Catastrophe models estimate financial losses from natural and man-made disasters. They are used by insurers, reinsurers, financial institutions, and public agencies for underwriting, capital management, and regulation. Historical loss records are short in comparison to the low frequency of catastrophic events. Extrapolating from history can only provide very uncertain estimates of tail risk. Catastrophe models look beyond history and fill the gaps where events could happen but have not yet occurred.

To this end, cat models typically simulate thousands of years of synthetic events. I will explain how these models are built and validated, from hazard to vulnerability to financial loss. I will focus on weather and climate models and show how physical models, historical observations, and statistical methods are used to build catastrophe models. I will highlight how academic research is used and translates into operational models.

Farid Ait-Chaalal, Moody’s Analytics.

Lieu
École normale supérieure – PSL
24 rue Lhomond – aile Erasme
salle Claude Froidevaux – E314