When we last wrote on the Strait of Hormuz in April, the United States had just imposed its naval blockade of Iran, Brent was pressing $100, and the question was how high prices would go. Two months on, the question has inverted — and the answer is more alarming for being counterintuitive.
Brent has fallen. North Sea Dated changed hands around $92 this week, down from dated peaks above $140 in March and $111 on 19 May. A casual reader concludes the crisis is easing. Two things are happening, and only one of them is reassuring. The acute war-risk premium has come off as tankers kept moving and a worst-case Hormuz shutdown didn't materialise — that part is real relief. But underneath it, the world is still meeting demand by drawing down inventory, and a softer screen price does nothing to refill the tanks.
The price has two stories in it
The Strait remains commercially constrained. The EIA's June Short-Term Energy Outlook records Middle Eastern producers cutting output by more than 11 million barrels per day, with global inventory draws averaging 6.3 mb/d through Q2 and an expected 7.6 mb/d in Q3; the IEA's reading is 8.5 mb/d in Q2. Those are extraordinary headline draw numbers — but they are total observed draws, and the bulk is being met from the parts of the system that never reach a British forecourt: non-OECD stocks, oil already on the water, and opaque Chinese inventory. The price is not "lying"; it is telling you the acute premium has eased and that the slack is coming out of the global tank, just not evenly.
The number that matters is what's left in the accessible tank — and there the EIA has put a marker down. It projects OECD commercial inventories falling to just under 2.3 billion barrels by December 2026 — the lowest since its dataset began in 2003 — equivalent to roughly 50 days of cover. That is the runway, and it is measured in months.
The runway calculation — and the headline that misleads
The reassuring figure you'll hear quoted is global observed stocks: ~8.2 billion barrels at the January peak. Divide by a multi-mb/d draw and you get years of cover. That figure is close to worthless. Half is OECD holdings; of those, roughly 1.85 billion barrels are strategic or held under government obligation and cannot be released without breaching the 90-day security floor. A quarter is oil on water, in transit, never simultaneously available. Around 15% is Chinese crude stock — opaque and not for sale. Accessible ≪ headline.
But the corollary cuts the other way too: the accessible cushion does not drain at the 8 mb/d headline rate, because most of that draw is coming from the non-accessible buckets. It drains at the OECD-paced rate implied by the EIA's own path — stocks falling from 2,838 mb in January toward 2,300 mb in December, about 1.6 mb/d. On that arithmetic the cushion is neither years (the headline illusion) nor weeks: it is months, and shrinking — comfort exhausting toward a two-decade low by roughly year-end, with genuine operating-stress territory a 2027 question. (Explore it yourself in the interactive Hormuz Inventory Runway; draw rates and inventory levels are EIA/IEA-sourced, the operating-floor levels are OilWatch estimates and flagged as such.)
That is still the story — it is just a slower-burning, more defensible one than a price chart suggests. A market drawing steadily toward its lowest stock cover in two decades is not a market that has resolved anything.
Why Britain is exposed at the wrong end of this
Three structural features put the UK ahead of the queue for stress, regardless of the exact runway length.
Diesel, not crude, is the pinch point. Britain is a structural net importer of middle distillates. Global refinery throughput is forecast by the IEA to fall by 4.5 mb/d in Q2, and the EIA notes the export squeeze has been most pronounced in diesel and jet fuel. A country whose road freight, agriculture and backup generation run on diesel competes for a shrinking pool of refined product — and that tightness can bite long before the crude runway runs out.
Refining thinness. Decades of rationalisation have left the UK with a handful of operating refineries and deep reliance on imported product from Rotterdam, the Gulf and Asia — precisely the lanes the closure has disrupted. There is little domestic slack to absorb a distillate shock.
The reserve obligation cuts both ways. IEA membership requires emergency stocks of 90 days of net imports. That sounds like comfort, but the genuinely drawable cushion is the slice above the obligation — and once a government begins drawing into the 90-day floor itself, it is signalling that the commercial market has already failed. The floor is not a safety net; in cascade terms it is the trigger.
What to watch — the cascade gates
The OilWatch framework treats this as a sequence, not a single event. The gates that would mark escalation from price event to availability event:
- A sustained widening of the diesel crack against crude — the clearest early signal that product, not barrels, is the binding constraint.
- A second IEA coordinated stock release following the 400 million barrels drawn down on 11 March — confirmation that commercial inventories alone can no longer hold the line.
- Forecourt rationing or priority-user allocation anywhere in north-west Europe — the point at which the runway model stops being theoretical.
The diplomatic picture offers thin hope: the UAE and Iran held rare direct talks this week, and Washington has hinted at a negotiating round even as President Trump threatens Iran's Kharg Island export terminal. But as the EIA's own administrator conceded, any return to pre-conflict flows must now reckon with a global market that has already been partly restructured. Drained inventories do not refill on a ceasefire signature; they refill on a schedule measured in quarters — and the rebuild competes with the same demand that emptied them.
Britain's fuel security, in other words, is not a function of today's pump price. It is a function of how much accessible inventory stands between the current draw and the operating floor — and on that measure the cushion is thinning toward a two-decade low, with the clock counted in months.
Methodology. Sourced inputs: OECD industry stocks 2,838 mb (IEA OMR, Jan 2026); EIA end-2026 projection ~2,300 mb / ~50 days, "lowest since 2003" (EIA STEO, Jun 2026); global observed stocks 8,210 mb (IEA OMR, Mar 2026); total global draw 6.3 mb/d Q2 / 7.6 mb/d Q3 (EIA STEO) and 8.5 mb/d Q2 (IEA OMR). The accessible-buffer runway uses the OECD-paced draw (~1.6 mb/d) implied by the EIA Jan→Dec path — not the total observed draw, which is met substantially from non-OECD, on-water and opaque-Chinese stock. Operating-floor levels (~2,050–2,200 mb) are OilWatch estimates; the runway is a range (toward the EIA's two-decade low by ~year-end; operating stress into 2027), not a single breach date. Interactive model: /runway.