Detection of price manipulation fraud through rational choice theory: evidence for the retail industry in Taiwan

Abstract

All enterprises, regardless of industry, are exposed to fraud risk. In the retail industry, the perpetrators of fraud manipulate the sales price of products instead of the quantity sold to avoid inventory discrepancies. Fraud examiners attempt to identify anomalous transactions using data analysis. This study analyzes the causes of fraudulent behavior, conceptualized based on the aspect of rational choice, and proposes an anomalous transaction detection model using variables identified as indicating fraud. A two-phase experiment analyzing a real-world data set from a retailer in Taiwan was designed to evaluate the performance of fraud variables. The findings demonstrate that these variables could slightly improve the results of the analysis and demonstrate that machine learning is applicable to fraud detection. In addition, this study contributes to rational choice theory to validate the applicability in fraud detection. Fraud examiners in the retail industry could reduce fraud losses by adopting the proposed approach, implemented using Weka software, to identify anomalies.