DETECTING FINANCIAL STATEMENT FRAUD: A COMPREHENSIVE REVIEW OF RESEARCH TRENDS AND FUTURE DIRECTIONS

Abstract

This paper surveys both domestic and international research on detecting financial statement fraud in firms, presenting a concise overview of the main findings. By comparing models that use financial ratios, non-financial information, and machine learning techniques, the study demonstrates that machine learning methods outperform traditional statistical approaches. It also identifies research gaps in Vietnam, such as the limited integration of non-financial data, the predominant reliance on traditional methods, and the narrow focus on companies listed on a single exchange. These findings contribute to the theoretical foundation for developing more effective fraud detection models and offer policy implications for relevant stakeholders.