A flexible model for actuarial risks under dependence

Methods for computing risk measures, such as stop-loss premiums, tacitly assume independence of the underlying individual risks. This can lead to huge errors even when only small dependencies occur. In the present paper, a general model is developed which covers what happens in practice in a realistic way. Moreover, it is also flexible, in the sense that it allows application in practice. Accurate and transparent approximations are presented, and the results obtained are illustrated through explicit examples.