基于资产组合相关的金融传染建模与仿真  

Modeling and Simulation of Financial Contagion Based on Correlated Assets Portfolio

作　　者：范宏[1] 刘春垚 Fan Hong;Liu Chunyao(Glorious Sun School of Business and Management, Donghua University, Shanghai 200051, China)

机构地区：[1]东华大学旭日工商管理学院

出　　处：《系统仿真学报》2019年第6期1062-1069,1084,共9页Journal of System Simulation

基　　金：国家自然科学基金(71371046)

摘　　要：金融机构投资的资产组合相关性是引起间接金融传染的重要原因,针对此问题的建模还未见。考虑金融机构投资的资产之间的相关性因素,构建基于资产组合相关的金融网络系统动态演化模型并设计仿真算法,研究资产相关性、资产多样性程度及市场密度(金融机构数与资产数的比值)对金融传染的影响。仿真计算结果表明,资产之间的正相关关系加剧了金融传染,而负相关关系缓冲并抵消了冲击影响;随着资产多样性程度的增加,传染发生的概率先增大后逐渐降低;市场密度为1(金融机构数与资产数相等)时传染发生的概率最大。该研究结果可为金融监管机构提供决策依据。The correlation of assets portfolio in financial institutions is an important cause of indirect financial contagion, but the modeling of this problem has not been studied yet. Considering the correlation between the assets invested by financial institutions, a dynamic evolution model and the simulation algorithm of financial network system based on correlated assets portfolio are constructed;and the impact of assets portfolio correlation, degree of assets diversity and market density(the ratio of financial institutions to assets) on financial contagion are studied. Simulation results show that the positive correlation between assets portfolio exacerbates financial contagion, while the negative correlation buffers and counteracts the impact. When the degree of assets diversity increases, the probability of contagion increases first and then decreases gradually. The probability of contagion is maximum when the market density is 1(the number of financial institutions is equal to the number of assets). These results provide the basis of decision making for financial regulators.

关 键 词：资产组合相关性 间接金融传染 动态演化模型 仿真算法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]