As flood risks grow worldwide, a well-designed insurance engaging various stakeholders becomes a vita instrument in flood risk management. This paper focuses on the design of a multi-pillar flood-loss sharing program involving partial compensation to flood vitims by the central government, the pooling of risks through a private insurance on the basis of location-specific exposures, and a contingent ex-ante credit to reinsure the liabilities. The analysis is guided by an integrated catastrophe risk management (ICRM) model consisting of GIS-based flood model and a stochastic optimization procedure with respect to location-specific risk exposures. To achieve the stability and robustness of the program towards floods with various recurrences, the ICRM uses stochatic optimization procedure, which relies on insolvency constraint and Conditional Value-at-Risk (CVaR) indicators. Two alternative ways of calculating insurance premiums are compared: the robust derived with the ICRM and the traditional average annual loss approaches. The applicability of the ICRM model is illustrated on a case-study of a larger Rotterdam area outside main flood protection system in the Netherlands. Our numerical experiments demonstrate essential advantages of the robust premiums, namely that they: (1) guarantee programs solvency under all (or a percentile) flood scenarios rather than one average event; (2) establish a tradeoff between the security of the program and the welfare of locations; (3) decrease the need for other risk transfer and risk reduction measures.