The bond insurer Assured Guaranty couldn't have asked for a more engaged judge than U.S. Senior District Judge Jed Rakoff of Manhattan, who threw some tough questions at a Flagstar underwriting manager on Monday, as Assured's $110 million bench trial against Flagstar nears an expected end this week. Rakoff was fixated on one of the loans underlying the Flagstar mortgage-backed securities that Assured insured, in which the borrower -- a Detroit police officer -- claimed to be the president of a mortgage brokerage. The judge was so disturbed by the loan-level details that emerged Monday that he twice mused whether to refer the matter to prosecutors in Michigan. It made for great theater, as you'll see below. But don't let the fraud kerfuffle fool you: Rakoff is poised to issue substantive rulings that will have a broad impact on bond insurers' attempts to recoup money from MBS issuers.

First, though, the kerfuffle. The excitement began when Flagstar's lawyers from Arnold & Porter launched their arguments for excluding testimony from Assured's underwriting and statistics experts. Those experts presented the bulk of the bond insurer's case accusing Flagstar of breaching representations and warranties on the loans underlying the securities Assured guaranteed, so the back-and-forth with the judge was quite intense. According to a transcript of Monday's proceeding, when Assured's lead counsel, Jacob Buchdahl of Susman Godfrey, stood up to defend the methodology of his underwriting expert, Rakoff brought up the suspect loan file, which he had selected at random for review last week when Assured's expert was testifying. At the time, the judge noticed that the loan applicant claimed to be both a Detroit cop and the president of a financial services company, which smacked of fraud to Rakoff.

Buchdahl told him Monday morning that there was even more to the story: The cop claimed to be the president of the very same company that issued the mortgage. (Flagstar subsequently acquired the loan for securitization.) Buchdahl's implication, which Rakoff immediately detected, was that the cop and the loan officer who authorized his mortgage had conspired to misrepresent the police officer's qualifications.

"Wow," Rakoff said. "The statute of limitations hasn't run. I wonder whether there is a good fraud case here against both the interviewer and the interviewee."

If Flagstar thought Rakoff might forget about the suspicious loan over Monday's lunch break, it thought wrong. The bank put on two witnesses in the afternoon session, as it began its defense. Rakoff asked both of them about the police officer's loan file. The first Flagstar witness was too senior to have answers to the judge's questions, but the second was a Flagstar underwriting manager named Marni Scott. Rakoff showed her the troubling loan file and asked whether Flagstar would have reviewed it, even though the loan was issued by a different mortgage broker. Scott said yes. The judge then asked whether it should have been obvious that the borrower wasn't actually the president of the mortgage brokerage that made the loan but was instead a police officer. "It appears to me that even the most modest checking of the sort you say should have been done would have revealed this to be a patently fraudulent application," Rakoff said.

Scott said she would have to look at the circumstances, that perhaps the borrower was both a police officer and mortgage brokerage president. Rakoff didn't buy it.

"You think that a Detroit police officer would also be the president of a substantial financial (firm), Regional Financial Group, (that) you were doing business with, that had offices in numerous states?" the judge said. "That that was sort of his sideline, he came home at 7:00 and as soon as he had dinner, he turned his attention to being the president of a major financial group? Is that what you think?" Scott eventually conceded that it probably wasn't plausible. (Flagstar's lawyers said they would return to the particulars of this loan later in the trial; a Flagstar representative declined to comment.)

It has to be considered a good omen for Assured that Rakoff found what he clearly considers a flagrant deficiency in a loan file he picked randomly from the 20 or so that the bond insurer presented at trial. But Rakoff well knows that this is one loan file from about 15,000 underlying the Flagstar MBS at issue in the case. In fact, the idiosyncrasies of the particular file Rakoff picked up could be seen as support for Flagstar's argument against the use of sampling. And if Rakoff decides that Flagstar is right and Assured can't extrapolate evidence of material breaches from the sample of loan files it reviewed, every monoline MBS case could be affected.

Here's why. Re-underwriting mortgage loan files is a tremendously expensive and time-consuming endeavor, since mortgage-backed securitizations typically bundle thousands of loans. Early on in the bond insurer put-back litigation, judges began granting monoline sampling motions, agreeing to decide whether to dismiss their cases based on material breaches the monolines found when they reviewed a subset of underlying loans, not the entire book of thousands of mortgages. For the purposes of preliminary motions, just about all of the judges presiding over these cases have permitted bond insurers to cite breach rates they extrapolated from samples of loans.

Assured's case is the first, however, to reach trial, so Rakoff will be the first judge to determine whether extrapolation holds up well enough to be the basis of a final judgment. Flagstar argued Monday that Assured's own underwriting expert, Rebecca Walzak, emphasized the subjectivity of the underwriting process, which undermines the rationale for sampling. "We are being asked to extrapolate liability, your Honor, from this sample all the way out into the world of thousands of loan files," said Flagstar counsel Veronica Rendon of A&P. "We are supposed to rely upon this approach as though it is valid across 16,000 loan files and be thinking about paying over $100 million based upon this testimony, when it boils down to nothing more than an unarticulated principle of I-know-it-when-I-see-it."

Assured responded that Flagstar was exaggerating the subjective element of its expert's methodology. "Look, our whole entire approach here is based on the fact that this is a sample and from this sample and the huge problems we see here we can conclude that throughout the entire (book of) loans there are abundant problems," Buchdahl said. "While each loan tells its own story, I think there is a very clear story that emerges of lack of attention to the most basic principles of underwriting.... They would blink at the same issues: overstated income, undisclosed debts, multiple credit inquiries that did not have the required letter of explanation, and other kinds of problems that Ms. Walzak identified."

Rakoff said Monday that would reserve a ruling on "what is a very interesting but complicated set of issues" until the end of the trial.

(Reporting by Alison Frankel)

Follow us on Twitter @AlisonFrankel, @ReutersLegal | Like us on Facebook