What it means
Reinsurance transfers part of an insurer's risk to another insurer, and pricing that transfer requires an estimate of the losses the reinsurer could pay under the treaty. Exposure rating approaches the problem through the portfolio's insured values, policy limits and other characteristics rather than relying only on its own past claims.
An insurer may have a new portfolio, a small number of large claims or insufficient years of reliable data, so its observed loss history may be a weak guide to the layer being priced, and an exposure-based method uses a broader view of comparable risks to estimate how losses could develop. The classic property approach expresses losses relative to insured values or policy amounts, because a loss of $100,000 means something different for a $120,000 building and a $10 million building.
Scaling losses to an exposure measure allows the model to connect severity patterns with the portfolio's distribution of values. An exposure curve describes how loss cost accumulates as the covered proportion of value increases; it is not simply the probability that a claim will occur, and the distinction between claim frequency and the share of total loss cost matters when allocating expected losses to insurance layers.
A reinsurance layer normally has an attachment point and a limit, and the reinsurer pays losses above the attachment up to that limit, subject to the treaty. Exposure rating estimates the share of losses expected to fall into that particular slice rather than treating every loss as fully reinsured.
Policy limits influence the estimate, because a large-loss layer receives little or no loss from policies whose limits cannot reach its attachment point, so a portfolio with more high-limit policies may have a different layer exposure even when its total premium matches another portfolio's premium. Selecting a curve requires judgment about the business being insured, since construction, occupancy, protection, location and coverage terms can change the loss pattern, and applying a curve from unrelated business can make an apparently precise calculation poorly matched to the real portfolio.
Exposure rating differs from experience rating, which begins with the portfolio's actual loss record and adjusts it for relevant changes. Exposure rating can estimate a layer with limited direct claims, but it still relies on empirical evidence or assumptions drawn from other risks.
Comparing the two approaches can reveal changing limits, weak data or an unsuitable model, so investigate a large disagreement rather than selecting whichever produces the preferred price. Expected loss is not the whole premium, because expenses, uncertainty, profit requirements and the cost of supporting the risk can affect pricing.
A reinsurer should not present a modelled loss estimate as if it were the final commercial price without those other considerations. For a non-finance manager, request the portfolio characteristics, chosen curve, attachment and limit, and the reason the model is considered suitable.
Ask how the estimate changes if insured values or severity assumptions are wrong. The model offers a structured estimate, not certainty about which claims will occur.
In practice
Real-world examples.
Example
An insurer has written a new commercial property portfolio with little claims history. A reinsurer studies the insured-value distribution and applies a relevant severity curve. The lack of prior losses is not treated as evidence that the portfolio has no risk.
Example
Two portfolios have equal premium but different policy limits. The second contains more policies large enough to reach a high reinsurance layer. Exposure rating can therefore produce different expected layer losses despite the matching premium totals.
Example
A pricing team applies a residential-property curve to industrial risks. A review identifies differences in construction and loss severity. The team tests a better-matched model instead of relying on the precision of the original spreadsheet.
Formula
Calculation
Illustrative layer payment for one loss = minimum of the layer limit and the amount by which the covered loss exceeds attachment, with no payment below attachment. A $700,000 covered loss in a $500,000 layer attaching at $300,000 produces $400,000 in that layer. This demonstrates the slice being modelled, not an exposure-curve pricing formula.Case study
Seen in the real world.
Fictional case: A reinsurer initially quotes a high layer using a broad curve and limited exposure information. The insurer supplies a detailed policy-limit distribution showing fewer policies capable of reaching the layer. The parties revise the estimate, then separately consider expenses and uncertainty before agreeing a price.
Watch out
Common mistakes.
- Assuming a low observed claim count proves that a high layer has little exposure.
- Using a loss curve without checking its fit to policy limits and risk characteristics.
- Confusing modelled expected losses with the full premium charged for reinsurance.
Questions
People also ask.
Does exposure rating require the portfolio's own claims?
It can be used when that history is insufficient, with suitable external loss evidence and exposure data.
Is an exposure curve a claim-frequency table?
Not necessarily. It can describe accumulated loss cost across proportions of insured value.
Can experience and exposure methods be compared?
Yes. Their differences can help identify data or modelling issues.
From the founder's library

Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%Related
