| China has officially entered the aging society for 20 years,and the aging of the population has brought many social problems,and the long-term care need is one of them.Many countries that have entered the aging society earlier in foreign countries have matured in medium and long-term care insurance,and China has only carried out long-term care insurance pilots for social insurance in some cities,and the development of commercial long-term care insurance is even more immature.This thesis will use data from the CHARLS from 2015 and 2018 to delve into pricing issues related to commercial long-term care insurance.This thesis first combines the general standards for the definition of long-term care status in relevant international studies with the disability standards for the elderly proposed in the "Notice on Carrying out Elderly Care Needs Assessment and Standardized Service Work" issued by China in 2019,and divides the long-term care state into a healthy state,a long-term care state,a severe long-term care state,and a death state.The group with long-term care needs will then be profiled and modeled using logistic-based regression,logistic regression with penalty,multiple adaptive splines and integrated methods,and the final best performing model is the XGBoost algorithm.XGBoost was used to sort the pre-selected portrait features by importance,and finally combined with the modeling effect,the top nine characteristics of the importance ranking were: age,whether there was medical insurance,whether there was housing,whether there was cohabitation with the spouse,gender,alcoholism history,whether home,self-assessment health,and place of residence.This was followed by modeling the transfer probability matrix and determining the long-term care insurance rate.From the 2015 and 2018 data,the approximate three-year mutual transfer probabilities between health status,long-term care status and severe long-term care status by age group and sex were calculated,which was transformed into a one-year transfer probability matrix.The main influencing factors were then used to calculate the one-year transfer probability of all ages and genders.From the perspective of policy support,this paper puts forward three policy recommendations:(i)expand the publicity of LTCI-related information,(ii)consider giving priority to the long-term care of the elderly into social insurance,and(iii)Strengthen the longterm care services related to the elderly in rural areas.From the perspective of commercial longterm care development,this thesis makes three suggestions for commercial insurance institutions:(i)actively try to issue LTCI,(ii)targeted publicity for target users,and(iii)can try to avoid moral hazard and other issues through urban pilots. |
