NextFin News - Copper is heading for a record high not because the world is short of metal, but because a looming U.S. tariff is pulling it into American warehouses and making it look scarce everywhere else. ANZ Research says policy uncertainty, rather than a genuine supply deficit, is now the dominant force in the market - a distortion that can push prices to new peaks even while global balances remain roughly balanced.
The mechanics are stark. U.S. refined copper imports exceeded 200,000 metric tons in July 2026 alone, the highest monthly total in at least 12 years, as traders rush to stockpile metal ahead of a possible tariff on refined copper. COMEX inventories have climbed for 46 straight days to a record 675,185 metric tons, while available inventories on the London Metal Exchange have fallen to roughly 90,000 tons, sending the premium on immediately deliverable copper to its highest level since 2021. Copper is being divided into two markets - one inside the United States, one outside - and the price is being set by the thinner, tighter side.
The Two-Market Squeeze
On the COMEX, the most-active September contract touched a record $6.7775 a pound on Wednesday before pulling back to about $6.56 a pound, a level still roughly 19% above where it began the year and about 50% higher than a year ago. In London, the picture is just as tight: the official cash price climbed to $14,525 a tonne at Wednesday's close, up $355 over the week, while the three-month contract settled at $14,336 - within striking distance of January's all-time peak of $14,527.50. The gap between the two benchmarks, the arbitrage that pays traders to ship metal across the Atlantic, has been running at roughly $550 a tonne, close to a 4% premium, and has been far wider at points this summer.
This is not a normal bull market. In a normal bull market, high prices call forth supply: mines ramp up, smelters run harder, scrap flows faster. What is happening now is the opposite. The International Copper Study Group reported that global mine output fell 1.6% in the first five months of 2026, hurt by production declines in Chile, the Democratic Republic of Congo and Indonesia, and it has forecast refined output growth of just 0.4% for the full year. At the same time, concentrate treatment and refining charges have collapsed into negative territory - around -$126.80 per dry metric ton by the end of June, versus a $0 annual benchmark - signaling that smelters are paying for the right to process ore.
Yet the price signal is not clearing the market. It is being trapped. Copper imported into the United States cannot easily flow back out to where it is needed, because once it is inside the U.S. tariff wall it becomes a different product from London-deliverable metal. Robert Edwards, principal copper analyst at CRU, put the arithmetic plainly: "If (U.S.) imports keep coming in as they have been, then it's going to look like a deficit market in reality." CRU's own baseline called for a 639,000-ton global surplus in 2026; the tariff threat has turned that into "at best a balanced market," assuming the metal stockpiled in the United States is no longer available to the rest of the world.
The distortion has a name in market-structure terms: it is a location-driven shortage, not a volume-driven one. The world has the copper. It is simply in the wrong place - and policy is the reason.
Why a Tariff Threat Can Move Prices More Than a Tariff Itself
The policy backdrop is a slow-moving deadline rather than a sudden shock. A 50% tariff already applies to U.S. imports of semi-finished copper products. The Commerce Department, in a review due by June 30, 2026, recommended a universal 15% tariff on refined copper starting January 1, 2027, rising to 30% on January 1, 2028. The White House has not issued a final ruling. That uncertainty is the engine of the distortion.
Uncertainty creates a one-way bet for traders. If the tariff is imposed, metal already inside the United States becomes more valuable than metal outside it, so the rational move is to ship early and hold. If the tariff is delayed or waived - as refined copper ultimately was last year - the importer has lost little beyond financing cost. This asymmetry explains why imports keep rising even as the probability of the tariff being imposed remains modest. Societe Generale modeling put the implementation probability reflected in the COMEX premium at just 14.6%.
"Record imports and a strong arbitrage mean the fragmented market should continue to support prices until the tariff outlook becomes clearer," said Ryan McKay, senior commodity strategist at TD Securities.
The incentive structure is doing something counter-intuitive: it is rewarding stockpiling in the very market that is supposedly most expensive. U.S. imports of refined copper cathodes totaled almost 885,000 tons in the first half of 2026, 3% more than the same period a year earlier and more than double the first half of 2024. The full-year 2025 total was a record 1.64 million tons. Every ton that crosses into the United States is a ton removed from the deliverable pool in London, which is why LME warrant cancellations - metal booked for withdrawal from exchange sheds - have stoked fresh concerns of supply shortages even as global balances remain intact.
ANZ senior commodity strategist Daniel Hynes pointed to the surge in LME warrant cancellations, metal booked for withdrawal from exchange sheds, saying the move "stoked fresh concerns of supply shortages."
The question hanging over the London market is why anyone would deliver metal there at all. "Why would you deliver to the LME," BNP Paribas head of metals strategy David Wilson asked clients this month, "when shipping into the U.S. ahead of the deadline still pays?"
The Second-Order Effect: Prices Signal Scarcity That Does Not Exist
Here is the second-order consequence that most market commentary misses. The first-order effect of the tariff threat is obvious: U.S. prices rise relative to London. The second-order effect is that the price signal itself becomes misleading for the real economy. When copper trades near a record, downstream users - cable makers, grid contractors, appliance manufacturers - read it as a sign of physical scarcity and either hoard or pass costs through. In reality, the scarcity is financial and locational, created by an arbitrage trade rather than by a shortfall in mine output.
This matters because it can trigger the very shortage the market fears. If manufacturers, seeing record prices, begin to over-order "just in case," they add another layer of precautionary demand on top of the tariff-driven stockpiling. The price then rises further, validating the hoarding behavior, and the loop feeds on itself. This is not a demand boom; it is a coordination failure, and it is reversible the moment the policy uncertainty resolves.
The tell is in the structure of the rally. A genuine demand-driven rally would show broad-based strength: rising consumption across regions, rising imports into consuming hubs, tightening inventories everywhere. What we see instead is a split. COMEX stocks are at a record high while LME available inventories are at multi-month lows. U.S. imports are at a 12-year high while Chinese demand shows cracks: industrial profits at Chinese firms rose 11.2% in July from a year earlier, slowing from June's 15.1% pace to the weakest reading of 2026, according to data released by the National Bureau of Statistics. A market short of copper does not look like this. A market being pulled apart by trade policy does.
The Counter-Thesis: What the Bears Are Right About
The strongest argument against the record-high call is not that the tariff threat is imaginary - it is that the price has already overshot the fundamentals that will matter once the uncertainty clears. Goldman Sachs Research estimates copper's fair value at around $11,500 a tonne and forecasts $11,200 for the fourth quarter of 2026, well below current levels. "We feel that the price has overshot its fair fundamental level," said Eoin Dinsmore of Goldman Sachs Research, adding that a clear decision on U.S. refined copper tariffs should serve as a "catalyst for a correction." Goldman has also raised its 2026 global surplus forecast to 300,000 tons from a previous 160,000 tons, and noted that the market recorded a 600,000-ton surplus in 2025, the largest absolute surplus since 2009.
TD Securities made a similar point in mid-August, arguing that tariff distortions are already unwinding and that prices should move lower as pre-emptive buying reverses. Glencore's chief executive, Gary Nagle, has argued that a tariff decision in either direction would cool prices by removing uncertainty. And there is a mechanical reason to be cautious about the COMEX stockpile: a Macquarie strategist noted it would take years to absorb the record inventory sitting in U.S. warehouses once the tariff motive fades.
Not everyone is bearish, however. Citigroup maintained its target of $14,500 a tonne for the zero-to-three-month contract and $15,000 for year-end, citing shrinking visible inventories outside the U.S., constrained mine supply and a soft scrap response to higher prices. "While demand growth remains tepid, supply is under much greater pressure," the bank said.
These competing views point to a timing question rather than a direction question. The bears are right that the current price embeds a policy outcome that has not happened and may never happen at the feared scale. They are wrong to conclude that this makes the near-term record call incorrect. In a distorted market, the price does not reflect what is true; it reflects what traders must do to protect themselves against what might be true. Until the White House rules, the incentive to hold metal inside the United States persists, and the incentive is what sets the marginal price.
What to Watch: The Signals That Would Break the Trade
The distortion is falsifiable. Three signals would tell investors the record-call thesis is breaking down:
- A final White House ruling. An exemption for refined copper - matching last year's decision - would unwind the COMEX-LME arbitrage and allow U.S.-held metal to return to the global deliverable pool. Total LME stocks rising back above the roughly 255,000 metric tons recorded at the end of July 2026 would confirm metal is returning to the deliverable pool.
- A sustained narrowing of the backwardation. The cash-over-three-month spread on the LME has been violently volatile: it widened from about $34 a tonne at the end of July to more than $200 a tonne by mid-August, collapsed as metal rushed into LME sheds, then rebuilt to $189 by late August. A sustained return toward the historical norm would signal that immediate-delivery scarcity is easing.
- COMEX inventory stabilization. The 46-day climbing streak in U.S. exchange stocks is the clearest read on the tariff trade. Consecutive weekly declines would indicate the stockpiling motive is fading.
Outlook: Record First, Reckoning After
The base case is that copper makes a new nominal record before the tariff decision lands, because the asymmetry of the trade favors continued stockpiling and the physical market outside the United States remains tight enough to support it. The upside case is that a confirmed 15% tariff accelerates pre-tariff flows, widens the COMEX premium further, and pushes London prices higher as metal is diverted away from non-U.S. markets - a scenario in which the record is not just touched but held. The downside case is that an exemption or a prolonged delay removes the stockpiling motive, the COMEX inventory overhang hits the market, and prices correct toward the $11,000-$12,000 range that fundamental forecasters see as fair value.
Split by time horizon, the picture is contradictory by design. In the short term - the next one to two quarters - sentiment, positioning, and the tariff deadline dominate, and the path of least resistance is higher. Over the medium term - six to twelve months - fundamentals reassert themselves: mine supply is still constrained, but so is demand growth in China's property sector, and the record COMEX stockpile becomes a source of supply rather than a sink. Over the long term, the structural case for copper remains intact: electrification, grid investment, and AI data-center buildout are real demand drivers that no tariff decision can erase. The question is whether today's prices are paying for the structural story or the policy distortion. Right now, they are paying for the distortion.
The uncomfortable truth for anyone betting on copper at a record is this: the market is not pricing scarcity. It is pricing uncertainty - and uncertainty, unlike a mine closure or a demand boom, can evaporate with a single announcement.
Explore more exclusive insights at nextfin.ai.
