Zero-Revelation RegTech: Detecting Risk through Linguistic Analysis of Corporate Emails and News

Abstract

Natural language processing is a fast-growing area of data science for the finance industry. The authors demonstrate how an applied linguistics expert system may be used to parse corporate email content and news to assess factors that predict escalating risk or the gradual shifting of other critical characteristics within the firm before they manifest in observable data and financial outcomes. The authors find that email content and news articles meaningfully predict increased risk and potential malaise. The authors also find that other structural characteristics, such as average email length, are strong predictors of risk and subsequent performance. Implementations of three spatial analyses of internal corporate communication, (i.e., email networks, vocabulary trends, and topic analysis) are presented. The authors propose a regulatory technology solution to systematically and effectively detect escalating risk or potential malaise without the need to manually read individual employee emails.

TOPICS: Big data/machine learning, legal/regulatory/public policy