Commercial auto posted a $4.9 billion underwriting loss in 2024, its 14th consecutive losing year, on AM Best's September 2025 market segment report. The loss narrowed from $5.5 billion in 2023, but not because losses improved. Premium simply grew faster than they did, and the line's two halves are now moving in opposite directions at record magnitude.

Key Takeaways

  • $4.9 billion lost in 2024 and roughly $40 billion cumulatively since 2011, averaging $2.9 billion a year over the past decade.
  • A record $6.4 billion liability loss alongside a record $1.5 billion physical damage profit in the same year, from the same line of business.
  • 8% average annual severity growth over nine years against roughly 3% economic inflation, which turns a $100,000 claim in 2015 into $215,900 by 2024 rather than $130,500.
  • $4 billion to $5 billion of estimated industry under-reserving in commercial auto liability, with 8.0% adverse one-year development in 2024 and adverse movement across every accident year from 2016 forward.
  • A 42-point combined ratio spread across the top 20 writers, from Progressive at the best to Sentry at 130.

The Improvement Came From Premium, Not Losses

Direct premiums written reached $72.2 billion in 2024, a 12.1% increase, with liability alone growing 12.3% to $43 billion. That is roughly three times the 4.0% average across all commercial lines. The combined ratio improved to 107.2 from 109.3, a 2.1-point move.

Almost none of that landed in liability. The liability combined ratio finished at 113.0 against 113.3, and has exceeded 100 every year since at least 2014, crossing 113 five times. The improvement is physical damage, which produced an 88.6 combined ratio, its best outside the pandemic year, and has not exceeded 100 since 2017.

The divergence reached a record in both directions in the same year: commercial auto liability posted its largest-ever single-year underwriting loss at $6.4 billion while physical damage posted its highest-ever profit at $1.5 billion. The loss and LAE ratio for liability hit 87.6, an 11-year high, and the gap between the two sub-lines' loss ratios reached 24.6 points.

The reason the two halves behave differently is that only one of them tracks something observable. Physical damage costs follow vehicle repair and replacement, which parts inflation and ADAS complexity have raised but which remain anchored to prices a pricing actuary can measure. Liability follows jury awards.

That makes the blended 107.2 a misleading figure to price against. A carrier setting rate on total commercial auto lets physical damage profit subsidize a liability sub-line that has not moved in three years, and the Triple-I and Milliman Forward View puts commercial auto and general liability as the only major lines still above a 100 net combined ratio while the broader industry ran at 91 in 2025.

The Rate Arithmetic Cannot Close a 5-Point Compounding Gap

Risk and Insurance's analysis of the AM Best data puts average commercial auto liability severity growth at 8% annually over nine years, against roughly 3% economic inflation. The 5-point spread is the social inflation component, and it compounds.

Run it forward. A $100,000 average claim in 2015 reaches $215,900 by 2024 at 8%. At 3% it reaches $130,500. The $85,400 difference, applied across hundreds of thousands of liability claims a year, is the loss the rate increases have been chasing.

The chase does not work at these magnitudes. Premium growing 12% against severity growing 8% nets four points a year, applied to a loss ratio that starts above 100. Reaching liability breakeven from 113 on four points a year takes most of a decade of sustained double-digit increases, and S&P Global Market Intelligence has revised growth below 10% while projecting the combined ratio rising from 104.4% in 2026 to 106.3% by 2029.

The reserves record the shortfall. AM Best estimates the line remains under-reserved by $4 billion to $5 billion. Milliman's statutory analysis found 8.0% adverse one-year development in 2024 with adverse movement across every accident year from 2016 forward, against a decade average of seven points. Adverse development from accident years 2021 and later alone exceeded $2.7 billion, and those are hard-market vintages that were supposed to be priced adequately.

Dispersion says the problem is not uniformly structural. Sentry ran 130 in 2024, Chubb 126.2, State Farm 123.6, while Progressive posted the best underwriting result among the top 20 on telematics-based segmentation and a willingness to non-renew. A 42-point spread means industry-average trend selections are close to useless for any individual book, and the 90th percentile loss ratio moving from 105% to 120% in 2024 says the tail of the carrier distribution is widening even where the median improves.

The Capital Behind the Claims Is Not Cyclical

Nuclear verdicts are the visible symptom. NAIC research puts auto accident cases at 22.8% of all verdicts above $10 million, second only to product liability at 23.6%. Marathon Strategies counted 135 corporate-defendant nuclear verdicts in 2024, up 52% and the highest since tracking began in 2009, totalling $31.3 billion. Verdicts above $100 million rose 81.5% to 49 cases, five exceeded $1 billion, and the median reached $51 million from $44 million. Trucking carries $4.1 billion of that mega-verdict exposure.

The TransRe overview identifies why this resists actuarial trending: non-economic damages, pain and suffering rather than medical cost or lost wages, now make up the majority of nuclear verdict amounts. A severity trend built on medical cost indices is measuring the smaller component.

Underneath the verdicts sits capital. Third-party litigation funding reached $17 billion globally as of 2021 with just over half in the United States, and industry estimates put the US market above $15 billion by 2025. Funded plaintiffs face no carrying cost, so a claimant who would have taken $500,000 within 18 months can wait years. Swiss Re's research finds funded cases settle higher and resolve slower, which is a loss development factor problem rather than a severity trend problem, and the two require different corrections.

The policy response has stalled. Eight states have enacted or proposed disclosure requirements for funding agreements. A proposed 40.8% federal tax on litigation funding appeared in early drafts of the 2025 One Big Beautiful Bill Act and was stripped before passage, removing the one measure that would have changed the asset's economics.

There is a precedent for how this resolves, and it is not encouraging. A 3.5-point combined ratio improvement arrived in 2014, and was followed by five pre-pandemic years running roughly 109 to 111. The improvement in the 2024 numbers is the same shape: premium outrunning losses for a year, in a line where the loss driver is funded by capital that does not respond to the underwriting cycle.

Further Reading

Sources