A Study of the Corporate Fraud Early Warning Models

• Main Authors: Yu-Jhih Lin, 林渝之
• Other Authors: 林金賢
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/15194296219930674108

碩士 === 國立中興大學 === 企業管理學系所 === 102 === From the past literature reviews about corporate fraud early warning models, paper used financial ratio and corporate governance variables .Our paper adopted financial ratio and corporate governance variables and conservatism principle variable by using Logistic Regression to construct three different models. And we wanted to discuss in the different time, how the models predicted. We adopted 71 companies of financial fraud and 142 healthy companies to be our samples during 1999-2011 including three kinds of fraud type. The empirical results show that financial ratios have significant relationship between fraud for one and two years before fraud. And broad size and conservatism principle variable have significant relationship between fraud for one, two and three years before fraud. Furthermore, the Model 3 adopted financial ratio, corporate governance variables and conservatism principle variable can predict better than financial Model 1 and non-financial Model 2. The power of Model 3 for one, two and three years before fraud is 61%、55%、46%. In addition,　the Model3 could predict 54 fraud companies and the percentage of prediction is 77.4% , which was better than model 1 and model 2 .It is necessary to consider all the aspects including financial index and non-financial index to predict fraud.