Thompson-Tau Outlier Detection Method for Detecting Abnormal Data of Listed Pharmaceutical Companies in China

Based on the quarterly financial statements from 2003 to 2014 of a listed pharmaceutical company with financial fraud records, this paper partitions the dataset of financial ratios into abnormal and normal groups by using Thomson-tau outlier detection method. Combined with the information revealed by China Securities Regulatory Commission (CSRC), in the empirical research we find that type I error rate and type II error rate are low and the overall accuracy rate and the accuracy rate of detected financial fraud data are relatively high. The results show that Thompson-tau outlier detection method is credible and useful for clustering financial data. From the detected outliers, we can further find the fraud financial data.