LEE-CARTER MODEL FOR CENTRAL MORTALITY RATE’S FORECAST

• Main Author: Septirana Putri, Amalina
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/36025
• Institution: Institut Teknologi Bandung

Insurance companies and pension funds have two biggest risks which have the relation to mortality, those are mortality risk and longevity risk. Mortality risk is any potential risk attached to the decreasing life expectancy of pensioners and policy holders, meanwhile longevity risk is any potential risk attached to the increasing life expectancy of pensioners and policy holders. Both of them may cause financial losses to the related companies. Potential losses are the result of higher pay-out ratios than expected for many insurance companies and pension funds which may be caused by the inaccurate forecasts of mortality rate that the related companies own. Lee-Carter model is known as mortality model which is easy to be applied, it also gives satisfying results to few countries. This model focuses on the log of central mortality rate’s modeling for the given observed age groups and observing years. The singular value decomposition approach is used to estimate the model’s parameters, and ARIMA is the model chosen for Lee-Carter’s forecasting method. In this thesis, the forecasting of central mortality rate will be done using Lee-Carter model with the help of singular value decomposition approach and ARIMA, in the hopes proper forecasts will be used to determine insurance premium and reserve for insurance companies and pension funds more accurately.