View source for Definition:Event set

📐 '''Event set''' is a foundational component of [[Definition:Catastrophe modeling | catastrophe models]], consisting of a large simulated catalog of individual hazard events — such as hurricanes, earthquakes, floods, or wildfires — each defined by physical parameters like location, intensity, path, and duration. In the insurance and [[Definition:Reinsurance | reinsurance]] industry, [[Definition:Catastrophe model | catastrophe model]] vendors such as Moody's RMS, Verisk, and CoreLogic generate event sets containing tens of thousands to millions of hypothetical scenarios that represent the full statistical range of what could plausibly occur over an extended time horizon. These synthetic events serve as the starting point for estimating the [[Definition:Loss | losses]] an [[Definition:Insurance carrier | insurer's]] portfolio might suffer under each scenario.

🔧 Within a catastrophe model's architecture, the event set feeds into the hazard module, which translates each event's physical characteristics into local intensity measures — wind speeds at specific coordinates, ground-shaking intensities at individual sites, flood depths at particular elevations. These intensity footprints then interact with the [[Definition:Exposure | exposure]] data (the properties, values, and construction types an insurer has on its books) and the [[Definition:Vulnerability function | vulnerability module]] to produce damage estimates. By running the full event set against a portfolio, the model generates an [[Definition:Exceedance probability curve (EP curve) | exceedance probability curve]] that tells the insurer, for example, the probability of losses exceeding $500 million in a given year. The richness and calibration of the event set directly determine the reliability of these outputs.

💡 Selecting and understanding the event set matters enormously for [[Definition:Capital management | capital management]], [[Definition:Reinsurance | reinsurance]] purchasing, and regulatory compliance. Different model vendors may produce materially different event sets for the same peril region, reflecting divergent scientific assumptions about storm frequency, fault-rupture behavior, or climate trends. Insurers and reinsurers must evaluate these differences critically — often blending outputs from multiple models — to avoid anchoring their [[Definition:Risk appetite | risk appetite]] decisions on a single view. Regulatory frameworks like [[Definition:Solvency II | Solvency II]] and [[Definition:Risk-based capital (RBC) | risk-based capital]] standards increasingly require companies to demonstrate that they understand the event sets underpinning their capital models, adding a governance dimension to what is fundamentally a scientific and actuarial exercise.

'''Related concepts'''
{{Div col|colwidth=20em}}
* [[Definition:Catastrophe modeling]]
* [[Definition:Exceedance probability curve (EP curve)]]
* [[Definition:Probable maximum loss (PML)]]
* [[Definition:Vulnerability function]]
* [[Definition:Stochastic modeling]]
* [[Definition:Aggregate exceedance probability (AEP)]]
{{Div col end}}