Research On Credit Risk Evaluation Of Chinese Listed Firms Based On The Semiparametric Models

How to build a suitable model and measure the credit risk with proper methods is a research hot spot in the field of current financial market risk management. The listed companies are important proportions of market economy, and main participants of our country’s financial security market, the credit situation of which directly determines the stability of market economy. At present, GARCH models have been mainly used to estimate the volatility by KMV model and VaR/CVaR model, through which we can study the credit risk of listed companies. However, normal GARCH models are insufficient in estimating volatility due to its rigor assumed conditions, lack of stability, and prone to trigger serious error. We found that semiparametric GARCH model and semiparametric EGARCH model can make up the defect of normal GARCH models and estimate volatility effectively. We proposed a new semiparametric model named semiparametric TARCH model based on semiparametric GARCH model and semiparametric EGARCH model, with which the volatility has been estimated. Secondly, based on the volatility obtained from semiparametric GARCH model, semiparametric EGARCH model and semiparametric TARCH model, we proposed two new models, semiparametric KMV model and semiparametric VaR and CVaR model, for studying credit risk of listed companies. Ultimately we testified the feasibility and superiority of the new model by using real evidence. We chose20special treatment (ST) listed companies and corresponding non-ST listed companies in different industries, of which the volatility has been estimated with semiparametric GARCH model, semiparametric EGARCH model and semiparametric TARCH model respectively, and applied the estimated volatility into KMV model and VaR/CVaR model to study the credit risk of listed companies. The result showed that semiparametric KMV model and semiparametric VaR/CVaR model are better than normal credit risk models on studying aspects of credit risk of listed companies and predicting operation situation in future, which is helpful to investors to make rational investments.