Analysis On Long Term Care Insurance Pricing

The 6th national census indicates that the population of the up 60-year-old is 178 million,accounting for 13.26%of the whole population.Furthermore,the population of the up 65-year-old is 119 million,about 8.87%of the whole population.According to the forecast of the United Nations,the proportion of population of the elderly person over 60 years old will increase to 30%by 2050.China,at present,is facing severe challenge of aging population,and the problem of long term care is also affecting the daily life of the elderly person.In this context,the "13th Five-Year" plan,the State Council and the Ministry of Human Resources and Social Security have shown their concern to this issue,and are prepared to solve this problem by establishing effective long term care insurance system.However,the long-term care insurance is under developed in China,and the research in this field is not enough.Thus,via both qualitative and quantitative methods,this thesis studies the suitable long term care insurance system and the pricing model in our country,and also conduct an empirical research on the transition probability and the actuarial premium of long term care insurance using the data of CHARLS.Based on the research about long term care insurance system in America,Japan and Germany,this thesis proposes some suggestions on the long term care system in China.In order to ensure the effectiveness and coverage of the system as well as the market dynamics,this paper argues that China should establish a system which contains both social and commercial long-term care insurance.Within this article,the state of long-term care is defined as 3 or above of ADLs failed or cognitive function failed.Four of the widely used models are studied,the united model,random state duration model,reduced scale model and Markov model.By comparing the models above,this paper selects Markov model for long term care insurance pricing because of more realistic assumptions and advantage of statistics and mathematics.Data of CHARLS is widely used by domestic and foreign scholars due to its quality.In this paper,the data of 2011 and 2013 is selected to calculate the probability of transition.An ordered logit model is built to find out that the significant factors of transition probability are age,gender,marital status and initial health status.After that,we calculate the two year transition probability matrix and obtain one year transition probability matrix by exponential and logarithm transformation,which is applied to the pricing of long-term care insurance.The conclusion is that advanced age,male and separation from spouse tend to cause long-term care risk.Besides,the insured are likely to transit to long-term care state(over 30%)when entering into middle and old age even they are healthy enough.By contrast,the probability of the transition from long-term care state to health state is low(about 10%).This thesis conducts empirical research on the pricing of long-term care insurance.The study assumes that insurance period is up to 80 years old,and works out premium rate table based on different age,gender,marital status and initial health state(see details in the appendix).The conclusion is in line with transition probability matrix:the insured which is male,separation from spouse or already in the long-term care state suffers higher premium rate.In general,the premium rate of long-term care insurance in China is relatively higher due to the corresponding higher long-term care risk faced by the aging population.