Model evaluation of KMV model

KMV is a default prediction model based on modern option pricing theory, which is an important revolution to the traditional credit risk measurement method. First of all, KMV can make full use of the information in the capital market to quantify and analyze the credit risk of all publicly listed enterprises; Secondly, because the data obtained by the model comes from the stock market data, rather than the historical data of the enterprise, it can better reflect the current credit status of the enterprise, and it is forward-looking and has stronger, more timely and more accurate forecasting ability; In addition, KMV model is based on contemporary corporate finance theory and option theory, and has a strong theoretical foundation.

However, KMV mode, like other existing modes, still has many defects. First of all, the application scope of the model is limited. Usually, this model is especially suitable for credit risk assessment of listed companies. When applied to non-listed companies, some important variables in the model are often replaced by some accounting information or other indicators that can reflect the characteristic values of borrowing enterprises. At the same time, through comparative analysis, the expected default probability of enterprises is finally obtained, which may reduce the accuracy of calculation to some extent. Secondly, the model assumes that the asset value of enterprises obeys normal distribution, but in practice, the asset value of enterprises generally presents non-normal statistical characteristics. Thirdly, the model can't distinguish different types of debts such as repayment priority, guarantee and contract, which makes the calculation results of model output variables inaccurate. According to the characteristics of China's transition economy capital market, Beida Company has developed a KMV model of listed companies with China characteristics, which is currently in the stage of stress testing.