Definition:Vulnerability function
📊 Vulnerability function is a mathematical relationship used in insurance catastrophe modeling that translates a given level of hazard intensity — such as wind speed, ground shaking, or flood depth — into an expected degree of damage for a specific type of structure or asset. Sometimes called a damage function or fragility curve, it sits at the core of the damage module within a catastrophe model, bridging the gap between the physical hazard an event produces and the financial loss an insurer can expect to pay.
🔧 Within a cat model's architecture, the vulnerability function takes inputs from the hazard module — for instance, peak gust wind speeds at a particular location — and maps them against building characteristics such as construction type, occupancy class, building height, and year of construction. The output is typically expressed as a mean damage ratio (the expected insured loss as a percentage of the total insured value) along with an associated uncertainty distribution. Reinsurers, primary carriers, and exposure management teams rely on these functions when running probable maximum loss analyses and setting risk appetite thresholds. Vendors like AIR, RMS, and CoreLogic each develop proprietary vulnerability functions calibrated against historical claims data and engineering studies, and differences among them can produce materially divergent loss estimates for the same portfolio.
🎯 Getting the vulnerability function right has outsized consequences for pricing accuracy, capital allocation, and reserve adequacy. An overly optimistic function underestimates tail risk and can leave an insurer exposed after a major catastrophe, while an overly conservative one inflates premiums and erodes competitiveness. As climate patterns shift and building codes evolve, insurers and modelers must continuously update these functions — incorporating post-event reconnaissance data, laboratory testing, and advances in structural engineering — to ensure they reflect current reality rather than outdated assumptions.
Related concepts: