The Comparision Of Credit Risk Evaluation Model Of China’s Listed Companies And Empirical Research

With the development of China’s banking industry and capital market,commercial banks,investors and listed companies are facing with an increasingly complex environment and restrict supervisory.Therefore,it is high time for them to improve their ability of risk management.Credit risk evaluation is the fundamental process of credit risk management.In order to improve of efficiency of credit utility by ensuring the high quality of the loan and lowering the bad loan ratio,commercial banks should improve their ability of credit risk evaluation.Investors can make better investment with the credit risk evaluation.Since the credit risk evaluation can be used to predict business failures and financial distress, the manager of the company can avoid the increase of credit risk by adjusting the operating strategy and financial structure according to the result of credit risk evaluation.According to different variables,credit risk evaluation model can be classified into three main types:financial ratio model,stock return model and cash flow model.Each kind of variable has its own advantages and disadvantages.Financial ratio is comprehensive and easily to get while it usually lags and can be easily counterfeited. Even though stock return is including more market information and able to reflect the investor expectation and industrial cycle quickly, it can be largely effected by the efficiency of stock market.Cash flow can reflect the real earnings and the flow of capital more quickly,but the cash flow system and the ability of cash flow management is imperfect in China. This paper aims at building financial ratio model,stock return model and cash flow model with the data of China’s listed companies. The best model can be found out to evaluate the credit risk of China’s listed companies by analysing the predicting accuracy, predicting speed and the fitting degree of different models. A comprehensive predicting model which combines different type of variable can be built.At first,this paper analyses the development,principle and features of each model.Then,an empirical analysis is conducted using the data of 118 China’s listed companies from January,1998 to February,2015.The first step of empirical analysis is to analyse the data by comparing the means of different variables in the model and then conduct univariate analysis.Besides, after finishing the test for multi-collinearity financial ratio model,stock return model and cash flow model have been built by using the method of logistic regression and the results of different models have been analysed.Finally,the comprehensive predicting model has been built with 6 financial ratios and cash flow variables which are selected from 18 variables based on the result of mean analysis and univariate analysis.Here are several main results of this paper:1.In terms of the predicting accuracy, financial ratio model is better than stock return model and cash flow model.The highest predicting accuracy of financial ratio model is 89.84%in the four years before the default. The average rate is 80.51%,nearly 10%higher than cash flow model, nearly 22%higher than stock return model. Cash flow model obtains the highest predicting accuracy of default companies, nearly 8%higher than financial ratio model. However, in terms of the total predicting accuracy financial ratio model is still better.2.In terms of predicting speed, financial ratio model can make the right prediction 3.51 years ahead of the default, half a year in advance than cash flow model and stock return model.Cash flow model is the best in predicting default companies since it can make the right prediction 0.26 year faster than financial model and 0.6 year faster than stock return model.But financial ratio model is still best when come to the total predicting accuracy.3.Although financial ratio model can do better in prediction,it does not mean all the variables in the model has stronger predicting ability than the variables in the stock return model and cash flow model.According to results of univariate analysis, the predicting accuracy of 6 variables among the 18 variables is above 60%,including the ratio of total debt to total assets, the ratio of operating capital to total assets, the ratio of net income to total assets,operating cash flow,change in liquidity and cash taxes paid.4.Comprehensive predicting model is built on 6 variables selected from the 18 variables.Emperical result shows that the comprehensive predicting model which combines the financial ratio variable and cash flow variable is better than the financial ratio model,stock return model and cash flow model.The total predicting accuracy in the year of t-1 and t-2 is above 90%. In the four years before default,the total predicting accuracy is nearly 7%higher than financial ratio model.Although these two models share the same total predicting speed,comprehensive predicting model is faster in predicting default companies.Therefore,comprehensive predicting model has better predicting speed since the cost of a false prediction of default companies is higher than the cost of a false prediction of non-default companies.In sum,financial ratio model is better than stock return model and cash flow model in predicting the credit risk of China’s listed companies.Comprehensive predicting model has better performance in predicting the credit risk than financial ratio model,suggesting combination of financial ratio viarables and cash flow virables is a way to improve the credit risk evaluation model.