A 9% growth quarter, a record reserve cushion, and a government dairy program that showed both the power and the limits of Skyward's data edge.
SKWD · Earnings Call · 2026-11-06
Skyward's Diesel Haircut
On the surface, Skyward Specialty's third quarter was another clean print: gross written premiums rose 9% to $803 million, fee-generating gross written premiums rose more than 18% to $517 million, diluted operating EPS increased 22% to $1.28, and the consolidated combined ratio was 89.9% — or 88.1% ex-cat. The expense ratio improved to 25.7%. Through nine months, operating income was roughly $175 million and annualized operating ROE exceeded 20%. The fundamental shape through July was already visible: total revenue reached $490 million, up 53% year over year, while operating margin was 13.0%, down 2.6 percentage points.
Yet the stock sits about 8% below its August 6 peak of $65.63 and is roughly flat over the last 90 trading days. The reason is not the underwriting result. It is a single voluntary federal program — and a diesel price chart.
A Clean Quarter, One Dirty Variable
The program is Skyward's dairy livestock quota-share solution, built with the major approved insurance providers. It is profitable, uncorrelated with the P&C cycle, and cleverly structured: Skyward cedes a portion of the risk to retrocessional reinsurers and captures value through ceding commissions. But the underlying federal product is voluntary. Farmers buy it to lock in margins. When diesel moved from roughly $3.50-$3.70 in January and February to $6.25 by October, farmer cash flow got squeezed and take-up fell.
“We effectively saw just a change in take-up rate amongst farmers as diesel prices started to move... we just simply saw a lower uptake.” — Andrew Robinson, Chairman and Chief Executive Officer · 2026-11-06
That produced a negative net earned premium true-up in the quarter. Management also preemptively haircut expected premium for the 2026-27 treaty year. They were careful not to call it a one-off, but they were equally clear that this is a demand-timing issue rather than a loss or credit problem.
In this market, as far as we can tell, we have created this product. We are one of one. It's an incredibly profitable product. It's uncorrelated for us.
That is the tension. The dairy program is exactly the sort of company-unique, high-return niche that makes Skyward interesting. It is also exactly the sort of program that can make a 9% growth quarter look noisier than the underlying business.
The Fee Engine Is No Longer a Sidecar
Apollo gross written premiums rose 3% to $135 million, with Specialty Risk Syndicate 1969 up 11%, partly offset by timing and seasonality in ibott Syndicate 1971. The more important number was fee-generating GWP: up 18% to $517 million, including 70% growth in partner syndicates. Two new partner syndicates launched in the quarter, including one with Generali, and Skyward is not seeding them with its own capital. That is the capital-light model working.
The partner syndicate platform now looks less like a curiosity and more like a genuine recurring-fee engine. Underwriting fee income was $11 million in the quarter and roughly $34 million through nine months, against full-year guidance of $30 million-$35 million. Andrew Robinson was explicit that managed premiums — the premium that drives fee earnings — should outgrow reported GWP:
“our growth in managed premium, which is the... capture of the premium that drives our fee earnings growth should outgrow sort of our reported GWP.” — Andrew Robinson, Chairman and Chief Executive Officer · 2026-11-06
This matters because it changes the quality of growth. Fee income is capital-light, recurring and less exposed to reserve risk. It also gives Skyward a way to participate in Lloyd's innovation without putting the balance sheet behind every dollar.
What Management Stopped Saying
Earlier this year, loss cost trends was the company's most prominent keyword, and commercial auto was a top-three concern. Neither appears in the current top-30 keyword set. That is not an oversight; it is a signal.
The reserve review explains why. Skyward did not take any prior-year reserve development in the quarter, IBNR is now 77% of total reserves — the highest in company history — and paid-to-incurred is running around 62%, the lowest it has been. Mark Haushill:
“We did not recognize any prior year reserve development during the quarter... initial indications are positive with emergence as expected or better from occurrence liability, property, surety, and A&H.” — Mark Haushill, Chief Financial Officer · 2026-11-06
The remaining pressure is in portfolios Skyward has already exited: commercial auto and claims-made professional liability. That is a very different message from the one the company was delivering a year ago, when loss-cost inflation dominated the conversation.
The new top concern is the competitive environment. In the E&S business, property submissions were up 60% in the quarter, but the hit ratio fell by a third and average premium fell 20%. Primary liability retention dropped 6 points. Robinson described a market where more business is being marketed, but not always sensibly:
“There's one particular competitor that effectively has kind of a fast track on certain liability classes... It's like welcome to the world of casualty where... you should be really responsible with your capital.” — Andrew Robinson, Chairman and Chief Executive Officer · 2026-11-06
This is a recurring complaint, not a new one. A prior call had Robinson saying “the speed at which things have moved is extraordinary” — Andrew Robinson, Chairman and Chief Executive Officer · 2026-05-07 about the softening market. The difference now is that the competitive pressure is showing up in measurable hit ratios and premium, not just rhetoric. Skyward is choosing to defend its book rather than chase.
The De-Rating vs. the Compounding
The quiet compounder is Accident & Health. Premium was up 57% in the quarter, and the business has more than doubled over two years. Its 76% loss ratio ranked fourth best among 30 carriers writing more than $250 million of stop-loss premium in the NAIC's 2025 analysis, roughly 15 points better than the industry average. The formula is a focus on smaller employers, group captives and distinctive medical cost management — not simply price.
That is not a new story, either. Robinson said in August: “We've kind of just hit it on the product market fit, right?” — Andrew Robinson, Chairman and Chief Executive Officer · 2026-08-05 The market may be underestimating how durable that fit is. The same goes for Bionic underwriting: submissions are up 30%-60% in places while headcount is not rising proportionally, and the expense ratio is still improving. AI is not a slide in the deck; it is a margin lever.
What the stock is doing with that is less generous. price to net income is 13.8x, down 22% year over year, and interest coverage has fallen to 7.2x from a 31.6x peak after the Apollo acquisition. Book value per share rose more than 15% from December to $28.74. Capital priorities are straightforward: fund the business first, then debt repayment or buybacks. Haushill:
“We would deploy capital first to the business... right now, I would lean more in toward repaying debt.” — Mark Haushill, Chief Financial Officer · 2026-11-06
The real risk is not the quarter. It is whether dairy is a timing issue, whether competition caps 2027 growth, and whether the fee engine can compound fast enough to offset a slower underwriting cycle. At 13.8x earnings for a business still generating over 20% operating ROE, the market is pricing in a lot of caution. If the dairy true-up proves transient and partner syndicates keep scaling, that caution may look like an opportunity. If it doesn't, the de-rating has further to go.