Consideration of Marital Status in a Mortality Model and its Application for Mortality Risk Management

As a variety of mortality risk factors have been identified by previous studies, it is desirable that these factors are reflected in mortality (longevity) risk assessments for life insurance, individual annuities, and pension plans. In this study, an extended traditional mortality table that accommodates marital status, which is one of the important risk factors in terms of mortality, is considered. A logistic regression method is used to model the mortality of groups of people with four different marital statuses–never married, married, widowed, and divorced. The analyses are based on data regarding mortality and changes of marital status in Korea. The results of the analyses that are based on the developed mortality model suggest that the information on risk factors must be reflected in an actuarial model to improve the evaluations and monitoring of risk for the portfolios of relevant insurance products.