基于尾部依赖的保险业系统性风险度量  

Systemic risk measure in insurance industry based on tail dependence

作　　者：谢远涛[1] 蒋涛[2] 杨娟[3] 

机构地区：[1]对外经济贸易大学保险学院,北京100029 [2]对外经济贸易大学应用金融研究中心,北京100029 [3]中国科学技术发展战略研究院综合发展研究所,北京100038

出　　处：《系统工程理论与实践》2014年第8期1921-1931,共11页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(71303045);对外经济贸易大学研究生科研创新项目(A2012054;201403);对外经济贸易大学学术创新团队(CXTD5-04)

摘　　要：从系统性风险的界定出发,综合考虑系统性风险的两个视角,基于保险公司的股票数据探讨保险业的系统性风险.通过构造Kendall协同系数检验整体依赖性和尾部依赖性,以(非对称)尾部依赖性为切入点,构造SV-t模型和厚尾的SV-GED模型,利用AIC准则和Hit检验法对不同依赖结构的copula模型进行筛选.通过实证分析,检测出中国人寿、中国平安和太平洋保险公司的非对称尾部依赖结构,认为保险业可能成为系统性风险传导链条上的一环,并且完全可能自身隐含和集聚系统性风险.结论部分分析了系统性风险的原因,针对如何降低下尾风险依赖提出政策建议,为我国保险业宏观审慎监管提供一些支持材料.The concept of systemic risk was defined in this paper to test the systemic risk in Chinainsurance market based on stocks data from the two different angles of view. Take tail dependence,by constructing Kendall coordinate coefficient test, as cut-in spot, SV-t models and fat-tail SV-GEDmodels were constructed to fit the data. Based on AIC information rules and Hit test, the appropriatecopula models with different dependence structures were chose. From empirical analysis, unbalance taildependence structure between China Life, Pingan and CPIC is positive, China insurance market wouldtransfer systemic risk as part of a chain and itself may assemble systemic risk. The reason why there existssystemic risk in insurance market, as well as the suggestions on how to reduce the lower tail dependence,was put out. It may contribute to our Macro-prudential Supervision.

关 键 词：系统性风险 尾部依赖 Hit检验 Joe-Claytoncopula函数 SV模型 

分 类 号：F224[经济管理—国民经济]