Model Comparison › Catastrophe Modeling

Actuarial Extreme-Event Physical Loss Modeling

GEV-distribution catastrophe model calibrated to Swiss Re sigma database. Occurrence Exceedance Probability (OEP) curves for 7 perils, climate loading by warming scenario, portfolio-level PML, TVaR-99, and insurance pricing metrics. 2–1,000 year return periods.

Warming scenario: Asset value ($bn): Detail peril:

—

Total AAL (base climate)

Annual average loss $bn

—

Total AAL (climate loaded)

Annual average loss $bn

—

100-yr loss (portfolio)

$bn at 1-in-100 yr return

—

250-yr loss (portfolio)

$bn at 1-in-250 yr return

—

TVaR-99 (portfolio)

$bn tail expected loss

Occurrence Exceedance Probability (OEP) Curves — Industry Loss ($bn)

Loss Breakdown by Component

Return Period Analysis (Portfolio)

Sources: Swiss Re sigma database 1970–2023 (industry AAL benchmarks, GEV distribution calibration) · Munich Re NatCatSERVICE 2023 (peril frequency/severity statistics) · IPCC AR6 WG1 Chapter 11 (climate loading rates by peril per °C additional warming) · Swiss Re CatNet Research (2022) climate-adjusted hazard intensities · Embrechts, Klüppelberg & Mikosch (1997) "Modelling Extremal Events" (GEV/GPD theory). Disclaimer: Indicative model only. Commercial catastrophe models (AIR, RMS, Verisk) use full stochastic event sets with high-resolution hazard footprints. This model uses simplified aggregate GEV fits; results should not be used for actual insurance pricing or regulatory capital calculations.