Mortality Index Simulation for Forecasting Malaysian Mortality Rates

Abstract

Mortality studies are very important in demography and actuarial areas because they assist policymakers and life insurers in managing longevity and mortality risks. In recent decades, many extrapolative mortality models have been developed following the Lee-Carter model. Despite the widely used Lee-Carter model for projecting mortality rates, the literature that has a thorough explanation of it is limited. In this study, we aim to provide a comprehensive explanation of the model with a focus on its fitting and simulation forecasting techniques. We fitted the mortality rates of the Malaysian population for the years 1991 to 2012 using the Lee-Carter model. We then projected the mortality rates for the years 2013 to 2018 using an autoregressive integrated moving average (ARIMA) (0,1,0) model by using a simulation of the mortality index. Findings showed that the Lee-Carter model performs well for this dataset based on the computed standard accuracy measures. The estimated age parameters exhibited a high mortality rate in the age group of 0-4 years, while the estimated time-varying parameter indicated a decreasing trend. This study presents a thorough interpretation of the Lee-Carter model and a detailed simulation of the ARIMA (0,1,0) model and hence provides a comprehensive reference for beginners in mortality studies.