Conditional mean risk sharing defines an allocation rule to distribute total losses among participants in an insurance pool. Under this risk-sharing scheme, the no-sabotage condition holds when conditional expectations of individual losses given their sum are comonotonic. This property has been widely studied considering independent risks, often assuming that they possess log-concave densities. This paper considers the no-sabotage condition for dependent-by-mixture risks which do not necessarily obey log-concave distributions. Sufficient conditions derived from three different approaches are proposed in order to fulfill the no-sabotage requirement. Several examples are given to illustrate the applicability of the results.
Denuit, M., Ortega Jiménez, P., & Robert, C. Y. (2024). No-sabotage under conditional mean risk sharing of dependent-by-mixture insurance losses (LIDAM Discussion Paper ISBA 2024/19). https://hdl.handle.net/2078.5/217300