Judge Rakoff issued an opinion with significant implications for parties litigating cases involving mortgage backed securities. In Assured Guaranty v. Flagstar Bank, Assured sought to recover some $89 million in insurance claims it paid to bondholders who invested in Flagstar bonds secured by what turned out to be lousy mortgages. When the mortgage market collapsed, the bonds secured by the mortgages also failed, leaving Assured on the hook. Assured claimed that Flagstar breached its contract with Assured by overstating the quality of the mortgages when it asked Assurance to insure the bonds — about 15,000 separate mortgage loans. After a bench trial, Judge Rakoff agreed with Assured and found that Flagstar was liable for the $89 million in insurance claims paid by Assured to bondholders. Judge Rakoff’s opinion is the first ever after trial holding a bank liable for breach of its representations and warranties to its monoline insurer about the quality of the morgages it originated.

Also of interest, both Assured and Flagstar relied on expert testimony to explain why – or why not – the loans suffered from material defects. But how do you actually try a case that involves the minute details of 15,000 different mortgage loans?  Notably, the court focused on a statistical sample of 800 loans rather than the total universe of 15,000 loans used for collateral. Assured’s expert found a compendium of unreliable loan documentation in the sampled loan files, representing an overview of much that went wrong in the housing market: loans made to borrowers who failed to disclose significant debt obligations, loans made to borrowers who inflated their incomes, loans made to borrowers who misrepresented their occupancy, and loans made with incomplete documentation. Assured found that a whopping 76% of the sampled loan applications were bogus.

In determining that Flagstar breached its contract with Assured, Judge Rakoff extrapolated these findings to all 15,00 of the insured loans. The use of a statistical sample to extrapolate the volume of defective loans in the total loan population is significant. While this is not the first time that the method has been used – in fact, Judge Rakoff cites to two other cases in New York courts where it has been done before – it is reportedly the first instance to do so in a housing crisis trial. According to Judge Rakoff:

“Although Flagstar argues that the fact determination of material breach in any given instance requires consideration of an entire loan file renders the loans ill-suited to proof by statistical sampling, this argument is unpersuasive. The very purpose of creating a representative sample of sufficient size is so that, despite the unique characteristics of the individual members populating the underlying pool, the sample is nonetheless reflective of the proportion of the individual members in the entire pool exhibiting any given characteristic.”

Finally, the central question in the case – whether underwriting standards were materially defective – of course is also at the heart of cases involving MBS proceeding in courthouses across the country. Here in the Southern District, for example, cases before Judge Cote alone brought by the Federal Housing Financing Agency on behalf of Fannie Mae and Freddie Mac seek up to $200 billion in combined damages from a who’s who of banks, including Bank of America, Citigroup, Inc., and Morgan Stanley. Undoubtedly, the Assured opinion – and appeal – will be followed closely.

(Thank you to my colleague Neal Kronley who helped prepare this entry.)