Ratih Kusuma Millennium, Rosita Kusumawati
This study aims to estimate possible losses that can be experienced by the company by calculating the Value at Risk (VaR) from the compound distribution of claims. The steps in constructing a compound distribution of claims are data exploration, determining the type of distribution, estimating the parameters of distribution with maximum likelihood method, testing goodness of fit of the claims data using Chi-Squared, Kolmogorov-Smirnov, and Q-Q Plots, claims distribution simulation, and calculating the VaR value at a certain level of trust. Example of numerical calculation are given using the XYZ Life Insurance Company daily claims data in 2019. The best distribution for claim frequency data is POI (1,511) with Mean Squared Error (MSE) 5.197 and for claim severity data is WEI (1.328e+07, 9.016e-01) with Mean Squared Error (MSE) 2.106e+14. And the VaR of simulation data claims for 2020 at a confidence level of 90%, 95% and 99% respectively are 55,135,149, 72,128,791 and 107,596,875. © 2022 Author(s).
Department of Mathematics, Universitas Negeri Yogyakarta, Jl Colombo No 1, Karangmalang, Depok, Sleman, Yogyakarta, Indonesia; Department of Statistics, Universitas Negeri Yogyakarta, Jl Colombo No 1, Karangmalang, Depok, Sleman, Yogyakarta, Indonesia
