The 2008 financial crisis included a wave of defaults on collateralized debt obligations, and some of those defaults caused enormous losses to monoline bond insurers. One result has been a second wave: over 30 lawsuits, the most prominent of which is MBIA’s pending case against Countrywide Home Loans in a New York state court, where the insurer is claiming some $3 billion in damages. MBIA Ins. Corp. v. Countrywide Home Loans, No. 602825/2008 (N.Y. Sup. Ct.). Earlier this month, after the non-jury trial of another closely-watched case, an insurer was awarded $90 million by Judge Jed Rakoff of the U.S. District Court for the Southern District of New York, in Assured Guaranty Municipal Corp. v. Flagstar Bank, FSB, No. 11 Civ. 2375 JSR (S.D.N.Y. Feb. 5, 2013). The decision was a clear win for the insurer on issues that resonate in other cases, but the court’s opinion still leaves room for argument about some fundamental questions.
The RMBS and the Bond Insurance
Like the contracts and instruments at issue, the facts of these cases are complicated.
Flagstar Bank and its subsidiaries originate residential mortgages and home equity loans and securitize them as residential mortgage-backed securities. The Assured Guaranty case involved two securitizations that were collateralized by approximately 15,000 home equity lines of credit. Flagstar originated the loans, then sold them (through a subsidiary) to two securitization Trusts. Investors in the Trusts, known as bondholders, entered into agreements under which the Trusts provided them with monthly interest payments and an eventual return of principal. These payments are secured and funded by payments on the underlying loans that flow into the Trust.
The Trusts were over-collateralized–that is, they contained more loans than were necessary to meet their obligations to the bondholders. Nevertheless, the bondholders were exposed to the risk of widespread defaults on the underlying loans, and that risk held the credit rating of the Trusts to BBB. To raise that rating, and to allow securitizations to be sold at lower interest rates, banks like Flagstar purchase bond insurance, and Flagstar obtained insurance for the two Trusts at issue from Assured Guaranty Municipal Corp. (AGM). Under an Insurance and Indemnity (I & I) Agreement, AGM guaranteed the Trusts’ timely payment of interest and principal to the bondholders. That guarantee had the effect (among others) of raising the Trusts’ credit rating to AAA.
The I & I Agreement incorporated a number of representations and warranties by Flagstar about the underlying loans–provisions that are set forth in full in related contracts. These included:
Each Mortgage Loan was originated in good faith and in accordance with [Flagstar’s] underwriting guidelines. . . .
No error, omission, misrepresentation, negligence, fraud or similar occurrence with respect to a Mortgage Loan has taken place on the part of any person . . . .
Flagstar was responsible for any material breach of these representations, regardless of whether it knew about the breach at the time the representations were made. Its responsibility was to cure such a breach “within 90 days of becoming aware of it.” Should it fail to do so:
The sole remedy of [AGM] . . . is [Flagstar’s] obligation . . . to accept a transfer of a Mortgage Loan as to which a breach has occurred and is continuing . . . or to substitute an Eligible Substitute Mortgage Loan for [the defective one].
The Trusts Go South
Beginning in 2008, as the real estate market collapsed, a large number of the loans underlying the Trusts went into default, and the Trusts were unable to meet their obligations to the bondholders. To date, AGM has had to pay more than $90 million to cover the shortfall.
As these losses began to mount, AGM conducted an internal review of a sample of defaulted loans, and it concluded that “the vast majority of the loans in the sample”–more than 900 loans in all–“violated Flagstar’s underwriting guidelines.” It also found that “many” of the files contained evidence of fraud. In two letters it sent to Flagstar in January 2009, AGM reported these conclusions and demanded that the bank “comply with the repurchase, reimbursement, and indemnification provisions” in the insurance contract. Flagstar essentially rejected that demand.
The Insurer’s Claims
AGM filed its action in 2011, asserting that Flagstar and its subsidiaries had “failed to perform” under the I & I Agreement (as well as related contracts), by breaching the representations and warranties concerning the underlying loans, by failing either to cure the defects in the charged off loans or to substitute or repurchase them, and by failing to reimburse or indemnify AGM.
In a ruling on Flagstar’s motion to dismiss, the court held that AGM’s remedies were “restrict[ed]” to (1) enforcement of Flagstar’s obligation to cure or repurchase defective loans, and/or (2) damages resulting from breach of that obligation. Assured Guaranty Mun. Corp. v. Flagstar Bank, FSB, No. 11 Civ. 2375(JSR) (S.D.N.Y. Oct. 31, 2011).
Consequently, when the case came to trial, AGM’s right to recover was not based directly on Flagstar’s acts or omissions with respect to the underlying loans. Unlike (for example) some of the fraud claims being pursued by MBIA, AGM’s case did not make Flagstar’s liability depend on the bank’s having employed shoddy underwriting practices, or having made false representations about those practices in the relevant agreements–even though allegations to that effect are arguably implicit in AGM’s case. Rather, the gravamen of AGM’s lawsuit (as modified by the court’s decision) was that Flagstar had failed to fix the problems (specifically, by repurchasing loans) that its underwriting practices, or fraud on the part of borrowers, might have caused, once those problems had been called to its attention by AGM’s letters of January 2009.
The Evidence at Trial
The heart of AGM’s case at trial consisted of expert testimony. One expert organized a selection of 800 sample loans and testified about the statistical validity of extrapolations that could be made from that sampling. A second expert oversaw a review of documents relating to the sampled loans and concluded that a large proportion of them had either been issued in violation of Flagstar’s underwriting guidelines or been obtained by misrepresentations about the borrower’s finances and/or the underlying property.
The final expert testified that he had constructed “an exact replica” of the Trusts–one that enabled him to simulate how the Trusts would have performed, ifFlagstar had repurchased “the proportion of loans estimated [by the work of the other two experts] to be defective in the two Trusts.” On the basis of this simulation, the expert concluded that “if Flagstar had begun repurchasing defective, charged-off loans in January 2009 . . ., there would have been sufficient cash coming into the Trusts to meet all the Trusts’ payment obligations to the bondholders . . . .”
Flagstar challenged all of this testimony with the opinions of its own experts, and the court acknowledged that they made some valid points. In the end, though, the court largely accepted the opinions of the insurer’s experts. Judge Rakoff found that a large proportion of the loans underlying the Trust violated the representations and warranties in the parties’ agreements, and that Flagstar could have prevented the insurer’s losses by repurchasing an equal proportion of loans.
The Missing Link
Flagstar did not, however, have an absolute obligation to repurchase every bad loan. Rather, as noted above, Flagstar’s obligation to cure or (if it failed to cure) to repurchase loans was activated only when it “became aware” of violations of the representations and warranties relating to those loans. Beyond accepting the conclusions of AGM’s experts, therefore, Judge Rakoff also had to find that Flagstar had “become aware” that certain loans were subject to repurchase before he could rule for the insurer.
The court based that finding on the two letters AGM sent to Flagstar in 2009. These letters, the court ruled, “rendered Flagstar constructively ‘aware’–or, at a minimum, put Flagstar on inquiry notice–of the substantial likelihood” that there were “pervasive breaches of the representations and warranties.” Judge Rakoff found, that is, that AGM could make Flagstar “become aware” of violations without identifying all (or even many) of the loans that were subject to repurchase; he reasoned that a contrary ruling would “inappropriately place the burden of notification on” AGM. In making this point, the court suggested, but did not explicitly state, that “awareness” (the term used in the parties’ contract) requires something less than actual notice.
The general question of whether the contractual requirement of “awareness” can be satisfied by “constructive” knowledge or “inquiry notice” was controversial—and it is currently being hotly contested in the Countrywide suit. But Judge Rakoff’s ruling also raised a second question: Whether the quality of the information contained in AGM’s demand letters was sufficient to establish such knowledge or notice.
AGM’s letters were based on an internal review of a significant sampling of loans, but these were a different review and a different sampling from the ones the court heard about at trial, and which it ultimately accepted. In its thoughtful, 100-page opinion, the court did not address the question of whether the review cited in the letters was detailed or credible enough to impose a duty of inquiry on a reasonable bank. Instead, in the place that discussion might have appeared, the court took up two different matters.
First, it acknowledged that Flagstar claimed to have “cured or rebutted” all of the assertions contained in the January 2009 letters, but it responded in an evasive manner: It said, “Flagstar’s response denying the need to repurchase . . . cannot be sufficient to absolve Flagstar of awareness that [AGM] was claiming pervasive breaches . . . .” In fact, Flagstar was required to act only if it became aware of the breaches themselves, not if it became aware AGM “was claiming” they existed. Yet the court did not address the question of whether Flagstar’s claim to have “rebutted” AGM’s claims was or was not correct.
Then, in a footnote, the court cited the existence of evidence that Flagstar “may have been aware of the problems in its underwriting at the time . . . the parties entered into” the I & I Agreement. But the court’s ruling was not based on that possible awareness; it was based on AGM’s letters. And the opinion simply does not explain what standard the court used to determine that those letters imparted “constructive awareness” to Flagstar.
It is possible that AGM’s letters were so clear, well-documented and well-reasoned that any finder of fact would have reached the same conclusion Judge Rakoff did about inquiry notice. In fact, that might be the reason the court’s opinion did not spend time on this issue. For purposes of future cases, therefore, a more important aspect of Assured Guaranty is its adoption of the principle that a bond insurer cannot be penalized for having failed to discover and document completely the kinds of pervasive misconduct that some mortgage lenders are now known to have sanctioned.
Still, courts that accept this principle might have to apply it to concrete facts that allow for different interpretations. For those courts, the rules for deciding what it takes to put a bank on “inquiry notice” have not yet been written.