Simulation of Resilience of an Insurance System to Flood Risk

The aim of this communication is to discuss the benefit of agent-based modelling to characterize the resilience of French insurance system to flood risk to extreme events. We first characterize the ”Cat Nat” system, the French system to manage recovery costs after major natural disasters. We then described the CatNatABM model. The main purpose of this model is to explore feedbacks between the dynamics of households settlement and the viability of ”Cat Nat” system. This behavior based model is meant “as simple as suitable” to capture main feedbacks of the system, with the purpose of enriching it gradually. Outputs of the model are dynamic, which allows to analyse the influence of eligible realisation of flood chronicles. In this paper, we focus on comparison between a benchmark state with no ”Cat Nat” system and a insurance state with a ”Cat Nat” system without regulation, to explore the conditions under which perverse effects of insurance may occur.