In this matter, the trustee alleged that defendant originator failed to cure or repurchase loans that allegedly failed to conform to representations and warranties. Vega expert Dr. Ethan Cohen-Cole performed an empirical analysis to determine whether the alleged loan-level defects, if true, would have resulted in a statistically significant increase in the risk for a given loan. Plaintiffs moved to exclude this testimony on several grounds.
First, plaintiffs argued that the methodology employed was not reliable and was “novel.” The Court noted that though the methodology combined a certain number of distinct statistical analysis, they were not inherently novel and unreliable. Second, plaintiffs argued that Dr. Cohen-Cole only accounted for 15 percent of plaintiffs’ expert’s breach allegations, and his analysis was thus unreliable. The Court noted that this issue could be resolved on cross-examination.
Next, plaintiffs argued Dr. Cohen-Cole’s opinion was contradicted by defendants’ underwriting expert. However, the Court pointed out that Dr. Cohen- Cole was retained to perform an entirely different analysis, from the point of view of an economist and not from the point of view of an underwriter. Therefore, the Court found there were no incompatible statements. Plaintiffs also argued that Dr. Cohen-Cole manipulated the results unfairly. The Court pointed out that this disclosure was made in the report, and therefore plaintiffs’ claim that defendants “hid” this information was disingenuous.
Finally, plaintiffs argued that Dr. Cohen-Cole should have performed more simulations. The Court dismissed this critique pointing out that he performed 101,000 Monte Carlo simulations for each loan under each scenario, 1,000 simulations for each of the 101 different data sets.
The Court ultimately denied the motion to exclude from the bench.