CITIGROUP, INC. v. USA, No. 1:2015cv00953 - Document 195 (Fed. Cl. 2023)
Court Description: REPORTED TRIAL OPINION Status Report due by 7/21/2023. Signed by Senior Judge Eric G. Bruggink. (jpk1) Service on parties made.
Download PDF
<![CDATA[
CITIGROUP, INC. v. USA Doc. 195 In the United States Court of Federal Claims No. 15-953T (Filed: June 29, 2023) ****************** CITIGROUP INC., Plaintiff, v. THE UNITED STATES, Defendant. ****************** Jean A. Pawlow, Washington, DC, for plaintiff. Benjamin C. King, Jr., Trial Attorney, United States Department of Justice, Tax Division, Washington, DC, with whom were David A. Hubbert, Deputy Assistant Attorney General, David I. Pincus, Chief, G. Robson Stewart, Assistant Chief, Stefan R. Wolfe, Trial Attorney, and Jason S. Selmont, Trial Attorney, for defendant. OPINION BRUGGINK, Judge. In 1989, Congress passed the Financial Institutions Reform, Recovery, and Enforcement Act (“FIRREA”) in response to the savings and loan crisis of the 1980s. FIRREA repudiated much of the early efforts of federal regulators to save the industry by incentivizing mergers and acquisitions between weaker banks and stronger ones. This change in the regulatory landscape spawned a series of lawsuits against the federal government that culminated in the Supreme Court’s landmark Windstar Corp. v. United States opinion, which settled the liability question in favor of the banks. Both this court and the Federal Circuit then resolved the damages question in the following years. Dockets.Justia.com The present case is one of the last legal ripples of FIRREA and the Winstar cases. It concerns the tax ramifications of both the damages award to Glendale Federal Bank, FSB (“Glendale”), a predecessor-in-interest to the present plaintiff, Citigroup, and the FIRREA-directed change in treatment of certain goodwill that was created as a result of one of the mergers undertaken by Glendale at the behest of the Federal Savings and Loan Insurance Corporation (“FSLIC”). Trial was held from March 21 to April 1, 2022, to determine whether Glendale had previously deducted amounts awarded as damages by Judge Smith and to determine what plaintiff’s basis was in the intangible rights rendered worthless by FIRREA. We hold that plaintiff is entitled to a deduction for the value of the RAP right and has proven a reasonable value for it and that plaintiff is entitled to a refund of taxes paid on the portion of its earlier damages award that corresponds to recapitalization costs. BACKGROUND In the late 1970s, persistent inflation had reached historic highs in the United States. It was popularly termed “stagflation” due to its effect on broader economic conditions. This prompted the newly appointed chairman of the Federal Reserve, Paul Volker, to declare a “war on inflation.” The Federal Reserve thus began raising interest rates at every opportunity to constrict the money supply and increase the purchasing power of the dollar. This series of rate hikes left certain financial sectors in peril, chief among which was the savings and loan (“S&L”) industry. Savings and Loan banks, also known as “thrifts,” were a subset of the banking industry that concentrated on mortgage lending. The danger posed by the rapid rise in interest rates was that the spread between rates on the long-term, fixed-rate mortgages held by the banks and the interest paid to retail depositors rapidly narrowed and then turned negative. Thrifts had to pay higher rates on deposits than they were receiving from the fixed-rate instruments that they had created over the prior 30 years. In essence, the main source of income for these banks turned into a monthly liability, which quickly challenged the liquidity of many of these thrifts. Recognizing this challenge, federal regulators stepped in. At the time, S&Ls were backstopped by the Federal Savings and Loan Insurance Corporation (“FSLIC”), which was regulated by the Federal Home Loan Bank Board (“FHLBB”).1 FSLIC, facing industry-wide liabilities exceeding 1 Both of these federal institutions were abolished by FIRREA and their 2 its own reserves, sought to stem the tide, in concert with the FHLBB, by encouraging the merger of healthy thrifts with weaker ones. In certain instances, the regulators offered assistance to the acquiring banks—in the form of cash, guarantees, indemnification provisions, and regulatory forbearance—to both induce the transaction and to make the surviving, combined entity a viable enterprise. These mergers/acquisitions were effected through tri-partite agreements between the merging banks and the federal government. One of these “assisted mergers” occurred between Glendale and First Federal Savings and Loan of Broward (“Broward”) on November 19, 1981. A Supervisory Action Agreement (“SAA”) between FSLIC and Glendale memorialized the government’s promises to Glendale. Two items of the FSLIC’s assistance are central to this case: branching rights and a guarantee of a certain regulatory accounting treatment of the resulting transactions, known as the “RAP right.”2 First, the branching rights were regulatory permission for the acquiring bank to operate in the state of the acquired thrift. Before these assisted mergers, federal regulators had allowed S&L institutions to operate only in their home states. As the Winstar trials made clear, branching rights had great value to these thrifts and thus were a significant inducement to merge with banks in other states. E.g., Trial Test. of David Hansen, Tr. A571-72. (ECF No. 123-1) (JX 40) (Glendale FSB, Inc. v. United States, No. 90-772C, Oct. 28, 1997).3 Second, the RAP right was, in short, a promise to protect the acquiring thrifts from the financial impact of any change to how the government viewed the accounting of these deals. Use of the purchase method accounting for the merger resulted in the creation of a “goodwill” asset on Glendale’s books which could be used as capital for regulatory reserve regulatory role transferred to the FDIC or vested in newly created regulatory agencies. “RAP” stands for “regulatory accounting principles,” which is shorthand for the government’s stance on, or as is the case here, acquiescence to, how acquiring banks were recording the thrift mergers on their financial books. 2 3 The deposition and earlier trial testimony of several unavailable witnesses was admitted prior to trial by our order dated February 1, 2022 (ECF No. 140 at 2). Those transcripts were later added to the trial record by order of March 30, 2023 (ECF No. 194). By agreement of the parties they were given joint exhibit numbers, but we cite to the versions available on the court’s docket. 3 purposes. At the time of this merger, the purchase method of accounting also allowed for up to a 40-year amortization period for the goodwill created by the merger (described below).4 The amortization period was a critical incentive to the merging thrifts because it created a positive spread between the goodwill depreciation each year (the asset value slowly diminishing on the books) and the accretion of value in the loan portfolio acquired from the failing thrifts (increased asset value).5 These “paper profits” were sought by the S&L banks as a way to bolster their balance sheets in the following years. Tr. 246 (Lowy) (describing the “phantom” or “paper income” as “what makes the deal go”). Much was made during trial, however, about whether the accounting of this deal was consistent with general accounting standards at the time. The United States argued that, if it was, the RAP right was not as valuable as plaintiff believes. But in Winstar, the Supreme Court foreclosed much of this argument when it held that the RAP right was indeed a valuable guarantee against loss in the event that accounting and/ or regulatory standards changed.6 See United States v. Winstar Corp., 518 U.S. 839, 890-91, 868-70 (1996). It is important to understand what is meant by “goodwill” in these transactions. Broward had liabilities exceeding its assets. Using purchase method accounting under the applicable Generally Accepted Accounting Principles (“GAAP”) in 1981, the purchase price of Broward was its negative Glendale’s writing down of the goodwill from the Broward acquisition was only for accounting purposes. Glendale did not depreciate the goodwill asset for tax purposes. 4 5 The loan portfolio increased in value each year because the purchase method required that all of the assets acquired in the deal be priced in the market as of the date of merger. Because interest rates had risen, the mortgage portfolio held by Broward was worth less than its face value. Each subsequent year, however, the value of those mortgages ca,e closer to their par value, thereby inflating Glendale’s books marginally faster than it was depreciating the goodwill asset. That spread, called the “accretion of loan discount,” created “paper profits” that were of great interest to the acquiring thrifts because they helped to shore up, at least on paper, the financial health of the banks. 6 As will be discussed later, although we agree with the government that the regulatory accounting principles matched the general accounting rules, that does not render the RAP right worthless. 4 total market value (assets over liabilities). Broward’s loan portfolio was priced as of the date of sale, also known as “marking to market,” and, because interest rates had risen, the value of its earlier-issued mortgages was less than its par value. This significantly widened the gap between assets and liabilities, which thereby raised the nominal purchase price. This delta was recorded on Glendale’s balance sheets as “supervisory goodwill” because, as plaintiff’s expert Dr. Mann explained, a transaction always had to balance to zero under the accounting rules. That goodwill was put on the other side of the ledger and “canceled” or “zeroed” out the excess liabilities assumed from Broward. Tr. 563-64. Consistent with GAAP principles at the time, and with the blessing of the regulators, the goodwill recorded from the merger could be amortized over 40 years and counted as capital for purposes of regulatory capital reserve requirements as long as it remained on the books.7 The deal between Glendale and Broward was consummated with a merger agreement between the two banks, but it also required approval, and in this case inducement, from the federal regulators, which took the form of the SAA between Glendale and FSLIC. The FHLBB also had to sign off on the SAA. FSLIC expressly offered the following items of assistance to make the deal work: interest rate protections, indemnification for any litigation losses stemming from the merger or from unknown Broward liabilities, promises to restructure loans to Broward from the FHLBB, and branching rights. The SAA itself was silent as to the RAP right, but it incorporated by reference “any resolutions or letters issued contemporaneously herewith.” JX 29 at 14. One such incorporated document was a letter from the FHLBB, dated November 19, 1981, which explained that the Bank Board would not bring enforcement proceedings against Glendale for failing to meet regulatory capital requirements. Id. at CITI0000267 (forbearance letter). 7 The concept of goodwill, and its effect as balancing out the accounting of a business acquisition, was not new. In general, the concept refers to the value of acquiring an operating business. “Going concern value” reflects the intangible benefits such as brand recognition and an already existing customer base that come along with a continuing business. Here, however, there is a certain tension between the economic realities of this merger (Broward was underwater financially) and the accounting treatment that was permitted by the accounting rules and the federal regulators signing off on the deal. This tension is apparent in several of the issues we are confronted with. Although the tax law largely tries to comport itself to economic reality, there are times, as here, where the two diverge. 5 Also incorporated was the FHLBB’s resolution approving the merger, FHLBB Resolution 81-710. That resolution referred to a Glendale-supplied but independent accountant’s letter that was to address the goodwill recorded from the deal and its amortization period. This took the form of an opinion letter from the accounting firm of Peat Marwick Mitchell (PX 14). The FHLBB resolution also required a stipulation from Glendale that its accounting for goodwill and its subsequent amortization complied with FHLBB Memorandum R-31b. The Supreme Court held that each of these documents was incorporated into the government’s agreement with Glendale. Winstar, 518 U.S. at 862-64. As discussed in more detail below, GAAP rules allowed Glendale to use the purchase method of accounting for the merger so that the resulting goodwill could be treated as an amortizable asset. Read together, these incorporated documents were a promise that Glendale would be protected from financial ramifications if the accounting and regulatory treatment of the deal were changed to Glendale’s detriment. Id. at 868-70. This promise is otherwise known as “the RAP right.” In 1989, Congress passed FIRREA in response to the S&L crisis. Pub. L. No. 101-73, 103 Stat. 18 (1989). This ended much of the favorable accounting treatment that FSLIC and the S&L industry used to incentive these mergers. Importantly, Congress aimed at the amortization of goodwill and the favorable treatment of it as an asset for capital reserve purposes. Winstar, 518 U.S. at 856-57. FIRREA eliminated supervisory goodwill as an asset on Glendale’s books. As a result, it had five years to depreciate goodwill off its books. Glendale thus found itself undercapitalized for regulatory purposes and was forced to raise capital and restructure—part of which involved the sale of its Florida operations in 1994. It recorded a gain of over $200 million on the sale and paid tax on that amount. It did not reduce that gain by offsetting against it the loss of its branching rights and other intangibles obtained from Broward against it. Glendale successfully stabilized, but it incurred many costs. It subsequently sued in this court for damages flowing from FIRREA’s changes to the accounting and regulatory treatment of goodwill. Judge Smith ultimately awarded $380,787,000 in reliance damages, termed “wounded bank damages,” which accounted for recapitalization costs and the increased cost of capital suffered by Glendale. Glendale Fed. Bank, FSB v. United States, 54 Fed. Cl. 8 (2002), aff’d 378 F.3d 1308 (Fed. Cir. 2004). One issue 6 in this case is whether plaintiff has already garnered a tax benefit from this award so that they were properly included in income and any tax paid thereon should not be refunded. See Citigroup, Inc. v. United States, 140 Fed. Cl. 283 (2018). Glendale, and later Citigroup as its successor in interest, claimed a deduction for all of Glendale’s supervisory goodwill as early as 1994 and repeatedly thereafter until 2003. Those claims were first denied because of the pendency of the damages litigation. After certiorari was denied in Glendale’s damages case in 2005, plaintiff again claimed a deduction for the supervisory goodwill and for the tax it had paid on its reliance damages award in returns filed with the IRS in 2010 and 2014 for the 2005 tax year. Those were denied in May 2015. Plaintiff filed suit here on September 1, 2005, claiming a refund of $412,940,394 as overpaid taxes for the 2005 tax year. Plaintiff moved for summary judgment in 2017, asserting that FIRREA rendered all of its supervisory goodwill asset worthless and that the tax paid on its reliance damages was improper because (1) the damages were statutorily exempt from taxation and (2) the damages payment was otherwise not income. We disagreed on the first point, holding that, although plaintiff had a tax basis in the supervisory goodwill it acquired in the Broward merger, it had to apportion the purchase price among the various intangibles it acquired and prove worthlessness. 140 Fed. Cl. at 290. On the issue of the taxability of reliance damages, we held that, although not statutorily exempt, reliance damage awards were not considered income unless the tax benefit rule applied. Id. at 297. Trial was necessary to apportion basis (an exercise of identification and valuation) and to determine whether plaintiff had already deducted amounts included as reliance damages. Trial was held March 21 to April 1, 2022, and post-trial arguments heard on December 12, 2022. DISCUSSION Plaintiff asserts two claims. The first is for a $798,291,000 deduction for the loss of all the supervisory goodwill, which Citigroup contends was rendered wholly worthless upon the passage of FIRREA. According to the complaint, that deduction would result in a refund of at least $279,401,850 for the 2005 tax year. Compl. ¶ 75. The second claim is for the return of approximately $133.5 million in taxes paid on the reliance damages award made by Judge Smith after the original post-Winstar trial. Citigroup originally included the approximately $380 million award as income and 7 paid tax on it. It then amended its return, seeking to exclude the damages award from income because it represented only uncompensated losses, not lost profits or any other type of income. That claim was denied by the IRS, and the issue is included in this suit. Although we held previously that reliance damages were not generally considered income, the tax benefit rule might apply, and trial was thus necessary. We begin with the worthless asset claim. I. Worthless Asset Deduction – Supervisory Goodwill Plaintiff’s primary theory is that, at the time of merger, Glendale’s books had, a $798,000,000 asset—supervisory goodwill.8 It points to Glendale’s accounting records at the time, the work papers of the Peat Marwick firm, and a Peat Marwick accounting opinion letter incorporated into the SAA. Plaintiff argues that when FIRREA eliminated the 40-year amortization period and the use of supervisory goodwill for regulatory reserve purposes, that asset disappeared from Glendale’s books, and Citigroup, as Glendale’s successor, was entitled to a worthless asset deduction under section 165 of the Internal Revenue Code (“IRC”). At summary judgment, we held that the worthless asset question was more complicated because supervisory goodwill represented an entire basket of intangible benefits acquired in the merger with Broward which were not separately identified. 140 Fed. Cl. at 289. Chief among these items of consideration, as the government pointed out, were the branching rights that were untouched by Congress’ action in FIRREA and therefore not rendered worthless. Id. at 288. We set for trial the issues of identifying all of the separately identifiable and valuable intangibles that were acquired, their valuation, and whether they were rendered worthless. Id. at 290. The bulk of the value of the supervisory goodwill is represented by 8 The $798 million figure is an adjusted cost basis. The amount recorded as supervisory goodwill at the time of the merger was lower, but plaintiffs are entitled to deduct their adjusted basis for any assets lost. I.R.C. § 165(b). Defendant has not objected to the use of plaintiff’s adjusted basis figures, and thus we adopt them. Plaintiff’s valuation experts, however, largely expressed their opinions in terms of fair market value at the time of deal. Those numbers have to be adjusted upwards to account for adjusted basis. Plaintiff has provided those figures for each of its offered values. Our conclusion for branching rights differs from plaintiff, and thus an adjusted figure for that intangible will have to be supplied prior to judgment. 8 two intangibles: (1) the RAP right and (2) the branching rights. These were unquestionably the main inducements to merge with Broward, though several others were identified and valued by plaintiff at trial. We begin with branching rights. A. Branching Rights Glendale wanted to enter the Florida mortgage marketplace. Purchasing Broward through a merger allowed it to do that with the FHLBB’s explicit grant of branching rights as part of the government’s deal with Glendale. The purchase price, or the delta between assets and liabilities, was recorded simply as supervisory goodwill. Plaintiff’s fundamental view of the case is that no further identification and allocation of intangibles is necessary. It views the entirety of the $798 million of “goodwill” as worthless the day the after enactment of FIRREA. Although we understand that, as of FIRREA, Glendale no longer had supervisory goodwill available as an asset for regulatory capital reserve purposes, some of the rights and other intangible benefits were not directly affected by FIRREA. The tax treatment of basis must follow as much as possible the economic realities of the deal that created the asset now alleged to be worthless. Citigroup, 140 Fed. Cl. at 289. Branching rights were a major inducement with significant value and they were not rendered ipso facto worthless by FIRREA. Plaintiff’s expert, in fact, did not balk at the task nor did they have any conceptual problem with valuing these rights. The same is true for several other, smaller, intangible assets valued during the trial. Relief in the form of a worthlessness deduction thus must take account of branching rights and any other assets not rendered worthless by FIRREA.9 Id. at 290. 1. Plaintiff’s valuation Plaintiff offered the testimony of its expert, Dr. Steven Mann, a professor of finance at Texas Christian University. Dr. Mann was qualified as an expert in financial economics. He had testified for the government during the Washington Mutual litigation in district court, providing a market analysis that rebutted Washington Mutual’s position that the premium it paid to acquire a failing thrift was attributable to the acquisition of branching Plaintiff does not concede this point. At trial, it claimed an abandonment loss and then, in its post-trial papers, a worthlessness deduction for the branching rights and some of the other assets. Those arguments are addressed below. 9 9 rights. Washington Mut., Inc. v. United States, 996 F. Supp. 2d. 1095, 111617 (W.D. Wash. 2014). His criticism of the plaintiff’s branching rights valuation carried the day in district court and the Ninth Circuit. 856 F.3d 711, 727 (9th Cir. 2017) (finding no error in the trial court’s reliance on Dr. Mann’s analysis to discount plaintiff’s valuation of the branching rights); 996 F. Supp. 2d. at 1117 (finding that Dr. Mann’s market analysis “significantly undercut[] the trustworthiness” of the plaintiff’s valuation). In this case, unlike Washington Mutual, Dr. Mann was called on to perform another market analysis, but this time to provide an affirmative value for the branching rights, a value above zero. He did so by first compiling prices paid by thrifts acquiring banks in other states. This was relatively straightforward because courts have uniformly held that the purchase price for thrift mergers is represented by the difference between the assets and liabilities of the acquired thrifts. E.g., Citigroup, 140 Fed. Cl. at 290 (explaining that the parties agreed with the Ninth and Federal Circuits’ approach to the purchase price). The point was to compare those acquisitions with mergers not involving branching rights to find the difference. To make that comparison useful, however, another step was necessary because the relative size of the acquisitions was not uniform. Dr. Mann thus divided the purchase prices by the liabilities of each transaction, yielding a percentage that was, in essence, another measure of the purchase price.10 To keep the comparison meaningful (i.e., between similar situations), Dr. Mann found five merger acquisitions in 1981-82 that involved branching rights in Florida passing to the acquirer, including the Glendale-Broward merger. The comparison of those transactions revealed a median price of 29% of assets over liabilities. Dr. Mann’s next step was to find similar Florida thrift acquisitions without branching rights to compare with mergers involving branching rights. He cited 15 such transactions, which included intrastate Florida mergers and interstate Florida acquisitions without branching rights.11 After 10 Dr. Mann explained that he could have divided by assets or liabilities to compare transactions. He chose liabilities because he viewed the accounting of the liabilities as less volatile than that of the assets. Tr. 579-80. 11 Dr. Mann included in his set of non-branching rights mergers, a transaction in which the acquiring thrift had already been granted branching rights from an earlier FSLIC-assisted merger. Tr. Tr. 524-25. 10 again computing a percentage for the price premium, he came up with a figure of 21% for these acquisitions. This number was then compared with the branching rights transactions and a delta of 8% emerged, which he viewed as the average premium for acquisitions of branching rights. Dr. Mann also took a nationwide survey of 124 thrift acquisitions without branching rights, which similarly yielded a purchase price of 21% or, when he limited the timeframe for the transactions to October 1981October 1982, the remaining 83 transactions revealed a 24% price.12 He then compared the two sets of mergers not involving branching rights with the set of five mergers that involved Florida branching rights to find the difference. The result was a 5% to 8% premium paid for branching rights. When applied to the economics of the Glendale-Broward merger ($2.5 billion in liabilities), we find a range of $125-200 million of the purchase price allocated to the branching rights. Dr. Mann opined that a 5% price was probably his best estimate because it was derived from a comparison of acquisitions all taking place before the passage of the Garn-St. Germain Act. The most likely result thus, according to Dr. Mann, was that Glendale paid a $125 million premium for the right to operate thrifts in the Florida market.13 Mr. David Plastino, an economic consultant and principal at the Brattle Group, also testified for plaintiff in this regard. He was qualified as an expert in financial economics and business valuation. He stated that Dr. Mann’s valuation was reasonable and comported with the overall economics of the Broward acquisition. He also found the bottom of Dr. Mann’s range to be the most likely, positing a value of $124.2 million for the Florida 12 Dr. Mann explained that the deregulation of the S&L industry after October 1982 changed market conditions—mortgage rates went down— such that the most apt comparison, in his opinion, was to limit transactions to the October 1981 to October 1982 timeframe. Tr. 523-24. The Garn-St. Germain Depository Institutions Act, Pub. L. 97-320 (Oct. 15, 1982), allowed adjustable-rate mortgages and prohibited due-on-sales clauses in certain instances. 13 Dr. Mann also explained that this figure could be off by as much as a percentage point, or $25 million or so, due to uncertainty as to whether other intangible rights acquired, but not separately identified, might be the source of some of that premium. See Tr. 538-40. He also suggested that the value might be as low as $110 million due to interest rate protection or FHLBB refinancing provisions of the SAA reflecting some of the value of the price premium paid in the Broward acquisition. Tr. 540-41. 11 branching rights. 2. Defendant’s Position Defendant did not offer its own valuation of the branching right, Instead, the government generally criticizes Dr. Mann’s work as imprecise due to the range of values produced and argues that the range ($125-$200 million, give or take $25 million) is too broad to meet plaintiff’s burden of proof. Defendant also offers two other responses. The first questions whether Dr. Mann’s valuation properly isolated the branching right from the other intangibles that were part of the deal. The second point concerns whether plaintiff is entitled to ask for a deduction for the branching rights in this litigation. a. What did Dr. Mann Value? Defendant points out in its post-trial briefs that Dr. Mann’s valuation formula, a comparison between assisted and unassisted mergers, did not perfectly isolate the value of the branching rights because there were several other FSLIC assistance items in the deal that had value to plaintiff. Another of plaintiff’s experts, Mr. Harms, in fact, provided valuations of those items to the court, and Dr. Mann admitted that his branching right value would include any other FSLIC assistance items. Thus, his value for only the branching right might be off by up to a percentage point ($25 million). Defendant takes the point further by arguing that Dr. Mann’s method is so imprecise that it inadvertently also valued the RAP right. If the RAP right was part of the deal offered by the government to Glendale, a comparison to unassisted mergers (which have no such valuable guarantee) must include that right, argues the government. Defendant thus concludes that Dr. Mann’s value for the branching rights is unreliable and that it therefore undercuts plaintiff’s case by inadvertently providing a valuation of the RAP right that is far below any offered by plaintiff’s RAP value expert, Professor McDonald. Plaintiff attempts to revive Dr. Mann’s work in two ways. First, it cites to his testimony to establish that he did not value the RAP right. Tr. 542-43; 589-90. On cross-examination, Dr. Mann explained that his analysis “controlled” for the RAP right because all the mergers he used in his comparison used purchase method accounting. Tr. 589. The RAP right was, 12 in Dr. Mann’s view, tantamount to using that accounting method for the merger. Thus, the difference in premium between assisted and unassisted mergers could not be due to the guarantee of purchase method accounting. Second, plaintiff pointed to the differences in the methods employed by Dr. Mann and Professor McDonald, arguing that they were not valuing the same thing. Dr. Mann examined the purchase prices of various transactions, including the Glendale-Broward merger, while Prof. McDonald valued future income streams made possible by the RAP guarantee, two very disparate tasks. Plaintiff also cites to Mr. Plastino, who was asked whether Dr. Mann was valuing anything other than the branching right and answered that Dr. Mann was not.14 Tr. 1273. Mr. Plastino also explained that the going concern value of Broward was preserved by the FSLIC assistance, i.e., without it, Broward would not have had going concern value. The unassisted mergers would also have had going concern value to account for the goodwill, because those mergers did not need FSLIC assistance to keep the entity afloat, according to Mr. Plastino. Tr. 1275. The conclusion then is that only the branching rights could account for the purchase premium found by Dr. Mann in his comparison between unassisted and assisted thrift mergers. b. The Branching Rights are not recoverable as a deduction in this lawsuit Defendant’s second argument regarding the branching rights is that they are not properly before the court other than as a deduction from the basis that Glendale had in the goodwill resulting from the Broward merger. Defendant argues that neither Glendale nor its successor Citigroup made a refund claim based on an abandoned asset deduction for the branching rights nor were they otherwise rendered worthless by the enactment of FIRREA. Thus, they were not part of any claim made to the IRS and cannot be part of this lawsuit. We entertained briefing on this issue when plaintiff moved after trial In Mr. Plastino’s view, Dr. Mann’s valuation was not inclusive of the other FSLIC assistance items, but he admitted that Dr. Mann’s testimony was inconsistent with his opinion. Mr. Plastino was nevertheless untroubled by the possible inclusion of the FSLIC assistance items in Dr. Mann’s valuation because he believed that his selection of $125 million out of Dr. Mann’s broader range of possible values (up to $200 million) would account for the value of those other FSLIC assistance items. Tr. 1273-74. 14 13 for clarification on the point or in the alternative to amend its complaint (ECF No. 167). We denied plaintiff’s motion, holding that the “2010 and 2014 claims filed with the Internal Revenue Service mark the contours of what plaintiff can litigate here.” Citigroup, Inc. v. United States, No 15-953T (Fed. Cl. Aug. 29, 2022) (order denying motion for clarification) (ECF No. 174 at 1). Although abandonment and worthlessness are both creatures of I.R.C. § 165, their proof is different. We thus held that, because no claim for an abandonment loss had been made, the claim was not before the court. Id. at 1-2. We also noted that the issue had not been tried by consent. In fact, the opposite was true as the court and defendant were operating under the assumption that branching rights would serve only as a deduction from plaintiff’s potential basis in the supervisory goodwill. Plaintiff attempts an end run around this jurisdictional deficit in two ways. First plaintiff resorts to its general theory of the case that all the goodwill generated from the Broward merger was rendered worthless by FIRREA, an argument we have already denied on summary judgment. 140 Fed. Cl. at 290. Plaintiff adds, however, that, because Glendale’s deal with the FHLBB required it to record the entire purchase price as supervisory goodwill, it could not have separately identified the branching rights and claimed any deduction for them. In essence, plaintiff argues that it was bound for tax purposes by the accounting of the merger performed at the time of the merger. See generally Comm’r v. Danielson, 378 F.2d 771 (3rd Cir. 1967) (holding that sellers in a deal were bound by values allocated in the agreement). Second, plaintiff argues that FIRREA did, in fact, render the Florida branching right worthless by forcing Glendale to recapitalize, part of which involved selling its assets in Florida. By ceasing Florida operations without an intent to operate there again, the branching rights were rendered objectively and subjectively worthless because they could not be sold and because Glendale had no intention of operating in Florida again. Even if Glendale was eventually forced to sell its Broward assets due to capital needs, there are intervening events in the causal chain that would require different proof from that offered regarding the worthlessness of the RAP right caused by FIRREA. Congress did not force Glendale from the Florida market by changing the law regarding the amortization of goodwill and the right to count goodwill as a capital reserve. It did not change the law regarding branching rights. Glendale continued to operate in Florida for several years after FIRREA. Plaintiff may have elected to abandon its Florida branching rights but that does not establish worthlessness, a point on which there was a total lack of proof. All we have in evidence is a 1994 sale of the Florida assets (for which it recorded a gain) and testimony that 14 Glendale had no intention of reentering that mortgage market. That is insufficient to establish worthlessness. 3. Glendale’s basis in the branching rights The branching rights value remains relevant, however, because plaintiff’s basis in the goodwill can be reduced by the value of the branching rights, and any other intangibles identified, to arrive at the value of the RAP right.15 We have Dr. Mann’s $100 million range between $125-200 million, Mr. Plastino’s pinpoint of $124 million, and plaintiff’s argument that the adjusted basis is $139 million.16 Although Dr. Mann’s testimony was credible and his valuation method made economic sense, the result is an undeniably broad range. The court cannot find a range of deductions. Although plaintiff is correct that the law does not require mathematic certainty, Citigroup has the burden of providing a reasonably certain figure. Mr. Plastino’s valuation of branching rights was unhelpful because his opinion was not the result of an independent analysis or alternative valuation method. His testimony about branching rights was, in large part, a bootstrapping of Dr. Mann’s work: Mr. Plastino said that he found it reasonable and further explained why the RAP right was not part of what Dr. Mann valued. His selection of the precise value for the branching rights appeared much more the result of bookkeeping than any exercise of independent judgment. He was called upon by plaintiff to fit the constituent elements into the grand total of the supervisory goodwill. Adding up Professor McDonald’s RAP right, Mr. Harms’ intangibles, Dr. Mann’s tax benefits, and his own valuation of intangibles garnered from Broward, Mr. Plastino was left with a $124 million dollar hole that he filled with the branching rights in order to arrive at the total adjusted basis of $798 million, on the assumption that it was all deductible. Plaintiff established that the branching rights were worth between $100 and $200 million. The government argues that Dr. Mann’s value is too 15 As discussed below, plaintiff argues alternative paths to value its lost RAP right. One is through its expert Dr. McDonald, who offered his affirmative opinion on value as an isolated intangible benefit. The other path is by process of elimination whereby all of the other known intangibles are applied against Glendale’s basis in the supervisory goodwill, and the remaining number is ipso facto the value of the RAP right. 16 It is unclear how that adjusted basis was derived. 15 imprecise because, among other things, it fails to isolate the branching rights from other benefits acquired in the deal. Dr. Mann admitted his valuation method did not control for the other intangibles granted by FSLIC. Thus, in his view, the lower end was a more likely result, $125 million. Plaintiff, however, provided values for those other FSLIC-granted benefits during trial. We find it a simple enough exercise to start with the top of Dr. Mann’s range ($200 million) and reduce that figure by the value of the three other FSLIC assistance items to arrive at a reasonable figure for the branching rights. As detailed below, plaintiff’s expert Mr. Harms valued the three other items of FSLIC assistance (indemnity, advanced refinancing, and interest rate protection) at a combined $21.7 million. We then take $200 million and reduce it by $21.7 million and arrive at a branching rights value of $178.3 million. The value of the branching rights in hand, we turn to plaintiff’s value for the RAP right. B. The Rap Right The bulk of the testimony in this trial concerns the value of the RAP right. It was indisputably lost by Glendale when FIRREA changed the treatment of supervisory goodwill for regulatory and accounting purposes.17 Plaintiff offered the court alternative paths to valuation. First, was an independent, affirmative value for the RAP right irrespective of any other intangibles gained by Glendale in the deal. The other path was to isolate the value of the RAP right from the rest of Glendale’s basis in the supervisory goodwill by reducing that overall number ($798 million) by the value for each of the other intangibles. The remainder, or the residual, is the value of the RAP right. 17 The government argues that the RAP right was not rendered worthless by FIRREA. It points to Judge Smith’s wounded bank damages award during the earlier damages trial, arguing that Glendale successfully enforced the guarantee (the RAP right), and thus it was not rendered worthless by FIRREA. That argument is unavailing because it is an incomplete view of the effect of FIRREA. The Act not only ended the treatment of supervisory goodwill as an asset for regulatory purposes, but it also ended the 40-year amortization period for supervisory goodwill. We view the earlier damages award as addressing the first change (capital compliance) but not the second (amortization period). There was substantial value in having a goodwill asset on the books for 40 years. That ended due to FIRREA’s mandate that such assets be depreciated within a maximum of five years. 16 Plaintiff presented its expert, Dr. Robert McDonald, who valued the RAP right independently with an economic model.18 The assumption behind the model was that the RAP right allowed the new Glendale-Broward combined entity to operate. In other words, without the RAP right, Glendale could not have made the deal work. This meant that any future income generated by the combined Glendale-Broward portfolio was directly attributable to the RAP right promise and, when properly measured, represents the value of that promise in 1981. He did not use actual future Glendale income, however, because the value of the RAP right is as of the date of the merger—when Glendale’s basis in any assets acquired was set— meaning that the promises had to be valued based only on what was known then, which means the reasonable market expectations going forward. Dr. McDonald devised a way to value expected future income streams as of the merger by projecting future interest rates, as of the date of the Broward acquisition, based on the historic movement of rates and the projected cash flows that those rates would generate from Glendale’s mortgage portfolio. Plaintiff also presented several witnesses regarding the issue of the regulatory and accounting treatment of goodwill at the time of the merger to buttress its position that the RAP right was of great value and that it allowed the combined entity to survive. We begin first with the question of what GAAP allowed at the time of the merger. 1. RAP follows GAAP Plaintiff argues that, at the time of the transaction, it was uncertain whether the purchase method could be used to account for the GlendaleBroward merger. Defendant counters that the opposite is true: it was certain that this treatment was allowed by generally accepted accounting practices in place in 1981. The parties also dispute whether Glendale could amortize the goodwill over a 40-year period absent assurances from the government. FIRREA, of course, foreclosed that avenue of accounting, limiting the goodwill write-off to a much shorter period, but the parties dispute what GAAP embraced prior to FIRREA as a reasonable period for amortizing the Broward merger goodwill. Plaintiff presented the testimony of Richard Fink, Glendale’s general 18 Dr. McDonald holds a PhD in economics and is a chaired professor in the Kellogg School of Management at Northwestern University. He was qualified and accepted as an expert in the valuation of financial instruments including guarantees by federal entities. 17 counsel at the time of the deal. He testified regarding his negotiations with the government about the SAA and what would be included regarding the regulatory accounting principles and how they would apply after the merger. Mr. Fink told the court that one item of particular importance during those negotiations was Glendale’s accounting for the merger with the purchase method. It was his experience that previous mutual thrift mergers were treated as pooling transactions, meaning that the two entities’ assets and liabilities were simply combined and that no goodwill was created to offset excess liabilities.19 Tr. 98-99, 102. This would have resulted, in his opinion, in a non-viable combined entity because post-merger Glendale likely would have fallen short of sufficient capital reserves for regulatory purposes. Tr. 99. The fact that the purchase method allowed for the creation of the supervisory goodwill that could be used as an asset on the positive side of the ledger for capital compliance purposes was central to the deal. That same opinion was echoed at trial by Martin Lowy, a banking lawyer and author of High Rollers: Inside the S&L Debacle, a history of the savings and loan crisis.20 It was his view that, at best, the accounting of the Broward merger left uncertain whether it ought to have been treated as a pooled merger or as a purchase. In his opinion, the deal met all the qualifications of a pooling transaction, meaning that the assurance of the accounting treatment as a purchase going forward was a promise in and of itself. E.g., Tr. 208-09. A “mutual” refers to a S&L bank that was owned by its depositors and borrowers—like a credit union—as opposed to being set up as a corporation owned by shareholders. 19 Mr. Lowy’s book was cited by the Supreme Court favorably in Winstar as a historical record of what happened during the crisis of the S&L industry in the 1980s. We were not overwhelmed, however, with his qualifications to opine on the application of accounting principles to financial transactions. Mr. Lowy, although a noted historian of this period, is a lawyer by trade and education. We heard insufficient experience during his direct examination and voire dire that would indicate he had on the job experience or informal education regarding the application of GAAP or regulatory accounting rules to thrift mergers. We thus do not weight his testimony as heavily as Mr. Hargett from the government who was qualified in this area. Mr. Lowy was allowed to testify conditionally with the court’s caveat that we might disregard much of his testimony when it came to applying the accounting rules to the Glendale-Broward transaction. See Tr. 193-94. 20 18 Plaintiff also presented testimony from two former federal regulators involved in this deal. It offered the testimony of Dr. David Croft, the Director of the Office of Examination and Supervision at the FHLBB, from the earlier damages trial. Croft testified that the use of purchase accounting for mutual thrift mergers was new during this time, albeit consistent with GAAP. Croft Trial Test. at A226 (ECF No. 123-1) (JX 45) (Glendale FSB, Inc. v. United States, No. 90-772C, Feb. 26, 1997); see also id. at A209. Next, Mr. Beasley, the Director of FSLIC in 1981, testified that he could not recall any assisted thrift mergers prior to Glendale-Broward that used the purchase method of accounting.21 Tr. 1513. After learning of the intended use of the purchase method for this merger, he spoke with Dr. Croft and members of the accounting industry regarding whether this was acceptable under GAAP. Tr. 1514. We take that to mean that the regulators were concerned with conforming this deal to the applicable accounting standards, especially given the apparent novelty of using the purchase method. The implication of this, for plaintiff, is that the RAP right was more than just a financial protection regarding change in the accounting treatment. Plaintiff submits that the RAP promise settled the uncertainty of which method was appropriate for this deal. Dr. McDonald built these assumptions into his valuation methodology by assuming that it was only the RAP right that made the accounting for the deal possible. Future income streams are thus attributable to it. Both sides cite to Accounting Principles Board Opinion (“APBO”) 16 as the controlling statement of the then-applicable standards regarding thrift merger accounting.22 APBO 16 (PX 289) covers many types of business combinations, but there is no dispute that it is the correct statement of the principles that applied at the time to this transaction. Several witnesses 21 Mr. Beasley testified that he would have been unaware of any unassisted mergers that used the purchase method. The Accounting Principles Board was an arm of the American Institute of Certified Public Accountants. It was eventually replaced by the Financial Accounting Standards Board (“FASB”). The mission of these groups was to assemble and disseminate accounting principles that were viewed as generally acceptable, known as “GAAP.” FASB issued an opinion shortly after the Broward merger clarifying that using a 40-year amortization period for goodwill in these transactions was inappropriate. The estimated remaining life of the loans acquired was the proper standard for amortization of thrift merger goodwill. PX-382 ¶ 30. 22 19 testified that both pooling and the purchase method was acceptable under APBO 16, but that, when the qualifications for pooling were met, it was the method to be used. Mr. Lowy read APBO 16 as clear that pooling was the best method for the Glendale merger with Broward, and thus the deal was not consistent with the general accounting rules at the time. The implication plaintiff offers is that, absent the deal with the FHLBB and FSLIC, Glendale could not have accounted for the transaction the way that it did. The government counters with the opinion of Joel Hargett, a financial consultant and CPA involved in the Savings & Loan crisis in the 1980s. He had extensive experience as an auditor of S&L thrifts and as a regulator in various government roles for several years in the late 1980s and early 1990s. He was qualified and accepted as an expert in RAP and GAAP accounting during the relevant period, the financial analysis of S&L thrifts, and “net worth requirements and regulatory capital standards in the savings-and-loan industry.” Tr. 1560. Mr. Hargett’s opinion was that RAP generally followed GAAP, unless specifically noted by the regulatory agencies, and that the purchase method of accounting was the proper method to be used for the Glendale-Broward merger23 (i.e., that there was nothing remarkable about using purchase method accounting for this deal). Defendant also offered two letters from the Peat Marwick firm in which Glendale’s outside accountants stated that the purchase method was appropriate for this merger. PX 14; PX 28 at CITI0000119. Thus, we note that neither Glendale’s accountants nor the federal regulators had a problem with the use of purchase accounting for Glendale’s acquisition of Broward. In fact, Glendale’s CFO, David Hansen, testified that there was no consideration by Glendale or the government given to using the pooling method. Hansen Dep. at A980-81 (May 24, 2017) (ECF No. 123-1) (JX 41). That was because, as he recalled, the pooling method was not used in supervised mergers.24 Id. The Peat Marwick accounting opinion was We are summary in our discussion of Mr. Hargett’s opinions regarding GAAP and RAP because the resolution is clear without a lengthy recitation of his testimony. We note, however, that we found him credible on these issues and his testimony was very helpful to the court. 23 We do not think that the testimony of Mssrs. Beasley and Fink are directly at odds with Mr. Hansen. Beasley and Fink primarily recalled transactions prior to the Glendale merger. Around the time of the Glendale-Broward deal, the industry was changing its view on the accounting of these deals. 24 20 unequivocal that the deal should be accounted for using the purchase method. The federal regulators were satisfied with that statement. Further, as Mr. Hargett explained, pooling was an ill fit for this merger because of Broward’s pending insolvency. The character of the deal as an acquisition and not as two businesses combining for mutual benefit was appropriate. Another document makes clear that the FHLBB intended for these mergers to be accounted for consistently with GAAP. The FHLBB Office of Examinations and Supervision issued Memorandum R-31b on September 1, 1981, before the Glendale-Broward merger. It provides that “business combinations involving insured institutions should be accounted in accordance with [GAAP].” PX 9 at CITI0000274. This court and the Federal Circuit have already construed this document as permitting, if not mandating, the purchase method of accounting for these types of transactions. See WMI Holdings, Inc. v. United States, 891 F.3d 1016, 1025 (Fed. Cir. 2018) (affirming this court’s finding that the rules in place required thrifts to use the purchase method). The memo goes on to explain the appropriateness of recording goodwill for these mergers when using the purchase method and details certain intangibles that should be recorded separately when possible. The document also briefly discusses the amortization of goodwill, recognizing that, although valuing intangible factors is inherently subjective, banks should analyze and determine an asset’s life to match “income derived from the accounts” of the acquired thrift. Id. at CITI0000275. The result should be “a reasonable estimate of the useful life” of the intangible asset but “should, not, however, exceed forty years.” Id. (citing APBO No. 17).25 This is what Glendale and its auditor Peat Marwick did when they accounted for the merger with Broward.26 25 APBO No. 17 dealt specifically with the subject of accounting for intangible assets. PX 422. Glendale believed that 40 years was appropriate for the goodwill generated in the Broward merger due to the indefinite life of the branching rights it was acquiring. 26 Mr. Lowy attempted to discount Memorandum R-31b by stating that it did not apply outside of the FHLBB staff, See Tr. 294, as if the FHLBB itself would somehow not be bound. Mr. Lowy urged that the case-by-case nature of the FHLBB’s approach to these transactions meant that this central staff memo—essentially instructions to the Washington, DC bank board staff— would not have been authoritative. We heard no good reason why that would be so, however. And, as Mr. Lowy himself testified, it was the central FHLBB staff that worked on this merger, not the regional supervisory agents. Further, the document relied on for the ad hoc approach to each deal predated 21 Based on the clear statement from Memorandum R-31b, which predated this merger, the Peat Marwick’s letter, Mr. Hargett’s opinion, and the apparent lack of any pushback from the FHLBB about the purchase method accounting for this merger, we conclude that, as of the date of the merger, GAAP was assumed to follow RAP. Although that itself does not discredit Prof. McDonald’s opinion, it casts significant doubt on one of the critical assumptions on which he constructed his model. 2. Dr. McDonald’s Valuation of the RAP Right Dr. McDonald testified that he set out to value the RAP right as a valuable guarantee of three things: (1) that goodwill could be amortized over 40 years; (2) that the loan discount could be accreted for a shorter period, thus producing paper profits; and (3) regulatory forbearance from the government. The government argues that he valued the wrong thing and that the inputs and outputs of his model are flawed in several respects. We begin with a summary of what Dr. McDonald did and then consider defendant’s argument that he valued the wrong thing. a. The Model Put in its simplest terms, Dr. McDonald valued the promises from the government as the difference in Glendale’s expected future cash flows as of the date of the merger, with and without the promises. Because Glendale would have been unable to continue in business absent those promises, per Dr. McDonald, the comparison is between zero income and whatever his model generated for anticipated future earnings. He put two constraints on his model: (1) that the newly combined portfolio would remain constant, meaning that it would not be rebalanced in any way except for the addition of new earnings generated by the portfolio existing as of the merger; and (2) the combined entity’s net worth would remain above zero to be able to continue in business. The first constraint protects against any hindsight bias and protects against growth in Glendale’s business that is not attributable to the RAP right. That second constraint was largely unexplained. Dr. McDonald’s model begins with the mortgage portfolio of the issuance of Memorandum R-31b. See Tr. 296 (Mr. Lowy mentioning an August 1981 communique requiring a case-by-case review). 22 Glendale immediately after the merger with Broward.27 He uses predicted future interest rates based on the yield curve as of the merger, known as the “Hull-White Model” to generate values for the portfolio going forward for 20 years, and then discounts those future streams back to November 1981. Based on the yield curve used, the Hull-White model predicts interest rates. Mortgage rates were then calculated by Dr. McDonald based on the treasury rates for the 1, 5, and 10-year bonds supplied by Hull-White. With each incremental advance in time, Dr. McDonald consulted the Hull-White Model based on the yield curve on that date to come up with three different interest future rate scenarios: rates went up, they went down, or they stayed level. Cash earned or spent is then calculated for each scenario at each increment in time.28 The model continues for twenty years in this fashion, thus blossoming the total number of scenarios by orders of magnitude. We visualize this approach as a sort of matrix, or lattice of matrices, of possible outcomes. Eventually, after running all the figures through a Monte Carlo simulation, the most likely resulting value after 20 years, discounted to November 1981 value, was $513 million.29 We are unable to adopt this value, however, because we agree with the government, as explained below, that he made critical errors in his assumptions of what precisely the government promised in the RAP right.30 He used Glendale’s June 30, 1981 financial statements, prior to the merger, as the starting point and then added Glendale’s holdings from a Peat Marwick letter dated March 15, 1982, which contained Broward’s audited balances as of the date of the merger. DX 3 (Glendale financial statements); PX 14 (Peat Marwick letter with audited Broward balance sheet). 27 28 Dr. McDonald detailed how he calculated the expense of servicing the interest on Glendale’s deposits at each time increment and how the cash, whenever generated from mortgage payments, was reinvested at 1-year rates going forward until the end of the simulation. A Monty Carlo simulation is a method of testing a model that has multiple variables. Using a computer, the model can be run many thousand times with all the possible variables to come up with all of the possible results. This then can show the most likely outcomes, the average outcome, median outcome, and other similar pieces of information. It is regularly used as a tool in the finance industry to arrive at net present values for projected cash flows. 29 We note, however, that this does not necessarily foreclose a result in plaintiff’s favor. As explained later, plaintiff presents an alternative basis for the valuation of the RAP right by proving the value of all of the other 30 23 b. The RAP Right Defendant does not take Prof. McDonald’s list of three guarantees within the RAP right at face value. According to the government, his model incorrectly ascribes the value of the accounting treatment of the supervisory goodwill to the RAP right. Or, as defendant puts it in its post-trial brief, he confused the guarantee with the asset itself. Second, defendant argues that the accretion of the loan discount was not part of the deal with FSLIC and thus cannot have been part of the RAP right. It is therefore inappropriate, according to the government, to ascribe the value of that benefit to the RAP right. Finally, defendant urges that the forbearance valued by Prof. McDonald is not what was granted by the SAA and is not related to this deal. Defendant’s conclusion is that the model is valuing the wrong thing entirely and cannot be relied on as a measure of loss suffered by plaintiff. We agree. There was imprecision in Dr. McDonald’s testimony regarding whether he viewed the goodwill asset itself as a grant from the government, (i.e., as part of the RAP right). During direct examination, he maintained that he had constrained his valuation to the three elements we listed above. When pressed during cross-examination, however, he wavered, stating that he was not taking a position “on where the supervisory goodwill came from.” Tr. 872. But then some clarity emerged during the follow-up questions from the defense. Counsel asked Prof. McDonald whether his model assumed that Glendale would have gone out of business, or been insolvent, the day after the merger absent the RAP right. The answer: “Yes. I mean, had there been no – had the supervisory treatment not been there, then they would have been unable to operate.” Tr. 874. Defendant seizes this answer as proof that Prof. McDonald was valuing the goodwill itself, which was not a byproduct of the deal with the government. We agree, at least as it concerns Dr. McDonald’s model. As we explained above, the accounting treatment of this deal was consistent with GAAP, meaning that it was proper for Glendale to record the deal as it did, including the goodwill asset. With the excess liabilities of Broward now on its books as an asset (supervisory goodwill), Glendale could remain in compliance with FSLIC capital reserve requirements, but it was not the government’s action that created the goodwill. Instead, it was generally accepted accounting rules that allowed Glendale to assume nearly$800 million in excess liabilities but to count them as an asset intangibles gained by Glendale. 24 Put another way, the incorporation of the Peat Marwick opinion was not tantamount to the government granting that accounting treatment as part of its inducement to Glendale. Because the purchase method of accounting was proper, it was the acquisition of Broward that created the asset used to keep the newly merged Glendale above water. See WMI Holdings Corp. v. United States, 891 F.3d 1016, 1025-26 (Fed. Cir. 2018) (affirming this court’s holding that GAAP permitted the creation of goodwill in the deal and thus that was not a component of the RAP right). As we held in Washington Mutual, Inc. v. United States, 130 Fed. Cl. 653 (2017), and consistent with the Supreme Court’s explanation in Winstar, “the nature of the RAP Right is a guarantee that [plaintiff] could amortize the goodwill created by FSLIC Mergers over a period of up to forty years, notwithstanding an adverse regulation regarding the amortization period for goodwill.” Id. at 691. For that reason, Judge Grigsby held that the valuation method offered by plaintiff in that case was unreliable because it assumed that the RAP right created the goodwill used to offset the liabilities acquired. Id. at 692-93. That conclusion was affirmed by the Federal Circuit. WMI Holdings Corp., 891 F.3d at 1025-26. We find that the same error undermines Dr. McDonald’s valuation here.31 Whether the accretion of loan discount is properly considered an element of the RAP right is immaterial. As the Supreme Court explained in Winstar, the accretion of discount was a feature of the 40-year amortization of the goodwill. Id. at 2443-44. The value of the accretion of loan discount, although not itself a grant from the government, was properly included in the value of the RAP right because Glendale had bought for itself a guarantee that it could reap those profits going forward. That does not matter, however, because we cannot isolate the loan discount value in Dr. McDonald’s We do not view the Supreme Court’s statement that the parties to the SAA intended to “settle the regulatory treatment of these transactions as a condition of their agreement,” Winstar, 518 U.S. at 2449, or similar statements, to indicate the contrary. Those statements must be read in light of what the Court explained earlier, that the regulators’ promises were “to insure the promise against loss arising from the promised condition’s nonoccurrence.” Id. at 869. The promised condition was treatment of goodwill as an asset for regulatory purposes along with its long amortization period. That treatment was already in place before the deal. The change, made by FIRREA triggered the government’s duty to compensate. FIRREA did not, however also include compensation, which results in the Winstar litigation and now the resulting tax controversy. 31 25 method. The issue of whether the forbearance valued by Prof. McDonald was precisely the same as that granted in the SAA, is also immaterial. Defendant is correct that the forbearance provision of the SAA had only a 10-year duration and covered only a portion of Glendale’s assets that pre-existed the merger. Prof. McDonald clarified during cross-examination that he was referring to how FSLIC was treating these mergers generally, see Tr. 89899, namely, forbearing from closing thrifts post-merger when liabilities would have resulted in negative net assets or a lack of sufficient capital reserves. We do not see any additional value added by this “forbearance” and, in any event, it is rendered unimportant by our conclusion that Dr. McDonald’s assumption about the goodwill asset undermines his whole model.32 Dr. McDonald’s assumption that, absent this regulatory treatment of goodwill, the newly formed entity would have been insolvent, leads him to attribute all future cash flows to the deal with FSLIC, i.e., but for the deal with the government, Glendale would not be permitted to operate, and its future income goes to zero. That assumption is wrong. RAP followed GAAP. As defendant put it, Dr. McDonald valued the wrong thing. He confused the guarantee with the assets acquired. We thus cannot rely on his methodology to find Glendale’s basis in the RAP right. That is not the end of the case for plaintiff, however, because it alternatively arrives at the RAP right by providing a value of all the other intangibles acquired in the deal and then reducing Glendale’s basis by those amounts to arrive at the remaining value which can only be attributed to the RAP right. C. Other FSLIC Assistance In addition to the RAP and branching rights obtained from the government in the merger, there were three other benefits granted Glendale 32 Defendant also presented a host of reasons that it takes issue with Dr. McDonald’s model, some of which were discussed by Mr. Hargett, but about most of which we heard no testimony because defendant elected not to present its own valuation case. As such, we have no reason to credit its arguments, and no reason to deal with them. And, in any event, we have already held Dr. McDonald’s model to be unreliable. We further note that, Mr. Hargett did not possess the necessary expertise, as a CPA and former bank regulator, to address the more esoteric issues of economics raised by Dr. McDonald. 26 in the SAA—interest rate protection, an indemnification provision, and a FHLBB refinancing provision. Plaintiff presented Mr. Travis Harms, a principal in Mercer Capital’s Family Business Advisory Services Group to value these rights. Mr. Harms is an expert in business valuations and financial consulting, a CPA, and a chartered financial analyst. He was offered and accepted as an expert in the valuation of intangible assets. The total value for these three “other items of FSLIC assistance” in 1981 offered by Mr. Harms was $21.7 million 1. Interest rate protection The interest rate protection agreement came from section five of the SAA. It was in essence a risk-sharing provision whereby both parties agreed to reimburse the other depending on the movement of rates. If on the yearly anniversary of the deal, the average weekly rates were above the weekly average of the transaction’s closing, FSLIC agreed to pay Glendale a differential reimbursement and vice versa. The risk-sharing provision continued for five years after the close of the merger. How the payment was to be calculated is unimportant for our purposes, but it is important to note that the risk was not shared evenly. The provision waived the first $20 million of liability that Glendale might incur should rates fall. Mr. Harms determined the fair market value of such an interest rate protection agreement using interest rate derivatives (a financial instrument that pays based on the movement of interest rates). He explained that could be done by modeling a portfolio of interest rate calls and puts (option contracts). Tr. 1115-17 (Mr. Harms describing the Black Model of interest rate derivatives and explaining what call and put options are). For each of the five anniversaries, Mr. Harms selected a call and a put option and put it into the portfolio, and then determined its fair market value using the Black Model of interest derivatives.33 The total value of this five-year portfolio was just over $13.4 million, which had to be adjusted to account for the first $20 million not owed by Glendale and to account for the date that the interest rate averages were calculated (anniversary date as opposed to year-end). Mr. Harms made these adjustments by modeling the portfolio using a binomial model. This model was open to adjustments, such as the first $20 million not owed by Glendale. After those parameters are set, the model 33 The Black Model is a tool for valuing option contracts and is standard financial tool used for pricing interest rate derivatives. 27 generates the future cash flows discounted back to present value based on the same set of options (puts and calls). Mr. Harms generated 250,000 possible iterations and arrived at an average value of $11,975,000 in 1981 dollars. He then took the values generated by each model, Black and Binomial, and averaged them, arriving at $12.7 million for the value of the interest rate protection agreement. 2. Indemnification In section six of the SAA, FSLIC promised to indemnify Glendale for any damages that might be awarded against Glendale due to the Broward acquisition or for any unknown liabilities of Broward that existed at the time of the merger. Mr. Harms valued this provision as “representation and warranties” insurance (“R&W”), the most analogous product in the market. R&W insurance covers the holder in the event of a loss caused by a breach of a representation or warranty offered by a counterparty. This type of agreement is common in mergers and acquisitions. Mr. Harms explained that R&W policies would include a coverage limit and that the premium was usually around 2% to 3.5% of that limit. Here, there was no cap on the government’s potential liability nor any deductible, and the agreement lasted 10 years. These factors boosted the value of such a policy upwards, in his view, to six percent. Tr. 1147. Mr. Harms used 10% of the total liabilities assumed by Glendale in the merger as a conservative estimate of the universe of liabilities that Glendale might face from the deal. He then took 6% of that $73 million and came up with a total value of the indemnification provision of $4.4 million.34 Tr. 1150. 3. Advance refinancing In section 8 of the SAA, the government agreed that Glendale could restructure or refinance advances made to Broward by the FHLBB for 90 days post-merger. Mr. Harms came up with $4.6 million for this value, and defendant stipulated to that amount at trial. 34 He further explained that the nature of the business that Broward was in, fixed rate mortgages, was a relatively low risk venture, and thus hidden liabilities were unlikely. That led him to select a relatively small universe of potential liabilities ($73 million) against which to apply the premium (6%). Tr. 1148-49. 28 4. Plaintiff cannot deduct them Neither defendant nor its expert challenged the valuation of these promises from the SAA. The government does not believe a tax deduction is owed, however, because these items were not identified and claimed in Citigroup’s 2005 tax return nor are they called out in the Complaint. Both of those documents sought a refund for supervisory goodwill as a “single, inseparable intangible asset.” E.g., Compl. ¶¶ 64-65 (ECF No. 1). These assets were unaffected by FIRREA and were not part of the breach lawsuits in the Winstar litigation. Defendant argues that the regulators involved had no problem with Glendale treating all these assets together as part of supervisory goodwill, but that tax law requires separate identification and valuation if a deduction for loss is claimed. See Ledger Co. v. United States, 507 U.S. 566 (1993). Defendant thus concludes that, not having made a separate claim for these intangible assets, plaintiff is barred from asserting a deduction for them here under the doctrine of variance. We agree but have a more basic problem with these items. We heard no evidence nor argument that they were rendered worthless by FIRREA or any other event. Plaintiff’s 2005 tax return is predicated on FIRREA as the breach or loss-causing event. Plaintiff’s claim was for the entire value of the supervisory goodwill recorded on Glendale’s books in 1981, which then remained on its books untouched (not depreciated for tax purposes) until FIRREA changed the treatment of goodwill from these mergers, at which point plaintiff pursued and has continued to pursue a tax deduction for that entire amount. We hold that these three items of FSLIC assistance, interest rate protection, indemnity provision, and advanced refinancing, are not deductible because plaintiff neither made a claim for their worthlessness to the IRS nor has it proven that they were made worthless. The result is the same for the other, non-SAA related, intangible assets that we discuss next. D. Other Intangible Assets Plaintiff’s experts identified an additional five intangible assets received by Glendale from the merger with Broward. Dr. Mann valued the tax benefits inherent in the acquisition of Broward’s portfolio resulting from built-in losses in mortgages that had lost value. Mr. Plastino valued the core deposit and assembled workforce intangibles acquired by merging with 29 Broward as well as the intangible value of two other relatively minor revenue streams that Broward had in its Florida Outlook Development Corporation and its real estate appraising business. 1. Tax benefit Dr. Mann performed an analysis of the tax benefits that Glendale could realize if it sold Broward mortgages that had lost value due to rising interest rates since their inception. This was a relatively straightforward exercise of figuring their market value as of the merger and calculating the loss. He found a value in the $45-50 million dollar range. Mr. Plastino, in assembling all of the values, picked the middle of that range, $47.5 million. Defendant presented no disagreement with these figures, but it argues that Dr. Mann left out the value of Florida state tax benefits and argues that, because Broward had already filed its tax return for the 1981 tax year as of the merger, Dr. Mann’s value fails to account for the refund generated to Glendale from Broward’s claim. We note that plaintiff does not claim a deduction for these tax benefits; it merely uses their value to isolate the value of the RAP right. We heard no evidence to question Dr. Mann’s figure, and defendant’s criticism regarding Broward’s 1981 tax year was accounted for by Dr. Mann in his testimony when he explained that Broward’s $19.5 million return for the 1981 year was included in Glendale’s accounting for the purchase in the “interest receivable and other assets” category called out in the Pete Marwick letter. Dr. Mann also explained that any Florida tax benefits would have been similarly accounted for in that same category. Mr. Plastino agreed and echoed these sentiments during his testimony. Defendant presented no expert opinion or other evidence to question these explanations. We thus take Dr. Mann’s valuation of $47.5 million to be accurate. 2. Core Deposits and Assembled Workforce Mr. David Plastino identified four valuable intangible assets that Glendale received from Broward that were features of Broward’s existing business. The first two related to the value of having existing depositors at Broward and the value of an existing workforce in place at Broward to carry out operations. Mr. Plastino valued these at $18 million and $3,904,000 respectively. As to the first, the “core deposit intangible,” Mr. Plastino explained that this represents customer loyalty. This was valued by quantifying the 30 difference that Broward would have to pay in interest to keep existing depositors versus the interest cost to acquire new ones. Pete Marwick had already placed a value on this intangible of $18 million. Mr. Plastino confirmed that value through two exercises: an income approach and a market approach. Both confirmed that the Pete Marwick valuation was within a reasonable range of potential values for the customer loyalty expressed in the core deposit intangible. We need not spend any time explaining these methods as defendant does not challenge them. The assembled workforce intangible represents the value of having skilled labor in place to carry out operations at the Broward locations. Such a workforce can generate greater income than a brand-new workforce if Glendale had to hire one when it acquired Broward. Mr. Plastino valued these workers by estimating what it would have cost Glendale to hire a new workforce based on the logic that Glendale would not pay more by way of a premium than what it would have cost it to hire new employees. He did so via case-study data, rendering a value of just over $3.9 million. This evidence was unchallenged by the government. 3. Outlook Development Corporation and Florida Real Estate Appraisers The last two intangible assets identified and valued by Mr. Plastino related to businesses that Broward was operating apart from its home mortgage enterprise. The Outlook Development Corporation was a small real estate development company. Its hard assets were already valued and included in the Pete Marwick opinion, but no going concern value was identified. Mr. Plastino reviewed Outlook’s earnings, applied a growth rate factor, adjusted for the cost of equity, and discounted back to the date of the merger to come up with a going concern value of $6.103 million. The same factors and rubric were applied to the Florida Real Estate Appraisers, Inc., producing a $300,000 value. Defendant did not question these values but raised the issue of whether Mr. Plastino’s survey of the benefits acquired by Glendale had missed several other intangibles. Mr. Plastino was asked whether he identified any other valuable intangibles acquired by Glendale. His answer was in the negative. He was then asked specifically about name brand value but concluded that Broward had none because it was not producing positive value for Glendale. See Tr. 1298 (opining that Glendale did not pay any premium for Broward’s brand). Mr. Plastino also did not believe that there was any “ordinary goodwill” that resulted from the Broward acquisition 31 because he did not view Broward as having any going concern value. Tr. 1248-49, 1314-15.35 4. Plaintiff cannot deduct them Plaintiff argues that, other than tax benefits, the intangibles gained from acquiring Broward, like the FSLIC assistance items, were rendered worthless by the chain of events set off by FIRREA. Because Glendale was forced to sell its Florida operations in 1994, those intangibles were rendered subjectively and objectively worthless. Adding them up, it claims $30.76 million for these deductions. Plaintiff argues that, because it did not reduce the gain Glendale realized when it sold the Florida assets to account for any loss of the goodwill associated with these intangibles, they are all deductible now. The government responds that Citigroup has not previously claimed a deduction for the loss of these assets and thus it is barred from doing so now. Plaintiff replies that it was estopped from claiming them separately due to the application of the Danielson rule. The Danielson rule, derived from the Third Circuit’s landmark decision in Danielson v. Comm’r, 378 F.2d 771 (3rd Cir 1967), holds a tax petitioner to the express bargain that it made, if it agreed to a certain tax treatment of assets or other items that were the subject of the bargain. That rule was adopted by our predecessor and is binding in the Federal Circuit. See generally Hartmann v. United States, 694 F.2d 96 (Fed. Cir. 2012) (although not relying on the rule, the circuit court noted the application of the rule in this circuit and its earlier adoption by the United States Court of Claims). Plaintiff thus concludes that it is proper to include them in this lawsuit now, in essence, because this suit was its first opportunity to do so. Defendant disagrees, arguing that Danielson is inapposite because it provides a defense for the government, at its option, and is not a shield for claimants seeking to wiggle out from the strictures of their tax claims before the IRS during litigation. Citigroup has not been the subject of an adverse construction of Glendale’s deal with Broward and the government. Nor has it been limited by the IRS to the accounting of the deal at the time of the merger. The IRS disallowed the 2005 claim because it found that Glendale had no basis in the supervisory goodwill. Although that conclusion was wrong as a matter of 35 Government counsel also posited in its post-trial briefing that there was value in having an existing customer base of mortgagors. No witness endorsed this view, however. 32 law—the acquiring thrifts have a basis equal to the purchase price (liabilities over assets) in whatever intangible assets were acquired, see, e.g., Washington Mutual, Inc. v. United States, 636 F.3d 1207, 1219 (9th Cir. 2011)—plaintiff can only claim a deduction for that portion of the basis which was lost. The supervisory goodwill represented a basket of intangible benefits gained by Glendale. It can only deduct them “if [it] can allocate the cost basis to each of those rights individually.” WMI Holdings, 891 F.3d at 1021. At summary judgment, we made clear that doing so was plaintiff’s task at trial and that we could not “restrict our view on valuation, or basis, to the deal between FSLIC and Glendale. 140 Fed. Cl. at 289. As with the FSLIC assistance items, we have not heard any testimony or other evidence of the worthlessness of these assets. We are thus in no position to assess these claims, and they fail for a lack of proof of loss. E. The Sum - Supervisory Goodwill We are unable to rely on Dr. McDonald’s valuation of the RAP right, but plaintiff proved a value for the branching rights, in the form of a range, established the tax benefits acquired, the value of the smaller FSLIC assistance items (indemnification, interest rate protection, and advanced refinancing), and provided a reasonable value for the other Broward intangibles: core deposit intangible, assembled workforce benefit, and the going concern value of the real estate development and appraisal businesses. Although we have concluded that the worthlessness of these other, smaller intangibles was not proven and are not deductions otherwise available in this lawsuit, they serve another purpose. Plaintiff argues that it can, in effect, back into the value of the RAP right by subtracting the value of these other intangibles from its total basis in the supervisory goodwill to arrive at the value of the RAP right, which was rendered worthless by FIRREA. We agree. See Jack Daniel Distillery v. United States, 379 F.2d 569, 579 (Ct. Cl. 1967) (holding that the “residuary method” was a proper way to value an asset acquired in a deal when the parties “have reasonably established the value of all other assets.”) Defendant offered no reason to question either the identification of these other intangible assets or their values, aside from its intimations on cross-examination of Mr. Plastino regarding name recognition and the customer list of Broward.36 He did not endorse that idea, nor did defendant 36 Defendant did argue that Dr. Mann valued the entire deal instead of the branching rights, but, as we explained above, Dr. Mann explained why that criticism was invalid. 33 present any other evidence for it. Thus, we find that plaintiff has identified and valued all of the intangibles acquired by Glendale as the purchase price, which was recorded as a single entry, “supervisory goodwill” on its books. The RAP right, the largest single constituent element, was valued by a process of elimination—reducing the value of the supervisory goodwill by all of the other intangibles to arrive at the remaining basis, which can only be accounted for by the RAP right. In that regard, this case is unlike either the Washington Mutual case in district court or in this court. In both of those cases, the plaintiff claimed abandonment or worthlessness of branching and RAP rights. It did not identify and value any other rights that might have been acquired in its mergers. Thus, the courts could not back into the value of either right simply by a process of elimination. Further, in the district court case, plaintiff’s expert arrived at a fair market value that exceeded the price paid to acquire the thrift, meaning that the value of each right had to be precisely known in order to deduct the proportional share of plaintiff’s basis in each intangible lost. Washington Mutual, Inc., 996 F. Supp. 2d at 1104. We also do not share the Federal Circuit’s concern in WMI that some of the basis might be attributed to traditional goodwill or other intangibles. See WMI Holdings Corp., 891 F.3d at 1028.37 Here, plaintiff has done all the attribution and apportionment necessary; none of the purchase price remains unallocated. Defendant has offered nothing to rebut plaintiff’s identification and valuation of the bundle of intangibles and offered no independent identification of other intangibles acquired by Glendale. We are thus left with a simple exercise of subtraction to arrive at plaintiff’s basis in the RAP right, or, put another way, after taking out the value of the other assets acquired, the remaining basis (remainder of the purchase price) is the value of the amount of supervisory goodwill attributable to the RAP right. 37 In that case, Washington Mutual argued on appeal that this court had improperly rejected its branching rights valuation solely because it had already rejected the RAP right claim. The Federal Circuit rejected that argument because plaintiff had not established that those were the only two assets acquired or that some of the purchase price could not be attributed to traditional goodwill. Id. If there were other assets acquired by Washington Mutual or some of the price related to traditional goodwill, then no process of elimination could be used to arrive at a value for either. That is not the case here. 34 II. Tax Benefit Rule – Previously Awarded Damages The remaining issues concern the taxability of Glendale’s previous damages award. Citigroup included the entire damages award, some $381 million, in income for the 2005 tax year. It later amended that return, seeking to exclude the entire amount. At summary judgment, defendant argued that the tax benefit rule might apply, and we agreed, setting the issue for trial. 140 Fed. Cl. at 296-97. At trial, plaintiff narrowed the issues for this claim. It now seeks to exclude from income only $23.234 million of recapitalization costs included in Judge Smith’s damages calculus and $48.713 million in cancellation of indebtedness income. A. Recapitalization Costs On the recapitalization costs, defendant stipulated that, if plaintiff had not previously capitalized these expenses to an account that generated amortization deductions, it would not oppose their exclusion from income. See Tr. 73-74 (statements of defense counsel). Mr. James Carrol, Citigroup’s head of tax controversy, clarified that such was not the case, and thus defendant concedes the appropriateness of excluding the $24.234 million from income. See Def.’s Post-Trial Br. 18. B. Cancellation of Indebtedness Income The second amount plaintiff seeks to exclude from income stems from another avenue of Glendale’s efforts to boost its books post-FIRREA. Glendale bargained to cancel certain debts in exchange for stock. This generated taxable income as discharged indebtedness, which was reported by Glendale in 1993. Plaintiff now claims that this was part of the damage it suffered as a result of FIRREA and can thus be properly apportioned out of Judge Smith’s “wounded bank” damages award. The government opposes the exclusion from income of the approximately $48 million in canceled indebtedness on the basis that it was not a component of Judge Smith’s wounded bank damages, despite plaintiff’s characterization, and because it has never been the subject of a tax return, and moreover not the subject of the two returns at issue in this suit. Such a claim is thus not properly part of this lawsuit. Like defendant, we cannot find any mention of cancellation of indebtedness in Judge Smith’s opinion. Further, Mr. Carrol’s testimony in 35 this regard was impermissible hearsay. Although the documents he relied on speak for themselves, absent his explanation, we could not independently ascertain from them the bona fides of plaintiff’s claim in this regard. Further, we take defendant’s point that no claim for overpayment of tax on this amount was ever made to the IRS. For all these reasons, this amount cannot be excluded from income now. CONCLUSION Plaintiff has proven the value of the supervisory goodwill attributable to the RAP right. It did so by proving the value of every other intangible asset Glendale acquired in its merger with Broward, both from the government and from Broward itself. That leaves us with a simple mathematics exercise to reach the amount of basis left over and attributable to the RAP right. For the various reasons explained above, the other rights are not deductible for failure of proof of loss or are not the subject of the tax controversy at issue in this case. Plaintiff’s adjusted cost basis in the Supervisory Goodwill, $798 million, is reduced by the value of the other intangibles, none of which are deductible. Dr. Mann gave us a reasonable range for the value of the branching rights of $100-200 million, and, as explained above, we view the value as $178.3 million, which we think is fair given the testimony in the prior cases concerning Glendale’s great desire to enter the Florida market. We also subtract the items of FSLIC assistance, which Mr. Harms valued at a combined $21.7 million, and the intangibles garnered from Broward, which were valued by Mr. Plastino at a total of $28,307,018. We must also take out the value of the tax benefits gained, which we found to be $47.5 million. Those sums are, however, fair market value at the time of the transaction and not the adjusted amounts. We do not have an adjusted basis for the branching rights value, and we must ask the parties to perform the final calculation of the adjusted basis. Plaintiff likewise proved that it should have been able to exclude $24,234,000 of recapitalization costs awarded by Judge Smith as reliance damages. No tax benefit previously was enjoyed by Glendale from these costs, and so plaintiff is entitled to a rebate of the tax paid on that amount included in income. All other claims are denied as explained above. Accordingly, the parties are directed to confer and file a joint status report on or before July 21, 2023, regarding appropriate language and the correct amount for judgment. 36 s/Eric G. Bruggink ERIC G. BRUGGINK Senior Judge 37
]]>
Some case metadata and case summaries were written with the help of AI, which can produce inaccuracies. You
should read the full case before relying on it for legal research purposes.
This site is protected by reCAPTCHA and the Google
Privacy Policy and
Terms of Service apply.
