Forecasting Longevity Gains Using a Seemingly Unrelated Time Series Model |
| |
| Authors: | César Neves Cristiano Fernandes Álvaro Veiga |
| |
| Institution: | 1. Brazilian Insurance Supervisor (SUSEP), Rio de Janeiro ‐ RJ, Brazil CEP: 20071–900;2. Department of Statistics and Actuarial Sciences, Rio de Janeiro State University (UERJ), Pavilh?o Reitor Jo?o Lyra Filho ‐ 6° andar ‐ Sala 6019 - Bloco B, Maracan?, Rio de Janeiro ‐ RJ;3. Brazilian College of Insurance (ESNS), Centro, Rio de Janeiro, RJ;4. Department of Electrical Engineering, Pontifical Catholic University of Rio de Janeiro (PUC‐Rio), Gávea, Rio de Janeiro ‐ RJ |
| |
| Abstract: | In this paper a multivariate time series model using the seemingly unrelated time series equation (SUTSE) framework is proposed to forecast longevity gains. The proposed model is represented in state space form and uses Kalman filtering to estimate the unobservable components and fixed parameters. We apply the model both to male mortality rates in Portugal and the USA. Our results compare favorably, in terms of mean absolute percentage error, in‐sample and out‐of‐sample, to those obtained by the Lee–Carter method and some of its extensions. Copyright © 2015 John Wiley & Sons, Ltd. |
| |
| Keywords: | SUTSE model structural model longevity gain mortality rates |
|
|
