In this Winstar-series case, the United States appeals the award by the Court of Federal Claims, Senior Judge Loren Smith, to Glendale Federal Bank, FSB ("Glendale") of $381 million in "wounded bank" damages; Glendale cross-appeals the trial court's denial of another $527.5 million in reliance damages.1

This is the second appeal on the damages question. The first appeal arose after a lengthy trial, which resulted in an award to Glendale of $909 million in restitution and "non-overlapping reliance damages." That trial itself was the result of a remand from an earlier consolidated appeal to this court in which the United States challenged the trial court's determinations regarding liability. The trial court had held that the Federal Savings and Loan Insurance Corporation ("FSLIC") had entered into regulatory capital contracts with three savings and loan "thrifts," Glendale, Statesman Savings Holding Corp., and Winstar Corp., and that the enactment of the Financial Institutions Reform, Recovery, and Enforcement Act of 1989 ("FIRREA"),2 and the regulations imposed thereafter, had breached those contracts. We affirmed the trial court's liability judgment, Winstar Corp. v. United States, 64 F.3d 1531 (Fed.Cir.1995) (en banc), and the United States Supreme Court affirmed ours, United States v. Winstar Corp., 518 U.S. 839, 116 S. Ct. 2432, 135 L. Ed. 2d 964 (1996).

Liability in favor of the plaintiff thrifts having been established, the case was returned to the trial court for determination of damages owed to plaintiffs. The trial court found for the plaintiff Glendale in the amount of $909 million. The Government again appealed. The appeal was viewed as a test of the difficult issues of measuring damages in what became known as the Winstar cases, of which there were some 120 in the queue. On appeal and after a thorough review of the record and the various damage theories propounded by the parties and employed by the trial court, we concluded that the award of $909 million was not supportable, and vacated the judgment. Glendale Fed. Bank, FSB v. United States, 239 F.3d 1374 (Fed.Cir.2001).

In the course of that first damages trial Glendale had sought "expectancy damages," the kind of damages often associated with lost profits. The trial court rejected that theory, concluding that Glendale's model for expectancy damages in the form of lost profits was implausible. We agreed. The theory did not fit comfortably with the history of these cases in the years following the breach; the problems of proof suggested that any award premised on expectancy damages would be too speculative to uphold.

Faced with a situation in which as a general proposition expectancy damages were ruled out, but with a legitimate claim by the thrift for a remedy, the trial court fashioned a restitution remedy based on the assumed risk that Glendale undertook when it acquired the salvaged bank (Broward Savings and Loan, a Florida thrift). The crux of the restitution component was the market value of the liabilities assumed by Glendale at the time of the transaction, which the trial court viewed as the value of the benefit conferred on the Government by Glendale. Glendale placed the proper amount of restitution damages at $798 million; the Government argued the proper amount was zero. The trial court sided with Glendale, and after adjusting the amount for the net gains it calculated Glendale had received from the transaction, awarded $510 million.

In our review we concluded that the restitution theory was basically flawed, because it was based on an assumption that the non-breaching party was entitled to the supposed gains received by the breaching party, when those gains in the context of these cases were both speculative and indeterminate. Accordingly we vacated the trial court's award of $510 million.

Recognizing that any damages award in these cases was necessarily a matter of post hoc reconstruction of a set of facts that may have never actually transpired, we did agree with the trial court that a viable theory on which damages could be based was a reliance theory. The trial court had considered reliance theory as a basis for an award, and had added specific reliance damages of $381 million to the total award granted plaintiff; these were denominated non-overlapping reliance damages, and identified as damages for post-breach events.

We noted that in reliance damages, the critical event is the breach itself. Consequently, damages resulting from that event may be the result of actions before or after the breach. But these damages had to be real, and reasonably ascertainable. We concluded that using reliance theory as the damages measure for the losses sustained by thrifts as a result of the Government's breach in FIRREA would provide a firmer and more rational basis than the alternative theories argued by the parties. We ended our opinion noting the enormous effort and dedication to these cases by Judge Smith, and suggested that much of the extensive record before the court — compiled over 150 days in trial and constituting some 20,000 pages — might lend itself to re-evaluation consistent with the guidance provided. Which leads to the second appeal, now before us.

Following the first appeal on damages and our remand to the trial court for further consideration of the issues, Glendale moved in the trial court for entry of judgment, asking the court to reinstate the post-breach "wounded bank" portion of its earlier opinion and judgment, based on the prior findings of the court. This was the $381 million. Glendale claimed an additional $527 million, which it alleged to be the actual out-of-pocket loss suffered by Glendale in its Florida division. The total, with certain adjustments, was something over $862 million. In response to Glendale's motion, the Government did its numbers, and, surprise, came up with zero.

As noted at the beginning, Judge Smith awarded Glendale the $381 million, representing the previously-determined amount of Glendale's "wounded bank" damages, and other incidental reliance damages. The trial court rejected Glendale's claim for the additional $527 million, ruling that Glendale was not entitled to recover on the remainder of its claim because its reliance model failed to measure the actual losses sustained by it as a result of the Government's breach. The Government appeals the $381 million award; Glendale cross-appeals the denial of its claim for the additional $527 million.

In response to the trial court's award of $381 million in reliance damages, the Government notes that the Glendale-invented notion of "wounded bank" damages is basically ($335 million of the $381 million) the increase in the cost of funds Glendale allegedly incurred as a result of its having been wounded by the Government's breach. These losses were said to have occurred because three years after, and as a result of, the breach, Glendale fell out of capital compliance; depositors and others became nervous about placing funds with it; the bank was required to pay more interest to attract depositors; and it was required to pay higher fees for deposit insurance.

But, argues the Government, in proving that it would not have fallen out of capital compliance but for the breach, Glendale relied upon the same model it presented in support of its earlier claim for lost profits under the now-discredited expectancy damages theory. According to the Government, since Glendale did not offer any other evidence, it failed to prove that it would have remained in capital compliance in the absence of the breach. The Government offered several other reasons why Glendale's proof on the issue failed, including that the proof regarding what happened three years after the breach was too remote and speculative, it was based on the costs of acquiring the Florida thrift, a cost element that had been rejected by the courts, and in any event Glendale actually benefited, rather than was hurt, by acquiring the Florida thrift.

The trial court was unimpressed with the Government's arguments. With regard to the reinstatement of the "wounded bank" damages, the trial court was of the view that there was nothing in our remand instructions that required the court to revisit its prior findings and award of post-breach reliance damages. The trial court is correct; indeed, we invited the court to proceed on the basis of the record before it. However denominated, the focus of a recovery based on the reliance interest is the real costs incurred for capital and services that the thrift would not have incurred but for the contract and its subsequent breach. Whether in a given case this properly includes the higher costs to a thrift of conducting its general business after FIRREA — the "wounded bank" claim — is a matter of proof; if too speculative, it can and should be denied as the burden lies with the plaintiff to establish its damages. In the case before us, we are unpersuaded that the trial court's factual findings and conclusions regarding reliance damages, based on the earlier record, are clearly erroneous; the judgment in that regard is fully consistent with the law as we enunciated it. The award of reliance damages in the amount of $381 million is affirmed.

With regard to Glendale's cross-appeal on the operating losses allegedly sustained by Glendale in its Florida venture, the trial court after hearing all the evidence concluded that there was a failure of proof: "plaintiff has failed to persuade the court that its reliance damage model shows any `actual losses sustained by plaintiff as a result of the Government's breach.'" 54 Fed. Cl. at 13 (quoting Glendale, 239 F.3d at 1383). The court explained with some care why that was so, including an extensive review of the report of plaintiff's expert witness, which, in the court's judgment, reflected an absence of an accurate accounting of the actual losses. The court went on to note that the framework for the calculation of reliance damages as articulated by the Federal Circuit focuses on actual out-of-pocket losses, not paper calculations of losses, and that much of Broward's paper deficit was eliminated by the reduction in interest rates.

These are factual assessments by the trial court based on a thorough review of the record and long-familiarity with the evidence, much of which related to a complex series of events occurring many years before. We are not inclined, indeed, it would be highly speculative for us on the limited record we have before us, to second-guess the court's determinations in this regard. Put in the traditional language of judicial review, we again are unpersuaded that there has been a showing of clear error by the trial court. The denial of Glendale's claim for additional damages is affirmed.

Finally, the Government as its ultimate fallback argues that, even if Glendale is entitled to reliance/"wounded bank" damages of $381 million, the Government is entitled to an offset of $243 million. This is an amount that Glendale incorporated as a gain to Glendale in its accounting for losses when it sold the Florida division, resulting in the net claim of $527 million. That claim of course was denied by the trial court, a denial which we have affirmed. Nevertheless, the Government argues that since Glendale acknowledged this amount in its calculation of net losses, it should be required to deduct it from its claim for reliance damages.

Glendale responds that the burden is on the Government to prove the offset, and that to prove an offset against the awarded reliance damages the Government would have to show that Glendale received a net benefit from its Florida franchise. We agree that Glendale's calculation of its denied claim for losses regarding the Florida division is not determinative of its other operational losses as a result of the breach. The trial court did not deduct the $243 million from its final award to Glendale, and we see no reason that we should do so.

In sum. A large part of the problem with which the trial court has had to wrestle has been finding a viable damages theory that fits the complex fact patterns of these cases, one that is fair to the damaged thrifts, but is based on real losses sustained so as not to overcompensate for the breach. The overall liability context of contract and breach in these cases is well established by the earlier litigation and appeals in this court and in the Supreme Court. Regarding the damages to be awarded, it was anticipated — and has proved to be the case — that the specific documentary and evidentiary bases on which damages would be based would vary from case to case.

Expectancy damages theory, based on lost profits, has proven itself impractical for these cases, and generally not susceptible to reasonable proof. We have not, however, barred as a matter of law the use of expectancy/lost profits theory, see Cal. Fed. Bank, FSB v. United States, 245 F.3d 1342, 1350 (Fed.Cir.2001), but, given the speculative nature of such a damages claim, one that has yet to be successfully established in any Winstar case, see, e.g., Cal. Fed. Bank v. United States, 54 Fed. Cl. 704 (2002) (remand), experience suggests that it is largely a waste of time and effort to attempt to prove such damages.3

Restitution as a generalized theory for recovery of assumed benefits to the Government, based on non-provable paper costs, has been rejected as explained above. See LaSalle Talman Bank, F.S.B. v. United States, 317 F.3d 1363, 1376-77 (Fed.Cir.2003). On the other hand, we have allowed restitution for the limited purpose of returning the acquiring thrift to the status quo ante when specific initial contributions to an acquired thrift have been established. See Landmark Land Co. v. FDIC, 256 F.3d 1365 (Fed.Cir.2001). In LaSalle, we recognized that when restitution damages are based on recovery of the expenditures of the non-breaching party in performance of the contract, the award can be viewed as a form of reliance damages. 317 F.3d at 1376.

Reliance damages, however, are supportable when based on actual losses that are fully proven. "Reliance is an ideal recovery in Winstar cases. Despite the landscape where alternative forms of recovery are speculative and loss models inherently unreliable, reliance damages can be ascertainable and fixed."4 Proof of reliance damages are factual determinations to be made by the trial court on the basis of the evidence presented by the plaintiff thrifts. Given the complexities of reconstructing the transactions and their consequences, absent a compelling showing of clear error, that should be the end of the matter.

We have noted that "`[i]f a reasonable probability of damage can be clearly established, uncertainty as to the amount will not preclude recovery,' and the court's duty is to `make a fair and reasonable approximation of damages.'" Bluebonnet Sav. Bank, FSB v. United States, 266 F.3d 1348, 1356-57 (Fed.Cir.2001) (quoting Locke v. United States, 151 Ct. Cl. 262, 283 F.2d 521, 524 (1960)). The Government indicates that there remain some forty-five Winstar cases in the Court of Federal Claims awaiting decision. Optimistically, in our first remand on the damages issue we noted: "Though factual differences may require some fine-tuning of results, we expect that the guidance given to the trial court in this opinion will assist the court, and the parties, in disposing of the remaining cases as well." Glendale, 239 F.3d at 1376.

We remain optimistic that with the additional guidance and support given the trial court in this opinion the remainder of the Winstar cases can be disposed of either by settlement or by trial based on the particular facts of the case, and without further dispute over the theory on which damages may be calculated. It would benefit the thrifts and the Government, since it is not in the interest of either to have endless litigation, for both to stop arguing extreme positions and promptly resolve these cases in a fair and even-handed manner.

AFFIRMED.