Where each premium dollar goes: claims, commissions and profit

About three-quarters of each dollar of private passenger auto premium goes to claims and the cost of handling them. The rest pays for selling the policy, running the company, taxes and fees, and whatever is left over as profit. In the NAIC's report on 2024 results, the industry's private passenger auto lines spent about 66 cents of each premium dollar on losses and about 9 cents on adjusting them, and kept an underwriting profit of about 4 cents before counting investment income.

Those are national averages for one year. Your own insurer's split will differ, and so will the split in any one state or year. This article explains each piece, shows the industry figures with their sources, works through an example with invented numbers, and spells out why comparing one company's ratio with another's is harder than it looks.

The five buckets a premium fills

Actuaries describe a rate with one equation: premium covers expected losses, loss adjustment expenses, underwriting expenses and a profit provision. The Casualty Actuarial Society's ratemaking text, Basic Ratemaking (fifth edition, 2016), states the equation exactly that way, and it is the logic behind every filed auto rate. The pieces are:

• Losses. Money paid to or on behalf of people with covered claims: repair bills, medical payments, settlements, stolen-car payouts. This is the biggest piece by far.
• Loss adjustment expense (LAE). The cost of handling claims: adjusters' pay, appraisers, claims offices, and legal defense. Defense costs for a liability claim count here even when the insurer owes nothing.
• Selling expenses. Commissions paid to agents and brokers, plus other acquisition costs such as advertising, mailings and the salaries of sales staff who work for a wage rather than commission.
• General expenses, taxes, licenses and fees. Overhead such as the home office and actuaries' salaries, and the premium taxes and state fees an insurer owes. Premium taxes and fees vary by state, so this slice is not the same in every state.
• Profit. What remains after those costs when the policy year is complete. Insurers also earn investment income on the money they hold between collecting premiums and paying claims, which is why "profit" has two parts, covered below.

A smaller item, policyholder dividends, appears in NAIC tables because mutual companies sometimes return part of their margin to policyholders. For private passenger auto it was about half a cent per dollar in 2024.

What the industry numbers show

The most direct national source is the NAIC's annual Report on Profitability by Line by State. The edition published in July 2026 covers calendar year 2024 and draws on annual statements filed by 2,634 property and casualty insurers, which the NAIC estimates represent well over 95% of U.S. premiums. The table below reproduces the countrywide line for private passenger auto, all company types, as a percentage of net premiums earned.

Net premiums earned for the whole auto line were about $339 billion. A few readings follow from the table.

First, liability and physical damage behave very differently. Liability, which pays other people's injuries and property damage, ran at an underwriting loss in 2024 while collision and comprehensive earned a double-digit margin. Combined into one line they average out, but a company that sells mostly liability-heavy policies will not look like one weighted toward physical damage.

Second, the selling expense is the same size for both coverages, about 13.5 to 13.7 cents, while the claims piece swings. Commissions and acquisition costs are mostly a percentage of premium, so they do not shrink when claims are low.

Third, the investment gain is real money but it sits outside the underwriting result. The NAIC explains that its column on investment gain on insurance transactions counts only gains attributable to premium and claim reserves and leaves out gains on the insurer's capital and surplus, and that it excludes unrealized capital gains. The report also cautions that it cannot be used to decide whether current rates are adequate for future costs. It shows what happened last year, not what next year's filing should say.

A second yardstick: all property and casualty lines

For perspective, the Insurance Information Institute (Triple-I) publishes industry totals drawn from NAIC data. Its industry overview shows that in 2025, across all property and casualty lines, insurers earned about $953.5 billion in premiums, incurred $634.8 billion in losses and loss adjustment expenses (about 67% of premium), and spent $251.1 billion on other underwriting expenses (about 26%). Auto is more claims-heavy and lighter on expenses than that average. In the 2024 NAIC table, losses plus LAE come to about 75 cents per auto dollar, and general expense, selling expense and taxes together come to about 20 cents.

A worked example with invented numbers

Say a hypothetical insurer needs to price a six-month policy for a typical driver. The numbers below are made up to show the mechanics, not market data. Its actuaries project:

• Expected claims and adjusting costs per policy: $1,100
• Fixed expenses per policy (printing, data entry, billing): $60
• Variable expenses as a share of premium: 15.5% (commission and acquisition 12%, premium tax 2.5%, other 1%)
• Target underwriting profit: 4% of premium

Because the variable costs and profit are percentages of the premium itself, the premium has to be large enough that what remains after them covers the fixed costs and the expected claims. The CAS text gives the formula: premium equals (losses plus LAE plus fixed expense) divided by (1 minus variable expense share minus profit share).

1. Add expected claims and fixed expense: $1,100 + $60 = $1,160.
2. Subtract the variable share and the profit target from 100%: 1 - 0.155 - 0.04 = 0.805.
3. Divide: $1,160 / 0.805 = about $1,441.

Where each dollar of that $1,441 goes:

Now test how thin the margin is. If real claims come in 5% above the projection ($1,155), the underwriting profit shrinks to about $2.60, or 0.2 cents per dollar. If they come in 10% above ($1,210), the insurer loses about $52 on the policy, a combined ratio of about 103.6%. To get back to the 4% target at those higher costs, the premium would have to be about $1,578, a 9.5% increase. That is a large share of why loss ratios drive rate filings and why a modest surprise in claim costs turns into a double-digit rate request.

Combined ratio, investment income and what "profit" means

The combined ratio adds the loss ratio and the expense ratio. A figure under 100 means the underwriting alone made money; over 100 means it lost money. In the example above, the baseline combined ratio is 96.0%, the 5% miss gives 99.8%, and the 10% miss gives 103.6%.

Triple-I reports combined ratios for private passenger auto from NAIC data, in its auto insurance facts page. The swing is large. Private passenger auto liability ran a combined ratio of 111.3 in 2022 and 100.2 in 2025, while collision and comprehensive combined went from 113.4 in 2022 to 81.0 in 2025. Those readings are after dividends to policyholders, and a drop in the ratio is an improvement. A stretch of combined ratios far above 100 is followed, with a lag, by rate increases, and a stretch far below 100 tends to invite competition and, in some states, questions from regulators.

Investment income is the second source of profit. Auto claims are paid within months or a few years, so the float is smaller than in lines like workers' compensation or medical malpractice, but it is not zero. The CAS text notes that for long-tailed lines investment income can be large enough to accept an underwriting loss, while for short-tailed lines the investment potential is lower and underwriting profit is a larger part of the total return. For how interest rates feed back into price, see do interest rates and investment income change your premium?. The takeaway for reading an annual report: an insurer with a combined ratio of 101 and a healthy investment return can still be profitable, while one at 96 may still miss its target return if it holds a lot of capital against the premium it writes.

Why direct writers and agent-based insurers look different

How a company sells its policies changes the mix of expenses even when the claims cost is identical. The CAS text makes the point directly: a national direct writer may incur significant other acquisition costs for advertising, while an agency-based company relies on agents to generate new business, so its other acquisition costs are lower, partly offset by higher commission expense.

The table below is hypothetical, built to show structure rather than any company's real numbers.

Neither model is automatically cheaper. A direct writer saves the commission but pays for advertising and its own call centers; an agent-based company shifts sales cost into commissions but may spend less on marketing. Agents also do work that would otherwise sit with the insurer, such as advice at renewal, help at claim time and policy servicing. Whether that is worth a commission is a consumer judgment, not an accounting fact. For the commission side in detail, see how insurance agents and brokers get paid, and why it matters to you.

The CAS text adds that commissions run lower in personal auto than in lines that require a detailed inspection when the policy begins, such as large commercial property.

What regulators let insurers put in a rate

Insurers cannot simply add any cost they incur. State laws set the test, and the details vary by state, so check your state's insurance department for how it applies where you live; the state car insurance requirements lookup tool is a starting point for your state's basics.

Florida is an example of a statutory approach. Under Florida Statutes section 627.0651 (2026 edition), motor vehicle rates must allow the insurer a reasonable rate of return. When the state's Office of Insurance Regulation reviews a filing it must consider, among other factors, past and prospective loss experience, past and prospective expenses, competition, trend factors, and investment income from investable premiums, which the statute says must not include income from invested surplus. Rates are deemed excessive if they are likely to produce a profit from Florida business that is unreasonably high in relation to the risk, or if expenses are unreasonably high in relation to services rendered. Portions of judgments and settlements tied to bad faith or punitive damages cannot be included in the rate base.

California goes further in listing what cannot be recovered. Its regulation 10 CCR 2644.10, on excluded expenses, bars these from ratemaking: political contributions and lobbying; executive pay above a regulatory formula; bad faith judgments and associated defense costs; costs of unsuccessfully defending discrimination claims; fines and penalties; institutional advertising (advertising not aimed at winning business for a specific insurer and not giving consumers information relevant to buying); and payments to affiliates above fair market value. The disallowance is applied by reducing the efficiency standard used in California's review. See California and Florida for each state's rate-review basics.

The CAS text notes that a few states restrict which expenses may be included in rates, such as charitable contributions or lobbying. In practice that means the expense line in a filed rate may be lower than the insurer's actual expenses, and the difference is paid out of profit.

Why one company's numbers differ from another's

When you compare a company's loss ratio or expense ratio with the industry, or with a competitor, line up these differences first:

• Coverage mix. Liability-heavy and physical-damage-heavy books have different loss ratios, as the NAIC table shows.
• Net or direct, written or earned. The NAIC report states that it allocates some expenses on a mix of written and earned premium, and that a net-basis table and a direct-basis table can show different profitability. A fast-growing insurer's expense ratio looks worse measured against earned premium than written premium, because acquisition costs are paid up front.
• Reserve changes. Last year's claims estimates may be revised this year; those adjustments move the loss ratio without any new accidents.
• Weather and catastrophes. A hailstorm year pushes comprehensive losses up for insurers in that territory only.
• Reinsurance. Ceding premium to reinsurers changes the net numbers.
• One year versus a cycle. The NAIC report itself offers long-term averages because single calendar years swing.
• Allocation. Some expenses are not reported by state or line and are allocated by formula, so a state-by-line figure is an estimate.

Comparisons across companies are rarely fair, and the NAIC's warning on adequacy applies to consumers too: a single ratio does not tell you whether your premium is fair. A better test is your own quote against others for the same coverages; see how car insurance rates are calculated for the pricing structure and Profitability by line and state: what NAIC data says about margins for how to read the profitability tables.

Questions to ask when you see a premium-dollar chart

Charts that claim to show "where your premium goes" often come from an insurer, a trade group or an advocacy group. Before trusting one, ask:

1. Which year, and is it one year or an average?
2. Is it one insurer, one state or the national industry?
3. Does it include loss adjustment expense with claims, or separate?
4. Is the denominator earned premium, written premium or something else?
5. Are investment returns shown, or only underwriting profit?
6. Who produced it, and what is its source?

Common questions

Does most of my premium really go to claims?

On average, yes. In the NAIC's 2024 figures about 66 cents of each auto premium dollar went to losses and about 9 cents to the cost of handling them, roughly three-quarters in total. The proportion depends on the coverage, the company and the year.

Is an insurer's profit margin capped?

Not by a fixed national percentage. States require rates not to be excessive, inadequate or unfairly discriminatory, and many look at whether profit is reasonable for the risk. In Florida the statute treats profit unreasonably high in relation to the risk as a sign of excessive rates. The exact test and the review process differ by state, so the state insurance department is the place to ask about a particular filing.

Do independent agents make coverage more expensive?

Not by themselves. Commission is a cost, but a company without agents carries advertising and call-center costs instead. Because real prices also reflect claims experience, underwriting rules and discounts, the only reliable comparison is quotes for the same coverages.

If an insurer reports a combined ratio under 100, are rates too high?

Not necessarily. A ratio below 100 means underwriting broke even or made money in that year. It can reflect a mild claims year, rate increases catching up after losses, or a conservative mix of business. Regulators look at multi-year results and expected future costs, not one year.

Key takeaways

• About three-quarters of an auto premium dollar covers claims and the cost of handling them; selling, overhead and taxes take roughly another fifth, and the underwriting profit is a few cents.
• Liability and physical damage differ sharply: in the NAIC's 2024 data liability ran at an underwriting loss while collision and comprehensive earned a double-digit margin.
• A small miss in claims costs wipes out a thin margin, which is why modest cost surprises turn into large rate requests.
• Regulators allow reasonable expenses and a reasonable return, and some states exclude specific items, so filed rates can understate the insurer's true costs.
• Ratios from different companies, years and sources are rarely comparable without checking coverage mix, premium basis and allocation.