基于银行-资产二分网络的金融风险传染研究

Research on Financial Risk Contagion Based on Bank-Asset Bipartite Network

为研究金融风险的传染问题,将二分网络理论引入到银行-资产网络模型中,提出风险传染模型。基于二分网络理论,对银行-资产网络进行拓扑分析,分析金融系统对单个资产类别外部冲击的敏感性,并评估系统中可能导致传染存在的特征。实证分析选取我国2021年第三季度各银行数据进行压力测试,结果表明,引入的模型能够捕捉银行的投资资产对不同外部冲击情景以及贬值效应的敏感程度。

Aiming at the problem of financial risk contagion,this paper introduces the bipartite network theory into the bank-asset network model and puts forward the risk contagion model.Based on bipartite network theory,this paper gives a topological analysis of the bank-asset network and analyzes the sensitivity of the financial system to the external impact of a single asset class and evaluates the existence of characteristics that may lead to contagion in the system.As a case,the model is applied to test the stress of China's banking data systems in September 2021.The results show that the introduced model can capture the sensitivity of banks'investment assets to different external shock scenarios and depreciation effects,and assets with a higher proportion are more sensitive to external shocks.Banks with different asset sizes are also sensitive to external shocks,banks with large asset sizes are relatively closely related to assets,which makes banks with large asset sizes more prone to systemic risks.