Application of the linear hazard transform to model survival probability based on the Indonesian mortality table

• Main Authors: Permana, F. J., Sugiarto, I., Johan, Y.
• Format: Conference Proceeding
• Language: English
• Subjects: Insurance; Life insurance companies; Maximum likelihood method; Mortality; Parameter estimation; Tables (data)
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0158214

Life insurance companies use the mortality table to calculate actuarial present values of benefits and life annuities. In Indonesia, life insurance companies widely use the Indonesia Mortality Table because it can represent Indonesia’s mortality rate. However, the Indonesian Mortality Table (TMI: Tabel Mortalita Indonesia) is not launched every year due to some limitations, e.g. time and costs. The last released mortality table is used to describe mortality rate at the year between two consecutive publication years of mortality table. There are some methods to transforms the mortality rate to the next published mortality table from the previous one. This transform assumes there is linear relationship between two mortality tables obtained from two consecutive publication years. The proportional hazard transform can be considered as the special case of the linear hazard transform. This paper will discuss the linear and proportional hazard transforms by using two approaches, those are linear relationship over a year and k-year periods. By choosing the right transform, the future mortality rate can be predicted too. Transformation parameters are estimated by the maximum likelihood method. Simulation results applying to TMI II and TMI III show that the best performance is given by linear hazard transform over k-year period.