通过对真实银行间借贷行为的数学抽象,建立了一个动态的银行间借贷网络模型.利用该模型,在计算机平台上,可以模拟仿真银行间市场.仿真的银行间市场能够展现出银行间资金流动等诸多规律,并能够用于研究银行间的流动性保护以及潜在的风险传染等规律性问题.同时,提出了分布式的银行间风险传染预警策略.利用分布式方法,将每家银行的风险特征映射成银行系统整体风险预警值,能够有效的解决中央预警方式的时滞性问题.仿真实验结果表明,银行的违约和风险传染并不是突发的,而是银行系统的风险累积的结果;分布式预警策略反映了银行系统内风险的累积过程,能够很好地预警银行系统性风险.
A dynamic interbank loan model is established by mathematical abstraction for real bank lending behavior. By using this model, the behavior of interbank can be realistically simulated via computer platfor- m. Through the simulation, it is found that a lot valuable regulation of interbank which can be used to study liquidity protection and potential risk. Meanwhile, this paper also presents a distributed banking systemic risk prediction strategy. Using the distributed method this strategy maps the risk character of banks to banking sys- tem risk index. This method can effectively solve time-delay of centralized prediction strategy. The simulation results show that the bank default and risk contagion is not an emergency even, but an accumulated result. The presented strategy can also reflect the accumulated process and precisely predict the banking system risk.