A simulation model for calculating solvency capital requirements for non-life insurance risk

To stay solvent, an insurer must have enough assets to cover its liabilities towards its policy holders. In this paper, we construct a simulation model that is able to generate solvency capital requirements (SCR) for non-life insurance risk. The only input to the model is assumptions about the distributions of payment patterns and ultimate claim amounts. These assumptions should ideally be based on findings in empirical data studies. We illustrate the modelling technique by considering a specific case with motor insurance data from the Swedish insurer Folksam. The SCR values generated by the simulation model with different distributional assumptions in this specific case are analysed and compared to the SCR value calculated using the Solvency II standard model. The most important finding was that the uncertainty in prediction of the trend in ultimate claim amounts affect the SCR substantially. Insurers and supervisory authorities should be aware of the effects of this trend prediction uncertainty when building and evaluating internal models in the Solvency II or other regulatory frameworks