First Heights Bank, FSB v. United States

OPINION

BRUGGINK, Judge.

This is an action for breach of contract. It is generally related to the Winstar¹ group of cases and more specifically to a subset of cases claiming breach in connection with the “Guarini” legislation.² Liability has previously been found in favor of plaintiffs. Pending are the parties’ cross-motions for summary judgment on damages. Also pending are plaintiffs’ October 7, 2002 motion to strike, plaintiffs’ February 10, 2003 motion for an order requiring defendant to state its position, defendant’s April 1, 2003 motion to strike, and defendant’s April 9, 2003 motion to strike portions of plaintiffs’ oral argument. The court held oral argument on March 24, 2003. For the reasons set out below, plaintiffs’ motion for summary judgment is granted in part and denied in part. Defendant’s motion for summary judgment is granted in part and denied in part. All other motions are denied as moot.

BACKGROUND

Background facts in this litigation can be found in First Heights Bank, FSB v. United States, 51 Fed.Cl. 659 (2001) (“First Heights I”) and First Heights Bank, FSB v. United States, 53 Fed.Cl. 195 (2002) (“First Heights II”). Familiarity with those opinions is presumed. In First Heights I, the court held that the government breached the implied covenant of good faith and fair dealing by enacting Guarini, because it eliminated tax deductions in a targeted, retroactive fashion, and that the doctrine of prior material breach did not bar plaintiffs from recovering. In First Heights II we held that Pulte Corporation,³ though it was not a party to the Assistance Agreement, could remain a party to this lawsuit. We now resolve the issue of damages.

For all relevant years, plaintiffs were members of a consolidated group for purposes of filing their federal income tax returns. (The name “Pulte” or the term “plaintiffs” will be used herein to refer to the consolidated group or to the current or prior parent, unless specific reference only to one of the plaintiffs is more appropriate in the context.) Electing to file as a consolidated entity means that all inter-company accounts were ignored and the assets, liabilities, and equity reported on the Pulte group’s consolidated balance sheet reflected the group’s combined amounts. Pulte (or its predecessor) was the common parent of the Pulte consolidated group and paid the consolidated group’s federal tax liability to the IRS.

In 1988, Pulte acquired First Heights Bank from the Federal Deposit Insurance Corporation (“FDIC”)⁴ and executed an Assistance Agreement in accord with that transaction. Pursuant to that contract, the FDIC agreed to reimburse plaintiffs for losses that plaintiffs realized upon disposition of covered assets. Under the contract, covered assets generally included all assets of the acquired institutions. The parties accounted for these covered asset losses (“CALs”) in a special reserve account maintained by First Heights. This account was audited and sometimes adjusted by the FDIC during the term of the Assistance Agreement.

Under the tax law then in place, i.e., preGuarini, the receipt of assistance prompted by covered asset losses did not preclude the thrift from claiming the loss for tax purposes. See Centex Corp. v. United States, 48 Fed.Cl. 625, 632-36 (2001). Under the terms of the agreement, the FDIC received a twenty five percent share of those tax benefits and thus had an interest in the accuracy of the special reserve account.

In 1993, Congress enacted Guarini, which prevented plaintiffs from claiming tax deductions relating to CALs which had been reimbursed by the FDIC. Guarini retroactively prevented these deductions back to March 1991. When Congress passed Guarini, the United States Department of the Treasury prepared revenue estimates to quantify its effect on the treasury. It estimated that, between calendar years 1989 and 1999, Guarini would generate an additional $213 million in income tax revenue from the Pulte consolidated group. The Department of the Treasury also estimated an associated $41 million drop in revenue to the FDIC.

Plaintiffs anticipated passage of Guarini and, in 1993 ____

In preparation for this litigation, plaintiffs began a project known internally as the “Tax Project File.” The purpose of this project was to determine and document First Heights’ CALs and tax charge-offs on an asset-by-asset basis for all covered assets. The completed Tax Project File Summary constitutes a list of all covered assets for each acquired institution with related CAL activity and tax charge-off history. Ms. Linda McCall, who submitted an affidavit as plaintiffs’ tax accounting expert, concluded that the Tax Project File Summary reflects, in all material aspects, tax charge-offs as documented in plaintiffs’ federal income tax return workpapers. It also reflects CAL activity as reported in the special reserve account filings reported to the FDIC.⁵ Based on her review, plaintiffs’ damages model assumes that there were no material differences between book and tax basis for the acquired covered assets. Documented book losses thus represent tax losses in the same amount. The FDIC never challenged this assumption about book and tax basis.

Using the information from the Tax Project File Summary and other supporting documents, Ms. McCall calculated plaintiffs’ total potential CALs to be $711,534,337. Plaintiffs’ actual CALs claimed pre-Guarini were $541,483,676. By subtracting one number from the other, Ms. McCall was able to arrive at the total CALs disallowed by Guarini, $170,050,661.⁶

To show the effect of the disallowed CALs on their finances, plaintiffs prepared a “with and without” calculation. These calculations are used for a variety of business and tax practices and procedures. The purpose of the with and without methodology is to isolate the incremental tax effect of a particular change in an existing or contemplated tax return position. The “without” column reflects the returns of the taxpayer as actually filed and the “with” column reflects the returns as they would look if one of the items on the return (e.g., a CAL deduction) were to change. Plaintiffs used the with and without calculation to isolate the incremental tax effects of Guarini on plaintiffs’ “as filed” tax returns for all relevant tax years.

Plaintiffs prepared a separate with and without calculation for each relevant tax year founded upon actual entries from plaintiffs’ consolidated federal income tax returns, amended returns, and supporting workpapers. The with and without calculations demonstrate that the $170 million in CALs that plaintiffs lost as a result of Guarini would not have begun to substantially reduce plaintiffs’ federal income tax liability until tax year 1998. Until that year, plaintiffs had enough pre-existing net operating losses (“NOLs”) to offset taxable income. The carryforward of the NOLs generated from plaintiffs’ CALs would have substantially reduced plaintiffs’ federal income tax liability in 1998 and would have continued to reduce liability in 1999, 2000, and 2001. The FDIC, which had a direct stake in plaintiffs’ returns, never challenged plaintiffs’ use of NOLs or other tax planning techniques.

First Heights met the thrift asset test of I.R.C. § 7701(a)(19) and qualified as a thrift for all tax years beginning in 1988 until it ceased banking operations in 1998. For those tax years, plaintiffs determined their bad debt reserve using the reserve method prescribed by I.R.C. § 593. Under the reserve method, bad debt deductions could be computed in one of two ways — the percentage of taxable income method or the experience method. Plaintiffs used the experience method in both 1988 and 1990 and the percentage of taxable income method at all other relevant times. As of December 31, 1990, the end of the last year in which plaintiffs were allowed to take CAL deductions, First Heights’ ending bad debt reserve balance totaled____⁷

Because of Guarini and its retroactivity to 1991, plaintiffs did not claim or take a bad debt deduction for tax years 1991 through 1996. In 1996, Congress enacted I.R.C. § 593(f), which terminated the reserve method for thrifts. This statute also required thrifts previously using the reserve method to recapture their excess bad debt reserve balances into the taxpayer’s taxable income ratably over a six year period beginning in 1996. This provision was known as the “bad debt reserve recapture requirement.”

Pulte, therefore, was required to recapture the ... reserve balance into taxable income. It recaptured one-sixth of the reserve into First Heights’ taxable income in both 1996 and 1997. In 1998, however, First Heights ceased all banking operations. Based on Pulte’s own independent research and the recommendation of its thrift tax adviser____ This position, actually documented on Pulte’s returns, is reflected in plaintiffs’ with and without calculations.

The with and without calculations also reflect Pulte’s actual position____⁸ For these calculations ____ Pulte believed this treatment was proper. The FDIC did not at any time question Pulte’s ... positions____

In sum, plaintiffs demonstrate that the passage of Guarini resulted in the payment of an additional $68,130,105 in federal taxes. In order to reflect the fact that the FDIC was entitled to twenty-five percent of the taxes avoided, this gross amount must be reduced by FDIC’s share, plus interest.

Some of the states for which Pulte filed returns required Pulte to use its federal tax return figures to compute state income tax liability. Guarini, therefore, directly affected Pulte’s state income tax liability in these states. Plaintiffs did with and without calculations for those states to determine Guarini’s effect on state tax liability. These calculations are based on actual income tax returns and supporting workpapers although, as discussed below, with respect to ... the damages figure must be based on future projections. The damage calculation for state taxes is net of the related federal tax benefit, because state taxes are deductible for federal income tax purposes.

The majority of plaintiffs’ claim for increased state taxes due to Guarini relates to the calculated state tax benefit in ... for 2004 to 2006 ($1,095,909 of the total $1,989,186 state income tax claim). Unlike calculations for federal taxes, which are based on historical data, the impact of Guarini on ... taxes can only be based on a projection of the future. This is due to the fact that NOL carryforwards attributable to ... have not yet been fully used. Those NOLs have to disappear before deductions can be taken. Plaintiffs’ current damages claim anticipates using current ... NOLs before they expire in 2007. In fact, plaintiffs now project that, based on current taxable income forecasts, the ... NOL carryforward will more likely than not be fully utilized even sooner-by 2003. These projections are based on plaintiffs’ recent acquisition of Del Webb Corporation and the purchase of a very substantial parcel of land in____These acquisitions increase the percentage of plaintiffs’ taxable income apportionable to ... and thus increase the likelihood of usage of the NOLs prior to expiration. Because the tax savings would only accrue in the future, plaintiffs discount their savings to the present at a six percent rate.

The FDIC and Pulte executed a Settlement and Termination Agreement to settle unrelated litigation. That contract requires Pulte to limit the request for damages in this case to a percentage of the total available. In computing damages here, plaintiffs have deducted the FDIC’s calculated share of damages from that contract. Net of that share, plaintiffs’ total claim for lost federal and state tax benefits is $48,683,620.

Another element of plaintiffs’ damages is a claim for $2,136,497 in borrowing costs which they assert were precipitated by the loss of income net of taxes. Plaintiffs alleged that they expected to use their calculated tax benefits for 1997 (or a portion of them) to pay down the revolving lines of credit during that year, which would have reduced plaintiffs’ borrowing costs. The weighted average rate of interest paid by plaintiffs on their revolving credit line during 1997 was ... percent. Plaintiffs alleged that Pulte would have used its calculated tax benefits for 1999 to fund homebuilding projects for which it instead used credit. The weighted average interest rate on Pulte’s lines of credit during 1999 was ... percent.

An additional major element of plaintiffs’ damages consists of lost profits which they assert could have been earned on the additional income retained net of taxes. They assert that lost profits amounted to $48,369,496 during 2000 through 2002. Part of plaintiffs’ proposed undisputed findings of fact thus relate to the negotiating background and to the evidence that there is a causal connection between the breach and those asserted lost profits. Unlike the facts set out above, which we find to be materially undisputed, the following facts are disputed in part. We set them out below, however, to assist in framing plaintiffs’ claim for consideration as a matter of law.

During negotiations, Pulte advised defendant by letter on March 31, 1988 that Pulte was one of the largest homebuilders in the United States, as well as a mortgage banker, with 32 straight years of profits. Pulte proposed sharing tax benefits based on the consolidated taxable income of all business operations. Pulte’s 1998 proposal to the FDIC states:

Since PHC proposes to share tax benefits based on the profits from all of its business operations, not just profits that may be earned from the acquired [thrifts], this tax sharing proposal is also a very substantial profit sharing proposal.

To demonstrate this clearly, we point to the fact that PHC, a New York Stock Exchange company, is one of, if [not] the, largest home builders in the United States, as well a very large mortgage banker. It has had an unparalleled and unbroken record of 32 straight years of profits. These annual operating profits have averaged over $50 million in the last 5 years.

This tax benefit and profit sharing proposal, therefore, is very real[,] not speculative. It compares favorably to proposals wherein the [FDIC] is to receive tax benefits and/or profit sharing from thrifts or other institutions that are undercapitalized, have created substantial unrealized losses on their books relating to underwater assets, such as real estate in depressed markets.

The final Federal Home Loan Bank Board meeting approving the transaction recognized the homebuilding revenues and profits of the entire Pulte consolidated group as one of many components of the transaction.

Plaintiff Pulte Diversified Companies, Inc. (“PDCI”) has been a wholly-owned subsidiary of Pulte since 1987. Since that time, the majority of homebuilding operations have been conducted by wholly-owned subsidiaries of PDCI and/or their wholly-owned subsidiaries or divisions. First Heights is also a wholly-owned subsidiary of PDCI. Neither First Heights, however, nor any of its subsidiaries conduct any homebuilding business. Although plaintiffs and their direct and indirect subsidiaries report their audited financial data on a consolidated basis, the term “Pulte” or “plaintiffs” used in the following paragraphs of this section should not be understood to include First Heights.

Pulte engages in a capital planning process each year that results in adoption or revision of a capital planning model for plaintiffs’ homebuilding business. Pulte’s geographic subsidiaries and divisions provide historical data regarding capital expenditures, revenues, expenses, and returns on invested capital (“ROIC”) in new and existing homebuilding projects in each of the markets in which plaintiffs operate. The geographic subsidiaries and divisions also identify prospective homebuilding opportunities in their respective markets and project capital expenditures, revenues, and ROIC for the upcoming year and the next five years. Pulte’s management assembles and analyzes this historical and projected data into a national business plan and presents it to Pulte’s Board of Directors at the end of each year. This national, or overall, business plan reflects actual and projected ROIC for the plaintiffs’ homebuilding operations. Pulte also prepares a quarterly “Competitor and Economic Package” that is presented to its Board of Directors, which includes information pertaining to Pulte’s ROIC and debt to total capitalization ratio.

The ROIC that Pulte realizes upon its homebuilding operations vary widely from project to project. When Pulte evaluates a potential homebuilding project, it does not assume in every case that it will receive a net positive ROIC soon after beginning the project. An expectation to receive a net positive ROIC within one quarter of beginning a homebuilding project would be “very aggressive.” Generally, Pulte experiences net cash outflows in the early phases of a homebuilding project and positive ROIC in the later phases. Some homebuilding projects undertaken by Pulte can be completed in a year, but it is not unusual for Pulte to complete projects in two or three years.

In a document created by Pulte entitled “Ranking of 2002 Annual Plan, Operational Statistics by Market,” Pulte ranked its home-building operations in ... as having some of the lowest ROIC of all its markets. This same document ranked ... market, with a ROIC of ..., as 48 out of its 63 markets and the ... market, with a ROIC of ..., as 51 out of its 63 markets. Pulte’s average ROIC in its homebuilding operations was ... during 2000, ... for 2001, and is projected to be ... for 2002. Pulte projected that its ROIC in 2002 on its Del Webb projects would be about ____

Plaintiffs have calculated the cash basis tax benefits that would have been available in each quarter during the years 1997 through 2002 had Guarini not been enacted. Mr. Roger Cregg, CFO of Pulte Homes Inc., assumed that, had Guarini not been enacted, the tax savings calculated would have been included in Pulte’s capital planning models. This would have permitted Pulte to plan on the basis that the funds would become available during the periods indicated (or sooner) and to engage in new and existing home-building projects. Plaintiffs’ lost profits and borrowing cost claims are based on this assumption.

Mr. Cregg was not willing to authorize plaintiffs to borrow against their lines of credit to fund these projects because of concern that this would cause Pulte to exceed its 40% debt to total capitalization ratio (“debt/ cap ratio”) and thereby risk a downgrading or elimination of Pulte’s investment grade rating. The debt/cap ratio is a fraction, the numerator of which is debt related to plaintiffs’ homebuilding business and the denominator of which is the sum of such debt plus equity. Pulte targets a 40% debt/cap ratio as part of its policy to maintain an investment grade rating. This goal allows Pulte the flexibility and risk protection provided by a conservative balance sheet and is viewed as more important than the short-term reduction in the weighted average cost of capital that might result from the utilization of higher debt.

Credit rating guidelines play an important role in the capital structure policy. Pulte consciously strives to maintain those ratings by adhering to a suggested debt level and coverage ratio and by maintaining good communications with the rating agencies. Pulte does not want to risk a rating below investment grade and strives to maintain an investment grade rating to preserve access to capital at affordable rates. This puts an upper limit on debt which, in turn, puts an upper limit on capital expenditures.

Representatives of the rating agencies have expressed to Mr. Cregg on several occasions the importance of Pulte’s 40% debt/cap ratio to the maintenance of Pulte’s investment-grade rating. On one occasion, Mr. Cregg pointed out to a representative of a rating agency that it had allowed another homebuilder to retain an investment-grade rating even though that homebuilder’s debt to total capitalization ratio substantially exceeded 40%. Mr. Cregg also asked the representative to disclose their “formula” — i.e., the various business and financial factors of a company that they considered in rating companies — so that he could prepare an analysis based on those factors that would justify a debt ratio that was higher than 40%. The representative advised Mr. Cregg that they were comfortable with Pulte at 40%.

Pulte chose 40% as its target debt/cap ratio because it believed that the credit rating agencies did not like to see a ratio in excess of 40 to 45% and “got nervous” when it exceeded 45%. On occasion, Pulte’s ratio would hit 45%, but so long as it was able to manage it back down to 40%, it believed the rating agencies were satisfied. Pulte manages its debt to capital ratio at a “macro,” consolidated level and not on an individual project basis.

The decision to keep the debt/cap ratio at or below 40% was never memorialized in writing. The 40% target is not a hard and fast rule, but rather considered by plaintiffs to be a prudent business practice. It was self-imposed and not required by any lending or contractual agreement or by the Securities Exchange Commission.

On August 31, 2000, Pulte executed a line of credit agreement with a creditor which provided that Pulte would maintain its debt/ cap ratio at or below 50%. One rating agency reported on May 2, 2001, that Pulte’s management stated that it had a “comfort range” for its debt/cap ratio of 35-45%. Pulte’s debt/cap ratio at the end of 1997 was approximately 42%; at the end of 1999, it was approximately 32%; and at the end of 2000, it was approximately 35%. At the end of December 1999, Pulte had over $51 million in cash on hand and at the end of 2000, it had over $183 million. At the end of 2000, Pulte had over $400 million in corporate credit lines available to fund its homebuilding projects.

In 2001, Pulte acquired another homebuilding company, Del Webb. As a result of the acquisition of Del Webb, Pulte’s debt to capital ratio increased to 47.4% at the end of the third quarter of 2001. At the end of 2001, Pulte’s debt/cap ratio was 44.8%. On March 31, 2002, Pulte’s debt/cap ratio was 43.7%. As a result of the acquisition of Del Webb, one credit rating agency downgraded Pulte’s bond rating, but it remained an investment grade rating. Another credit rating agency affirmed its investment grade rating of Pulte after the announcement of the intention to acquire Del Webb.

At least one rating agency publishes a Corporate Rating Criteria Guide that details the criteria used by that agency to rate issuers and specific bond issues. It states in this guide that bond ratings are based on more than just debt/cap ratios and lists numerous criteria that are considered. The guide states that the single most critical aspect of all credit rating decisions is cash flow and cautions against managing a business to achieve a prescribed rating.

In 2000, Pulte made the conscious decision to repurchase $66 million of its own stock, rather than invest in homebuilding projects. Pulte’s repurchase of its stock resulted in an increase in Pulte’s debi/cap ratio. The purpose of this stock repurchase was to increase the value of Pulte’s stock.

For the purpose of addressing plaintiffs’ lost profits and borrowing cost claims, therefore, we assume that plaintiffs attempted to engage in profitable homebuilding projects, subject to the self-imposed constraint of a 40% debt/cap ratio. At the consolidated level, sustaining this ratio helped to maintain Pulte’s investment-grade rating. Plaintiffs’ homebuilding projects were, on average, profitable and plaintiffs allege that they would have used additional tax benefits to fund additional projects. We can also assume that, at the time of the acquisitions, federal regulators were aware that Pulte was in the home building business, was successful at it, and would continue to try and make a profit building houses.

DISCUSSION

Six issues are raised in the parties’ cross-motions for summary judgment: (1) whether plaintiffs have standing as a consolidated group to pursue a claim against the government; (2) whether First Heights actually possessed $170,050,661 in CAL deductions that would have resulted in a $68,130,105 decrease in federal taxes; (3) whether plaintiffs may recover state income taxes unnecessarily paid as a result of the government’s breach; (4) whether plaintiffs are entitled to lost profits; (5) whether plaintiffs may recover for borrowing expenses incurred as a result of the government’s breach; and (6) whether plaintiffs are entitled to a “grossed up” recovery in order to offset income tax liabilities.

I. Standing

Defendant argues that plaintiffs, as individual corporations, have suffered no injury, because First Heights Bank, Pulte Diversified Companies, and Pulte Homes had no taxable income to offset during the relevant period. We disagree. Pulte is the parent of the consolidated group and may assert claims on behalf of the entire consolidated tax entity. See Centex Corp. v. United States, 55 Fed.Cl. 381, 387-88 (2003) (discussing consolidated groups). It is undisputed that the Pulte consolidated group had taxable income during the relevant period.

II. Plaintiffs’CALs

The first component of plaintiffs’ damages consists of the increased federal income tax liability incurred as a result of Guarini. Plaintiffs computed this liability by (1) determining the CAL tax charge-offs and bad debt deductions that Guarini prevented, (2) calculating their income tax liability with and without the disallowed tax charge-offs and deductions, and (3) comparing the with and without liabilities so calculated. Plaintiffs compute their additional federal income tax liability at $68,130,105, as set out above.

Defendant’s expert, Mr. William Wolf, disagrees with plaintiffs’ figure in several ways. Mr. Wolf is a CPA and tax accounting partner with White, Zuckerman, Warsavsky, Luna & Wolf with thirty years of experience in public accounting. We discuss each of his criticisms below.

A. Book/tax difference at acquisition

Plaintiffs concluded that there were no material differences in the book and tax basis of covered assets from the five acquired thrifts. Mr. Wolf states that his review of the evidence demonstrates that there may have been substantial differences in the book and tax basis of covered assets. Write-downs of covered assets for book purposes may be taken at different times than for tax purposes and, in Mr. Wolfs experience, thrifts often claim tax charge-offs for assets when a specific loss reserve was established for book purposes. The book write-down, however, was often not claimed until the asset was sold. Mr. Wolfs concern is that potential tax charge-offs could be taken twice — once by an acquired thrift and then later by First Heights.

Mr. Wolf believes that his review of plaintiffs’ documents reveals instances in which the thrifts’ tax charge-offs may not have been taken for book purposes. He claims that these differences result in at least $4.8 million in overstated charge-offs with a concomitant reduction in tax benefits of $1,689,600. Mr. Wolf believes that this is not the only reduction that must be made because of the book and tax differences, but he makes his estimate based on the limited evidence available. Defendant therefore argues that plaintiffs have failed to meet their burden of proving damages to a reasonable degree of certainty.

Although Mr. Wolf is free to criticize plaintiffs’ tax reporting to the IRS, the court simply uses numbers from plaintiffs’ actual returns to calculate contract damages in this case. See First Nationwide v. United States, 56 Fed.Cl. 438 (2003). Mr. Wolfs adjustments to his “with” column (the “as filed” returns) are misplaced.⁹ ... Mr. Wolfs speculation must be disregarded, because breach damages here are based on a “but for” world based on actual figures. Because plaintiffs are entitled to base their with and without analysis on their returns as filed, in which the book basis and tax basis were equivalent, Mr. Wolfs criticisms are irrelevant.¹⁰

B. Additional charge-offs

Mr. Wolf also believes that CALs should be reduced by an additional $32,688,973 because Ms. McCall’s damages report claimed fewer pre-Guarini tax charge-offs than seemed to be available as of March 31, 1991. According to Mr. Wolf, this shows that First Heights had the ability to further mitigate by accelerating an additional $32.7 million in tax charge-offs into a pre-Guarini year, as Mr. Wolf did. Mr. Wolf says that this contradicts Ms. McCall’s claim that “there were no instances where a tax charge-off should have been claimed that was not claimed.”

We disagree. Initially, even if Mr. Wolf is correct that plaintiffs had $32.7 million in additional charge-offs, his suggestion that plaintiffs should have taken those $32.7 million in charge-offs in pre-Guarini years is misplaced. The law requires only reasonable efforts to mitigate. Mr. Wolf acknowledges

C____adjustment issue

During the evaluation of plaintiffs’ damages claims, Mr. Wolf analyzed plaintiffs’ “with and without” calculation of____Within that calculation, Mr. Wolf found that plaintiffs had____He believes this is contrary to IRS regulations and therefore adjusted plaintiffs’____He then concluded that these adjustments do not affect the amount of lost tax benefits claimed by plaintiffs. He still believes the issue is material, however, because the changes affect the timing of when the tax benefits would be available. This timing, in turn, affects plaintiffs’ claims for lost profits, increased borrowing costs, and a small environmental tax issue.

Once again, Mr. Wolfs position is inconsistent with the fully informed and supported position taken by plaintiffs on their actual returns and is therefore inconsistent with a proper “with and without” methodology. In correcting the plaintiffs’ “errors,” Mr. Wolf testified that he was simply “doing the IRS’s job for them.” ... We do not accept Mr. Wolfs criticisms of the plaintiffs’ tax returns. Instead, we use the numbers from the returns as filed.

D. Bad debt reserve recapture issue Mr. Wolf accurately notes that Guarini was not related to, and did not cause, the bad debt reserve recapture requirement of 1996. He concludes that any loss of CALs resulting from the recapture requirement should not be attributable to Guarini. Mr. Wolf also assumes that First Heights’ decisions to dispose of its assets in 1994 and cease to operate as a thrift in 1998 were not related to Guarini. Mr. Wolf therefore makes two related adjustments to plaintiffs’ damages.

Mr. Wolf first reduces plaintiffs’ damages by the amount of their bad debt reserve recapture. This amount ... represented plaintiffs’ bad debt reserve balance at the end of December 31, 1990 and remained unchanged until recaptured into taxable income from 1996 through 1998. Mr. Wolf explains that, because Guarini did not cause the bad debt reserve recapture requirement, plaintiffs should not be allowed to claim damages for lost CALs equivalent to the amount of their bad debt reserve.

Mr. Wolfs adjustment, however, is incorrect as a matter of law. Plaintiffs’ claim is unaffected by the recapture requirement. Both plaintiffs’ “with” and “without” columns take into account the recapture of the ... bad debt reserve balance, reflecting real-world circumstances. Because both columns reflect this transaction, plaintiffs’ damages are unaffected. Even Mr. Wolf admits that his adjustment results in something “not in essence a with and without calculation.”¹¹

Mr. Wolf next proposes reducing plaintiffs’ claim by $66.6 million on the theory that plaintiffs could have claimed a substantially larger addition to the bad debt reserve than they did as of December 31, 1990. If plaintiffs had used the reserve up to his computed maximum at that time, they would have had additional deductions in 1990 of $66.6 million.¹²

We disagree. Mr. Wolf is once again improperly challenging plaintiffs’ actual tax returns. Even assuming that plaintiffs should have made the suggested changes and deducted an additional $66.6 million in 1990, adding to the reserve balance, Guarini would have prevented subsequent charge-offs to the reserve. Without the ability to lower the reserve balance, plaintiffs would have been forced to include any additional deductions into taxable income when the recapture requirement was enacted in 1996. Plaintiffs, therefore, would have received no actual benefit from Mr. Wolfs hypothetical additional deductions in 1990.

E. Separate Return Limitation Year NOL issue

Finally, Mr. Wolf contends that plaintiffs incorrectly handled the issue of their separate return limitation year (“SRLY”) NOLs — those belonging to the acquired thrifts which were incurred prior to acquisition by plaintiffs. Under SRLY rules, these NOLs can be used only to offset the taxable income of First Heights — not any other members of Pulte’s consolidated group. Ms. McCall concludes that Guarini prevented First Heights from using the SRLY NOLs that expired. Mr. Wolf on the other hand believes that plaintiffs could have used several techniques to utilize the SRLY NOLs before they expired. Mr. Wolf identifies five ways in which the plaintiffs could have avoided these damages. These arguments in effect contend that plaintiffs should have made different business and tax judgments to minimize the impact of Guarini.

Each of Mr. Wolfs suggestions would have required the plaintiffs to alter their “as filed” tax returns as well as their business practices or corporate structure. These suggestions, however, merely reflect his view of how plaintiffs should have conducted their affairs. He has not presented any proof that plaintiffs, in fact, unreasonably or illegally conducted their affairs. His criticisms cannot be taken into account.

III. State Income Tax

Plaintiffs claim damages for their increased state income tax liability in four states. Mr. Gregory Nelson, plaintiffs’ witness, calculated these damages. He prepared a separate with and without calculation for each state. Mr. Wolf, however, disputes plaintiffs’ calculation for____¹³

In ..., plaintiffs had actual existing NOL carryovers of $46 million as of December 31, 2000. Plaintiffs’ damages model also contains $28 million of additional incremental NOL from their 1991-1993 CALs. Plaintiffs’ NOLs would expire in 2007. To show that their damages were caused by Guarini, plaintiffs’ model would have to utilize existing NOLs prior to using deductions derived from CALs. Historically, plaintiffs have utilized such ... — related NOLs at a rate of $3 million to $8 million per year. At those rates, plaintiffs likely would not be able to utilize their NOLs before expiration. Of necessity, plaintiffs’ damage claim was based on projected figures. These damages calculations anticipated using roughly $11 million per year over the next six years (double the historical average), thus making room for a projected future tax benefit attributable to CALs. Based on more current figures, plaintiffs now anticipate a probability of using all ... NOLs by 2003.¹⁴ Plaintiffs then computed the present value of these denied future tax benefits as of December 15,2002.

Mr. Wolf makes three arguments that plaintiffs’ projections are speculative and unrealistic. First, if plaintiffs used their historical figure for ... NOL usage — $6 million per year — then the ... NOLs would expire before they could be used and plaintiffs would have no Guarini damages. Mr. Wolf next suggests that plaintiffs’ figures for apportionment of income to ... are too high. He notes that taxable income attributable to ... declined from 8.2% in 1992 to 2.9% in 2000. Finally, Mr. Wolf disagrees with plaintiffs’ present value computation. He contends that both the present value should be computed no later than the date of trial on damages, not December 15, 2002, and that the present value discount rate should be higher. Mr. Wolf argues for a discount rate of 15%.

This issue was incompletely presented in the briefing. Plaintiffs do not deal directly with Mr. Wolfs criticisms. The court is left with differing opinions as to the proper way to make future projections. The issue, at least as presented, cannot be resolved on summary judgment. We reserve this issue for subsequent determination.

IV. Lost Profits

Plaintiffs believe that breach of the Assistance Agreement caused them to lose not only their tax benefits, but also the income that those benefits would have generated. Plaintiffs contend that they would have used the additional income to invest in home-building projects during 2000-2002. Defendant challenges plaintiffs’ lost profits because they do not flow directly and immediately from the breach of the Assistance Agreement. Defendant argues that, as a matter of law, the lost profits are too remote to be considered a “natural result” of Guarini.

In addition, defendant challenges certain assumptions in plaintiffs’ lost profits model. First, defendant’s expert Dr. James A. Wilcox, who has a Ph.D. in economics from Northwestern University and is a professor at the University of California, Berkeley, attacks plaintiffs’ figures for yearly average ROIC. He argues that plaintiffs’ figures fail to account for the risk inherent in such projects. He suggests instead using the “risk-free” rate represented by the rate of return on a three-month U.S. Treasury bill. As a second option, he suggests a rate of 5.04%.

Defendant also questions whether the lost tax benefits were the sole cause of plaintiffs’ inability to undertake additional homebuilding projects. Defendant claims that plaintiffs’ debt/cap ratio was not a hard and fast rule preventing plaintiffs from investing in other projects. It was not reduced to writing and plaintiffs were often willing to exceed this self-imposed ratio for worthwhile opportunities. Therefore, defendant suggests that Guarini was not the sole cause of plaintiffs’ inability to invest in additional homebuilding projects but that their limitation was self-imposed.

Defendant also offers the report of Mr. James J. Martell, Jr. to rebut plaintiffs’ claim that they were precluded from engaging in available homebuilding projects. Mr. Martell is a certified public accountant with an MBA. Among other business ventures, he founded and served as president of The Fortress Group, a merger of several intermediate-sized homebuilders, and has over ten years of experience in the homebuilding industry. He argues that if plaintiffs had the opportunities in 2000, 2001, and 2002 which they claim to have lost, plaintiffs would have undertaken them despite the lack of additional income from CAL deductions. He hypothesizes that plaintiffs had no profitable opportunities in those years and actually invested in activities other than homebuilding.

The implication is that plaintiffs consistently were shifting capital to other places and not necessarily to the most profitable home-building projects. For example, both Mr. Martell and Dr. Wilcox cite the fact that plaintiffs used $66 million in 2000 to repurchase stock.¹⁵ Moreover, both experts highlight plaintiffs acquisition of Del Webb in 2001, which raised plaintiffs’ debt/cap ratio well above 40%. This acquisition had a projected ROIC of ..., but plaintiffs’ claim an average ROIC for homebuilding projects in 2001 of over____Lastly, Mr. Cregg stated at his deposition that if plaintiffs had received the lost tax benefits in 2002, they would have used it to return to a 40% debt/ cap ratio by paying down credit lines.

Dr. Wilcox also challenges plaintiffs’ assertion that financing all additional homebuilding projects would have unacceptably raised plaintiffs’ debt/cap ratio. He points out that if plaintiffs had financed all of the projects claimed to be forgone, they almost never would have raised their debt/cap ratio above 45%.¹⁶

Defendant argues that five other issues prevent summary judgment for plaintiffs. First, Dr. Wilcox suggests that the averaged ROICs are not helpful, because plaintiffs would have initially invested in projects with the highest rates of return, leaving only the lower ROICs on the table. The average, therefore, is an inflated indicator.¹⁷ Second, plaintiffs have failed to identify forgone homebuilding projects with any specificity. Third, in his opinion, plaintiffs’ projections of how fast they would have profited from investment in homebuilding projects is excessively optimistic and at odds with the facts. Fourth, though plaintiffs believe that no tax sharing payments to the FDIC were required under the Assistance Agreement until October 2001, Dr. Wilcox believes that these payments would have been due sooner, decreasing the potential lost profits projections. Lastly, he believes that plaintiffs’ failure to make ... thereby affecting plaintiffs’ lost profits projections.

We agree with defendant that, if the lost profits claimed were recoverable as a matter of law, there remain multiple disputed issues of fact and expert opinion precluding summary judgment. It is unnecessary to preserve this issue, however, because we also agree with defendant that the precise lost profits claimed are not recoverable as a matter of law.

The contract the government breached was the promise, in good faith, not to deal unfairly with plaintiffs by interfering with their ability to reap the benefits of the Assistance Agreement. This agreement included the favorable tax treatment flowing to the consolidated group. What the parties carefully negotiated, at least insofar as relevant here, was the opportunity to reap those benefits without interference. Both parties, as demonstrated by the tax sharing plan, understood that the arrangement would produce tangible monetary benefits to each. We can readily posit, moreover, that the representatives of the United States knew that plaintiffs were heavily involved in the homebuilding business. Nothing plaintiffs have offered the court, however, supports an understanding by the United States that the plaintiffs needed the net tax savings for purpose of investing in homebuilding projects. The type of negotiating evidence offered by plaintiffs supports only the inference that the government knew about and was interested in the profits from plaintiffs’ homebuilding business only insofar as they would support the anticipated tax benefit sharing.

We hold, as a matter of law, that the subject of the bargain was confined to the opportunity to take advantage of the tax laws; that it would not have been forseeable to federal negotiators that plaintiffs were entering into the arrangement to secure future funds for homebuilding projects. Absent some specific reference, those projects, and the potential profits on them, were simply too remote to be within the parties’ contemplation as potential collateral damage from a breach. See Wells Fargo Bank v. United States, 88 F.3d 1012, 1023 (Fed.Cir.1996), cert. denied, 520 U.S. 1116, 117 S.Ct. 1245, 137 L.Ed.2d 328 (1997); Glendale Federal Bank, FSB v. United States, 43 Fed.Cl. 390, 397-98 (1999).

V. Borrowing Costs

Plaintiffs argue that they should be compensated for the additional borrowing expenses resulting from the breach. Plaintiffs contend that they would have used some portion of the tax benefits denied by Guarini to pay down their revolving lines of credit in 1997 and to invest in homebuilding projects in 1999, instead of using credit.¹⁸

Plaintiffs, however, are not entitled to these damages as a matter of law. Although we agree that borrowing costs are theoretically recoverable and are not necessarily forbidden pre-judgment interest, see Centex Corp., 55 Fed.Cl. at 390, (discussing borrowing costs under comparable circumstances), plaintiffs must prove both foreseeability and causation with reasonable certainty. See Bluebonnet Sav. Bank, FSB v. United States, 266 F.3d 1348, 1355-58 (Fed.Cir.2001). We find plaintiffs’ evidence too insubstantial to support an award.

Plaintiffs present no evidence that borrowing costs were foreseeable at the time of the acquisition. Even if we accept all of plaintiffs’ factual allegations as true, there is no indication that the United States could have anticipated that plaintiffs would have incurred additional borrowing costs as a result of the breach. In a corporate structure as large and diverse as the Pulte consolidated group, the organization could have applied the tax benefits in any number of ways.¹⁹ It does not necessarily follow that increased tax benefits would allow plaintiffs to either pay down old credit or avoid using new credit. Plaintiffs’ borrowing costs were too remote from the breach. There is nothing in the contract terms or in the negotiations from which the government negotiators could have inferred the assumption of risk of increased borrowing costs in the event of breach.

VI. Gross-up

Plaintiffs urge that their damages need to be “grossed-up” to account for income taxes that they might have to pay as a result of this judgment. In the alternative, they urge that a separate award be made to the IRS on their behalf. Though a gross-up is sometimes appropriate, we disagree that it is proper in this case. See Centex Corp., 55 Fed.Cl. at 888-89 (denying a comparable gross-up claim). Here, it would be unfair to treat plaintiffs’ judgment as new income and the court’s award should be tax-free.²⁰ See id. No gross-up or separate award of taxes is necessary.

CONCLUSION

The parties’ cross-motions for summary judgment are granted in part and denied in part as set out above. In sum, plaintiffs are entitled to recovery of the net amount of lost federal income tax benefit. That amount cannot be finalized because the amount sought for lost tax benefits ($48,683,620) includes the as-yet-unresolved claim for state tax benefits. Plaintiffs’ other claims for relief are denied. Plaintiffs’ October 2, 2002 motion to strike, plaintiffs’ February 10, 2003 motion for an order requiring defendant to state its position, defendant’s April 1, 2003 motion to strike, and defendant’s April 9, 2003 motion to strike portions of plaintiffs’ oral argument are denied as moot. The parties are directed to consult with each other and propose in status reports filed no later than June 23, 2003 further proceedings for resolving the sole remaining issue of the claim for state tax benefits.

1 . United States v. Winstar Corp., 518 U.S. 839, 116 S.Ct. 2432, 135 L.Ed.2d 964 (1996).

2 . This legislation, the Omnibus Budget Reconciliation Act of 1993, Pub.L. 103-66, § 13224 (1993), is named for its primary sponsor, Representative Frank Guarini. The law is hereinafter referred to as “Guarini.”

3 . We subsequently allowed the substitution of Pulte Homes, Inc. for Pulte Corporation.

4 . Throughout this opinion, we use "FDIC" to represent both the Federal Deposit Insurance Corporation and its predecessor organization, the Federal Savings and Loan Insurance Corporation.

5 . During preparation of Ms. McCall’s expert report, she found documentation errors in plaintiffs’ tax workpapers for their 1988-1990 returns, as filed and amended. She reviewed the work-papers and found some selected assets within the documentation for which the workpapers recorded a tax charge-off in an amount less than what should have been reported. In addition, she found some assets where plaintiffs had reported amounts greater than what could or should have been reported. Ms. McCall then "trued-up" plaintiffs’ tax workpapers to accurately correlate to plaintiffs’ actual tax return position.

6 . Ms. McCall double-checked this number by preparing a pro forma calculation of plaintiffs’ bad debt reserve. Her exercise resulted in the same value for CALs disallowed by Guarini, $170,050,661.

7 . The government disputes that this number is the correct amount, but not that it was the number actually recaptured by the plaintiffs.

8 . Beginning in 1990, 26 U.S.C. § 56(g) (2002) required certain taxpayers to compute an ACE adjustment as a component of the taxpayer's AMT calculation.

9 . Mr. Wolf purports to be "fixing IRS audits."

10 . In any event, we find they are too speculative to put at issue Ms. McCall’s conclusion based on her own careful review.

11 . Plaintiffs correctly note that Mr. Wolf arbitrarily adds the ... reserve balance into taxable income in 1991. There is no logical connection between the 1996 recapture requirement and plaintiffs’ 1991 taxable income.

12 . Mr. Wolf claims that plaintiffs may still choose to amend their 1990 tax return to take this additional deduction.

13 . Mr. Wolf also initially complained of errors in plaintiffs’ ... and ... returns, but these criticisms were withdrawn at oral argument.

14 . Mr. Gregory Nelson, plaintiffs’ witness, explained that plaintiffs’ forecasts and projections are extremely conservative and thus reasonably attainable.

15 . This stock repurchase may also have directly increased plaintiffs’ debt/cap ratio, in direct opposition to plaintiffs’ stated unwillingness to exceed their self-imposed 40% ratio.

16 . With the exception of two quarters shortly after acquiring Del Webb.

17 . In other words, if plaintiffs’ average ROIC for a certain year was ..., then the higher-priority projects for that year were necessarily above ... and the lower-priority projects were below---Almost by definition, if plaintiffs could have invested in additional projects that year, they would have to reach down into the lower-priorily projects — those with a lower ROIC. They likely had already selected the higher-priority, higherROIC projects.

18 . Defendant disputes that plaintiffs would have used the proceeds to invest in homebuilding projects.

19 . We note here that plaintiffs’ claim for borrowing costs seems to conflict with their claim for lost profits. In their borrowing costs claim, plaintiffs allege that they would have used tax benefits to pay down its credit lines and to forego credit to invest in home building projects. In plaintiffs’ lost profits claim, they allege that they would have used tax benefits to invest in home building projects that they could not otherwise invest in, because they were prevented by their self-imposed 40% debt/cap ratio from drawing upon credit.

20 . We note, however, that plaintiffs could use RCFC 60(b) in the event this assumption is incorrect.