An Empirical Study On The Validation Of KMV Model: Does It Work On Chinese Listed Companies In Credit Risk Analysis?

The domestic financial institutions have actively seeked to construct the credit risk assessment model, since the Basel Capital Accord started in the year 2006. In fact, early detection of potential bankruptcy for financial institutions is a very important issue. In recent years, many financial theories have been developed. One of the most famous models is KMV Credit Monitor Model. However, attempts to demonstrate the effectiveness are quite few. The purpose of this study is to put the KMV Credit Monitor Model into a credit risk assessment model of financial institutions, and to demonstrate the effectiveness. The sample data is from 2008 to 2010, which has 52 corporations, with 31 corporations in financial crisis and 21 normal corporations. The analysis objects include the financial data and the data on stocks. The empirical analysis proceeds can be divided into two stages. First, we use KMV model to estimate the DD (Distance to Default, DD) of the sample companies, and then integrate the DD into Logit regression model to establish an early warning system.The findings suggest that the DD of KMV model is not significantly associated with probability of default in 1ST, 2nd, 3rd and 4th quarters prior to the financial crises of sample firms. In some extent, DD can be an early warning system. However, it can't predict exactly such a bankrupt will occur or not. As to the integration of the DD into Logit Regression Model, it does not enhance the early warning system, because there is no significant correlation between DD and credit risk in the Logit Regression Model. This paper tried to improve accuracy of the Credit Risk Model by modification time, correditing Default Point and revision the market value expression of the public company. Unfortunately, it is still no significant correlation between DD and credit risk in the Logit Regression Model. The reason may be that the stock market is not a mature market and the empirical study on the validation of KMV Model need to be further studied.