OpVaR

Functions for computing the value-at-risk in compound Poisson models. The implementation comprises functions for modeling loss frequencies and loss severities with plain, mixed (Frigessi et al. (2012) doi:10.1023/A:1024072610684) or spliced distributions using Maximum Likelihood estimation and Bayesian approaches (Ergashev et al. (2013) doi:10.21314/JOP.2013.131). In particular, the parametrization of tail distributions includes the fitting of Tukey-type distributions (Kuo and Headrick (2014) doi:10.1155/2014/645823). Furthermore, the package contains the modeling of bivariate dependencies between loss severities and frequencies, Monte Carlo simulation for total loss estimation as well as a closed-form approximation based on Degen (2010) doi:10.21314/JOP.2010.084 to determine the value-at-risk.