FORECASTING MORTALITY RISK MEASURES BASED ON AR(1)-APARCH(1,1) MODEL

• Main Author: Sabrina, Darin
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/77311
• Institution: Institut Teknologi Bandung

Mortality is the number of deaths observed in a group of individuals in a specific area and time. The change in mortality over time is called the mortality rate. An increase in life expectancy or a downward trend in the mortality rate occurs from year to year. These conditions pose risks in the actuarial field, especially for pension funds or life insurance companies that provide life annuities. It is called a risk because it can provide losses to the company since the costs incurred to pay claims are potentially greater than what has been prepared. In this thesis, the mortality rate is theoretically modeled using a stochastic model that accommodates the effects of homoscedasticity and heteroscedasticity to model the mean dynamics and conditional variance (volatility) on mortality rates, namely the AR(1)-APARCH(1,1) model. The model was chosen after conducting an analysis of compatibility with the behavior of the mortality rate data. Furthermore, in predicting the risk of future mortality rates, the risk measures used are MaR and CMaR based on the concepts of VaR and CVaR risk measures. Eventually, a coverage probability test was carried out from the predicted results of MaR and CMaR on the basis of the AR(1)-APARCH(1,1) model