Summary
expressing disagreement with the Sixth and Tenth Circuits and concluding that "it is simply not a correct reading of Raleigh to say that nonbankruptcy law determines only the abstract validity of the claim" and that " Raleigh is very clear that a creditor's claim `in the first instance' is a function of the nonbankruptcy law giving rise to the claim"
Summary of this case from In re Kaiser Aluminum Corp.Opinion
No. 02-83984-SSM
December 29, 2003
John K. Lyons, John Wm. Butler, Jr., Skadden, Arps, Slate, Meagher Flom, Chicago, IL, Lawrence E. Rifken, J. Eric Crupi, James H. Lister, McGuireWoods LLP, McLean, VA, Alexander Williamson Powell, Jr., David E. Carney, Skadden, Arps, Slate, Meagher Flom, Washington, DC, Daniel F. Blanks, Douglas M. Foley, McGuireWoods LLP, Norfolk, VA, John H. Maddock III, Joseph S. Sheerin, Sarah Beckett Boehm, McGuire Woods LLP, Richmond, VA, Michael Arthur Umayam, Fried, Frank, Harris, Shriver Jacobson, Washington, DC, for U.S. Airways Group, Inc.
Dennis J. Early, Office of the U.S. Trustee, Alexandria, VA, U.S. Trustee.
Scott L. Hazan, Brett H. Miller, Otterbourg Steindler Houston Rosen PC, New York City, Byron L. Pickard, Malcolm M. Mitchell, Jr., Vorys, Sater, Seymour and Pease LLP, Alexandria, VA, for Official Committee of Unsecured Creditors Creditor Committee.
MEMORANDUM OPINION
Before the court are the objections of the reorganized debtors and the post-confirmation creditors committee to the $2.084 billion claim of the Pension Benefit Guaranty Corporation ("PBGC"). The claim arises from the termination of the defined-benefit pension plan for the pilots of U.S. Airways, Inc. ("US Airways"). As a result of the termination, the liabilities of the plan have been taken over by the PBGC. The debtors and the committee contend that the PBGC claim is approximately three times greater than the amount the PBGC actually needs to pay the pilots their promised benefits. The central point of dispute is whether the value of the unfunded future benefits should be determined by applying the PBGC's own valuation regulation, or whether the court instead should independently discount those benefits to present value using a hypothetical "prudent investor" rate of return and an expected retirement age ("XRA") reflecting the financial disincentives for pilots to retire early. For the reasons stated, the court determines that the valuation regulation is controlling and that the PBGC's claim should be allowed as filed.
With a few exceptions, the dollar figures in this opinion have been rounded to the nearest million to avoid losing the forest for the trees.
Findings of Fact A. US Airways files for chapter 11 reorganization
US Airways, together with its parent holding company and five affiliates, filed voluntary chapter 11 petitions in this court on August 11, 2002. A joint plan of reorganization was confirmed on March 18, 2003, and became effective on March 31, 2003. Under that plan, general unsecured creditors — which includes the PBGC — will share pro rata in a pool of stock in the reorganized debtor having a projected value at confirmation of something less than 2 cents on the dollar.
B. Termination of the Pilots' Pension Plan
In the course of the chapter 11 case, U.S. Airways moved the court to approve a "distress" termination of the defined-benefit pension plan it maintained for its pilots. The facts surrounding that motion are set forth at length in the court's written opinion supplementing its oral ruling and need not be repeated in detail. In re U.S. Airways Group, Inc., 296 B.R. 734 (Bankr. E.D.Va. 2003). Suffice it to note that in the October to November 2002 time frame, U.S. Airways identified a serious funding shortfall with respect to the pension plan that it maintained for its pilots. Two factors conspired to create the funding shortfall. First, protracted poor performance by the stock market had resulted in a significant decline in the value of the plan assets. Second, the decline in long-term interest rates to a 40-year historic low had increased the amount of the current liabilities for the plans, since current liabilities are determined based on the cost of an annuity to pay the specified benefit, and that cost in turn rises as long-term interest rates fall. As projected by the plan's actuaries, the required contributions to the pilot's pension plan over the seven years of the debtors' business plan totaled $1.659 billion. This sum, coupled with a dramatic reduction in projected revenues for 2003 and an increase in fuel costs, made the business plan unworkable. One solution investigated by the debtors involved "restoration funding," under which — as the debtors viewed the law — the PBGC could approve a 30-year amortization of the funding shortfall. For the pilots' plan, restoration funding would have required the debtors to pay approximately $122 million per year, or $854 million over the seven years of the business plan, instead of the $1.659 billion that would be required without such relief. Another solution explored by the debtors was a waiver by the Internal Revenue Service ("IRS") of the funding contributions that would have been required in 2004 and subsequent years. Both the PBGC and the IRS refused to approve the relief requested by the debtors, and an attempt at legislative relief likewise proved unsuccessful.
It was in that context that this court made a finding that unless the pilots' pension plan was terminated, the debtors would be unable to pay all its debts pursuant to a plan of reorganization and would be unable to continue in business outside the chapter 11 reorganization process. That finding was a required condition for a "distress" termination of the plan under the Employee Retirement Income Security Act of 1974, 29 U.S.C. § 1001 et seq., ("ERISA"). The debtors did agree with the Air Line Pilots Association — the bargaining representative for the pilots of U.S. Airways — to establish a follow-up defreed-contribution pension plan for the active duty pilots. That plan is funded by contributions that approximate what the debtors would have paid had "restoration funding" been approved and is intended to approximate the benefits that active duty pilots would have received under the terminated plan. However, it provides no benefits to retired pilots.
C. The PBGC Claim
When a pension plan is terminated as a result of a distress termination, the liabilities of the plan are assumed by the PBGC. The PBGC timely filed a contingent claim (Claim No. 2835) on November 4, 2003, in the amount of $2,083,600,000.00 for unfunded liabilities under the pilots' pension plan. The assets of the plan — which all parties agree were worth approximately $1.223 billion on the date the plan was terminated — have been turned over to the PBGC. Using the assumptions in the PBGC's valuation regulation, the PBGC's consulting actuaries have calculated the present value of the plan's liabilities on March 31, 2003 — the date the plan was terminated — as being $3,441 billion. After credit for the assets turned over, the resulting unfunded liability would be $2.219 billion. The debtor's own actuaries concede that the figure is at least that amount if the PBGC assumptions are applied. Those assumptions include use of the 1983 General Annuity Mortality table ("the 1983 GAM") to determine life expectancy, a discount rate of 5.1% for the first 20 years and 5.25% thereafter, and an expected retirement age of 56.
Although the proof of claim was never formally amended after the plan was terminated, the PBGC's consulting actuary later calculated the amount of the claim to be $2,218,596,000 as of the plan termination date. Because the proof of claim had never been amended to reflect the higher figure, the court ruled that the claim would be limited to the amount set forth in the filed proof of claim.
The amount of the unfunded liabilities, as calculated by the debtor's actuary using the assumptions required by the PBGC valuation regulation, is $2.362 billion, or slightly higher than the figure calculated by the PBGC's consulting actuaries.
The debtors, through their actuary, presented calculations that the present value of the plan liabilities would be significantly lower using what the debtors contend are more reasonable assumptions. The debtors used a more recent mortality table ("the 1994 GAM"), a discount rate of 8%, and an expected retirement age of 60. The resulting calculation of plan liabilities as of March 31, 2003, was $2.093 billion. After credit for the plan assets turned over to the PBGC, the unfunded liability, according to the debtors, was no more than $894 million.
D. The PBGC Valuation Regulation
The PBGC first promulgated a regulation to value the liabilities of a terminated pension plan in 1976. 41 Fed. Reg. 48,484 (1976). The regulation was most recently amended in 1993, 58 Fed. Reg. 50,812 (1993), and is currently codified at 29 C.F.R. § 4044.41 to 4044.75. The valuation regulation specifies use of the 1983 GAM, a discount rate which undated monthly to reflect current annuity pricing, and a table of expected retirement age for plan participants. The stated goal of the regulation is to generate a value which will "accurately approximate the cost of [single-premium] group annuity contracts" that would pay the benefits promised under the terminated plan. 58 Fed. Reg. 5128.
To determine that cost, the PBGC conducts what has been described as a "double-blind" quarterly survey of current prices charged by insurance companies for single premium annuities. Because insurance companies consider such pricing information highly sensitive, the PBGC relies on a trade association, the American Council of Life Insurers ("ACLI"), to send the survey forms to the insurance companies. ACLI's involvement is purely voluntary on its part, there being no written contract between it and the PBGC, nor any payment for its services. Return by the companies of the completed survey forms is also voluntary. The survey responses are returned in a sealed envelope which ACLI forwards, unopened, to the PBGC. The survey responses are identified only by a code letter. The PBGC does not know either the identity of the companies to which the surveys are sent or the responding companies, and ACLI never sees the contents of the responses. The number of surveys sent out each quarter is in the "mid-twenties," and the number of actual responses varies from quarter to quarter but is typically between 9 and 12. Although the PBGC applies certain informal tests to weed out inconsistent or unlikely data ("outliers"), there is no procedure for independently verifying that the information provided on the survey responses is accurate. The responses are nevertheless remarkably consistent. To take just one example, the reported prices for the September 2002 survey for an annuity paying $10 per month for life beginning immediately at age 60 were as follows: $1,450.91; $1,436.59; $1,485.58; $1,465.62; $1,492.61; $1,434.60; $1,582.00; and $1,482.00.
In order to establish the discount rate for the valuation regulation, the PBGC works backwards from the pricing information in the survey. That is, the PBGC first calculates the average price reported on the surveys for single-premium annuities at a specific set of ages. It then calculates, by an iterative process, what interest rate would most closely result in those prices under the mortality assumptions of the 1983 GAM. Specifically, it takes a range of plausible interest rates and calculates what the cost of an annuity would be if each rate were used in conjunction with the 1983 GAM and the PBGC's expected retirement age table. The interest rate that comes closest to generating a cost matching the average reported cost from the survey is then adopted as the discount rate. The survey responses for June 30th and September 30th of each year are used to calculate the published discount rate , to be used on January 1st of the following year. The rate is updated monthly throughout the year in the same proportion as the change in the Moody's corporate bond index rate. As noted, the discount rate in effect on March 31, 2003, was 5.1% for the first 20 years and 5.25% thereafter. This discount rate, it must be noted, is not necessarily the rate used by insurance companies when pricing annuities. A commercial insurance company is likely to use a more recent mortality table, such as the 1994 GAM, in pricing annuities. The price of an annuity is a function both of mortality assumptions and of discount rates. The PBGC's goal is not to discover the discount rate actually used by the insurers, but rather to arrive at a discount rate that, combined with the mortality assumptions embodied in the 1983 GAM, give the same price as the prices currently being quoted by insurance companies. Thus, the interest rate cannot be viewed in isolation. Put another way, the question is not whether 5.1% for 20 years and 5.25% thereafter is an appropriate discount rate but whether those rates, when used with the 1983 GAM, give an annuity pricing that is reflective of the market place on the date the U.S. Airways pilots' pension plan was terminated.
E. The "Prudent Investor" rate
The debtors presented the testimony of Howard Mark Crane, an investment consultant with Watson, Wyatt Company, most of whose clients are pension plan sponsors, to establish that the expected long-term rate of return that could be earned by a "prudent investor" on an asset pool consisting of 60% stocks and 40% bonds — which he testified was typical of large pension plans — was in the range of 7.1% to 9.1%, with an adjusted mid-point of 8%. This figure was derived from a survey of investment expectations by professional fund managers and was then tested by running a series of 15,000 "stochastic" simulations to determine whether — given the random fluctuation in actual returns that might be expected over a 30-year period — the plan assets would be sufficient to make the required payments when due. Using an interest rate of 8% gave a greater than 50% probability that the assets would be sufficient. The debtor further presented evidence that the PBGC itself, in administering the trust fund which holds the assets of terminated pension plans, invests in an asset pool of approximately 80% stocks, 10% short-term fixed income instruments, and 10% cash. Finally, the debtor presented evidence that over the last 20 years the PBGC has achieved an annualized rate of return on its invested trust funds of approximately 10%, and that it has received projections from its investment advisors of an 8% long-term future rate of return on equities.
The PBGC has a separate revolving fund which holds the insurance premiums collected by the PBGC from pension plan sponsors. The revolving fund invests only in government bonds.
The PBGC in turn presented testimony from Dennis E. Logue, the dean of the business school at the University of Oklahoma, that an economist would not value a future contingent liability by reference to the expected rate of return on assets set aside to meet that liability. Rather, according to Dean Logue, a liability should be valued at the amount a willing buyer would pay a willing seller to assume the liability. That view was echoed by Victor Medugno, a consulting actuary with experience in pricing annuities. Mr. Medugno testified that there is an active market for close-out annuities for terminating pension plans. He further testified that he had conducted a published study for the Society of Actuaries in an effort to find a proxy for the PBGC rate and concluded that the two rates that most closely tracked the PBGC rate over time were the 30-year swap rate and the 30-year Fannie Mae bond yield. These were, respectively, 5.13% and 5.48% on the date the U.S. Airways plan was terminated. Mr. Medugno further opined that the liability values calculated by reference to the PBGC regulation were "reasonably close" to the prices that a commercial insurance company would charge for an equivalent annuity. In reaching this opinion, he relied on his own general experience as well as a study that had been conducted by the American Academy of Actuaries dated May 4, 2000. The stated purpose of that study was "to compare, for a sample of transactions, actual costs to terminate a plan with those that would have resulted had PBGC assumptions been used." The study concluded that the PBGC assumptions resulted in a "relatively small overstatement of termination liability (averaging 3% to 4%) . . . apparently primarily due to mild interest rate conservatism.'' However, the terminations examined by the study all involved plans that were considerably smaller than the U.S. Airways pilots' plan (the highest reported termination cost being $30.3 million and the mean being less than $5 million). Additionally, Mr. Medugno conceded that except in very small cases the final price actually charged for close-out annuities is usually set by negotiation.
F. The Expected Retirement Age ("XRA")
As noted, the cost of a group retirement annuity is influenced by the expected age at which employees will retire. The higher the age, the lower the cost of an annuity. It is undisputed that U.S. Airways pilots may retire as early as age 50. The PBGC valuation regulation includes a table developed by the PBGC setting forth retirement age assumptions based on the amount of the participant's benefit. 29 C.F.R. § 4044 App. D. Generally, the higher the retirement benefit, the earlier the expected retirement age. For the pool represented by the U.S. Airways pilots, that assumption is 56 years of age.
US Airways presented the opinion testimony of Gerald A. Glass, its vice-president for employment relations, and Mark Taylor Dungan, its actuary, that a more reasonable expectation of the age at which its pilots were likely to retire was 60, which also happens to be the mandatory retirement age for pilots. The reason the pilots are not likely to retire any earlier than they must, according to these witnesses, is the strong financial incentive to keep working as long as possible. First, pilots are, in general, a highly-compensated group. Additionally, in the airline industry seniority is everything in terms of salary, schedules, and routes. A pilot could not leave U.S. Airways and fly for another carrier without taking a substantial cut in pay and intangible benefits (schedules, choice of routes, etc.). Finally, the structure of the defined-contribution plan that was created to replace the terminated plan is such that pilots will not come anywhere near recouping the benefits they would have received under the terminated plan unless they work until age 60.
The PBGC presented evidence that from 1998 through 2002, a total of 728 U.S. Airways pilots retired, and that 323 (or approximately 44%) of those were less than 60 at the date of retirement. Even after excluding pilots who retired under an early-retirement program ("window") no longer in effect, 63 of 340 pilots, or nearly 19%, were under 60 when they retired. Since the defined benefit plan was terminated, 93 active duty pilots have retired, of whom 10 were less than age 60. Although this amounts to 11% of those retiring from active duty, most were nevertheless close enough to age 60 that they brought the mean retirement age down to only 59. Even Mr. Medugno, although he thought U.S. Airways's prospects were sufficiently clouded that a significant number of pilots might retire prior to age 60, was of the opinion that 58 was a more probable average retirement age than 56.
An additional 65 retired from furlough stares. Fully 59 of these were under age 60, with the mean age of retirement being 57.
Conclusions of Law and Discussion
The central point of contention is whether the PBGC's claim for the unfunded liabilities should be determined by applying its own valuation regulation, or whether the court is free to make an independent determination based on its own findings as to an appropriate discount rate and likely retirement age. The debtors' position, in a nutshell, is that a bankruptcy court is not legally bound by the valuation regulation and is always free to make its own finding in order to prevent the PBGC from receiving a "windfall" and to ensure equal treatment of creditor claims. Additionally, the debtors assert that the valuation regulation — at least as implemented — is invalid because it improperly delegates duties to an industry trade group and relies on an unscientific anonymous survey. Each of these arguments will be examined in turn.
The "windfall" rhetoric must be taken with a large grain of salt. Since claims are being paid with stock having an estimated value equal to less than two cents for each dollar in allowed claims, full allowance of the PBGC's claim could never result in the PBGC receiving more than the amount of its actual loss.
In a pre-trial brief filed 9 days prior to the evidentiary hearing, the debtors for the first time attempted to raise an additional defense that the PBGC failed to mitigate its damages when it refused to allow "restoration funding" of the pilots pension plan. The court declined to allow the late assertion of that defense based on a finding of unfair prejudice to the PBGC. Without intimating any view as to the possible merit of such a defense had it been timely asserted, the court notes it would hardly be fair to impugn to the pilots — whose non-guaranteed benefits make up a substantial portion of the PBGC claim — the sins (if any) of the PBGC.
A. Whether the valuation regulation should control
Pension plans governed by ERISA may be voluntarily terminated in one of two ways. In a "standard termination" under 29 U.S.C. § 1341(b), the employer must distribute the plan assets to plan participants in the form of annuities purchased from a private insurer. In a "distress termination" under 29 U.S.C. § 1341(c), the plan assets are turned over to the PBGC, which assumes responsibility for paying the promised benefits under the plan. Although there is a relatively modest amount the PBGC guarantees to pay, the remaining non-guaranteed portion will be paid by the PBGC to the extent the funds turned over or subsequently recovered by the PBGC are sufficient to do so. Indeed, in the present case, approximately 73% of the amount the PBCG recovers on its claim will go to pay the non-guaranteed portion of the promised benefits.
Upon a distress termination, the employer and each member of its controlled group become jointly and severally liable to the PBGC for the amount of the "unfunded benefit liabilities." 29 U.S.C. § 1362(b)(1)(A)-(B). The term "unfunded benefit liabilities" is specifically defined by ERISA as the excess of the benefit liabilities "determined on the basis of assumptions prescribed by the [PBGC]" over the value of the plan assets on the date of termination. 29 U.S.C. § 1301(a)(18)(A). The PBGC's argument, put simply, is that since ERISA defines an employer's liability in terms of the valuation regulation, the regulation should control for claim allowance purpose. As support for its position, it points to the Supreme Court's opinion in Raleigh v. Illinois Dep't of Revenue, 530 U.S. 15, 120 S.Ct. 1951, 147 L.Ed.2d 13 (2000). The issue in that case involved the burden of proof on an objection by a chapter 7 trustee to a proof of claim for state taxes. In deciding that the standard of proof under state law applied, Justice Souter, writing for the Court, observed, "Creditors' entitlements in bankruptcy arise in the first instance from the underlying substantive law creating the debtor's obligation, subject to any qualifying or contrary provisions of the Bankruptcy Code." 530 U.S. at 20, 120 S.Ct. at 1955. See also IRS v. Levy (In re Landbank Equity Corp.), 973 F.2d 265, 270 (4th Cir. 1992) ("[T]he Bankruptcy Code . . . does not endeavor to supplant the substantive law under which the claim against the estate . . . arose.").
The debtors, for their part, rely on two opinions at the federal court of appeals level — one pre-dating Raleigh, and the other post-dating it — that rejected application of the PBGC valuation regulation in favor of a present value computation that applied a "prudent investor" discount rate to the future liabilities. In re CF I Fabricators of Utah, Inc., 150 F.3d 1293 (10th Cir. 1998), cert. denied 526 U.S. 1145, 119 S.Ct. 2020, 143 L.Ed.2d 1032 (1999); In re CSC Indus., Inc., 232 F.3d 505 (6th Cir. 2000), cert. denied 534 U.S. 819, 122 S.Ct. 50, 151 L.Ed.2d 20 (2001). In CF I Fabricators, the Tenth Circuit observed that ERISA's definition of "unfunded benefit liability" by its terms applied "`for the purposes of this title [ERISA]' only" and therefore did not necessarily extend to bankruptcy. 150 F.3d at 1301. It therefore held that the district court properly determined the present value of the unfunded liabilities to be $124 million rather than the $223 million amount that would have resulted from using the valuation regulation, reasoning that using the PBGC assumptions would cause its claim to be treated differently than other claims in the same class in violation of § 1123(a)(4), Bankruptcy Code. Id. In CSC Industries the Sixth Circuit similarly held that the bankruptcy court's "authority . . . to determine the amount of claims in the bankruptcy proceedings and treat creditors in the same class equally gives it the authority to value unfunded benefit liabilities claims using a `prudent investor rate.'" 232 F.3d at 509. It rationalized that holding with Raleigh by explaining that although nonbankruptcy law governed the validity of a claim, "bankruptcy courts have the statutory authority to determine the allowability and amount of the claim." Id. (emphasis in original). On that basis it upheld a ruling that valued the PBGC's unfunded benefit claim at $1.8 million based on a 10% "prudent investor" discount rate rather than the $49.7 million value that would result from use of the PBGC valuation regulation.
This court must respectfully disagree with the Sixth and Tenth Circuits. Raleigh is very clear that a creditor's claim "in the first instance" is a function of the nonbankruptcy law giving rise to the claim. It is true, of course, that bankruptcy law determines how the claim will ultimately be paid (if at all). Thus, creditors with allowed claims may nevertheless be paid only pennies on the dollar. Such indeed is the treatment of claims under the debtor's confirmed plan in the present case. Additionally, creditors who file late claims or fail to disgorge an avoidable transfer may receive nothing even though their claims are otherwise valid under non-bankruptcy law. §§ 502(b)(9) and 502(d), Bankruptcy Code. Finally, an enforceable claim under nonbankruptcy law may be subordinated to other claims in bankruptcy on equitable grounds and thereby receive a reduced or no distribution. § 510(c), Bankruptcy Code. But it is simply not a correct reading of Raleigh to say that non-bankruptcy law determines only the abstract validity of the claim — that is, whether the debtor has some liability to the creditor — as divorced from the amount of the claim. Here, both the debtor's liability to the PBGC and the amount of that liability are not only creatures of statute, but of the same statute.
It is, of course, a common-place that "courts of bankruptcy are essentially courts of equity, and their proceedings inherently proceedings in equity." Local Loan Co. v. Hunt, 292 U.S. 234, 240, 54 S.Ct. 695, 697, 78 L.Ed. 1230 (1934). These equitable powers, as the Supreme Court has explained, "have been invoked to the end that fraud will not prevail, that substance will not give way to form, [and] that technical, considerations will not prevent substantial justice from being done," Pepper v. Litton, 308 U.S. 295, 305, 60 S.Ct. 238, 244, 84 L.Ed. 281 (1939). As a result, "[a] bankruptcy court in passing on allowance of claims sits as a court of equity," and in the exercise of its equitable jurisdiction "has the power to sift the circumstances surrounding any claim to see that injustice or unfairness is not done in administration of the bankrupt estate." Id. at 307-08, 60 S.Ct. at 245-46.
A bankruptcy court's equitable powers must nevertheless be exercised within the contours of the Bankruptcy Code. Norwest Bank Worthington v. Ahlers, 485 U.S. 197, 206, 108 S.Ct. 963, 969, 99 L.Ed.2d 169 ("[W]hatever equitable powers remain in the Bankruptcy Courts must and can only be exercised within the confines of the Bankruptcy Code"); see also In re Lapiana, 909 F.2d 221, 223-24 (7th Cir. 1990) ("[B]ankruptcy, despite its equity pedigree, is a procedure for enforcing pre-bankruptcy entitlements under specified terms and conditions[.]") (emphasis added). Under the Code, when an objection is made to a claim, the court, after notice and a hearing, "shall determine the amount of such claim in lawful currency of the United States as of the date of the filing of the petition, and shall allow such claim in such amount[i]" § 502(b), Bankruptcy Code. In fixing the amount of a claim "as of the date of the filing of the petition," there is no dispute that the court must discount future damages to present value. Thus, if the PBGC held merely a common-law indemnity claim for future benefits owed to the pilots, there can be little doubt that this court would have full authority to determine an appropriate discount rate as well as related assumptions as to mortality and expected retirement age. In this connection, 8% would hardly be an unreasonable choice for a discount rate. The court notes, for example, that the Code of Virginia uses an 8% discount rate for the purpose of commuting a life estate to present value. Va. Code Ann. § 55-269.1 et seq.
The distinction here is that the court is not simply valuing a contingent future loss. Rather, Congress, by statute, has expressly given the PBGC a present right to recover an amount determined in accordance with the valuation regulation. Nor can this court agree with the debtors and the committee that application of the regulation would somehow offend the bankruptcy policy requiring that all claims in the same class be treated equally. There is nothing in the record to suggest that the other claims in the general unsecured creditor class are being allowed in an amount different from what the creditor would be owed under applicable non-bankruptcy law. So long as all claims are determined in accordance with applicable nonbankruptcy law, there cannot be any genuine issue of disparate treatment.
B. Does the valuation regulation dramatically overstate the PBGC's actual loss?
At bottom, the "unequal treatment" argument rests on the premise that the real loss likely to be suffered by the PBGC is so much less than the amount calculated using the valuation regulation that the PBGC will recover a hugely greater percentage of its loss than will other unsecured creditors. Accordingly, it is necessary to examine the debtors' contention that using the PBGC assumptions overstates by a factor of nearly three the actual amount needed by the PBGC to pay the benefits that were promised under the terminated plan.
The reason this is so, the debtors argue, is that $894 million, added to the $1.223 billion in assets already turned over, will — if "prudently" invested — provide sufficient funds to pay the promised benefits as they become due. As defined by the debtors, a "prudent" investment strategy is not a risk-free strategy, but rather one that assumes some risk for a greater rate of return. In particular, it relies on an investment asset pool that includes a significant amount of equities on the ground that over a sufficiently long period of time the rate of return on stocks is higher than would be realized by investing solely in risk-free instruments such as government bonds. Such an asset pool, according to Mr. Crane, would likely achieve a long-term rate of return of 8% per year, and it is this rate that the debtors have dubbed a "prudent investor" rate of return.
As the debtors' own actuary acknowledged, the term "prudent investor rate" is not one used by actuaries. The term appears to have its genesis in a thoughtful opinion written by then-Chief Bankruptcy Judge Burton R. Lifland in In re Chateaugay Corp., 126 B.R. 165 (Bankr.S.D.N.Y. 1991), vacated, op. withdrawn sub nom. The LTV Corp. v. PBGC, 1993 WL 388809 (S.D.N.Y. 1993). As in the present case, the issue was how to value the PBGC's claim for unfunded benefits under a terminated pension plan. Judge Lifland had previously ruled that "bankruptcy law controls the applicable discount rate to be used in determining the value of a claim based on . . . liability for cash payments subsequent to the [f]iling [d]ate." Id. at 171. Evidence was presented in support of five approaches to determining an appropriate rate. The PBGC first presented testimony that application of its regulation resulted in the full economic value of its claim. It also presented evidence as to the price that would have been charged by insurance companies, in a competitive bidding situation, for an annuity contract to meet the benefit obligations. Finally, it presented evidence as to the risk-free rate of return on a long-term treasury bond. The debtors presented evidence in support of a "risk-adjusted discount rate" that attempted to measure "the risk that [the debtor] would not meet its obligations . . . by reference to the interest rates observed in the market for bonds of similar duration and priority issued [by the debtor] at the same time as the pension obligations arose." Id. at 171-72. Finally, the creditors committee presented evidence in favor of using "the rate of return achievable by a reasonable, prudent, long-term [pension fund] investor who seeks to achieve the best long-term return on his investment consistent with preserving his capital and minimizing risk," an approach Judge Lifland termed the "pension fund portfolio theory." Id. at 172. The hypothetical "prudent investor" who lies at the heart of this last approach "would choose investments-which would produce a yield reflecting thereturns achievable in the market at a whole" — in other words, "an `average' investor earning `average' returns" — and "would invest in a `prudent' portfolio of diversified investments to minimize risk." Id. at 176. The committee's expert used the "average asset allocation of private pension funds" as "most accurately reflect[ing] the risk profile of the reasonable prudent investor" and calculated an expected rate of return based on the four most recent years of historical returns. Id. at 176-77.
Because the statute was written differently then, the PBGC's claim in Chateaugay was only for the guaranteed benefits.
In its pretrial brief, the PBGC observes with some force that it was precisely the debtors' pursuit of such an investment strategy that caused the pilots pension plan to become underfunded and resulted in its termination. Although the stock market had done very well for a number of years, the value of the plan's assets had dropped 13% in the year prior to the plan's termination. Nor was the PBGC — which itself invests heavily in equities in managing the assets of terminated pension plans — immune from the effects of the stock market decline. The losses experienced by the PBGC trust fund — which is invested in a mixture of equities and bonds similar to those of most pension plans — were 26.1% in 2001 and 15.5% in 2002.
Even with those losses (as well as smaller losses in two earlier years), it is true that the PBGC has done pretty well over the long haul. The reported returns on the PBGC's trust fund assets over the 18-year period from 1985 through 2002 are approximately 10% on an annual basis. Thus — assuming a willingness to tolerate some degree of risk — it would not be unreasonable for an investor in pension fund assets to expect an 8% long-term rate of return. Even the PBGC witnesses admitted that the administrator of an ongoing pension plan would breach no duty of care by investing in such an asset pool and assuming an 8% long-term rate of return. The difference, argues the PBGC, is that when an ongoing plan experiences investment losses in a particular year, the shortfall can be made up by increased contributions in subsequent years. With a terminated plan, by contrast, there are no future contributions, and thus, only one chance to get it right. That stark reality, according to the PBGC, requires that the liabilities of a terminated pension plan be valued based on a risk-free or nearly risk-free investment strategy.
This view is supported by the testimony of the PBGC's economic experts, who took the position that the fair value of a liability is the amount a willing purchaser would charge in order to assume it, not the amount of money that would have to be invested to pay the liability when it became due. To be sure, expected rates of return are hardly irrelevant. After all, the willing purchasers in the marketplace of pension liabilities are insurance companies. Insurance companies invest the premiums they receive and must necessarily make projections of investment returns when they price their policies, since it is out of those earnings that benefits will be paid.
The real issue is one of risk. Annuity issuers base their pricing on returns offered by low-risk investments (typically high-quality corporate bonds). Those returns are lower than the returns that might be achieved by investing in the stock market. The stock market, however, is highly volatile and far from certain, as amply demonstrated by both the debtors' and the PBGC's investment results in 2001 and 2002. While it is easy to run computer simulations, the simple fact is that no one can predict with certainty what returns the stock market will produce over the next 50 years. Given the strong societal interest in protecting pension benefits, a risk-free or nearly risk-free rate to value the pension liability is more appropriate than a rate based on optimistic projections (even if those projections are widely-shared by fund managers) as to the stock market's future long-term performance. Thus, the court is unable to conclude that use of a "prudent investor" rate is necessary to avoid overstating the actual loss resulting from the distress termination.
The limitations of judicial determinations as to a "prudent investor rate" are strikingly illustrated by one of the cases relied upon by the reorganized debtors. In CF I Fabricators, the trial court found that 12.3% was a prudent investor rate in 1992. Of course, in the euphoric climate of 1992 — which, after all, was only a decade ago — perhaps few investment professionals foresaw the sustained bear market of the last few years. But it happened anyway. And, needless to say, the testimony even of the debtors' investment expert, Mr. Crane, does not project 50-year returns anywhere near the 12.3% found in CF I Fabricators.
The use of 56 as an expected retirement age is more problematical. Although that assumption may be valid across a generic range of industries, it does not appear to mesh well with the reality that pilots are well paid; are already required by law to retire at an earlier age than most workers; and cannot, if they retire earlier than they must, readily translate their experience and skills to other employment providing a similar level of income and benefits. The very limited experience in the seven months since termination of the defined-benefit plan and U.S. Airways's emergence from chapter 11 is probably not a sufficient statistical basis upon which to project future early retirements, but the court would be inclined, even being conservative, to find that 58 or 59 is a more reasonable long-term expected retirement age for the cohort represented by the U.S. Airways pilots than 56. Thus, application of the XRA tables in the valuation regulation may mildly overstate the actual amount the PBGC will need to pay the promised benefits under the terminated plan. The testimony presented at the hearing was insufficient for the court to determine what the actual effect would be, in dollars, of using an age 58 or 59 XRA instead of age 56, but the difference would not appear so great as to warrant ignoring the valuation regulation in favor of an independent computation by the court.
C. Whether the valuation regulation is invalid
The debtors take the position that even if this court were otherwise obligated to follow the valuation regulation, the court should not do so in this case because the regulation is invalid, or at least has been implemented in an invalid manner. As noted, the statute defines "unfunded benefit liabilities" as being the amount determined by reference to the PBGC valuation regulation. That regulation was already in effect when the statute was amended to its present form, and the court must therefore presume that Congress knew and approved of the PBGC's general methodology.
No argument is made that adoption of the regulation failed to comply with the notice and comment requirements, 5 U.S.C. § 553(b) and (c), of the Administrative Procedure Act.
The stated goal of the regulation is to determine a value for the unfunded benefit liabilities that approximates the cost of purchasing an annuity contract from a commercial insurer that would pay the same benefits. To determine that amount, the PBGC, with the assistance of the ACLI, conducts an anonymous quarterly survey of annuity prices. That the survey has serious limitations is hardly to be questioned. Because of the anonymous manner the survey is conducted, the PBGC has no way of knowing whether it is receiving responses from market leaders or small players, nor does it have any way of independently verifying that the values reported on the survey forms accurately reflect what the responding company is actually charging for the annuity policies it sells. Indeed, because it does not itself send out the survey forms, the PBGC does not even know the specific universe of companies to which the forms were sent. As the General Accounting Office has recently observed in a report to Congress, the anonymous nature of the survey necessarily creates "ambiguity about the extent to which the PBGC interest rate factors reflect the current broad market for group annuities." These limitations are an unavoidable consequence of the unwillingness of commercial insurance companies to release sensitive pricing information (or at least, to release it in such a form that specific prices can be identified with specific companies). Short of Congress enacting a statute permitting the PBGC to subpoena the information, it is difficult to see how the PBGC could overcome the inherent limitations of the present survey.
The testimony, however, was that small players were more likely to underprice annuities. Thus, any error in including responses from companies without a substantial presence in the market would work in favor of the debtors.
United States General Accounting Office, "Private Pensions: Process Needed to Monitor the Mandated Interest Rate for Pension Calculations" (GAO-03-313, February 27, 2003) at 25, 2003 WL 687114.
But the issue is not whether the survey could be improved. The question is whether, notwithstanding its limitations, the survey is nevertheless a rational way for the PBGC to implement the stated goal of determining a value that approximates the cost of a commercial annuity. Here, considerable deference must be given to the PBGC, as the agency responsible for carrying out a legislatively-delegated function, for determining how best to implement the statutory mandate. Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843-44, 104 S.Ct. 2778, 2782, 81 L.Ed.2d 694 (1984) ("We have long recognized that considerable weight should be accorded to an executive department's construction of a statutory scheme it is entrusted to administer[.]"). This is particularly true given the lack of evidence that the liability values derived from the PBGC regulation differ in any substantial way from the price of commercial annuities. Put another way, while the debtor has pointed to problematical elements in the PBGC methodology, it has presented no evidence that the methodology actually reaches an incorrect result or — more to the point — that U.S. Airways could have purchased a close-out annuity from a commercial insurance company at a price significantly below the value calculated by reference to the PBGC regulation. The best the debtors can show is that no one can be absolutely certain that the survey answers are correct. Under the Administrative Procedure Act, however, an agency's decision must be upheld unless it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2)(A). Given the presumption of validity that must be accorded the PBGC regulation, the debtor has simply not presented sufficient evidence for the court to find that the survey is an irrational or arbitrary means for the PBGC to measure the close-out liability of a terminated pension plan. Accordingly, the court declines to find that the regulation is invalid.
Conclusion
In summary, the court concludes that the PBGC's claim for unfunded benefit liabilities should be determined using the PBGC valuation regulation, since Congress has chosen to define the claim by reference to that regulation. Although the amount calculated under the regulation may exceed the amount a hypothetical "prudent investor" would have to set aside to pay the promised benefits as they became due, the use of a "prudent investor" rate impermissibly shifts the risk of loss from adverse stock market performance — such as led to the termination of the U.S. Airways pilots plan in the first instance — to the retirees. Because the PBGC's valuation regulation — which seeks to replicate the cost of a private-sector annuity paying the promised benefits — gives proper weight to Congress's goal of protecting the health of the nation's private pension system, it is to be preferred over the use of discount rate premised on uncertain projections of future stock market returns. For these reasons, a separate order will be entered allowing the PBGC's claim as filed.