In re Foreign Exchange Benchmark Rates Antitrust Litigation

Background

On August 26, 2022, defendants moved to exclude certain proposed trial testimony from plaintiffs’ experts Dr. Hal J. Singer and Mr. Eric Robin. On September 30, 2022, the court issued a short order resolving the motions and stated that a separate opinion would explain its reasoning. This opinion provides those reasons. The rules discussed are the Federal Rules of Evidence.

Dr. Singer’s Testimony About Chats

The application to preclude Dr. Singer from presenting factual narrative about the chats he examined was granted in part and denied in part. He was permitted to testify about how many chats were collected, what kinds of chats were collected, and why. He could explain, from an antitrust economist’s perspective, what kinds of information exchange are inconsistent with competition and likely to produce wider spreads, and why the searches used to collect the chats were appropriate. He could also give limited illustrative discussion of particular chats.

Dr. Singer was precluded from going beyond what was needed to explain the basis for those opinions. In particular, he could not interpret chats, opine on whether particular chats involved competitors sharing price information, or read excessive numbers of chats into the trial record.

The application to preclude Dr. Singer from describing the alleged conspirators’ spread chats as “routine” was denied. The court recognized that “routine” is not a technical term, but found it would be odd and counterproductive to prevent him from using ordinary language to characterize the number and frequency of chats.

Guilty Pleas and Regulatory Settlements

The plaintiffs’ experts were precluded from offering opinions about the likely significance of guilty pleas and regulatory settlements. The court stated that defendants’ motions in limine to preclude evidence of regulatory settlements were for the most part granted because settlement is generally inadmissible to prove liability under Evidence Rule 408. Expert testimony explaining why someone might rationally settle was therefore irrelevant under Rule 402.

Guilty-plea evidence itself was permitted at trial under the court’s earlier order because its value in helping the jury outweighed the risk of unfair prejudice. The experts nevertheless could not testify about the likely reasons for those pleas. The court found that testimony only minimally helpful and likely to increase prejudice, so it was excluded under Rules 702 and 403.

The court also excluded Dr. Singer’s proposed use of settlements and guilty pleas as a basis for his economic analysis. Although class counsel argued that this material could help show that spreads widened during the period covered by the settlements and pleas, the court found that using the material this way would be an improper indirect method of introducing prejudicial and inadmissible settlement evidence. The court also noted that Dr. Singer had acknowledged that settlements and guilty pleas were not necessary to assess the economic effects of the alleged conspiracy.

Mr. Robin likewise was precluded from testifying about guilty pleas and regulatory settlements. His proposed opinion—that regulators would have been unlikely to impose those outcomes if the chatroom activity had been proper—was not based on specialized expertise and risked encouraging the jury to substitute regulators’ conclusions for its own judgment.

Mr. Robin’s Qualifications and Factual Narrative

Defendants argued in a footnote that Mr. Robin was not qualified to testify about “market color.” The court held that defendants forfeited that argument by raising it only in a footnote. The court also rejected the argument on the merits, finding that Mr. Robin’s experience qualified him to give opinions about the foreign-exchange industry, trading practices, and whether practices were routine or problematic. His having worked at only three of the sixteen alleged co-conspirator banks and not having worked in a compliance department did not change that conclusion.

The request to preclude other unidentified factual narrative from Mr. Robin was denied.

Dr. Singer’s Regression Analysis

The application to preclude several of Dr. Singer’s conclusions about the alleged narrowing of spreads was granted in relevant part. The court accepted that regression analysis is generally a reliable method for estimating relationships between variables, but found defendants’ criticism of this particular analysis persuasive.

The court concluded that Dr. Singer’s “Conduct” coefficient measured variation in spreads only between the last six months of 2013 and the first six months of 2014, rather than the average difference between the entire alleged conspiracy period and the post-conspiracy period. The court also found that Dr. Singer had not rebutted the assertions that the coefficient was unchanged when the Conduct variable was activated only for that twelve-month period and that the Conduct variable and year-fixed-effects variables were perfectly overlapping for the other years. Under Rule 702, Dr. Singer was therefore precluded from offering the broader interpretation of the coefficient.

The court overruled defendants’ other challenges to Dr. Singer’s opinions, but the challenged testimony was not offered at trial. Those challenges were therefore moot. The testimony included his opinion that spreads were wider during the last six months of the alleged conspiracy and the first six months afterward, opinions from his “cointegration regression” analysis, and his opinion that chats sharing spreads coincided and correlated with wider spreads.

Transmission Theories

The application to preclude testimony about “Transmission Theories” was denied. The parties used that term for proposed mechanisms by which an agreement in one chat about one spread could affect other spreads. Defendants identified vertical and horizontal correlations, adverse selection, and feedback loops as such theories.

Because the latter two theories were not addressed in expert testimony at trial, challenges to them were moot. The court rejected defendants’ relevance challenge to testimony about correlations between spreads, finding that this testimony was directly relevant to whether plaintiffs had proved a broad, global conspiracy to widen spreads.

Spread Durability or Stability

The challenge to expert testimony about the purported durability or stability of spreads was denied. Although defendants argued that Mr. Robin’s opening merits report gave only brief treatment to spread stability and relied on a questionable citation, the court found that his reply report provided a detailed discussion supported by his experience, record evidence, and scholarly sources. The court also found that Mr. Robin adequately based his opinion on spread matrices and explained their relationship to the foreign-exchange data used by another expert, Dr. Melvin. The court rejected the argument that Mr. Robin’s and Dr. Singer’s opinions lacked sufficient facts or data.

Disposition

The opinion concludes that, with respect to Dr. Singer’s and Mr. Robin’s testimony, the motion was granted in part and denied in part. The body of the opinion generally describes defendants’ applications to exclude testimony, while the conclusion refers to “Plaintiffs’ motion.”