Centerpoint Energy, Inc. F/K/A Reliant Energy, Incorporated, and American Electric Power Company Inc. v. Public Utility Commission of Texas ( 2004 )

IN THE SUPREME COURT OF TEXAS

IN THE SUPREME COURT OF TEXAS

════════════

No. 03-0396

════════════

CenterPoint Energy, Inc. f/k/a Reliant Energy, Incorporated

and American Electric Power Company, Inc., Petitioners

v.

Public Utility Commission of Texas, Respondent

════════════════════════════════════════════════════

On Petition for Review from the

Court of Appeals for the Third District of Texas

════════════════════════════════════════════════════

Argued February 18, 2004

            Justice Owen delivered the opinion of the Court, in which Justice Hecht, Justice O=Neill, Justice Jefferson and Justice Wainwright joined.

Justice Brister filed a dissenting opinion, in which Chief Justice Phillips, Justice Schneider and Justice Smith joined.

In a regulated environment, electric utility companies made very large expenditures to build generation plants, some of which were nuclear power plants.  Under regulation, those utilities and their shareholders were entitled to, and had, a reasonable opportunity to recover through rates not only their reasonable and prudent investments of capital in those plants, but also a reasonable, regulated return on those investments.[1]  In 1999, the Texas Legislature decided that it was in the public interest to partially deregulate the electric power industry.[2]  The Legislature recognized that in fundamentally changing the industry, it was altering the assumptions that had led utilities to invest large sums in power generation assets.  The Legislature understood that the cost of these assets likely would be recovered in a regulated environment, but might well become uneconomic and thus unrecoverable in a competitive, deregulated electric power market.  The Legislature called such uneconomic assets stranded costs.[3]  The term Astranded costs@ has a specific definition in the Public Utility Regulatory Act (APURA@ or Athe Act@),[4] but generally speaking, it is the extent to which the book value of generation-related assets and purchased power contracts exceeds their market value.[5]

The Legislature concluded that if generating plants became uneconomic as a result of legislatively mandated deregulation, it was in the public interest for utilities to be made whole by recovering their full investment in those generation plants, although the utilities would no longer receive a return on those investments.[6]  The Legislature determined that utilities should not be required to forfeit their investments in generating plants with the advent of deregulation.  The Legislature thus said in the PURA that if there are stranded costs, an electric utility Ais allowed to recover all of its net, verifiable, nonmitigable stranded costs incurred in purchasing power and providing electric generation service.@[7]  The Legislature set forth a comprehensive scheme for estimating, finalizing, and recovering those costs.[8]  Stranded cost recovery, if any, will occur over a period of years rather than in a lump sum.[9]  No one disputes that the Legislature intended electric utilities to recover carrying costs on stranded costs to compensate for the financing costs incurred during the stranded cost recovery period.  Nor does anyone dispute that prior to deregulation, carrying costs on investments in generation plants were included in rates.  The only issue before us is the date from which carrying costs may be recovered once deregulation commenced:  January 1, 2002, which was the first day of deregulation, or two or more years later, at the end of final true-up proceedings.

In a rulemaking proceeding, the Texas Public Utility Commission determined that carrying costs on a true-up balance must be calculated from the later date, the date of a true-up final order (sometime after January 10, 2004).[10]  CenterPoint Energy, Inc. (formerly known as Reliant Energy, Inc.) and American Electric Power Company, Inc. (AAEP,@ a public utility holding company whose Texas operating company was formerly known as Central Power and Light Company) contend that this rule is invalid, arguing that carrying costs should be recovered from the date that regulated rates ended and competition commenced, which was January 1, 2002.  The court of appeals rejected the generation companies= arguments and upheld Rule 25.263(l)(3).[11]

The court of appeals also rejected a related challenge to Rule 25.263(l)(3).  In separate proceedings not before us, the Commission directed CenterPoint and AEP to reverse early efforts to mitigate potential stranded costs.[12]  If it is ultimately determined in an appeal from those proceedings that the generation companies have stranded costs and the Commission erred by reversing early mitigation efforts, the generation companies argue that Rule 25.263(l)(3) does not permit them to recover interest for the period of time that amounts associated with early mitigation efforts were incorrectly refunded to customers.[13]  The court of appeals in this case held that Aa utility=s right to fully recover its stranded costs does not encompass a right to early mitigation.@[14]  The generation companies take issue with this determination, but advise us that the matter would be moot if this Court concludes that they are entitled to carrying costs on stranded costs from January 1, 2002.

We hold that Rule 25.263(l)(3) is inconsistent with the Legislature=s intent, expressed in Chapter 39 of the PURA, that utilities fully recover their Anet, verifiable, nonmitigable stranded costs incurred in purchasing power and providing electric generation service,@[15] that Aexist on the last day of the freeze period [December 31, 2001].@[16]  A two- or three-year gap in recovery of carrying costs would not permit generation companies full recovery of their stranded costs as the Legislature envisioned.  However, the capacity auction true-up procedure set forth in the Act[17] may include a component for return of or on stranded costs in 2002 and 2003, a determination that cannot be made from the record in this rulemaking proceeding.  The amount of stranded cost recovery, if any, through capacity auction true-ups will have to be considered in determining the amount of carrying costs on stranded costs from January 1, 2002 to ensure that there is no overrecovery of stranded costs.[18]  We accordingly remand this issue to the Commission for further consideration of whether to address carrying costs in a rule or in contested case hearings applicable to each electric utility and its affiliates.

Because Rule 25.263(l)(3) is invalid and we are remanding this matter to the Commission, we do not address whether or under what circumstances generation companies might be entitled to interest on refunds of early mitigation credits if those refunds were to be reversed.

I

We first consider the standard of review.  The Commission=s order adopting Rule 25.263[19] reflects that the rule was promulgated under section 14.002 of the Act, which provides:  AThe Commission shall adopt and enforce rules reasonably required in the exercise of its powers and jurisdiction.@[20]  The order also cites sections 39.252 and 39.262 of the PURA, which the Commission said Aaddress[] a utility=s right to recover stranded costs@ and Arequire the commission to conduct a true-up proceeding for each . . . utility after the introduction of customer choice.@[21]

Section 39.001 of the PURA has separate provisions governing review of the validity of a competition rule.  Section 39.001 provides that A[j]udicial review of competition rules adopted by the commission shall be conducted under Chapter 2001, Government Code, except as otherwise provided by this chapter [39].@[22] Section 39.001 provides for a direct appeal to the Third Court of Appeals for A[j]udicial review of the validity of competition rules,@[23] and for expedited procedures in such an appeal.[24]  The Commission=s order does not contain a reference to section 39.001.

CenterPoint and AEP nevertheless filed a direct appeal in the Third Court of Appeals, and the court of appeals= opinion recites that the court had before it a direct appeal under subsections 39.001(e) and (f) of the Act.[25]  No one has taken issue with the characterization of Rule 25.263 as a Acompetition rule@ or the applicability of subsections 39.001(e) and (f).  Regardless of whether those subsections apply, the validity of Rule 25.263(l)(3) is at issue, and under the Administrative Procedure Act[26] and our case law,[27] the rule is invalid if, among other things, it is in violation of a statutory provision, in excess of the agency=s statutory authority, or arbitrary, capricious, or characterized by abuse of discretion or clearly unwarranted exercise of discretion.  We turn to an analysis of Rule 25.263(l)(3) with these considerations in mind.

II

This is the fourth case in which we have addressed issues arising out of the partial deregulation of the electric power industry, including issues concerning stranded costs.[28]  Stranded costs include regulatory assets,[29] which are essentially bookkeeping entries that reflect a charge that was to be included in a utility=s future rates in a regulated environment.[30]  Stranded costs also include the reasonable excess cost above the market value of assets such as generating plants, including nuclear power plants.[31]  As we have previously explained, under the regulatory scheme that existed before 1999, an electric utility had an opportunity to recover prudent capital investments in generation assets and a reasonable return on those investments through rates.[32]  The Act recognizes that, generally, there are financing or carrying costs associated with generation assets and that these carrying costs were historically recovered in rates.[33]  Accordingly, under traditional rate regulation, ratepayers would pay carrying costs as a utility recovered its investment in generation assets over the useful life of those assets.

The Commission recognizes that if costs are stranded in a deregulated environment, a generation company is entitled to recover carrying costs on those stranded costs, which are recovered over time either through a competition transition charge or securitization.  The dissent does not dispute that the Act implicitly, if not explicitly, assumes that there will be carrying costs on stranded costs. The only issue is whether the Act contemplates roughly a two-year gap in recovery of carrying costs between the date regulation ceased (January 1, 2002) and the date of a final true-up order (2004 or perhaps beyond).

The Commission says such a gap is permissible. The Commission determined in Rule 25.263(l)(3) that carrying costs on stranded costs should be recovered by electric utilities only from the date of the final true-up order.[34]  Carrying costs are to be calculated based on each utility=s individual cost of capital established in that utility=s unbundled cost of service (UCOS) proceeding.[35]  Final true-up orders will be entered sometime after January 10, 2004, which is specified in section 39.262 as the date after which each transmission and distribution utility, its affiliated retail electric provider, and its affiliated power generation company must jointly file to finalize stranded costs.[36]  We must decide whether the Commission=s failure to permit the recovery of carrying costs for approximately a two-year period after regulated rates ended and customer choice began violates the Act, is in excess of the Commission=s statutory authority, or is arbitrary, capricious, or a clearly unwarranted exercise of discretion.

In its order adopting Rule 25.263(l)(3), the Commission explained why it chose the date of a final true-up order as the date from which carrying costs should accrue by saying Aa utility=s true-up balance becomes due upon the issuance of a final order in that utility=s true-up proceeding.@[37]  In its briefing on appeal, the Commission elaborated, contending that stranded costs did not come into existence on the first day of customer choice, but instead will come into existence only after a true‑up proceeding is concluded.  The court of appeals agreed with this rationale.[38]  In post-submission briefing in this Court, the Commission contends that in addition to this logic, disallowing carrying costs from the date customer choice commenced (January 1, 2002) until the date of a final order sometime in 2004 or later is justified because otherwise generation companies may receive a double recovery.  The Commission contends that the capacity auction true-ups that will be conducted under section 39.262(d) of the PURA[39] include a component that will allow generation companies to recover debt service on their book generation assets in addition to operating costs.  Texas Industrial Energy Consumers (TIEC), an intervenor in the court of appeals and in this Court, likewise contends that because of the capacity auction true-up, generation companies would receive a double recovery if they are permitted to receive carrying costs from the date customer choice began.

The generation companies counter that the date as of which stranded costs are to be determined is December 31, 2001, as reflected throughout chapter 39.  The generation companies also contend that under section 39.262(d), they are entitled to recover the amount the capacity auction true-up yields, without regard to whether they have stranded costs.  Capacity auction true-up proceeds and stranded cost recovery are entirely separate, the generation companies contend, and there would be no double recovery if carrying costs on stranded costs are permitted from January 1, 2002.

Because of the complexity of the issues, we think it helpful to outline our conclusions before examining the Act in greater detail. We conclude that the Commission=s construction of chapter 39 was incorrect regarding the date as of which stranded costs are to be determined.  Chapter 39 reflects that the amount of stranded costs, if any, is to be determined as of the day before competition beganCDecember 31, 2001Cor earlier in some cases.[40]  However, the Act recognizes that a determination of whether there were stranded assets as of December 31, 2001, and a meaningful valuation of those assets as of that date, could not be made until a deregulated market had a period of time to develop and then to stabilize.  Because the Commission=s rule is based on an incorrect construction of the Act in this regard, it is infirm.

That does not mean that generation companies are entitled to carrying costs on the entire positive balance of stranded costs, if any, from January 1, 2002.  Based on the record before us, it appears that the design of the capacity auction true-up may have permitted generation companies to recover during 2002 and 2003 at least a portion of their fixed costs, including stranded costs, if any. That determination cannot be made from this record.  Preventing an overrecovery of stranded costs requires a determination, on a company-by-company basis, of whether proceeds from a capacity auction true-up had a component for return on or of stranded costs and of the quantity of any such return.  We accordingly remand this proceeding to the Commission for further consideration.

The dissent asserts that our holding Apotentially@ entitles utilities to Abillions of dollars in interest.@[41]  As the dissent concedes, however, it is unknown at this point whether there will be any stranded costs at all and thus any carrying costs.  If it is determined that stranded costs did exist on December 31, 2001, the amount of any such costs is likewise unknown, as is the extent to which stranded cost recovery has or will occur through a capacity auction true-up.  This Court must give effect to legislative intent as expressed in the PURA.  We must ensure that the Commission implements the statutory scheme set forth in Chapter 39 of the PURA without regard to speculation about how deregulation may or may not affect market values of generation assets and thus may or may not affect rates.

The pertinent sections of Chapter 39 and the record in this case are considered more thoroughly below.

III

The deregulation process has many components, some of which can be briefly summarized for purposes of this appeal. The Legislature determined that the production and sale of electricity should no longer be regulated in Texas, except for transmission and distribution services and the recovery of stranded costs.[42]  The Legislature chose January 1, 2002 as the date on which customer choice would begin, with a few exceptions not relevant to the issues in this case.[43]  Correspondingly, December 31, 2001 was the date traditional regulation of wholesale rates would end.[44]  The Legislature recognized that deregulation could not be accomplished overnight. Accordingly, beginning September 1, 1999, and continuing until January 1, 2002, the Legislature froze retail electric rates.[45]  The Legislature also directed that by January 1, 2002, each electric utility must separate its business activities into a power generation company, a retail electric provider, and a transmission and distribution utility.[46]  For a five-year period after the beginning of customer choice, from January 1, 2002 until January 1, 2007, retail electric providers are required to offer residential and small commercial customers rates that are six percent less than those in effect on January 1, 1999, with certain adjustments.[47]  This is the Aprice to beat@ in the retail market.[48]

Resolution of the issues raised by Rule 25.263 requires a more detailed focus, however, on provisions of chapter 39 that govern generation companies and stranded costs.

A

In enacting deregulation legislation, the Legislature had before it a 1998 report prepared by the Public Utility Commission that analyzed potential stranded costs.[49]  The term AECOM@ was used in that report, meaning excess costs over market.  That report identified a number of generation companies that, in a deregulated market, were projected to have unrecoverable or Astranded@ costs, principally nuclear power plant investments.  The Legislature required electric utilities identified in this 1998 report as having projected stranded costs to use Aa number of tools . . . to mitigate stranded costs@ by Areduc[ing] the net book value of, otherwise referred to as >accelerat[ing]= the cost of recovery of, its stranded costs@ before customer choice began on January 1, 2002.[50]  For each of the three years preceding customer choice (1999, 2000, and 2001), electric utilities were required to file annual reports showing whether they had excess earnings from charging frozen rates,[51] and if so, to apply those excess earnings to reduce the book value of potentially stranded investments.[52]  During the period leading up to customer choice, before utilities were unbundled, utilities also had the option of redirecting depreciation expense relating to transmission and distribution assets to generation assets as another tool for reducing the book value of potentially stranded assets.[53]

The Legislature recognized that these early mitigation efforts might not be sufficient to eliminate stranded costs. In the period leading up to customer choice on January 1, 2002, the Legislature gave electric utilities another option.  At any time after September 1, 1999 (the start of the retail rate-freeze period), a utility was permitted to securitize 100 percent of its regulatory assets[54] and up to 75 percent of its other estimated stranded costs.[55]  This meant that a generation company could begin to recover stranded costs even before January 1, 2002, and before the final dollar amount of those stranded costs, if any, could be quantified.

The only explicit reference to carrying costs on stranded costs appears in a section of the Act regarding securitization.[56]  The securitization provisions reflect that the Legislature implicitly, if not explicitly, assumed that carrying costs on stranded costs are to be borne by ratepayers over the entire life of the generating assets under either conventional financing methods or securitization.[57]  Section 39.301 sets forth the purposes of allowing securitization.  Securitization is intended to Alower the carrying costs of the assets relative to the costs that would be incurred using conventional utility financing methods.@[58]  The Acosts that would be incurred using conventional utility financing methods@ are the carrying costs on generation assets that customers would otherwise pay to the generation company.  The Legislature commanded the Commission to Aensure that securitization provides tangible and quantifiable benefits to ratepayers, greater than would have been achieved absent the issuance of transition bonds.@[59]  In this same vein, the Legislature said that the Commission could permit securitization only if it found Athat the total amount of revenues to be collected under the financing order is less than the revenue requirement that would be recovered over the remaining life of the stranded costs using conventional financing methods.@[60] Securitization could occur under the statute, and did occur (at least for regulatory assets),[61] prior to January 1, 2002.[62]  When securitization is used, carrying costs on stranded costs are recovered from the date of securitization forward.[63]  There is no two-year gap in payment of carrying costs for 2002 and 2003.

In making the determination of whether securitization benefitted ratepayers, the Legislature directed the Commission to look at the entire remaining life of stranded costs, beginning as early as September 2, 1999.[64]  There is no suggestion that in making that calculation, the Commission could drop out the carrying costs for the two-year period between the onset of customer choice (January 1, 2002), and the date a final true-up order could be entered (2004 or beyond).  If stranded costs never come into existence until 2004, as the Commission argues, then the Legislature=s securitization scheme compared apples to oranges and allowed securitization to proceed using a much less stringent test from the ratepayers= perspective.  That is not a reasonable construction of the Act.

B

For estimated stranded costs that had not been mitigated or had not been or could not be securitized, the Legislature provided that those costs should be recovered through competition transition charges starting on the first day of competition, January 1, 2002. The Legislature directed in section 39.201 that between April 1, 2000 and January 1, 2002 the Commission was to determine any expected competition transition charge and make it effective on January 1, 2002.[65]  The actual dollar amount of stranded costs could not be known at the time those charges were to be determined, so in section 39.201, the Legislature directed the Commission to calculate competition transition charges using the ECOM administrative model in the Commission=s 1998 Report, but using updated, company-specific inputs and market-based natural gas forward prices, in addition to other specified updates.[66]

It is highly significant to the question before us today that the Legislature said in section 39.201 that the pertinent date for quantifying stranded costs was December 31, 2001, Athe last day of the freeze period@ and the last day before customer choice began on January 1, 2002.[67]  In implementing competition transition charges, the Commission was directed to calculate Athe amount of stranded costs as defined in Subchapter F that are reasonably predicted to exist on the last day of the freeze period [December 31, 2001].@ Accordingly, the Commission was told that sometime between April 1, 2000 and January 1, 2002, it was to estimate Athe amount of stranded costs as defined in Subchapter F that are reasonably projected to exist on the last day of the freeze period [December 31, 2001]@ and to put nonbypassable rates in effect on January 1, 2002 to begin recovery of these amounts.[68]  The Legislature did not tell the Commission to estimate the amount of stranded costs that were projected to exist as of a date in 2004, or at the end of a true-up proceeding, which would have been appropriate dates if the Commission=s interpretation of the Act were correct. Instead, the Legislature directed the Commission to permit generation companies to begin recovery of stranded costs on January 1, 2002, through competition transition charges, if they were projected to have stranded costs as of December 31, 2001.     

C

The Legislature=s use of the words Aremaining stranded costs@ in section 39.201(l) is also significant.[69]  That section provides that, A[t]wo years after customer choice is introduced [meaning two years after January 1, 2002],@ the Commission is to determine in final true-up proceedings whether there are any Aremaining stranded costs.@[70] This indicates that the Legislature thought that stranded costs would have been in existence before the final true-up and only remaining stranded costs would be recovered going forward.  If, as the Commission contends, stranded costs could not come into existence until after the true-up proceedings were concluded, then the Legislature would not have referred to Aremaining stranded costs@ that are to be quantified during the final true-up.  The Commission=s position is contrary to the Legislature=s directive that the 2001 Astranded cost estimate@ must Abe reviewed and, if necessary, adjusted to reflect a final, actual valuation in the true-up proceeding.@  The 2001 projection of what stranded costs would exist as of December 31, 2001 was to be reviewed and adjusted.

D

As it turned out, the calculations made by the Commission in 2001 using the ECOM model showed that no generation company was projected to have stranded costs as of December 31, 2001. Accordingly, no competition transition charges were implemented for any generation company.  That fact seems to have obscured the Commission=s view of the date as of which section 39.201 says stranded costs are to be measured. Rule 25.263(l)(3) is contrary to what the Legislature contemplated could happen under section 39.201.  If stranded costs had been projected in 2001 to exist on December 31, 2001 for a generation company, then that company was entitled to begin collecting stranded costs.[71]  Of course, the stranded costs would not have been collected in a lump sum.  The Legislature gave the Commission factors to consider in deciding Athe length of time over which stranded costs@ may be recovered through competition transition charges.[72]  No one, including the dissent, disputes that there would have been a component for recovering carrying costs in competition transition charges.

If company A had been projected in 2001 to have $5,000,000 in stranded costs, A would have begun recovering $5,000,000 plus carrying costs through a competition transition charge from January 1, 2002 over a number of years.[73]  If the 2004 true-up confirmed that the 2001 projection was an accurate predictor of the amount of stranded costs, A would continue to receive the competition transition charge.  The net result would be that A recovers carrying costs on stranded costs from January 1, 2002.

But, as has happened, assume that A was projected in 2001 to have no stranded costs and therefore did not receive a competition transition charge in 2002 and 2003.  Further assume that in a 2004 true-up proceeding, it is determined that A has stranded costs of $5,000,000.  The Commission says that A could begin recovering $5,000,000 plus carrying costs from 2004 over a period of years.  The net result would be that A recovered carrying costs only from 2004.

We must ask, why would the Legislature, planning in 1999 for various contingencies, have intended for company A to recover two years of carrying costs if the 2001 projection turned out to be an accurate predictor of actual stranded costs, but not if the 2001 projection was not an accurate predictor of actual stranded costs?  It is extremely unlikely this was the Legislature=s intent, particularly when it is undisputed that if the 2001 projection overestimated rather than underestimated stranded costs, overrecovery dating back to January 1, 2002 would be reversed.[74]  It seems more likely that the Legislature intended for its scheme to be symmetricalCrequiring adjustments for both overrecovery and underrecovery if the 2001 projection was not an accurate predictorCrather than arbitraryCallowing adjustments only for overrecovery.  The Act unquestionably provides that if the 2001 projection proved to have overestimated stranded costs when the true-up was conducted in 2004, then that overrecovery would be rectified through 1) a reduction in the competition transition charge, to the extent it had not been securitized, 2) a reversal, in whole or in part, of depreciation expense redirected under section 39.256, 3) a reduction in the transmission and distribution utility=s rates, or 4) a combination of these measures.[75]

If the Commission and TIEC were correct that no stranded costs could come into existence until the end of a true-up proceeding, which would be sometime in 2004 or perhaps beyond, then a generation company that collected competition transition charges under section 39.201 would be required under the rationale of Rule 25.263(l) to refund all carrying costs collected as part of those charges between January 1, 2002 and a final true-up order.  Nothing in chapter 39 suggests such a result.  For example, suppose that the 2001 ECOM model calculations made pursuant to section 39.201 had projected that a generation company=s stranded costs as of December 31, 2001 were $5,000,000, and that company began collecting competition transition charges over a fifteen-year period to recover that amount.  Additionally assume that in 2004, the final true-up showed that the company=s stranded costs were $5,000,000, and that $1,000,000 of those costs had been recovered through competition transition charges.  Applying the Commission=s reasoning, the generation company would have to refund the carrying cost component in the transition charges collected from 2002 until 2004.  Indeed, applying the Commission=s reasoning, the company would have to refund interest on the carrying costs to make up for the time value of the carrying costs that the company collected before 2004.

The Commission=s contentions in this appeal regarding carrying costs are inconsistent with its own rule.  Rule 25.263(g)(2)(A) recognizes that under the example in the paragraph above, a company that began collecting carrying costs in 2002 as part of a competition transition charge would keep those carrying costs if, in a 2004 true-up proceeding, it is found to have stranded costs.  Rule 25.263(g)(2)(A) provides that in a final true-up, any generation-related invested capital recoverable through a competition transition charge, exclusive of carrying costs, projected to be collected through the date of the final order in the true-up proceeding, is to be deducted from the December 31, 2001 book value of generating assets.[76]

The Commission=s rule creates an anomaly.  Whether carrying charges can be collected from January 1, 2002 depends entirely on whether the 2001 ECOM model projected stranded costs.  If the model did, then unquestionably, section 39.201 required the Commission to put into effect competition transition charges through which generation companies would begin recovering stranded costs and carrying costs on those stranded costs.[77]  If the ECOM model projected no stranded costs, but in 2004, market valuations reveal that the 2001 ECOM projection was not a good predictor of actual stranded costs, then the Commission=s rule does not permit carrying costs.  Carrying cost recovery under the Commission=s rule can turn entirely on the accuracy of the 2001 ECOM projections.  This is not a reasonable construction of sections 39.201 and 39.262.

E

Other parts of section 39.201 indicate that the Legislature considered December 31, 2001 to be the date as of which stranded costs would finally be calculated in a true-up proceeding.  Subsection 39.201(l) says:  ATwo years after customer choice is introduced [which would be 2004], the stranded cost estimate under this section shall be reviewed and, if necessary, adjusted to reflect a final, actual valuation in the true-up proceeding under Section 39.262.@[78]  A[T]he stranded cost estimate@ in subsection (l) refers back to the estimate performed under subsection (h) that was to apply the ECOM model with updated inputs in order to calculate Athe amount of stranded costs as defined in Subchapter F that are reasonably projected to exist on the last day of the freeze period@ as required by subsection (g).[79]  If stranded costs did not and could not exist as of December 31, 2001, as the Commission and TIEC contend, then why did the Legislature direct that the estimate of stranded costs as of December 31, 2001 be adjusted?  If the Legislature had meant to substitute a calculation of stranded costs as of a date in 2004, it would have said so.  It did not.  Even for final true-up purposes, section 39.201 refers back to stranded costs projected to exist as of December 31, 2001.[80]

Section 39.262 sets forth in greater detail how the final true-up proceedings are to be conducted in 2004.  Section 39.262(c) refers to Afinaliz[ing]@ Athe estimated stranded costs used to develop the competition transition charge in the proceeding held under Section 39.201.@[81] Here again, the Legislature is directing that the 2001 estimates used to calculate stranded costs Aprojected to exist on the last day of the freeze period [December 31, 2001]@ be finalized.[82]  The Legislature is directing that a final determination be made of the stranded costs that existed on the last day of the freeze period.  It did not direct the Commission to determine stranded costs that exist Aon the first day a final true-up order is issued.@

F

The reference in section 39.201(g) to the definition of stranded costs in Subchapter F leads to another reference to the December 31, 2001 date.  Stranded costs are defined in Subchapter F as follows:

AStranded cost@ means the positive excess of the net book value of generation assets over the market value of the assets, taking into account all of the electric utility=s generation assets, any above market purchased power costs, and any deferred debit related to a utility=s discontinuance of the application of Statement of Financial Accounting Standards No. 71 (AAccounting for the Effects of Certain Types of Regulation@) for generation‑related assets if required by the provisions of this chapter. For purposes of Section 39.262 [regarding true-up proceedings], book value shall be established as of December 31, 2001, or the date a market value is established through a market valuation method under Section 39.262(h), whichever is earlier, and shall include stranded costs incurred under Section 39.263.[83]

The Legislature defined stranded costs by using December 31, 2001, the day before customer choice was to begin, as the benchmark for book value, or an earlier date if assets were sold or exchanged.[84]

The Commission and TIEC point out that the definition of stranded costs in section 39.251(7) has two components, book value as of December 31, 2001, and market value, which may not be determined for some companies until 2004 or beyond.  This, they say, is justification for concluding that stranded costs do not come into existence, and therefore the company has no right to carrying costs, until the date of a final order in a true-up proceeding.  This reasoning has several flaws.  The first is the wording of the Act itself.

Section 39.251(7) recognizes that stranded costs may be finally determined even before January 1, 2002, the date that competition began, and certainly before 2004.[85]  The definition of stranded costs provides that if, under section 39.262(h), a company sells or exchanges assets to establish market value before December 31, 2001, then for purposes of a true-up proceeding (section 39.262), book value shall be established on that earlier date of sale or exchange.[86]  Accordingly, any Apositive excess of the net book value of generation assets over the market value of the assets@[87] could be known even before January 1, 2002.  How can it be said that such stranded costs did not come into existence until 2004 or beyond?

Similarly, a company may have been projected to have no stranded costs when the Commission performed the ECOM model calculation in 2001 required by section 39.201.  That company may sell or exchange assets sometime in 2002 or 2003.  Stranded costs can be finally quantified once that sale or exchange occurs.  Yet the Commission says that these stranded costs could not come into existence until 2004 and that the generation company is not entitled to accrue any carrying charges on these stranded costs until the date the Commission issues a final order in a true-up proceeding.  Here again, the statutory language does not support such a result.

G

TIEC and, to some extent, the Commission argue that because of fluctuations in market prices from December 31, 2001 until the date of final orders in true-up proceedings in 2004, stranded costs could come in and out of existence.  Therefore, they say, it is reasonable to choose a date in 2004 rather than December 31, 2001.  This contention ignores the fact that no gain could be realized from upswings in the market value of generation assets unless those assets were sold or exchanged. Interim market swings therefore have nothing to do with the stranded cost equation (net book value of generation assets over the market value of the assets)[88] unless generation assets are sold or exchanged. If they are sold or exchanged, then, as discussed above, the amount of stranded costs is determinable on the date of sale or exchange, and there is no justification for deferring the accrual of stranded costs until sometime in 2004 or beyond, at the end of a true-up proceeding.

There is no cause for concern that rises in gas prices during the interim between January 1, 2002 and December 31, 2003 would translate into excess profits for power companies, even if their nuclear power plants operated profitably.  As will be discussed in more detail in section IV below, the return that a power company could earn during 2002 and 2003 was predetermined in 2001.  If that predetermined margin is exceeded, then the excess will be refunded by the power company pursuant to the capacity auction true-up under section 39.262(d)(2).[89]

H

This Court did not resolve the issue now before us in In re TXU.[90]  That case concerned the Commission=s reversal of early mitigation efforts.  A majority of the Court held that mandamus relief was unavailable and did not reach the merits of the controversy.  The dissent in that case would have reached the merits of whether the Commission had the statutory authority to reverse early mitigation efforts before a final true-up.  At one juncture, the dissent said, A[t]he Legislature has required early mitigation of stranded costs not because those costs actually exist now but because it has been estimated that they will exist after the 2004 true-up and waiting until then to begin recovery threatens competition.@[91] Read in context, the dissent was explaining that Athe unquestioned fact is that stranded costs cannot be determined with any accuracy until one knows what the retail price of electricity is in a competitive market, and no such market exists.@[92] Neither the Court nor the dissent purported to decide whether the Commission could require a two or more year gap in recovery of carrying costs on stranded costs.

I

Importantly, neither the Commission nor TIEC has offered any rationale to explain why the Legislature chose to use the book value of generation assets on December 31, 2001 (or even earlier) in calculating stranded costs if it intended for stranded costs to come into existence only after a final true-up proceeding in 2004 or beyond.  But conversely, there is a compelling reason to determine the amount of stranded costs that existed as of December 31, 2001 and yet use the market value in 2002, 2003, or 2004 of the stranded assets. That compelling reason is that the Legislature knew with certainty that there would be no valid market indicators on December 31, 2001, the day before customer choice began, or for up to two years thereafter.

The fact that the Legislature permits the actual market value of assets in 2002, 2003, or 2004,[93] or 2004 projections of the value of nuclear assets,[94] to be used to calculate the value of stranded costs that existed as of December 31, 2001 is entirely consistent with the rationale underlying the capacity auction true-up proceeding,[95] to which we now turn.

IV

It is no coincidence that the capacity auction true-up proceeding[96] covers roughly the same period of time between the start of customer choice, January 1, 2002, and the date on which generation companies could first file to finalize stranded costs in a true-up proceeding, which was January 11, 2004.[97]  By definition, stranded costs include generation assets= excess book value over market.[98]  The Legislature recognized that on the first day of deregulation, January 1, 2002, there was no way to validly quantify stranded costs, if any, because a market for electricity, both wholesale and retail, would need time to develop, and there would be interim distortions and fluctuations, perhaps severe ones.  The Legislature was also concerned that distortions and fluctuations in the market price of power during the first two years of deregulation could harm consumers and generation companies alike.  The Legislature accordingly designed the capacity auction true-up proceeding because of the likelihood that no stable market would exist until up to two years after the first day of deregulation.

There are two objectives accomplished by the capacity auction true-up proceeding that are pertinent to this appeal.  The first is that a generation company is limited to a set margin that it will receive for sales of power, no matter how high or how low gas prices and fuel costs might be during 2002 and 2003.  The second is that a generation company is permitted to earn a return on its generation assets during this period. What cannot be determined from this record is how much of that return is a return of or on stranded costs.

Section 39.153 requires a generation company to auction entitlements to at least 15 percent of its total power generation capacity, commencing at least 60 days before the beginning of customer choice.[99]  This auction obligation continues until the earlier of 60 months (five years) after the beginning of customer choice or the date the Commission determines that 40 percent or more of the electric power consumed by residential and small commercial customers within the affiliated transmission and distribution company=s service area before the onset of customer choice is provided by nonaffiliated retail electric providers.[100]

At the end of the first two years that this auction obligation is in effect (essentially 2002 and 2003), as part of the true-up proceeding in section 39.262,[101] a determination will be made of the difference between the price of power obtained through the capacity auctions and the power cost projections that were employed in the 2001 ECOM model for the years 2002 and 2003 to estimate stranded costs under section 39.201 (which determined whether there would be competition transition charges).[102]  This essentially guarantees consumers and power companies that the power company will receive no more and no less than a margin predetermined by the Commission in 2001 when the ECOM model was run in compliance with section 39.201.  The former electric utility=s fuel balance determined under section 39.202(c) (which is not at issue in this appeal) is netted with this margin.[103]  If the sum of these two items shows that the power company has overrecovered, the transmission and distribution utility is credited.[104]  If it shows the power company has underrecovered, the transmission and distribution utility is billed.[105]

The court of appeals held that the Commission erred by requiring in Rule 25.263 that any amount owed to the power company resulting from the calculation under subsection 39.262(d) be netted against any Anegative@ stranded cost calculation.[106]  No one has appealed that ruling.  The generation companies contend, however, that the court of appeals= holding forecloses any comparison of the capacity auction true-up to the stranded cost calculation.  The Commission and TIEC counter that the margin guaranteed by the capacity auction true-up was intended by the Legislature to be the only means of recovering any part of stranded costs between January 1, 2002 and the date of a final order in a true-up proceeding.  The correct construction of the Act lies between these two polar positions.

In the rulemaking proceeding that led to the adoption of Rule 25.263, the power companies, TIEC, and others disputed how the capacity auction true-up determination should be made.  The same order that decided that carrying costs on stranded costs should be recoverable only from the date of a final true-up order also decided how the capacity auction true-up would be calculated.[107]  The Commission essentially accepted the power companies= position regarding the calculation of the capacity auction true-up.[108]  The Commission determined that actually re-running the ECOM model was not required by section 39.262(d) of the PURA, but instead, that it was appropriate to use Aaggregated capacity auction revenues, actual fuel costs, and sales amounts [which] are compared to data from the ECOM model.@[109]

It is not clear from the Commission=s order in this rulemaking proceeding precisely what is calculated by the capacity auction true-up, but filings by the power companies as part of the process do shed some light on the matter.  Reliant Energy, Inc., now known as CenterPoint, said in written public comments (part of the record in this case) that the capacity auction true-up calculation resulted in a Amargin predicted to be available to contribute to fixed costs and therefore to reduce stranded costs.@  Reliant explained in greater detail:

For purposes of the true-up, the ECOM model has two main components: the price of power and the price of fuel.  The difference between those components is the margin predicted to be available to contribute to fixed costs and therefore to reduce stranded costs.  Assume, for example, that the ECOM price of power is $43/mwh, and the ECOM price of gas is $33/mwh.  The margin that is available to reduce stranded costs in this example is $10/mwh.

The capacity auction will also yield a price of power and a price of fuel.  The purpose of the PURA ' 39.262(d)(2) true-up is to ensure that the [power generation company] ultimately receives the same margin from the capacity auction process as the ECOM model predicted.  The [power generation company] may recover part of, all of, or more than that ECOM margin through the bid premiums.  In addition, the [power generation company] will experience some gain or loss on fuel when the capacity auction strike prices are compared to the [power generation company=s] actual costs.  The remainder (or overcollection) of the margin should be recovered from (or paid back to) ratepayers in the true-up proceeding.  Thus, at the time of the true-up, the [power generation company] can be made whole by the following formula:

(ECOM market revenues - ECOM fuel costs) - ((capacity auction price x total busbar sales) - actual fuel costs)

Maintaining the assumption that the margin between the ECOM price of power and the ECOM price of gas is $10/mwh, the [power generation company] should retain that margin in the capacity auction true-up, assuming sales remain the same.  For example, suppose the capacity auction price is composed of a $2/mwh bid premium and a $33/mwh fuel cost, for a total capacity auction price of $35/mwh. Assuming that the actual fuel cost is $33/mwh, the [power generation company] would recover from the entitlement holder all of its fuel costs and $2/mwh to apply against stranded costs. But to retain the net margin of $10/mwh in the ECOM model, the [power generation company] should be allowed to recover $8/mwh from ratepayers.

This method could work to the benefit of ratepayers as well.  For example, assume that the capacity auction price was $42/mwh and the price of gas was $30/mwh.  In that instance, the [power generation company] would overrecover its expected margin by $2/mwh and would owe that amount to ratepayers.        

The Commission adopted a formula for calculating the capacity auction true-up amount that is substantially the same as that proposed by Reliant, except that the Commission limited the true-up to the years 2002 and 2003, omitting any calculation for the months in 2004 before a final true-up order is issued for each power company.  No one challenges Rule 25.263 with regard to the capacity auction true-up calculation.

The capacity auction true-up calculation will be company-specific, based on a margin developed in each company=s unbundled cost of service (UCOS) proceeding.[110]  This information appears to be confidential because of competition concerns.  It is not part of our record and is not available on the Commission=s website.

What can be gleaned from the record in this proceeding is that some portion of the margin that results from the capacity auction true-up may contain a component that allows a return of or on stranded costs. The court of appeals held that if a power company is entitled to bill the transmission and distribution utility for the amount that netting the final fuel balance and capacity auction true-up yields, then that amount cannot be netted against a stranded cost calculation that results in a negative number.[111]  The court of appeals= determination has not been challenged in this Court and is final.  However, that determination does not foreclose the Commission from taking into account any return of or on stranded costs that the margin from the capacity auction true-up contains in determining the appropriate carrying costs on stranded costs.  Section 39.262, which addresses true-up proceedings, provides at the outset that A[a]n electric utility . . . may not be permitted to overrecover stranded costs through the procedures established by this section or through the application of the measures provided by the other sections of this chapter.@[112]  In setting a competition transition charge or allowing securitization of stranded costs at the conclusion of a final true-up proceeding, the Commission can ensure that there is no overrecovery of stranded costs or carrying costs on stranded costs if the capacity auction true-up margin has already provided a return of or on part of those stranded costs.[113]

TIEC is incorrect when it contends that the margin yielded in the ECOM model worksheets for each company with regard to the capacity auction true-up was intended by the Legislature to be the only means of recovering carrying costs on stranded costs until 2004.  Sections 39.201 and 39.262(d) contemplate that a company may recover both competition transition charges from January 1, 2002, as well as the margin contemplated in the capacity auction true-up.[114]  As recognized by the court of appeals, Athe [L]egislature chose not to include [the capacity auction true-up amount] in its definition of stranded costs or to incorporate it into the methods it prescribes for calculating stranded costs.@[115]  But there may be some overlap of recovery of carrying costs on stranded costs under these sections.  The extent to which carrying costs on stranded costs have been recovered in the margin provided by the capacity auction true-up for 2002 and 2003 remains to be determined.

* * * * *

For the reasons considered above, we hold that Rule 25.263(l)(3) is invalid, and we remand this proceeding to the Commission for further consideration.

_____________________________________

Priscilla R. Owen

Justice

OPINION DELIVERED: June 18, 2004

---

^[1] See generally City of Corpus Christi v. Pub. Util. Comm=n, 51 S.W.3d 231, 238 (Tex. 2001).

^[2] See generally Chapter 39 of the Texas Public Utility Regulatory Act, Tex. Util. Code '' 39.001-.910.

^[3] Id. ' 39.001(b).

^[4] Id. ' 39.251(7).

^[5] Id. ' 39.001(b)(2); see also City of Corpus Christi, 51 S.W.3d at 237-38.

^[6] Tex. Util. Code ' 39.001(b)(2); see also City of Corpus Christi, 51 S.W.3d at 238.

^[7] Tex. Util. Code ' 39.252(a).

^[8] Id. '' 39.201, .251-.254, .256-.265, .301-.313.

^[9] Id. '' 39.201(k), .262(c).

^[10] 16 Tex. Admin. Code ' 25.263(l)(3).

^[11] 101 S.W.3d 129, 146-47.

^[12] See Tex. Pub. Util. Comm=n, Application of Reliant Energy for Approval of Unbundled Cost of Service Rate Pursuant to PURA ' 39.201 and Public Utility Commission Substantive Rule ' 25.344, Docket No. 22355 (Oct. 4, 2001) (order), available at http://interchange.puc.state.tx.us (accessed June 17, 2004); Tex. Pub. Util. Comm=n, Application of Central Power & Light Company for Approval of Unbundled Cost of Service Rate Pursuant to PURA ' 39.201 and Public Utility Commission Substantive Rule ' 25.344, Docket No. 22352 (Oct. 5, 2001) (order), available at http://interchange.puc.state.tx.us (accessed June 17, 2004).

^[13] 101 S.W.3d at 147.

^[14] Id.

^[15] Tex. Util. Code ' 39.252(a).

^[16] Id. ' 39.201(g); see also id. '' 39.251(7), .201(l).

^[17] Id. ' 39.262(d).

^[18] Id. ' 39.262(a) (utilities may not be permitted to overrecover stranded costs).

^[19] 26 Tex. Reg. 10498, 10520 (2001) (citing Tex. Util. Code ' 14.002); see also 26 Tex. Reg. 4359, 4360 (2001) (proposed June 15, 2001) (stating in the Notice of Proposed Rulemaking that Rule 25.263 was proposed under Tex. Util. Code '' 14.002, 39.252, and 39.262).

^[20] Tex. Util. Code ' 14.002.

^[21] 26 Tex. Reg. 10498, 10520 (2001) (citing Tex. Util. Code '' 39.252, .262).

^[22] Tex. Util. Code ' 39.001(e).

^[23] Id.

^[24] Id. ' 39.001(f).

^[25] 101 S.W.3d at 132.

^[26] Tex. Gov=t Code ' 2001.174(2).

^[27] See Pub. Util. Comm=n v. City Pub. Serv. Bd. of San Antonio, 53 S.W.3d 310, 315-16 (Tex. 2001); see also Office of Pub. Util. Counsel v. Pub. Util. Comm=n, 104 S.W.3d 225, 232 (Tex. App.BAustin 2003, no pet.).

^[28] The prior decisions are City of Corpus Christi v. Public Utility Commission, 51 S.W.3d 231 (Tex. 2001), TXU Electric Co. v. Public Utility Commission, 51 S.W.3d 275 (Tex. 2001), and In re TXU Electric Co., 67 S.W.3d 130 (Tex. 2001).

^[29] Tex. Util. Code ' 39.302(5).

^[30] City of Corpus Christi, 51 S.W.3d at 238.

^[31] Tex. Util. Code ' 39.001(b)(2).

^[32] City of Corpus Christi, 51 S.W.3d at 238.

^[33] Tex. Util. Code ' 39.301.

^[34] The rule says:

The TDU [transmission and distribution utility] shall be allowed to recover, or shall be liable for, carrying costs on the true-up balance.  Carrying costs shall be calculated using the utility=s cost of capital established in the utility=s UCOS [unbundled cost of service] proceeding, and shall be calculated for the period of time from the date of the true-up final order until fully recovered.

16 Tex. Admin. Code ' 25.263(l)(3).

^[35] Id.

^[36] Tex. Util. Code ' 39.262(c).

^[37] 26 Tex. Reg. 10498, 10519 (2001).

^[38] 101 S.W.3d at 146-47.

^[39] Tex. Util. Code ' 39.262(d).

^[40] Id. '' 39.201, .251(7), .262(c), .262(h).

^[41] ___ S.W.3d at ___ (Brister, J., dissenting).

^[42] Tex. Util. Code ' 39.001(a).

^[43] Id. '' 39.101(a), .102.

^[44] Id. ' 39.001(b).

^[45] Id. ' 39.052.

^[46] Id. ' 39.051(b).

^[47] Id. ' 39.202.

^[48] Id. ' 39.202(a).

^[49] See, e.g., id. '' 39.254, .262(i) (referring to the Report to the Texas Senate Interim Committee on Electric Utility Restructuring, Potentially Strandable Investment (ECOM) Report: 1998 Update (Apr. 1998)).

^[50] Id. ' 39.254.

^[51] Id. ' 39.257.

^[52] Id. ' 39.254.

^[53] Id. ' 39.256.

^[54] ARegulatory assets@ are defined in section 39.302(5) of the Utilities Code.  See generally City of Corpus Christi v. Pub. Util. Comm=n, 51 S.W.3d 231 (Tex. 2001); TXU Elec. Co. v. Pub. Util. Comm=n, 51 S.W.3d 275 (Tex. 2001).

^[55] Tex. Util. Code '' 39.201(i), .301-.303.

^[56] Id. ' 39.301.

^[57] Id. '' 39.301, .303(a).

^[58] Id. ' 39.301.

^[59] Id.

^[60] Id. ' 39.303(a).

^[61] See generally TXU Elec. Co. v. Pub. Util. Comm=n, 51 S.W.3d 275, 277 (Tex. 2001); City of Corpus Christi v. Pub. Util. Comm=n, 51 S.W.3d 231, 235-36 (Tex. 2001).

^[62] Tex. Util. Code ' 39.201(i)(1).

^[63] Id. '' 39.301, .302(4), .302(7), .303.

^[64] Id. '' 39.201(i), .301; see also generally TXU Elec. Co., 51 S.W.3d at 281-84.

^[65] Tex. Util. Code ' 39.201(a), (c), (d).

^[66] Id. ' 39.201(h) (referring to Utility Code section 39.262(i)).

^[67] Id. ' 39.201(g).  The freeze period is defined in Tex. Util. Code ' 39.052(a) as September 1, 1999 until January 1, 2002.

^[68] Id. ' 39.201(d), (g).

^[69] Id. ' 39.201(l).

^[70] Id.

^[71] Id. ' 39.201(d).

^[72] Id. ' 39.201(k).

^[73] Id. ' 39.201(a)-(k).

^[74] Id. ' 39.201(l).

^[75] Id. '' 39.201(l), .262(g).

^[76] 16 Tex. Admin. Code ' 25.263(g)(2)(A).

^[77] Tex. Util. Code ' 39.201(a), (b), (d), (f), (g), (h), (k).

^[78] Id. ' 39.201(l).

^[79] Id. ' 39.201(g), (h), (l) (emphasis added).

^[80] Id.

^[81] Id. ' 39.262(c); see also id. ' 39.262(h) (describing the methods of quantifying stranded costs Afor the purpose of finalizing the stranded cost estimate used to establish the competition transition charge under Section 39.201@).

^[82] Id. ' 39.201(g).

^[83] Id. ' 39.251(7).

^[84] Id.; see also id. ' 39.262(h) (outlining the methods for quantifying stranded costs to finalize the stranded cost estimate used to establish the competition transition charge).

^[85] Id. ' 39.251(7).

^[86] Id. '' 39.251(7), .262(h).

^[87] Id. ' 39.251(7).

^[88] Id.

^[89] Id. ' 39.262(d)(2).

^[90] 67 S.W.3d 130 (Tex. 2001).

^[91] Id. at 166 (Hecht, J., dissenting).

^[92] Id. at 165-66 (Hecht, J., dissenting).

^[93] See Tex. Util. Code ' 39.262(h).

^[94] See id. ' 39.262(i).

^[95] Id. '' 39.153, .262(d).

^[96] Id.

^[97] Id. ' 39.262(c).

^[98] Id. ' 39.251(7).

^[99] Id. ' 39.153(a).

^[100] Id. ' 39.153(b).

^[101] Id. ' 39.262(d).

^[102] 16 Tex. Admin. Code ' 25.263; see also 26 Tex. Reg. 10498, 10498-501, 10524 (2001).

^[103] Tex. Util. Code ' 39.262(d).

^[104] Id.

^[105] Id.

^[106] 101 S.W.2d at 138-41.

^[107] 26 Tex. Reg. 10498 (2001).

^[108] Id. at 10501.

^[109] Id.

^[110] 16 Tex. Admin. Code ' 25.263(i)(1).

^[111] 101 S.W.3d at 138-41.

^[112] Tex. Util. Code ' 39.262(a).

^[113] Id. '' 39.201(l), .252, .262(c).

^[114] Id. '' 39.201, .262(a), (d).

^[115] 101 S.W.3d at 140.