《保险研究》20210608-《死亡率模型选择下职工基本养老保险的长寿风险估算》（贺磊、马昕）

［摘   要］本文采用残差分布情况、生物合理性、预测稳健性、参数估计稳健性、模型简洁性和预测准确性六种指标,从六种经典死亡率模型中选择适合中国0~89岁男女的最佳随机动态死亡率模型,构建人口发展模型和城镇企业职工基本养老保险精算模型。基于死亡率不确定性利用VaR和CVaR估算基本养老保险制度财务缺口上限和超过上限的尾部风险,评估各省城镇企业职工基本养老保险的长寿风险,并模拟分析了渐进式延迟退休政策对不同省份养老保险长寿风险的影响。研究发现:经济发达、人口密集的省份未来面临的长寿风险较大；渐进式延迟退休会降低养老保险的长寿风险,但是政策的效果会逐渐减弱。

［关键词］长寿风险；随机动态死亡率模型；养老保险；延迟退休；区域异质性

［基金项目］湖南省社科基金资助项目(17YBQ0067)。

［作者简介］贺磊,湖南师范大学商学院副教授、硕士生导师,研究方向:长寿风险；马昕,湖南师范大学商学院硕士研究生,研究方向:长寿风险。

The Longevity Risk of The Basic Social Pension Insurance of Urban Employees Based on Mortality Model Selection

HE Lei,MA Xin

Abstract：The optimal stochastic dynamic mortality model for Chinese males and females aged between 0-89 years was selected from six classical mortality models by using the six indexes of residual distribution,biological rationality,prediction robustness,parameter estimation robustness,model simplicity and prediction accuracy.Based on this,the population development model and the actuarial model of the basic social pension insurance for urban employees were constructed.In view of the uncertainty of mortality rate,VaR and CVaR were used to estimate the upper limit of the financial gap in the basic pension system and the tail risk of exceeding the upper limit,and then evaluated the longevity risks of provincial urban employee social pension programs.Finally,the impact of longevity risk on the pension system of different provinces under the gradual retirement delay policy was simulated and analyzed.The results show that the economically developed and densely populated provinces and cities face greater longevity risk in the future.The gradual delay of retirement will reduce pension longevity risk,but the effect of the policy will gradually weaken.

Key words：longevity risk;Stochastic Dynamic Mortality;pension insurance;retirement delay;regional heterogeneity