Corrective factors for longevity projections in a dynamic context

Abstract

In light of the fact that forecasting longevity by updating data, dynamically in time, holds per se the capacity to improve the longevity projection itself with respect to the real phenomenon’s behavior, our interest is focused on the analysis and, specifically, on the measurement of the improvement effect resulting from implementing the dynamic forecasting procedure other than the static one. Actually, the study presented in this paper leads to the detection of corrective factors, of simple use and interpretation, allowing to increase the predictive accuracy of the static forecast. The proposed corrective methodology is particularly suitable in the context of actuarial issues, especially those sensitive to the impact of dynamic factors such as longevity. For this reason, the numerical applications are based on the Cairns-Blake-Dowd model and on the old ages .