Arch's 83.5% Combined Ratio Hides a Primary Loss-Pick Slip

Arch Capital Group's headline 83.5% group combined ratio for the second quarter of 2026 blends two segments moving in opposite directions: reinsurance ran at 77.5% while the insurance segment jumped to 98.5% from 93.4% a year earlier (Arch Capital Q2 2026 earnings release, July 28, 2026). Strip out catastrophes and prior-year reserve releases and the group's underlying accident-year loss ratio still widened to 82.5% from 80.9%, a signal the composite number conceals.

Arch reported the results on July 28, 2026, in its earnings release and accompanying financial supplement, filed with the SEC as exhibit 99.1 to a Form 8-K. Net income available to common shareholders came to $1.047 billion, or $3.00 per share, an 18.0% annualized return on average common equity, and book value per share rose 2.8% from the first quarter to $68.04. Those are strong headline numbers for a Bermuda-based specialty writer that blends primary insurance, reinsurance, and mortgage guaranty under one roof. The group combined ratio of 83.5%, though, is a premium-weighted average of a reinsurance book still generating outsized margin and a primary insurance book that has stopped doing so, and the two halves of that average are telling different stories about where Arch is in the underwriting cycle.

Decomposing the 83.5%: Two Segments, One Number

The group combined ratio is the number that leads every wire story on Arch's quarter, and on its own it looks like a continuation of a strong run: 83.5% is well inside underwriting profit territory, and it compares to 81.2% in the second quarter of 2025 (Arch Capital Q2 2026 earnings release, July 28, 2026). But a combined ratio is a premium-weighted blend across segments, and Arch's three segments did not move together. The insurance segment, Arch's primary commercial and specialty book, posted a 98.5% combined ratio in the quarter, up more than five points from 93.4% a year earlier, with underwriting income collapsing to $27 million from $129 million (Insurance Business America, July 2026). The reinsurance segment ran at 77.5%, still comfortably profitable though down from a segment that had produced record pretax underwriting income in the prior-year period, with segment underwriting income of $410 million against $451 million a year earlier. Mortgage, the smallest of the three, posted a 22.8% combined ratio, its usual structurally low-loss profile, with $220 million of underwriting income.

Catastrophe losses explain part of the insurance segment's deterioration but not all of it. Catastrophic activity added 7.6 points to the insurance segment's loss ratio in the quarter, against 2.9 points in the prior-year period, a swing of 4.7 points that alone would explain most of the year-over-year combined ratio gap (Insurance Business America, July 2026). Group-wide pretax catastrophe losses net of reinsurance and reinstatement premiums totaled $201 million, with 8.0 points of the group's cat load sitting in insurance and 2.8 points in reinsurance (Arch Capital financial supplement, July 2026). The insurance segment's expense ratio also crept higher, to 35.5% from 33.6%, a rise the company attributed partly to transitional integration costs tied to the 2024 acquisition of the U.S. MidCorp and Entertainment insurance businesses from Allianz (Insurance Business America, July 2026). Cat load and expense drag are both real, but neither is the number that should worry a reserving actuary reading this filing. That number sits one layer deeper.

The Number That Survives the Cat and Reserve Noise

Cat losses are volatile by nature and prior-year development is, definitionally, about business already written; neither speaks directly to whether Arch is pricing the current accident year adequately. The metric that does is the accident-year loss ratio excluding both catastrophes and reserve development, and at the group level it moved from 80.9% in Q2 2025 to 82.5% in Q2 2026, a roughly 1.6-point widening (Arch Capital financial supplement, July 2026). That is a small number in isolation, and Arch's own trade-press coverage barely mentions it next to the more dramatic segment combined-ratio swing. But it is the one figure in the release that cannot be explained away by a bad cat quarter or a generous reserve release, because both of those have already been stripped out. A 1.6-point group-level drift, on $3.985 billion of net premium earned in the quarter, works out to roughly $64 million of incurred loss that a cat-free, development-free comparison did not have a year earlier.

The segment breakdown sharpens the picture further. Arch's own supplement shows the insurance segment's combined ratio excluding catastrophes and prior-year development running at 91.6%, versus 79.9% for reinsurance (Arch Capital financial supplement, July 2026). A 12-point gap in underlying, noise-stripped combined ratio between the two segments is not explained by mix alone; it points to the primary book absorbing loss trend, social inflation, and casualty severity at a rate its 2026 rate actions have not fully offset, while the reinsurance book, priced off cedents' own loss experience and typically repriced annually at renewal, has kept pace. That gap is the real story inside the 83.5% headline, and it is the kind of divergence a group combined ratio is structurally built to hide.

Sizing the $165 Million Reserve Cushion

Favorable prior-year reserve development added $165 million to the quarter's underwriting result group-wide, a release that flatters every ratio discussed above (multiple sources citing Arch Capital's Q2 2026 disclosures, July 2026). Broken out by segment as a percentage-point loss-ratio benefit, mortgage carried the largest favorable adjustment at 15.7 points, reinsurance carried 5.3 points, and insurance carried a comparatively modest 1.4 points (Arch Capital financial supplement, July 2026). That distribution matters for how much of the group's reported margin is earned in the current year versus released from prior years. Mortgage's reserve releases are a known and recurring feature of that line, where loss emergence patterns are short and well modeled; the reinsurance segment's 5.3-point release is more consequential, because it means a meaningful share of reinsurance's already-strong 77.5% combined ratio is not current-year underwriting profit but a favorable revision to how past accident years are expected to develop. Strip that release out of reinsurance and the segment's own noise-adjusted combined ratio of 79.9% is the more honest read of its current pricing adequacy, still comfortably better than insurance's 91.6%, but not by quite the margin the headline 77.5% versus 98.5% comparison suggests.

The insurance segment's own 1.4-point favorable development is small enough that it barely offsets the segment's underlying deterioration, which is itself informative: a primary book with slipping current-year loss picks is not being rescued by reserve releases from its own back book the way reinsurance is. That asymmetry, a shrinking cushion on the segment that needs one most, is the clearest evidence that Arch's group margin in Q2 2026 leaned more heavily on reinsurance and mortgage releases than on primary insurance's own underwriting.

Segment	Combined ratio, Q2 2026	Ex-cat, ex-dev combined ratio	Cat load (points)	Favorable PYD (points)
Insurance	98.5%	91.6%	7.6	1.4
Reinsurance	77.5%	79.9%	2.8	5.3
Mortgage	22.8%	39.8%	—	15.7

Reading the 6.9% Premium Decline as a Discipline Signal

Net premiums written fell 6.9% group-wide to $4.05 billion, and net premiums earned fell 8.1% to $3.985 billion (Insurance Business America, July 2026). In the insurance segment specifically, net premiums written were 5.1% lower year over year, though Arch has said the underlying decline, excluding non-renewals of certain acquired programs, was closer to 1.8% (Insurance Business America, July 2026). Group gross premiums written were roughly flat, down 1.1% to $6.13 billion, which means the net decline is concentrated in what Arch chose to retain and cede, not in what it wrote at the top line. In a market where property-catastrophe reinsurance rates have been softening through 2026, a diversified underwriter pulling back on net retained primary premium while gross volume holds is at least consistent with a company declining to chase rate-inadequate business rather than one simply losing market share. Arch's own P&C soft-market playbook, discussed in actuary.info's coverage of reserve adequacy through the 2026 pricing downturn, has flagged premium retreat as one of the more reliable tells that a carrier's actuaries, not just its marketing language, are driving renewal decisions.

Arch CEO Nicolas Papadopoulo characterized the quarter as "a strong quarter, driven by solid underwriting performance across our three segments, reflecting the continued strength of our diversified platform and disciplined execution across the enterprise" (Arch Capital Q2 2026 earnings release, July 28, 2026). That framing is defensible at the group level and at the reinsurance and mortgage segment level specifically. It is a harder sentence to apply unmodified to the insurance segment, where the underlying, noise-adjusted combined ratio of 91.6% and a 12-point gap versus reinsurance suggest primary loss picks are the part of "disciplined execution" still catching up.

Why Segment-Level Accident-Year Picks Are the Better Adequacy Gauge

For a group like Arch that blends primary and assumed risk under one reported combined ratio, the headline number is a poor proxy for pricing adequacy precisely because it can offset a deteriorating segment with a strong one, and because both segments carry different amounts of reserve cushion available to smooth the current year's result. A reinsurer with a large back book of short-tail property business can release reserves from a benign recent cat year to flatter this year's combined ratio without that release saying anything about how the current treaty year is priced. A primary casualty writer, by contrast, has less of that cushion available in the years immediately following a soft-market underwriting vintage, because casualty loss development is slow enough that adverse trend in a recent accident year often has not fully emerged by the time the reserve is tested. That asymmetry is exactly what shows up in Arch's numbers: reinsurance carrying 5.3 points of favorable development against insurance's 1.4, on a primary segment whose ex-cat, ex-development combined ratio is already 12 points worse than its reinsurance counterpart.

The practical implication for anyone modeling Arch, or benchmarking a comparable diversified writer, is to weight the segment-level, noise-adjusted accident-year loss ratio far more heavily than the group combined ratio when assessing whether current pricing is keeping pace with loss trend. Arch's own group ex-cat, ex-development loss ratio drifting from 80.9% to 82.5% is the aggregate warning sign; the segment split showing that drift concentrated in a primary book running a 91.6% underlying combined ratio, with a shrinking reserve cushion to lean on, is the more actionable one. RenaissanceRe's own Q2 2026 quarter showed a comparable divergence between its property and casualty retro books, a pattern actuary.info examined in its coverage of RenaissanceRe's 72.8% combined ratio, and Chubb's Q2 2026 release showed the same casualty-versus-property split working through its own book, covered in actuary.info's analysis of Chubb's 83.8% combined ratio meeting a softening casualty book. The pattern recurring across three large, diversified carriers in the same reporting cycle is itself a signal: 2026's soft market is pressuring primary casualty and specialty pricing broadly, even where reinsurance and property lines still look comfortably profitable on a blended basis.

What the Segment Split Implies for Reserving and Pricing

The reserving implication is straightforward: an actuary reviewing Arch's insurance segment reserves for adequacy should treat the 1.4-point favorable development on that segment as thin cover, not confirmation that current accident-year picks are conservative. A segment with a 91.6% underlying combined ratio and a widening loss ratio trend has less room to absorb a future adverse development finding than a headline 98.5% figure, inflated by a 7.6-point cat load that will not repeat every quarter, might suggest to a casual reader. The pricing implication runs the other direction: if Arch's insurance segment rate actions in the back half of 2026 do not close the gap implied by an 82.5% group accident-year loss ratio running above 80.9%, the segment's combined ratio in coming quarters will show it again, cat losses or not, because the ex-cat, ex-development figure is specifically designed to strip out the noise that made this quarter's headline number look worse than the underlying trend actually is.

None of this reads as a crisis for Arch. Reinsurance and mortgage are carrying the group comfortably, book value per share grew, and the company's 18.0% annualized return on equity is a result most of the P&C sector would take. The point is narrower and more specific to how the release should be read: the 83.5% group combined ratio is the number every headline led with, and it is the least informative number in the release for judging whether Arch's primary underwriting is keeping pace with loss trend in a softening market. The 91.6% insurance-segment, ex-cat, ex-development figure, and its 12-point gap against reinsurance's 79.9%, is the one worth tracking into the third quarter.