基于合成数据的系统性风险传染研究

The Study of Contagion Mechanism of Systematic Risk Based on Synthetical Data

本文基于合成数据的方法和复杂网络理论,对银行间系统性风险的传染问题进行了研究。首先在只能获取支付系统和银行年度报表数据的约束下,采用合成数据的方法,估算出我国支付系统中银行间的交易数据,从而得到交易矩阵。以此合成数据为基础,进一步借助复杂网络理论,利用计算机仿真技术对系统性风险传染问题进行了研究。研究结果表明,国有商业银行抵御风险能力最强,城商行和农商行这类小型银行对国有商业银行有着很强的依赖性。最后形成了若干政策建议。以上成果有可能为金融风险传染理论的发展和宏观审慎管理制度的完善等起到一定的推动作用。

This paper studies the contagion mechanism of systematic risk based on synthetical data and network model.First,under the circumstance that only the data of payment system and banks' annual accounts are available,we adopt the method of synthetical data to estimate the data of transactions between banks in payment system,and then get the transaction matrix.Base on the synthetical data,we make research on the contagion mechanism of systematic risk by using complex network model and computer simulation technology.We find that State-owned commercial banks have the strongest ability to defend against the systematic risk,while urban commercial banks and rural commercial banks have a strong dependence on state-owned commercial banks.Then we come up with related suggestions and policy implications which may improve the macro-prudent management both in theory and practice.