An evolutionary approach to fraud management

Building on several contributions to the analysis of insurance fraud, we propose a dynamical model of the fraudulence game, where three typologies of players interact: the insurance company, the fraudsters and the honest insured (who may be tempted to become dishonest), each one taking decisions on the basis of an adaptive strategy. It follows from the mathematical analysis that several scenarios and different asymptotic outcomes of the game are possible. In all cases, managerial/actuarial interpretations and implications are provided, suggesting how insurers can adapt proper control policies both to evolving behaviours of policyholders and to different external (economical, geographical, social) contexts.