Predicting Securities Fraud Settlements and Amounts: A Hierarchical Bayesian Model of Federal Securities Class Action Lawsuits

Authors


  • We are grateful to Adam Savett for providing the data examined in this article and for several helpful discussions.

Blakeley B. McShane, Assistant Professor, Kellogg School of Management, Northwestern University, 2001 Sheridan Rd., Evanston, IL 60208; email: b-mcshane@kellogg.northwestern.edu. Watson is Principal and Vice President at Juridigm, Inc.; Baker is William Maul Measey Professor of Law and Health Sciences, University of Pennsylvania Law School; Griffith is T.J. Maloney Chair in Business Law, Fordham University School of Law.

Abstract

This article develops models that predict the incidence and amount of settlements for federal class action securities fraud litigation in the post-PLSRA period. We build hierarchical Bayesian models using data that come principally from Riskmetrics and identify several important predictors of settlement incidence (e.g., the number of different types of securities associated with a case, the company return during the class period) and settlement amount (e.g., market capitalization, measures of newsworthiness). Our models also allow us to estimate how the circuit court a case is filed in as well as the industry of the plaintiff firm associate with settlement outcomes. Finally, they allow us to accurately assess the variance of individual case outcomes revealing substantial amounts of heterogeneity in variance across cases.

Ancillary