Imprecise inference for warranty contract analysis

This paper presents an investigation into generalised Bayesian analysis of warranty contracts, using sets of prior distributions within the theory of imprecise probability. Explicit expressions are derived for optimal lower and upper bounds for the expected profit for the manufacturer of a product, corresponding to an imprecise negative binomial model for which two sets of prior distributions are studied. The results can be used to set a maximum value of compensation such that the manufacturer׳s expected profit remains positive, under vague prior knowledge.