Optimization And Application Of M-score Model Based On Kingenta’s Financial Fraud Case

In recent years,there have been frequent incidents of financial fraud in my country’s listed companies,and corporate financial fraud is a serious challenge to the corporate information disclosure system.Serious damage can be described as a "cancer" in the securities market.Therefore,it is necessary to severely crack down on the financial fraud of enterprises,and only by accurately and efficiently identifying the financial fraud of enterprises can lay a good foundation for the effective management of financial fraud.This paper selects the M-Score model,which is widely used in foreign countries,by sorting out the research literature of domestic and foreign scholars on the identification of financial fraud.On the basis of in-depth analysis of the motives and methods of Kingenta’s fraud,the M-score was tested by using Kingenta’s financial statement data from 2011 to 2020.It was found that the comprehensive accuracy of the model for the identification of Kingenta’s financial fraud was only 60%.It cannot achieve its application effect to US listed companies,so it is necessary to optimize the design of the model.In order to improve the validity of the model,by studying the domestic scholars’ research literature on the fraud of Chinese listed companies,on the basis of the 8financial indicators of the original M-Score model,7 new financial indicators that can characterize the fraud characteristics of Chinese listed companies,6 indicators that can reflect the characteristics of the industry and 4 non-financial indicators are added.And finally,building a financial fraud identification indicator system including 25 indicators.According to the chemical raw material and chemical product manufacturing industry where Kingenta is located,this paper selects 15 listed companies that received administrative penalty decisions from the China Securities Regulatory Commission due to financial fraud in the industry from 2010 to 2020 and 15non-fraud companies in the same industry as research objects.After data cleaning,46 fraudulent samples and 281 non-fraudulent samples were obtained.After performing descriptive statistical analysis,significance analysis and correlation analysis on the above sample data,through binary logistic regression analysis,12 indicators with strong explanatory power were screened out,including accounts receivable incremental indicator,gross profit margin incremental indicator,asset quality incremental indicators,sales revenue incremental indicators,financial leverage incremental indicators,accrual items incremental indicators,inventory ratio,difference between inventory growth and revenue growth,cash ratio,accounts receivable turnover industry deviation,ownership concentration,and types of audit opinions.Finally,a financial fraud prediction model with these 12 indicators as the core is constructed.Using Kingenta’s financial data from 2011 to 2020 to test the optimized M-Score model,the comprehensive accuracy rate reached 80%,which is20% higher than the original model.In the end,this paper puts forward suggestions for the use of the M-Score model,and puts forward specific measures to control the financial fraud problem of enterprises based on the theory of fraud risk factors,which can be used for reference.Different from the empirical analysis of the fraud of listed companies in the entire A-share market,this paper is based on Kingenta’s financial fraud case and the construction principle of the original model,starting from an industry perspective,adding industry analysis indicators and non-financial information analysis indicators,building a more targeted financial fraud identification model for listed companies.