Why Loss History Is the Biggest Lever in a Renewal
When a commercial insurance renewal is priced, the largest single input is almost always the account's own loss history. Capacity conditions, the rating cycle and the broker's negotiation all move the number, but the foundation underneath the premium is the underwriter's estimate of what the account is likely to cost in claims, and that estimate is built largely from what the account has cost before. Understanding how the underwriter turns loss history into a premium is therefore the most useful thing a broker or buyer can do, because it identifies where the real bargaining power in the negotiation sits.
The core idea is the loss cost: the expected value of claims the account will generate, before the insurer adds its expenses, the cost of capital, and its profit margin. The premium is, at its simplest, the loss cost loaded for expenses and profit and adjusted for the market environment. Of these components, the loss cost is both the largest and the one most specific to the individual account, which is why it is where the pricing argument is won or lost. An account that can credibly demonstrate a lower expected loss cost than the underwriter's first estimate has the strongest possible basis for a lower premium, because it is arguing about the biggest number on a foundation the underwriter respects.
There are two broad ways an underwriter estimates the loss cost, and most commercial pricing uses a blend of both. Experience rating builds the loss cost from the account's own past claims, on the logic that the account's history is the best predictor of its future. Exposure rating builds the loss cost from the account's risk characteristics (its size, activity, location, and the hazard of its operations) benchmarked against the broader portfolio, on the logic that accounts with similar characteristics produce similar losses. Each has strengths and weaknesses, and the art of pricing is in weighting them appropriately for the account in question.
This post sets out the mechanics: burning-cost pricing as the basic experience-rating method, the difference between experience and exposure rating and when each dominates, loss development and why recent years cannot be taken at face value, credibility weighting as the formal way of blending an account's own experience with the wider book, and how a broker or buyer uses all of this to negotiate. The aim is to make the underwriter's loss-cost model legible, so that the buy side can engage with the actual basis of the premium rather than arguing around it.
Burning Cost: The Basic Experience-Rating Method
Burning cost is the simplest and most widely used experience-rating method, and it is the natural starting point for understanding how an account's own losses translate into a premium. At its core it is a historical average: what the account has actually cost in claims, expressed as a rate.
How burning cost is calculated
The burning cost is the account's historical losses over a defined period divided by a measure of its exposure over the same period, giving a loss rate. The losses are the claims the account incurred (paid plus outstanding reserves) over, typically, the last several years; the exposure base is a measure of the account's size that scales with risk (turnover, wage roll, sum insured, vehicle count, or another appropriate driver depending on the line). Dividing one by the other gives the burning cost: the historical claims cost per unit of exposure. Loading that loss rate for the insurer's expenses and profit, and applying it to the exposure projected for the coming period, produces an experience-based premium. The logic is direct: this is what the account has cost per unit of exposure, so this, adjusted, is what it should pay.
Why raw burning cost is not the answer
A naive burning cost (raw historical losses over raw exposure) is rarely the final number, because several adjustments stand between the historical experience and a fair forward-looking rate. Past losses must be brought to current values, because a claim from several years ago, in today's terms, costs more than it did then; inflation in claims costs (medical, repair, rebuild, legal) means historical losses understate what the same event would cost now. Recent years' losses are not yet fully developed, because claims take time to be reported and to settle, so the most recent years understate the eventual cost. Large or unusual losses may need to be capped or smoothed so a single freak claim does not distort the rate. And the exposure base itself must be projected forward to the period being priced. Each of these adjustments is a place where the experience-rated number can move, and each is a place where a broker who understands the mechanics can engage.
The number of years and the exposure base
Two choices shape the burning cost materially. The first is the number of years of experience used: too few years and the rate is volatile and dominated by chance; too many and it is dragged by old experience that no longer reflects the account as it is now. The choice trades stability against relevance, and the right number depends on the line and the account's volume of claims. The second is the exposure base: it should be the measure that genuinely drives the account's losses, so that the rate scales correctly as the account grows or shrinks. A mismatch between the exposure base and the true loss driver distorts the rate, especially when the account changes size. Both choices are legitimate subjects for discussion between underwriter and broker, because both affect the resulting premium.
Experience Rating Versus Exposure Rating
Burning cost is experience rating: it prices the account from its own past. The alternative is exposure rating, which prices the account from its characteristics benchmarked against the wider market. Most real pricing blends the two, and understanding when each dominates is central to understanding how any given account is rated.
What exposure rating does
Exposure rating estimates the loss cost not from the account's own claims but from its risk characteristics: its size, the nature and hazard of its activities, its locations and their exposures, its protections, and how these compare with the insurer's book of similar risks. The insurer holds, implicitly or explicitly, a view of what an account with a given set of characteristics should cost, derived from the portfolio, and applies that view to the individual account. Exposure rating is how an underwriter prices an account with little or no relevant loss history (a new operation, a new line for the account, a risk too infrequent to have produced claims) and how the underwriter sanity-checks an experience-based number against what the account's characteristics suggest it ought to cost.
When each method dominates
The weight between experience and exposure rating depends mainly on how much credible loss experience the account has. A large account with high-frequency losses (a big fleet, a large workforce for an employers'-liability or workers-compensation-type exposure, a large property schedule with regular attritional losses) generates enough claims that its own experience is a reliable predictor, and experience rating dominates. A small account, a new risk, or a low-frequency high-severity exposure (where years can pass with no claim and then a single large loss occurs) has little credible experience, so exposure rating dominates, because the account's own quiet history says little about its true exposure to the large loss. Most commercial accounts sit between these poles, and the underwriter blends the two methods according to how credible the account's own experience is, which is exactly what credibility weighting formalises.
Why the buy side must understand both
A broker or buyer who understands which method is driving the price knows where the argument lies. If the account is being experience-rated, the discussion is about the loss history: its development, the treatment of large losses, the inflation adjustment, the exposure base, and whether recent risk improvements not yet reflected in the data should lower the forward estimate. If the account is being exposure-rated (because its own experience is thin), the discussion is about its characteristics and benchmarking: whether the underwriter's view of accounts like this one is fair, whether the account's specific features make it better than the benchmark, and whether the hazard assessment is correct. Arguing loss history on an exposure-rated account, or benchmarking on a heavily experience-rated one, misses where the price is actually being set. Diagnosing which method dominates is the first step in a productive renewal discussion.
Loss Development and Credibility Weighting
Two technical refinements turn a crude burning cost into a defensible loss-cost estimate, and both are places where the buy side can engage substantively. Loss development corrects the fact that recent claims are not yet fully known; credibility weighting blends the account's own experience with the wider book in proportion to how much the account's experience can be trusted.
Loss development
Claims are not fully known when they happen. A claim takes time to be reported, and once reported it takes time to settle, with the reserve estimate moving as more is learned. The consequence is that the most recent years of an account's loss experience are immature: they show less than the claims will eventually cost, because not all claims are reported yet and the open ones are not yet fully reserved or settled. Taking recent years at their current reported value understates the true loss cost, sometimes substantially, especially for long-tail lines like liability where claims develop over years. Underwriters apply loss-development factors to project immature years to their ultimate expected cost, based on how claims for that line have historically developed over time. For the buy side, loss development cuts both ways: an underwriter may load recent years for development that the buyer believes is overstated for this account, and a buyer may need to accept that a recent clean year is not as clean as it looks because its claims have not yet emerged. Engaging with the development assumptions, rather than the raw figures, is where a sophisticated discussion happens.
Credibility weighting
The account's own experience is rarely either fully reliable or wholly irrelevant; it is partially credible, and credibility weighting is the formal method for blending it with the wider book in proportion to that credibility. The estimated loss cost is a weighted average of the account's own experience-based rate and the exposure-based rate from the broader book, with the weight (the credibility factor, between zero and one) reflecting how much the account's own experience can be trusted. A large account with abundant, stable claims experience gets high credibility, so its own experience drives the rate. A small account with thin or volatile experience gets low credibility, so the book's exposure-based rate dominates and the account's own noisy history is given little weight. Credibility weighting is what stops a small account's lucky quiet spell from producing an unsustainably low rate, and what stops one bad year from permanently penalising an account whose underlying risk has not changed.
Why these refinements matter to the buyer
Loss development and credibility weighting are not just actuarial machinery; they determine how much the account's recent record actually affects its price. A buyer who has had a good recent run wants that experience to carry weight, which it will only do if the account is large enough to be credible and the recent years are not heavily loaded for development. A buyer who has had a bad year wants that year given no more than its due weight, which credibility and the treatment of large losses govern. Understanding these mechanics lets the buy side argue the right points: whether the development assumptions are reasonable for this account, whether the credibility given to the account's own experience is appropriate to its size and stability, and whether a single large loss is being allowed to distort a rate it should not.
Using the Loss-Cost Model in the Negotiation
Once the buy side understands how the underwriter builds the loss cost, the renewal negotiation changes character. Instead of arguing about the headline price, the broker and buyer engage with the components of the loss-cost model, which is where the largest and most defensible movements are found. This is the practical payoff of understanding the mechanics.
Engaging with the model, not the number
The most effective renewal argument is not a plea for a discount; it is a reasoned challenge to the inputs of the loss-cost estimate. Is the loss history being developed reasonably for this account, or are recent years loaded for development the buyer can show is unlikely to emerge? Are large or one-off losses being capped and smoothed, so that a single non-recurring event does not permanently inflate the rate? Is the inflation adjustment reasonable? Is the exposure base the right driver, and is it projected forward fairly? Is the credibility given to the account's own (favourable) experience appropriate to its size? Each of these is a specific, evidence-based point that engages the underwriter on the actual basis of the price, and each can move the number more than a generic negotiation ever will, because each addresses the largest component of the premium.
Bringing forward what the data does not yet show
The loss history is backward-looking, and one of the buyer's strongest levers is evidence of risk improvement that the historical data does not yet reflect. Loss-control investment, a loss-adjuster or surveyor's recommendations implemented, new protections, process changes, or the removal of the cause of past losses are all reasons the account's future loss cost should be lower than its past experience suggests. The underwriter prices from history, but a credible, documented case that the risk has genuinely changed gives a basis for a forward estimate below the raw experience. The discipline is to present this concretely (what changed, when, and why it reduces the loss cost) rather than asserting that the account has improved. A well-evidenced risk-improvement story is the buyer's route to a price that reflects the account as it is now, not as it was.
The data quality lever
Underwriters price uncertainty into the loss cost: where the loss data is incomplete, ambiguous, or poorly presented, the underwriter assumes the worse interpretation and loads accordingly. Clean, complete, well-structured loss data (full claims listings with dates, status, paid and outstanding amounts, and cause) lets the underwriter price the account on what actually happened rather than on a cautious assumption, and it removes the uncertainty loading that poor data invites. Improving the quality and presentation of the account's loss data is one of the most reliable ways to improve the price, because it directly reduces the uncertainty the underwriter is pricing. A broker who presents the account's experience cleanly, with large losses explained and development addressed, is doing real pricing work, not just administration.
Where the model meets the market
The loss-cost model sets the technical floor, but the final premium also reflects the market environment, the insurer's appetite, and the competition for the account. Understanding the loss-cost model lets the buy side separate the two: how much of the price is the account's own technical cost, which the buyer influences through its risk and its data, and how much is market loading, which the buyer influences through competition and the structure of the placement. A buyer who knows its account's defensible loss cost can tell whether a quote is technically fair or carrying market loading, and can direct the negotiation accordingly, pressing on the loss-cost inputs where the technical number looks high and on the market loading where the technical number is fair but the price is not. That separation, made possible by understanding the rating mechanics, is what turns a renewal from a haggle into a reasoned negotiation.
Negotiating on the loss-cost model depends on understanding exactly how each insurer's wording defines the covered losses, structures deductible and large-loss treatment, and feeds the experience that drives the rate. Sarvada gives commercial insurance brokers structured, searchable access to insurer policy wordings, so the broker can compare how different insurers define and price the cover behind the loss experience and build a renewal argument grounded in the actual basis of the premium. Request Access to evaluate how structured wording access supports loss-cost-driven renewal negotiation.
