A New York court ruled yesterday that a bond insurer claiming Bank of America’s Countrywide unit fraudulently induced it to insure $21 billion of mortgage-backed securities can use statistical sampling to prove its case.

Judge Eileen Bransten granted a motion to allow the insurer, MBIA, to use statistical sampling, turning down the objections by Bank of America.

Bank of America

BAC loading...

Quote  |  Chart  |  News  |  Profile

[BAC  Loading...      ()   ] had told investors that it intended to fight repurchase requests—also known as put-backs—on a “loan-by-loan basis.” That process would have required MBIA and others seeking to force Bank of America to buy back loans it pooled into mortgage securities to proceed one loan at a time, a costly and time-consuming process.

MBIA

MBI loading...

Quote  |  Chart  |  News  |  Profile

[MBI  Loading...      ()   ] has proposed that for each of the 15 securitizations—each of which contains thousands of home loans—it is challenging, it will sample 400 loans. Those loans will be subdivided according to borrowers credit score and loan to value ratios. The samples will then be tested in an effort to prove MBIA’s claims of fraud, breach of insurance contracts, and repurchase demands.

The ruling does not mean that MBIA will automatically prevail if the sampling shows that the loan pools were flawed. Bank of America may still argue that the sampling methodology was too flawed to amount to proof of MBIA’s claims. Alternatively, it could present statistical sampling of its own to counter the findings of MBIA’s sampling. But whether these arguments will prevail will be made by the finder of fact at trial—which means either the jury or—if Bank of America opts not to have a jury trial—the judge.

MBIA has been handed a powerful tool for establishing its claims. And Bank of America is a bit closer to facing the put-back apocalypse it hoped it could avoid—or at least slow down—by fighting the claims one loan at a time.

Want more on mortgage put-backs?