Our model’s estimate for Sunday 1 November 2026 (UTC) — extrapolated past our tested range
6.3
million barrels per day
Working range 4.9 to 15.4. This range is wider than usual. Our back-test only validates this model up to 92 days past the anchor quarter (April–June 2026); we are now 124 days past it. We widen the range further for every extra day in this zone, because the one signal we have about longer-range error — measured with hindsight data our live model does not have in real time — shows the miss growing faster than the one-quarter error we actually validated. Treat this as a wider-than-tested best guess, not a backtested figure. It will tighten back to our normal range as soon as EIA publishes July–September 2026.
This estimate is extrapolated past our tested horizon — see how we get to a number.
No change since our 2026-10-31 estimate: still 6.3 million barrels per day.
See every past estimate → — table, chart, and a JSON/CSV/Atom feed of our own daily figures.
New to the strait itself? Geography, what normally flows through it, and why →
Nobody publishes a daily figure for the Strait of Hormuz that is free to republish. The public record arrives a quarter at a time, months after the fact. The trackers that watch the strait day by day are locked behind commercial licences.
So we build our own. Our model takes the public evidence we can legally stand behind, calibrates against the quarterly record, moves with monthly Gulf oil production data, and carries the result forward to the most recent completed day (UTC) — with a range that reflects how much we genuinely know rather than how confident we would like to sound. That range widens while the strait is unsettled and narrows toward ±3% in calm periods — flows fell from 21.6 to 4.9 million barrels per day in three quarters during 2026, and our model has been estimating through that disruption. When our back-tested horizon runs out before the next published quarter arrives, we stop publishing a number rather than guess past what we have tested; see above if that is what you are seeing right now.
This is our estimate, not an official measurement — and that is the product. The method is ours, it is published in full, and it is scored against every new figure that lands. Widening what feeds it — more public data, official reports, credible reporting — is continuous work, and the sources and method page always says exactly what is in the model today.