When Ratios Fall Short: The Case for Reconsidering Risk Proxies in Corporate Crime Research

• In Event: Digital Crime and Decentralized Finance: Artificial Intelligence, Cryptocurrency, and Large Language Models

Abstract

Due to corporate emphasis on performance outcomes, financial metrics are widely adopted as proxy measures for organization-level risk factors such as strain or distress. However, these metrics can be conceptually ambiguous, and can introduce myriad analytical challenges, such as extreme multicollinearity. Using a machine learning approach to analyze enforcement data of corporate financial crime, this study finds that commonly employed metrics also tend to underperform in predictive accuracy when compared to more disaggregated alternatives. These findings demonstrate the need for a critical reexamination of how risk factors are operationalized in corporate crime research.

Author

• Fiona Chan, Indiana University