Brent has stopped spiking on Hormuz and started climbing

Twenty-two percent in seven sessions with no down day among them — the only week in the crisis's 133 trading days that nobody sold into.

Published 2026-09-11 · analysis

Between the close on 28 August and the close on 9 September, Brent went from $89.75 to $109.51, up 22.0%, and it rose on every one of the seven sessions in between. The crisis has not done that before. Across its 133 trading days since 28 February there is no other seven-session stretch with no down day in it — not one. Size is not what is unusual here; the seven days to 20 March were bigger, $90.98 to $118.42 for 30.2%. What is unusual is that nobody took the other side.

This crisis has always paid for its gains with violence, and this rally has not been paid for. Thirteen seven-day windows since February have added 20% or more. The other twelve ran a mean daily move between 3.1% and 6.3%, with March's worst weeks thrashing at over 6% a day in both directions. The window just closed ran 2.9%: the 92nd percentile of crisis weeks for gain, the 35th for daily movement. Strip out the opening 7.0% jump on 1 September and the six sessions after it averaged 2.2% and never exceeded 3.2%. Nor is that first day the president's doing — the post announcing strikes near the strait is timestamped 17:25Z, after the 16:30 London assessment, so it cannot have priced into that close. The market got there on its own.

Widen the frame past this war and the shape gets rarer still. In 9,973 Brent closes going back to May 1987, a run of seven consecutive up sessions has added 20% or more exactly three times. Two of them came a fortnight apart in May and June 2020, starting from $24.20 and $32.73 a barrel as the price climbed out of the collapse that bottomed at $9.12. Both were rebounds from a crash. This is the first one going the other way — into a crisis rather than out of one.

The rhyme it breaks is the fade. Four times before this, Brent has run up four sessions or more since February, and all four were lower within five sessions of the peak. After 23 July it gave back 18.8% in three sessions, $105.32 to $85.51. After 30 April it lost 18.0% in five, $124.24 to $101.82. Selling the Hormuz spike has been the most reliable trade of the war. The official series ends at the 9 September close and has not printed 10 or 11 September, so the seventh session is the last thing here that can be checked — and through it, the give-back has not started.

The wire is not what is doing this. Tehran's channels carried 61 items across those seven sessions, 38 of them about the strait, including the Supreme National Security Council secretary declaring it completely closed on 6 September and a count putting transits at their lowest since May. By this site's own words-and-deeds test, statement days move Brent 0.90 times as much as quiet days at p = 0.80 — a null, not a signal. Meanwhile the account that spent August calling the strait open and asking whether to rename it has not mentioned Hormuz since 3 September. What changed is not the volume of claims but the shape of the move. For six months the correct read on a Hormuz rally was that it would be sold inside a week; that read has now failed seven sessions running, and it failed quietly, which is the part worth noticing.

Open in the interactive site → · the feed · email me when something breaks