具有随机网络结构的银行系统稳定性研究

The Random Network Structure and Systemic Risk in Interbank Market

摘要: 随机网络是银行间拆借市场的重要特征，本文首先基于随机图构建了动态冲击强度下随机银行网络系统模型，随后对该模型进行了仿真研究，分析了不同存款冲击强度下银行系统的稳定性情况。结果表明，1) 存款冲击强度越大，银行系统越不稳定，银行之间的连接可以提高系统面临存款冲击时的稳健性；2) 随机银行网络系统在面临存款冲击时，存在最优连接度，且不同强度的存款冲击条件下，其最优连接度不同；3) 在最优连接度范围内，连接度越大，第一次倒闭越晚发生；此外，连接度越大的银行系统越容易发生银行倒闭的“雪崩现象”，因此长远来看，银行之间过度的连接不利于银行系统的稳定性。

Abstract: The random network is the important characteristics of the interbank market. Therefore, this paper uses the data of China interbank market to construct a random interbank network based on random graph. Then, based on the network constructed, the dynamical bank network system model is constructed. Considering the case of deposits shocks of the banks in the network system, we investigate the stability of random interbank network. The research results show that the random interbank network gets more unstable as deposits shocks increase while the connectivity among banks can make the network more stable. However, there is no monotonic linear relationship between connectivity and stability but an optimal connectivity for different deposits shocks. Within the optimal range of connectivity, the higher the connectivity is, the later the default occurs. Besides, banking system with high levels of connectivity tends to show avalanches, which means that excessive connectivity makes banking system unstable.