Cracks in the Foundation
The most prominent of the ongoing cases involving a single MBS issuer is Bank of America's proposed settlement of putbacks relating to mortgages that were originated, packaged, and sold as mortgage-backed securities by Countrywide. BofA acquired the liability for Countrywide's MBS losses when it bought Countrywide for $4.1 billion in 2008. It looked as though BofA would deflect the brunt of those liabilities after it agreed in June 2011 to settle putback claims with respect to 530 separate Countrywide MBS deals which were projected to have $108 billion of losses for the relatively small price of $8.5 billion. However, a recent flurry of legal rulings and evidentiary developments has threatened to blow up this accord. Much is at stake, as Bank of America's current loss reserves rest on the assumption that this deal will be approved, and other banks are beginning to use this settlement to estimate their own MBS liabilities.
BofA's settlement was not the typical legal compromise reached by adversarial parties after an arm's-length negotiation. Instead, it was a deal between BofA, The Bank of New York Mellon Corporation (BNYM) as trustee, and a small group of institutional investors who had purchased Countrywide's MBS and had reasons to keep BofA happy. The settlement was cobbled together in a relatively short amount of time (less than a year), without any lawsuit having been filed. In the legal world this is akin to a war treaty being hammered out without a single shot fired.
The suit was initiated by a group of institutional investors represented by Kathy Patrick of Gibbs & Bruns. This group wrote a letter to BNYM, which had been hired by Countrywide to serve as the trustee of each of Countrywide's 530 nonprime MBS trusts. (BNYM derives more than 60 percent of its MBS trustee business from BofA.) A trustee essentially receives a fee to oversee the flow of funds into and out of the trust. Unless it becomes aware of a particular type of default in the obligations of certain parties to the trust agreement (known as an "event of default"), the trustee has little obligation to take any overt action on behalf of investors.
Patrick's letter identified several areas in which Countrywide had apparently breached its obligations to investors, but carefully did not invoke an official event of default. Other MBS trustees had received and ignored similar letters, but BNYM decided to sit down with Patrick and BofA's attorney, Theodore Mirvis of Wachtell, Lipton, Rosen & Katz. Notably absent were any investors who might have pushed for an aggressive settlement. At the end of the process, the parties came up with a settlement of $8.5 billion, which sounded like a lot until it was recognized that this applied to all investors, not just Patrick's group. In effect, Countrywide MBS investors would get about 8 cents on the dollar for their losses. BNYM submitted the settlement for approval through a little-used special vehicle under New York state law called Article 77, which did not allow any investors to opt out, as they would be able to in a traditional class action. This had the effect of allowing BofA to put all of its repurchase liabilities behind it in one fell swoop. Article 77 had never been used to settle MBS liabilities of this nature, let alone 530 MBS trusts at once; it was a novel and unprecedented approach that smacked of a legal Hail Mary.
Fast-forward to today, where we find the settlement currently awaiting approval from Judge Barbara Kapnick in New York state court. Battle lines have been drawn, with BofA, BNYM, and the institutional investors that support the deal on one side, and on the other, the steering committee of institutional investors who oppose the deal, joined by the attorneys general of New York and Delaware. Central to the outcome of the case, which is currently set for a merits hearing before Kapnick at the end of May, is the question of whether the settlement is reasonable. The answer will have significant ramifications. The same group of institutional investors that initiated the BNYM settlement still represented by Patrick has also begun negotiations or sent legal notices to Morgan Stanley, Wells Fargo & Company, and JPMorgan Chase & Co. over their MBS liabilities.
BNYM has since admitted in the Article 77 proceedings that it reached the $8.5 billion number entirely on the basis of the work of one expert, Brian Lin of RRMS Advisors. Lin is a former Merrill Lynch trader with no disclosed experience in evaluating MBS settlements and no disclosed clients other than BNYM. He reached his figure based on several assumptions that resulted in "haircuts" applied to the original loss estimates for the 530 trusts. One assumption was that a smaller percentage of mortgages would default than actually did. Another was that it wasn't necessary to review individual mortgages from the Countrywide MBS for defects; instead, Lin looked at repurchase data based on BofA's experience with Freddie Mac and Fannie Mae on higher-quality conforming mortgages.
But perhaps his biggest assumption was that BofA would be compelled to repurchase only 40 percent of defective mortgages, which was based on the idea that plaintiffs would have to prove "loss causation." Under this theory, MBS investors, in order to recover, would have to prove that their losses were caused by a representation and warranty breach. This version of the "global catastrophe" defense which has been adopted in some form by every major MBS defendant rests on the alluring argument that MBS losses are the result of many factors (including the global credit crisis, the housing slump, deadbeat borrowers, and high unemployment rates), and that issuing banks cannot be blamed for all of these things.
The problem is that the contracts governing MBS say otherwise. They provide that a mortgage must be bought back if it breaches its guidelines in a way that "materially adversely impacts" the bondholder's or bond insurer's interest in the mortgage. As a result, the loss causation defense has now been rejected in three cases brought by monoline insurers against MBS issuing banks: MBIA v. Countrywide, Syncora v. EMC, and Assured v. Flagstar. The judges in the first two cases relied largely on insurance law to reach their conclusions. But U.S. District Judge Jed Rakoff's decision in Assured v. Flagstar was the first to take a purely textual approach to the language, and his finding could not have been more clear. Specifically, Rakoff held that "it is irrelevant to the court's determination of material breach what Flagstar believes ultimately caused the loans to default. . . . It is the fact that Assured faced a greater risk than was warranted that is at issue for the question of breach."
One further assumption that BofA made in calculating its settlement amount is that it would not be held to have successor liability for Countrywide. But any day now, in New York state court, Judge Eileen Bransten will issue her opinion on this very issue when she rules on a motion for summary judgment in MBIA v. Countrywide. Should she decide that BofA has successor liability for Countrywide, yet another pillar of BofA's settlement will come crumbling down.
The settlement has also come under heavy fire in recent months from the Countrywide MBS investors who oppose it. In a March 13 filing by investor American International Group Inc., expert Charles Cowan argues that Lin had better evidence at his disposal than the Freddie and Fannie repurchase experience, and he could have used this other evidence to estimate the breach rate for the pool of Countrywide mortgages, instead of using the nonanalogous number provided by BofA. He also questioned why Lin had not used proper estimates of default and loss severity rates, and contended that Lin's 40 percent "success rate" haircut was wrong. Using what he represented as a "scientifically valid approach to the use of the limited information provided to Mr. Lin," Cowan estimated BofA's total repurchase liability to be $56.3 billion a far cry from Lin's $8.5 billion calculation.