Kingstone Companies placed its 2026/2027 catastrophe program at $500 million of limit, up $60 million, with the core catastrophe excess-of-loss layer priced more than 15% cheaper on a risk-adjusted basis and total program cost falling from about 13% of projected direct premiums earned to roughly 11% (Kingstone Companies, July 2026). It also added wildfire cover, at a $3.5 million first-event retention, to a book that has not yet written a California policy.

$500M
Total 2026/2027 catastrophe reinsurance limit
-15%+
Risk-adjusted price change on core CAT XoL
11% DPE
Program cost as % of projected direct premiums earned
34+
Reinsurers on the tower, 6 of them new to the panel

Key Takeaways

  • $500 million of limit, up $60 million from $440 million, at a core CAT excess-of-loss price more than 15% cheaper on a risk-adjusted basis.
  • 200 basis points of ceded premium released, with program cost falling from about 13% of projected direct premiums earned to roughly 11%.
  • 25% of the tower is a 2025-vintage 144A cat bond at a locked spread, so this cycle's softening flowed only through the un-bonded 75%.
  • A $3.5 million first-event wildfire retention on a book with no California experience behind it. Vendor catastrophe models carry the entire calibration weight.
  • 34-plus reinsurers, six new to the panel, on a placement that required no higher retentions, tighter hours clauses or per-risk sublimits.

The Cost-Per-Limit Math

The release lets an outside reader run the price-per-limit arithmetic without a broker deck. Total limit rose from $440 million to $500 million, a 13.6% increase in coverage, while the core CAT XoL cost fell more than 15% on a risk-adjusted basis, meaning after normalizing for exposure growth in the underlying book. Those two together back out a gross rate-on-line decline larger than 15%, because the exposure base itself grew and the risk-adjusted comparison holds exposure constant.

The program-cost ratio confirms it from the income-statement side. Falling from 13% of projected direct premiums earned to 11% is a 200-basis-point release of ceded premium as a share of the top line, which flows straight to net earned premium and, subject to loss experience, to underwriting margin (Artemis, July 2026). Artemis notes the wildfire cover was priced inside that overall reduction rather than added as separate premium.

Program measure2025/2026 treaty2026/2027 treaty
Total catastrophe limit$440 million$500 million
Multi-year layer via 1886 Re cat bond$125 million (issued 2025)$125 million (rolled)
Program cost as % of projected DPE~13%~11%
Core CAT XoL risk-adjusted priceBaseDown 15%+
First-event wildfire retentionNot applicable$3.5 million
First-event named-storm retention$5 million$5 million
First-event winter-storm retention$6 million$6 million
Reinsurers on towerPrior panel34+, incl. 6 new

For a carrier this size the release is not incremental. On a book with a mid-teens catastrophe contribution to the loss ratio and a combined ratio target in the low 90s, two points of ceded premium is roughly two points of net combined ratio.

Retained volatility moves too. Raising limit by $60 million lifts the ceiling on covered events, but first-event retentions and working-layer size govern how much of a typical accident-year cat load stays with the cedent. Kingstone's disclosed retentions of $3.5 million for wildfire, $5 million for named storm and $6 million for winter storm sit where a smaller primary can absorb a hit without it defining a quarter (Coverager, July 2026).

A Quarter of the Tower Did Not Reprice

The 1886 Re Ltd. cat bond issued in 2025 supplies $125 million of the program, so 25% of Kingstone's catastrophe limit now comes from capital markets rather than traditional reinsurance. That share is high for a primary of this size, against dedicated alternative capital sitting closer to 20% of total reinsurance capital at Gallagher Re's most recent tally.

A 144A bond is a multi-year contract that fixes price and terms on its layer regardless of what the traditional market does. The advantage is stability against sudden hardening after a loss year, or against capacity shrinking for a specific peril. Neither pressure is operative at July 2026, so the bond currently reads as a mild drag on blended program cost rather than a benefit, and the softening reached only the un-bonded 75%.

That changes how the headline number should be read. Kingstone's disclosed figures imply the traditional layers absorbed a reduction larger than 15%, which is what pulled the blended risk-adjusted price down to the reported figure. The un-bonded portion, not the whole tower, is what the market repriced.

Market context puts the outcome at par rather than ahead. Guy Carpenter's July 1 renewals report puts the global property catastrophe rate-on-line index down 16% at mid-year, the largest annual decline since the late 1990s, and Kingstone's "more than 15%" lands within 100 basis points of it. That is unusual precision for a small primary: it indicates market-average terms rather than a discount for scale it does not have, or a premium for a book that only recently returned to profitability.

The supply side agrees. Thirty-four-plus reinsurers on the panel with six new participants, on a straight combination of a cat bond and a traditional excess-of-loss tower with no fronting layer, is structurally simpler than a 2023-era hard-market placement where cedents often accepted MGA-fronted risk transfer to complete a tower.

Meryl Golden put the terms on the record: "While we raised the limit purchased, added wildfire to the mix and improved terms, the cost of our core catastrophe excess of loss coverage decreased more than 15% on a risk-adjusted basis" (Yahoo Finance, July 2026).

The Wildfire Retention Is a Bet on the Model

Adding wildfire cover is unusual for a New York-domiciled specialty personal lines carrier, and it only makes sense against the company's disclosed 2026 California entry (Insurance Business, July 2026). Buying ahead of the exposure is ordinary capital planning: a cedent that waits until it has bound premium either accepts a very high year-one cat load or pays above market for last-minute capacity.

The complication is what the $3.5 million retention is priced against. It is not Kingstone's California wildfire experience, because there is none. It is what the vendor catastrophe models say the loss will be for a portfolio still on the drawing board, and that model output is what the reinsurance market underwrote. The retention sits wherever the model's exceedance-probability curves put it for a book that does not yet exist.

Those models have moved. Wildfire vendor models went through significant calibration cycles after the 2017-2018 California fire seasons, with the 2025 season adding another round. A cedent buying in July 2026 is buying against the post-2025-recalibration view, and the market's willingness to write at this retention says models and market have converged enough to price without a step-change premium. It does not say the converged view is right.

That leaves a specific hole in the capital model. Any adverse deviation from the model's central estimate lives in the retained layer up to $3.5 million per event, plus whatever share Kingstone co-participates in above the retention, and a capital model for the entering book has to load that retained volatility explicitly rather than assume reinsurance absorbs the tail.

Roughly 60 basis points on the mid-teens loss picks a wildfire program has to defend is a thin margin for error on a peril whose model history is this short. The same capital-planning frame from a much larger cedent's side runs through Travelers' July 2026 cat bond and Northeast excess-of-loss retention, where the retained layer is the number that moves.

Further Reading

Sources