Palomar Holdings' adjusted combined ratio worsened 3.6 points to 76.7% in the second quarter of 2026, from 73.1% a year earlier, while gross written premium grew 27.0% to $630.5 million and adjusted net income rose 31.4% to $63.8 million (Reinsurance News, August 2026).

That is not a contradiction. It is what a fast-growing specialty book looks like in the year before its loss picks either prove out or get corrected.

Key Takeaways

  • The loss ratio moved 8.8 points, to 34.5% from 25.7%, while the combined ratio moved only 4.5. The difference is an expense-ratio offset from net earned premium growing 59.5%, not an underwriting improvement.
  • Prior-year favorable development was $14.3 million against $99.4 million of current-quarter attritional losses, roughly 14% of the pick, so reserves landed close to where they were set rather than well above.
  • Crop premium grew 96% and surety 236%, the two least-seasoned lines in the book, against 1% year-to-date growth in earthquake, the line with more than a decade of accident years behind it.
  • Commercial earthquake faces average rate decreases exceeding 20% in large layered accounts, which CEO Mac Armstrong described as "below what we consider technical pricing levels."
  • Casualty carries the one clean rate-over-trend figure: ISO general liability loss cost up 4.4% across the top five states against a blended 7.5% rating increase achieved.

Decomposing the 3.6-Point Move

The total loss ratio came in at 34.5%, up from 25.7% a year earlier (TradingKey, August 2026). The expense ratio went the other way, falling to 48.8% from 53.1% as net earned premium of $287 million, up 59.5%, outpaced fixed-cost growth and diluted acquisition costs across a larger earned base. Netting the two produces the reported combined ratio of 83.3%, up from 78.8%.

The adjusted figure management leads with, 76.7% against 73.1%, strips non-cash items including changes in fair value of crop reinsurance derivatives and intangible amortization.

Underneath both sits $99.4 million of attritional losses, the current-accident-year pick before any development is applied (StockStory, August 2026). That is the figure least distorted by what was released from prior years. A carrier growing net earned premium 59.5% in a year will show expense-line improvement almost mechanically; the loss line carries the pricing signal, and it moved 8.8 points the wrong way.

MetricQ2 2026Q2 2025Change
Gross written premium$630.5M$496.3M+27.0%
Adjusted combined ratio76.7%73.1%+3.6 pts
Reported combined ratio83.3%78.8%+4.5 pts
Loss ratio34.5%25.7%+8.8 pts
Expense ratio48.8%53.1%-4.3 pts
Total favorable PYD$14.3M

A Release That Has Thinned to 14% of the Pick

Favorable development is a point in the carrier's favor: the 2025 picks were not too thin. The quarter carried $14.3 million, split $14.1 million attritional and $0.2 million catastrophe, concentrated in inland marine and property and crop.

Size matters as much as direction. Against $99.4 million of current-quarter attritional losses, a $14.3 million release is roughly 14% of the pick. That reads as reserves landing close to where they were set, not a book reserved conservatively and now paying out on that conservatism.

The distinction matters because releases are the mechanism a growing specialty carrier uses to hold a combined ratio steady while a book grows into its true loss cost. A large, reliable redundancy absorbs a worsening current-year pick for several quarters before the ratio shows it. A release at 14% of attritional losses leaves much less room, at the exact point the book is growing fastest.

Where it is growing compounds that. Crop gross written premium surged 96%, with full-year guidance lifted above $400 million from roughly $320 million, and CFO Chris Uchida told analysts that "crop losses show up earlier than the premiums," guiding to a full-year crop loss ratio in the "mid to upper 30s" (Investing.com, August 2026). Casualty grew 37% across seven niche lines, surety and credit 236% to roughly $39 million on the Gray Surety acquisition, and motor truck cargo 22% behind a newly approved 13% California rate increase. None of those carries the development maturity of the earthquake franchise, where residential retention exceeds 96%.

The pattern is not Palomar's alone this quarter. Kinsale cut commercial property premium 32.7% to protect a 75.5% combined ratio, RLI's property ratio improved 5.3 points to 56.8% while casualty grew 10.6% to $339.0 million on a deteriorating segment ratio, and W.R. Berkley held a 1.9-point gap between underlying and reported. In each case the healthiest-margin segment was the slowest-growing one.

The Line With the Most History Is the One Being Declined

Palomar's own account of where it is not growing carries more pricing information than the 27% headline does.

Commercial earthquake, roughly 36% of the earthquake franchise, faces average rate decreases exceeding 20% in large layered accounts. Armstrong described pricing in certain large-account segments as having fallen "below what we consider technical pricing levels," adding that the company hopes "pricing at these levels is indicative of market approaching a bottom." Year-to-date earthquake written premium rose just 1%.

That near-freeze is the company declining volume in the one line where it has the longest loss history and the clearest read on technical adequacy. "We will always sacrifice premium for profitability, but we will not sacrifice profitability for premium," Armstrong said. On casualty, management offered the supporting arithmetic: ISO general liability loss cost across the top five states ran 4.4% against a blended 7.5% rating increase achieved, more than 300 basis points of rate over trend.

The 27% blends those two environments rather than describing either. Rate adequacy is verifiable in earthquake and in casualty, and in earthquake the verdict was to stop writing. It is hardest to verify in crop, surety and builders risk, which is where the growth sits, and where the loss picks rest on the shortest internal claims history. The 3.6-point deterioration is consistent with a company pricing conservatively where it has data and absorbing loss-ratio drag from scaling lines faster than the earned premium base matures around them.