Research Of The Financial Institutions' Systemic Risk Contribution In China Based On The Analysis Of Complex Network

The financial systemic risk has become the focus of the study since the American sub-prime mortgage crisis and the European sovereign debt crisis occurred in succession.The connection between financial institutions and financial departments has become increasingly closer,and the areas involved in the operations of the institutions intersect each other.This results in the complexity of the modern economic and financial system.It has attracted great attention to identify systemically important institutions based on the research of characteristics of the correlation among financial institutions,which contributes to enhancing the supervision and dealing with the risk.The theory of the complex network provides a new method for identifying the conduction system of financial crisis and systemically important institutions.It is advantageous to define the financial system as a whole and from the macroscopic point of view.Also,the supervision department can extracting important information from the complex system.The process of China's financial market has lagged behind,so there are some shortages in our system.At present,under the influence of the transform of the economic construction and the international process,it is of great practical significance to study the systemic risk among financial institutions in China's financial sector.In this paper,with a perspective of the complex network,we study the relationship between network structure and systemic risk contribution.Based on the dynamic conditional correlation by applying DCC-GARCH model,first,we measure the systemic risk contribution with CoVaR.Then,financial networks are constructed from dynamic conditional correlations with graph filtering method of Planar Maximally Filtered Graph so that we can calculation the dynamics of financial institution's network topology structure.Finally,we analyze the relationship between the network topology structure and systemic risk contribution with panel data regression model.We find that financial institutions with greater node degree,larger node betweenness centrality,larger node strength,larger node clustering coefficient tend to be associated with larger systemic risk contributions.Therefore,regulators should pay attention to this kind of financial institutions.