The Selection of Claims in Securities Arbitration for Settlement: A Text-Based Analysis

Abstract

This paper develops text-based measures of case strength and similarity using Statements of Claims filed in securities arbitration concerning investments in Puerto Rico municipal bonds to predict arbitration outcomes. We find that stronger customer claims and claims more similar to previous claims are more likely to settle than result in an award. These claims are also associated with higher total customer payout, including both settlements and awards. Our findings offer unique insights into the role of settlements as an alternative method of dispute resolution, and the selection bias associated with claims which result in an award.

Keywords: Securities Arbitration, Settlement, Statement of Claim, Textual Analysis

JEL Classification: D14, D18, G18, G24, G28, G3, K41