Abstract:
|
Unearned premium, or more particularly the risk associated to it, has gained in popularity over the last decade or so. Unearned losses occur after the valuation date for policies written before the valuation date. The risk linked to these losses is that premium acquisition is assumed uniform when it is not in fact uniform, which creates a discrepancy in the proportion of outstanding risk linked to unearned premium. In this talk, we will introduce an individual parametric approach allowing for modeling of losses linked to unearned premium and we evaluate the risk arising from misevaluating the acquisition pattern of premium. We show a proof of concept through simulation to illustrate the modeling accuracy gained by isolating losses linked to unearned premium and we provide real data examples. Finally, we discuss how we could include dependency between claims from the same policy.
|