“Lies, Damned Lies, and Statistics”

Judge Chesler issued an Opinion in The Prudential Insurance Company of America et al. v. Bank of America, National Association et al., No. 13-1586 (Apr. 17, 2014), deciding the motion to dismiss the Complaint. Claims relating to fraud, a theory of underwriting abandonment and the 1933 Securities Act survived the motion to dismiss. The Court summarized the case as follows:

In a nutshell, this case arises from a dispute over the sale of certain residential mortgage-backed securities (“RMBS”) by Defendants [various Bank of America parties , including Merrill Lynch parties] to Plaintiffs [various Prudential parties]. The Complaint alleges that Defendants obtained the underlying mortgages, created the securitizations based on them, issued “offering Materials” for their sale, and sold them to Plaintiffs.

*     *     *

The Complaint alleges a variety of statistics in support of its claims. It is often not clear, however, what the basis for a particular statistic is. (1-2)

The Court’s description of the Complaint is pretty damning. But the Court does not find that the poor use of statistics in the Complaint is fatal to all of its claims.

Here are some highlights of the Court’s assessment of the Complaint:

  • “this Court does not find that the Analysis, as described in the Complaint, is such obvious junk research that it fails to constitute relevant factual allegations which, considered along with the other factual allegations in the Complaint, make plausible certain of the assertions of misrepresentation.” (8)
  • “The Complaint alleges that Defendants knowingly misrepresented that they would properly transfer title to the underlying mortgage loans to the particular trusts. The sole factual allegation made in support is: ‘Prudential’s forensic loan-level analysis revealed that across the Offerings Prudential tested, 43% of the Mortgage Loans were not properly assigned to the Trusts.’ Yes, if true, that is an astonishing fact– but there is not even a suggestion in the Complaint of a theory of how this gives rise to the inference of a knowing misrepresentation.” (13)
  • “The Complaint has so little explanation of the AVM [automated valuation model] methodology that this Court has no idea of how the computer used what information to generate property appraisals.” (15)

Notwithstanding the Court’s critique, it ends up finding the Complaint persuasive in the main:

The claim that Defendants’ representations about the underwriting practices and standards used in the issuance of the underlying mortgage loans were fraudulent because of a systemic abandonment of such underwriting standards is perhaps the central claim in this case. in brief, this Court has carefully examined the Complaint and finds that it states an abundance of factual allegations supporting this claim. (21)

The drafters of the complaint might reckon, ‘no harm no foul’ from the Court’s conclusion. But the rest of us might better see this as their having dodged a bullet, a bullet that the Plaintiffs’ attorneys shot at themselves. Mark Twain had said that “There are three kinds of lies: lies, damned lies, and statistics.” Not sure what he would have said about those in this Complaint — damned statistics?