Studying The Connectedness And Systemic Importance Of China's Bank System Based On The Network Topology

Systemic risk management has become the main content of the financial regulatory system whose main feature is macro prudential regulation. Because systemic risk stems from inter-agency relevance in the system, its measurement and analysis thus became the core of systemic risk regulation.In this paper, we adopt the method of the network topology of variance decompositions, which was proposed by Diebold F.X. and Yilmaz K. in 2014. We regard variance decompositions as weighted, directed networks, and then use the concept of node degree and centrality in network to define the connectedness and systemic importance in finance field.The main researchs of this paper include the following contents:(1) Introduces the network topology, node degree and centrality, generalized variance decomposition theory, as well as theprocessing of data andthe establishment of model.(2) Uses stock volatilities of Chinese 13 major listed banks in 2012-2015 years to get generalized variance decomposition, and then get dynamic network adjacency matrix of the banking system. Further more,uses the concept of network node degrees to analyze the connectedness between theinstitutions of banking system.We found that the highvalues of pairwise directional connectedness measures almostlytake placebetween the state-owned banks. Bank of China, Agricultural Bank of China, China Construction Bank, Industrial and Commercial Bank, theirstock's total directional connectedness to othersare successively ranked the first, the third,the fourth, and the fifth, so we get the conclusionthat they are in more central positions within the system. We can also find that the measure of total connectednessin the banking system is very high, and the overall level of risk is high.(3) Uses stock volatilities of Chinese 13 major listed banks from 2012 Q4 to 2015 Q4 to get static generalized variance decomposition, and then based on node's degree centrality and eigenvector centrality to analyse systemic importance of each banks in system. Then quarterly intervals, using panel data model to analyze the influencing factors of systemic importance. The empirical results show that the size of banks, financial networks correlation and liquidity interactively affect the systemic importance of the listed banks.