Poisson Regression Analysis for Risk Classification and Derivation of Mortality Rate Estimation in a Life Insurance Company (Case Study of a Life Insurance Company in Indonesia)

摘要

This study aims to classify risk and derive the best estimate mortality rate for life insurance company PT ABC by using a Poisson regression model. With the Poisson regression model, this study modeled the number of deaths with policy duration, underwriting treatment, extra mortality enactment, and the sum insured amount. The sample data used in this study is the death count of insured males over the age of 18 years during period 2012-2016 in life insurance company PT ABC. After modeling the baseline mortality with a Poisson regression model and then performing exploratory analysis using a Standardized Mortality Ratio (SMR) for classifying risk, the mortality rate estimation based on the risk profile of life insurance company PT ABC is obtained. This estimated mortality rate is used to construct a mortality table. The results of the study showed that mortality rates for three-year policy duration of insured males over the age of 18 years' underwritten insurance policy with sum insured less than IDR 100 million and no extra mortality are greater than mortality rates for zero-year policy duration with the same profile. By getting the mortality rate in accordance with the company's risk profile, life insurance companies are expected to be more precise in determining the required reserves and deriving premiums in line with the risk profile of the insured lives.