<![CDATA[The distribution of health insurance claims that tend to be asymmetric and have heavy tails, requires a method that can offer more flexible and robust solutions than traditional frequentist approaches, such as the Bayesian Method. The advantage of this method lies in its ability to comprehensively account for uncertainty in parameter estimation, thereby producing a posterior distribution that can capture complex pattern of claim data. This study aims to apply the Bayesian approach with the Markov Chain Monte Carlo (MCMC) Metropolis-Hastings method in modeling health insurance claims. The claim data used are divided into outpatient and inpatient claims, with the lognormal distribution fitting proven to be the most appropriate for both types of claims. Risk estimation through Value at Risk (VaR) and Conditional Tail Expectation (CTE) using the Bayesian approach showed more moderate results compared to empirical estimates, indicating that this approach can reduce risk overestimation]]>
Copyrights © 2024
