Credit Risk Measurement And Empirical Analysis Based On A-Share Listed Companies

Since 1980s,the volatility of financial markets has been increasing,on the other hand,financial globalization has become the general trend of develop-ment,banks and non-bank financial institutions are facing severe challenges in the region of credit risk.Under the situation of global low interest rates and even negative interest rates,China's financial market has accumulated a large number of asset price bubbles,and bank assets are gradually accumulat-ed risk during the rapid expansion period.With the slowdown in GDP growth,the Federal Reserve lifting interest rates,and the increasing of enterprise risk,the asset quality of China's banking is declining,and the liquidity risk is al-so increasing.Therefore,strengthening the management of credit risk is an important aspect to improve the bank's ability to operate steadily.One of the difficulties in credit risk management is measuring credit risk.Credit risk is usually closely related to the company's financial situation,by establishing a link between predicting credit risk and company history data,We construct a structural model of the default probability using stock data and financial data.After a further comparison and statistical analysis of dif-ferent models,the proper structural risk measurement model of China's listed companies can be determined.The above research can help the bank set up a proper credit risk model,and identify the credit crisis.Merton model is one of the major models to predict default probability and measure credit risk.In the 40 years since Merton model was published,in order to improve the prediction accuracy or expand the application scope,a lot of scholars have extended the basic model.In this paper,based on the data of 250 risk warning companies and 250 control companies over the past 10 years,we give an empirical analysis to investigate the Merton model and other alternative models,study the sensitivity of default prediction capability of Merton model,and further analyze the causes of different sensitivity.This paper compares and evaluates the performance of various optional models in predicting the default risk of listed companies in China.Several schemes that can improve the accuracy of the model are proposed,and a simplified mod-el with both excellent discrimination and small computational complexity is proposed.