The Research Of Longevity Risk On Commercial Endowment Insurance Of China

With the development of social economy and the improvement of medical and health care,the population mortality is declining and the average life expectancy continues to increase.Many countries,including China,have stepped into an aging society.The average life of world population increased from 45.9 years in 1955 to65.4 years in 2000,which is expected to reach to 74.3 years in 2050.Moreover,the expected average life in China has reached to 74.83 years according to the sixth census population.Under this condition,the aging population continues to increase.At the same time,there is a downturn in the birth rate,with the result that an ageing population society will come.Large elderly population,as well as the developing actual life,has a huge impact on the pension payment system.Moreover,longevity risk arises.As an important part of the old-age security system in China,commercial endowment insurance,will play an increasingly important role in the future.At the same time,due to the Business insurance company's own nature of profit and loss,the risk of longevity will bring more impact on it.Therefore,this paper will study the longevity risk faced by commercial pension insurance as the object of study.This article is mainly for longevity risk measurement of commercial endowment insurance in China,and some advices are proposed based on the analysis of deficiencies of longevity risk in our country at present.Research about longevity risk measurement in our country mainly stays on the theoretical at present,which is lack of quantitative analysis.Therefore,the article adopts the quantitative analysis method for longevity risk measurement,thus,longevity risk can be obtained quantitatively.Longevity risk measurement could be divided in two steps:Firstly,the mortality is predicted.In the choice of mortality prediction model,this paper chooses the Lee-Carter model,which has the advantages of good prediction accuracy and universality.The prediction of parameters in the model plays an important role in the model application.Therefore,this paper uses the newest achievement in the research field of mortality,which is the Bayes MCMC methods for the parameter estimation of mortality model.Different from the traditional two-phase parameter estimation method,all parameters could be estimated by one-time in the unified framework using this method and the problem of limited mortality data in China is solved,as well as the quality is poor.Also,compared with conventional parameter estimation methods,it is found that the Bayes MCMC method is moresuitable for death data characteristics of the population in China.At the same time,with the elderly population in longevity risk analysis become more and more important,the Lee-Carter model could only predict mortality of0-89-year-old group based on the data in our country.Therefore,the high-age mortality prediction study is carried on in this article.The mortality prediction model for the high age population is mainly selected in the Gompertz model and the Coale-Kisker model.Through comparison and analysis,we get the conclusion that Gompertz model is more suitable for extrapolating the mortality of the elderly population in our country.Therefore,Gompertz model was used to predict the mortality of the population over 90 years old.Secondly,based on the date of all-ages population mortality obtained in the previous step,the quantitative measure of longevity risk is carried out in the two aspects respectively,which are the capital requirements of longevity risk and premiums and the difference value of profit and loss brought by longevity risk.The first innovation of this paper is to introduce the Bias MCMC method into the quantitative measurement of longevity risk,so as to achieve a more accurate measurement of longevity risk in china.The second innovation is different from previous studies using age-based mortality data,in this paper,the mortality rate of China life insurance experience life table is based on the mortality rate of China life insurance industry,empirical mortality data were selected.This choice is more in line with the characteristics of China's commercial pension insurance.The third innovation is the use of WINBUGS software for Bayesian estimation.It greatly reduces the calculation and calculation process,and facilitates the application of Bayesian method.