APPLICATION OF PROPORTIONAL HAZARD TRANSFORM IN PREMIUM CALCULATION

• Main Author: ERSALINA NOVYANTI (NIM: 10113004), NADYA
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/23415
• Institution: Institut Teknologi Bandung

An insurance company receives insurance premium in exchange for the risk it undertakes. There are a number of risk measures which may be used to determine the risk premium. One of the them is known as the premium-based risk measure and proportional hazard transform is one risk measure within this category. If the risk premium is determined using proportional hazard transform, for aggregate loss data modeled by compound models, then the risk premium may not be able to be calculated analytically. In this final project (tugas akhir), for aggregate loss data modeled by a compound model, the risk premium is determined using proportional hazard transfrom risk measure and the calculation is done numerically. For a data used as a case study in this final project, the aggregate loss is modeled by a compound negative-binomial model with pareto severity. A non-parametric approach may be used to determine the risk premium. In this final project, for a data used as case study, the risk premium calculated using a non-parametric approach and that obtained when the data is assumed to follow a pareto distribution, are compared. The non-parametric approach used in this final project resulted in a biased estimate. A bias-corrected estimator using a bootstrap method is then applied. It is found that, based on the data used as a case study, the non-parametric approach and the bootstrap method lead to similar results.