信用风险相关性度量的MRS Copula模型构建及实证研究

Construction of MRS Copula Model for Credit Risk Correlation Evaluation and an Empirical Study of China's Stock Market

在构建行业信用风险指数的基础上,将马尔科夫机制转换引入到信用风险相关性的度量中,建立了信用风险相关性度量的MRS Copula模型。以1990-2012年电力、煤气及水的生产和供应业,批发、零售、贸易业,石油、化学、塑胶、塑料业和信息技术业为样本的实证研究表明,行业信用风险相关性表现出较为明显的机制转换特征和非对称效应,在高风险状态,信用风险相关系数达到了0.7以上,而在低风险状态,信用风险相关系数在0.2以下.同时,信用风险"一损俱损"的特征比较明显,行业信用风险的下尾相关系数较为显著,而上尾相关系数则并不显著.商业银行可据此调整信贷资产结构,防范信用风险传染,以及优化信贷组合管理.

Based on the construction of industry credit risk index,this paper introduces Markov regime switching（MRS） process into the Copula model,and develops MRS Copula models for credit risk correlation evaluation.By applying the Chinese Listed Companies data from 1990 to 2012,the industry indexes of credit risk are calculated,and the empirical results shows that there are obvious regime switching processes of correlations among Electricity-gaswater production and supply industry,Wholesale-retail-trading industry,Petroleum-chemical-plastic-plastic industry and information technology industry.Furthermore,in high credit risk regime,the correlation coefficients are all higher than 0.7,and lower than 0.2 in low regime.Meanwhile,there is evidence that industry credit risk has contagion and spillover effects,which could be deduced from the result that the lower tail dependence is significant,while the upper is not.These findings are useful for commercial bank to adjust the structure of credit assets,guard against credit risk contagion and optimize Credit Portfolio management.