The Reserves Nobody Counts
The claim that Iran can no longer rebuild what it has lost is the most consequential judgement anyone is making about this war, and it rests on a quantity nobody measures.
Iranian oil loadings ran at roughly two million barrels a day in March. Through August they ran between 220,000 and 260,000. Four different baselines are in circulation and the answer barely moves: the fall lands somewhere between 85 and 89 percent. That is the strongest number in the entire file, and it is the one figure about this war that behaves the way a measurement should.
It measures a flow. At prices under 100 dollars a barrel, something like 145 million dollars a day has stopped arriving in Tehran. Call it 4.4 billion a month. None of it tells you what the regime already holds.
The judgement everyone is actually making turns on that second thing. Iran cannot reconstitute its coastal radars and its mining capability, the argument runs, because it is short of foreign currency and cut off from dollars. Washington asserts the shortage. Iran's central bank governor calls the claim baseless, says the reserves were never frozen and says more than 18 billion dollars went to essential imports since March. He offers no detail for the figure. Two interested parties, opposite answers and no instrument standing between them.
Meanwhile the number doing most of the rhetorical work is being printed backwards. One widely carried account says the rial has lost 63 percent of its value since March. Sixty-three percent is precisely how much the dollar rose against the rial over that move, from a base near 1.35 million. The rial's own loss across the same distance is closer to 39 percent. Both figures are correct and they describe one event from opposite ends. The larger one is being published as the smaller one.
The starting points conflict too. One account puts the pre-war dollar at 1.75 million rials in late February. Another puts it at 1.35 million in late March. Both hold only if the rial gained nearly 30 percent during the opening three weeks of a war, which nobody claims happened.
There is a structural reason this stays messy. Iran runs two exchange rates, and the gap between the official rate and the street rate has widened past a factor of two since the fighting began. No single rate exists to be right about.
The deeper problem sits underneath all of it. Not one of these figures rests on a published series. Five currency readings, four loading baselines, four separate estimates of how much traffic moves through the strait. Every one is a point quoted at a moment, with nothing beneath it showing direction or speed.
Two readings remain open and this piece takes neither. Tehran's reserves may genuinely be close to gone, which would make the prevailing assessment right for reasons it cannot demonstrate. Or capability reconstitution may not run through hard currency much at all. Radars and mines built domestically, or supplied by Moscow and Beijing against oil rather than against dollars, would not bend to an export collapse the way the published chain assumes.
What would settle it is unglamorous and mostly already exists somewhere. Reserve data, a daily rate series, tracker figures stated separately instead of merged, Chinese customs imports by origin. Until somebody pulls them, the loudest claim about this war is a guess with good manners.