Strait of Hormuz Monitor — 13 Sep 2026

READ — Openness * phase * velocity

• THE CORRECTION — the US is running CONVOY ESCORT, and we have been saying the opposite — We withdraw a standing claim. This file has repeatedly stated that the operative US activity is a blockade and interdiction posture and NOT convoy escort, on the basis that Operation Project Freedom launched on 4 May and paused on 5 May and that the Strait was judged too tight for alongside escort. The evidence is now unambiguous the other way. On 1 September CENTCOM escorted up to forty commercial vessels carrying roughly eighteen million barrels of oil through the Strait — about 90% of a normal pre-war day's flow and the highest single-day wartime crude passage of the war. The operation was conducted under fire, with simultaneous strikes on Iranian air-defence sites, radar, maritime assets, mine-laying capability and communications, and several anti-ship cruise missiles destroyed during the escort. Commercial operators remain cautious even under protection, and war-risk premia and charter rates have risen regardless. BUT THE POSTURE CLAIM WAS WRONG, and a monitor whose whole value is the gap between what is announced and what physically happens does not get to be sloppy about which physical operations are running.
• WHAT IT EXPLAINS — the VLCC-exit stall is CONVOY BATCHING, the third and best explanation — On 10 September we armed a Phase-1 trigger on the observation that no VLCC had cleared the Strait outbound since 2 September. On 11 September we re-specified it, having established that crude moves by shuttle plus ship-to-ship transfer. Today supplies the cleaner mechanism: flow is convoy-batched. Under an escorted regime the outbound record is necessarily lumpy — long gaps punctuated by large, escorted batches — so a week of zero laden VLCC exits followed by a day of forty hulls is the expected shape rather than a signal. That retrospectively validates the re-specification: a trigger keyed to DIRECT VLCC exits was measuring the gaps between convoys. The re-specified form — total crude CLEARANCE, direct plus STS, requiring a flow print rather than a transit count — survives this evidence intact, and is now the only version of the test that could be meaningful. Note the arithmetic too: an 18 mb/d escorted day against our ~8.6 mb/d running estimate is entirely consistent, because 8.6 is an average across convoy and non-convoy days.
• WHAT IT MEANS FOR THE BADGE — escorted third-party transit is evidence AGAINST closure — This is the consequential point. The phase taxonomy in this file turns on one question: physical control of THIRD-PARTY transit through the waterway. A navy escorting ninety per cent of a normal day's oil flow through the Strait, against active opposition, is the strongest piece of against-closure evidence we have recorded. It does not mean the corridor is healthy — flow still runs at roughly 43% of baseline on our estimate, war-risk is ~40x the peace level at about $10m per VLCC passage, and 459 vessels are holding position. It means the corridor is CONTESTED AND TRAFFICKED rather than closed, which is precisely what Phase 2 describes. The badge holds more comfortably today than it did on Thursday, and we say so rather than quietly enjoying having been right for a different reason than we gave.
• AGAINST IT — the screened dark share SPIKED to 34.3%, and the queue lengthened to 459 — We do not net the good news against the bad. Two readings deteriorated hard. The live screened-tanker AIS-gap jumped to 34.3% — 86 of 251 screened tankers dark in the 24 hours to 13 September 0525Z — from 10.3% (17 of 165) on 11 September. That is a reversal inside forty-eight hours on a metric that had been arguing FOR normalisation, and it sits above the provider's own recent seven-day run-rate. Stranded vessels holding position away from berth rose to 459 from 381, near the top of a 219-512 seven-day range. Crisis pressure is unchanged at 95 and the escalation forecast eased a point to 57. HOW TO READ THE DARK SPIKE ALONGSIDE THE CONVOYS: vessels running dark in batches around escorted transits is a different behaviour from steady-state evasion, and is plausibly defensive rather than evasive. We flag that as the likely explanation without asserting it — we have just been wrong once this week by inferring a mechanism we had not established.
• DARK 62.7% — RE-SEARCHED, all-traffic scalar HELD, with the screened reversal recorded beside it — Re-searched fresh this morning per the standing daily rule. The all-traffic scalar holds at 62.7% (Kpler, 148 of 236 transits, 1-19 August) because there is still no published September all-traffic dark share anywhere in the open record — the same flag we have carried for a week. OPACITY (>60) therefore continues to fire on an August vintage, and we continue to say so rather than treat the firing as current evidence. What is new is the direction of the LIVE screened metric, which has now moved sharply in BOTH directions inside a week: 23.4% on 8 September, 10.3% on 11 September, 34.3% today. A metric that swings by a factor of three in five days on a screened universe that itself changes size (205, then 165, now 251 vessels) should be read as a noisy behavioural indicator, not a trend. We record all three readings so the volatility is visible rather than smoothed.
• THE PRICE — Brent CLOSED $104.61, up 3.5% but well off its ~$108 intraday high; the reversal rallied equities — Brent closed Friday at $104.61 — UP 3.5% on Thursday's $101.06, but roughly 3% BELOW the ~$108 it printed intraday on Friday morning. Both framings are true and the difference matters. The sequencing matters for anyone reading across from this file to the macro: the US August CPI printed HOT on Friday — core +0.3% m/m against a 0.2% consensus, headline +0.4%, taking implied September hike odds to roughly 88% — and equities rallied anyway, snapping a four-session losing streak, with the S&P up 0.86% to 7,656.98. The barrel did that, not the inflation print. A three per cent reversal off the intraday high relieved precisely the cost channel that had driven the previous four sessions. It is also a reminder of how much of the equity tape is currently a derivative of this corridor: the same market that fell on $108 rallied on $104.6 while absorbing a hawkish CPI without complaint.
• VELOCITY — DETERIORATING -> STALLED, and the offset is genuine rather than a fudge — We moved to DETERIORATING on 10 September and held it on 11 September. We move back to STALLED today. FOR: convoy escort at scale is a materially better corridor than we had assessed, the barrel fell 4%, and the escalation forecast eased. AGAINST: the screened dark share tripled, the vessel queue lengthened to 459, and Mokha consolidated with government forces retreating fifty miles. These are real on both sides and they genuinely offset, which is what STALLED means in this file — not 'nothing happened', but 'the vectors cancelled'. We resist the temptation to call it improving on the convoy news alone, because the convoy operation is dated 1 September and the deteriorating readings are dated today.

OFFICIAL RECORD * SoH — Statements * decisions - as regards the Strait

• 1 Sep — CENTCOM escorts ~40 vessels and ~18 MILLION BARRELS through the Strait, under fire — The record we should have carried a fortnight ago. US Central Command escorted up to forty commercial vessels carrying approximately eighteen million barrels of oil through the Strait of Hormuz — roughly 90% of a normal pre-war day's oil flow, and the highest one-day wartime peak of crude passage since the war began more than six months ago. The escort was conducted alongside offensive operations: CENTCOM announced strikes on Iranian air-defence sites, radar systems, maritime assets, mine-laying capabilities and communications installations, fended off waves of drone attacks, and destroyed several anti-ship cruise missiles during the escort itself. Commercial caution persists despite the protection, with war-risk premia and charter rates elevated. Book this as the correction it is: our standing 'interdiction, not escort' line is withdrawn.
• Commercial practice — selective Gulf transits under CONVOY ESCORT, and escorted-only calls — The commercial side corroborates the military one. Major container lines are operating on an escorted basis — one continues selective Gulf transits under convoy escort, another limits Gulf calls to escorted convoys only. That is a materially different operating regime from the one this file has described, and it explains a pattern we had treated as puzzling: traffic that appears to stop for days and then move in clusters. It also has a pricing implication worth stating. If transit is available but only in escorted batches, the binding constraint is SCHEDULING and QUEUE rather than access — which is consistent with 459 vessels holding position while eighteen million barrels move in a single day.
• 10 Sep — Mokha CONSOLIDATES: government forces fall back ~50 miles south to Dhubab — The second chokepoint is hardening rather than reversing. The Houthis took Mokha from Yemeni government-aligned units on 10 September after sustained fighting, and those forces have fallen back roughly fifty miles south to Dhubab. Mokha sits about 80km from Bab el-Mandeb, through which roughly 12% of world goods and a tenth of seaborne oil normally pass. Saudi Arabia answered the wider Houthi offensive with nearly forty strikes, and Pakistan has warned Tehran that continued Houthi action risks triggering a joint defence arrangement with Riyadh and Ankara. WE CONTINUE NOT TO BOOK THIS AS HORMUZ. It belongs in the regional premium. But the consolidation matters for the alternative-routing assumption: a fifty-mile retreat is territory changing hands, not a raid.
• The gas channel — still the genuine near-zero, and still the cleanest de-escalation test — Unchanged and therefore worth restating rather than dropping. No LNG carrier has transited the Strait since 11 July. Qatari force majeure has been running since Ras Laffan was struck and has been extended repeatedly. The asymmetry is physical: crude can be shuttled and transferred ship-to-ship, and — as we now know — escorted in convoy; LNG can do none of those things. That is why oil flows have partly recovered while gas has not moved for two months. IT ALSO MAKES THE GAS CHANNEL THE BETTER TEST. An escorted crude convoy is a military achievement; an LNG carrier transiting would be a statement that the corridor is genuinely safe rather than navigable under protection. Watch for force majeure being LIFTED rather than extended.
• The carried ledger — war-risk ~40x at ~$10m, 459 holding, crisis pressure 95 — War-risk cover remains around 40x the peace baseline at roughly $10.0m per VLCC passage against ~$250k in peace, with six P&I clubs withdrawn, though the Lloyd's Market Association maintains that safety rather than insurance availability is what suppresses traffic. Vessels holding position away from berth: 459, up from 381, seven-day range 219-512. Crisis pressure 95 (unchanged), escalation forecast 57 (down one). Aggregator day-count 196. The Oman safe-passage lane has still NOT registered at the IMO — the 'within days' of 7 September is now six days over, and no map, date or legal definition has been published for the declared exclusion zone.

WHERE WE'VE COME FROM — Trajectory

• 2026-09-05 / 09-08 — TWO VECTORS AT ONCE AND OUR OWN LINE BREAKS. CENTCOM strikes three vessels; Jizan is hit again; Iran declares a Gulf prohibited zone. Brent SETTLES $97.00 through our ~$96-100 trigger. We log it FIRED and RE-RATE throughput 14.0 -> 8.6 mb/d.
• 2026-09-09 / 09-10 — THE ATTRITION BECOMES A CAMPAIGN. CENTCOM destroys FIVE Iranian carriers in one action — eight hulls in four days; Iran answers with twenty missiles at Jordan's al-Azraq. Brent settles $101.21. No VLCC has exited since 2 Sep: we ARM a fifth trigger, cut transits 12 -> 10/day, move velocity to DETERIORATING.
• 2026-09-11 — WE RE-SPECIFY OUR OWN TRIGGER. Ship-to-ship shuttle transfer is established as a mechanism behind the VLCC stall, so the trigger moves to total crude clearance. The genuine near-zero is located in GAS — no LNG transit since 11 July, Qatar -96%. The Houthis seize MOKHA. Brent marks $108.08. Kharg terminal CONFIRMED not hit.
• 2026-09-12 / 09-13 — WE CORRECT THE POSTURE. CENTCOM is running CONVOY ESCORT and moved ~40 vessels and ~18m barrels on 1 Sep — ~90% of a normal pre-war day, under fire. Flow is convoy-BATCHED, which is the third and best explanation of the VLCC stall and validates the re-specification. Escorted third-party transit is evidence AGAINST closure. Against it: screened dark spikes 10.3% -> 34.3%, stranded rises 381 -> 459, Mokha consolidates. Brent closes $104.61, off a ~$108 intraday high. Velocity DETERIORATING -> STALLED. HOLD Phase 2.
• The frame — Four consecutive editions carrying a correction against ourselves: a throughput cut, an armed trigger, a re-specified trigger, and now a withdrawn posture claim. That is not a file losing its grip — it is a file being marked to evidence at the pace the evidence arrives. The badge has not moved through any of it, because the badge tracks physical control of third-party transit and that question has had a stable answer throughout. What has changed is the quality of our explanation for it, and today's is the best we have had.

WHERE IT'S HEADED — Direction * accelerants * reversers

• DIRECTION — a CONVOYED corridor: contested, trafficked, expensive, and scheduled — The right mental model is no longer a throttled pipe. It is a convoy system: access available but rationed by escort scheduling, priced through war-risk rather than denial, and punctuated by large batched movements against a long queue. That reconciles every number in this file — 43% average openness, an 18 mb/d escorted peak, 459 vessels waiting, war-risk at 40x, and a dark share that spikes around convoy windows. It also changes what to watch. Convoy FREQUENCY and SIZE now matter more than any daily transit count, and the constraint to monitor is escort capacity and political willingness to keep running it under fire.
• TIP TO FULL PHASE 1 — unchanged, but the crude-clearance test just got harder to satisfy — A confirmed STRIKE on Kharg or an export terminal; a THIRD-PARTY tanker SUNK in-channel; a US strike on Oman; a secondary-sanctions hit on China answered in the Strait; or total crude CLEARANCE (direct plus STS) confirmed at near-zero. Note what the convoy evidence does to that last one: with escorted batches capable of moving 90% of a normal day's oil, near-zero clearance now requires the ESCORT REGIME ITSELF to stop — which makes the trigger both harder to satisfy and far more meaningful if it ever is. The cleanest single indicator of that failure would be a suspension of escorted convoys, whether from losses, capacity, or a political decision.
• PATH BACK TO PHASE 3 — the gas test, and now an escort-normalisation test — The Oman lane registering at the IMO remains the legal route back and has now missed its own 'within days' by six. The better tests are physical. FIRST and best: an LNG CARRIER TRANSITING, or Qatari force majeure being lifted rather than extended — gas cannot be shuttled, transferred or convoyed around the problem, so it only moves when the corridor is genuinely safe. SECOND, new today: escorted convoys becoming routine and unopposed, with war-risk premia falling from 40x and the 459-vessel queue draining. An escort regime that stops needing to strike air-defence sites while it runs is a corridor on the way back.

SCENARIOS — Base * Bear * Bull

• BASE — a convoyed corridor at ~8-10 mb/d average, gas shut, Brent $98-110; Phase 2 HELD — Escorted convoys continue at intervals, moving large batches against a persistent queue; average flow holds 8-10 mb/d with a lumpy daily distribution; LNG stays halted and Qatari force majeure rolls again; war-risk stays near 40x; Mokha raises the regional premium without closing Bab el-Mandeb; Brent settles $98-110. Phase 2 held, velocity STALLED, OPACITY firing on an August vintage we have flagged. The modal path — and the one in which convoy frequency, not transit counts, is the number to track.
• BEAR — the ESCORT REGIME breaks, or a terminal strike lands, or Bab el-Mandeb closes — The new and most informative bear path is the loss of the escort regime: a successful attack on an escorted convoy or on an escorting warship, a capacity limit, or a political decision to stop. That would convert a scheduled corridor back into a denied one and would satisfy the crude-clearance trigger quickly. Alongside it the standing physical triggers — a Kharg terminal strike, a third-party sinking in-channel, a China hit answered in the Strait — and the regional path in which Mokha is used to close Bab el-Mandeb, removing the alternative route. Brent $115-130, onto a market that has just taken September hike odds to ~88%.
• BULL — an LNG carrier transits, or escorted convoys become routine and unopposed — Either the cleanest signal fires — a laden LNG carrier clears for the first time since 11 July and Qatar lifts rather than extends force majeure — or the escort regime normalises: convoys run without accompanying strike packages, war-risk falls from 40x, and the 459-vessel queue drains. European gas re-rates hard downward, which matters more for euro-area inflation than another ten dollars of Brent. Iran's incentives still argue for an off-ramp — inflation above 60%, its own loadings collapsed, and eight hulls destroyed in four days. The constraint is that Washington is escorting AND striking simultaneously, and that combination has no obvious de-escalatory equilibrium.

Bottom line. HOLD — Phase 2 (Fragile — RE-ESCALATING), velocity DETERIORATING -> STALLED, and the edition is a CORRECTION TO OURSELVES. We have carried for weeks that the operative US activity here is a blockade and interdiction posture and explicitly NOT convoy escort. THAT IS WRONG AND WE WITHDRAW IT. On 1 September CENTCOM escorted up to FORTY commercial vessels carrying roughly EIGHTEEN MILLION BARRELS through the Strait — about 90% of a normal pre-war day's oil flow, the highest single-day wartime crude passage of this war — while striking Iranian air-defence, radar, maritime, mine-laying and communications assets and destroying several anti-ship cruise missiles during the escort itself. TWO CONSEQUENCES. FIRST, it is the third and best explanation of the VLCC-exit stall we armed a trigger on (10 Sep) and re-specified (11 Sep): flow is CONVOY-BATCHED, so a week of no laden exits followed by forty escorted hulls is the expected shape, not an anomaly — which retrospectively validates moving the test onto total crude CLEARANCE. An 18 mb/d escorted day is entirely consistent with an 8.6 mb/d average across convoy and non-convoy days. SECOND, escorted convoys are the United States moving THIRD-PARTY traffic THROUGH the waterway, which is evidence AGAINST closure. The taxonomy turns on exactly that, so THE BADGE HOLDS MORE COMFORTABLY TODAY THAN IT DID ON THURSDAY. WE DO NOT NET THE BAD NEWS AWAY: the live screened dark-share SPIKED to 34.3% (86 of 251) from 10.3% on 11 Sep — a reversal inside 48 hours on a metric that had been arguing for normalisation; stranded vessels rose to 459 from 381; and Mokha CONSOLIDATED, with government forces falling back ~50 miles to Dhubab. The all-traffic dark scalar HOLDS at 62.7% because no September print exists, so OPACITY continues to fire on an August vintage and we keep saying so. THE PRICE: Brent CLOSED $104.61 — up 3.5% on Thursday but ~3% off a ~$108 intraday high — and that reversal — not the CPI, which ran hot at core +0.3% and took hike odds to ~88% — is what let equities rally after four down days. The right model is no longer a throttled pipe but a CONVOY SYSTEM: access rationed by escort scheduling, priced through war-risk, punctuated by batched movement against a long queue. Convoy FREQUENCY and SIZE now matter more than any daily transit count. TIP TO FULL PHASE 1: the standing physical set, plus — newly harder and newly meaningful — total crude clearance at near-zero, which now requires the ESCORT REGIME ITSELF to stop. TIP BACK TO PHASE 3: an LNG carrier transits, or escorted convoys become routine and unopposed with war-risk falling from 40x and the queue draining.