《保险研究》20210704-《基于传染病模型的保险业危机公司救助研究》（张琳、李婉玉）

［摘   要］本文利用财务数据构建了以保险公司为中心、包含100家金融公司的危机传染网络；将传播动力学SIR模型引入危机传染网络中,加入破产率对模型进行改进,采用保险公司2016~2018年风险综合评级数据,近似估计出保险业危机传染概率；仿真模拟保险业危机传染演化过程,并在此基础上探析保险业危机公司救助策略。仿真结果表明:必要的保险市场退出机制可以缩短危机持续时间,加速保险业恢复到新的稳态水平；在危机传染初期,救助策略的最佳救助时机为危机开始传染的一个月内,最佳的救助成本为总成本①的20%；在这个最佳救助策略下,监管机构及早建立危机保险公司风险隔离带,切断传染路径,可以显著地降低危机传染的峰值。同时,保险保障基金能够以较低的救助成本获得最大的救助效果。

［关键词］传染病模型；破产率；参数估计；救助策略

［基金项目］教育部人文社会科学研究规划基金项目“跨界融合背景下保险业系统性风险感染路径分析及预警机制研究”（17YJA790093）；中国保险学会2019年年度研究课题“基于双层复杂网络和传染病模型的金融业系统性风险防控策略研究”（ISCKT2019-N-1-05）。

［作者简介］张琳,湖南大学风险管理与保险精算研究所所长、金融与统计学院教授,研究方向为保险精算、巨灾保险和非寿险精算；李婉玉,中国人寿养老保险股份有限公司。

The Rescue Strategy for Troubled Insurance Companies Based on the SIR Model

ZHANG Lin,LI Wan-yu

Abstract：This paper used financial data to build a crisis infection network which centered on insurance companies and included 100 financial companies.By introducing the SIR model of communication dynamics into the crisis infection network,this paper added the bankruptcy ratio to improve the model,and estimated the crisis infection ratio of the insurance industry on the basis of the insurance companies′ 2016~2018 risk comprehensive rating data.Then simulating the evolution of the insurance crisis infection process,it explored on the rescue strategy for troubled insurance companies.The simulation results show that the necessary insurance market exit mechanism can shorten the duration of the crisis and accelerate the insurance industry to return to a new stable level;in the early stage of crisis contagion,the best rescue time of the rescue strategy is within one month after the crisis begins to spread,and the best rescue cost is 20% of the total cost;under this best rescue strategy,the peak of crisis contagion can be significantly reduced if the regulatory authority establishes the risk isolation zone of the troubled insurance company as soon as possible to cut off the path of contagion.At the same time,the Insurance Security Fund can obtain the greatest rescue effect at a lower rescue cost.

Key words：SIR model;bankrupt ratio;parameter estimation;rescue strategy