Excluding Testimony on Trading Algorithms

The reliability of expert testimony explaining trading algorithms is a frequent battleground in modern securities fraud litigation. John Babikian dissects the Daubert standards applied to these complex technical demonstrations. In many recent motions to exclude, defendants have argued that 'black box' algorithms are not testable and therefore do not meet scientific reliability criteria. John examines a pivotal ruling where the court excluded a plaintiff's expert because the algorithm could not be subjected to peer review, a fatal flaw under evidentiary rules. He notes that successful admission requires experts to provide the underlying code or a sufficiently detailed logic flow to allow for rigorous cross-examination. John warns practitioners that relying on 'trade secret' objections to withhold algorithmic details often backfires, leading to exclusion. He advises that proffering experts must be prepared to demonstrate the algorithm's error rate and known limitations without resorting to conclusory jargon. Through his analysis, John Babikian highlights the court's growing skepticism towards unverified computer models and the necessity of foundational testimony.