Publication: Hierarchical Lee-Carter model estimation through data cloning applied to demographically linked countries
Loading...
Identifiers
Files
Publication date
2015-05-01
Defense date
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
Some groups of countries are connected not only economically, but also social and even demographically. This last fact can be exploited when trying to forecast the death rates of their populations. In this paper we propose a hierarchical specification of the Lee-Carter model and we assume that there is a common latent mortality factor for all of them. We introduce an estimation procedure for this kind of structures by means of a data cloning methodology. To our knowledge, this is the first time that this methodology is used in the actuarial field. It allows approximating the maximum likelihood estimates, which are not affected by the prior distributions assumed for the calculus. Finally, we apply the methodology to some France, Italy, Portugal and Spain data. The forecasts obtained using this methodology can be considered as very satisfactory.
Description
Keywords
Bayesian inference, Data cloning, Hierarchical model, Lee-Carter model, Longevity risk, Projected life tables