A Study On Fitting Degree Between Actual Value And Theoretical Value Of China’s Listed Companies And Its Prediction

In the21st century, the influence of the credit risk to economic entity is deepening. In2007, the subprime crisis erupted in America, then it affected the global economy and formed the global financial storm. By the end of2009, the European debt crisis was beginning to show and continued to send force. Accompanying the deepening of the crisis, its effect also expanded from the economic field to political field. A series of credit events make people come to realize profoundly that the importance of credit risk management. How to improve the measurement accuracy of credit risk level is the most important aspect in the management of the credit risk.In the transformation of our country’s economic, in order to achieve the reasonable distribution of resources and lower their risk, financial institutions will need to find a method in microcosmic field that using foreign mature credit risk measurement model in China’s capital market after modification.On theory, this paper based on the economic theory of credit risk, introducing the modern credit risk evaluation model, comparing with the structure model, strength model and hybrid model. After comprehensive comparison, I choose the KMV model in structure model as the main object. In order to improve the applying degrees of the KMV model in our country, I did some necessary correction.In empirical, this article choose12companies that run across the credit risk horizon (ST or not is the standard) as the sample data. Measure the theory credit risk level of the company in2011using KMV model, finding the company’s actual credit risk level of2011in2012, then make the two level fitting operation, acquiring the threshold that the theoretical value fluctuation around the actual value. The acquiring of the threshold can effectively increase the company’s forecasting accuracy of the credit risk level. But the results of this measuring method is without the theoretical threshold, the threshold have to choose according to the research target completely.