Guides
The combined ratio: reading home insurer results
What a combined ratio of 95 or 108 means for a home insurer, how net, direct and underlying versions differ, and why no single ratio proves a company is sound.
The combined ratio is claims costs plus expenses, divided by premium, expressed as a percentage. Below 100% the insurer made an underwriting profit; above 100% it paid out more than it collected, before any investment income. A homeowners combined ratio of 88% means about 12 cents of underwriting profit per premium dollar, and 108% means an 8-cent loss.
The number is useful and easy to misread. Different sources compute it differently, one storm can swing it 20 points, and a company can post a 120% in a year it fully survives. This guide shows how to read it.
The formula
The combined ratio is the sum of two ratios, with a third small piece in some cases.
| Piece | What it measures | Typical contents |
|---|---|---|
| loss ratio (loss and loss adjustment expense) | Claims plus the cost of handling them, divided by earned premium | Paid and reserved claims, adjuster fees, inspections, defense costs |
| expense ratio | Operating costs divided by premium | Commissions, salaries, technology, marketing, premium taxes |
| Policyholder dividends | Dividends paid to mutual policyholders, divided by premium | Usually under 1 point |
The Insurance Information Institute (Triple-I) published the pieces for 2025 using NAIC data compiled by S&P Global Market Intelligence: loss and loss adjustment expense 61.1%, expenses 26.4%, dividends 0.4%, for a homeowners combined ratio of 87.8%. For 2023 the same table showed 84.5%, 26.1% and 0.4%, which added up to 110.9%. Notice that nearly all of the movement between the two years came from the loss piece. Expenses barely changed.
Net, direct, and "underlying"
There is not one combined ratio. When you read a number, find out which one it is.
Net combined ratio. Uses premiums and losses after reinsurance. This is what the Triple-I and Milliman figures are: 110.9 (2023), 99.7 (2024) and 88.1 (2025). It shows what the insurer actually kept.
Direct combined ratio. Uses premiums and losses before reinsurance. State regulators and AM Best use it to compare states, because reinsurance is purchased at the company level, not state by state. AM Best reported direct combined ratios above 100 in 17 states in 2023 and in 10 states in 2024.
Loss ratio only. Often mistaken for a combined ratio. The NAIC's 2025 market share report gives a homeowners direct loss ratio of 55.14% (losses divided by earned premium, with no expenses), against 61.51% in 2024 and 70.52% in 2023. The NAIC's loss and cost containment ratio for the year was 56.52%. These are not combined ratios; add expenses of roughly a quarter of premium to get close to one.
Catastrophe and underlying. Public insurers split the number into "catastrophe losses" and an "underlying" ratio that excludes catastrophes and prior-year reserve changes. A homeowners combined ratio of 94.6% might be 55 points of ordinary claims and expenses plus 40 points from named storms. Allstate's homeowners line illustrates the swing: its Q2 2026 homeowners combined ratio was 94.6, against 102 in Q2 2025, according to Insurance Journal's report on the results, with catastrophe losses falling in the quarter.
Statutory versus GAAP. Regulators' statutory accounting and the GAAP accounting in a public company's reports differ in how they treat acquisition costs and reserves. The two ratios for the same company can differ by a point or more.
What counts as a good number
There is no universal threshold, but a few reference points help.
- Under 100. An underwriting profit. For homeowners, 2019 was the last year before 2025 that the industry was clearly below 100, per Triple-I and Milliman's note that 2024's 99.7 was the first sub-100 result since 2019.
- 95 to 100. Roughly breakeven once investment income is added. Not alarming in a storm year.
- Over 105. A loss that surplus has to absorb. One year is survivable. Several in a row lead to rate increases, tighter underwriting or an exit from the state.
- Under 80. Strong, but a regulator may question whether the rates are excessive. Florida domestic insurers posted a net combined ratio of 76.8% in 2025, according to Guy Carpenter's June 2026 Florida renewal report, which came after the state's 2022 and 2023 litigation reforms; see Florida.
A single number deceives for a reason hidden in the ratio itself: it depends on storms. A company with a 90 in a quiet year and a 130 in a hurricane year is not a 110 company. Look at five years, not one.
A worked example (hypothetical)
Say an insurer reports these homeowners results.
| Year 1 | Year 2 | |
|---|---|---|
| Earned premium | $500 million | $540 million |
| Non-catastrophe losses and adjusting | $190 million | $205 million |
| Catastrophe losses (net) | $60 million | $230 million |
| Expenses | $135 million | $142 million |
| Combined ratio | 77% | 107% |
| Underlying ratio (ex-catastrophe) | 65% | 64% |
Year 2 looks like a collapse until you separate the pieces. Non-catastrophe claims and expenses behaved. A single catastrophe season moved the number by 30 points. An analyst would ask a different question: is the insurer's catastrophe load, the amount built into rates for average storms, big enough, and does its reinsurance limit the damage? The figures here are invented to show the mechanics.
What the ratio hides
- Reserve development. If an insurer under-reserved for old claims, the shortfall shows up as an extra point or two in later years, and a favorable release flatters the current year. Public insurers report it as prior-year development.
- Reinsurance cost. The reinsurance premium is deducted from net premium. When reinsurance gets cheaper, as it did in 2025 and 2026, the net combined ratio improves without a single claim changing. reinsurance: the insurers behind your insurer covers that mechanism.
- Mix of business. A company writing homeowners, auto and umbrella reports one blended ratio. Insurers that break out homeowners separately, such as Allstate in its quarterly reports, let you see the line on its own. Erie Insurance and other regional mutuals often report only a total.
- Float. A 103% combined ratio can still be profitable after investment income. See how home insurers make money, and lose it.
- Timing. Premium is earned gradually, while a catastrophe loss lands all at once. Quarterly numbers jump around, and a half-year figure for a company that has a storm season starting in June is not a forecast.
- Capital. The ratio says nothing about whether the insurer has enough policyholder surplus to take a bigger hit. Look at the rating; see Financial strength ratings: AM Best, S&P, Moody's and Demotech.
How to use it as a homeowner
You will not find a combined ratio on your declarations page, but it informs real decisions.
- Predict rate pressure. A state where direct combined ratios exceed 100 for several years tends to see rate increases and tighter underwriting. A state with ratios in the 70s and 80s often sees filings for decreases.
- Judge a company's staying power. Check a few years of results, not one, together with AM Best or Demotech ratings.
- Read news in context. A headline claiming "record insurer profits" refers to an underwriting result after years of losses, usually averaged across states that may not include yours.
- Ask your agent. If your insurer is retrenching, the agent often knows which states or ZIP codes are being restricted. See home insurance market outlook.
Key takeaways
- The combined ratio is the loss ratio plus the expense ratio; under 100% means an underwriting profit.
- Net, direct and underlying ratios are different measures, so always find out which one a source is quoting.
- Homeowners results swing on catastrophes: the line's net ratio was 110.9% in 2023, 99.7% in 2024 and about 88% in 2025 per Triple-I.
- Judge a company on several years plus its surplus and rating, not one ratio from one year.