Can Corporate Governance Information Facilitate Accounting Fraud Detection? Machine Learning Evidence for Chinese Listed Firms

Synopsis The research problem We examined whether corporate governance information, as nonfinancial information, is useful to regulators—a crucial yet underexplored information user—for accounting fraud detection. Motivation or theoretical reasoning Detecting accounting fraud has significant implications for all participants in the financial market. Given the economic significance of China’s financial market as the second largest economy and the prevalence of accounting fraud in China due to its relatively weaker institutional environment, efficient and accurate detection of accounting fraud is of paramount importance to regulators as well as other participants in the Chinese capital market. While numerous empirical studies have underscored the importance of examining financial information in detecting accounting fraud, it remains unclear whether nonfinancial information, specifically corporate governance information, has incremental value in helping regulators with accounting fraud detection in China. The test hypothesis Based on the corporate governance literature, we hypothesize that incorporating corporate governance information alongside financial information enhances regulators’ ability to detect accounting fraud. Target population We studied a sample of Chinese listed firms from 2007 to 2019. Analyses Using accounting fraud data from 2007 to 2019, we examined whether corporate governance information, on top of raw financial information, can help regulators improve the accuracy and efficiency of accounting fraud detection. We obtained accounting fraud cases for Chinese listed firms from the China Securities Regulatory Commission (CSRC), and raw financial and governance variables from China Stock Market & Accounting Research (CSMAR). Findings Using machine learning techniques, we found that incorporating corporate governance information with raw financial information improves accounting fraud detection. Specifically, across the five corporate governance categories, board structure and the personality traits of executives are crucial for identifying fraudulent activity. Moreover, chairperson age, performance-based composition, and ownership concentration are particularly relevant corporate governance factors when it comes to accounting fraud detection. Our findings suggest that nonfinancial information, such as corporate governance factors, provides incremental value beyond financial statements that can be particularly useful to regulators, who are important yet ignored information users in the capital market.