Hormuz: the volume shock and the price response are two different events

What is observable

The Strait of Hormuz carried around 20 million barrels a day of crude and products in 2025 — roughly a quarter of seaborne oil trade — together with just over 112 billion cubic metres of LNG, close to a fifth of the global trade, on the IEA's figures. Traffic has been disrupted since late February 2026. Reported daily transits ran at eight to fifteen vessels in the first week of August, against roughly 130 before the disruption began.

The cost of moving a ship through has moved by a different order. Additional war-risk premia for Gulf transits were quoted at 7.5 to 10 per cent of hull value in late July, against 1 to 3 per cent shortly beforehand. Underwriters have reportedly become reluctant to write spot cover at any price, which is a distinct problem from cover being expensive: a premium can be budgeted, an absence of cover cannot.

Brent, meanwhile, was around $84 a barrel on 10 August, some sixteen per cent above where it sat when the disruption began. That is a real move. It is not the move you would infer from a ninety per cent fall in transits, and the difference is the analytically interesting part of this episode.

Why the price has moved so much less than the volume

Three mechanisms account for most of the gap, and each one has a different shelf life.

Physical bypass absorbs part of it. Saudi Arabia's Petroline and the UAE's ADCOP together offer something in the region of 3.5 to 5.5 million barrels a day of available capacity that avoids the strait entirely. Against a normal flow near 20 million, that is a meaningful cushion and a hard ceiling at the same time — it is the part of the problem that engineering can solve, and it is under a third of the flow.

Inventory and floating storage absorb another part, but they are a stock, not a flow: they buy time and then stop. And the third mechanism is expectation. A market that assigns real probability to a negotiated reopening prices the expected value of the disruption, not its current severity. That is rational, and it means the benchmark is carrying information about the odds of resolution rather than about physical availability today.

The practical consequence is that the crude price is the wrong single indicator for a company trying to understand its own exposure. It is a probability-weighted number. A shipper without cover, or a manufacturer whose input arrives on a vessel that did not sail, is not experiencing a probability-weighted outcome.

Where this actually lands on a business

The exposures that matter here are rarely the ones in an energy hedge. In our framing they sit in four places, in roughly this order of neglect.

• Cover availability, not cover cost. Charterparties and sale contracts written on the assumption that war-risk cover is obtainable behave differently when it is not. This is a contractual and a legal exposure before it is a financial one.
• Delivery reliability rather than unit price. A schedule that slips by weeks reprices working capital, safety stock and, in some sectors, contractual service levels — none of which move with Brent.
• Pass-through policy agreed after the fact. Whether a freight and energy surcharge is absorbed or repriced is a decision most groups take market by market under pressure. Agreeing it in advance is cheap; agreeing it during is not.
• Second-order inflation in exposed importing markets, which reaches consumer demand and pricing power on a lag of months, well after the shipping story has left the front page.

What we do not claim

We do not have a view on how or when this resolves, and anyone offering one with confidence is selling something. The relevant analytical distinction is between the branches, not the point: a sustained-restriction branch and a negotiated-reopening branch have very different cost profiles, both are live, and a company can prepare for both without knowing which arrives.

This note is dated. It describes a situation that has moved materially several times since February and can be expected to move again. Treat the figures as of the date at the top of this page, and check them against the sources below before acting on them.