Skip to content

Rating & pricing

How the premium is actually computed — and where the schedule-rating credit (the cleanest Glacis hook) lives. All STABLE.

Premium = Rate × Exposure units

  • The rate is the price per unit of exposure.
  • Exposure units measure the amount of risk (per $1,000 of property value, per $100 of payroll for workers comp, per vehicle, per employee, etc.).
Method How it sets price Best for
Manual / class rating Look up the rate in a published table by class. Large, homogeneous groups (auto, homeowners).
Experience rating Adjust the manual rate up/down by the insured’s own past loss history. Mid/large commercial accounts (e.g., the workers-comp experience mod).
Schedule rating Apply debits and credits for specific risk characteristics of this account. Commercial risks with notable individual features.
Retrospective rating Final premium trues up to actual losses during the period (within a min/max). Large accounts willing to share risk.
Judgment rating Underwriter prices from expertise when there’s no credible loss data. Novel/unique risks — e.g., emerging AI exposure.

Schedule rating lets the underwriter apply credits (discounts) or debits (surcharges) for specific, identifiable characteristics of an individual risk — better management, safety programs, loss-control measures, etc. STABLE

This is the cleanest single bridge from “Glacis software” to “lower premium for the customer / better book for the MGA”:

An AI vendor demonstrates strong governance via Glacis evidence (attested human-in-the-loop, bias testing, model inventory) The underwriter applies a SCHEDULE-RATING CREDIT (a documented, defensible reason to charge less) Better-controlled risks pay less AND are less likely to have a loss → lower LOSS RATIO → higher MGA profit share (page 03, 15)

The customer gets a lower price for a real reason; the MGA gets a better-selected, lower-loss book. Both sides win because the credit reflects genuinely reduced risk.

Judgment rating = the right tool for a novel AI risk

Section titled “Judgment rating = the right tool for a novel AI risk”

Because AI/healthcare Tech E&O has no credible loss history, early pricing leans on judgment rating (and the E&S market, which permits freedom of rate/form). Over time, as the book generates loss data, pricing can shift toward experience/class rating — and Glacis’s accumulated telemetry makes that data richer. See 01 The flywheel.

Rate regulation (concept; specifics VERIFY)

Section titled “Rate regulation (concept; specifics VERIFY)”

Admitted carriers must generally file rates with the regulator, which polices that rates are not inadequate, not excessive, and not unfairly discriminatory. E&S (non-admitted) carriers enjoy freedom of rate and form (one reason novel risk goes there). STABLE

03 Loss ratio & combined ratio

Drill this page →13 bank questions stand behind what you just read. Check it while it’s warm.