For a company that just agreed to pay $23 million to resolve Securities and Exchange Commission allegations that it defrauded investors by overpaying for the mortgage-backed securities underlying the collateralized debt obligations it managed, ICP Asset Management and its founder, Thomas Priore, sure came off well in Gretchen Morgenson's column in The New York Times on Sunday. In a piece called "How to Find Weeds in a Mortgage Pool," Morgenson detailed ICP's efforts, on behalf of its Triaxx CDOs, to identify dud loans originated by Residen t ial Capital -- the mortgage-lending arm of Ally Financial -- and securitized in mortgage-backed trusts. Triaxx invested billions in ResCap mortgage-backed certificates, Morgenson reported. Using its propriety database of loan-level information, Triaxx pinpointed thousands of mortgages that appeared to breach ResCap's representations and warranties. Morgenson said Triaxx's technology "could finally help us get to the bottom of troubled mortgage investments."

With all due respect to Morgenson, I doubt Triaxx's database is a game changer, at least not in the current state of play of litigation over alleged breaches of MBS representations and warranties. No matter how sophisticated Triaxx's tech is, there are already plenty of data miners busily generating the same kind of information about deficiencies in individual mortgages underlying mortgage-backed certificates, from the Federal Housing Finance Agency to private data experts like CoreLogic that have provided analytics for certificate holders and bond insurers with reps and warranties claims. We've seen plaintiffs assert breach rates of 60, 75 and 80 percent based on those analyses, so it's hard to imagine that Triaxx's database could generate a higher percentage of deficient loans. (Or that any plaintiff demanding that an originator buy back one of those loans would have any better success than the investors already pursuing such claims.) I don't see how incrementally better data can make a big difference in investors' recoveries.

But that's not to say that Triaxx won't have an impact on put-back litigation. As Morgenson's story noted, ICP founder Priore described the database in an affidavit supporting Triaxx's objection to a proposed settlement of reps and warranties claims against Residential Capital. (I've previously written about the proposed deal, in which ResCap agreed to allow a claim of $8.7 billion in its Chapter 11 by MBS certificate holders. The settlement has garnered the support of 19 major institutional investors represented by Gibbs & Bruns and Ropes & Gray, as well as clients of noted MBS investor lawyer Talcott Franklin.)

Triaxx, which has also objected to Bank of America'sproposed $8.5 billion settlement with investors in Countrywide mortgage-backed notes, first laid out its objections to the ResCap proposed settlement in June, then elaborated on its key point in a brief in late August. The crux of its argument is that the proposed formula for allocating whatever MBS investors ultimately collect in the ResCap bankruptcy unfairly favors investors in second-lien and subprime mortgage-backed notes, as opposed to notes backed by prime mortgages. The Triaxx CDOs, which invested primarily in prime mortgage-backed notes, asserted that because the allocation formula among the 392 ResCap MBS trusts is based on the trusts' losses -- and not trust-by-trust breach rates -- prudent investors who have suffered fewer losses will recover less than they should for reps and warranties breaches. Investors in the riskier subprime-backed trusts, according to Triaxx, may have fewer actual claims for breaches of reps and warranties but will reap a disproportionate share of the ResCap MBS recovery because of their higher losses.

"There is no rational basis for the proposed settlement's assumption that the greater losses in the subprime ... trusts mean that there were more representation and warranty breaches in those trusts," wrote Triaxx's lawyers at Miller & Wrubel. "The allocation formula's assumption that the losses in the riskier trusts are equally likely to have been caused by representation and warranty breaches as the losses in the safer trusts ignores economic reality. The proposed settlement is inequitable and disproportionately favors investors in the riskier trusts."

Triaxx even suggested that the institutional investor group represented by Gibbs & Bruns and Ropes & Gray -- which includes BlackRock, Pimco, MetLife and the New York Federal Reserve's Maiden Lane entities -- bought discounted subprime-backed ResCap notes on the secondary market in anticipation of a windfall from the settlement. The CDOs moved to subpoena the institutional investors for information on what ResCap notes they own and what they paid for them. Triaxx subsequently withdrew the motion for procedural reasons but has said it intends to seek the same information through ordinary Chapter 11 discovery requests.

Both ResCap and the investor groups backing the settlement have said there's no substance to Triaxx's objections. In a supplemental brief in support of the proposed settlement, ResCap's lawyers at Morrison & Foerster said the deal is the fairest and most efficient way to resolve reps and warranties claims. Gibbs & Bruns' institutional investor clients were more forceful, asserting that Triaxx's "conspiracy theory" about the group's concentration in subprime-backed notes is "simply wrong." According to Gibbs & Bruns' brief, the major investors in the group supporting the settlement have holdings across the spectrum of ResCap mortgage-backed trusts and don't operate as a monolith. They don't even know what their fellow group members hold, the brief said, so there's no means to conspire.

More fundamentally, Gibbs & Bruns argued that using losses as a proxy for reps and warranties breaches is the most sensible way to distribute recovery. Representations and warranties about underlying loan pools didn't vary much from trust to trust, according to this reasoning. And prime loans, which were subjected to higher underwriting standards, are less likely to experience breaches than subprime and second-lien loans. So the bigger losses in trusts backed by riskier loans should correlate with a higher breach rate.

Before the fairness hearing on the proposed settlement takes place in November, you can expect the investors who support the deal to point out that Triaxx hasn't even asserted any put-back claims of its own but has just dropped into the Bank of America and ResCap cases to demand a bigger share for itself.

But if Triaxx's allocation argument -- which it has not yet asserted in its BofA objections -- gains traction with U.S. Bankruptcy Judge Martin Glenn in the ResCap case, that could affect how other MBS sponsors negotiate global settlements. I've speculated that Gibbs & Bruns is in late-stage talks with at least some of the banks its clients have asserted put-back claims against. Those deals, like the ResCap settlement, will presumably be based on the BofA prototype, which included the same loss-based allocation formula as the one Triaxx has found objectionable. If by some chance Glenn agrees with ResCap that allocations should be based on a trust-by-trust assessment of breach rates, that could slow global settlement negotiations with other banks.

(Reporting by Alison Frankel)

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