Abstract: Credit derivatives play dual functions of investment trading and risk management. Their investment trading behavior is bound to bring credit risk contagion in the credit risk transfer (CRT) market. Existing studies are mostly limited to single-layer financial networks, and most of them ignore the heterogeneity of economic actors. In view of this, the latent default status of counterparties in the CRT market and counterparty heterogeneity are considered, a SCIRS model of counterparty credit risk contagion in the single-layer network is constructed, and the effect and influence mechanism of different mechanism probabilities and counterparty heterogeneity on the contagion effect of counterparty credit risk in CRT market are explored. Secondly, based on the SCIRS model of single-layer network counterparty credit risk contagion, further considering the heterogeneity of inter-layer and intra-layer contagion, the SCI1I2RS model of double-layer coupled network counterparty credit risk contagion is constructed. The influence mechanism of different network structures (single-layer network and two-layer coupled network) on counterparty credit risk contagion in the CRT market is compared and analyzed. On this basis, the evolution characteristics of counterparty credit risk contagion in the CRT market under different connection modes and different mechanism probabilities are compared and analyzed. The following main conclusions are obtained using computational experiments and simulation to analyze the evolution characteristics of counterparty credit risk contagion in the CRT market. First, in the single-layer network model, the mechanism probability exhibits a “global effect.” Second, the counterparty heterogeneities of counterparty credit risk contagion have a “global enhancement” effect in the CRT market. That is, the counterparty credit risk preference increases, the counterparty credit risk cognitive level reduces, the information disclosure coefficient decreases, the counterparty leverage increases, the counterparty risk overflow capacity decreases, and the counterparty influence increases. Third, the scale of default counterparty of bilayer-coupled networks is significantly larger than that of single-layer networks when the networks are stable. Moreover, in the process of contagion, the counterparty credit risk contagion of single-layer networks in the CRT market reaches the steady state of the networks more quickly than that of bilayer-coupled networks. Fourth, in the three interlayer connection modes of counterparty credit risk contagion under bilayer-coupled networks, the assortative link exhibits the “global enhancement” effect, while the disassortative link has “global suppression” effect. Fifth, in the bilayer-coupled network model, the intralayer and interlayer contagion probabilities promote the transformation of the latent default counterparty to the default counterparty, which enhances the credit risk contagion globally. The intralayer and interlayer immunity probabilities promote the conversion of default counterparties to other counterparties, which inhibits the credit risk contagion globally.

Key words: bilayer-coupled network; counterparty credit risk; default contagion; heterogeneity; contagion model