Model Stokastik Dengan Pendekatan Generalized Linear Model Untuk Mengestimasi Cadangan Klaim Incurred but Not Reported

Abstract

One of the crucial things in the insurance business is determining the amount of IBNR claim reserves. The amount of IBNR's claim reserves is uncertain so it is necessary to estimate as accurately as possible. The estimation results of IBNR's claim reserves will affect the solvency and sustainability of the company. To calculate the estimated IBNR claim reserves, several approaches are used both deterministically and stochastically. This study uses a stochastic model with the GLM approach for data that is assumed to have an ODP distribution. Besides, this study also uses 2 different methods to calculate parameter estimates in the model, namely by performing parameter transformations and using the Verbeek algorithm. This study will compare the results of the IBNR claim reserve estimation obtained using these two methods in estimating the parameters in the model. The estimation results obtained indicate that the value of the IBNR claim reserves is the same. The advantage of the Verbeek algorithm is that the resulting parameter values ​​have interpretations.