The war that began on February 28, 2026, when American and Israeli strikes killed Iran's Supreme Leader Ali Khamenei, has already run through a ceasefire that collapsed, a naval blockade that was lifted and then reimposed in practice, and a June memorandum that both sides had signed before either side stopped honoring it. Six and a half months in, and by CENTCOM's own estimate to Congress, forty-three point six billion dollars in direct military costs later, with a one-and-a-half-trillion-dollar defense budget request already sitting in front of Congress for the following fiscal year, the war is once again being talked toward an ending. What it has not produced, at any point in those six and a half months, is a single number that every party to it accepts as true.
On September 3, 2026, the president of the United States posted on Truth Social that Hormuz volumes were back, comparing twenty million barrels a day before the war to eighteen million now. The figure was not invented. An escorted-day total near that size had already circulated in reporting two days earlier, when American forces guided a large convoy through the strait. What Trump's post did was convert that one peak day into a claim about the strait's ordinary condition. His own energy secretary, Chris Wright, had told CNBC the same week that Monday's combined flow was just over seventeen million barrels, close enough to read as confirmation, while adding, without naming a figure, that the multi-day rolling average was running well below it. Two days after that, an Energy Department spokesperson gave reporters a third number for a single day: nearly twelve million barrels. Four figures, one administration, one week, and no two of them describing the same strait.
Six months into this war, nobody at the Pentagon, the Energy Department, or the shipping-intelligence firms that track the Strait of Hormuz by satellite and transponder has published a figure the others would sign, because nobody has agreed on what a barrel that has "come through" the strait actually means, over what interval it should be counted, or which of the half-visible tanker fleets now moving through the Gulf should be counted at all.
This is the condition under which the war is now being negotiated to a close. On September 19, Mohsen Rezaei, secretary of Iran's Supreme National Security Council, told Al Jazeera that Tehran had sent three conditions to Washington through Qatar: an end to the war on every front, the release of Iran's frozen funds, and an end to the naval blockade. The third condition assumes a fact. It assumes the blockade is still intact enough to end. Three days earlier, a United States Navy captain had told the same network that the strait had never really stopped functioning at all.
September the Third, Three Numbers
Eighteen million barrels a day was not a modest claim. Before the war, the Strait of Hormuz carried roughly twenty million barrels daily, more than a fifth of the world's oil supply. If the president's number held as a description of the strait's ordinary condition, rather than of one escorted day, the strait that two air forces and a navy had spent seven months fighting over was back within ten percent of its pre-war ceiling. That is the kind of number a White House can use to describe a war it is trying to close out on favorable terms.
It did not survive the week it was made in. Wright had told CNBC that Monday's combined crude-and-products flow was seventeen million barrels, close enough to the president's figure to read as confirmation. On the same appearance he added that the rolling average across several days was running well below either number, without naming what it actually was. Two days later, an Energy Department spokesperson gave reporters a third figure for a single day: nearly twelve million barrels. Inside the same week, the administration cited eighteen million, then seventeen million, then an unnamed lower average, then twelve million, and none of the four was ever reconciled with the other three in public.
No number in this account survived long enough for the government itself to defend it past a single news cycle.
Wright is the cabinet secretary who controls the Strategic Petroleum Reserve, the official the market is built to trust on exactly this kind of figure, not a spokesman reading a line someone else wrote for him.
It is possible that Monday's seventeen million and Wednesday's twelve million were both honestly measured, describing different days in a strait where tanker traffic moves in convoys and pulses rather than a steady stream, so that the appearance of contradiction is partly an artifact of how oil actually travels rather than evidence that someone was shading a number. That softer reading cannot be ruled out from the public record, and it deserves to be stated plainly rather than assumed away.
What the softer reading cannot resolve is the use to which the number was put. The dispute is not primarily over whether eighteen million barrels could cross the strait in a single escorted day. It is over whether one day can honestly describe the condition of the strait across the weeks a peace deal has to cover, and on that question the president's own energy secretary, in the same week, gave a different answer.
What Vortexa and Tanker Trackers Actually Saw
The independent side of the ledger tells a tighter story, and a lower one. Vortexa, a commodity-tracking firm, put the twenty-eight day moving average at roughly eight million barrels a day as of August 31, according to senior analyst Rohit Rathod. Tanker Trackers, working from a different data pipeline, reported crude transit averaging just under six million barrels a day over the seven days ending that Wednesday.
Different windows and different definitions, a twenty-eight day average against a seven-day one, crude and products against crude alone, still placed both independent trackers in the same broad single-digit-million range, far below the eighteen-million headline.
Bob McNally, who runs the Rapidan Energy Group and has spent two decades pricing supply disruption for clients who cannot afford to guess wrong, put the honest version of the problem to Axios on September 4. "We will know the truth, or closer to it, in a few weeks," he said. "If all that oil came out, it has to go somewhere." Gregory Brew of the Eurasia Group offered a narrower reading the same week: more oil was probably moving than before, he said, just not as much as the administration claimed, though likely enough to represent a real American and Saudi advantage over Iran's position. Neither analyst called the numbers fabricated. They called them unverified, which was enough on its own to make the numbers unusable as a settled fact for anyone pricing risk in real time.
The structural reason sits underneath both estimates. Commercial tankers running Gulf routes now routinely disable their AIS transponders, the satellite signal that lets Vortexa, Tanker Trackers, and every other commercial monitor place a vessel on a map in real time. A dark tanker does not vanish from every method a tracker has; satellite imagery, cargo-loading data, and eventual port and customs records can still place it, days or weeks later. What a disabled transponder removes is the cheap, automatic, same-day layer that would let a tracker publish a number before the news cycle that number was meant to serve has already closed. Every count made inside that window, government and independent alike, is a count of the vessels that were easiest to place in time, and no agency has published what share of Gulf traffic that represents.
Turning off a transponder does not make a tanker disappear. It makes the tanker harder to turn into an agreed number before the political use of that number has already expired.
The scale of the problem is not marginal. A Ukrainian government intelligence catalog counted 1,337 vessels in the global sanctions-evasion fleet as of February 2026, more than double the roughly six hundred tracked in 2022, and by some sector estimates equivalent to roughly a fifth of the world's tanker numbers. Iran has run ships this way longer than any other sanctioned exporter, moving crude to Chinese refinery buyers under evasion arrangements that predate this war by nearly a decade. This fleet stays hard to place using several techniques at once, flag changes, opaque ownership, ship-to-ship transfers, and AIS disabling, each one degrading a different link in the tracking chain. A fleet operating at that scale is large enough to make real-time reconciliation materially harder, rather than treating the disputed range between the low and high Hormuz estimates now in public circulation as ordinary measurement noise.
Maysan Province, September the Thirteenth
Three weeks after Wright's numbers, the argument over how much oil was moving through Hormuz stopped being academic. On September 13, drones struck Saudi Arabia's East-West pipeline in the Riyadh and Medina regions, injuring workers and forcing the kingdom's energy ministry to suspend the line as a precaution. Investigators traced the drones' launch point to Maysan province in southeastern Iraq, on the Iranian border. The pipeline did not restart that week. Estimated repair time ranged from three to six weeks depending on the source, with partial resumption possible sooner and full restoration likely to take longer.
The East-West pipeline exists for exactly the situation the war had created. It carries Saudi crude overland from the Gulf coast to the Red Sea, letting the kingdom's oil reach the world without ever touching the Strait of Hormuz. At full capacity it can move seven million barrels a day, a figure large enough to matter to the global market on its own. In August it had been running at barely two million barrels a day, its lowest monthly level since January, as the line absorbed whatever extra load the Hormuz closure had pushed onto it. Then the one route built to make the Hormuz question moot went offline, with full restoration measured in weeks rather than days, at the exact moment three separate agencies could not agree on what was moving through the strait it was supposed to relieve.
The math is unforgiving. Saudi Arabia now has less capacity to route around the strait than it did in July, and a peace deal is being negotiated on the assumption that the strait itself is carrying somewhere close to its historic volume. If the trackers are right and the true figure sits near eight million barrels rather than eighteen, the loss of even part of a seven-million-barrel bypass is not a rounding error against a disputed range that, on some estimates, spans barely that much oil to begin with.
The one route built to make the Hormuz number irrelevant went dark, with full restoration measured in weeks, at the exact moment that number became the only one that mattered.
A Market That Has to Price Somebody's Number
Brent crude crossed one hundred dollars a barrel on September 9, the first time it had held that level since the war's early weeks, and touched one hundred eight dollars the following day as the East-West pipeline strike compounded the uncertainty. By the third week of September it was trading in the low one hundreds, elevated but not at crisis levels, a range that itself tells a story: traders were pricing meaningful risk to supply, not the near-total closure the war's worst months had produced, and not the fully restored flow the administration's higher figures implied either.
A war-risk premium is itself a number nobody outside London's insurance syndicates ever sees broken down by name. The Manifest has traced this mechanism before: an earlier account of the June closure found that no warship ever shut the strait so much as an insurance committee's listed-area designation did, by making passage commercially irrational for any vessel whose financing required continuous coverage. The same private, opaque pricing layer is still doing quiet work underneath the September numbers, translating a war nobody can measure into a premium nobody can see, which the market then treats as though it were information.
Put the gap in dollars rather than barrels and it stops looking like a rounding argument between analysts. The distance between the independent trackers' roughly eight million barrels a day and the administration's eighteen million is ten million barrels. At a Brent price sitting near one hundred dollars, that gap alone represents close to a billion dollars of daily throughput that either exists or does not, depending entirely on which government official's number a reader chooses to believe on a given morning. No war bond, no appropriations hearing, no shareholder call treats a billion-dollar daily discrepancy as background noise in any other context. This one has been sitting in public view since the third of September.
A market does not wait for a reconciled number before it prices one. Every trading desk, every refiner hedging next month's crude purchase, every insurer setting a war-risk premium on a Gulf transit has to choose, every morning, which of the competing figures to weight, and none of them can wait for CENTCOM, Vortexa, and the Energy Department to agree, because the agreement may never come. The price sitting in the low one hundreds represents the market's own private average of the competing claims, arrived at and traded on within a day, the way markets always resolve a question that official verification is not going to answer on a trading day's timescale. That average has never been audited by anyone. It is simply the number capital could not afford to wait for.
Nine Hundred Million Barrels Since May
Three days after the pipeline strike, CENTCOM offered its own number, and it was bigger than any of the others. Captain Tim Hawkins, the command's spokesman, told Al Jazeera on September 16 that US forces had assisted commercial vessels moving more than nine hundred million barrels of crude through the Strait of Hormuz since early May, that the Navy had redirected more than one hundred ships attempting to violate the blockade, and that mine-clearing operations had finished sweeping the primary shipping lanes several weeks earlier. Traffic through the strait, he said, continues to flow, and Iran does not control the waterway.
Run the arithmetic on Hawkins' own figure and something unexpected happens. Nine hundred million barrels across roughly four and a half months, from early May to mid-September, works out to just over six and a half million barrels a day on average. That number sits inside the same range Vortexa and Tanker Trackers had been reporting for weeks. Stated as a long-run average rather than a single-day headline, the cumulative CENTCOM claim is broadly consistent with independent tracking.
Hawkins built his figure on a scale nobody could spin for a single news cycle: a four-and-a-half-month average instead of a daily headline.
The contradiction sat in the daily numbers chosen for public consumption, not in the four-and-a-half-month totals: the eighteen million posted to a social platform, the seventeen million cited on a Monday, the twelve million attributed to a spokesperson on a Wednesday, each closer to a political ceiling than to any rolling average its author was prepared to stand behind for more than a day. Hawkins, briefing on a scale nobody could cherry-pick around, ended up closer to the independent trackers than his own government's own talking points had managed in weeks.
What "Redirected" Means
The barrel count is not the only figure in Hawkins' briefing that rests entirely on the word of the institution reporting it. "More than one hundred ships" redirected for attempting to violate the blockade is an operational claim, not a customs record. No manifest, no ship name, no port of origin accompanies it. Mine-clearing operations "finished sweeping the primary shipping lanes several weeks" before September 16 is a similarly unverifiable claim: no independent maritime authority, no insurer, and no allied navy has published a parallel assessment confirming the lanes are clear, and marine underwriters, whose entire business model depends on pricing exactly this kind of risk correctly, have not moved war-risk premiums on Hormuz transits down to a level that would suggest they believe it.
None of this requires believing Hawkins invented the number. Military public affairs officers are trained to state operational facts with confidence, because ambiguity in a war zone briefing reads as weakness to adversaries and allies alike, and a spokesman who hedges over mine-clearing status is a spokesman who has just told Iran's navy exactly where the uncertainty sits. The claim also serves a deterrent function, stated for an adversary audience rather than measured for an auditor, and nothing in six months of CENTCOM briefings suggests anyone expected it to be treated as a sourced figure. What changes in Doha is the use to which that confident, unaudited claim gets put: a peace process now treating operational confidence as though it were operational proof.
Doha, September the Nineteenth
Rezaei's three conditions, delivered through Qatar on September 19, were specific enough to be testable: an end to fighting on every front, the release of Iran's frozen funds, and an end to the naval blockade. The third condition is the one that exposes the whole arrangement. Iran is asking Washington to end something Washington insists, on the record, is not actually closing anything. The next day, September 20, Tehran was warning of a "painful" response if Washington launched further attacks, in the same breath as calling for the war's end and the blockade's lifting, a combination that reads less like a ceasefire in progress than like two governments each negotiating from a version of the map the other does not recognize.
Even the deals meant to resolve the dispute have needed their own clarifying statements. Iran and Oman announced a temporary transit arrangement for the strait on August 26. On September 13, Iran's foreign minister, Abbas Araghchi, felt the need to say publicly that the arrangement "in no way means that the Strait of Hormuz will be opened." A government spent a public statement correcting the inference that its own agreement had settled the question the agreement was reported to settle. Two days after that clarification, a UN fact-finding mission published findings that a February 28 American strike on a primary school in Minab, killing more than one hundred fifty people including roughly one hundred twenty children, failed to meet the obligation to verify a civilian target before firing. Trump had said in June that "nobody" targeted the school on purpose. The Pentagon has never published a competing account. On the ledger of contested facts this war has generated, oil is simply the one with a price attached.
There is a second missing figure in this war, and its absence tracks the same pattern from a different angle. Mojtaba Khamenei, who became Iran's Supreme Leader on March 9, nine days after the strike that killed his father, has not appeared in public since. His late-August statement on the economy, calling on officials to address "the chain of economic and livelihood challenges" and warning against language that would "harm social cohesion," was issued in writing and read by state media. No footage of him delivering it has surfaced. Iran's own leadership is, in this specific and literal sense, asking the world to take its highest office on faith in the same month Washington is asking the world to take its most basic maritime claim on faith.
Both sides have learned to keep functioning without the fact that would normally anchor them.
Iran's domestic position gives the negotiation its real urgency, and that urgency has little to do with the barrel count. Food inflation above one hundred twenty-seven percent, a rial trading near two point two million to the dollar, roughly two months of gasoline supply remaining, and a Basij mobilization of up to three hundred thirty thousand people ordered on September 18 describe a state under internal pressure that would exist whether or not a single tanker ever crossed the strait again. That pressure is the deeper mechanism, and it is not this piece's subject. What belongs here is the layer sitting on top of it: the specific external fact, an open or closed Strait of Hormuz, that the peace process treats as settled and that none of the institutions responsible for settling it can actually confirm.
The Architecture Nobody Is Required to Reconcile
None of the actors producing these numbers are behaving irrationally. A president closing out an unpopular war is rewarded, politically, for a figure that reads as victory, and Truth Social carries no editor positioned to check that figure against a shipping manifest before it posts. A cabinet secretary briefing a market operates inside an administration's broader messaging, where an optimistic range threads more easily than a precise one that might contradict the president who appointed him. A military spokesman briefing on operational success is institutionally rewarded for confidence, not for footnotes. A commodity-tracking firm selling subscriptions to hedge funds and refiners is rewarded for being right more often than the free numbers, which means its entire business model depends on official claims staying unreliable enough to need a paid alternative. A tanker operator moving crude out of a sanctioned system is rewarded, quite literally, for not being counted at all. Every incentive in this account points away from convergence and none of them points toward it, which is a stronger explanation than deception for why six months of war have produced six months of numbers that never once lined up. The same administration asking Congress for one and a half trillion dollars in defense spending for the following fiscal year has a straightforward interest in a strait that reads as secured rather than contested, whatever an auditor might eventually conclude, and no appropriations committee is currently positioned to check the barrel count against the budget line.
What actual reconciliation would require is not exotic. A single published methodology, applied consistently, covering both flagged and dark tonnage with a stated margin of error, cross-checked against port-of-origin and port-of-destination customs data on both ends of a voyage, updated on a fixed schedule rather than released selectively when a number happens to be favorable. Every element of that description already exists somewhere in the current apparatus, split across CENTCOM, the Energy Department, the IMF, and two private trackers who compete rather than cooperate. What does not exist is an institution required to assemble those pieces into one number and publish it before the political value of ambiguity has been fully spent.
Officials measure for the audience in front of them that morning. Trackers measure for the clients paying a subscription for a better number. Tankers measure for the sanctions regime each owner is individually trying to avoid. No treaty requires any of these figures to converge. No court adjudicates between a Navy captain's four-and-a-half-month average and a data firm's rolling one. No regulator fines an energy secretary for citing four different numbers in five days. The war will end on a signed date, specific and dated, the way the Islamabad memorandum was specific and dated in June. That is not a data gap. That is a machine for letting a war end without anyone ever having to prove what it cost to keep the strait open.
None of this is unique to Hormuz, or to this war. Any future conflict fought over a chokepoint, the Taiwan Strait, the Suez Canal, the Bab-el-Mandeb, will produce the same structure the moment commercial operators realize they can disable their own tracking and officials realize the daily headline matters more than the audited average. The portable version of the finding travels past oil entirely: a war measured by self-interested parties, none of whom answer to a shared referee, tends to produce several right-sounding numbers that were never designed to be compared against each other, rather than one clearly wrong one, and a peace process that has to proceed anyway.
The strongest version of the counterargument does not dispute any fact in this account. It accepts that Wright cited inconsistent numbers, that Hawkins' cumulative figure lines up with independent tracking, and that AIS spoofing is real and well documented across the shipping industry. It simply says this is what fog of war has always looked like in a contested chokepoint conflict, and that treating six weeks of noisy wartime data as evidence of institutional failure mistakes the ordinary friction of measurement under fire for something more structural than it really is. That objection is serious. Wartime logistics data has never been clean, and analysts have built careers assembling rough consensus out of exactly this kind of scattered, self-interested reporting.
The reading offered here does not claim that any official lied, or that the war is larger or smaller than it appears. It claims something narrower: that the specific fact a ceasefire is now being built on, whether the Strait of Hormuz is open, has no institution anywhere tasked with certifying it, and that this is not a temporary condition of the fighting but the standing condition of a chokepoint war fought after commercial shippers learned to disable their own tracking. The fog is real. What is new is that nobody on any side, including the side with the Navy, is even trying to burn through it in public.
Two Facts That Were Never Supposed to Sit This Close Together
The strait everyone is fighting to declare open carries, by the most careful independent count available, somewhere between six and nine million barrels a day, a figure that has barely moved even as the diplomacy around it has moved constantly. The number everyone keeps citing to prove the war is ending, eighteen million, came from a social media post that outlasted barely a single news cycle before the president's own department stopped repeating it.
The architecture accounts for every actor in this dispute. The president has a platform and an incentive to declare victory early. The energy secretary has a department and an obligation to a market that treats every word he says as tradeable information. CENTCOM has a mission and a spokesman trained to describe it in the most favorable available terms. Vortexa and Tanker Trackers have subscribers who pay specifically because official numbers cannot be trusted at face value. Rezaei has a negotiating position that requires the blockade to still exist on paper. The only actor in this account without a designed role is the reader who took the headline at face value, with no instrument of their own to check it against.
Wars have generated contested casualty counts and contested victory claims for as long as wars have been fought, and by that standard a disputed shipping figure barely registers as new. What is different this time is the specific promise that broke. The instruments that were supposed to end this kind of dispute, satellite tracking, AIS transponders, commercial analytics sold as the corrective to exactly the sort of government spin that clouded body counts in earlier wars, exist and are running in the Strait of Hormuz right now. They have not produced the convergence they were built to produce. The technology that was supposed to make a chokepoint's throughput as countable as a bank balance instead produced three institutions with three separate numbers, none of them obviously wrong, none of them reconcilable either. The International Monetary Fund's own PortWatch program tracks exactly this kind of chokepoint traffic and has not, as of this writing, published a figure that settles the dispute among the others. A tool built specifically to end this ambiguity has so far declined to.
The strait was measured to fit the peace. The peace was never measured to fit the strait.
Trump said the barrels are back. His own energy secretary, on the record, in the same week, said the honest number was lower. Both statements can be true. Neither tells us whether the strait is open.
Evidence Map
Facts, interpretations, forecasts, and disconfirming signals.
Core claim. The Trump administration and independent shipping trackers have not converged, and cannot currently converge, on a verified figure for oil flow through the Strait of Hormuz, because the underlying tracking data (AIS transponder signals) is voluntarily disabled by a significant and unmeasured share of the vessels involved, and no institution on any side is tasked with reconciling the resulting range. The peace negotiation and the global oil market are both proceeding on unverified numbers.
Evidence level. Facts (high, on the record): Trump's September 3 Truth Social claim of eighteen million barrels a day; Wright's three inconsistent CNBC-era figures across the same week, including his own acknowledgment that the rolling average was lower; Vortexa's and Tanker Trackers' independent estimates of six to eight million barrels a day; Captain Hawkins' September 16 cumulative claim of nine hundred million barrels since May; the September 13 drone strike on the Saudi East-West pipeline and its three-to-six-week estimated suspension; Rezaei's September 19 three conditions sent via Qatar; CENTCOM's estimate of forty-three point six billion dollars in war costs through September 3, reported to Congress on September 18. Interpretation (medium, marked): that some daily figures cited by officials were politically selected from measured observations rather than presented as representative averages, while the cumulative CENTCOM average is broadly consistent with independent tracking. Forecast (speculative): that no reconciled figure will be published before the war formally ends.
What would confirm this. A future independent audit (IEA, IMF PortWatch, or a comparable body) publishing a reconciled Hormuz throughput figure that the administration's daily claims fail to match; continued divergence between official daily statements and tracker averages through the remainder of the negotiation.
What would disprove this. The Energy Department publishing a single, sourced, methodology-disclosed daily figure that independent trackers subsequently corroborate within a normal margin of error; a verified sharp narrowing of the gap between official and independent estimates.
Watchlist. Whether the East-West pipeline resumes on schedule in late October; whether the Doha talks produce a memorandum that specifies a verification mechanism for the blockade's status, something the June memorandum did not include; IEA and IMF PortWatch reporting in the weeks following publication.
Jerry van der Laan writes The Manifest Archive, forensic journalism on the systems beneath power, money, and history. He traces the structures beneath them.
