保险公司长寿风险度量  

Measurement of Longevity Risk of Insurer:Model,Method and Empirical Research

作　　者：赵明[1] 王晓军[2,3] 

机构地区：[1]中国人民大学统计学院 [2]中国人民大学应用统计科学研究中心 [3]中国人民大学风险管理与精算中心

出　　处：《统计研究》2015年第12期76-83,共8页Statistical Research

基　　金：国家社会科学基金重大项目“我国养老保障体系应对人口老龄化挑战的对策研究”(13&ZD164);国家自然科学基金项目“社会保障预算管理研究”(71173230)的资助

摘　　要：本文梳理了几种重要的动态死亡率预测模型,给出了长寿风险度量的三种方法,选取了保险公司及国家官方公布的人口死亡率数据,度量了保险公司的两类长寿风险,对不同方法下长寿风险的度量结果进行比较,并分析了极限年龄与折现率变动对长寿风险影响的敏感性。研究发现:基于保险公司经营稳健性的视角,第一类长寿风险的度量应采用随机模拟法,第二类长寿风险的度量应采用标准公式法;极限年龄的变动对长寿风险的影响较小,保险公司无上调经验生命表极限年龄的必要;长寿风险对折现率的变动较为敏感,提高折现率,保险公司经营中的长寿风险显著降低。This paper summarizes the several important dynamic mortality models, proposes three methods of measuring longevity risk, and measures the two kinds of longevity risk of insurer by using the mortality data from insurance companies and National Bureau of Statistics. The study shows that the results of longevity risk are different in these methods. Based on the prudence principle, insurer should select the stochastic simulati6n method to measure the first kind of longevity risk, and select the standard formula method to measure the second kind of longevity risk. Moreover, the longevity risk is not sensitive to the terminal age, but sensitive to the discount rate. The longevity risk will decrease significantly by increasing the discount rate.

关 键 词：保险公司 长寿风险 死亡率模型 风险度量方法 

分 类 号：C812[社会学—统计学]