What is a loss ratio in dealer reinsurance, and what is a good one?
A loss ratio measures claims against premium. In a dealer reinsurance program it is usually stated as incurred losses divided by earned premium for a defined period, expressed as a percentage. There is no universal target, because a loss ratio is only meaningful next to the pricing it was built on: a product priced with a wide margin and a product priced thin can produce the same underwriting result at very different loss ratios. What matters more than the level is whether the ratio is calculated on earned rather than written premium, whether it is grouped by underwriting year, whether it includes reserves for claims not yet reported, and whether it is moving in a direction the underlying data explains.
The loss ratio is the number dealers ask about first and the one most often read incorrectly. Three mistakes account for most of the misreadings: comparing claims against written premium instead of earned premium, mixing underwriting years together so that immature years dilute mature ones, and treating a ratio that excludes unreported claims as though it were final. This guide defines the measure precisely, shows how the same book can present three different loss ratios that are all arithmetically correct, explains why early ratios in a new program are systematically misleading, and sets out what to examine before concluding that a ratio is good or bad. It does not publish benchmark loss ratios, because a benchmark detached from the pricing, product mix, term length and reserve basis behind it is more likely to mislead than to inform.
- Loss ratio is incurred losses divided by earned premium. Substituting written premium for earned premium understates the ratio, sometimes dramatically, in a growing book.
- Incurred losses should include reserves for claims already reported but not yet paid, and an estimate for claims incurred but not yet reported. A paid-only ratio is not a loss ratio.
- Group by underwriting year. Blending years lets a large immature year mask a deteriorating mature one.
- A new program almost always shows a flattering early loss ratio. That is a timing artifact, not performance.
- There is no universal good loss ratio. The same ratio can be excellent on one product and unsustainable on another depending on how each was priced.
- A ratio is a symptom. Frequency, severity, product mix, eligibility, term length and adjudication are the things that actually moved it.
- The loss ratio is not the dealer's return. Fees, premium tax, reserve timing and investment income all sit between the ratio and the result.
The Definition, Precisely
In its standard form the loss ratio is:
Incurred losses divided by earned premium, for a stated period and a stated group of business. Incurred losses means claims paid, plus the change in reserves for reported claims not yet paid, plus the change in the estimate for claims incurred but not yet reported.
Every element of that sentence is doing work. The period must be stated or the number cannot be reproduced. The group of business must be stated or two ratios are not comparable. Earned premium, not written. And incurred, not paid.
Programs also report related measures. An expense ratio compares operating costs to earned premium. A combined ratio adds the loss ratio and the expense ratio, and is the more complete picture of whether the business is profitable before investment income. A dealer looking only at the loss ratio can be looking at a book that is losing money on an all-in basis, or making money on one that appears mediocre. Ask for all three.
Earned Premium Is Not Written Premium
A vehicle service contract sold today may run for several years. The premium is written now, but it is earned across the life of the contract as the coverage period elapses, and the unearned portion sits as a liability until it does. This is the single most consequential distinction in reading reinsurance results.
Consider what happens in a growing book. New contracts add written premium immediately and add claims only slowly, because claims arrive over the life of the contract rather than at the point of sale. A ratio calculated as claims over written premium in a growing book will look excellent, and will keep looking excellent for exactly as long as the book keeps growing. When production flattens, the same ratio can deteriorate sharply with no change whatsoever in the underlying quality of the business.
This is not a hypothetical failure mode. It is the most common reason a program that appeared to be performing well for several years appears to deteriorate suddenly. Nothing deteriorated. The denominator stopped growing.
| Basis | What it divides by | What it counts as loss | Typical distortion |
|---|---|---|---|
| Paid over written | Written premium | Claims paid to date only | Flatters most; understates on both sides of the fraction |
| Paid over earned | Earned premium | Claims paid to date only | Flatters; ignores claims reported but unpaid and unreported |
| Incurred over earned | Earned premium | Paid, plus case reserves, plus IBNR | The standard basis; still an estimate, and revised as claims develop |
All three can be presented in good faith. Only the third answers the question a dealer is actually asking. When a statement shows a loss ratio without stating the basis, the basis is the first thing to establish.
Claims That Have Happened but Have Not Arrived
Some claims have occurred and not yet been reported. A component failed last week and the customer has not yet brought the vehicle in. These are incurred but not reported claims, and the reserve estimate for them is why an honest loss ratio is always partly an estimate.
That estimate is not guesswork, but it is a judgment, and it can be set optimistically or conservatively. Two consequences follow. First, a program that carries no such reserve at all is reporting a loss ratio that will rise later by construction, regardless of how the business performs. Second, when a loss ratio moves between reports, the first question is whether claims changed or whether the reserve estimate changed. Those are entirely different events and they call for entirely different responses.
This is one of the places where a dealer is entitled to ask a direct question and expect a direct answer: what reserving basis is used, who sets it, and has it changed during the period being reported?
Group by Underwriting Year
An underwriting year groups contracts by when they were written, and follows that cohort as its premium earns and its claims develop. It is the only grouping that lets a dealer see whether business written under a given set of rates, products and eligibility rules is performing.
Reporting that blends all years into one figure destroys that visibility. A large recent year, which has earned little premium and developed few claims, will pull a blended ratio down and can hide a mature year that is developing badly. The blended number is not wrong, it is simply answering a question nobody asked.
A useful program report shows each underwriting year as a row, with earned premium, incurred losses and the resulting ratio, updated each period so that a dealer can watch each cohort mature. When a cohort's ratio moves, the change is attributable to a specific set of business rather than to the book as a whole.
Why a New Program's First Numbers Flatter
In the earliest period of a program, very little premium has earned and very few claims have been reported, and many products have a waiting period or overlap with a manufacturer warranty during which claims are unlikely. The result is a loss ratio that can look extraordinarily good and means almost nothing.
Claims frequency on most vehicle service business tends to build as vehicles age and mileage accumulates, and the reporting lag adds further delay. A cohort therefore typically looks better than it is early, and settles toward its true level over several years. Judging a program on its first year, or accepting a proposal built on a first-year figure, is judging a race at the first turn.
The practical rule is to require multiple mature underwriting years before drawing conclusions, and to treat the most recent year as informational rather than evaluative. This applies with particular force to seasonal books such as powersports, where written premium is concentrated in part of the year and single-season readings are especially unstable. See powersports dealer reinsurance for how that plays out.
What Actually Moves a Loss Ratio
Treating the ratio as the problem leads to the wrong interventions. It is an output. The inputs worth examining, roughly in the order they are usually productive:
| Driver | What to look at | What it would look like |
|---|---|---|
| Frequency | Claims per contract in force | More claims, similar cost each |
| Severity | Average cost per claim | Similar claim count, higher cost each; parts and labor inflation is a common cause |
| Product mix | Share of premium by product | Ratio moves without either frequency or severity changing within any single product |
| Vehicle and term mix | Age, mileage, term length at sale | Older or higher-mileage units, or longer terms, entering the book |
| Pricing | Rates relative to expected cost | Ratio elevated consistently across cohorts rather than in one |
| Eligibility and coverage | What is covered and what is excluded | Change coincides with a coverage or eligibility revision |
| Adjudication | Authorization and denial patterns | Change coincides with an administrator or process change, not with the book |
| Reserving | Case reserve and IBNR basis | Ratio moves on a prior period as well as the current one |
| Cancellations | Cancellation rate and refund timing | Earned premium reduced without a matching reduction in claims already incurred |
The last row in that table, reserving, is worth a moment. If a ratio changes for a period that has already been reported, the cause is almost always a reserve revision rather than new claims, and that is a reporting question rather than an operating one.
Reading a Loss Ratio in Ten Minutes
- Establish the basis. Incurred over earned, or something else? If it is not stated, ask before reading further.
- Check the grouping. By underwriting year, or blended? Blended figures go back for regrouping.
- Confirm reserves are included. Paid-only figures will rise later by construction.
- Read each cohort separately. Look for the mature years first; treat the newest as provisional.
- Compare to the pricing assumption. The relevant benchmark is the loss ratio the product was priced to support, not an industry figure.
- Split frequency from severity. They call for different responses; a single ratio conceals which one moved.
- Look at product mix before concluding. A mix shift can move a blended ratio while every product is stable.
- Ask what changed operationally. Administrator, coverage, eligibility, rates, staffing or the type of unit being sold.
- Get the expense and combined ratios. The loss ratio alone does not say whether the book is profitable.
- Note whether a prior period moved. If it did, the reserving basis is part of the story.
The Loss Ratio Is Not the Return
A favourable loss ratio does not translate directly into money available to the dealer. Between the two sit the costs covered in the fee guide, premium tax where applicable, the timing of reserve releases as premium earns, investment results on reserves held, and the tax treatment of the company itself. A program can have a good loss ratio and a disappointing distribution because the fee stack consumed the margin, or because the profit is real but is still held as reserve against contracts that have years left to run.
That second case is worth stating plainly, because it produces genuine frustration. Underwriting profit on a long-term product is earned slowly by design. Money that is not yet distributable is not money that has been lost, and a dealer who expects a mature-book distribution from a young book will be disappointed by a program that is working exactly as intended. See dealer reinsurance and long-term dealership value for how that accumulation actually behaves.
Loss ratio targets circulate widely in the F&I industry, usually without the pricing basis, product mix, term profile, reserving convention or underwriting-year maturity that would make them interpretable. A dealer measuring a book against a borrowed figure can conclude that a well-priced program is failing, or that a thinly priced one is succeeding. The comparison that carries information is against the loss ratio the products were priced to support, which the administrator or carrier can state, and against the dealership's own prior cohorts on a consistent basis.
This is general education on how loss ratios are constructed and read in dealer reinsurance programs. It is not actuarial advice, and reserve adequacy, reserving methodology and the sufficiency of any specific estimate are matters for a qualified actuary and the program's auditors. It is not tax, legal or investment advice.