Research On Identification Of Systemically Important Financial Institutions In China From The Perspective Of Complex Network

The extensive and prolonged economic recession and social suffering brought about by the financial crisis indicate the fragility and interconnectedness of the financial system.The report of the 19 th National Congress of the Communist Party of China proposed that improving the dual pillar regulatory framework of monetary policy and macro-prudential policy,improving the financial supervision system,and holding the bottom line of avoiding systemic financial risks.At the same time,in the context of the new normal of the economy,the downward pressure on the economy is increasing,and the hidden systemic risks are gradually emerging and easy to spread through the complex financial system network.Identifying systemically important financial institutions is of unprecedented significance.This is not only the goal of designing the macro-prudential regulatory framework,but also the key process of carrying out regulatory work and projects,as well as the inevitable result of preventing and resolving systemic financial risks.Because the complex network can be used as the path carrier of risk contagion,and the financial system is normally considered as a typical complex system composed of a large number of interacting individuals.Gale and Kariv(2007)pointed out in the American Economic Review,the top journal of economics,that network structure can be used to understand financial crisis and risk contagion.Schweitzer et al.(2009)published a paper in Science,pointing out that complex networks are vital for financial crisis research.Barabási(2012)also mentioned in the review of Nature Physics that network research has crucial application value in complex systems such as finance.Therefore,the identification of systemically important nodes,the measurement of systemic financial risks and the exploration of the sources of systemic financial risks in the financial system and financial institutions by complex network method have increasingly become the focus of the theoretical and practical scholars in the field of financial risks at home and abroad.This paper takes the daily stock data of 40 listed financial institutions in China from2014 to 2021 as samples to build monthly PMFG financial institution networks.On this basis,the systemically important financial institutions are measured and identified with the help of the existing centrality and the new strength gravity centrality proposed in this paper,and the effectiveness of the centrality is tested from the two aspects of discrimination and accuracy with the Individuation Rate,Imprecision Function and Kendall Correlation Coefficient.In addition,this paper also measures the systemic risk contribution of each financial institution through d Co Va R,and uses panel regression model to profoundly explore the impact of network centrality on the systemic risk contribution from both the overall and epidemic impact perspectives.The main results of this paper include:(1)Most systemically important financial institutions are concentrated in the securities and banking industries,especially the securities industry,whose network influence has caught up with or even exceeded that of the banking industry,which conforms to the view proposed by many scholars that supervision should shift from “too big to fail” to “too connected to fail”;(2)Some financial institutions have the characteristics of "importance inertia",showing that they are in a highly contagious state in consecutive months.This “inertia” makes the identification method of important nodes have some forward-looking functions.In addition,some financial institutions exhibit “high frequency systemically important”,which provides the basis for targeted fine regulation;(3)The strength gravity centrality proposed in this paper has better performance in terms of differentiation and accuracy,but the combined analysis of different centrality is more conducive to comprehensively identifying important nodes;(4)The average risk spillover rate of the banking industry is still the highest,and the systemic risk contribution changes with time,except that the systemic risk contribution will increase significantly when the crisis occurs,which is similar in the period of stock market surge,indicating that risks accumulate in the rising market and emerge in the crisis.(5)The regression model found that network centrality was positively correlated with the systemic risk contribution,indicating that institutions with centrality had greater financial risk spillover effect on the financial system and a higher contribution to the overall systemic risk,and the positive correlation was more obvious during the epidemic period.It may be that the impact of the epidemic intensified the risk of intra industry and cross industry infection and resonance.Based on the above conclusions,specific policy suggestions are put forward:(1)Formulate appropriate regulatory policies for different industries,and implement differentiated and guiding regulatory work;(2)Combine the static evaluation of index method and the dynamic monitoring of network method,interactively guarantee the effectiveness and timeliness of identification;(3)Pay attention to small and mediumsized institutions with high relevance and strong infectivity,carry out key monitoring,set up risk early warning,and improve financial regulatory standards in a timely and appropriate manner;(4)Relying on “importance inertia”,the regulatory framework and main objects are formulated in advance for the early importance nodes;(5)Focus on cross-industry linkages,attach importance to excessive risk sharing and business interconnection,and strengthen the coordination and cooperation ability between departments;(6)Guard against the accumulation of systemic financial risks in prosperity,and appropriately cool down to reduce potential risks.