Assessing the Performance of the Discrete Generalised Pareto Distribution in Modelling Non-Life Insurance Claims

The generalised Pareto distribution (GPD) offers a family of probability spaces which support threshold exceedances and is thus suitable for modelling high-end actuarial risks. Nonetheless, its distributional continuity presents critical limitation in characterising data discrete forms. Discretising the GPD, therefore, yields derived accommodates count while maintaining essential tail properties GPD. In this paper, we model non-life insurance claims under three-parameter (DGP) distribution. Data study on reported settled claims, spanning period 2012–2016, were obtained from National Insurance Commission, Ghana. maximum likelihood estimation (MLE) principle was adopted fitting DGP to yearly aggregated data. involved two steps. First, propose modification μ id="M2"> + 1 frequency method literature. proposal provides an alternative routine generating initial estimators MLE, cases varied intervals, as characteristic claim study. Second, bootstrap algorithm implemented obtain standard errors parameters. performance compared negative binomial using Akaike Bayesian information criteria. results show that appropriate better fit regulatory considered.