Description: This paper describes risk-factor-dependent multiple decrements and actuarial cost functions as a general compartment model, a discrete time and discrete state stochastic model of morbidity and mortality processes. For large numbers of risk factors, the model is extended to its natural continuous time and state form. From Transactions of Society of Actuaries 1991, Vol. 43.

Description: The investment risk generated by a stochastic interest rate is recognized as often the greatest risk associated with a whole life or retirement policy. Recent research has investigated the mathematics of this risk when the interest rate follows a stochastic process and the time to death is considered to be random. In this paper we focus on the use of the zero coupon bond as an investment instrument to fund a whole life policy. It is unlikely that a company will entirely fund a particular single policy with a zero coupon bond.

Description: Accurate forecasts of future health care expenditures for cohorts of elderly individuals can be especially difficult. Current techniques use average per capita costs, based on local experience, adjusted for the cohort's age, sex, institutionalization, and welfare status distribution. This paper presents a method to adjust such per capita cost estimates to reflect more accurately the likely health status of a particular cohort. The adjustment uses the cause-of-mortality profile in the local population, which is a readily available tabulation known to be highly correlated with a major proportion of total medical care expenditures for people beyond age 65. From Transactions of Society of Actuaries 1984, Vol. 36.

Description: This paper proposes the use of stochastic compartment model methods for estimating both the current direct and current indirect costs of specific cancers and for forecasting future direct and indirect costs. From Transactions of Society of Actuaries 1982, Vol. 34.

Description: This paper presents a method of testing second-order stochastic dominance of risks. This methodology allows risks to be partially ordered by value with available data - requiring only that the risks have yields distributed as one of the location-scale family of distributions and that the actuary can specify that family.