Marsh released its Q2 2026 Global Insurance Market Index on July 23, and the headline is a soft market almost everywhere: global commercial rates fell 6%, the eighth consecutive quarter of decline, with property down 12% (Insurance Journal, July 23, 2026). One line broke the pattern. US casualty rates rose 7%, and 11% excluding workers’ compensation. Casualty rates fell in every region outside the United States. Marsh attributed the increase to “ongoing claims severity and litigation pressures in the US.”
That single divergence is the story. When capacity is abundant and carriers are competing rates down across property, marine, financial, and cyber, and they are still pushing US casualty up double digits net of comp, they are telling you the loss-cost trend under liability is running ahead of what price relief elsewhere can offset. The Q2 print follows a 9% US casualty increase in Q1 2026, so this is the second straight quarter of high-single to double-digit casualty pricing while the rest of the market corrects downward.
Who it affects
This matters most to buyers who retain casualty rather than transfer it: self-insured employers carrying general liability and auto liability inside a large deductible or SIR, single-parent and group captives writing third-party or excess casualty, public-entity pools reserving their own GL and auto layers, and hospital captives carrying professional liability. These buyers never receive a premium quote for the retained layer. The carrier rate increase they read about in the trade press does not hit their income statement as a higher premium. It hits their balance sheet as a higher expected loss, or it should.
The reserve mechanism
A carrier casualty rate change is a repriced view of forward loss cost. For a self-insured, that same view belongs in two places: the a priori (expected) loss ratio in a Bornhuetter-Ferguson estimate, and the severity assumption behind the tail factor. If your casualty a priori is still anchored to 2019 to 2021 development and carriers have repriced the same exposure up roughly 9% then 7% in two quarters, your expected loss ratio is stale and your IBNR is understated for the most recent accident years. The excess and umbrella layers are where this bites first. Selective capacity at the top of the tower pushes retentions up at renewal, and every dollar of attachment increase moves expected loss out of the transferred layer and into the SIR the buyer reserves directly.
The workers’ comp exception is the tell. US casualty rose 7% with comp and 11% without, so WC is dragging the number down. That gap says the pressure is severity-driven liability (auto, GL, umbrella, medical professional), not frequency-driven comp, which stays soft on benign frequency. Reserve the divergence, not the blended number.
What this means for your next review
Do not let a falling property renewal convince the board that risk is falling. Property rate relief and casualty reserve strengthening can, and right now should, coexist. For the next study, ask your actuary whether the casualty a priori loss ratio reflects current carrier severity selections or pre-2023 development, and model how much expected loss shifts into your retained layer if umbrella attachment rises at renewal. Watch the Q3 Marsh index: a fourth straight quarter of US casualty increases would confirm severity, not a one-time correction. See also What’s Actually Driving Your IBNR Higher and the Five Leading Indicators of Adverse Development, and our read on the Q1 print, Marsh: US Excess Casualty Up 18% as the Tower Reprices.
Sources
- Insurance Journal, “Q2 Global Commercial Insurance Rates Keep Dropping, Except for US Casualty,” July 23, 2026: https://www.insurancejournal.com/news/national/2026/07/23/878716.htm
- Marsh, Global Insurance Market Index Q2 2026: https://www.marsh.com/en/services/international-placement-services/insights/global-insurance-market-index.html
- CAS and Triple-I, “New CAS and Triple-I Analysis Quantifies Impact of Legal System Abuse on Liability Insurance”: https://www.casact.org/article/new-cas-and-triple-i-analysis-quantifies-impact-legal-system-abuse-liability-insurance