Abstract:

Government bailout during the financial crisis is a time-varying dynamic process， while most literature assumes that it is static， and finds that government intervention is an effective tool to block risk contagion and achieve financial stability. Systems subject is formulated and studied to a contagion process over time， with the system’s current state and the degree of government interventions affecting its future state. Then， the contagion risk under different intervention intensities is quantified and it is found that government bailouts can both reduce and magnify the contagion risk. Specifically， the government bailout can reduce the number of defaulting individuals in the bank contagion network， which in turn changes the risk spillover path in the network and the contagion cycle when the banking system reaches the steady state. When the intensity of government intervention is large relative to the external shock， bailouts can reduce the risk spillover path and contagion rounds， and then reduce the contagion risk of the banking system. On the contrary， government bailouts will increase spillover paths and contagion rounds， thus amplifying contagion risks.Furthermore， the bank risk contagion under different combinations of external shock strength and bailout intensity is simulated， and it is found that there is an efficient frontier of government bailout. Given external shocks， only when the intervention intensity exceeds the boundary can the government bailout play a role in blocking risk contagion and stabilizing the banking system； Conversely， the banking system could experience cascading defaults and even the loss of bailout funds already invested. In addition， it is found that the risk characteristics of the banking system will affect the location of the effective bailout frontier. Specifically， the greater the leverage skewness and direct correlation of the banking system， the more government intervention funds are needed to achieve effective rescue. Moreover， when the strength of external shocks is small， the impact of leverage skewness on rescue costs is higher than that of direct correlation. The effect of direct connectedness is greater when the strength of the external shock is large.Finally， based on the above research results， two policy implications are drawn. First， bank failure contagion is a dynamic process. Therefore， the setting of government rescue policy should pay attention to its comprehensive utility in the process of multiple rounds of risk contagion， rather than only the effect at the current time point， otherwise it may aggravate the contagion loss. Second， higher leverage skewness and direct correlation are the main factors that increase the cost of crisis management. Therefore， the regulatory authorities should strengthen the supervision of banks with high leverage and large-scale interbank business. The results of this paper have certain reference significance for the regulatory authorities to make crisis rescue decisions.

Key words: dynamic government intervention, risk contagion, bank system stability, effective bailout frontier