On 28 February 2026, the opening day of the war, Iran declared the Strait of Hormuz closed to normal commercial traffic.
The strait did not close. It has not closed since. No party has sealed it, no party has physically blocked it, and through every week of the six months that followed a ship that chose to sail could sail. And yet within a fortnight commercial transits had fallen by more than eighty per cent, roughly twenty thousand seafarers were sitting aboard vessels that were not going anywhere, and production across the region was being shut in at a rate that would peak above eleven million barrels a day.
Something closed. It was not the water.
What “closed” actually means
A strait is not one thing that is either open or shut. It is three, and they come apart.
Legal access is whether passage is permitted, and by whom. Physical access is whether a vessel can get through the water. Commercial access is whether anyone will underwrite, finance, charter and crew the voyage. All three are necessary. Any one of them can fail on its own. And in this war all three moved — at different times, in different directions, and for different reasons.
The reason this matters is not taxonomic. A reader who holds only the binary — open or closed — cannot answer the question that follows from it, which is what would have to change for traffic to resume. If the strait were physically blocked, the answer would involve clearance operations. If it were legally closed, the answer would involve a legal instrument. Neither is the answer here, and neither would help.
Physical passage never fully disappeared
Iran’s declaration was not accompanied by a physical closure and no evidence suggests one was attempted. The strait’s geography makes total closure hard and its maintenance harder: it is wide enough, deep enough and long enough that sealing it is a sustained operation rather than an act, conducted under the guns of everyone who wants it open.
What did happen was degradation. In April the International Maritime Organization stated that there was no safe transit anywhere in the strait — a statement about navigational conditions, issued by the body whose business those conditions are, and not a statement that the strait was closed. Mine clearance later appeared among the commitments reported in the June memorandum, which is evidence that the parties treated a hazard as real. It is not a survey, and this note publishes nothing about where any hazard is or was, because that is operational geography rather than analysis.
Degraded is not sealed. The distinction held for six months and it is the reason the second mechanism mattered so much.
Insurance became the transmission mechanism
In early March, protection-and-indemnity clubs issued seventy-two-hour notices terminating war-risk extensions for the area. Replacement cover was quoted at rates that made most voyages uneconomic.
That is the whole event. There is no strike in it, no mine, no blockade and no warship. A vessel without war-risk cover does not sail — not because anyone prevents it, but because the charterer will not load a hull it cannot insure, the financier will not fund the cargo, and the master will not take the ship. The decision is made in London and Oslo by underwriters reading the same news everyone else is reading, and it propagates through the market in days.
Within weeks commercial transits had fallen by more than eighty per cent. The water was unchanged.
This is the note’s central claim and it is worth stating plainly: the operative closure of the Strait of Hormuz in 2026 was a commercial event. It was produced by the withdrawal of a financial instrument, and it would be reversed by the return of one. That is why this module’s first observable is not whether the strait is open but whether cover is being written at rates charterers accept.
A caution belongs here, and it is a real one. It is tempting to quantify the move — to say premiums rose by some multiple — and this note does not, because the pre-war baseline for Hormuz transit cover could not be obtained from an underwriting or broking source. Wartime quoted rates are documented; the denominator is not, and the figures in circulation disagree with each other. A ratio with an unknown denominator is not a measurement, and the argument does not need one. The mechanism is the finding.
The bypasses are real — and insufficient
Two pipelines reach open water without passing the strait. Saudi Arabia’s East–West system runs to Yanbu on the Red Sea. The UAE’s route runs to Fujairah on the Gulf of Oman. Both are genuine infrastructure, both carried more than usual, and both matter.
They are also not a replacement, for three separate reasons that are often merged into one.
The first is scale. Normal throughput of the strait runs at roughly twenty million barrels a day of crude and products — about a third of global crude oil trade. Assessed bypass capacity is a low single-digit multiple of a million barrels a day. The bypasses are a mitigation of a fraction.
The second is coverage. There are six Gulf Arab states and the bypasses belong to two of them. Kuwait, Bahrain and Qatar have no pipeline alternative at all, and Oman’s exports leave outside the strait for reasons of geography rather than engineering.
The third is the one most easily lost, and this module keeps three quantities apart everywhere for it: capacity is not availability, and availability is not flow. A pipeline’s design figure states what it was built to move. What it can move today depends on pumping, maintenance, storage and the terminal at the far end. What it actually moved last month is a third number. Treating the first as the third is how a real partial workaround becomes an imaginary full one — and it is the single most common error in commentary on this system.
Then Bab el-Mandeb closed to Saudi-linked shipping under a Houthi declaration, which is worth stating precisely: the state that owns the principal Hormuz bypass found the other end of that bypass constrained by a different actor with its own agenda. Redundancy that shares a threat environment is less redundant than it looks on a map.
Liquefied gas has no escape route
For oil the bypasses are insufficient. For gas they do not exist.
Close to a fifth of world LNG exports must transit the strait, and there is no pipeline that substitutes for liquefaction. A molecule of Qatari gas becomes exportable at a liquefaction train and nowhere else; a train that has been damaged is not rerouted, it is rebuilt. When capacity at Ras Laffan was damaged — around seventeen per cent of it, for a period the operator has estimated at up to five years — no maritime solution addressed any part of that.
This is the sharpest illustration of why the module treats the strait as a system rather than a place. Oil’s exposure to Hormuz is severe and partially substitutable. Gas’s exposure is total, and the constraint that bit hardest was not in the water at all.
Legal claims do not create physical control
In May 2026 Iran announced a Persian Gulf Strait Authority, asserting a legal régime over passage.
All six Gulf Arab states rejected it, Oman included — and Oman is the strait’s other riparian, the state through whose territorial water the outbound traffic lane runs, and the one Gulf state that maintains working relations with Tehran. Its rejection is the most informative of the six, because it marks where Omani neutrality stops.
The Authority altered neither physical passage nor commercial access. Nothing moved differently because of it. It is a clean natural experiment in the independence of the three states of access: a legal claim was made, the legal state changed, and the other two did not follow.
The converse held earlier in the war and is just as instructive. The insurance withdrawal changed no legal status whatsoever — no instrument was signed, no authority asserted, no right of passage denied — and it closed the strait in the only sense that moved cargo.
Why the distinction matters beyond this war
The mechanism is not new. The Security Council said so itself: resolution 2817 of 2026, condemning attacks on Gulf states, recalls resolution 552 of 1984 by name. Forty-two years earlier, during the Tanker War, the same repertoire was tried for the first time — attacks on commercial shipping, the reflagging of tankers, and outside navies in the Gulf. What was learned then was that a maritime economy can be attacked through the terms on which it is insured and crewed rather than through its hulls.
What is different now is the concentration. The system that 1984 tested was one where the affected trade had alternatives. The system of 2026 runs about a third of global crude trade and close to a fifth of world LNG through one waterway, with fixed infrastructure at both ends and no substitute for the gas at all.
The general lesson travels. Any chokepoint whose users depend on insurable, financeable, crewable passage can be closed without being blocked, by a party that never touches the water. The instrument is availability of cover, and it responds to perceived risk rather than to demonstrated interdiction — which means it can move faster than any military action, and can fail to reverse long after the military action stops.
That is the finding this module was built to hold, and it is why its first question about the Strait of Hormuz is not whether ships can pass. It is whether anyone will insure them.