Glossary

Catastrophe Modeling

Also known as: cat modeling, probable maximum loss modeling

The use of probabilistic computer models — incorporating hazard data, exposure characteristics, and building vulnerability — to estimate the likelihood and financial magnitude of losses from events such as hurricanes, earthquakes, wildfires, and floods.

Insurers use catastrophe ('cat') models to price property insurance premiums and set reinsurance capacity, and increasingly lenders and institutional investors run the same class of models independently during underwriting to stress-test physical climate risk beyond what a FEMA flood map or historical loss data alone would show. Cat models output metrics such as probable maximum loss (PML) and average annual loss (AAL), which feed directly into insurance cost assumptions, reserve requirements, and in some cases loan sizing for assets in higher-hazard geographies. Because model outputs can vary meaningfully between vendors depending on the underlying climate and vulnerability assumptions, sophisticated CRE investors increasingly treat a single cat model score as one input among several rather than a definitive risk determination, particularly as insurers themselves have begun retreating from or repricing coverage in high-risk coastal and wildfire-prone markets faster than models have historically predicted.