Modelling and forecasting of stochastic mortality models

Abstract

This dissertation paper sought out to model and forecast the mortality in the Maltese Islands using stochastic models. The modelling of mortality using a stochastic approach is a relatively new, but rapidly growing, area of research. In recent years, a vast literature on the subject has been developed. The focus was set on the National population data for the Maltese Islands for both males and females aged between 65 and 84 years old. The data used was obtained from the National Statistics Office of Malta. Firstly, an overview of the different mortality models proposed in the literature was provided. Then, two stochastic mortality models were modelled. Both models gave a relatively good fit to the data. Finally the models were used to forecast future mortality rates. The method used for forecasting was that of a random walk with drift. The random walk with drift is most commonly used in most mortality literature. Adequate results were obtained.