For parametric MGAs

Calibration credential for your trigger-probability claims.

Your trigger-probability claims are pricing-load-bearing. When you tell a state regulator or a reinsurance capital provider that the 1-in-50 probability of a Cat-3+ landfall in a specific basin grid cell is X, the credibility of that number determines your capital efficiency. Today there is no neutral third-party credential supporting that probability. Forecast Registry provides one.

See live ratings Read methodology v1.0

What Forecast Registry provides for your business. A neutral, third-party calibration credential keyed to the same trigger types you write contracts on: cat-at-landfall, central pressure, max sustained wind at a parametric grid cell, USGS gauge exceedance, USGS ANSS earthquake magnitude. The credential is published, citable in offering materials, and audit-defensible under NAIC Model Bulletin 30.

Peril classes wired into the scoring engine

Atlantic basin tropical cyclone — IBTrACS v04r01 resolution; CSU baseline; trigger types: cat at landfall, central pressure, max sustained wind at grid cell. Retrodictive backfill complete: n=20, BSS +0.199 vs climatology.
California wildfire — CAL FIRE incident archive; NIFC baseline.
Pacific typhoon — JMA Best Track; JTWC baseline.
US severe convective — SPC Storm Reports; SPC climatology baseline.
US flood — USGS NWIS; gauge return-period baseline.
Earthquake (CA + Cascadia) — USGS ANSS; UCERF3 + Cascadia time-dependent hazard baseline.
European windstorm — PERILS Industry Loss Index; Met Office reanalysis baseline.
Drought / agricultural commodity — USDA NASS QuickStats; IRI ENSO baseline.

Use cases for your business

Capital-provider negotiations

Cite the registry's audit-defensible calibration credential when negotiating with reinsurance / ILS capital. The credential is independent of your in-house modeling, which reduces the perceived information asymmetry.

State regulatory filings

Reference the methodology and your registry-side calibration row in NAIC Bulletin 30 disclosure templates. Demonstrates third-party validation.

Cedent education

Hand cedents the registry's per-peril leaderboard to make the case that parametric triggers are scoreable in a way traditional indemnity covers are not.

API integration

Pull GET /pricing/{peril} + GET /forecaster/{id} directly into your pricing pipeline as a calibration cross-check on internal models.

What we are asking

A 50-day pilot in which your firm contributes one peril forecast set to the registry as a contributing forecaster (in addition to the reference forecaster).
Participation in the 60-day v1.0 public comment period — the comment on the trigger-types coverage shape from a working parametric MGA is exactly the input we need.
An optional advisory seat on the standards board once the entity is formed.

Reference forecaster calibration

Founding contributor's reference forecaster (Chorus Public Ledger): 1,963 SHA-locked forecasts, 82 resolved, mean Brier 0.0399 on n=18 high-conviction in-headline forward. Atlantic basin per-peril retrodictive: n=20, mean Brier 0.2002, BSS +0.199. Live ratings dashboard →

Contact

Addie Conner, founding contributor. hello@forecast-registry.org (provisional).

Last reviewed: 2026-06-05.