基于共同加权网络的银企信贷与银行系统风险

Bank-firm Loan and Banking System Risk Based on Weighted Banking Shared Network

每年都会有大量的企业从不同商业银行获得信用贷款,这些复杂的信贷关系会对银行的系统风险带来一定影响。通过将中国2007—2016年1.1万笔贷款汇总到327家商业银行总行级别,并采用二分网络映射方法,构建银行间共同加权网络,对信贷网络的结构性和传染性及其对银行系统风险的影响进行分析。结果表明:2007—2016年,银行信贷的共同网络从簇状“核心-边缘”层级结构逐步演变为“小世界群落”特性。银企间的信贷关系越紧密(连通比越高),越有利于降低银行系统传染性风险。信贷业务类型的异质性越强(度相关性越大),银行系统越容易产生结构性风险。银企间信贷的去中介化和网络异质性不利于降低银行系统的结构性风险。网络传染性指标中的连通比、网络结构性指标中的度相关性对银行系统风险有独立的显著性影响,适宜作为银行系统风险监控的先行指标。

There are many bank-firm credit loans between commercial banks and corporations every year.The complexity ofthe bank-firm credit relationship affects the banking systemic risk.By aggregatingeleven thousand loan data from 2007 to 2016 to the head office level of 327 commercial banks,this paper uses the binary network mapping method to set up weighted banking shared network,so as to research the structure and contagion of loan network and its effect on the banking systemic risk.The results show that,during 2007 and 2016,the shared network of banking credit gradually evolved from cluster“core edge”hierarchy to“small world community”characteristics.The closer the credit relationship between banks and firms(the higher the connectivity ratio),the more conducive to reducing the risk of contagion in the banking system.The more heterogeneous(the greater the degree correlation)of credit business types,the more likely the banking system is to generate structural risk.The disintermediation and network heterogeneity of bank-firm loans are not conducive to reducing the structural risks of the banking system.The connectivity ratio of the contagion index of the shared network and the degree correlation of the structural index of the network have independent significant impacts on the banking systemic risks,so they are suitable as monitoring indicators for the banking systemic risk.