A note on computing bonus-malus insurance premiums using a hierarchical Bayesian framework

Abstract

In this paper we consider statistical problems arising from applications concerning insurance-premium calculation. We describe an integrated set of Bayesian tools for modelling bonus-malus systems (BMS) for insurance premiums. This paper describes a bonus-malus system (BMS) applicable to insurance claims procedures, constructed using a hierarchical Bayesian model. We then address notions and techniques of robust Bayesian analysis in the context of problems arising in BMS.