CONJUGATE PRIOR PARETO DAN APLIKASINYA DALAM PERHITUNGAN PREMI UNTUK KLAIM BESAR PADA REASURANSI

• Main Author: RAHMANIA, ATSILA
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/19939
• Institution: Institut Teknologi Bandung

An insurance company may insure itself to a reinsurance company. There is a possibility that the claims reported to an insurance company are large claims. For an excess of loss (XL) reinsurance contract, the amount of coverage paid by the reinsurance company to the insurance company is the excess of the claims severity from a speci?ed threshold. By determining a su?ciently large threshold, the excesses of claims severity could be modelled by a Generalized Pareto distribution. This ?nal project (skripsi) attempted to discuss the topic in a scienti?c paper by Vilar-Zanon and Lozano-Colomer (2007). A distribution from the Generalized Inverse Gaussian (GIG) family distribution is chosen as a prior distribution for the shape parameter of a Pareto distribution. It turned out that the prior is a conjugate prior for the Pareto distribution. The estimation of parameters was carried out using Bayesian approach by applying the mean squared error as the loss function. In this ?nal project, the data used consists of 10 portfolios found in Vilar-Zanon and LozanoColomer (2007). The resulted Bayesian estimator for the parameters were used to determine the pure premium in an excess of loss reinsurance contracts (XL).