Copper's Record Lasted Five Days. The Shortage Didn't.

On September 10, three-month copper on the London Metal Exchange traded at $14,875 per metric ton — the highest price in the metal's history. Five days later, it touched $13,958, a 6% slide that erased three weeks of gains. Shanghai's most-active contract followed it down.

Here is what did not happen in those five days: no mine reopened. No new supply reached the market. The structural deficit forecasts that justified the record did not move by a single ton.

What happened instead was a headline. On September 10 — the day of the record — Reuters reported that the White House was hesitating on tariffs for refined copper, weighing the benefit of encouraging domestic mining against the cost of making an already-expensive industrial input more expensive for American manufacturers. The report removed, in one stroke, the urgency that had driven buyers to front-load purchases ahead of a tariff that might never arrive. The selloff followed within hours.

That sequence tells you something uncomfortable about this year's copper rally: a meaningful part of the record price was built in Washington, not in the mines. The question that matters for anyone holding copper exposure — metal, miners, or the industrial companies paying these prices — is how much.

A Record Built in Washington

The policy mechanics matter here, because they created the buying behavior that just unwound.

In July 2025, the White House added copper to the Section 232 national-security tariff program — the same authority used for steel and aluminum — imposing a 50% tariff on semi-finished copper products. But refined copper, the cathode that feeds the entire downstream industry, was deferred: the Commerce Department was ordered to report back on domestic refining capacity by June 30, 2026, with an escalating tariff schedule — 15%, then 30% — under consideration.

That deadline passed without a decision. And in the vacuum, the market did what markets do: it front-ran the possibility. Buyers accelerated imports into the United States to get ahead of a tariff that might land at any moment, pulling deliverable metal across the Atlantic and leaving inventories outside the U.S. tighter than global headline numbers suggested. American COMEX prices traded at a persistent premium to London — a de facto import duty imposed by uncertainty itself, collected by no one.

The September 10 Reuters report punctured that logic. If the tariff is stalled indefinitely, there is no reason to hoard metal in New Orleans warehouses. The unwind was mechanical.

Briefings like this land in members' inboxes before the market prices them in. Join free →

The Shortage Is Real. The Question Is Where.

None of this means the copper bull case was fiction. The supply side of this market has genuinely broken.

Start with Grasberg. In September 2025, roughly 800,000 metric tons of mud flooded the block-cave section of the world's second-largest copper mine, killing workers and forcing Freeport-McMoRan to declare force majeure. Freeport expects the incident to erase roughly 270,000 tons of copper from 2026 output — about 2% of global mine supply from a single event — with a phased restart targeting only 85% of capacity by the second half of this year.

Grasberg was not alone. The Kamoa-Kakula complex in the Democratic Republic of Congo cut output by an estimated 300,000 tons after flooding. Codelco's El Teniente lost 33,000 tons to an accident. Teck trimmed guidance by 60,000 tons across its operations. Cobre Panama — over 300,000 tons of annual capacity — remains shut. Unplanned outages historically run about 5% of global supply; this cluster blew through that, in a market with far less inventory flexibility than it had a decade ago.

The demand side is just as structural: AI data centers, grid modernization, and rearmament are all copper-intensive, and none of them are cyclical whims. Treatment charges — the fees smelters earn for processing concentrate — collapsed to unprecedented lows this cycle, the clearest possible signal that there is not enough mined concentrate to feed the world's smelting capacity.

And yet the world's major forecasters cannot agree on what this adds up to. J.P. Morgan projects a 330,000-ton refined copper deficit for 2026. The International Copper Study Group sees a 150,000-ton deficit. Goldman Sachs forecasts a 300,000-ton surplus. That is a nearly 650,000-ton spread between the two ends of the street — on the same market, with the same public data.

Meanwhile, the physical signals just flipped. LME warehouse stocks jumped 3.6% in a single session to 242,900 tons. The futures curve — the most reliable real-time gauge of physical tightness in any commodity — moved from backwardation, where buyers pay a premium for metal now, into contango, with spot trading at a $67.50 discount to three-month delivery. Scarcity pricing, by that measure, is gone.

So which is it? Either the shortage was never what it appeared, or the tape is temporarily lying about a deficit that is still coming. The difference is worth roughly $2,000 a ton — and it is resolvable, because the two sides of that forecaster split are not actually talking about the same market.

Get the next briefing free in your inbox: Join AlphaBriefing

The rest of this briefing is for paid members: why Goldman's surplus and J.P. Morgan's deficit are both defensible — and which one the curve is voting for, the three-scenario framework for the White House tariff decision with price zones for each, the positioning logic on miners versus metal after a 74% year for the big producers, and the five-date catalyst calendar through year-end.

AlphaBriefing Paid gets you every investment thesis, scenario framework, and catalyst brief we publish — the analysis private intel clients pay four figures for, at a fraction of that.

Unlock the full briefing →