Courtroom Valuation, Rulified Finance, and the Promise (and Perils) of Machine Learning

Abstract

About four decades ago, financial valuation methodology started to play an increasingly pivotal role in the litigated outcomes of high-stakes commercial disputes. In doing so, however it inadvertently smuggled in a Trojan Horse of its own making, unleashing several undesirable collateral consequences that have come to undermine the practical value of valuation in litigation (notwithstanding its enduring popularity). This paper makes three contributions. First, using simulation evidence based on actual firms’ value and other financial variables, we show that current practices allow considerable expert discretion—what we refer to as “expert degrees of freedom.”13 Second, we argue that emerging tools from machine learning (ML) may be able to augment, improve, and even supplant prevailing courtroom valuation practices. Third, we argue not only that judges can admit expert evidence based on such approaches, but also that they can and should begin to expect them from litigants and experts.