KMV模型在我国商业银行信用风险管理中的适用性分析及实证检验

KMV Model in China's Commercial Bank Credit Risk Management Analysis and Empirical Applicability

在介绍KMV模型、Credit Metrics模型、Credit Risk+模型和Credit Portfolio View模型这四种国际流行的信用风险管理方法的基础上,基于定性和定量分析相结合,对这四种信用风险管理方法进行比较分析,认为KMV模型最适合我国目前的国情。以2013年45家ST公司和与之配对的45家非ST公司以及2014年20家ST公司和与之配对的20家非ST公司为样本,对样本的违约距离进行实证检验。实证结果表明KMV模型基本上能够识别上市公司的信用状况,但是也有一些企业的违约距离不符合实际情况,这也说明该模型在我国商业银行信用风险度量中的识别能力有限,究其原因可能与该模型所要求的一些假设条件在我国尚不能得到有效满足等因素有关。因此,我国商业银行在对债务企业进行信用评价时,综合利用KMV模型与债务公司的财务数据会使信用风险的度量结果更加可靠。

This paper introduces four popular international credit risk management methods：the KMV,Credit Metrics,Credit Risk＋,and the Credit Portfolio View.Based on qualitative and quantitative analysis,this paper compares these four credit risk management methods and finds out that the KMV is the most suitable for our current situation.The sample consists of 45ST（specially treated）listed companies and paired with 45non-ST companies in 2013 and 20ST companies and paired with 20non-ST companies in 2014.We also have an empirical test on the distance of default of the sample.The empirical results show that the KMV model can identify the listed company＇s credit situation,but there are some companies default distance is not realistic.It also shows that the model＇s ability to identify credit risk is limited.Among other factors,the reason may be related to that some assumptions of the model required in China is still not effectively met.Therefore,it is more reliable for China＇s commercial banks use both the KMV model and the company＇s financial data.