Copper's Tariff-Scare Plunge Was a Policy Bet Unwinding, Not a Demand Break

A report that Washington's plan to tax copper imports had stalled sent the metal's premium sharply lower on Thursday — and copper producers fell with it, Freeport-McMoRan down roughly 6.6% and Southern Copper more than 7%. For anyone holding or watching these names, the first instinct is to read this as a copper reality check. It is worth pausing before that. The thing that actually collapsed was a policy bet, not the metal's supply-and-demand math — and separating the two is where the investment read lives.

The premium that just went negative was a tariff gauge

Copper trades in two places that usually stay close: COMEX in the U.S. and the LME, whose prices anchor the global market. Shipping is cheap and arbitrage keeps the two roughly aligned. What broke that alignment — and has for fifteen months — is the expectation of a tariff.

Since February of last year, when President Trump ordered an investigation into copper imports under Section 232, traders have been pulling metal into the U.S. ahead of a possible duty. The economic logic runs through the price gap itself: the more the market expected a tariff, the wider the premium U.S. buyers were willing to pay to secure metal before the duty landed. Societe Generale has read that gap like a probability table — in mid-August, the premium implied roughly a 15% chance of a 15% tariff in January 2027 and a 37% chance of a 30% duty in January 2028.

That is why the front-month COMEX-LME spread for copper, which had already been whipsawing, traded around $400 a ton in early June — and had spiked to roughly $2,900 a ton in late July 2025, when the market feared a sweeping 50% levy on all copper.

On Thursday the spread fell about $200 a ton intraday and went negative, closing near minus $20 a ton. A negative premium is a blunt statement: at that level, shipping metal to the U.S. is uneconomic, because the bet that Washington would tax imports has faded.

The inventory "overhang" was a hoard, not oversupply

The tariff scare did not just move prices — it moved physical metal. COMEX-registered copper inventories climbed for 57 straight days to a record of roughly 767,000 short tons, or about 696,000 metric tons. Macroeconomic bears point to that pile as evidence of glut, but that reading gets the direction wrong. That stockpile was built deliberately: traders hauled mostly LME-warehoused metal into the U.S. to get it in before a tariff could land. In other words, the "record" inventory in America was drawn down from warehouses everywhere else.

Strip that distortion out and the underlying picture is one of genuine tightness, not surplus. The International Copper Study Group projected a refined-copper deficit of about 150,000 tons for 2026; J.P. Morgan's estimate was more than twice that. Demand from AI data centers, grid modernization, and electrification is the driver, and mine supply has not kept pace. Copper was in price discovery at a record high near $14,875 a ton on Thursday — the record it had just set before the tariff news knocked 3% off the LME benchmark and 4.7% off COMEX.

The distinction matters because the whole reason the premium existed was policy speculation rather than consumption. The market is not pricing that copper demand broke. It is pricing that a government decision the market had banked on now looks less likely.

What the tariff stall actually costs — and what it does not

Be clear-eyed about what is genuinely lost. A phased tariff on refined copper — Commerce had recommended 15% from January 2027 rising to 30% a year later — would have been a tailwind for domestic producers and for U.S.-market prices. The White House's reported hesitation, driven by fears that a duty would raise manufacturing costs and offset the benefits of encouraging more domestic mining, removes that support. The metal that traders had locked into U.S. warehouses also becomes an overhang: if the tariff never arrives, consuming that much copper domestically would take years, as Macquarie has noted, and some of it could flow back to global markets. All of that is a real, if modest, subtraction from the U.S.-only story.

But here is the part that should govern a decision about copper miners. The largest U.S.-listed producers do not sell their entire output at a U.S. premium into a tariff-protected market; Freeport, Southern Copper, and Teck sell a global book at world prices. Their cash flows track the global copper price and the deficit that supports it, not the COMEX premium or the fate of a Washington policy vote. Those cash flows did not change on Thursday. What changed was that the market subtracted one speculative, U.S.-specific tailwind from stocks that had already run hard — Freeport is up about 40% year to date and more than a third over the last four months.

That is the honest tension, and it cuts against an automatic "buy the dip." Value only exists below intrinsic value, and nothing about this dip makes these names cheap. Freeport trades near 12 times EV/EBITDA with a trailing P/E above 30, a full price that already embeds the optimistic version of high copper holding; Southern Copper is richer still. The tariff stall did not rescue that valuation. It removed one supporting pillar and left the expensive part intact.

So the tariff scare is best treated as what it is: the unwinding of a policy bet layered on top of a structural copper story. The structural story — a supply deficit against AI and grid demand — is intact and independent of Washington. That keeps the bull case for copper alive. But it is a deficit thesis priced at premium multiples, and the margin-of-safety question — whether these cash-flow streams are worth what you would pay today — was unresolved before the premium turned negative, and remains unresolved now.