Analyzing persistence of life insurance policy using survival model

Abstract

The objective of this study is to identify the risk factors those caused for increasing the persistence of life insurance policy, based on one of the wellknown insurance company in Matara. In this study, the life insurance policies which were opened during the period of 1st of January 2010 and 31st of December 2011 are considered. The data set is formed of a sample of 570 life insurance policies. Gender of the policy holder (male and female), type of policy (with 14 categories) and mode of payment (single, monthly, quarterly, half yearly, and annually) are considered as affecting factors. The method of analysis is based on statistical approach, survival analysis. Kaplan-Meier Estimator and Life Table Analysis are used as survival techniques. According to the present study, persistence of life insurance of females is likely to be higher than males. The policies those have higher survival time have the less of surviving probability and hence they have higher risk of lapsation. Some policies are identified as they have higher persistence. Single mode of payment has the constant surviving rate and policies paid half yearly, have the lowest survival rate while those paid annually have the highest survival rate. The survival model provides much more information to the management of the insurance company than what the regression model can offer.