7. Earnings Curve and Associated Shared-Savings Rate(s)

The earnings curve and associated shared-savings (or "sharing") rate(s) create the opportunity for earnings under the risk/reward incentive mechanism, once the MPS is achieved. All parties agree that this opportunity should be based on a percentage of the verified net benefits achieved by the portfolio of energy efficiency programs. For example, if verified portfolio net benefits are $100 million when a utility achieves 100% of the savings goals, and the sharing rate is 10% at that level of performance, then ratepayers would retain $90 of the $100 million in net benefits and pay the utility $10 million in earnings. Parties also agree that the earnings curve should be "tiered" in structure, that is, the shared-savings rate (the percentage of net benefits shareholders receive) should increase at higher levels of portfolio achievement with respect to the savings goals.

However, parties fundamentally disagree on how to establish the shareholder earnings potential under the incentive mechanism, the appropriate level for such earnings and the associated shared-savings rates. The utilities and NRDC argue that supply-side "comparable earnings" is a key benchmark for this purpose. Accordingly, these parties calculate what the utility would otherwise earn on the portfolio of supply-side resources avoided by energy efficiency to establish this benchmark. ⁴⁰ DRA, TURN, CE Council and CLECA, on the other hand, argue that using a supply-side return is an excessive benchmark for establishing an energy efficiency incentive where ratepayers, not shareholder, dollars are at risk. They use different approaches to establish their recommended shared-savings rates, which result in substantially lower earnings potential than under the risk/reward incentive mechanisms proposed by the utilities and NRDC.

The parties also disagree on the appropriate methodology for calculating supply-side comparable earnings, should the Commission adopt this approach. In particular, there is disagreement over whether: 1) comparable earnings calculations should impute any earnings associated with purchased power to reflect debt equivalence, 2) supply-side investment returns should be reduced to account for alternative use of capital, and 3) comparable earnings should be evaluated with respect to the same amount of savings, or the same amount of investment.

Finally, parties disagree on how the tiered shared-savings rates should be structured beyond the MPS, and whether they should include a larger incentive (in the form of a higher earnings rate) for achieving higher levels of kW savings.

In the following sections we address these and other issues related to the level of sharing of net benefits between ratepayers and shareholders. We begin with a description of the utilities' supply-side comparable earnings analysis, which other parties refer to extensively in their filed comments and testimony.

7.1. Utility Supply-Side Comparable Earnings Analysis

In developing proposals for a shared-savings rate, each utility conducted what we refer to as a "supply-side comparable earnings analysis" or simply, a "comparable earnings analysis." This analysis calculates the level of shareholder earnings associated with procuring supply-side resources that are displaced by energy efficiency. As discussed in this decision, there is disagreement among the parties on several key assumptions that go into performing this analysis. Moreover, parties disagree on the relevance and purpose of using supply-side comparability as a benchmark for the shared-savings rate. However, before discussing those disagreements, it is useful to review how the supply-side comparable earnings analysis is performed.

Mechanically, the utilities perform this analysis by estimating the amount and type of supply-side resources avoided if they achieve 100% of the savings goals over the 2006-2008 program cycle with energy efficiency. They then determine the revenue requirement they would need to recover from customers as a result of these avoided supply-side procurements. Some of the procurements would be from "steel-in-the-ground" supply-side resources (e.g., avoided generation, transmission and distribution facilities) and some would be from avoided power purchases. For their base case analysis, each of the utilities assumed a 50-50 split between these utility-build and utility-buy scenarios.

The utilities' revenue requirement calculations for "steel-in-the-ground" supply-side resources reflect that these costs would be rate-based. That is, the utility finances such infrastructure projects through a combination of debt and equity capital, and then rates are set to recover the original investment plus an authorized cost of capital (interest on debt and return on equity). The revenue requirement calculations, including the return to shareholders for their capital (equity) investment, are also grossed up for all applicable taxes.

For purchase power contracts, the revenue requirement is recovered contemporaneously from ratepayers through balancing accounts. Therefore a cost of capital (debt or equity) is not included in those revenue requirement calculations. However, in producing their revenue requirement calculations, the utilities impute what is referred to as "debt equivalence" for the purchased power assumed in the analysis.

Debt equivalence is a term used by credit rating agencies for treating long-term non-debt obligations, such as power purchase agreements, as if they were debt in assessing a utilities' credit rating.⁴¹ The utilities assume that 30% of the dollar value of the purchase power contracts included in their supply-side comparable analysis is equivalent to additional debt in their capital structure. This increases the proportion of equity capital required in the utility's capital structure which, in turn, increases the total return on equity included in calculation of the avoided revenue requirement.

The dollar level return on equity that is included in the revenue requirement calculations described above is what we refer to as "supply-side comparable earnings." This is the amount of earnings to shareholders that, but for energy efficiency, the utility would be authorized to collect in rates. Dividing this level of earnings by the net benefits (PEB) expected from the energy efficiency portfolio (at 100% goal achievement) yields the "supply-side comparable" shared-savings rate.

Table 2 below summarizes the results of each utility's supply-side comparable analysis, which incorporate the following base case assumptions: 1) a 50-50 split between the "utility build" and "utility buy" scenario, 2) average energy efficiency measure life of 12-years and 3) debt equivalence for purchased power. The comparable earnings shared-savings rate is calculated by dividing the comparable earnings by the PEB (net benefits).⁴² All numbers are presented on a pre-tax basis.

7.2. Position of the Parties

Parties' proposals for the earnings curve, i.e., the shared-savings tier structure and associated shared-savings rates, are compared in Attachment 3.⁴³ As indicated in that attachment, DRA, TURN and CE Council propose shared-savings rates in the 1.5% to 3% range beyond the MPS, plus a higher earnings rate of up to 3.5% if kW goals are exceeded by 25-50% (depending on the proposal).⁴⁴ NRDC proposes shared-savings rates in the 6% to 12% range beyond the MPS, and the utilities propose rates in the 10% to 30% range, depending on the tier.

Attachment 2 presents these proposals in terms of the dollar earnings at different levels of savings goal achievement, based on a PEB that includes all portfolio costs.⁴⁵ Some parties to this proceeding suggest that we view proposed dollar earnings relative to portfolio costs, that is, as a ratio of earnings to portfolio funding levels. This perspective fails to recognize that the sharing rate and associated earnings are not tied to the energy efficiency portfolio costs, but rather to the much larger dollar value of avoided supply-side costs. We continue to endorse the yardstick we set in D.03-10-057 that the earnings levels we establish under a shared-savings mechanism should be compared to "how much ratepayers would have had to pay if the program savings had not been realized."⁴⁶

From this perspective, we present below (1) the shareholder earnings level for all four utilities combined at 100% of savings goals, and (2) the ratepayers' portion of net benefits under each party's sharing proposal:

TABLE 3
Shareholder Earnings/Ratepayer Net Benefits by Proposal
Sharing Proposal Of:	Shareholder Earnings at 100% of Goals (all utilities combined) ($ million)	Ratepayer Net Benefits at 100% of Goals47 ($ million)
PG&E	$538	$2,151
SCE	$538	$2,151
SDG&E/SoCalGas	$403	$2,286
NRDC	$323	$2,366
DRA	$81	$2,608
TURN	$54	$2,635
CE Council	$54	$2,635

Before turning to the specific disputed issues on how to establish those levels, we note that the expected ratepayer "return" (net benefits) from the $2.2 billion ratepayer investment during the 2006-2008 program cycles if the savings goals are met is expected to range from 107% to 132% under the various shared-savings proposals before us. Hence, there is no question that, under any of the parties' proposals, achievement of the Commission's 2006-2008 goals for energy efficiency savings is expected to produce an extraordinary monetary return to ratepayers. And in the process, such achievement will create an unprecedented level of net resource benefits to all Californians-- on the order of $2.7 billion.

It is within this context that we consider one of the fundamental, and most controversial, issues in this proceeding: What level of earnings potential under the risk/reward incentive mechanism should be adopted to ensure that this type of return to ratepayers on their investment, and associated net resource benefits to all Californians, are achieved or surpassed? In the following sections, we summarize parties' positions on the level of earnings potential and associated shared-savings rates for the risk/reward incentive mechanism.

7.2.1. PG&E, SCE, SDG&E and SoCalGas ("The Utilities")

The utilities argue that supply-side comparability provides a relevant benchmark for establishing the earnings potential and shared-savings rate(s), pointing to the language of the National Action Plan for Energy Efficiency, California's Energy Action Plan, the Energy Policy Act, recent studies and prior Commission decisions as clear and strong policy support for this benchmark. In their view, these policies are based on the common sense understanding of what it takes to ensure the sustained commitment of utility management to meet and exceed the Commission's savings goals with cost-effective energy efficiency. They argue that this takes a potential stream of earnings from energy efficiency that both the investment community and utility management will view as a comparable earnings opportunity relative to the utility's core business of creating and operating reliable energy infrastructure.

In addition to the numerical results of their comparable earnings calculations, the utilities testified that they took other factors into account in developing their proposed shared-savings rate(s). In particular, they considered what dollar level of potential earnings for their company would sustain long-term management commitment to energy efficiency at all management levels and across all utility departments. They also considered which sharing rates at different levels of performance would produce an equitable allocation of net benefits between shareholders and customers.⁴⁸

PG&E also evaluated its proposed shared-savings rate from the standpoint of incentive regulation theory, exploring the fundamental trade-off between the policy objectives of protecting ratepayers against excessive utility profits and ensuring the delivery of effective results and reduced costs by the utility. In its direct testimony, PG&E describes how "sliding scale" or "shared-savings" regulation represents a hybrid form of regulation used widely by regulators to balance these objectives. PG&E concludes that the magnitude of its proposed shared-savings rate is consistent with hybrid regulatory schemes considered to be reasonable based on a review of the economic literature.⁴⁹

Using these considerations, the utilities propose a multi-tiered earnings rate structure for the risk/reward incentive mechanism. The highest shared-savings rate applies when the utility meets or exceeds the savings goals, ranging from 20% to 30% under the utility proposal. Lower shared-savings rates apply when performance is below the savings goals, but at or above the MPS. Utility proposals for these lower rates range from 10% to 15%. (See Attachments 3 and 4.)

7.2.2. NRDC

NRDC concurs with the utilities that the risk/reward incentive mechanism should consider comparability with returns the utilities are currently allowed on investments in supply-side resources. In NRDC's view, consideration of supply-side comparability is reasonable because utility portfolio managers have the option of directing resources and personnel to different types of resources. Without energy efficiency incentives, NRDC argues that the utilities may be more inclined to devote those resources to supply-side options, for which their shareholders earn a return. However, as discussed below, NRDC also believes that supply-side comparability should be viewed as only one benchmark that the Commission should consider.

In arriving at its estimate of comparable supply-side earnings, NRDC takes a somewhat different approach than the utilities. NRDC first calculates a comparable supply-side earnings rate, and then applies that rate to the level of ratepayer investment in energy efficiency. The utilities, on the other hand, first establish the level of supply-side resources (e.g., steel-in-the-ground investments/power purchases) needed to achieve the savings produced by the energy efficiency portfolio, and then derive the earnings foregone from that level of supply-side procurement. All other things being equal, NRDC's approach results in a lower level of earnings foregone, since the energy efficiency portfolio (in order to be authorized for funding) produces a comparable level of savings at lower costs than the avoided supply-side resources.

More specifically, NRDC first calculates an average pre-tax return on capital of 14.52% based on the Commission's approved cost of capital in D.05-12-043 and a 40.8% tax rate. NRDC then calculates the net present value of the total earnings on a power plant, assuming straight-line depreciation of the investment over 12 years. NRDC's calculations produce an effective earnings rate equal to 54% of the original capital investment.⁵⁰

NRDC does not support the utilities' position that additional earnings should be imputed for the "debt equivalence" of power purchases, and therefore assumes an effective earnings rate for power purchases of zero. Weighting the earnings rates for utility-build and utility-buy by their respective proportions of the utility portfolio (assumed to be a 50-50 split), NRDC calculates a 27% comparable earnings rate. To produce illustrative foregone earnings, NRDC multiplies this earnings rate by the total energy efficiency 2006-2008 portfolio costs that NRDC believes should be included in the PEB. This calculation produces a comparable earnings estimate of $497 million for all four utilities combined.⁵¹

NRDC recommends that the Commission also consider the level of performance (% of goal achievement) for which this level of earnings should be awarded. In NRDC's view, comparable supply-side earnings should only be fully awarded at a level of excellent performance well above the forecasted level of performance. Accordingly, NRDC recommends that at 100% of goal achievement, the utilities share 12% of the net benefits, which corresponds to earnings of $323 million for all four utilities combined. Beyond the MPS and until 100% achievement of goals, NRDC proposes an earnings rate half that large (or 6%). In NRDC's view, these earnings rates will provide a level of reward that is material to the utilities, while also requiring them to stretch significantly to achieve excellent performance beyond the savings goals and forecasted PEB.⁵²

7.2.3. DRA

In DRA's view, past policies of establishing energy efficiency sharing rates based on calculations of foregone supply-side earnings lack any factual predicate.⁵³ DRA argues that such an approach erroneously assumes that shareholders are harmed through the loss of shareholder earnings when energy efficiency programs displace utility supply-side investments. DRA contends that this is not the case, based on fundamentals of financial and economic theory.

More specifically, since utility shareholders do not actually invest in ratepayer-funded energy efficiency programs, DRA argues that shareholders retain the option of investing the money they do not invest in utility supply-side resources elsewhere, i.e., in an alternative investment of comparable risk in the marketplace. And since utility shareholders can earn a return that is "presumptively equal to the utility's authorized cost of capital,"⁵⁴ DRA concludes that the earnings rate for energy efficiency programs that achieves supply-side comparability is effectively zero.⁵⁵

Moreover, DRA contends that establishing the incentive levels proposed by the utilities represents an excessive amount of compensation, based on historic data. In particular, DRA points to the shareholder performance incentive for PY2000 and PY2001 that was capped at 7% of the program budget. DRA claims that because the utilities delivered superior results under this incentive mechanism there is no justification for the higher incentive rates proposed in this proceeding.

For these reasons, DRA rejects the comparable earnings approach and calculations proposed by NRDC and the utilities and recommends a benchmark based on a salary-based bonus system. DRA concludes that a 3% sharing rate (at 100% goal achievement) would be comparable to this benchmark.

More specifically, in its September 8, 2006 filing, DRA presents calculations that it claims shows how a sharing rate of 3% (producing $81 million in earnings for all utilities combined) is in line with the level of management fees that would be paid to mutual fund managers if they managed a fund equal in value to the 2006-2008 portfolio budget.⁵⁶ For this calculation DRA assumes that salaries for energy efficiency program staff and contractors comprise 25% to 30% of the portfolio budget, and a salary-based bonus scale would range from 3% to 15% for average to exemplary performance. For a three-year portfolio of $2 million, DRA calculates that the performance-based incentives would then range between $15 million to $90 million for average to exemplary performance (or 0.75% to 4.5% of the $2 million portfolio budget on a pre-tax basis.)

DRA also presents a Managerial Bonus model in its direct testimony to support its earnings rate proposal. As DRA explains, this model relies on two assumptions to obtain the basic incentive level (3% of PEB) at 100% of goals for the three-year program cycle: labor costs as a percentage of energy efficiency program budget and a salary-based bonus rate that should motivate superior performance.

Using data from 2006-2008 budgets, 2006 actual program costs and additional data on managerial compensation, DRA calculates that if all the utilities reached 100% savings goals and the incentive earnings were returned to energy efficiency program staff, each staff member would a bonus of approximately 35% of their base salary. Using the utility's compensation surveys included as part of each utility's general rate case filing, DRA also compared the results of its proposal to the actual bonus rates (total cash compensation to base salaries) paid in the survey years, which include data on the utilities and comparable companies. DRA concludes that an average incentive rate of 35% is greater than the average bonuses paid to 1) employees in the manager/supervisor category for the utilities and those comparable companies included in the utility surveys and 2) a weighted average of all managers, including executives for the utilities.⁵⁷

Based on these results, DRA concludes that its proposal to establish the shared-savings rate at 3% of PEB when 100% of the goals are met will motivate the utilities towards achieving or exceeding the Commission's energy efficiency goals, and at the lowest cost to ratepayers. This translates to earnings of $81 million over the 3-year program cycle for all four utilities, assuming that they reach 100% of their savings goals.⁵⁸ DRA recommends that for performance between the MPS and 100% goal achievement, the 3% be reduced by half, to 1.5%. DRA also recommends a higher earnings rate of 3.5% if the utility achieves over 125% of the Commission's kW goals.

Finally, DRA reviewed and compared energy efficiency incentive levels in nine other states with those being proposed in this proceeding, based on a 2006 survey by the American Council for an Energy-Efficient Economy (ACEEE). DRA concludes from this information that the incentives proposed by the utilities exceed every existing incentive program by a substantial margin. While its proposal is on the low end of the spectrum, DRA maintains that the size of California's energy efficiency programs should warrant a lower incentive rate.⁵⁹

7.2.4. TURN

TURN's general position is that supply-side comparability does not work in the context of establishing a shared-savings rate for ratepayer-funded energy efficiency, and that the Commission should establish the level of shared savings based on other considerations, as discussed further below.⁶⁰ However, if the Commission were to adopt a benchmark based on supply-side comparability, a position that TURN strongly opposes, TURN contends that the utilities' calculations would need to be reduced to reflect "alternative uses of funds." That is, TURN argues that the utilities' analysis of comparable supply-side earnings ignores the fact that a utility does not hide the money that it doesn't spend on supply-side resources "under the mattress." ⁶¹ Instead, TURN argues that the utility has alternative uses for these funds that create substantial shareholder value.

To illustrate this point, TURN describes what the utility could do with shareholder funds if they are not needed for supply-side investments. If the utility did not have enough cash for these investments in the first place, TURN posits that the utility would raise capital in the markets by selling new shares of stock (in the case of California utilities, through its holding company). TURN concludes that if the utility can avoid the issuance of new stock due to expensed energy efficiency, then its earnings-per-share will be higher, all other things being equal. Therefore, even though earnings per share increase when a utility undertakes a supply-side investment (because the utility earns a return on equity and the numerator increases), the resulting increase in earnings per share is diluted because the utility would have had to issue more shares (increase the denominator) to raise the investment capital for these projects.

Using PG&E's 2006 year-end values for earnings, number of shares, book value and stock price, TURN calculates that this dilution factor would reduce the positive impact of new supply-side investments on earnings-per-share on the order of 57%. Therefore, TURN suggests that PG&E's calculations of foregone earnings should be reduced by 57% to reflect this earnings-per-share dilution, assuming that PG&E would have sold additional shares of stock to finance the supply-side resources that are no longer needed.⁶²

Where the utility would have adequate equity to finance such investments, TURN describes the following alternative investments available to it: 1) reducing short-term borrowing or increase short-term investments, 2) paying down debt, 3) accelerating replacements of aging equipment or 4) paying dividends to the holding company. TURN argues that investment alternative #1 represents a relatively unprofitable uses of funds and would only be undertaken over the very short-term. With respect to alternative #2, TURN observes that paying down debt is better than investing in a money-market fund (at about a 4% after-tax rate), but is not an extremely profitable use of cash in terms of quantifiable benefits. With regard to alternative #3, TURN concurs with DRA's assessment that this use of funds would result in zero "lost" equity return, since the investments would receive the same return on equity as the supply-side resources no longer needed.

For investment alternative #4, TURN describes four options available to the holding company for the funds it receives (in the form of dividends) from the regulated utility. The holding company could: 1) pay dividends to shareholders, 2) use the additional dividends from the utility as equity to invest in unregulated projects, 3) use the additional dividends from the utility as equity for the pursuit of mergers and acquisitions either of regulated or unregulated entities, or 4) buy back stock with the money. For the first three of these options, TURN concludes that, from the point of view of the shareholders, there is no lost equity return. According to TURN, this is because shareholders (under option #1) have the ability to reinvest the money in assets that presumably have the same risk-adjusted cost of capital as the utility (otherwise they would not buy them). Under options #2 and #3, shareholders also have no lost equity return because the holding company invests in projects that provide good if not better risk-adjusted returns than California regulated utility projects (otherwise they would not make them).

With respect to option #4, TURN asserts that stock buybacks benefit the utility and its shareholders in several ways, including by increasing earnings per share (all other things being equal), and by providing money to shareholders who choose voluntarily to sell in order to make alternative investments that earn the same amount or more on a risk-adjusted basis. TURN calculates that if PG&E's holding company took the cash no longer needed for supply-side investments, and instead bought back shares of its stock, the "foregone earnings" from not building those supply-side resources would be 45% less than PG&E's comparable earnings calculations suggest, all other things being equal.⁶³

TURN concludes that the utilities' comparable earnings analysis is flawed because it ignores elementary principles of finance, accounting and economics by not considering potential uses of equity that would not be invested in energy efficiency. TURN also agrees with DRA that the theoretically "right" answer to a calculation of foregone shareholder earnings is "probably zero." ⁶⁴ Nonetheless, to provide some boundaries on the utility's calculations, TURN recalculates PG&E's numbers assuming that the Commission establishes a return on equity that is actually greater than the cost of equity.⁶⁵ Under these circumstances, TURN suggests that foregone earnings would be non-zero because shareholders would not be able to find a return on their equity for comparable risk investments as high as the foregone utility supply-side investments. To quantify this scenario, TURN recalculates PG&E's numbers by subtracting PG&E's authorized return on equity (11.35%) by the 8.5% Standard and Poor (S&P) 500 return.⁶⁶ In addition, TURN argues that the utilities have overstated comparable earnings by imputing debt equivalence for power purchases, a practice that TURN contends the Commission had not approved in establishing the utility's cost of capital. TURN removes the debt equivalence calculation from PG&E's numbers and makes other adjustments and corrections to PG&E's calculations. The result is what TURN characterizes as a maximum comparable pre-tax earnings rate of 3.4%.

In sum, TURN concludes that a supply-side comparability analysis does not justify anything higher than a 3.4% shared savings rate, and is probably zero. Moreover, TURN argues that there is no evidence on the record to demonstrate a correlation between high incentive levels based on supply-side comparability and better utility performance, and in fact asserts that historical data suggests quite the opposite. Therefore, TURN recommends that the Commission evaluate the appropriate level of sharing of energy efficiency net benefits from a different perspective altogether.⁶⁷

TURN submits that the appropriate perspective is one that views the energy efficiency risk/reward incentive mechanism as analogous to performance-based ratemaking (PBR) mechanisms that the Commission has established in the past to either 1) prevent harmful consequences from cost-cutting measures, or 2) reward shareholders for cutting costs. TURN specifically refers to the following PBR mechanisms adopted by the Commission:⁶⁸

· The nuclear and coal power plant performance incentive mechanisms ("target capacity factor" mechanism) adopted in the early 1980s for SCE;

· The Core Procurement Incentive Mechanism (CPIM) and Gas Cost Incentive Mechanism adopted for PG&E and SoCalGas respectively, beginning in 1994;

· SCE's base rate PBR revenue sharing mechanism adopted in 1996 until 2003 for sharing profits or losses from transmission and distribution operations; and

· The Employee Safety and Distribution Reliability Performance Incentive Mechanism adopted for SCE beginning in 2003.

Based on its review of the sharing rates and caps for these other targeted incentive mechanisms, TURN developed shared-savings rates of 2-3% in this proceeding.⁶⁹ More specifically, TURN proposes a sharing rate of 2% between the MPS and 100% of goal achievement, 2.5% if the goals are achieved and exceeded, and a higher rate (3%) rate if the utility exceeds 125% of the kW goal.

7.2.5. CE Council

CE Council attended the Phase 1 workshops, filed a proposal for a specific risk/reward incentive mechanism and an opening brief, but did not participate in evidentiary hearings.⁷⁰ Therefore, CE Council did not present sworn testimony, subject to cross-examination, on the position it has taken with respect to the level of shareholder earnings and earnings rates. In its comments, CE Council argues that there is very limited risk to the utilities under energy efficiency programs, and therefore the earnings should be commensurately limited. In particular, CE Council believes that the savings goals should not be particularly difficult to meet and, based on past experience, the risk of penalties is small.⁷¹ Moreover, CE Council argues that the utilities benefit from energy efficiency in intangible ways, such as the benefits associated with brand labeling. CE Council also supports and reiterates several of the arguments presented by DRA and TURN in their workshop comments. CE Council suggests that a better measure of appropriate earnings potential is what has been effective in other states.

Based on these considerations, CE Council recommends sharing rates increasing from 2% to 3% under a two-tier structure.⁷² Like DRA and TURN, CE Council also recommends a higher earnings rate tied to achieving more than 125% of the kW goals.

7.2.6. Position of CLECA

CLECA did not participate in the workshop process or present a specific design proposal for the risk/reward mechanism in post-workshop filings. However, CLECA did sponsor a witness and testimony during evidentiary hearings on the overall level of potential earnings.⁷³

In that testimony, CLECA contends that the potential earnings proposed by the utilities and NRDC are much too large. CLECA argues that the only risk to shareholders under the pending proposals is if they fail to successfully pursue cost-effective energy efficiency portfolios, which CLECA asserts would be a low risk given the projected benefit-cost ratios for these programs. In addition, CLECA argues that energy efficiency expenditures are not depriving the utilities of earnings opportunities and, in fact, provide them with significant institutional marketing benefits without the need for any shareholder investment. For these reasons, and based on a review of DRA and TURN's submittals, CLECA supports a shared-savings rate in the 3-5% range.⁷⁴

7.3. Discussion

TURN, DRA, CE Council and CLECA ask that we reject any consideration of a supply-side comparable earnings analysis, unless we dramatically reduce the resulting numbers. For the reasons discussed below, we find their arguments for rejecting this analysis unpersuasive, and discuss the shortcomings of their proposed alternative benchmarks. We conclude that supply side comparability should be one, among other relevant considerations, in establishing the earnings potential under the incentive mechanism we adopt today.

7.3.1. Supply-Side Comparability: History and Purpose

As discussed further below, some parties contend that the purpose and regulatory context that gave rise to using supply-side earnings comparability as a benchmark for energy efficiency incentives in the past has fundamentally changed. Before turning to those specific arguments, we present a brief overview of that history.⁷⁵

The concept of providing utilities with an opportunity to earn from energy efficiency and other demand-side management (DSM)⁷⁶ efforts was developed in the late 1980s in response to the Commission's stated need to take a fresh look at the role of DSM in utility resource procurement. This was a time prior to electric industry restructuring when California's investor-owned utilities met their customers' energy needs by acquiring and delivering energy resources on their behalf, as they do again today. The Commission embarked on a proceeding to create positive financial incentives that would produce a "win-win" alignment of ratepayer and shareholder interests in achieving least-cost, integrated resource planning objectives.⁷⁷

After conducting several years of experimental programs on energy efficiency incentive mechanisms, in 1993 the Commission reviewed the results of the experiments and concluded:

"On balance, there are disincentives to DSM created by both regulation and the private profit-making nature of the firm that limit utility shareholders and management's interest in pursuing all practicable, cost-effective and reliable DSM." (D.93-09-078, Conclusion of Law 1.)

"Under the current regulatory framework, DSM shareholder incentives are necessary and appropriate to increase the private value of DSM to a utility by bringing that value more in line with its social value." (Ibid., Conclusion of Law 3.)

"Regulatory mandates and rate of return penalties do not create potential "win-win" situations for shareholders and ratepayers. Rather they create a "ratepayers win or else shareholders lose" approach to DSM regulation." (Ibid., Finding of Fact 11.)

More specifically, the Commission identified the following financial and regulatory biases against energy efficiency (or in favor of supply-side resources), and concluded that shareholder incentives were an effective way to address them:⁷⁸

(1) Utilities only earn on supply-side investments under current regulatory practices absent energy efficiency incentives;

(2) Cost-effective energy efficiency investments will increase rates in the short-term, even though it will minimize revenue requirements and customer bills over time.

The Commission then proceeded to evaluate what level of earnings opportunity would be "sufficient (and not too much) to offset these biases" in a lengthy evidentiary process to which most of the same parties to this proceeding also participated, including TURN and DRA.⁷⁹ To this end, the Commission directed parties to calculate the "effective earnings rate" associated with supply-side resources deferred or avoided by DSM investments. Parties also presented other proposals for the Commission to use instead of or in conjunction with its consideration of supply-side earnings comparability. However, the Commission rejected those proposals for various reasons.

For example, the Commission rejected recommendations to rely on historical evidence of utility management interest in establishing the earnings potential. In doing so, the Commission found that levels of earnings achieved in the past would not be accurate indicators of the level of earnings opportunity that would be needed to overcome disincentives to DSM in the future. The Commission also rejected DRA's recommendation to reduce the supply-side effective earnings rate by 40-50% based on its assertions that utility management was biased in favor of demand-side resources over supply-side resources, contrary to the Commission's own findings. In addition, the Commission rejected TURN's position that instead of supply-side comparability, the utility shareholders were entitled only to a minimal management fee for managing ratepayer-funded energy efficiency.⁸⁰

Instead, the Commission elected to develop calculations of supply-side comparability as a general benchmark, and then assess the appropriate level of target earnings within the context of that benchmark and the incentive mechanism being proposed, taking into consideration the relative risks and rewards associated with supply-and demand-side alternatives. In addition to the differences in risk due to who funds the initial investment, the Commission identified other relative risks to be considered, such as how shareholder earnings vary with project performance and who bears the risk of non-cost-effective investments. The Commission described these different (and changing) risk/reward profiles for demand- and supply-side resources in D.94-10-059. In doing so, the Commission considered the relative risks and rewards to shareholders and ratepayers under traditional cost-of-service and performance-based ratemaking that was in place for supply-side resources, or being contemplated in the near future. Not surprisingly, the Commission found that comparisons between the earnings opportunity from DSM and supply-side resources were difficult to make, given the differing performance, earnings and investment characteristics of demand- and supply-side resources.⁸¹

On balance, taking supply-side comparability and other factors into consideration, the Commission adopted an earnings rate at target performance on the lower end of the supply-side comparability analysis presented in the proceeding. "Target" performance (and the MPS) under that earlier shared-savings mechanism was based on the projected level of net benefits (PEB) for the energy efficiency portfolio, as there were no established goals for kW, kWh or therm savings goals at that time.

More specifically, parties to the proceeding presented a range of 26% to 52% for the effective earnings rate associated with supply-side resources deferred or avoided by DSM investments, based on the capital costs of avoided generation. This corresponded to target earnings levels of $77 to $153 million for all four utilities combined. The Commission adopted a shared-savings rate of 30%, which translated into potential shareholder earnings of $88.7 million out of the expected $295 million in net benefits at target performance (for all utilities combined).

In this proceeding, DRA argues (and TURN concurs) that a supply-side comparability analysis should produce a comparable return of zero because shareholders do not "lose" any earnings when the utility undertakes energy efficiency. In fact, no party argues that utility shareholders are made worse off by energy efficiency, since everyone acknowledges that investors can put their investment dollars elsewhere and earn a return comparable to the one they would have earned on the displaced supply-side investments.⁸² However, the purpose of a supply-side comparability analysis is not, and has never been, to prove or disprove the tautology of zero foregone shareholder earnings posed by DRA and TURN in this proceeding.

Instead, as discussed above, a comparable earnings analysis provides a numerical benchmark for addressing the very heart of the bias that stands in the way of successful implementation of California's energy policies by the utilities we regulate: Utility investors are attracted by opportunities to earn returns, and absent energy efficiency incentives, utilities only earn on supply-side investments. Recognition of this disincentive to energy efficiency has been expressed in the federal Energy Policy Act of 1992--a statute that is still in effect-and California's 2003 Energy Action Plan.

More specifically, the Energy Policy Act of 1992 requires state commissions to consider the following standard:

"The rates allowed to be charged by a State regulated electric utility shall be such that the utility's investment in and expenditures for energy consideration, energy efficiency, and other demand side management measures are at least as profitable, giving appropriate consideration to income lost from reduced sales due to investments in and expenditures for conservation and efficiency, as its investments in and expenditures for construction of new generation, transmission and distribution equipment."⁸³

California's 2003 Energy Action Plan identified the following action as one of "critical importance" for optimizing energy conservation and resource efficiency:

"Provide utilities with demand response and energy efficiency investment rewards comparable to the return on investment in new power and transmission projects."⁸⁴

In 2006, over the objections of both TURN and DRA, this Commission reiterated the need to address this barrier in the context of its adopted Procurement Incentive Framework:

"...the record in this proceeding persuades us that financial incentives for preferred resources are worthwhile to pursue in conjunction with a [greenhouse gas] cap. Doing so is entirely consistent with the policies articulated in prior Commission decisions, as well as with the action items outlined in the [Energy Action Plan] (I and II). In particular, those policies articulate the need to bring energy efficiency and demand-side resource investments in line with traditional supply-side resources when it comes to the opportunities to earn returns on those investments."⁸⁵

No party to this proceeding presents evidence to dispute that this fundamental bias exists in today's regulatory environment, now that investor-owned utilities have been returned to role of managing both supply-and demand-side resource procurement on behalf of their ratepayers. As discussed further below, no party presents convincing evidence to overturn the finding we made in 1993 concerning the short-term rate impacts associated with energy efficiency that also serve to bias utilities towards supply-side options.

More generally, as NRDC and others point out, a comparable earnings benchmark recognizes that utilities as portfolio managers make day-to-day decisions on how to direct their resources and personnel that regulators cannot directly control or mandate. Without an energy efficiency incentive, given the focus of investors and utility management on increasing shareholder value, utilities will on balance be more inclined to devote scarce resources to procurements on which they will earn a return, and not on meeting or exceeding the Commission's energy efficiency goals, or maximizing ratepayer net benefits in the process. As one witness describes:

"Senior management has the job of assigning limited resources including human capital and senior management attention on doing some things with a great degree of attention and other things kind of as business as usual....By having a payout that's much much less than what we would earn on supply side...sends the message that is, quite frankly, less important, and perhaps you shouldn't invest as much attention and resources in that area as you would in other areas that your...investors are going to be demanding from the business enterprise....What we are attempting to do in the state of California-and I think it's all something we all ought to be very proud of-is to treat energy efficiency not just in terms of words, not just in terms of policy, but hard and fast investment dollars, resources and attention as our primary resource for the state of California...as an alternative to supply-side resources...that satisfies the state energy policy as expressed in California Energy Action Plan...and...is the least cost and quickest way to address the global warming issues that this country faces and that this planet faces."⁸⁶

We agree. Therefore, knowing how much investors would have earned on supply-side procurements, if not for energy efficiency, is useful information: It helps us to consider, among other factors, what level of earnings potential will be sufficient to overcome the biases in favor of supply-side resource procurement and achieve our policy objectives for energy efficiency.

In arguing against supply-side comparability, DRA and TURN generally assert that financial and regulatory biases against energy efficiency are significantly less today than in the past. However, upon close inspection of the record, these parties do not actually refute the fact that utilities continue to earn only on supply-side investments under current regulatory practices, absent energy efficiency incentives. Instead, their comments suggest that this is not really a bias at all, since it is shareholders who put up the initial capital for supply-side investments, whereas ratepayers fund demand-side expenditures without requiring up-front capital from investors.⁸⁷ TURN and DRA made this same argument in the proceedings leading up to D.94-10-059, where we rejected it as a rationale for either discontinuing shareholder incentives altogether or for reducing the earnings potential to minimal levels. As we concluded in D.94-10-059, in addition to who funds the initial investment, there are multiple dimensions to relative risk to consider including: (1) how shareholder earnings vary with project performance and (2) who bears the risk of non cost-effective investments.⁸⁸ Moreover, these considerations do not alter the fact that under cost-of-service regulation utilities earn on "steel-in-the-ground" investments, and not on the successful procurement of energy efficiency.⁸⁹

More importantly, in considering what is fair to ratepayers, we observe that ratepayers "invest" in both supply-side and energy efficiency resources, irrespective of who puts up the initial capital. The only difference is that for steel-in-the-ground investments (generation, transmission, distribution) ratepayers have to pay not only the cost of the facilities, but also the financing costs (debt service and return-on-equity, and associated taxes) to compensate those that put up the initial capital. In contrast, since energy efficiency expenditures are "expensed" and reflected in rates immediately, energy efficiency saves ratepayers substantial financing costs. Those cost savings are magnified because a dollar of energy efficiency can displace far more than a dollar of supply-side investment to meet the same GWh, MW and MTherm energy needs.⁹⁰ Hence, the critical question is not "who puts up the capital" for energy efficiency, but rather, "how can we ensure that the potential return on ratepayers' investment in energy efficiency is actually realized."

DRA also argues that a major disincentive to energy efficiency has been removed with the "decoupling" of sales from revenue requirements in California.⁹¹ As described in the record, this disincentive arises due to forecasting errors when utility base rates are set to recover the utility's fixed-cost revenue requirements. The rate charged to customers is derived from the authorized revenue requirement divided by forecasted sales. Hence, if actual sales fall lower than the forecasted levels (because of greater than expected energy efficiency installations, for example), then the rates charged do not recover the utility's fixed costs. Therefore, without any decoupling of revenues from sales or other approaches to address this forecasting risk, the utility faces a strong disincentive to reduce sales through energy efficiency beyond the forecast (and, in fact, a strong incentive to promote sales volumes before the next general rate case.) Decoupling sets up a mechanism to track the difference between actual and forecasted base rate revenues, whereby overcollections are refunded to ratepayers and undercollections are recovered in subsequent rate adjustments.⁹²

However, decoupling merely eliminates a financial penalty for pursuing energy efficiency-it does not make it the preferred resource from a shareholder, investment community or utility management perspective.⁹³ Moreover, decoupling in the form of an Electric Revenue Adjustment Mechanism (ERAM) was in effect when we considered and adopted a risk/reward incentive mechanism taking into account supply-side earnings comparability in 1994. In fact, decoupling and other methods to address this "lost revenue" deterrent to energy efficiency were in place in several other states with energy efficiency programs prior to electric industry restructuring, dismantled in conjunction with industry restructuring, and then resurrected in 2001 in California and in some other states in more recent years.⁹⁴ As DRA itself notes: "The Commission eliminated the ERAM in 1996 with the advent of deregulation, but following the energy crisis of 2000-2001, California returned to its policy of decoupling sales from revenues..."⁹⁵ Our reinstatement of a decoupling mechanism is not a reason to ignore our earlier findings on the existence of significant disincentives to energy efficiency or our policy determinations since the energy crisis to address them.

In fact, the only specific change in regulatory circumstances that TURN and DRA identify relates to the manner in which costs for energy efficiency are funded and recovered through rates. They point out (and no party disputes) that since 1996 energy efficiency has been funded through a "non-bypassable" public goods charge and procurement dollars earmarked for energy efficiency and fully recovered immediately from ratepayers.⁹⁶ TURN concludes that this change has addressed a "critical disincentive" to utility energy spending, referring to the Commission's finding that DSM expenditures funded in rates "have initial rate impacts that pose competitive risks to the utility in the form of potential bypass." ⁹⁷ DRA's cross-examination of utility witnesses similarly suggests that it too believes that this disincentive has been addressed by the changes in how energy efficiency is funded, relative to the cost recovery methods in place in 1993 and 1994.⁹⁸

We do not find this conclusion to be logical or supported by the record. Funding of energy efficiency through a non-bypassable charge on distribution rates does not change the fact that there are bypass options available in California, that expenditures on cost-effective energy efficiency results in initial rate increases, and that higher rates increase the risk of bypass. California's investor-owned utilities face a risk of bypass today through community choice aggregation and municipalization proposals and the Commission is considering issues related to the reinstatement of direct access, another form of bypass, in R.07-05-025. As explained by one of SCE's witnesses, spending on energy efficiency is different from supply-side resources in terms of short-term rate impacts. This is because energy efficiency (unlike supply-side resource additions) reduces the number of kWh, kW and therms over which fixed costs are spread.⁹⁹ No party has convincingly refuted these facts. Therefore, we stand by the following finding we made in 1993: "Even though energy efficiency may have a higher ratepayer and societal value, other options (e.g., inter-utility power purchases) may have a higher private value to utilities because they generally do not initially increase rates."¹⁰⁰

Our finding that fundamental disincentives to energy efficiency persist today is corroborated by two recent national studies discussed in testimony and submitted into the record. The July 2006 National Action Plan for Energy Efficiency, facilitated by the U.S. Department of Energy and U.S. Environmental Protection Agency, describes the bias against energy efficiency in this way:

"When utilities invest in hard assets, they depreciate these costs over the useful lives of the assets. Consumers pay a return on investment for the un-depreciated balance of costs not yet recovered, which spreads the rate effect of the asset over time. Utilities often do not have any opportunity to earn a return on energy efficiency spending, as they do with hard assets."¹⁰¹

In March 2007, the United States Department of Energy (DOE) issued its report to the United States Congress pursuant to Section 139 of the Energy Policy Act of 2005. That report identifies "the ability to earn a rate of return on physical assets" as a disincentive facing investor-owned utilities to "implementing energy efficiency programs, rather than investing in physical assets such as power plants and transmission lines."¹⁰²

The Lawrence Berkeley National Laboratory report cited by both DRA and PG&E specifically noted--similar to the 2007 DOE Report-that one of the two key "hidden costs"" that must be mitigated for a "fair incentive"

"...consists of the opportunity cost associated with utility activities foregone by pursuit of DSM programs...includ[ing]...foregone earnings from alternative supply-side investments that would have been made in the absence of a DSM program. We believe that these costs increase with the scale of DSM programs and, therefore will be greatest in the later phases of DSM program implementation. This second type of hidden costs, while still difficult to measure, is more well defined....The primary analytic issue is determining earnings comparable to those that would have been earned through the acquisition of resources in lieu of DSM."¹⁰³

In sum, we conclude that the fundamental regulatory and financial biases we identified in D.93-09-078 also exist under the current regulatory framework, in which utilities have returned to their traditional role as resource portfolio managers. Within the context of this history and purpose for supply-side comparability, we address the various issues raised by the parties to this proceeding on the appropriate level of earnings potential under an energy efficiency incentive mechanism. We start with a discussion of alternative benchmarks presented for our consideration.

7.3.2. Consideration of Alternative Benchmarks

As discussed in Sections 6.2.3 and 6.2.4 above, TURN and DRA developed their shared-savings rates based on considerations or benchmarks other than supply-side comparability. Below, we discuss these alternative approaches ("benchmarks") and our findings and conclusions.

7.3.2.1. Mutual Fund Management Fees

Initially in this proceeding, DRA explained that its proposed 3% sharing rate was based on calculations of the range of management fees earned by mutual fund portfolio managers. DRA asserts that such fees range from $0.75 to $4.50 per $100 dollars, and then notes that its 3% sharing rate falls within that range.¹⁰⁴ DRA provides no references or evidence to support its assertion that this is actually what mutual fund managers charge in the way of fees. Moreover DRA makes an "apples-to-oranges" comparison here, because fund management fees are based on the total portfolio value, not the "profits" or net benefits of the portfolio. In contrast, the shared-savings rate that DRA proffers as being within the range of those fees represents a percentage of net benefits of the energy efficiency portfolio, not its total value.¹⁰⁵ As PG&E showed in it rebuttal testimony, if you assume (as TURN does) that the typical return from a portfolio is 8.5%, then DRA's calculations when presented as a percentage of net benefits would yield a comparable sharing rate on the order of 35% off the flow of earnings, and not 3%.¹⁰⁶

As we observed in D.94-10-059, when we rejected a similar proposal, we surmise that mutual fund managers would demand considerably more than the single-digit fees DRA calculates if faced with the risk/reward profile associated with today's adopted incentive mechanism. That is, if they earned only in proportion to portfolio gains, as measured over a multi-year period, and if they were also required to pay financial penalties for missing targets and for losses on their clients' investment.¹⁰⁷ Here too, the record does not support adopting this recommendation.

7.3.2.2. Management Bonus Model

To further support its post-workshop recommendation for a 3% sharing rate, DRA presented a Management Bonus Model in its prepared testimony. Based on the results of this model, DRA claims that a 3% shared-savings rate translates into a 35% managerial bonus equivalent, which DRA concludes would be a sufficient financial incentive to motivate utility behavior under the risk/reward incentive mechanism. (See Section 6.2.3 above.)

However, as the record shows, this claim is premised on a calculation that includes only those employees dedicated exclusively to energy efficiency activities in the denominator. This leaves out employees in other departments that directly and indirectly support the development and implementation of energy efficiency programs, including the executives who made the broad policy decisions related to energy efficiency policy and shareholder returns. In particular, DRA's calculations exclude compensation to members of the Board of Directors, officers, senior managers, field personnel and other resource-planning and procurement staff who also have a duty to consider the opportunity to increase earnings in making their decisions about deploying resources. Moreover, the 3% bonus equivalent is calculated using "base" salaries alone, and does not consider other salary and non-salary benefits received by employees.¹⁰⁸ Because of these exclusions, DRA's model results (i.e., that $81 million in earnings over three years represents a 35% bonus equivalent) is substantially overstated.

We concur with NRDC, SCE and others that all levels of management and personnel throughout the company, and not just within the energy efficiency division, need to be motivated to view energy efficiency a core business activity in order to achieve the aggressive energy efficiency and environmental goals of the state.¹⁰⁹ DRA Witness Roberts appears to agree with this perspective, based on his testimony that DRA's proposal would be "large enough to motivate the entire organization."¹¹⁰ However, when DRA's calculations are corrected to actually reflect this perspective, the numbers that result do not resemble a 35% managerial bonus equivalent.

A simple calculation illustrates how DRA's model results would differ substantially if the denominator reflected the base pay for the entire company, as opposed to the limited group included in DRA's model. Using DRA's 35% figure, PG&E's incentive at 100% of energy efficiency goals would be $9.7 million per year.¹¹¹ Based on the data presented in DRA's testimony, PG&E's base pay for all employees in 2004 was $702.6 million.¹¹² Dividing $9.7 million by the 2004 base pay yields a company wide "bonus" rate of only 1.4 percent, rather than the 35% presented in DRA's testimony. This calculation does not correct for other salary and non-salary benefits that should appear in the denominator of DRA's calculation, which would lower the bonus equivalent further. However, it serves to generally illustrate that DRA's model substantially overstates the bonus equivalent of $81 million by limiting its focus on a small subset of the utility's employees.

In addition, DRA bases its model on the premise that the utilities' short-term bonus compensation plans are sufficient to motivate the investor-owned utilities of California to aggressively pursue energy efficiency, a premise that is unsupported in theory and unproven in practice.¹¹³ Moreover, when corrected for the limited scope of employees considered by DRA, the Managerial Bonus Model actually reveals why DRA's proposed 3% shared-savings rate would result in a "bonus" that would be virtually imperceptible. For these reasons, we do not adopt it as a benchmark for shared-savings.

7.3.2.3. Energy Efficiency Incentives Offered in Other States

DRA and CE Council argue that the incentives offered by other states present the appropriate measure or range that California's energy efficiency incentives should be compared to for benchmarking purposes.¹¹⁴ TURN also joins this argument in its opening brief.¹¹⁵

However, comparisons of the incentive levels offered in other states fail to address the characteristics of individual states that may make them have greater or lesser relevance for California policy makers.¹¹⁶ CE Council refers to earnings rates found in other venues and merely asserts that the administrative structures are comparable. Differences not supporting its position are neglected. For example, CE Council fails to discuss that the minimum performance threshold for the incentive mechanism in Massachusetts has been 70% and was only recently changed to 75%. CE Council also does not explore whether the MPS is applied against a savings goal that has been increased relative to Massachusetts's recent performance to the same extent as in California.¹¹⁷

Similarly, in assessing the nine other states' energy efficiency incentives presented in its testimony, DRA did nothing to evaluate numerous important factors that are essential to a valid comparison to California. Specifically, DRA did not consider the level of MW, GWh and MTherm goals (if any) established for the utilities in other states and whether these goals were established by utilities or regulatory/legislative organizations. DRA also did not consider whether verification efforts, if they were in place, were conducted ex post (post installation) and independently of the utility in question. DRA did not consider differences in retail sales, energy efficiency budgets and expenditure levels, or whether the investor-owned utilities in the other states had the option of investing in supply-side resources rather than energy efficiency programs.¹¹⁸ Nor did DRA know or consider whether any of these other states' incentive mechanisms also included financial penalties, as did all the proposals in this proceeding.¹¹⁹

A further important variable that the record shows is difficult to assess is to what degree the current regulatory and institutional structures for other states with energy efficiency incentives are indeed analogous to California's. For example, some of these states have restructured their electric markets, whereas others have not, and utilities in some states are given more responsibility for the delivery of energy efficiency resources than in others.¹²⁰ All this makes comparisons particularly difficult without far more information than DRA or CE Council provided for the record here. In fact, the nine states listed in DRA's testimony represent vastly different utilities, in different service areas, with different economic determinants of the power marketplace and the energy efficiency market there, as well as critical institutional differences.¹²¹

Importantly, the ACEEE report that DRA cites reviewed energy efficiency incentives after electric restructuring, during which time incentive rates for those states that still retained energy efficiency incentive mechanisms were observed to decline considerably.¹²² DRA's survey of other states' incentive rates reflects where most of those states have ended up after this decline. ¹²³ A survey of other state's energy efficiency incentives would have looked much different if DRA had considered the incentive rates in place prior to electric restructuring, when investor-owned utilities across the country managed resource portfolios as our California investor-owned utilities do again today. In fact, the survey conducted by Lawrence Berkeley National Laboratories prior to restructuring produced much higher range for the incentives in place in 1992 and those anticipated for the 1993-1994 period based on initiatives underway in ten states.

That survey shows incentives in the range of 8.2% to 50.3% as a percentage of program costs in 1993-1994, as compared to the ACEEE survey results of 3.3% to 15.3% (also as a percentage of program costs).¹²⁴ To put the results of the two surveys in the context of the earnings rate proposals in this proceeding, DRA's proposal represents incentives of 3.7% of program costs and the highest proposal for shared-savings in this proceeding (PG&E's) represents 24.5% of program costs (at 100% of goal achievement).¹²⁵

In addition to ignoring the relevance of electric restructuring history on the ACEEE survey results, DRA further asks that we ignore the higher end of the range of incentive levels that can be observed from that particular survey for Nevada, Arizona and Wisconsin, arguing that they are less appropriate points of comparison.¹²⁶ However, as PG&E notes, several key variables (including expected rate of population growth) make Nevada and Arizona potentially the most comparable, if indeed any state can be validly compared to California.¹²⁷ Moreover, DRA's assertion that because California's programs are more mature, less financial incentive is needed to improve performance is not founded in basic economic theory or logic. Where energy efficiency has been underway for some time, past achievements have generally pushed the utilities further up on the supply curve, as in California, thus increasing the level of difficulty of achieving future targets.¹²⁸

In sum, DRA and others ask that we benchmark earnings using the range of incentive levels adopted in other states since industry restructuring, and in doing so, to ignore the upper end of that range. For the reasons discussed above, this benchmarking approach is not reasonable, and we do not adopt it.

7.3.2.4. Non-DSM Performance-Based Incentive Mechanisms

As discussed in Section 6.2.4, TURN argues that non-DSM performance-based ratemaking (PBR) mechanisms adopted in California should be used to benchmark the earnings potential for energy efficiency. In its Opening Brief, TURN presents graphs depicting average and maximum dollar incentive awards received by utilities under various PBR mechanisms. TURN submits that its proposal for shared-savings is reasonable, noting that it produces an earnings level for PG&E that falls within the middle of these ranges.¹²⁹

We find that TURN's analysis is flawed, for several reasons. Most significantly, TURN restricts its analysis to just the absolute dollar amount of these non-DSM mechanisms, and never discusses what each mechanism is designed to achieve or the value of success to ratepayers.¹³⁰ However, each of them was developed in the context of the objectives for the particular mechanism, and each have very different risk/reward parameters. How can one, for example, compare a mechanism designed to establish penalties and earnings related to the average duration of customer outages (e.g., under SCE's performance PBR) with one that is intended to motivate the procurement of an unprecedented level of energy efficiency savings estimated to produce almost two billion dollars in net benefits to ratepayers?

In addition, TURN's comparison looks at the achieved results under the various non-DSM PBR mechanisms, whereas it is the potential results for energy efficiency that is established by the shared-savings rates and reflected in parties' proposals.

For example, in the only PG&E mechanism TURN mentions-the CPIM, or Core Procurement Incentive Mechanism-the objective is for PG&E to meet or beat the "market" on core natural gas costs. Outside a tolerance band, PG&E shares in the benefits of lower gas costs or pays for gas costs above that level. This mechanism replaces administrative review of the reasonableness of core gas costs. TURN is implicitly comparing all aspects of this mechanism with the challenge of significantly increasing energy efficiency savings on a sustained basis. However, as PG&E points out, the challenges may not be comparable, and the rewards may not all be in the explicit financial opportunity CPIM provides (e.g., the elimination of after-the-fact reasonableness reviews). In particular, TURN does not mention that if PG&E's purchased gas costs are less than the applicable benchmark, and are less than the tolerance deadband, PG&E gets to keep 25% of the savings, and customers get the other 75%.¹³¹

Similarly, while TURN focuses on what was actually earned under the Coal Plant and Nuclear Unit Incentive Procedures, they each provide for a 50% sharing of net benefits earned from successful plant operations with shareholders, which is well in excess of the sharing rates proposed by any party to this proceeding. Moreover, the "maximum" profit under SCE's PBR mechanism presented in TURN's graphs is significantly understated, since SCE could actually have earned under that mechanism almost three times the amount depicted in TURN's graph.¹³²

Nor is there any discussion about the maximum penalty provisions under these mechanisms, or the thresholds of performance established before penalties are imposed or rewards can be earned. As we discussed in D.94-10-059, and reiterate today, the potential for earnings under the energy efficiency performance incentive mechanism should take these other design factors into consideration. TURN's focus on the level of dollar rewards previously earned under non-DSM PBR mechanism completely ignores these considerations.¹³³

TURN's analysis is further flawed by the lack of reasonable criteria for deciding what PBR mechanisms to include in its analysis, and the apparent exclusion of ones that could be relevant. When asked what criterion or criteria it used to select the non-DSM PBR mechanisms it relied on in developing its incentive proposal, TURN responded that it "did not develop any specific criteria to include or exclude incentive mechanisms."¹³⁴ Interestingly, TURN did not include the Hazardous Substances Clean-up Mechanism adopted in 1993. A key feature of this mechanism is that insurance recoveries relating to these sites are shared by shareholders and ratepayers at a 30/70 sharing rate to incent the utility to aggressively pursue such upsides that benefit everyone.¹³⁵ Nor did TURN include SDG&E's base revenue PBR earnings in its analysis. Had it done so, TURN's analysis would have revealed that its proposal for potential earnings for SDG&E (approximately $6 million at 100% of goals) over a three-year period was only about 40% of what SDG&E actually earned under its base case PBR over the same timeframe.¹³⁶

For these reasons, we reject TURN's assessment that non-DSM PBR incentive levels provide a reasonable benchmark for establishing incentive levels in this proceeding.

7.3.2.5. Energy Efficiency Incentive Levels During Electric Restructuring and the Energy Crisis

TURN neglects to consider in any context the performance-based DSM incentive mechanism adopted in 1994 as a relevant PBR benchmark for its purposes, even though it was designed to address energy efficiency and procurement objectives similar (although not identical) to those articulated in this proceeding. TURN and DRA argue (with CE Council concurrence) that such consideration would not be meaningful because the utilities have shown no statistical correlation between the higher incentives offered prior to electric restructuring and improved performance.¹³⁷ Instead, these parties conclude that the lower incentive levels adopted by the Commission during electric industry restructuring and the energy crisis would serve as a more appropriate benchmark for earnings in this proceeding.

TURN and DRA make much of the fact that the utilities did not perform a statistical analysis to correlate incentive levels with performance. In fact, the record indicates that such an analysis would be extremely difficult (if not impossible) to perform due to the numerous variables that have affected portfolio performance--as well as differences in how energy efficiency performance has been defined--over the last 15 years. Still, TURN spends 13 pages of its opening brief presenting figures to support its position that utility spending and utility savings do not correlate to the higher incentive levels provided by D.94-10-059. Because all but one of these figures first appeared in briefs, there was no opportunity for other parties to test their assumptions and bases, or confirm that they accurately represented information supplied in utility data responses that were included in the record.¹³⁸ However, as PG&E and others note, these figures are problematic in many ways, including the fact that they do not reflect many other variables that affect performance.¹³⁹

In particular, DRA, CE Council and TURN attempt to infer from historical data on incentive levels and budgets that there is no correlation between the two, and therefore, no correlation between incentive levels and performance. However, the Commission has established funding levels for energy efficiency over the years taking a variety of factors into consideration. As PG&E points out:

"...from 1990 through 1997, the Commission approved energy efficiency funding in rate cases every two to three years, and annually reviewed and approved program plans and expected savings. Between 1997 and 2004, the Commission approved funding (consistent with legislative requirements) as well as programs and expected savings every year. Thus, every budget, program plan and expected savings reflected public input and ultimate approval by the CPUC. Each year's planned budget and expected savings therefore balanced the interests of all active stakeholders (including DRA and TURN) and the policy goals of the Commission at that time."¹⁴⁰

Therefore, it is not reasonable to conclude that because budgets and spending levels do not appear to be correlated with incentive levels, then higher incentive levels are no more effective than lower incentive levels, as these parties suggest.

We also rejected this line of thinking in D.03-10-057. In 2003, TURN presented similar graphical depictions of budgets and incentives when it argued that the Commission should reopen and repeal the shared-savings incentive mechanism adopted in D.94-10-059. Pointing to the drop in program spending in 1995 relative to previous years, TURN asserted that the shared-savings mechanism we adopted in 1994 did not provide incentives to the utilities to aggressively pursue cost-effective energy efficiency, despite the continuation of substantial shareholder incentives. In rejecting TURN's position, we explained:

"....the reasons for the reduction in program spending are certainly debatable. TURN fails to point out one very plausible factor to explain this reduction, namely, that we authorized reductions in DSM expenditures in order to continue an electric rate freeze that eventually became the basis for the electric rate freeze codified in [Assembly Bill] 1890."¹⁴¹

Moreover, as we noted in D.03-10-057, the lack of correlation between incentives and spending levels, for whatever reason, does not mean that the incentive mechanism has not produced sizable net benefits to ratepayers. Interestingly, none of the graphs presented by TURN even look at net benefits (savings minus costs) to ratepayers as a performance metric to consider. In fact, the only figure presented in this proceeding that does suggest a positive correlation between incentive levels and the production of savings at the highest efficiencies (or lowest total costs) is the one produced by SDG&E in its testimony and subjected to cross-examination.¹⁴²

Nonetheless, as the record indicates, it is difficult to draw definitive conclusions from graphing historical data on incentives and savings or net benefits, even if a more comprehensive graphing of data were available. This is because of fundamental differences in reporting and measurement practices, as well as very different purposes and incentive structures over the 15 year period. During some years "commitments" were counted in reporting savings achievements, while in others they were not.¹⁴³ In addition, in most of the years between 1990 and 2005 savings were not subject to ex post verification.¹⁴⁴ Therefore, the MWh achievements (and other metrics based on those achievements) depicted in these graphs are not directly comparable.

In light of these fundamental differences, it is difficult to determine exactly what the figures in TURN's Opening Brief represent. In particular, TURN's argument appears to rest largely on its Figure 4 and 5, which are captioned "total DSM spending" versus "total DSM incentives" and "total first-year electric savings (MWh)" versus "total DSM incentives," respectively. The utility incentives are those associated with the program for that year, but it is not clear whether the figures chart the original utility incentives claim for a given year, or the actual amounts that were collected after the subsequent years of measurement, or the amount awarded in that year for previous years' activities.

In addition, for years 1998-2000 it is not clear whether TURN included both the current program year expenditures plus the pre-1998 commitments that were paid in subsequent years. TURN does not indicate if the expenditures for 1998 and beyond are "actual" as were reported for pre-1998 years or "recorded" expenditures which included commitment dollars to be paid in future years as projects were completed. Moreover, since it presents only first-year savings, Figure 5 does not indicate how incentive levels relate to lifecycle savings for the measures counted in each year.

DRA and TURN ignore these differences and inconsistencies, as well as others. Observing that the utilities exceeded their savings targets in 2001 with incentives capped at 7% of energy efficiency program budgets, DRA argues that this level should be sufficient incentive to the utilities to achieve the Commission's goals.¹⁴⁵ In its opening brief, TURN makes a similar assertion.¹⁴⁶ However, this and other inferences made by these parties from historical data ignore the fundamental differences in the role of utilities in energy efficiency and resource procurement as well as the changes in Commission policy on energy efficiency over the past 15 years.

As we described in previous decisions,¹⁴⁷ the history of energy efficiency over the last 15 years can be divided into several different "eras". From 1990-1997 (the "pre-restructuring era"), the Commission viewed DSM including energy efficiency as a resource, and utilities administered the programs and earned incentives for successful achievement of resource savings and positive net benefits to ratepayers. During the first three years of this period, incentives were at relatively low experimental levels and tied to savings achievements based on ex ante estimates of load impacts. Beginning in 1993, the Commission adopted rigorous ex post measurement protocols and a shared-savings incentive mechanism designed to encourage the achievement of maximum dollar net benefits for ratepayers, and not tied to specific kWh, MW or therm savings goals.

Then, by the end of 1997, electric restructuring brought significant change. The focus of energy efficiency became market transformation with the emphasis on making energy efficiency a normal part of market transactions, and eventually phasing out ratepayer-funded DSM entirely. Utilities became temporary administrators, under short-term extensions through 2001. Milestone-based incentive mechanisms were adopted in lieu of shared-savings. Milestone incentives were based on spending levels, program activity levels (e.g., the number of audits performed) and on measuring market effects, with only a small portion of the incentive payments based on ex ante energy savings-and none on verified net resource benefits.¹⁴⁸ Overall incentives were capped at 7% of total energy efficiency budgets, and were eventually discontinued as of program year 2002.

The energy crisis year of 2001 brought with it emergency responses, including a "summer 2001" solicitation by the Commission to ramp up energy efficiency spending targeted to addressing energy shortfalls as quickly as possible. Since that crisis, and the reinstatement of the utilities as both supply-side and demand-side portfolio managers, we have entered into a new era of energy efficiency policy heralded in by the Energy Action Plan of 2003 and Commission policy determinations in this rulemaking and its predecessor (R.01-08-028).

Given this history, we find it unreasonable to infer that the energy efficiency incentive levels adopted during restructuring (or during the peak of the energy crisis in 2001) are appropriate for the risk/reward incentive mechanism we are adopting today. Nor is it reasonable to infer correlations between the stability of the pre-1998 industry structure and earnings mechanism and the rapid changes of the following few years. Conclusions drawn by comparing only one or two variables (e.g., energy savings and budgets) across these years fail to address the significance of the other variables in play during those years, such as restructuring, the energy crisis, and subsequent recovery.

For the reasons discussed above, we do not find merit to the inferences and recommendations made by TURN, DRA and CE Council concerning historical energy efficiency incentive levels and performance.

7.3.3. Supply-Side Comparability Benchmark: Adopted Range of Values

As discussed in previous sections, supply-side comparability provides a relevant numerical benchmark for the earnings potential under an energy efficiency incentive mechanism, in the context of other considerations. Most of the base case assumptions used by the utilities in their analysis were not disputed in this proceeding, such as the use of a mid-range value (12 years) for average energy efficiency measure lives, or the 50-50 split between "utility build-utility buy" scenarios. Based on the record in this proceeding and our review of the utilities' long-term procurement plans, we find these assumptions to be reasonable for our purposes today, namely, to estimate a comparable supply-side earnings benchmark.¹⁴⁹

DRA notes that several of the input assumptions are subject to change over time and may vary by utility, but does not present alternate values for calculating the base case numbers "since DRA recommends against using this model for the current program."¹⁵⁰ Instead, DRA suggests that specific input assumptions be updated and debated at each earnings claim (i.e., "at such time that shareholder funded incentives are proposed.")¹⁵¹ This recommendation presumes that the supply-side comparable earnings calculations developed in this proceeding represent a "model" where the input and output values directly establish the shared-savings rates. This presumption is not valid: As discussed in this decision, supply-side comparability serves as a general benchmark for a range of values, and is considered along with other factors in establishing the earnings potential and associated shared-savings rates under the adopted incentive mechanism.

DRA's proposal also suggests that the shared-savings rates would be modified at each earnings claim (through the "update and debate" that would occur). This would introduce an unreasonable level of uncertainty as to the design parameters of the mechanism itself, ongoing litigation and associated delays into the earnings recovery process. For these reasons, we do not adopt DRA's approach to establishing the input assumptions for supply-side comparable earnings calculations. Instead, for today's purposes of establishing a benchmark range of values, we use the utilities' base case assumptions with variability around certain parameters, as discussed below.

There was considerable debate in this proceeding over whether the utilities' return-on-equity assumptions should be adjusted downwards to reflect alternative use of funds, as proposed by TURN. We conclude that TURN's alternative-use-of-funds analysis is premised on an assumption that is not supported by the factual record in this proceeding or by a persuasive conceptual rationale, as we explain below. Therefore, we do not adopt this proposed adjustment to supply-side comparability calculations.

More specifically, TURN's analysis is premised on the assumption that the utility would have "in its pocket" the amount of cash that it otherwise would have used to invest in the supply-side resources, if not for energy efficiency.¹⁵² According to TURN's Witness Marcus, the utilities would not leave this large amount of available cash "sitting under their mattress," but would instead invest it in alternative ways.¹⁵³

But where does this large amount of cash come from? For PG&E alone, the equity that TURN assumes the utility would have on hand because it did not need to invest in supply-side infrastructure due to energy efficiency is on the order of $500 million.¹⁵⁴ TURN suggests that the utility has this amount of cash on hand through the accumulation of "retained earnings" over time, that is, what is left over in cash (customer bill payments less utility expenses) after what the utility spends to meet its capital needs and to pay out dividends.¹⁵⁵ However, as PG&E Witness Patterson testified, the utility does not accumulate large amounts of cash on hand to make investments other than in its own capital infrastructure, which currently costs PG&E approximately 2-½ to 3 billion dollars per year.¹⁵⁶ Other utility witnesses corroborated this testimony.¹⁵⁷ Common sense as well as the factual record refute TURN's premise that the utility would make it a practice to raise money in the capital markets to cover supply-side investments that it does not need to make, in order to retain those funds so that they could be used for alternative investments.¹⁵⁸ In fact, TURN's Witness Marcus acknowledges that the utilities are unlikely to issue new shares of stock to raise capital for the investments that they are actually planning to make over the next 3-5 year timeframe, if not longer.¹⁵⁹

In prepared sworn testimony and under cross-examination, utility witnesses explained how their companies actually plan and manage their cash requirements, based on first-hand experience as corporate planners. As they explained, the utility does not plan to have more cash than is needed for the plant and equipment that it will be building (or for working cash requirements), and carefully manages its cash reserves accordingly. The utility also does not sell shares or issue debt to raise cash for a capital investment it does not need to make, such as the supply-side resources that energy efficiency is planned to defer or displace. Granted, as one utility witness pointed out, there may be instances where the original forecast of cash requirements may overstate the need for capital infrastructure, resulting in more cash than is actually needed. However, that is certainly not something the utility plans for, and when it does occur, the utility generally uses that extra cash to buy back enough equity and debt to maintain its authorized equity/debt structure. It does not follow that the utility has "alternate uses" for equity on a dollar-for-dollar basis that was not needed for supply-side resources due to energy efficiency, as TURN's analysis assumes.¹⁶⁰

There is no factual dispute that PG&E, SCE, SDG&E and SoCalGas use profits from the capital investments and utility operations they do undertake for a variety of purposes described in TURN's testimony, including the pay-out of dividends to investors or stock repurchases (via their holding companies).¹⁶¹ However, such practices would be expected for any investor-owned utility company that is profitable based on the investments it does make in infrastructure (for which it is authorized a return) and the prudent management of its operating costs. It does not follow (as suggested by TURN's data requests and cross-examination on what the utility has done with earnings it has accumulated in the past or may accumulate in the future)¹⁶² that these profits originate from cash available to the utility because of supply-side investments displaced through energy efficiency.¹⁶³ Moreover, TURN's assertion that the utility's net income "will increase as a result of lower capital expenditures" defies a basic tenet of cost of service ratemaking: When capital expenditures are lower, by definition rate base is lower and so too is net income, all else remaining equal.¹⁶⁴

In addition, TURN's analysis of "alternative uses of funds" presents a fundamental contradiction to the position TURN also takes in this proceeding, namely, that the "right" answer to a calculation of foregone shareholder earnings is probably zero.¹⁶⁵ If shareholders are no worse off when energy efficiency displaces supply-side resources because they can take their investment funds elsewhere to earn a comparable return, how is it possible that the utility can use those funds for the alternative investments that TURN describes in its testimony?

Finally, we observe that TURN's economic and financial theories on basically the same issue have taken very interesting twists and turns over the years. For example, in 1993, observing that the utilities' market to book ratios were greater than 1.0, TURN asserted that the utility's return-on-equity was set too high by the Commission. TURN then postulated that because the utilities could improve earnings per share just by issuing additional shares under these circumstances, they were motivated to promote sales growth (and a corresponding growth in utility plant investment). TURN's recommendation at that time was to get rid of this "growth incentive" by setting the rate-of-return lower, rather than authorizing shareholder incentives for energy efficiency. D.93-09-078 presents a detailed discussion on the lack of factual, logical or policy support for TURN's thesis and rationale for opposing the continuation of energy efficiency shareholder incentives.¹⁶⁶

Having its previous theory rejected over ten years ago, TURN is now suggesting that utilities are motivated to raise equity in the market for supply-side investments no longer needed (or keep cash raised for that purpose "in its pocket"), so it can use those funds for alternative investments that TURN argues would yield an equal if not greater return to its shareholders.

Interestingly, in this proceeding TURN again observes that the utility's market-to-book ratios are high, and also expresses the view that the Commission is setting the return-on-equity too high in its cost of capital proceedings. However, TURN draws quite different conclusions from these observations in this instant case. According to TURN's Witness Marcus, the existence of authorized returns that are higher than the cost of equity, and the high ratio of market-to-book values are precisely the reasons for any positive foregone earnings that TURN calculates under its alternative-use-of-funds analysis.¹⁶⁷

In sum, TURN's theory on utility behavior in this proceeding lacks the support of a factual record and a persuasive conceptual rationale. For the reasons discussed above, we do not make adjustments for alternative-use-of-funds to the calculations of supply-side comparability.

There was also debate in this proceeding over whether comparable earnings should be calculated based on "comparable performance" versus "comparable costs." As described in Section 6.2.2 above, NRDC argues that the latter approach should be taken, i.e., that comparability should be calculated based on the energy efficiency portfolio costs. We do not agree. The utilities' approach is consistent with the purpose of the comparable earnings analysis, namely, to calculate earnings associated with the supply-side resources avoided by energy efficiency. NRDC recommends an approach based on comparable costs, rather than performance, but does not provide a persuasive conceptual rationale for doing so. Nor has this Commission taken that approach in the past when comparing the earnings potential of supply-side resources in the context of an energy efficiency shared-savings mechanism.¹⁶⁸ In fact, when questioned by the assigned ALJ on this subject, NRDC's witness was unaware of any precedent at this Commission or in other states for the approach to comparable earnings that NRDC advocates in this proceeding.¹⁶⁹ For these reasons, we do not calculate comparable supply-side earnings based on the costs of the energy efficiency investment.

Parties also debated whether debt equivalence should be imputed for power purchases in the utilities' comparable earnings calculations. In addition, TURN disputes the use of only combined-cycle natural gas turbines (CCGT) for the avoided generation capacity, and argues for replacing 24% of that amount with lower-cost combustion turbines (CTs).¹⁷⁰

We have carefully reviewed the record on these issues, and conclude that there are persuasive arguments on both sides of these disputes. For example, as NRDC, TURN and DRA point out, the Commission has not adopted a formal debt equivalence policy in cost-of-capital proceedings and has chosen instead to consider the issue of debt equivalence on a case-by-case basis in those proceedings. At the same time, as the utilities note, the Commission has authorized the utilities to impute debt equivalence in assessing the risks of power purchase agreements.¹⁷¹ With regard to the CT substitution issue, SCE raises reasonable resource planning and dispatch issues in response to TURN's arguments for such substitution.¹⁷² These specific issues are more appropriately addressed in cost-of-capital proceedings (for debt equivalence policies and calculations) or resource planning/avoided cost proceedings (for the CCGT versus CT avoided cost question). Rather than attempt to resolve them here, we will acknowledge that they create a range of possible outcomes around the base case assumptions for the purpose of calculating comparable earnings.

Imputing debt equivalence to power purchases and assuming avoided generation capacity based exclusively on CCGT costs produces the upper range of comparable earnings estimates. Based on the scenario analysis presented on the record, this upper range is approximately $700 million for all utilities combined, over the three-year program cycle.¹⁷³ Removing debt equivalence and substituting 24% of avoided CCGT costs with CT capacity costs, as TURN recommends, will produce the lower range of the calculations. Based on the record, we estimate this lower range at $450 million for all four utilities combined, over the three-year program cycle.¹⁷⁴

Finally, DRA questions the applicability of the utilities' supply-side comparability analysis to natural gas energy efficiency activities, since natural gas companies do not avoid the same type of supply-side investments as assumed in the analysis for electric utilities. We agree with PG&E that the supply-side comparability benchmark should not be separated by fuel or type of program activity, but rather should serve as a general numerical guide for setting the appropriate share of the combined net benefits from electric and natural gas efficiency programs. In fact, by including only the earnings from the electric supply-side resources foregone, and none from any gas supply-side resources foregone, it could be argued that the analysis understates the resulting shared savings rate. As PG&E points out, this approach avoids the need to debate the size of gas supply-side resource investments, about which there is no record.¹⁷⁵

In sum, we conclude that the supply-side comparability analysis in this proceeding yields comparable supply-side pre-tax earnings in the range of $450 to $700 million for all four utilities combined, based on the energy efficiency program activities planned for the 2006-2008 program cycle.

7.3.4. Adopted Shared-Savings Rate(s)

As most parties to this proceeding acknowledge, establishing the level of earnings opportunity for a shareholder risk/reward incentive mechanism is ultimately a judgment call that the Commission must make, and not a precise science.¹⁷⁶ Generally speaking, we believe that the earnings potential under such a mechanism should be designed both to balance the potential penalties under the mechanism and to offset existing financial and regulatory biases in favor of supply-side procurement. In this context, consideration should be given to what level of earnings potential will provide a clear signal to utility investors and shareholders that achieving and exceeding the Commission's savings goals (and maximizing ratepayer net benefits in the process) will create meaningful and sustainable shareholder value. At the same time, we should weigh and consider differences in the risk/reward profile of utility resource choices in applying the comparable earnings benchmark to our incentive mechanism. In addition, consideration should be given to the level of performance expected in return for higher and higher earnings potential. Moreover, these considerations should be balanced by a sense of what is "fair" to ratepayers in terms of the return on their investment in energy efficiency.

In our judgment, providing the opportunity to earn using supply-side comparable earnings as a benchmark creates the sustainable "clear signal" we are looking for on the earnings side of the incentive mechanism. However, we are also persuaded by the arguments made by DRA and others that this benchmark should be applied very conservatively to our adopted incentive mechanism, given the relative risk/reward profiles of utility resource choices and the additional considerations discussed below.

As we recognized in D.94-10-059, it is challenging to compare the different (and changing) risk/reward profiles among utility resource choices in considering how to apply a supply-side comparable earnings benchmark to the design of an energy efficiency incentive mechanism.¹⁷⁷ Nonetheless, we were able to make certain observations relevant to the different risk/reward profiles that were applicable at that time, and reached conclusions from them as to the appropriate level of earnings potential for energy efficiency relative to that benchmark.¹⁷⁸

Today, we observe that since the energy crisis, utilities must now manage supply-side resources to meet resource adequacy requirements subject to the risk of significant financial penalties.¹⁷⁹ With the passage of Senate Bill 1368, utilities also now face the risk of non-compliance for any long-term supply-side procurement that does not meet the GHG performance standard required by that statute. Moreover, although its implementation is several years out into the future, the statewide GHG emissions cap pursuant to Assembly Bill 32 will further increase the financial risks associated with any supply-side resource procurement that increases GHG emissions.¹⁸⁰ These developments serve to increase the risks associated with procuring at least certain types of supply-side resources and, in turn, reduce the relative risks of energy efficiency in the context of utility resource portfolio management, even if the impact of these changes cannot be quantified.

Working in the other direction, however, are the fundamental disincentives to energy efficiency that we discuss in Section 6.3.1 relative to "steel-in-the-ground" supply-side investments of any type (renewables or non-renewable generation, transmission and distribution), and the short-term rate impacts associated with energy efficiency relative to all supply-side procurements, including power purchase agreements. Moreover, as discussed in Section 4.2, energy efficiency is not as readily "deployed" as supply-side resources, since it involves customer-side appliance changes, arrangements with manufacturers and distributors, etc. Hence even when the portfolio funding is authorized, there is more uncertainty as to the actual deployment results of energy efficiency than for supply-side resources. In addition, energy efficiency requires independent verification of load impacts that extends beyond program implementation. These attributes introduce uncertainty and performance risks to shareholders that are not shared by supply-side choices.

Although ratepayers put up 100% of the investment capital for energy efficiency programs, shareholders are at risk under the adopted incentive mechanism for losse