Dependent Enterprises Of Default Group's Credit Risk Measurement

The enterprise group with correlated defaults credit risks are focused by the bank and the Supervisory department intensely. Due to the frequency of the related transaction and the disorder of the guarantee and the mutual clud, the credit risks of enterprise is infectious, retardant and burst. Besides, there are such situation as "multi-thread giving credit" and "giving credit excessively" when the bank gives the credit existence to enterprise group. The factors of information symmetry weakens the recognition capability of enterprise groups credit risk, which causes its risk to proliferate to or affect the related enterprise and financial organs. The risk is extremely destructive. Therefore, it is necessory to establish model that can measure accurately the violation dependence enterprise group's credit risks.This article introduces the mixed model under the complete information condition according to our imperfect market. This model hasn' t considered the situation of other related enterprise group but only a single enterprise. After the research and analysis, the writer applies appropriate copula function to the mixed moder under the incomplete information condition, so that to protray the structure of violation and relying between the enterprises in the enterprises group more conveniently and accurately. As the fluctuation rate of the assets value is the important input variable which affects the assessment of the model, the accuracy of the assessment of the fluctuation rate will have much to do with the violation measurement result of the model. But in current measurement research on the violation probability, the fluctuation rate is always regarded as a constant but rarely the fluctuation rate heteroscedasticity, therefore, we can forecast its fluctuation rate through the GARCH model.The real diagnosis result indicated: Comparing with the other three kind of Copula function, the Gumbel contiguous function can portray the rear part dependence accurately in the credit risks model; When the rating is based on the entire "enterprise group" and adopts Gumbel Copula to the measurement of the cartel violation probability in the enterprise group, we can discovere that considered relies on one another the violation accumulation violation probability is obviously bigger than the situation which hasn' t considered the violation relying on one another, which is to say the neglects to the violation dependence consideration can underestimate the potential risk.