We propose a model to select the optimal portfolio which underlies insurance policies with a guarantee. The objective function is defined in order to minimise the conditional value-at-risk (CVaR) of the distribution of the losses with respect to a target return. We add operational and regulatory constraints to make the model as flexible as possible when used for real applications. We show that the integration of the asset and liability side yields superior performances with respect to naive fixed-mix portfolios and asset based strategies.We validate the model on out-of-sample scenarios and provide insights on policy design.
|Publication status||Published - 2004|