Research On Systematic Risk Transmission From The Perspective Of Bank Network

Modern financial system is highly interdependent because of all sorts of connections and forms a complex financial networks, interconnection between financial institutions can lead to shock transmit through the financial network to form systemic risk, the subprime crisis is a typical example.Nearly more than ten years, network science has developed rapidly and gradually applied to financial contagion research, promotes its becomes a new research direction in the field of financial stability.From network perspective, this dissertation based mainly on bank network formed from lending to each other between banking institutions (system), simulates risk contagion effects of different shock in the bank network, in order to further understanding of financial systemic risk formation and contagion mechanism. The main research work and conclusions are as follows:1. The theory analysis of systemic risk contagion mechanism. Credit shock is shown as banking institution defaults its debt and damages creditor bank’s assets, and may cause konck-on effects; market liquidity shock is usually shown as asset fire sales and affects all banks, produces network externalities; funding liquidity shock’s typical characteristics is rise in interbank debt rollover risk, individual bank’s liquidity hoarding behavior because of funding liquidity shocks may result in the interbank market freeze, and leads to network externality. Contagion analysis based on network model shows that the contagion effects of credit shock has attenuation characteristics, market liquidity shock has amplification characteristics, funding liquidity shock has not attenuation characteristics. Subprime crisis is result of risk contagion caused by the interaction of credit shock, market liquidity shock and funding liquidity shock.2. Cross-border bank network risk contagion analysis. This dissertation depics cross-border bank network by means of cross-border financial claims data of BIS, founds that cross-border bank network density shows the characteristics of significant rised before the global financial crisis and then moderate declined. The banking system of US and UK have always been the most important node in the network, however, some banking system, such as Turkey and Greece’s are in the periphery of cross-border bank network. This dissertation simulates and compares contagion effects from two dimensions of the data type (IBB and URB) and shock type (credit shock, credit shock and liquidity shock) with network analysis method based on the balance sheet. Simulation results shows that:(1) Under the condition of different shocks, US and UK banking system are the most systemically important, French and German banking system failure can also cause more serious influence. Belgium and the Netherlands banking system are the most vulnerable, this is because its external claim is relatively high and so concentrated in a small number country that it is weak to resist external shocks.Greek and Turkish banking system have very strong immunity in the face of external shocks due to its foreign claim are relatively small, US banking system is also resilient.(2) Under the condition of credit shock and liquidity shock, all the banking system becomes more fragility, it shows the importance of liquidity. Among them, the importance of France and Germany and Spain’s banking system increase significantly, and the contagion effects of Spanish banking system bankruptcy appears the phenomenon of tipping point.(3) simulation also finds the contagion phenomenon result from the indirect connection, and the simulation results seems to have warning role for the outbreak of the subprime crisis.3. China bank network risk contagion analysis. This dissertation depics China bank network by means of China’s interbank market lending data, and finds that large commercial banks, part of joint-stock banks and CDB are in the core of the network, the bank of China is the most important nodes in the network, while, part of joint-stock banks, all the city commercial banks, rural commercial banks and foreign capital banks are in the periphery of the network. This dissertation simulates and compares contagion effects of different network structure and different shock types with network analysis method based on the balance sheet. Simulation results shows that:(1) On the whole, China’s interbank market contagion risk is small. Different simulation both show that the bank of China is the most systemically important bank, followed by the industrial and commercial bank of China. Higher capital levels can effectively enhance the bank’s ability to resist risk.(2) Compared with the credit shock alone, contagion effects aggravates and contagion threshold decreases after add liquidity shocks. When considering the dual shocks, systemic importance of Remaining bank (on behalf of the common shock) is particularly significant, in extreme cases, liquidity shock of the Remaining bank alone would cause other banks fail, this is result of the rest of the other bank financing depend on Remaining bank seriously. Moreover, contagion effects of Remaining bank and the bank of China also appears the phenomenon of tipping point.(3) The interbank market structure affects risk contagion. When market structure changes from complete structure for incomplete structure, contagion effects caused by credit shocks aggravates, dual shocks can lead to the contagion threshold of the bank failure further decreases, when interbank market concentration is very high, different shock can lead to number of bank have a contagion effect increases.4. Macroprudential regulation under the network perspective. Comparative analysis of the basic characteristics of microprudential regulation and macroprudential regulation. In view of the microprudential regulation defects, the macroprudential regulatory reform after the global financial crisis from network perspective is discussed, it mainly includes comprehensive and systemic regulation of financial institutions (shadow banking system should be brought into the regulatory scope), controlling the financial network structure, dealing with the issue of too big to fail and too interconnected to fail.This dissertation’s main innovations are as follows:1. Study financial risk contagion from a new perspective;2. Puting forward a new method to expand the application range of the RAS algorithm;3. Constructing systemic risk index to measure the systemic risk of cross-border banking network;4. Find the phenomenon of tipping point of contagion effects and the evidence of liquidity shocks important in empirical network simulation.