论文部分内容阅读
生活方式的改变、医学的进步和遗传学的新发现都会使人的预期寿命变得不确定。本文针对中国人口死亡率历史数据(0~89岁男性数据),利用贝叶斯信息准则和嵌套模型的似然比检验等方法,比较了8种目前流行的随机死亡率模型的拟合效果;同时,检验了这8种随机死亡率模型预测结果的生物合理性和稳定性,并比较了它们的预测效果。结果表明,由Lee-Carter模型拓展而来的Age-Period-Cohort模型最适合于拟合和预测中国的人口死亡率,这为我国寿险企业和养老金机构的死亡率风险管理提供了科学依据。
Changes in lifestyles, advances in medicine, and new discoveries in genetics all make people’s life expectancy uncertain. According to the historical data of population mortality in China (male data of 0-89 years old), this paper compares the fitting effects of eight popular random mortality models with the Bayesian information criterion and the likelihood ratio test of nested models At the same time, we tested the bio-rationality and stability of the prediction results of the eight random mortality models and compared their prediction results. The results show that the Age-Period-Cohort model extended from the Lee-Carter model is the most suitable for fitting and forecasting China’s population mortality rate, which provides a scientific basis for the risk management of mortality of China’s life insurance companies and pension institutions.