narrower zone of reasonableness may result in a point of central tendency within the "upper half" of the zone of reasonableness that is close enough to the overall point of central tendency to support the selection of that higher ROE. 

In sum, although two aspects of Opinion No. 531 will apply in all future public utility rate cases, the largest identifiable change in the ROE in that order was based on a case-specific analysis—performed in the context of "anomalous" capital market conditions—of the level of return needed to encourage investments in transmission where that ROE will apply to a group of public utilities rather than to a single public utility. Had FERC's ROE policies remained unchanged, the result would have been lower ROEs. Therefore, Opinion No. 531 is likely to increase public utility returns in cost-based rates, but the scope and magnitude of this effect as applied to other public utilities remains unclear.