Research On Systemic Financial Risk Based On VaR Multiplex Network In The Context Of Epidemic

The emergence of COVID-19 in early 2020 has had a severe negative impact on Chinese financial market,and the spread of the epidemic was quickly curbed under the country’s strong prevention and control measures.But globally,as the epidemic has not been fully controlled,there are still uncertainties in economic development and the possibility of further accumulation and outbreak of systemic financial risks in China.Therefore,to strengthen the control of systemic risk,it is significant to analyze the characteristics of risk association network in financial market in the context of the epidemic,identify important nodes of financial risk transmission,and explore the impact of risk association network structure on the level of systemic financial risk spillover.We select 36 listed financial enterprises including banks,securities and insurance,and empirically study the correlation mechanism of systemic financial risks among financial enterprises under the background of COVID-19 from the perspective of multiplex network.First,we measure the ΔCoVaR of financial institutions in the sample period based on the improved GARCH-DCC-t-Copula method,and focus on the abnormal changes in the level of systemic risk spillovers of financial institutions before and after the outbreak of COVID-19.Then,for the obtained VaR sequences,we select Pearson,Spearman and Kendall correlation coefficients to calculate the corresponding correlation matrix,so as to reflect the risk correlation between each institution.Meanwhile,the minimum spanning tree(MST)method is used to filter the information of the distance matrices,and the VaR multiplex risk value association network is further constructed.The clustering function is used to obtain the centrality structure characteristic index of the multiplex network.Finally,with ΔCoVaR as the explained variable and the central structural indexes of VaR multiplex network as the core explanatory variable,a panel regression model is constructed to analyze the impact of structural characteristics of VaR multiplex network on systemic risk spillover.The empirical results show that :(1)The level of systemic risk spillover in Chinese financial market reached its peak twice in January and July 2020,corresponding to the outbreak of large-scale epidemics in China and the United States respectively.In 2021,the volatility of systemic risk spillover level decreases,but the overall level is still high;(2)The characteristic indexes of VaR multiplex network can significantly reflect the negative impact of the epidemic and other emergencies on financial market,and the network topology changes greatly under the impact of the epidemic;(3)Some characteristic parameters obtained from VaR multiplex network have significant correlation with the level of systemic financial risk spillover,the eigenvector centrality has positive correlation with the level of systemic financial risk spillover,while the asset size of an institution has no significant influence on the level of systemic risk spillover.