The Triple-I and Milliman Forward View released May 14 projects P&C net premium growth at negative 3.7% for the first half of 2026, against positive 1.6% in 2025. It landed days after AM Best confirmed a 92.9% net combined ratio for full-year 2025, the strongest underwriting result in nearly two decades. The industry is contracting at the top of its own cycle, and the profit that defines the top is partly borrowed.

Key Takeaways

  • Negative 3.7% net premium growth projected for H1 2026 against positive 1.6% in 2025, the first industry-wide contraction since 2020, with no recovery projected until 2027.
  • 92.9% net combined ratio for 2025 on AM Best data, 93.0% on Fitch's, with net underwriting income tripling to $60.9 billion.
  • $18.1 billion of favorable prior-year development put the accident-year combined ratio at 94.9, nearly two points worse than the calendar-year figure.
  • Workers' compensation reserve redundancy fell from $16 billion to $14 billion and its accident-year combined ratio stands at 102.
  • Replacement cost growth of 2.1% for H1 2026, down from the 6% to 8% range of 2022 and 2023, but projected to re-accelerate through 2028.

What the Headline Result Contains

Net underwriting income tripled to $60.9 billion on a 92.9% combined ratio, which Fitch put at 93.0% and called the best since 2006. Swiss Re labelled 2025 the cyclical peak. The quarterly figure was lower still: Q3 2025 came in at 89%, the lowest single quarter since at least 2001.

Two points of it were borrowed. AM Best reported $18.1 billion of favorable prior-year reserve development, nearly double the prior year, and excluding it the accident-year combined ratio was 94.9. That gap is the difference between what the industry earned on business written in 2025 and what it reported after releasing reserves held against earlier years.

Replacement cost growth has moderated to a projected 2.1% for H1 2026, flat with 2025 and well down from the 6% to 8% range that drove severity through 2022 and 2023. The Forward View expects it to re-accelerate through 2028 and eventually outpace general inflation as tariff effects reach construction materials and auto parts.

Line of Business2025 Net Combined RatioTrendForward Outlook
Personal auto91.8Improved 3.5 pts from 2024Favorable into 2026; NWP growth slowing to 4.0%
Homeowners88.1Decade low despite LA wildfiresReplacement cost re-acceleration could pressure margins
Workers' compensation91 (CY, per NCCI)12th consecutive year below 100Low 90s through 2028; AY CR at 102
General liabilityAbove 100Persistently unprofitableGradual improvement; above 100 through 2026
Commercial autoAbove 100Persistently unprofitableForecast at 104.4 in 2026; no near-term breakeven

The average conceals a two-tier industry. Personal auto at 91.8 and homeowners at 88.1 carry the result; homeowners held that level despite roughly $40 billion of insured losses from the January 2025 Los Angeles wildfires. General liability and commercial auto remain above 100, which means the aggregate is a diversification subsidy running from personal lines toward casualty.

Why Volume Falls When Margins Peak

The contraction has three sources and none of them is demand. Rate adequacy removed the justification for further increases: personal auto filings that ran 10% to 15% annually through 2023 and 2024 have decelerated to low single digits, and at a 91.8 combined ratio neither regulators nor competitors will support more. Exposure growth slowed alongside it, in auto unit counts, housing starts, and commercial property values. And competition accelerated, with property catastrophe reinsurance rates falling 14% to 25% at the January and April 2026 renewals, which transmits straight into primary property pricing. Chubb has been non-renewing business where rates fell 30% to 40%.

The pricing consequence is that a rate indication built on calendar-year results will understate the need. The number that describes the cost of writing business in 2025 is the accident-year 94.9, not the 92.9, and the reserve development bridging them is finite. Fitch expects releases will not repeat at the same scale in 2026 and 2027.

Run it on a personal auto book. A line at 91.8 may need 3% to 5% of rate simply to stay there once three things are accounted for: tariff-driven parts and materials inflation running 2% to 3% above the embedded trend assumption, with APCIA estimating a 2.7% increase in collision repair costs and $3.4 billion of additional premium needed across auto lines; frequency drifting back toward pre-2020 levels; and favorable development on the 2022 and 2023 accident years not repeating on the 2024 and 2025 vintages.

Contraction adds a mechanical effect on top. The combined ratio is a ratio, and when premium falls 3.7% while losses do not fall proportionally, which they will not where the driver is litigation severity, the ratio deteriorates with no change in absolute loss dollars. A carrier concentrated in general liability or commercial auto experiences that differently from one weighted to personal lines, and the industry figure will not show which.

Both Sources of the Subsidy Are Draining at Once

Workers' compensation has been the other reliable offset, and it is turning. NCCI reported a calendar-year combined ratio of 91 for 2025, up five points from 86, while reserve redundancy fell from $16 billion to $14 billion. The accident-year figure is 102, so current business is marginally unprofitable before any release. Lost-time claim frequency fell 2% but medical and indemnity severity each ran 4%, and California posted an accident-year combined ratio of 129, above 100 for five consecutive years in a state carrying 20% of the national market.

That matters for carriers who used workers' compensation releases to absorb casualty adverse development. Casualty deterioration on the 2021 through 2024 accident years has reached $15.8 billion, spreading past the soft-market vintages that were expected to carry it. The offset shrinking and the deterioration spreading are happening in the same reserve review.

The historical shape of what follows is consistent. The 2006 peak preceded the 2007 to 2011 soft market. The 2013 trough in combined ratios set up competitive softening through 2014 to 2017. The 2019 improvement reversed through 2020 to 2022. Each peak deteriorated within 12 to 18 months, and in each case the mechanism was the same: strong results attract capital, capital funds share competition, and rate adequacy erodes before the loss trend does.

What is different here is that premium growth has already turned negative, which in prior cycles arrived after the peak rather than with it. AM Best projects the combined ratio moving from 92.9% to 96.9% in 2026 and Fitch to 96% or 97%, with Swiss Re taking ROE from 15% in 2025 to 12% in 2026 and 10% in 2027. Those are mid-cycle numbers rather than distress. The exposure is that the reserve releases and the workers' compensation redundancy that made 92.9% possible are both being consumed while the casualty lines they were subsidising are still above 100.

Further Reading

Sources