Open in interactive viewer → charts, metric popovers & call review

Pelagos Cashes In on Conflict — Then Asks the Market to Rerate the Machine

A 91.5% combined ratio and a fresh $400M buyback show FIHL's capital-allocator model working. The stock still trades below the book value it keeps compounding.
FIHL · Earnings Call · 2026-11-12

The war book, and the bend in the curve

Fidelis — the Bermuda specialty insurer that now goes to market as Pelagos Insurance Capital, though the ticker is still FIHL — turned in an annualized operating ROAE of 15.3% and a combined ratio of 91.5% in Q3 2026, growing gross premiums written 5.3% to $840 million. On the surface it is a quiet quarter. Underneath, the profit engine is loudly geopolitical: marine war and political-risk lines underwritten into live conflict zones. Dan Burrows framed it as “a combined ratio of 91.5% in the third quarter and 89.6% over the last 12 months” — Daniel Burrows, Chief Executive Officer · 2026-11-12. That trailing number is the real tell — this is a book running well inside its own long-term loss-ratio assumptions even with a fat catastrophe line. And the catastrophe line was fat: 14.8 points, or $116 million, all of it inside Insurance. Where it came from matters.

These losses were primarily related to losses from the ongoing conflicts in the Black Sea and Middle East, with the Black Sea losses related to business written before the recent reescalation in the region.

Allan Decleir, Chief Financial Officer · 2026-11-12
That is the interesting bit. The loss was legacy premium — written before the region re-ignited — not the fresh, richly priced stuff. The post conflict business (a phrase genuinely new to this company's vocabulary this quarter) is running at a sub-20% loss ratio per group MD Jonny Strickle; the whole war book, pre- and post-conflict blended, is "kind of mid-40s for the year." So marine war pushed premium to the top end of guidance without being a profit booster. It is a growth story wearing a catastrophe costume.

The machine under the hood

The strategic frame management keeps selling is the capital allocator model: rather than one underwriting team, a curated network of them, screened, sized and compared. The network of underwriting partners now numbers 16, and outside The Fidelis Partnership it has “already bound more than $400 million of premium outside of The Fidelis Partnership in the first 3 quarters of 2026” — Daniel Burrows, Chief Executive Officer · 2026-11-12 — roughly 10% of year-to-date GPW, on a path toward 20% to 30%. This is also where the quarter's quiet complication lives. New partners skew toward lower volatility lines — homeowners, E&S, mortgage reinsurance — which carry light cat exposure but heavier attritional loss and a different expense shape. The tell: policy acquisition expense attributable to TFP fell to 12.4 points from 14.5, exactly as expected as the mix dilutes. CFO Allan Decleir waved off the concern — look at the combined ratio "in its entirety," he said, and the hurdle rates still hold. Fair, but "hurdles hold" is a claim, not a number, and the market will want to watch it as more partner premium earns through. The guidance cadence is the freshest signal. In August, Burrows had warned not to expect Q3 to repeat the prior quarter's growth: “I don't think we expect Q3 to deliver the same sort of growth” — Daniel Burrows, Chief Executive Officer · 2026-08-13. He was right — 5.3% is squarely mid-single-digit land. But now Strickle is calling Q4 “our strongest growth quarter of the year” — Jonathan Strickle, Group Managing Director · 2026-11-12, putting full-year 2026 at the high end of the mid-single-digit range. That is a re-acceleration, not a fade — and it depends on a pipeline management says it filters hard, letting fewer than 5% of candidate partners through.

Capital return, and the stubborn discount

The third pillar is the buyback. In Q3 the company repurchased $83 million of stock at an average $24.55, including $35 million negotiated privately with founding shareholder Pine Brook; year-to-date buybacks reached $363 million and added $0.99 to diluted book value per share. The board then replenished the authorization to $400 million and lifted the quarterly dividend to $0.22, a roughly 3.2% yield. Diluted book value ex-AOCI stands at $27.11, up 12% since year-end. Which sets up the recurring tension of these calls: the stock trades below the book value it keeps growing, and management knows it. “We're very pleased with the momentum in the stock price, but we know there's a long way to go until we unlock the true value of the business... we think we should be trading at a significant multiple to book.” — Daniel Burrows, Chief Executive Officer · 2026-11-12 This is not new — long-time holder Leon Cooperman has pressed the same point for several quarters — and the answer has not changed: compound the results, let the multiple follow.

The friction worth watching

Two things broke the pattern this quarter. First, adverse development of $5 million (0.6 points of the combined ratio) out of the Political Risk, Violence and Terror book — unusual because most prior-year development here emerges in the first half of the year. Strickle was candid: “it tipped into our layer that was unexpected, and we booked it” — Jonathan Strickle, Group Managing Director · 2026-11-12. A 2025 accident-year loss, no change to long-term assumptions, but a reminder that Political risk tails can bite late. Second, the conflict opportunity is maturing. Strickle expects a "sizable market loss" from the current situation to pressure rates in marine and political-violence lines going forward — the very rate adequacy that justified leaning in will erode as the market catches up. The Middle East conflict premium spike that powered 2026 is, by management's own telling, closer to its late innings than its beginning. Alongside a Reinsurance property book that shrank on a single client's outwards-program restructuring (expected to partly return early next year), it is a reminder that this is a momentum model, not a compounding annuity. And one theme has simply vanished. Russia-Ukraine aviation litigation, once the single loudest worry on these calls, is now scarcely mentioned. Back in November 2025, Burrows said of the legacy exposure, “we can draw a line under it” — Daniel Burrows, Chief Executive Officer · 2025-08-14. It has stayed drawn. The absence is itself the story — the overhang that once defined the discount is gone, and the multiple has yet to fully notice.

Bottom line

FIHL is not a complicated quarter to read, but it is a revealing one. premium growth is re-accelerating into Q4, the partner network is scaling faster than the target laid out a year ago, and capital is coming back to shareholders at scale. The risks are equally legible: mix drift toward attritional business, a conflict book whose pricing edge is fading, and a valuation the company insists is wrong. Management's bet is that consistency closes the gap. On this evidence the machine is running — the market just hasn't fully paid for it yet.