From the archive · Thursday, August 13, 2026
A ceasefire at ninety-nine and a reopening at four, on deadlines two days apart — and a market that sold volatility to a six-week low into them
The market priced how bad, not how long
Priced level, unpriced duration — the closure moves from supply shock to demand shock
The most useful thing that happened on Wednesday was a disagreement, and it was conducted in public between two sources that could both be checked.
In the morning, two wire reports described diplomats as signalling progress toward reopening the Strait of Hormuz. Later the same day a senior Iranian source told reporters there had been no progress at all, that Tehran and Washington remained at loggerheads over reviving the interim deal agreed in June, and that the talks were at a fresh impasse. The equity market took most of the session to decide between them and settled it in the final hour, closing lower as peace-deal hopes diminished.
Our own layer had already decided, and it is worth being precise about what it had decided and when. A US-Iran agreement on the strait by 15 August is marked at 0.0495. A US-announced end of the blockade by the same date is marked at 0.045, down from 0.675 on 5 August. The Iran-Oman track — the mediating channel, and the last one still live — is at 0.150, having been 0.815 on 6 August. That last one did not fall in a day. It printed seven consecutive marks, each lower than the one before it, through 0.730, 0.720, 0.450, 0.295 and 0.225. A collapse can be one bad headline; seven one-way sessions is a market that revised in the same direction every day it looked.
Set against that, the other side of the same table went to certainty. An effective US-Iran ceasefire by 14 August is marked 0.9865, by month-end 0.990, with the Israel-Iran band through month-end at 0.905. All three are series highs. This is the split this brief first published on 9 August, when the same two questions read 0.900 and 0.085 on a 31 August deadline. It was right, and what has changed is not the shape but the clock: both legs now sit on deadlines forty-eight hours away. On Saturday it stops being something we argue and becomes something that happened.
What the market did with all of this was nothing, and the nothing is the story. July inflation landed exactly on consensus — 0.1% on the month, 3.4% on the year, core 0.2% and 2.5%, each annual figure a tenth below June. The VIX closed 14.55, its lowest in 40 sessions. Crude finished +0.08%. It is worth being exact about the pattern, because it is the whole argument: every bit of repricing this closure has produced in the last week happened in ONE session, Monday's +5.05%, and the two sessions since have delivered nothing at all. A market that reprices once and then stops has formed a view about how bad something is. It has not formed a view about how long it lasts — and duration is the only variable Saturday touches.
The obvious objection is that the metals are pricing it, and that gold and silver rising on a risk day is the Gulf premium showing up where it always does. The evidence runs against that. Silver rose +1.18% on the session against gold's +0.59%, and +3.47% against +1.54% over the three sessions from Friday. Copper fell -0.28%. Silver is a poor haven and a good monetary beta — it carries an industrial leg that a genuine war-and-recession bid punishes — so silver leading gold is close to disqualifying for a war-premium reading on its own, and copper falling removes the broad supply-shock reading as well. The positioning data points the same way with its own limitations attached: silver managed-money net stood at 11,974 in the 2026-08-04 table against 13,782 on 30 June, a smaller position while the price rose 9.6% over the same stretch. That is one reporting category rather than a census of buyers, and it stops on 4 August. A survey reported this week reaches the same conclusion from the opposite direction, finding Fed policy under the new chair outweighing both geopolitical tension and Chinese central-bank buying in gold forecasts. The metals are trading the discount rate, not the strait.
We should mark the book honestly, including where it went the wrong way. On 10 August this brief argued the market had stopped pricing Hormuz as an event and started pricing it as a condition; on Wednesday the International Energy Agency did precisely that in a published forecast, cutting world oil demand for this year by a further 510,000 barrels a day, to a contraction of 1.6 million, and naming the closure. An agency revising demand rather than supply is that call arriving at institutional scale. Against it, the toll leg: on Tuesday we quoted the year-end band on an Iranian transit charge at 0.560 and called it rising, its highest since 4 August. It ticked down to 0.520 on Wednesday. Small, and it went against us, and it is the second time in a week this brief has been right about a direction and wrong about the mechanism that delivers it.
The part that deserves more attention than the deadline is what the IEA revision implies about category. A supply shock and a demand shock look identical while the price is rising and have opposite second halves. The first ends with barrels returning to an economy that still wants them. The second ends with a price that does not recover, because the consumption has been re-based — routes retimed, refineries re-tooled for different grades, industrial users switched or shut. Six months of closure is roughly where the second process starts to dominate, and that is where the Gulf now is. It also dissolves the apparent complacency in Wednesday's tape: a barrel sitting still through a diplomatic collapse is not a market ignoring the news, it is a market where the risk premium and the demand write-down have grown to similar size and are cancelling each other out.
Two other things are worth carrying out of the session. The first is that the second chokepoint has stopped being quiet: Houthi forces killed six aboard a cargo ship in Bab el-Mandeb on Tuesday, the first Red Sea shipping fatalities in over a year, and US forces fired on a container ship in the Gulf of Oman within hours. This brief has leaned for four editions on a normal Red Sea as the evidence that the disruption was Hormuz-specific, and that evidence is now weakening. The second is that the long end declined the all-clear the volatility market accepted. The thirty-year held 5.25% and the ten-year closed unchanged at 4.68% through an in-line print, with food inflation described as unpriced in the bond market and the Treasury market's absorption capacity being questioned directly. One of those two markets has the July print wrong, and the long end has now held above five per cent through a hawkish Fed, a credibility fight and a soft inflation reading in succession.