AM Best published its annual commercial auto review on October 29, 2026 under the title “Commercial Auto May Finally Be Seeing Light at The End of The Tunnel.” The AM Best release puts the line’s 2025 underwriting loss at $2.2 billion, a 55% improvement over the prior year, and says: “Overall, loss reserve deficiency is down, and the most recent accident year appears to be adequately reserved.” The full report is behind a paywall, so the reserve detail below comes from Risk & Insurance’s November 4 summary and is flagged as such.
That summary carries the number a fleet risk manager will hear repeated at renewal: AM Best’s estimate of the commercial auto liability reserve deficiency narrowed from $4 billion to $5 billion at year-end 2024 to roughly $2 billion to $3 billion at year-end 2025. The temptation is to read that as permission to ease recent-year reserves. The detail underneath points the other way.
What the report says
The release splits the 2025 result cleanly. Liability lost more than $4.6 billion. Physical damage earned a $2.4 billion gain. The direct loss ratio for the first half of 2026 was 66.1, which AM Best calls a modest improvement on the same period a year earlier.
As reported by Risk & Insurance, the liability combined ratio was 108.6 in 2025. A year ago, Insurance Journal’s coverage of the 2025 edition put the 2024 liability combined ratio at 113, so the improvement is about 4.4 points and the line is still more than eight points from breakeven. Liability is 77% of commercial auto net premiums written.
The reserve figures, all per the Risk & Insurance summary:
- Adverse prior-year development on liability was just under $2 billion in 2025, the lowest since 2018.
- More than half of it came from accident year 2023.
- Accident years 2021 and earlier contributed only $97 million, which AM Best reads as a sign that older years are adequately reserved.
- Adverse development added four points to the 2025 calendar-year loss ratio, against roughly seven points in prior non-COVID years.
- The average loss per claim rose to $31,537 in 2025 from $28,771 in 2024.
Who it affects
Self-insured trucking, delivery, and transit fleets retaining auto liability; group captives and risk retention groups writing fleet auto; public entity pools with vehicle exposure. Anyone whose actuary benchmarks auto liability development or severity trend against industry data will see this report cited in the year-end 2026 review.
The reserve mechanism
Start with why the deficiency narrowed. Take the midpoints of AM Best’s two ranges: $4.5 billion at year-end 2024 and $2.5 billion at year-end 2025. The gap closed by about $2 billion. Carriers recognized just under $2 billion of adverse liability development in the same year. (The midpoint comparison is our arithmetic, not an AM Best statement.) On that reading, AM Best’s view of ultimate losses for prior years barely moved. Carried reserves moved up to meet it. The deficiency shrank because carriers paid for it through calendar-year 2025 earnings, not because claims turned out cheaper.
That distinction decides the fleet question. An industry estimate that fell because ultimates came down would support lower selections. An estimate that fell because carriers strengthened only tells a fleet that others have booked what it may not have booked yet.
Next, where the development sits. With $97 million of just under $2 billion coming from 2021 and earlier, roughly 95% of 2025’s adverse development landed on accident years 2022 and later, and more than half on 2023 alone. Accident year 2023 moved from 24 to 36 months of maturity during 2025. That is the window in which represented bodily injury claims leave adjuster-level reserves and get priced by demand and suit, and it is where age-to-age factors picked from pre-2020 history are most likely to run light.
The lever is severity. The average loss per claim rose about 9.6% in one year by our arithmetic on the two figures above, and AM Best attributes the rise to social and economic inflation, including more attorney-represented claimants and higher jury awards (per Risk & Insurance). A favorable reserve story and a near double-digit severity increase in the same report do not reconcile unless the favorable part belongs to old years and the severity part to recent ones.
Finally, the remainder. AM Best still sees $2 billion to $3 billion of deficiency at year-end 2025. If 2021 and earlier are adequate and 2025 appears adequate, simple elimination puts most of what remains in accident years 2022 through 2024. That is our inference from the published statements, not a breakdown AM Best has released in the material we could read.
The carrier-level evidence agrees
The carrier disclosures this site has covered since midyear match the accident-year pattern rather than the headline. On W.R. Berkley’s October 19 call, management called commercial auto “the primary source of adverse development in the insurance operation,” including exposure reaching excess and umbrella coverage, per MarketBeat’s call summary; our read of Berkley’s nine-month filing traces that development to recent accident years. The Hartford added $26 million to commercial auto reserves in the second quarter of 2026, citing higher severity trends, per Investing.com; see our Hartford and Travelers comparison. Both charges were booked in 2026, after the year-end 2025 position AM Best measured.
This is also consistent with what was visible in July. actuary.info’s read of AM Best’s earlier 2025 industry results described about $2.0 billion of new commercial auto deficiency recognized in 2025, concentrated in accident years 2023 and 2024. The October report confirms the vintage. Our June piece on the $4 billion to $5 billion estimate covers the starting point.
Where this shows up in your reserves
Pull the auto liability development exhibit from your last two actuarial reports and look at accident year 2023 between 24 and 36 months, and 2024 between 12 and 24. Compare your actual incurred emergence with what the prior report expected. If your 2023 year did not move while the industry’s absorbed more than half of a $2 billion charge, either your claims are better or your case reserves are slower. The five leading indicators help tell those apart, and Commercial Auto and Fleet IBNR explains why bodily injury must be read separately from physical damage.
What this means for your next review
- Ask your actuary which accident years the industry benchmark improvement applies to, and confirm no credit is taken on 2022 through 2024.
- Ask for accident year 2023 and 2024 liability ultimates on paid and incurred bases, with the severity trend assumption stated next to the 9.6% one-year industry change.
- If a Bornhuetter-Ferguson expected loss ratio is used for 2025 and 2026, ask whether it was lowered on the strength of this report.
Our read
The industry improvement is real for accident years 2021 and earlier and unproven for everything after. We would not ease 2022 through 2024 fleet liability selections at year-end 2026. Fleets carrying explicit margin on 2021 and prior years have better support for letting that margin run off as claims close.
We expect year-end 2026 statutory data, which AM Best will review in fall 2027, to show another year of adverse commercial auto liability development, smaller than 2025’s, with accident year 2024 replacing 2023 as the largest contributor as it passes through the same 24 to 36 month window. The “adequately reserved” verdict on accident year 2025 was reached at 12 months of maturity; accident year 2023 is the reminder of how much can change by 36. With second-quarter 2026 price increases the lowest in more than 10 years (per Risk & Insurance) while severity runs near 10%, we read 2026 as the accident year most likely to be the next one under-picked.
Sources
- Best’s Market Segment Report: US Commercial Auto Insurance Segment May Finally Be Seeing Light at End of Tunnel, AM Best, October 29, 2026
- Commercial Auto Underwriting Results Show Signs of Long-Awaited Improvement: AM Best, Risk & Insurance, November 4, 2026
- AM Best: Commercial Auto Liability Drags Down Segment and it Could Get Worse, Insurance Journal, September 22, 2025
- W. R. Berkley Q3 earnings call highlights, MarketBeat, October 19, 2026
- Hartford Q2 2026 slides, Investing.com, July 24, 2026
- AM Best’s Decade-High P&C Profit Tests Reserve Durability, actuary.info, July 2026