When a line is worth a billion: Wording lessons from the Russian aircraft judgment

The largest aviation insurance dispute since the first Gulf War has been decided, and the reinsurance market should read it closely — not for the headline number, though that is arresting enough, but for what it reveals about the wordings the market trades every day and rarely negotiates.

In AerCap Ireland Ltd v AIG Europe SA [2025] EWHC 1430 (Comm), the Commercial Court found that lessors of aircraft stranded in Russia after the 2022 invasion were entitled to recover under their war risks cover, awarding the lead claimant some $1.035bn, with more than $240m in interest on top.

The losses will run through reinsurance and retrocession in the ordinary way, and the parallel operator-policy trial — listed for October 2026 — proceeds by cut-through directly against reinsurers. This is a reinsurance story as much as a direct one.

The most valuable lesson sits in a single line of wording. The lessors’ policies provided “contingent” cover, which responds where the insured is “not indemnified” under the operator’s own policy.

But what does “not indemnified” require — that the insured has actually not been paid, that it could not recover on the balance of probabilities, or that it has exhausted the primary?

The court held it meant “had not been paid,” so that the lessors’ unresolved operator-policy claims were no bar to recovery. That construction is now before the Court of Appeal.

Whichever way it lands, the point for underwriters and their reinsurers is the same: the trigger relative to the primary is typically drafted in a line, and here it has been worth close to a billion dollars. It is the most consequential and least negotiated term in the wording.

The second lesson is that the boundary between covers is no technicality when the covers carry different limits.

The lead claimant argued that its loss was an all risks loss rather than a war risks one — not out of doctrinal preference, but because its war risks cover was capped in the region of $1.2bn while its all risks cover was materially more generous.

It lost on precisely that point. Where two standard covers carry different limits, the characterisation of the peril does not merely classify the loss; it determines the recovery, and the difference can exceed the entire value of the assets in dispute.

Reinsurers whose exposure follows the characterisation of the peril should know exactly where that boundary falls in their own wordings, and whether they intend it to drive quantum. Frequently it does; frequently no one decided that it should.

The third lesson concerns the architecture of derivative protection. A lessor does not usually control the insurance of the asset it owns.

Its protection runs through a chain — a lease covenant to insure, an operator policy naming it as additional insured through a standard AVN67B or AVN67C endorsement, a cut-through to the reinsurers, and its own contingent cover behind all of it.

Each link was assumed to work until it was tested, and the litigation tested them all at once. For reinsurers, the cut-through is the link that matters: it is the mechanism by which a party you never underwrote comes to you directly, and its limits are discovered, as here, in the most expensive way available.

The operator-policy trial will test those cut-through claims squarely, under Russian law, in an English court — a combination worth watching whatever one’s position in the tower.

There is a broader observation the market should not miss. These were not bespoke disputes. They turned on standard forms — the same contingent-cover language, the same AVN67 endorsements, the same war and all risks architecture — that the market trades in volume.

A wording question of that kind, once litigated in open court, is answered for everyone: within weeks of the judgment, wordings across the aviation market were being revised on the strength of it.

Had the same question gone to confidential arbitration, as a great deal of coverage and reinsurance business does, the market would have got no such answer — and would have carried on trading a form it could not reliably construe, pricing the uncertainty into premium.

The reinsurance market, which arbitrates far more of its wording disputes than it litigates, might reflect on how rarely it gives itself the benefit of a public answer to a question that affects the whole book.

The aviation market has had its answer, subject to appeal. It got it because these claims happened to be litigated rather than arbitrated. The wider market should ask why it so seldom does.

By Russell D. Pulver

The author is an independent arbitrator and mediator and expresses no view on the merits of any coverage or indemnity question in the proceedings discussed.