Judge Allows Statistical Evidence of Trading Patterns in Fraud Case Against Former Western Asset Trader

MANHATTAN (LN) — U.S. District Judge Gregory H. Woods on Friday denied both the government’s and the defense’s motions to exclude expert testimony in the securities fraud case against former Western Asset Management Company trader S. Kenneth Leech II, ruling that the parties’ competing statistical analyses of his trading patterns are sufficiently reliable to reach a jury.

The government alleges that Leech, who served as Chief Investment Officer for fixed-income strategies at Western Asset, engaged in a fraudulent “cherry-picking” scheme from January 2021 through October 2023. The indictment accuses him of executing trades and then allocating well-performing positions to favored accounts, such as the “Macro Opportunities” strategy, while allocating underperforming trades to disfavored accounts, such as the benchmarked “Core Strategies.”

To prove the scheme, the government proffered experts Harvard Business School’s Professor Lauren Cohen and New York University’s Courant Institute Professor Petter Kolm. They plan to testify that statistical analyses of Leech’s trade blotters show a pronounced skew in first-day returns: trades allocated to Macro Opportunities averaged a $0.24 million gain, while those allocated to Core Strategies averaged a $0.31 million loss.

The defense countered with experts from Charles River Associates and NYU’s Stern School of Business, Aaron Dolgoff and Professor Bruce Tuckman. They proposed a “blind allocation exercise” in which Tuckman allocated Leech’s trades without knowing their performance. Dolgoff testified that Tuckman’s allocations matched Leech’s 56.8 percent of the time, suggesting that portfolio management constraints—not fraud—explained the trading patterns.

Woods rejected the defense’s argument that the government’s experts improperly opined on Leech’s state of mind under Federal Rule of Evidence 704(b). The judge ruled that terms like “consistent with” and “driven by” described conduct and cause-and-effect relationships rather than impermissible conclusions about intent.

However, Woods ordered the parties to meet and confer on mutually agreeable language to ensure the experts do not cross the line into stating final conclusions about Leech’s mental state. He also flagged the defense’s use of the term “null hypothesis” to describe the government’s case as potentially confusing and likely to be sustained as objectionable if elicited at trial.

The court also denied the government’s motion to exclude the defense’s CRA exercise, finding the methodology sufficiently rigorous. Woods noted that while the exercise did not perfectly replicate Leech’s trading environment, the differences went to the weight of the evidence, not its admissibility.

Leech faces five counts including investment adviser fraud, securities fraud, commodity trading advisor fraud, commodities fraud, and making false statements to the SEC. He previously testified to the SEC that he typically had an allocation in mind when placing trades, a statement the government alleges was false.

Woods is a judge on the U.S. District Court for the Southern District of New York.