Longitudinal Analysis of Mortality Risk Factors for Actuarial Valuation

Today, people are living longer and the world population is getting older. Recent statistics indicate that a 65-year-old female in the United States is estimated to live to 88.8 years old, while a 65-year-old male to 86.6 years old. This translates to about a two-year increase in life expectancy from birth compared to that more than a decade ago. Understanding these trends and their potential impact are ever more relevant to the insurance industry. In this thesis, we emphasize the longitudinal modeling framework of pension and long term care insurance using some advanced techniques to analyze patterns and trends in longevity and to investigate potential covariates to further characterize the nature of the risk. Using data from the Health and Retirement Study (HRS), our work finds that factors that incorporate demographic, health, lifestyle, and financial information help improve model projections of mortality. We used multiple state framework to develop models for understanding the utilization of long term care. Some key findings indicate that female tends to be more vulnerable to exposure for long term care needs, and so with low educated people. Finally, motivated by the data obtained from an insurer, this thesis also examined the effect of policy termination on the survival of policyholders with life insurance contracts. We modeled the time until a policy lapses and its subsequent mortality pattern and found some evidence of mortality selection. We subsequently examined the financial cost of policy termination. The lack of available data precluded us from extending this analysis to pension plans and long term care insurance products; such can be done as further studies

long term care insurance