THE ELECTRIC INDUSTRY CONTINUES to evolve in different forms across the country. Some proposed changes are radical, others are more evolutionary, but few utilities remain unaffected. Competition in the generation sector sharpens the focus on cost-effectiveness, market share, and new lines of business. Investor-owned utilities (IOUs) are reorganizing their current business structures and regulators are struggling either to lead the charge or to keep up.
One of the biggest challenges facing regulators is to encourage the benefits of competition while protecting electric consumers from excessive rates that produce windfall profits for shareholders. For example, over the past few years, California's IOUs have asked for a higher allowed return on common equity (ROE) to compensate shareholders for the additional risk associated with competition in the electric generation sector. In 1994, following the landmark "Blue Book" restructuring proposal from the California Public Utilities Commission (CPUC), the IOUs cried even more loudly for higher returns in their annual rate proceeding on cost of capital.
In that docket, the Independent Energy Producers Association (IEP) asked the CPUC to consider a new approach to regulating the cost of capital for California's electric utilities. In parallel with the unbundling of electric services and associated rates, the IEP asked the CPUC to unbundle the cost of capital.1 The premise of IEP's argument was that unless utilities disaggregate the risks of providing unbundled services, along with rates and the allowed shareholder returns, IOU customers would end up overpaying for certain services, such as transmission and distribution (T&D) activities, while underpaying for others, such as generation. In addition, utility generation services would gain an unfair advantage over other
generation-only competitors. In a decision issued in November 1994, the CPUC agreed with the argument presented by IEP and charged that any future unbundling of services should include an unbundling of the ROE and rates associated with those services.2
Since the CPUC's 1994 order, California's IOUs have made significant progress towards disaggregating and unbundling their electric services. For example, Pacific Gas & Electric (PG&E) and San Diego Gas & Electric (SDG&E) have filed applications with the CPUC to reorganize utility assets and form holding companies. Once restructured, those companies will operate separate business entities for the production, marketing, and delivery of electric services. In addition, these two utilities and Southern California Edison (SCE) have each filed open-access transmission tariffs with the Federal Energy Regulatory Commission (FERC) to provide transmission-only services to customers. Consistent with the CPUC's decision, it would appear that now is the time for unbundled electric tariffs to include ROE components that reflect the risk of providing unbundled services.
THE CHALLENGE AT HAND
The challenge in unbundling the cost of capital comes not in proving the theory to be fair and equitable, but in determining how and when to do it. As concerns how, the traditional methods of determining cost of capital require identifying and collecting data for comparable groups. Since most electric services have been vertically integrated for a considerable time, truly comparable groups do not exist for unbundled services, especially for T&D. Absent
comparable groups, the job of accurately setting unbundled ROEs requires a detailed review and breakdown of the costs and risk exposure of the various components of the utility's assets, separated by business function: generation, T&D, and (where applicable) gas services. This analysis would likely require collecting a massive amount of data, some of which is not readily available.
As for when, some utility representatives have argued that unbundling risks now would amount to nothing more than an accounting exercise, since rates are not yet unbundled, and would hinder the IOUs' abilities to compute their annual revenue requirements. Further, several jurisdictional issues between the FERC and the CPUC regarding transmission pricing, in light of the CPUC's decision regarding unbundling, complicate the implementation process.
GENERATION IS RISKIER
Both financial analysts and industry players agree that the bulk of the risk faced by IOUs today lies primarily with the generation function. First, the generation function is being opened to competition; the T&D function, at
present, remains largely a monopoly insulated from competitive risk. Second, the generation function is inherently riskier due to the variability associated with its inputs, such as fuel prices and availability, and construction risks associated with new generating capacity.
For example, the criteria Standard & Poor's (S&P) developed for its bond ratings of electric utilities demonstrate that the majority of business risks faced by utilities arise from the generation function (see Table 1 on next page).3
This is not to say that the T&D function does not carry its own risks, such as those associated with prudent use of the T&D system and recovery of capital investment in new facilities and upgrades of existing facilities. However, on balance, the risks of producing kilowatt-hours (generation) are significantly greater than the risks of delivering them (T&D).
In fact, the FERC's 1991 ruling on the Nevada Sun-Peak case,4 which allowed a ceiling of 15 percent for that EWG's ROE, indicates that exempt wholesale generators (EWGs) bear more risk than an electric utility, which provides both generation and T&D.
As a third example of relative risk, for illustrative purposes, we examined the difference in betas (as a proxy for risk) between the competitive and regulated arms of the natural gas and telephone industries, which have recently been restructured. In the natural gas industry, pipeline companies (which operate in a competitive industry) post a higher beta (em 0.92 on average (em than local distribution
companies (regulated natural monopolies), which show an average beta of approximately 0.64.5 Similarly, during the last decade, the telephone industry was deregulated and long-distance carriers were exposed to competitive risks. ROEs for long-distance carriers increased steadily due to increased competition, while ROEs for the regional Bell operating companies (the regulated arm) remained relatively constant.6
More competition, and less of a safety net once direct access has been established, will continue to increase the risks associated with the generation side of the business for integrated utilities. Therefore, unbundling of the utility's functions and the cost of capital should occur sooner rather than later.
Unbundling now would prevent utilities from gaining an unfair
advantage (from possible cross-subsidies) as they compete in the generation-only business.
CALCULATING THE UNBUNDLED COST OF CAPITAL
This increased risk in the generation function, coupled with recently filed open-access transmission tariffs, argues for unbundling the cost of capital now (em even though both functions are still regulated.
Traditional cost-of-capital models rely on comparable groups. In today's world, however, identifying comparable groups for unbundled electric utility functions is problematic, if not impossible. Currently, the closest thing to a generation comparable group in the United States is the universe of independent power producers (IPPs); as a "comparable" to utility generation, however, this group presents problems. First, only a small subset of the group are publicly traded as stand-alone companies. Many, if not most, of the major players are themselves
subsidiaries of electric utilities. Second, these companies do not have the same franchise rights enjoyed by regulated utilities. Third, IPPs employ different financing and ownership structures that make comparisons to utilities difficult. For a transmission comparable group, there are currently no "pure-play" T&D companies in the United States, except perhaps for municipal utilities. The ownership and financing structures of the municipals, however, are even further removed from those of the IOUs.
In light of the difficulties associated with uncovering comparable groups, one may ask: Can proxies be used? In our opinion, certain proxies are valid for use in determining unbundled ROEs. In fact, we have developed a simple method, using two proxies, to estimate the potential magnitude of the ROE differential between generation and T&D functions.
Proxy 1. First, we looked for available proxies for pure-play generation company ROEs (see the caveats listed above). One such proxy is from the FERC decisions granting market-based rates for EWGs, which are at least an approximation of generation-only companies. In the Ocean States Power I and II cases7, for example, the FERC approved market-based rates that included an ROE component 15 percent higher than the FERC's generic benchmark ROE for utilities. In the Nevada Sun-Peak case, the FERC stated that a reasonable ROE would approach or exceed 15 percent, the upper end of the range it considered appropriate for traditional electric utilities.
Another such proxy is from reported ROEs for publicly traded IPPs. At the time of this analysis, those ROEs ranged from 15.1 to 17.9 percent for AES, California Energy, Destec, and Magma. These results are also consistent with a recent paper by David P. Wagener that estimated an implied cost of equity for IPPs of 16.9 percent.8
Based on those proxies, we estimated that, at a minimum, the ROE for the generation function is at least 15 percent higher than the overall ROE for the combined generation/T&D utility. Consider SCE as an illustrative example.
Based on the split between generation rate base and T&D rate base for the utility and the assumption of identical capital structure for the generation and T&D functions, we calculated an
implicit ROE for SCE's T&D function.9 That analysis produced a differential of 326 basis points, as shown in Table 2. We believe this figure represents a conservative estimate, since this method yields relatively low generation ROEs compared to our proxies. Using a generation-only ROE of 15 percent would have increased the ROE differential between generation and T&D by more than 500 basis points.
Proxy 2. Second, we looked at beta differentials in two recently deregulated industries: natural gas (transmission and distribution) and telecommunications (long-distance and local service). In the gas industry, the betas of pipelines and local distribution companies differed by about 0.28. In the telecommunications industry, betas of AT&T and the regional Bell companies vary by 0.21 to 0.26.10 Using these as proxies for the generation and T&D functions, we estimated a likely beta differential between generation and T&D of 0.25. ROE differentials can be derived from beta differentials using the Capital Asset Pricing Model (CAPM):
ROE difference = Beta difference x (Rm-Rf)
This analysis produced an ROE differential of 175 basis points for SCE:
1.75 = 0.25 x 7.0FN
Based on these results, with bundled ROEs, customers purchasing T&D services only would be overcharged and would subsidize other utility functions. The cross-subsidy that would occur is not insignificant. For 1995, for example, the ROE component of total revenue requirements is $678 million for SCE, at an allowed ROE of 12.1 percent. Based on unbundled ROEs of 13.92 percent for generation and 10.66 percent for T&D, the resulting cross-subsidy is roughly $45 million per year.
Our proxy methods and example provide only a first cut approximation of the potential magnitude of the differential in appropriate ROEs and the resulting cross-subsidy for the different utility functions. It assumes that the appropriate unbundled capital structure for the utility's generation and T&D function is the same as its current bundled ratemaking capital structure. In fact, full unbundling would appropriately include adjusting each function's capital structure based on relative risk, which could in turn affect the overall cost of capital associated with each function.
Looking forward, if and when generation is operating in a truly competitive market, generation ROEs could be determined using a market-based approach as opposed to traditional ratemaking models. Such an approach would work only if: (i) regulated utilities had divested their generation assets such that the resulting GenCos are competing on a level playing field with other power producers; or (ii) to the extent that divestiture does not occur, assurances are made that there is no cross-subsidization across utility functions (i.e., the generation arm of the utility is not subsidized through inappropriate allocations of utility costs to the T&D function).
In addition, the current method of using comparable utilities and market betas to calculate the allowed ROE for utilities is just as inappropriate as maintaining bundled ROEs for the generation and T&D components of these IOUs. For example, many of the IOUs provide gas services, which certainly have different risk profiles than the electric-related services component. Most IOUs also have subsidiaries that are involved in unregulated activities. The risks associated with gas services and these unregulated subsidiaries, however, are included in the betas used to determine the overall cost of capital for these IOUs. A method that more accurately
reflects the actual risks associated with the electric services portion of the utilities on an unbundled basis is recommended.
In the final decision for the 1995 cost-of-capital proceedings, the CPUC stated that "unbundling costs of capital is economically sound, will send correct price signals to energy markets, and will mitigate cross-subsidies. The utilities should address unbundling of costs of capital in any future proceedings that aim to unbundle electric or gas rates and services."11 So far, little progress has been made toward this end.
Since this decision, PG&E, SDG&E, and SCE have filed open-access transmission tariffs with the FERC. In SCE's proposal, for example, transmission rates were determined using SCE's weighted-average (i.e., bundled) cost of capital, as specified by the FERC. Because the unbundled cost of capital for transmission would likely have been lower than the utility's weighted-average cost of capital, it is likely that the transmission rates approved by the FERC will be higher than appropriate. In fact, unbundling the cost of capital would not only result in a lower ROE for transmission services, but would also likely result in lower debt costs and/or higher debt/equity ratios, all of which would reduce the overall cost of capital component of revenue requirements for transmission services.
We performed an illustrative calculation for SCE to provide a rough estimate of the savings that could result by determining transmission rates on an unbundled basis. Using conservative estimates for ROE, return on debt, and capital structure for the transmission function alone, we estimated that unbundling the cost of capital could result in savings on the order of 2 to 8 percent of the overall cost of service under the open-access tariff. In terms of dollars, for a customer buying 500 MW of transmission service on SCE's system, a 6-percent reduction in the tariff would result in savings of roughly $750,000 per year.
As long as utility electric services remain bundled, and as long as none of these services is opened to competition, the difference in risk between the generation function and T&D function of an IOU does not really matter much for ratemaking. However, competition exists in the generation sector today, and unbundling of electric services is occurring now. As a result, the cost of capital component of each utility service function needs to be determined separately. By doing so, resulting rates will be appropriate for the services provided, and customers will pay for only the services they use.
As demonstrated, estimating ROEs using traditional models requires a series of complicated, data-intensive calculations. And, since there currently are no truly comparable groups, the exercise becomes that much more problematic. The concept of finding pure-play comparables among utility services, even those in a holding company structure, seems unrealistic as well. Until market-based solutions are available, simple proxies can be used.
Finally, acceptance of unbundling the cost of capital must extend beyond California and should be considered at the FERC as well. Because certain tariffs, such as most open-access transmission tariffs, are under the FERC's jurisdiction, the FERC must rethink the manner in which it requires utilities to calculate rates. Doing so will help maintain proper price signals and promote fair competition among power generators. t
Susan Stratton Morse, CFA, is a financial analyst/advisor and a co-founder of the firm of Morse, Richard, Weisenmiller & Associates, Inc., Oakland, CA. Her testimony and this article were developed jointly with Meg Meal, CFA, also a financial analyst and a principal at MRW. Melissa Lavinson is an associate at MRW.
The Case for Unbundling Cost of Capital
Despite moves toward electric deregulation, state regulators typically do not distinguish between the generation and T&D functions in setting capital structure or ROE. The costs of capital for each segment are quire different. Failing to recognize this difference produces misleading risk profiles.
As long as utilities retamin regulated and have bundled capital structures, regulators must measure risk appropriately for unbundled services to ensure accurate pricing. Unbundling the cost of capital hand-in-hand with electric services will:
. prevent cross-subsidization between generation and T&D
. prevent cross-subsidization between full-service customers (i.e., those who buy generation, transmission, distribution, and other ancillary services) and customers buying only certain services (e.g., transmission)
. send clear and correct price signals to customers and generation providers
. allow market forces to determine the appropriate ROE and investment returns for the generation sector.*
*From Testimony of Susan Stratton Morse on Behalf of the Independent Energy Producers Association on Recommendations for Determining the Allowed Return on Equity for California's Electric Utilities in Light of Restructuring. A.94-05-009, page 6, August 15, 1994.
1. From Testimony of Susan Stratton Morse on Behalf of the Independent Energy Producers Association on Recommendations for Determining the Allowed Return on Equity for California's Electric Utilities in Light of Restructuring, A.94-05-009, August 15, 1994.
2. Re Sierra Pacific Power Co., Application 94-05-009, Decision 94-11-076, Nov. 22, 1994, 158 PUR4th 217.
3. Criteria taken from "Credit Comments," Standard and Poor's Creditweek, October 11, 1993, pages 7-12.
4. FERC Docket No. ER91-11-001.
5. From Testimony of Susan Stratton Morse on Behalf of the Independent Energy Producers Association on Recommendations for Determining the Allowed Return on Equity for California's Electric Utilities in Light of Restructuring. A. 94-05-009, page 29, August 15, 1994.
7. FERC Docket Nos. ER88-478-000, ER91-576-000.
8. Wagener, David P., "Letting Go of Electric Generation," PUBLIC UTILITIES FORTNIGHTLY, Feb. 15, 1995, pp. 33-35.
9. For the purposes of this article, we chose to use information from SCE to demonstrate ROE differentials since SCE offers only electric-related services, as opposed to PG&E and SDG&E which offer both electric- and gas-related services. However, applying our methods to these other utilities yields similar results.
10. From Testimony of Dr. Lawrence A. Kolbe on Behalf of Southern California Edison on Cost of Capital Restructuring Effects, CPUC A.94-05-017, May, 1994, pp. 10-12.
(FN) Represents the difference between the expected return on the market portfolio (Rm) and the risk-free rate (Rf), as presented in SEC's Open-access Transmission Tariff filing at the FERC (Dkt. ER96-222-000).
11. From CPUC D.94-11-076, page 24.
Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.