Q2 2026 P&C Earnings: Property Softens, Casualty Won't Follow

Q2 2026 property and casualty earnings arrive with commercial pricing decelerating unevenly. Large-account property rates are down more than 30% while US commercial auto excess-of-loss layers rose 10% to 15% at mid-year renewals.

The blended index does not show that. The CIAB's Q1 2026 survey put the overall average premium change at negative 1.2%, and a book falling by a third can sit inside that figure.

Key Takeaways

• Negative 1.2% blended against more than 30% off large-account property is the gap: one is every commercial line a broker places, the other is the cat-exposed layer where reinsurance capacity sets the price.
• A 33-quarter streak of increases ended, with large accounts down 2.7%, medium down 1.9% and small accounts still up 1.1%, a 60% deceleration from the prior quarter's 2.8%.
• Guy Carpenter's US property cat rate-on-line index fell 16%, matching the global index, with Asia-Pacific down 19% and Europe down 15%, the steepest annual drop since the late 1990s.
• $648 billion of dedicated reinsurance capital at year-end 2025, up 11%, against $38 billion of natural catastrophe losses through June 15, below the ten-year average.
• $15.8 billion of adverse prior-year casualty development in 2024, the highest on record for those segments, is why casualty is not following property down.

The Blended Index and the Property Number Are Not the Same Number

Autonomous Research's Q2 preview centers on the pace of commercial pricing declines, a personal lines tradeoff between growth and profitability, and whether brokers can sustain organic growth. Large accounts are taking the steepest cuts, with property reductions exceeding 30% in places, while mid-year reinsurance renewals brought reductions of 15% to 20%.

The broker survey reads far calmer. Overall average premium change came in at negative 1.2%, ending a 33-quarter streak stretching back nearly nine years. Large accounts, those generating more than $100,000 of annual commissions and fees, fell 2.7% for a second consecutive quarter. Medium accounts slipped 1.9%. Small accounts still rose 1.1%, though that was a 60% deceleration from the prior quarter's 2.8%.

Both figures are right. One blends property and casualty across every line placed for an account size; the other isolates the cat-exposed property layer. A blended index barely more than a point soft is fully consistent with a property book down a third, provided casualty inside the same blend is flat or firming.

On the personal side the softening is a choice rather than a constraint. Progressive noted that consumers who shop infrequently had begun returning as premium reductions took hold, with combined ratios still ahead of long-term averages.

Earned Rate Lags Written, and the Lag Just Inverted

Written rate moves the moment a policy renews. Earned rate catches up only as the in-force book turns over, typically across four to twelve quarters depending on term and renewal-date distribution.

Through 2023 and into 2024 that lag ran in carriers' favor. Earned rate sat below written rate, understating margin improvement and leaving reserves conservative against the rate actually being charged on new business. A property book renewing down 30% inverts it. Earned rate now runs above written rate for several quarters, so a reported accident-year property loss ratio looks better than the business being written today purely as an artifact of renewal timing.

That is where the reserving trap sits. A loss trend selected off a book still earning through the tail of the hard market is calibrated to a rate environment that no longer exists at the point of sale. As the softer written rates earn in, the accident-year loss ratio drifts upward even with frequency and severity flat, because earned premium is falling faster than the loss side is improving. A single blended trend selection across the earned book, without tracking the rate vintage behind each cohort of earned premium, sets Q2 and Q3 reserves against a market the written book has already left.

The capital picture explains why property moved this fast. Gallagher Re put dedicated reinsurance capital at a record $648 billion at year-end 2025, up 11%, with non-life insurance-linked securities capital at a record $135 billion and $15.6 billion of catastrophe bond issuance by mid-June. Aon separately reported global reinsurer capital of $790 billion at March 31. Natural catastrophe losses of $38 billion through June 15 ran below the ten-year average.

Record capacity meeting a light loss year is a supply story, not evidence that property loss cost trends improved, and the same mechanism ran through the January renewals before accelerating at mid-year.

Casualty Is Firming While Property Softens, and the Blend Nets Them

Casualty came through the same renewal season differently. Gallagher Re found pricing broadly stable rather than soft, with reinsurers differentiating between cedants more sharply than in prior cycles. US commercial auto stayed hardest, with loss-affected excess-of-loss layers up 10% to 15%. Healthcare liability kept climbing with severity, led by hospital professional liability. Workers' compensation stayed profitable even as large-loss frequency and severity rose.

Swiss Re's behavior is the carrier-level confirmation. It posted $2.6 billion of H1 net income and an 81.1% combined ratio and still cut casualty treaty volume 5.9% at the June and July renewals. A reinsurer beating its own targets does not give up premium unless it has concluded the price does not cover where loss costs are going.

The reserve record supports that read. Milliman put adverse prior-year development across other liability occurrence, commercial auto liability, non-proportional reinsurance liability and product liability at $15.8 billion in 2024, a record for those segments, and Swiss Re Institute has measured social inflation at 5.4% a year between 2017 and 2022 against 3.7% economic inflation. Triple-I and Milliman forecast general liability and commercial auto as the only major lines still above a 100 combined ratio through 2026 and 2027.

The two effects do not net symmetrically, which is what the blended combined ratio cannot show. Property is the larger share of most diversified commercial books, so its decline pulls the blend harder than a smaller casualty layer pushes back, and it arrives gradually as rate earns in. Casualty's reserve risk arrives at once, the way the 2024 development did. Verisk's Q1 industry result, a $15.8 billion underwriting gain and a 92.4% combined ratio against an $864 million loss a year earlier on net written premium growth of 2.9%, is what those two trends look like once averaged into one figure.