Credit Risk Early-warning Model Of ST Listed Companies In China

The market economy is based on credit. Accompanied by the globalization of the economy, credit has been highly socialized, and credit relationship has become a kind of the basic economic relations. With the high-speed development of economy and society, the scale of credit transactions has been increasingly large. Accompanied by the credit transactions, credit risk has penetrated into all aspects of the economic activities. Subprime mortgage crisis had swept through the United States, the European Union and Japan who are belong to the world’s major financial markets in August2007.This crisis had come from the credit risk. It was reasoned from a large number of defaults by subprime home mortgage borrowers.The three major international rating agencies had cut Greece’s sovereign debt rating from2009, and that led the European sovereign debt crisis to break out. The root cause of the debt crisis is that the government had burdened much debt more than what they could afford. These crises had a profound experience and understanding of credit risk for people in the world. It could be said that the credit risk is not only can cause such a wide range of serious impact as the subprime mortgage crisis and the European debt crisis, its impact exists in all aspects where there is credit risk.After the development of20years, there are a great number of listed companies to play the backbone of the national economy. Listed companies have become the backbone of our national economy, and play a pivotal role in promoting the development of the national economy. With the rising influence and the driving force in the national economy, listed companies as the barometer of the stock market are more important. It can be said that the listed companies have an irreplaceable role in Chinese national economy. Listed companies as Chinese commercial banks’major credit objects, not only are they the main financing bodies of the stock market, but also are the bond market financing participants. The listed companys’ credit risk is not only reflected in the credit bond market, but also be found in the macro-environment. Listed companies as the important borrowers of funds in the credit markets in China, undoubtly also contain the serious credit risk and crisis. In the increasingly fierce market competition and the increasingly turbulent economic environment, taking the control of credit risk is very important for commercial banks and investors in the capital market. Credit risk does not happen overnight, but there is a gradual process. So we can take effective measures to control the risk in order to reduce the economic losses.In the most of the traditional credit risk early-warning models, the assumptions are too strict. Although modern credit risk early-warning models have the theoretical foundation; the solid data requirements are highly. The neural network model has the black box. These characteristics limit the use of the appropriate model. The logistic regression model does not require variables follow a normal distribution and having the same covariance, so it’s more suitable for the actual situation of relative credit risk early-warning model. So in this article, we use the logistic regression model to study.It is more suitable for building the credit risk early-warning model for Chinese listed companies.In the empirical analysis part, we use the financial data to built credit risk early-warning models in different periods. We have selected54financial indicators as the initial variables. Using the Kolmogorov-Smirnov test method, the T-test method and the Mann-Whitney U test method, we get the appropriate indicators. Then we use the principal component analysis to reduce the dimension of indicative variables. Finally we use the back gradually removed method to determine the last principal component factors to build the credit risk early-warning model. In this article, we get three credit risk early-warning models.From these models, we get the following conclusions:t-1-year credit risk early-warning model correctly predicts of97.1%of non-ST companies, ST companies correctly of95.1%, and the overall predictive accuracy rate of96.6%; t-2-year credit risk early-warning model correctly predicts of non-ST93.4%, of ST companies52.2%, the overall predictive accuracy is83.3%; in the t-3-year credit risk early-warning model,it predicts correctly of99.5%for non-ST, the overall predictive accuracy rate is74.8%. Overall, the accuracy rates of the credit risk early-warning models for non-ST’s prediction are relatively high.