Property Releases, Not Pricing, Are Carrying Q2 2026 Combined Ratios

Short-tail property redundancy, not improved current-year casualty pricing, is what kept Q2 2026 P&C combined ratios looking strong: RenaissanceRe booked $257.5 million of favorable property development against a 103.3% combined ratio in Casualty and Specialty (RenaissanceRe, July 2026), and the same split shows up across nearly every Q2 reporter with disclosed prior-year detail.

FactSet's Schedule P analysis, published as the second-quarter reporting wave got underway, flagged the industry-wide version of this pattern: incurred losses escalating from accident year 2024 to accident year 2025 even as the industry's headline combined ratio improved to 92.4% in the first quarter of 2026 from 99.2% a year earlier (FactSet, June 2026). The 2024 calendar-year combined ratio of 96.8% was the only reading below 97.0% since 2013, against a ten-year average of exactly 100.0% (FactSet, June 2026). Fitch Ratings has already priced in where that trend runs out: its full-year 2026 combined ratio projection sits at 96% to 97%, built on a more typical hurricane season and materially less favorable reserve development than carriers have been booking through the first half of the year (Fitch Ratings, 2026). The gap between a sub-90% quarterly print and a high-90s full-year forecast is the release dependence turning into a forecasting problem.

The Property-Casualty Split Across Four Q2 Reporters

Reading Schedule P detail at the carrier level, rather than the industry aggregate, makes the mechanism concrete. Four Q2 2026 reporters with disclosed segment-level prior-year development show the same pattern in different proportions: short-tail property and specialty lines releasing reserves from redundant vintages, casualty and long-tail commercial lines running flat to markedly worse.

Carrier	Segment	Q2 2026 combined ratio	Q2 2025 combined ratio	Prior-year development
RenaissanceRe	Property	27.1%	27.4%	$257.5M favorable
RenaissanceRe	Casualty & Specialty	103.3%	101.8%	$58.0M adverse
RLI Corp	Property	56.8%	62.1%	concentrated within $35.1M favorable total
RLI Corp	Casualty	99.3%	96.5%	concentrated within $35.1M favorable total
Travelers	Business Insurance	86.8%	93.6%	$319.0M favorable (vs. $79.0M a year earlier)
W.R. Berkley	Consolidated	90.0%	91.6%	near breakeven ($1M net favorable, Q1 2026)

RenaissanceRe's split is the cleanest read because the company discloses prior-year development by segment: $257.5 million of favorable development in Property against $58.0 million of adverse development in Casualty and Specialty, a swing large enough on its own to explain most of the gap between the segments' 27.1% and 103.3% combined ratios (RenaissanceRe Q2 2026 earnings release, July 2026). A portion of that Casualty and Specialty figure is a $54.0 million reclassification of reserves tied to the Baltimore Bridge collapse rather than fresh current-accident-year deterioration, an insured marine liability loss this site has tracked as it keeps migrating between segments as carriers finalize which lines absorb it (actuary.info's Baltimore Bridge marine reserving benchmark). Strip that reclassification out and the underlying casualty deterioration is smaller than 103.3% implies, but it is still a deterioration, not an improvement, and it sits on the opposite side of the ledger from Property's release.

RLI does not disclose prior-year development by segment, but the direction of its two largest books tells the same story. Property improved 5.3 points to a 56.8% combined ratio even against active catastrophe activity in the quarter, while Casualty, RLI's largest segment by premium, widened from 96.5% to 99.3% as underwriting income fell to $1.7 million from $8.3 million (RLI Q2 2026 earnings release, July 2026, as detailed in actuary.info's RLI Q2 2026 reserve-dependence analysis). RLI's $35.1 million of total favorable development, up from $24.4 million a year earlier, is not broken out by line, but reserve credits in a specialty book like RLI's concentrate disproportionately in casualty, the segment with the longest claim tail and the most room for loss picks set three to five years ago to prove conservative. Strip the total development out entirely and RLI's implied current-accident-year combined ratio runs closer to 94%, not the reported 85.6%, a gap that has widened from roughly 6.1 points a year earlier to 8.4 points in the current quarter.

Travelers looks different on the surface, since its $319.0 million of favorable Business Insurance development, more than four times the $79.0 million booked a year earlier, sits inside a segment that is itself casualty-heavy (general liability, umbrella, and commercial auto alongside property and package business). That figure complicates a clean property-versus-casualty read, and it is the exception worth naming directly: not every casualty-adjacent book is deteriorating, and Business Insurance's 86.8% combined ratio, a 6.8-point improvement, benefited from both lower catastrophe losses of $518 million (down from $927 million) and that larger reserve credit (Travelers Q2 2026 earnings release, July 2026). Whether that $319.0 million reflects genuine casualty improvement or the same short-tail property and auto physical damage lines inside a blended commercial segment doing the releasing is not disclosed at the line level in the earnings release, which is itself evidence for why Schedule P's statutory line-of-business triangles, not the segment reporting carriers choose for investor communications, are the more reliable diagnostic.

A Depleting Asset: Why the 2021-2023 Property Redundancy Has a Shelf Life

The property releases carrying Q2 2026 results trace back to a specific and closing window of accident years. Property and other short-tail lines written in 2021 through 2023, during a hard market defined by post-pandemic demand, rapidly rising reinsurance costs, and loss picks set conservatively against elevated catastrophe uncertainty, have consistently emerged better than reserved as those years matured. Short-tail property losses typically reach 85% to 95% of their ultimate value within 24 to 36 months of the accident year closing, per the loss-development patterns NAIC's Schedule P Parts 2 through 4 are built to expose (NAIC Schedule P instructions). That means most of the redundancy embedded in the 2021 to 2023 property vintages has already been recognized in the calendar years since, and what is left to release in any given 2026 quarter is a shrinking residual rather than a renewable source of margin.

Casualty lines do not work the same way. General liability, commercial auto liability, and umbrella and excess coverage typically take 8 to 10 or more years to reach ultimate, which means the 2021 through 2023 accident years are still comparatively immature and exposed to loss-trend surprises, chiefly social inflation and outsized jury verdicts, that were still building when those years were priced. That asymmetry is the mechanical reason property releases and casualty firming can appear in the same Schedule P filing for the same carrier in the same quarter: one set of accident years is old enough to be known, the other is not. W.R. Berkley's chief executive framed the casualty side of that trade directly on the company's Q2 call: "We continue to see attractive opportunities across select liability lines" (W. Robert Berkley, Jr., W.R. Berkley Q2 2026 earnings release, July 2026), a comment that reads as confidence in current pricing on lines other carriers are reserving more cautiously against, not as evidence the reserving caution is misplaced.

The practical question for a reserving actuary is how many more quarters of that property cushion remain. If 24 to 36 months is the typical window in which short-tail losses mature to their ultimate value, then accident year 2023, the youngest vintage still capable of producing a meaningful release, is already past that window as of mid-2026. Accident years 2024 and 2025 are still developing, but they were written into a softening property-cat rate environment, down 14.7% at the January 2026 renewal by one broadly cited industry benchmark, which means the loss picks underlying those years carry less of the conservative cushion that made 2021 to 2023 releases so reliable. The redundancy is not disappearing on a fixed calendar date, but its most productive vintages are aging out faster than new, equally redundant vintages are being written behind them.

The Calendar-Year Versus Accident-Year Gap Is the Soft-Market Tell

The clearest quantitative signal of how much work reserve releases are doing is the gap between a carrier's reported, calendar-year combined ratio and its current-accident-year combined ratio once development is stripped out. At RLI, that gap widened from 6.1 points a year ago to 8.4 points in the second quarter of 2026, even as the reported combined ratio itself moved less than a point, from 84.5% to 85.6% (RLI Q2 2026 earnings release, July 2026). A reported ratio that looks nearly flat year over year while its underlying, ex-development ratio widens by more than three points is a soft-market signature: current-year pricing is losing ground to loss-cost trend, and reserve credits are absorbing the difference before it reaches the headline number.

Travelers shows the same mechanism from the other direction. Its underlying combined ratio, which strips out both catastrophe losses and prior-year development, improved only 0.6 points in the second quarter even as the reported combined ratio improved 6.7 points, with 2.5 of those points attributable to higher net favorable prior-year development and 3.6 points to lower catastrophe losses (Travelers Q2 2026 earnings release, July 2026). Put differently, roughly two-thirds of Travelers' headline improvement came from catastrophe timing and reserve credits, not from the current-year loss ratio getting better. That is not a criticism of Travelers' discipline; the underlying ratio's 0.6-point improvement is itself real progress. It is a reminder that the number moving the stock and the analyst consensus, the 83.6% reported figure, is not the number that predicts where the combined ratio lands once catastrophe activity normalizes and the release cushion thins.

Capital and Dividend Math When the Release Tailwind Fades

Reserve releases are not merely an accounting artifact; they are funding capital return decisions in real time. RenaissanceRe repurchased $350.0 million of common shares in the same quarter its Property segment released $257.5 million of reserves, and W.R. Berkley returned $223 million in dividends and $112 million in buybacks in a quarter where its consolidated combined ratio improved to 90.0% partly on the strength of a catastrophe load that fell to $62.4 million from $99.2 million a year earlier (RenaissanceRe and W.R. Berkley Q2 2026 earnings releases, July 2026). None of that capital return is improper; both companies remain well capitalized by any conventional rating-agency metric. But a capital-return pace set against a combined ratio that is being flattered by a finite reserve cushion is a pace that has to reset lower, or be justified by stronger current-accident-year margin, once the cushion runs out.

Fitch's 96% to 97% full-year 2026 combined ratio projection is the number that makes that reset concrete. If the industry's calendar-year combined ratio converges toward that range as 2026 progresses, quarterly prints in the mid-80s to low-90s that current investors are pricing as the new normal will look instead like a temporary release-driven trough. Boards setting buyback authorizations and dividend increases against today's headline numbers, rather than against Fitch's forward view or their own underlying, ex-development combined ratio, are effectively spending a one-time reserve credit as if it were recurring underwriting margin. That is precisely the dynamic this site flagged in the broader industry reserve picture earlier this year, when a decade-high profit figure sat alongside early signs that reserve durability, not underwriting improvement, was doing the heavier lifting (actuary.info's Q1 2026 reserve-quality read on the soft-market combined ratio).

Short-Tail Specialty Versus Long-Tail Commercial: Divergent Exposure

How exposed a carrier is to the fading of this tailwind depends heavily on its line mix. RLI and RenaissanceRe's Property segment sit at one end of the spectrum: short-tail books where losses are known relatively quickly, where the 2021 to 2023 redundancy has already mostly emerged, and where the next source of reserve margin has to come from newer, less conservatively reserved accident years written into a softening property-cat rate environment. Casualty-heavy commercial writers, including RenaissanceRe's own Casualty and Specialty segment and the general liability and umbrella books embedded in Travelers' Business Insurance and W.R. Berkley's insurance segment, sit at the other end: reserves for 2021 to 2023 are still developing, which means both the upside (further redundancy, if severity trend cooperates) and the downside (further strengthening, if social inflation does not) remain live for several more years.

That asymmetry is exactly what the industry's broader casualty triangles have been showing since 2024, when several major commercial writers first began booking material adverse development on general liability and umbrella accident years written between 2016 and 2021 (actuary.info's 2024 casualty triangle adverse-development analysis). The medical malpractice sector, where the industry-wide combined ratio pushed past 105% as jury verdict severity accelerated, is the extreme case of the same long-tail dynamic playing out years ahead of where general liability and umbrella now sit (actuary.info's medical malpractice reserve-crisis coverage). A carrier writing predominantly short-tail property and specialty lines can expect its reserve position to normalize within a few years of a hard-market vintage maturing out. A carrier writing predominantly long-tail commercial casualty is still absorbing the consequences of pricing and reserving decisions made five to eight years earlier, which is why the Q2 2026 combined ratio split by segment is a more useful predictor of 2027 and 2028 results than the consolidated number any single carrier reports this quarter.