网络视角下的金融结构与金融风险传染

基于不同的银行间市场网络结构假设，利用中国银行业数据，使用最大熵方法估计银行间资产负债关系，建立银行间市场网络以研究我国单个银行破产引发的金融风险的传染概率和影响程度，进而通过建立异质性银行的多主体仿真模型研究银行间市场结构和银行资产负债表结构对金融风险传染的影响. 研究结果表明：相比于完全连接网络，中心-边缘的层级网络将增大金融风险传染的范围和程度. 此外，对于银行资产负债表结构而言，所有者权益占比提高将增强金融机构的抗风险能力，降低传染概率，而银行间资产和负债占比提高则会扩大金融风险传染的影响. 这为我国银行业系统风险监管中对金融结构的关注提供了重要启示.

Financial structure and financial contagion from the network perspective

This paper estimates the interbank bilateral asset and debt positions based on different assumptions of interbank network structures by using data of China's banking sector. The network structure of interbank market is modeled to analyze the financial contagion caused by a singular bank failure. A multi-agent simulation model including heterogenous banks is built to study how interbank network structures and banks' balance sheet structures affect financial contagion. The results indicate that a core-periphery interbank structure would increase the influence of financial contagion compared with a complete network structure. Moreover, the balance sheet structure would affect financial contagion, with a higher equity ratio increasing financial stability and decreasing probability of financial contagion, and a higher interbank asset and debt ratio the opposite. This implies that regulation of systemic risk in banking sector should pay more attention to financial structure.