Research On Banking Systemic Risk Based On Multilayer Network Theory

With the development of the modern interbank market,there are complex creditor-debtor relationships amomg banks,which provide a diffusion path for bank risk.Along with the advancement of marketization of China’s banking system and the deepening of the degree of global economic uncertainty,the external environments of China’s banking industry will be more complicated.In addition,the reform and innovation of the financial system of our country also bring new impact to the banking industry.On the one hand,the liberalization of the interest rate and the emergence of the internet financial tools narrow bank spreads and squeeze the profit margin of bank deposit and loan,which force banks to transform and upgrade.On the other hand,the deposit insurance system is part of China’s financial reform,and it is conducive to the development of the banking industry in the long run,but in the short term,losing the stealth guarantee of the government,the bank will be impacted and faced with local risk if the bank is mismanaged,along with the effect of external competition and other factors.If risks diffuse in the interbank market,it is likely that the risk will be induced in the banking system.According to the existing research,this paper investigates banking systemic risk based on the multi-layer network theory,which is useful to understand systemic risk and provide decision-making basis for systemic risk management.Firstly,the dynamic model of bank multi-layer network is constructed.This paper presents a network structured representation of interbank market based on multi-layer network theory.According to the secured and unsecured lending relationships,an arbitrary structure of interbank directed,weighted multi-layer network is constructed.Starting from two kinds of systemic risk contagion channels of banks,namely,the interbank lending network and the overall loss caused by the change of external asset prices,the bank behavior is modeled based on the bank balance sheet.Combining with the dynamic change of external asset price to describe the dynamic change mechanism of interbank balance sheet,and according to the mechanism of bank withdrawal and bankruptcy liquidation,the paper describes the process of bank systemic risk contagion,and then constructs the dynamic model of bank multi-layer network.It provides the foundation for the further study of the banking systemic risk.Secondly,banking systemic risk is simulated based on the multi-layer network model.Based on the analysis of the formation mechanism of banking systemic risk under the influence of bank multi-layer network,this paper selects the random multi-layer network structure among banks,small world multi-layer network structure among banks and scale-free multi-layer network structure among banks through the dynamic model of interbank multi-layer network.The impact of initial shock and different network structures on systemic risk is simulated.With the increase of loan heterogeneity,the number of bankrupt banks in the small world multi-layer network and scale-free multi-layer network is less,and there is no significant upward or downward trend,while in the random multi-layer network,the number of bankrupt banks is more,showing a significant monotonous upward trend.With the increase of the average degree of the lending network,the number of bank failures in random,small world,scale-free multi-layer networks all show the trend of rapid increase first and then slow decrease;With the increase of the proportion of guaranteed lending,the number of bankruptcies in random,small world and scale-free multi-layer networks shows a downward trend as a whole;With the increase of price fall index,the number of bank failures in random,small world,scale-free multi-layer network is gradually rising.In addition,the impact of the partial loss of external asset price is more significant to the bank bankruptcy,followed by the total loss of unsecured borrowing,and the impact of the total loss of secured borrowing is the least.Finally,this paper empirically studies the influence of the multi-layer network structure on the systemic risk based on the stock data of China’s listed banks.Based on the correlation coefficient method,the multi-layer network model of China’s banks is constructed and its structural characteristics are analyzed.Based on the time-varying Copula-△Co Va R model,this paper measures systemic risk of listed banks.Based on this study,a panel regression model is built to test the impact of multi-layer network structure on systemic risk of banks.Among them,systemic risk is the explained variable,and the degree centrality and the betweenness centrality of the multi-layer network are the core explanatory variables.The results of the above study show that: the degree centrality has significant influence on systemic risk,while the betweenness centrality does not.In a word,this paper systematically studies the characteristics of banking systemic risk based on multi-layer network theory from the perspective of simulation and empirical research.The relevant research results are helpful to deepen the understanding of the multi-layer network structure and banking systemic risk,and provide the basis for the decision-making of banking systemic risk management,which has important theoretical and practical significance.