This is a Winstar related case in which the Court of Federal Claims held that the government had breached an agreement with a savings and loan company that the latter could use a particular method of accounting in determining its capital for regulatory purposes, and awarded damages of approximately $18,600,000. In its appeal the government has challenged only the award of damages. It contends that the breach of the agreement did not cause the injury for which the plaintiffs were awarded damages. We affirm.
I
A. The facts relating to the financial problems of the savings and loan industry in the early 1980s and the federal government’s attempts to alleviate the situation are well known and need only to be briefly summarized here. At that time a large number of savings and loan companies (also known as thrifts) were in serious financial straits and facing insolvency. Federal regulators devised a program under which economically healthy thrifts would acquire financially-distressed ones. To encourage such action, the regulators offered various benefits to the acquiring thrifts, which usually were incorporated in written agreements with them. These included treating the excess of the amount paid for the acquired thrift over that entity’s value (known as “regulatory goodwill”) as an asset in determining the acquiring thrift’s compliance with the thrift’s regulatory capital requirements. The agreements also permitted the acquiring thrifts to amortize their “regulatory goodwill” over a substantial period, generally 25 years.
In 1989 Congress enacted the Financial Institutions Reform, Recovery and Enforcement Act of 1989 (“FIRREA”), Pub.L. No. 101-73, 103 Stat. 183 (Aug. 9, 1989), which made significant changes in governmental regulation of the savings and loan business. These included prohibiting thrifts from using “regulatory goodwill” as a capital asset for regulatory purposes and requiring them to phase out that practice over a five-year period. In the Act, Congress also prohibited thrifts from using subordinated debt as part of their regulatory capital.
As a result of FIRREA, many thrifts became insolvent and the regulators liquidated them.
A number of thrifts filed suit in the Court of Federal Claims seeking damages on a variety of theories. In United States v. Winstar Corp., 518 U.S. 839, 116 S.Ct. 2432, 135 L.Ed.2d 964 (1996), the Supreme Court upheld this court’s decision that the United States was liable for breach of contract to thrifts with which it had agreed to permit the use as regulatory capital of “regulatory goodwill” that had been created in connection with the thrifts’ acquisition of a failing thrift. In Winstar, the Supreme Court did not address “the appropriate measure or amount of damages” for such breach. Id. at 910, 116 S.Ct. 2432. In a number of subsequent decisions, this court has addressed various damages questions arising in such cases. See, e.g., First Heights Bank, FSB v. United States, 422 F.3d 1311 (Fed.Cir.2005); Granite Mgmt. Corp. v. United States, 416 F.3d 1373 (Fed.Cir.2005); Westfed Holdings, Inc. v. United States, 407 F.3d 1352 (Fed.Cir.2005).
*1317B. The basic facts in this case, as found by the Court of Federal Claims, are largely undisputed.
In the mid 1980s, the appellee Citizens Federal Bank (“Citizens”) and its predecessor and affiliates were a successful and well-managed thrift. At the request of the federal regulators of the thrift industry (“the Regulators”), in 1986 and 1988 Citizens acquired two financially-troubled thrifts. In connection with those acquisitions and to encourage them, the Regulators entered into written Acceptance Agreements with Citizens under which Citizens was authorized to treat the excess of the acquired thrift’s liabilities over the purchase price as “regulatory goodwill,” which would be amortized over 25 years and could be used to satisfy Citizens regulatory-capital requirements. Citizens obtained $35.9 million of “regulatory goodwill” in connection with the 1986 acquisition and $17 million in connection with the 1988 acquisition. Citizens Fed. Bank v. United States, 59 Fed.Cl. 507, 510 (2004) (“Liab.Op.”); Citizens Fed. Bank v. United States, 51 Fed.Cl. 682, 685 (2002) (“Breach Op.”).
Following the enactment of FIRREA, in 1989 Citizens still was in compliance with the regulatory-capital requirements, although its coverage margins were greatly reduced.
Prior to the enactment of FIRREA, Citizens had a large amount of subordinated debt, which its depositors held. FIRREA prohibited the use of such debt as part of a thrift’s regulatory capital. To deal -with that problem, Citizens issued non-cumulative preferred stock in exchange for the subordinated notes. While this exchange improved Citizens regulatory-capital situation, it had an adverse tax consequence for Citizens because although the interest it had paid on the notes was deductible, the dividends paid on the preferred stock that was substituted for them were not.
Citizens also reduced its assets which, in conjunction with the substitution of preferred stock for subordinated debt, enabled it to comply fully with the new regulatory capital requirements that FIRREA mandated.
Citizens then filed the present suit in the Court of Federal Claims, seeking damages for the government’s alleged breach of the provisions in the Assistance Agreements permitting it to use regulatory goodwill as part of its regulatory capital and to amortize it over 25 years. The court first held that the government had breached the Assistance Agreements—a ruling which, as indicated, the government does not here challenge.
The case then proceeded to the damages phase, in which Citizens initially sought damages exceeding $350 million, based on various theories of damages: lost profits, restitution, reliance, and cost of capital replacement. After three rounds of summary judgment motions, in which most of Citizens’ theories were rejected, a trial was held to determine what mitigation damages Citizens was entitled to for the costs it incurred in replacing the regulatory capital it had lost as a result of FIRREA.
The court awarded Citizens’ damages of $18,683,901 consisting of the following elements: (1) its transaction costs of $3,802,901 for the exchange of preferred stock for debt; (2) $266,000 for the difference between the dividend rate on the preferred stock and the interest rate on the subordinated notes; and (3) $14,615,000 as compensation for the diminished cash flow resulting from the adverse tax consequences of the exchange.
In so holding, the trial court ruled that the proper standard for determining whether the government’s breach of contract caused Citizens’ damages was whether the breach was a substantial factor in causing the damages rather than, as the *1318government urged, whether the damages were attributable to and resulted entirely from the breach.
II
As just noted, in assessing damages the trial court applied the “substantial factor” theory of causation, i.e., that a defendant that has breached a contract is liable for those damages the other party to the contract suffered for which the breach was a substantial factor in causing the damages. The government contends that the court used the wrong standard. The proper standard, the government argues, is “but-for” causation, under which the breaching party is liable only for those damages that it directly and entirely caused.
According to the government, the proper standard governing causation in breach of contract claims is that announced by our predecessor, the Court of Claims, more than 100 years ago in Myerle v. United States, 33 Ct.Cl. 1 (1897). There, the Court of Claims held that a “plaintiff can only recover those items of damage which are the proximate result of the acts of the Government____ For a damage to be direct there must appear to be no intervening incident (not caused by the defaulting party) to complicate or confuse the certainty of the result between the cause and the damage; the cause must produce the effect inevitably and naturally, not possibly or even probably.... There must not be two steps between the cause and damage.” Id. at 27. The government contends that under this standard, which it urges this court consistently has followed, it should have been awarded summary judgment and Citizens should not have recovered anything.
Our cases dealing with the proper standard of causation may appear superficially somewhat inconsistent in applying the “substantial factor” and “but for” theories. We discern a common thread among them, however: the selection of an appropriate causation standard depends upon the facts of the particular case and lies largely within the trial court’s discretion. The standard is comparable to that governing selection of an appropriate methodology for determining damages, which lies within the trial court’s discretion. See Cyber Corp. v. FAS Techs., Inc., 138 F.3d 1448, 1461 (Fed.Cir.1998) (en banc) (“The amount of damages determined by a district court is a question of fact that is reviewed for clear error on appeal, while the method used by a district court in reaching that determination is reviewed for an abuse of discretion.”) (citation omitted).
In its most recent case dealing with the issue, Indiana Michigan Power Co. v. United States, 422 F.3d 1369 (Fed.Cir.2005), this court stated that
Damages for breach of contract are recoverable where: (1) the damages were reasonably foreseeable by the breaching party at the time of contracting; (2) the breach is a substantial causal factor in the damages; and (3) the damages are shown with reasonable certainty.
Id. at 1373 (citing Energy Capital Corp. v. United States, 302 F.3d 1314, 1320 (Fed.Cir.2002)).
In Energy Capital, the Court of Federal Claims applied the “substantial factor” causation theory. It explained:
Analyzing Ramsey [Ramsey v. United States, 121 Ct.Cl. 426, 101 F.Supp. 353 (1951)] and other cases distinguishes between cases where the lost profits were claimed under other contracts and cases where lost profits were claimed directly under the contract with the United States. Because in this case Energy Capital seeks lost profits flowing from the breach of the contract with the United States, Ramsey does not impose *1319a high burden with regard to causation____ For these reasons, this Court rejects the Defendant’s argument, based on Ramsey, that the Plaintiff must prove that the breached [sic] caused its losses “inevitably.” Instead, the Court will require the Plaintiff to prove that the breach was a “substantial factor” in causing its losses, the test in the majority of jurisdictions.
47 Fed.Cl. at 395 (citation omitted). This court affirmed the Court of Claims’ award of lost profits without commenting specifically on the use of the substantial-factor standard. Energy Capital Corp., 302 F.3d at 1328-29 (“We do not agree [with the government] that Energy Capital’s lost profits were overly remote and speculative as a matter of law.”).
In its earlier decision in Bluebonnet Savings Bank v. United States, 266 F.3d 1348 (Fed.Cir.2001), this court also upheld and apparently approved the Court of Federal Claim’s use of the “substantial-factor” test in a Winstar-related case. There the thrift “s[ought] to recover the increase in financing costs caused by the passage of FIRREA which breached the capital, subordinated debt, and dividend forbearances.” Id. at 1355. In rejecting the trial court’s failure to award damages, this court stated: “The Court of Federal Claims properly determined that the breach of the forbearances was a substantial factor in Bluebonnet’s increased financing costs....” Id. at 1356.
There have been a number of Winstarrelated cases in which this court, in sustaining the Court of Federal Claims’ award or rejection of lost profits damages, approved that court’s use of the “but-for” theory of causation. See, e.g., First Heights Bank, FSB v. United States, 422 F.3d 1311 (Fed.Cir.2005); La Van v. United States, 382 F.3d 1340 (Fed.Cir.2004). We do not read those cases, however, as announcing any broad rule that the “but-for” theory of causation must always, or even generally, be used in determining damages in Winstar-related cases or prohibiting the trial court from using the “substantial factor” test in appropriate cases.
California Federal Bank v. United States, 395 F.3d 1263 (Fed.Cir.2005), is not inconsistent with this analysis and conclusion. There the Court of Federal Claims applied a “definitively established” standard of causation. The thrift contended that the trial court should have applied the “substantial factor” standard. In affirming the trial court, this court stated that “the Court of Federal Claims correctly rejected the ‘substantial factor’ test advocated by CalFed.” Id. at 1268. Although this statement is somewhat broad, we do not read it as announcing a rule that in all Winstar-related cases the standard of causation always is the “but for” one. We view it as another instance in which this court has upheld the causation standard for a particular case as within the court’s discretion. Indeed, less than 8 months after the opinion in California Federal issued, in the Indiana Michigan Power opinion already discussed, this court stated that contract damages are recoverable “where ... the breach is a substantial causal factor in the damages.” 422 F.3d at 1373.
In the present case, the Court of Federal Claims did not abuse its discretion in using the “substantial factor” theory of causation. The court adequately explained the reasons for its action. As the court stated:
The “substantial factor” standard has been reviewed with approval by the Federal Circuit in both Bluebonnet and Energy Capital even though Myerle has not been explicitly overruled. Because the Federal Circuit has accepted this standard, it is consistent with this *1320Court’s analysis in Energy Capital, and because the standard has been adopted in numerous Winstar-related cases before other judges on the Court, this Court concludes that it is the appropriate standard to be applied in the present case. As a result, the Court finds that Plaintiffs may proceed to attempt to establish that the exchange offer was caused by the breaching provisions of FIRREA.
Liab. Op., 59 Fed.Cl. at 515.
Ill
The trial court awarded Citizens approximately $18.6 million in what are called “mitigation damages.” These are intended to reimburse a non-breaching party to a contract for the expenses it incurred in attempting to rectify the injury the breach caused it. See Restatement (Second) of Contracts § 347 cmt. c (1981) (“[T]he injured party is entitled to recover for all loss actually suffered.... [I]nclud[ing] costs incurred in a reasonable effort, whether successful or not, to avoid loss.”). In this case, the damages were awarded to compensate Citizens for the expenses it incurred in replacing its regulatory capital after FIRREA had precluded thrifts from using regulatory goodwill or subordinated debt as regulatory capital.
The government challenges the damages award on two grounds. (A) Since the Assistance Agreements did not explicitly authorize Citizens to continue using its subordinated debt as regulatory capital, Citizen’s refinancing of that debt cannot be attributed to the government’s breach of those agreements; to the extent that the damages reflected those expenses, the government is not liable for them. (B) It was not foreseeable that the government’s breach of the Assistance Agreements would result in the adverse tax consequences that Citizens had suffered as a result of replacing the unsecured debt with preferred stock. Neither argument can prevail.
A. We have upheld the Court of Federal Claims’ use here of the “substantial factor” test for determining whether the breach of contract caused Citizens’ damages. Here the trial court found that “the exchange [of preferred stock for subordinated debt] was due in part to mitigate the effects of lost goodwill and capital credit” and that the record did not support “the Government’s contention that the exchange offer was solely in response to the non-breaching provisions of FIRREA.” Liab. Op., 59 Fed.Cl. at 516. Compare Bluebonnet, where this court ruled that “[t]he Court of Federal Claims properly determined that the breach of the forbearances was a substantial factor in Bluebonnet’s increased financing costs because it forced Bluebonnet to raise capital at a time when FIRREA had made investments in thrifts riskier and considerably less attractive.” 266 F.3d at 1656.
The government argues, however, that the damages award was improper because the trial court did not require Citizens to prove which shares of preferred stock were used to replace “regulatory goodwill” and, further, that whatever costs Citizens incurred to replace “regulatory goodwill” should be separated from its cost of replacing the subordinated debt.
This court’s decision in Home Savings of America, FSB v. United States, 399 F.3d 1341 (Fed.Cir.2005), requires rejection of this contention. There a thrift, following the enactment of FIRREA, issued stock to replace the “regulatory goodwill” it had lost and to preserve the large capital cushion it had maintained as a conservatively-run institution. Home Sav., 399 F.3d at 1352-53. The government argued that the Court of Federal Claims erred in awarding damages covering the costs associated with the sale of stock because the thrift “raised *1321capital for a variety of reasons and did not specify which portion of the capital it raised to replace supervisory goodwill.” Id. at 1353-54. In upholding the damages, this court stated that the thrift’s “strategy of raising capital through various types of financing was a commercially reasonable effort to maintain its debt-to-equity ratio, and fair and reasonable efforts to mitigate are all that the law requires,” id. at 1353, and that the trial “court’s approach to calculating the benefits of cash was ... within the court’s sound discretion” even though the court did not require the thrift to specify which portion of the capital was used to replace the lost “regulatory goodwill,” id. at 1354-55.
Here, as in Home Savings, the trial court did not err in awarding damages for Citizens’ cost of issuing preferred stock, because that was a reasonable effort to mitigate the loss of “regulatory goodwill” even though Citizens had not shown which particular shares were used to replace the regulatory goodwill. Indeed, it is difficult to see how Citizens could have allocated particular portions of the preferred stock to such replacement.
B. “Foreseeability is a question of fact reviewed for clear error.” Bluebonnet, 266 F.3d at 1355 (citing Landmark Land Co. v. Fed. Deposit Ins. Corp., 256 F.3d 1365, 1378 (Fed.Cir.2001)). The trial court found that “it was foreseeable at the time of the contract that Citizens would have to replace the capital credit and goodwill in order to continue to be a self-sufficient institution” if the government breached the “regulatory goodwill” provisions of the Assistance Agreements. Liab. Op., 59 Fed.Cl. at 520. That finding, which the record supports, is sufficient to sustain the damage award. Citizens was not required also to show that it was foreseeable that, in replacing such capital, it would incur negative tax consequences. The foreseeability requirement reflects the principle that a breaching party should not be liable for damages that “it did not at the time of contracting have reason to foresee as a probable result of such a breach.” Restatement (Second) of Contracts § 351 cmt. a (1981). If it was foreseeable that the breach would cause the other party to obtain additional capital, there is no requirement that the particular method used to raise that capital or its consequences also be foreseeable. See Joseph M. Perillo, 11 Corbin on Contracts § 56.7 at 108 (2005 rev. ed.) (“What is required is merely that the injury actually suffered must be one of a kind that the defendant had reason to foresee and of an amount that is not beyond the bounds of reasonable prediction.”).
The decisions of this court reflect and apply that principle. In the present case, the Court of Federal Claims relied on our decision in Bluebonnet in finding that the “tax consequences here are similar to other costs that have been found to be foreseeable and caused by the breach.” Liab. Op. at 520. In Bluebonnet, this court affirmed the Court of Federal Claims’ finding that it was foreseeable that a thrift would be “forced to seek even more capital to meet heightened regulatory requirements” and that it was also foreseeable that the costs and risks associated with meeting these requirements would have other negative economic consequences such as “increasing] the costs of securing debt or equity financing.” 266 F.3d at 1356; see also Home Sav., 399 F.3d at 1355 (Fed.Cir.2005); cf. S. Cal. Fed. Sav. v. United States, 422 F.3d 1319, 1336-37 (Fed.Cir.2005).
Old Stone Corp. v. United States, 450 F.3d 1360 (Fed.Cir.2006), on which the government relies as establishing that the trial court’s finding of foreseeability was clearly erroneous, does not compel that conclusion. In Old Stone, this court ruled *1322that the Court of Federal Claims had erred when it found that a thrift’s seizure by the government was foreseeable at the time of contracting. Id. This court pointed out that for the government’s seizure to be foreseeable, the parties, at the time of contract formation, would have had to foresee four factual situations. It was the lack of evidence to support any of these assumptions that led this court to reverse the trial court’s finding of foreseeability. Id. The situation in that case was quite unlike that in the present case, and Old Stone does not support the government’s contention.
CONCLUSION
The judgment of the Court of Federal Claims awarding Citizens damages of $18,683,901 million is
AFFIRMED.