AXA XL's Galileo Re Cat Bond Prices the First US Terrorism Risk

For 24 years, US terrorism risk has sat almost entirely on a federal backstop rather than in private capital markets. That changed this month: AXA XL sponsored the Galileo Re Ltd. (Series 2026-1) catastrophe bond, securing $67.5 million of retrocessional protection that shares a single limit across US named-storm risk and, for the first time in cat bond history, US terrorism risk (Artemis.bm, July 2026).

$67.5M

Size of the Galileo Re Series 2026-1 notes.

2.16%

Initial combined modeled expected loss across both perils.

Spread paid to investors on the notes.

A Peril Cat Bonds Have Always Avoided

Catastrophe bonds have securitized hurricane, earthquake, wildfire, and even pandemic risk for two decades, but terrorism has been the conspicuous absence. The reason is structural, not incidental. Terrorism has no stable base rate the way wind or seismic activity does; a single attack can shift the entire loss distribution overnight in a way no amount of historical data can anticipate, and the peril's severity tail is driven by human intent rather than a physical process a catastrophe model can simulate from first principles. Add to that the Terrorism Risk Insurance Act (TRIA), the federal backstop Congress created in 2002 after the September 11 attacks specifically because the private market withdrew from terrorism coverage almost entirely, and there has been little commercial pressure on insurers to look past TRIA for capacity. Galileo Re is the first test of whether capital markets will price the risk directly instead.

How the Bond Actually Models Two Unrelated Perils

The mechanics matter here as much as the headline. Galileo Re's $67.5 million of notes provide AXA XL a single shared limit of retrocessional protection across both US named-storm risk and US terrorism risk, meaning a loss from either peril draws down the same pool of collateral rather than each peril carrying its own dedicated tranche. Verisk modeled the hurricane component using its AIR risk model, the same modeling framework underpinning most property cat bonds; Moody's modeled the terrorism component using its RMS risk model, a probabilistic terrorism model that simulates attack scenarios (target selection, weapon type, casualty and damage distributions) rather than observed loss history, since there is no meaningful actuarial loss history for the peril to fit a frequency-severity curve to in the first place. Combining a physically modeled peril with a scenario-modeled peril under one limit is itself a modeling choice: it assumes the correlation between a major hurricane and a terrorism event affecting the same exposed portfolio is low enough that pooling their capital requirement is more efficient than holding separate collateral for each, a bet that has never been tested in a live cat bond before this one. The notes carry an almost two-year term, maturing in early June 2028 (Artemis.bm, July 2026).

An Industry-Loss Trigger Is the Real Innovation

The more consequential design choice is the trigger, not the peril bundling. Galileo Re pays out on an industry-loss basis, meaning the notes trigger once the wider insurance industry's aggregate loss from a qualifying event crosses a set threshold, as measured by a third-party index, rather than on AXA XL's own indemnified loss. Every terrorism cat bond written before Galileo Re used an indemnity trigger instead, tying payout directly to the sponsor's actual incurred claims. An industry-loss trigger is faster to settle (the index reports a market-wide estimate rather than waiting for AXA XL's own claims department to adjust and finalize losses) but introduces basis risk: AXA XL's actual terrorism losses in a real event could run higher or lower than its pro-rata share of the industry index implies, leaving a gap the sponsor bears itself. Choosing an index trigger for the first-ever terrorism cat bond suggests the market prioritized settlement speed and modeling simplicity over precision, a reasonable tradeoff for a proof-of-concept transaction this size.

The index itself is Property Claim Services' Global Terror Index, and a loss only qualifies once the event has been certified as an act of terrorism under the Terrorism Risk Insurance Act, the same federal statute the bond is implicitly testing an alternative to. That certification dependency creates a structural wrinkle worth flagging directly: TRIA is due to expire in 2027, roughly a year before Galileo Re's June 2028 maturity (Artemis.bm, July 2026). If Congress does not reauthorize TRIA on schedule, or reauthorizes it with a materially different certification process, the legal mechanism the bond's own trigger depends on could change mid-term, an unusual form of legislative basis risk layered on top of the already-novel modeling basis risk. It is a reminder that even a capital-markets alternative to a federal backstop can end up contractually tethered to that same backstop's statutory machinery.

Pricing a Risk With No Loss Distribution to Anchor It

The initial combined modeled expected loss across both perils was 2.16%, and investors are being paid a 6% spread, roughly two and a half times the modeled loss rate. That multiple is the market's price for model uncertainty, not just for the modeled loss itself. Wind risk is modeled probabilistically from decades of storm track and structural vulnerability data, so a buyer has real confidence in the shape of the loss distribution even if the exact figure is uncertain. Terrorism modeled through RMS scenario analysis carries a different kind of uncertainty: the model is only as good as its assumptions about who might attack, what target, and with what weapon, assumptions that cannot be back-tested against a real historical record the way a hurricane model's frequency assumptions can. A wide spread-to-expected-loss multiple is how ILS investors price that gap between "the model produced 2.16%" and "we actually believe the tail risk is bounded at 2.16%."

A Record Market Looking for New Perils to Price

Galileo Re landed inside a cat bond market that just posted its strongest half-year on record. First-half 2026 issuance reached $17.98 billion, topping the prior H1 2025 record of $17.56 billion, across 83 new transactions, itself a record versus 72 deals a year earlier, and the outstanding market hit $65.6 billion as of June 30, 2026, the largest quarter-end total ever (Artemis.bm, 2026). Twelve first-time sponsors entered the market in the same period. Against that backdrop, an industry-loss trigger stands out: 81% of H1 2026 risk capital used indemnity triggers, up from 79% a year earlier, meaning Galileo Re's index-based structure sits in a shrinking minority of the market at the exact moment it is being applied to a brand-new peril category for the first time. A market this flush with capital and this dominated by first-time sponsors and record issuance is precisely the environment where investors have both the appetite and the pricing power to demand a premium for underwriting something genuinely untested.

Why Now, and Why AXA XL

Timing here is not incidental. Property catastrophe reinsurance pricing has been falling sharply through 2026, with Guy Carpenter's Global Property Catastrophe Rate-On-Line Index down roughly 16% for the year, so ILS sponsors and investors alike have an incentive to look for uncorrelated risk that does not trade at softening-market spreads. Terrorism risk, priced almost nowhere in capital markets until now, is about as uncorrelated with property catastrophe losses as a peril can get, which is exactly the diversification argument ILS funds have been making for cat bonds generally since the asset class began attracting pension and endowment capital. AXA XL, meanwhile, already carries a large global property and specialty book with meaningful terrorism accumulation, making it a natural first mover for a structure that tests whether capital markets, not just TRIA and the traditional excess-of-loss terrorism treaty market, will take the risk.

What This Tests for TRIA

TRIA is up for reauthorization again, as it has been roughly every five to seven years since 2002, and the program's design has always assumed that private capital cannot or will not absorb catastrophic terrorism losses without a federal backstop sitting behind it. A single $67.5 million bond does not settle that question, the notional is a rounding error against TRIA's federal loss-sharing structure, which triggers only after an individual insurer's losses cross a deductible tied to prior-year premium and after the industry-aggregate trigger of roughly $200 million in insured terrorism losses is met. But Galileo Re is a real data point that ILS capital, hunting for yield uncorrelated with softening property cat pricing, is willing to underwrite terrorism risk directly when a name-brand sponsor and two established modeling agencies package it credibly. If more sponsors follow, the reauthorization debate over how much of the terrorism tail the federal government should keep absorbing gets a genuine market alternative to point to, rather than a hypothetical one.

A Precedent From a Different Uninsurable Peril

Terrorism is not the first peril capital markets have had to price with essentially no usable loss history. Pandemic risk faced the identical problem before the World Bank's pandemic bonds, issued in 2017 with no frequency-severity data to anchor a model, ultimately triggered and paid out during COVID-19 in 2020. Those bonds were criticized afterward for trigger design flaws that delayed payouts precisely when funds were needed fastest, a lesson that appears to have shaped Galileo Re's choice of a faster-settling industry-loss index over an indemnity trigger. The broader pattern across both pandemic and terrorism securitization is the same: the first transaction in a new peril category tends to prioritize getting a workable structure to market at all over getting the trigger mechanics precisely right, and the refinements come in the second and third deals, if there are any.

Why This Matters for Actuaries

For pricing actuaries on standalone terrorism or package policies with a terrorism peril, Galileo Re's spread-to-expected-loss ratio is a rare, real market-clearing data point for how capital providers price terrorism tail uncertainty when they are not relying on TRIA's backstop, useful context even for lines where TRIA coverage still applies, since it shows what the risk costs when the government is not standing behind it. For reserving and capital actuaries at reinsurers or ILS funds evaluating whether to follow AXA XL into this structure, two open questions matter more than the headline pricing: whether the wind-terrorism correlation assumption embedded in a shared limit holds up under a scenario that touches both perils at once, and whether the TRIA-certification dependency baked into the PCS Global Terror Index trigger creates a coverage gap if Congress lets the 2027 reauthorization slip past its deadline before the bond's 2028 maturity. Watch whether other sponsors bring a second terrorism-inclusive cat bond to market in the next 12 to 18 months, and whether that second deal keeps the industry-loss trigger or reverts to indemnity; either signal would say more about the structure's staying power than Galileo Re's pricing alone.