Category: Pricing & rating · Reviewed by Apex Insurance Brokers · Last reviewed
Catastrophe model
A catastrophe model (or “cat model”) is a computational framework that estimates the probability distribution of losses to insured exposures from natural or man-made catastrophic events. The output drives pricing, capital allocation and risk transfer for property, marine, energy and specialty lines.
Core components
Stochastic event set — tens of thousands of physically plausible hypothetical events with annual frequencies, calibrated to historical and scientific data.
Hazard module — for each event, modelled physical intensity (wind speed, ground motion, water depth) by location.
Exposure module — geocoded portfolio of insured locations with values, occupancy, construction, year built.
Vulnerability module — damage functions translating hazard intensity to damage ratio by exposure type.
Financial module — applies policy and treaty terms (deductibles, limits, sub-limits, coinsurance) to gross damage to produce net loss.
Output metrics
Average Annual Loss (AAL) — expected loss per year.
Exceedance Probability (EP) curve — return-period losses.
Occurrence Exceedance Probability (OEP) — largest single-event loss at each return period.
Aggregate Exceedance Probability (AEP) — total annual loss at each return period.
Tail Value at Risk (TVaR / TCE) — average loss in the tail beyond a quantile.
Solvency II internal models and the SCR standard formula (for cat risk) reference vendor model output, although the firm remains responsible for model validation, suitability, parameter overrides and view-of-risk adjustments. The PRA published SS18/13 Solvency II: Catastrophe Risk on supervisory expectations.
References
Grossi, P. and Kunreuther, H. (2005). Catastrophe Modeling: A New Approach to Managing Risk. Springer.
