Toward an Integrated Framework With Corporate Culture for Financial Fraud Detection in China’s A‐Share Market
Zhaolong Zhang, Ying Wu, Fengmin Xu, Xiangyu Chang · Accounting and Finance · 2025
ABSTRACT We propose an integrated framework for financial fraud detection in China's A‐share market by incorporating corporate culture alongside traditional financial indicators. The framework includes five cultural dimensions: Integrity, Governance, Corporate Social Responsibility, Transparency, and Teamwork, integrated into fraud detection models, using advanced machine learning techniques such as Logistic Regression, Extreme Gradient Boost, and RuleFit. Our study shows that including cultural indicators improves fraud detection performance, enhancing accuracy, precision, and recall. The RuleFit model also uncovers interaction effects between cultural and financial factors, providing deeper insights into financial misconduct. This approach underscores the role of corporate culture in promoting market integrity, sustainability, and fraud prevention.