An extension of relational methods in mortality estimations

Abstract

Actuaries and demographers have a long tradition of utilising collateral data to improve mortality estimates. Three main approaches have been used to accomplish the improvement- mortality laws, model life tables, and relational methods. The present paper introduces a regression model that incorporates all of the beneficial principles from each of these approaches. The model is demonstrated on mortality data pertaining to various groups of life insured people in Sweden.

Keywords

• applied mathematics