Speed read

The U.S. Bankruptcy Court for the Southern District of New York has issued a decision interpreting the definition of "Loss" under the 1992 ISDA Master Agreement.


The U.S. Bankruptcy Court for the Southern District of New York (the Court) recently issued a decision interpreting the definition of "Loss" under the 1992 ISDA Master Agreement, holding that where the parties have selected "Second Method" and "Loss," the Non-defaulting Party has the power to calculate its Loss by any method, provided that such method is reasonable and selected in good faith.

At issue in the case was a forward share repurchase transaction entered into by Intel Corporation (Intel) and Lehman Brothers OTC Derivatives Inc. (LOTC) under a 1992 ISDA Master Agreement (the ISDA)1 and a trade confirmation (the Confirmation) in which the parties selected "Second Method" and "Loss" as the methodology for calculating an Early Termination Payment in connection with the transaction.

"Loss" is defined in the 1992 ISDA Master Agreement as an amount that a party reasonably determines in good faith to be its total losses and costs (or gain, if applicable) in connection with the ISDA and a terminated transaction or transactions. The alternative method for calculating an Early Termination Payment under the 1992 ISDA Master Agreement is "Market Quotation," which is defined as an amount determined on the basis of quotations from leading dealers in the market to enter into a replacement transaction that would have the effect of preserving for the Non-defaulting Party the economic equivalent of any payment or delivery that would have been required under the transaction being terminated.

The Court's decision interprets the definition of "Loss" in accordance with its plain meaning and finds that a Non-defaulting Party has discretion and flexibility in selecting a methodology for calculating its Loss, so long as the selection of the methodology and the calculation of Loss are ultimately performed "reasonably and in good faith."


In August 2008, Intel entered into a forward share repurchase agreement with LOTC to allow Intel to repurchase a number of its own shares during its "quiet period," which generally begins several weeks before the end of each fiscal quarter and runs until Intel's earnings announcement. During a quiet period, Intel's managers often possess material non-public information and as a result, Intel's in-house traders do not repurchase Intel shares during this time. Under a forward share repurchase agreement entered into prior to the quiet period, Intel could arrange for its counterparty to purchase Intel shares and deliver them to Intel during the quiet period.

Pursuant to the Confirmation, Intel paid $1 billion to LOTC as a prepayment for a number of shares of Intel common stock (determined according to an equation set out in the Confirmation) to be delivered by LOTC to Intel on a future date. LOTC delivered $1 billion of collateral to Intel to secure LOTC's future obligation to deliver shares under the Confirmation. By the end of the transaction, LOTC was obligated to deliver approximately 50.6 million shares of Intel common stock. Before this date occurred, however, Lehman Brothers Holdings, Inc. (together with LOTC, Lehman), LOTC's guarantor under the ISDA, declared bankruptcy. This constituted an Event of Default under the ISDA. After LOTC failed to deliver any shares and based on the continuing Event of Default, Intel terminated the transaction and designated an Early Termination Date under the ISDA. At the time that Intel designated an Early Termination Date, there were no remaining payments or deliveries outstanding under the Confirmation other than the past-due delivery of Intel shares by LOTC.

To determine the Early Termination Payment, Intel, as the Non-defaulting Party, calculated its Loss with respect to the transaction2. Intel calculated its Loss, and the Early Termination Payment, to be $1 billion plus interest. Under the ISDA, Intel was entitled to set off an Early Termination Payment against the collateral posted to Intel by LOTC. Accordingly, Intel set off the Early Termination Payment owed by LOTC against the entirety of the $1 billion of collateral (plus accrued interest) posted by LOTC under the ISDA. Lehman challenged Intel's Loss calculation and set-off, arguing that the only reasonable method to determine Intel's Loss was to use the fair market value of the undelivered shares as of the scheduled settlement date, which would have been only $873 million.

The threshold question considered by the Court, therefore, was whether the definition of Loss prescribes a specific methodology for calculating Intel's Loss, as argued by Lehman, or whether the definition of Loss permitted Intel to select any methodology for calculating its Loss on a transaction, so long as such methodology produces a calculation performed reasonably and in good faith.


The Court rejected Lehman's argument that the definition of Loss required Intel to calculate Loss based on the fair market value of the undelivered shares, and applied the general rule for calculating Loss described above.

The court also rejected Lehman's argument that the "cross-check principle," developed in the English law case of Anthracite Rated Investments (Jersey) Limited v. Lehman Brothers Finance S.A., [2011] EWHC 1822 (CH), should apply in this instance. Lehman argued that while Loss and Market Quotation are different concepts, they are generally intended to achieve the same outcome, and consequently, the result from one can be useful as a "cross-check" to test the accuracy of the other. Lehman argued that applying Market Quotation as a cross-check on Intel's calculation would result in a calculation that only included Unpaid Amounts (namely, the fair market value of the undelivered shares), lending support to its calculation of Loss, which it said should also be limited to the fair market value of the undelivered shares. The Court, however, held that this principle was only applicable to situations where there are deliveries or payments to be made after the Early Termination Date. The principle was, therefore, inapplicable to this case because the Confirmation did not call for any future deliveries or payments to be made after Early Termination Date.

As Intel's good faith was not challenged, the final determination for the Court was whether Intel's method of calculating its Loss was reasonable. Because this was a motion for summary judgment, part of that determination was whether there were disputed issues of material fact regarding the reasonableness of Intel's calculation. Intel offered six methods to calculate Loss in an amount equal to $1 billion plus interest. The Court found four of these methods persuasive.

First, the Court found that Intel's use of its upfront costs plus interest to calculate Loss was consistent with the plain meaning "total losses and costs," a phrase included in the definition of Loss. Lehman argued that this phrase should be inapplicable in this case on the basis of its argument that the only proper calculation of Loss was the fair market value of the Intel shares that LOTC failed to deliver, but the Court rejected this argument and Lehman did not offer a different view of Intel's losses and costs. Therefore, because Intel's view was undisputed and consistent with the plain meaning of "total losses and costs," the Court concluded that this approach to calculating Loss was reasonable.

Second, the Confirmation included a provision that reduced the cash amount of the Early Termination Payment by the total "Agreed Value" – which was not defined simply as the fair market value – of shares already delivered. The total Agreed Value of the undelivered shares, as calculated according to a formula contained in the Confirmation, was $1 billion. Although the provision containing the definition of Agreed Value dealt with the manner in which the Early Termination Payment could be satisfied rather than the method by which Loss should be calculated, the Court found that since Intel's calculation of Loss, excluding interest, was equal to the Agreed Value of the undelivered shares, Intel's Loss calculation was consistent with the Confirmation and the manner in which the parties conceptualized the value of the shares to be delivered.

Third, under the Confirmation, Intel's exposure was expressly set at $1 billion dollars, but could be reduced by the Agreed Value of any shares delivered early. Intel argued that both its exposure and LOTC's outstanding performance were valued at $1 billion on the Early Termination Date because no shares had been delivered. Although exposure is not synonymous with Loss, the Court found that because the parties' agreement gave Intel the right to set off $1 billion, reduced by the Agreed Value of any shares delivered early, Intel's exposure could be reasonably used as a proxy for Loss. The fact that the exposure remained at $1 billion supported the reasonableness of Intel's calculation of Loss.

Fourth, since Intel advanced $1 billion, but LOTC failed to deliver any shares, the Court found that Intel had a restitutionary interest in the funds it advanced. Since the restitutionary interest was in funds worth $1 billion, that interest also supported the reasonableness of Intel's Loss calculation.

The remaining two methods of calculating Loss offered by Intel were considered less persuasive by the Court.

Intel argued that its reason for entering into the transaction – and, therefore, the value of the shares to Intel – was the ability to reduce shareholder equity by $1 billion. Intel likened the repurchase program to a dividend, and its goal was to return cash to remaining shareholders. The Court noted, however, that the actual value of the transaction was not necessarily the $1 billion repurchase, but the aggregate increase in share prices of the remaining shares as a result of the repurchase program. Since Intel gave no evidence on the projected or actual share price increase, the Court did not view Intel's claims regarding a $1 billion reduction in shareholder equity as undisputed evidence that could be relied upon in the Court's determination.

Finally, Intel pointed to an internal Lehman e-mail, which acknowledged that Lehman would not be able get the $1 billion in collateral back from Intel, to show that Lehman agreed with Intel's calculation of Loss. The content of the e-mail, however, was contradicted by several Lehman employees in deposition testimony. Thus, Lehman's internal view was not undisputed, and even if it were, the Court stated that it would not be relevant to the reasonableness of Intel's calculation of Loss.


Based on the above, the Court concluded that Intel's Loss calculation under the 1992 ISDA Master Agreement was reasonable.

Many market participants have moved toward the 2002 ISDA Master Agreement, which replaces "Loss" with the term "Close-out Amount," or have amended their existing ISDA Master Agreements to apply the "Close-out Amount" concept instead of "Loss" (for example, by way of ISDA's Close-out Amount Protocol). It would be interesting to see how the arguments that were successful in this case might be evaluated by a court interpreting the definition of "Close-out Amount" under the 2002 ISDA Master Agreement. The definition of "Close-out Amount" allows the Non-defaulting Party to determine its losses and costs "in good faith and us[ing] commercially reasonable procedures in order to produce a commercially reasonable result." Given the similarity of the language in both definitions that describes the standard with which the Non-defaulting Party must comply when determining its "Loss" or the "Close-out Amount," this decision may provide useful precedent to future disputes over the calculation of a "Close-out Amount" under the 2002 ISDA Master Agreement.