融合CDS网络的银行间信用风险传染模型研究

信用违约互换(Credit Default Swap,CDS)作为银行信用风险缓释的重要工具,也是信用风险传染的重要渠道。因此,考虑到CDS抑制效应与传染效应的双重作用,本文构建了一个含有CDS网络交互作用的银行间信用风险传染模型,并运用计算实验与仿真模拟分析了在不同网络结构下银行网络信用风险传染与CDS网络抑制率,以及CDS网络信用风险抑制效应与银行网络传染率之间的关系,以及信用风险传染下银行网络和CDS网络的演化特征。研究结果表明:(1)银行网络的信用风险传染阈值与CDS网络抑制率呈单调递增的正相关系。而且,与同质网络结构相比,异质网络结构下的银行网络信用风险传染阈值相对较小。(2)CDS网络的信用风险抑制阈值与银行网络传染率呈单调递增正相关关系。而且,与同质网络结构相比,异质网络结构下CDS网络的信用风险抑制阈值明显较大。(3)银行网络最终破产规模与CDS抑制率呈单调递减的负相关关系。而且,异质网络结构下的银行破产规模要大于同质网络结构下的银行破产规模。(4)CDS网络破产规模与银行传染率呈单调递增的正相关关系。而且,同质网络结构下的CDS网络破产规模大于异质网络结构下的CDS网络破产规模。

Research on the Model of Inter-bank Credit Risk Contagion by Fusing CDS Networks

Through payment, settlement, bill, loan and other businesses, banks have established a connected inter-bank market, which amplifies the contagion effect and damage degree of credit risk. However, Credit Default Swap (CDS) is an effective way to mitigate the spillover effect of inter-bank credit risk. In this article, with the help of complex network theory, the inter-bank credit risk contagion effect and the credit risk inhibition effect of CDS on Banks are considered, and a two-layer network model of inter-bank credit risk contagion that integrates CDS network is constructed. The results show that: (1) the credit risk contagion threshold of the bank networkλ_B^*presents monotonically increasing characteristics with the increase of CDS network inhibition rateλ_C. Moreover, compared with homogeneous network structure, the contagion threshold of bank networkλ_B^*credit risk under heterogeneous network structure is relatively small. (2) The credit risk inhibition threshold of the CDS networkλ_C^*presents a monotonically increasing feature with the increase of the contagion rate of the bank networkλ_B. Moreover, the credit risk suppression threshold of CDS networksλ_C^*with homogeneous network structure is higher than that of CDS networks with heterogeneous network structure. (3) With the increase of CDS inhibition rateλ_C, the final bankruptcy scale of the bank networkR_B(∞)presents a monotonic decreasing characteristic; withthe increase of bank infection rateλ_B, it presents a monotonic increasing characteristic. Moreover, the scale of bank bankruptcyR_B(∞)under heterogeneous network structure is larger than that under homogeneous network structure. (4) With the increase of bank contagion rateλ_B, the bankruptcy scale of CDS network presents a monotonically increasing feature. Moreover, the bankruptcy scale of CDS networkR_C(∞)under homogeneous network structure is larger than that under heterogeneous network structure.The model and research results designed in this paper will provide new ideas for the application of two-layer network in the study of Banks and CDS counterparty credit risk contagion.