The Great Copper Arbitrage - A 2026 Update

In August 2025 I wrote about one of the most extraordinary physical copper trades in history. The threat of US import tariffs had pushed CME copper to a premium of more than $3,000/mt over the LME. Traders bought LME-priced cathode around the world, moved it to the United States and locked in margins that would normally be unthinkable in refined metals.

Then on July 30th 20225, the details were announced. Refined copper cathode was exempt from the new duties. The CME/LME arbitrage collapsed from around $2,500/mt to flat almost instantly. Traders who had already locked in the arb protected their profits. Those still waiting for a 50% tariff saw thousands of dollars per ton of potential margin disappear. At that time it looked like the great copper arbitrage was over, but it turns out it wasn't.

The tariff threat never disappeared.

The July 2025 proclamation excluded refined cathode from the immediate 50% tariff applied to semi0finished copper products, but it also left the cathode question open.

The Commerce Department had recommended a phased universal duty on refined copper of 15% from January 1, 2027, rising to 30% from January 1, 2028. The proclamation required the Commerce Secretary to provide the President with an update on the domestic copper market by June 30th, 2026, so that the President could determine whether those duties were warranted.

That distinction is very important - the June 30th deadline was for Commerce to provide an update to the President. It was not a statutory deadline for the administration to make a public announcement.

That date came and went without a public decision. The market was left with no definitive tariff rate, no implementation schedule, and no guarantee that cathode would be tariffed at all. The 15% number is more than a market rumor - it came from last year's official Commerce recommendation - but it is still not current policy.

That ambiguity has recreated the arbitrage opportunity.

In late March 2026, the CME premium over LME copper began rising again as the June review date approached. In the last few months we have seen the CME trade as much as $800/mt over the LME in the nearby months with months, and by May 2026, March 2027 CME premium was closer to $1,000/mt.

At an LME copper price of $14,000/mt, a 15% duty would equal $2,100/mt. That does not mean the CME/LME spread must mechanically move to exactly $2,100/mt. Freight, physical premiums, available domestic supply, tariff exemptions, spreads, and the date any tariff would apply all influence the relationship. But $2,100/mt is a useful theoretical anchor, and it is till far above the arbitrage traders can lock in today.

This leaves traders with a very familiar decision. Lock in the margin now, or wait for the potentially much larger one.

Assume a trader already owns LME-priced cathode that can ultimately be sold into the US market - either warranted in the CME or delivered to a consumer. If they lock they arb today, they secure the gross margin, but they give up any potential upside from a tariff announcement. However, they remove the risk that the arb collapses before they execute, which would make the trade unprofitable.

But, the trader may already have paid a substantial premium over the LME price to buy the cathode. They have to cover freight, handling, storage, insurance, and financing. The current LME backwardation is also adding another element to the equation.

The cash-3s spread is nearing levels of $200/mt backwardation, the largest since October 2025. Unless they had previously borrowed the spread that is a significant cost. And at $14,000/mt copper, every additional month spent holding and not delivering copper costs an additional ~$80/mt.

At the lower end of the current arbitrage, these costs could essentially eat the entire potential profit if the trader locks in today.

If the trader waits then the upside is obvious. A 15% tariff could send CME premiums over LME close to $2,100/mt. On an inventory of 10,000mt, every additional $100/mt in arb value is worth $1 million. And there are rumors some traders are substantially longer than 10,000mt of physical copper right now.

However, the downside risk is just as real. If the administration confirms that cathode will remain exempt for now, or simply allows the tariff threat to fade (they do have some other pretty important issues to deal with right now), the CME premium could collapse just as it did in 2025. In this instance, waiting is not a free option, it is an option with a very expensive daily theta.

Where should the copper sit?

Whether to execute arb at current levels is only one part of the trade, the other is geography. A trader can import cathode into the US today and clear customs before any tariff is announced. If a duty is later imposed, that metal is already inside the tariff wall. It can be sold to a domestic consumer, or if eligible, warranted in a CME warehouse.

But importing early means committing the metal to a market with record inventory. If no tariff materializes the trader could be left financing copper in the most oversupplied visible market in the world.

Alternatively, traders can hold copper outside the US, preserving the ability to redirect it to Europe, China, or an LME warehouse if the US opportunity disappears.

Market participants appear to be assuming that any cathode tariff would be announced in advance and that there would be a window between announcement and implementation. But this is not guaranteed, particularly for an administration that is known to make impulse decisions.

If a duty became effective immediately - or before a vessel could arrive and clear customs - the trader could be left facing an extremely large import duty. They would be faced with the option of paying the tariff, or diverting the cargo and trying to unwind the arb at whatever level was available. Neither one of these options sounds great.

Last year the question was whether copper could reach the United States before the deadline. This year there is not even a definitive deadline to race against.

The US has already accumulated copper

The tariff threat has transformed the location of global copper inventories. US imports of refined copper jumped by 80% to 1.64 million mt in 2025. The pace has accelerated again this year, with imports rising 13% to 763,000 mt during the first five months of 2026. Shipping data from early August showed more than 200,000 mt arriving in July alone.

As of August 11th, CME warehouses held ~664,000 mt of copper warrants on stock. That includes around 419,000 mt of registered metal and another 245,000 mt of eligible stock that could be warranted at a given moment.

By comparison, LME stocks have fallen to a ~215,000 mt, just half of what they were in early May. Nearly 50% of all visible LME inventory has disappeared in just over 3 months. Of what remains, ~125,000 mt (58%) is already cancelled and awaiting load-out.

Perhaps one of the most interesting consequences of this trade is that the US government does not necessarily need to impose a cathode tariff to attract copper into the country. The threat of a tariff is already doing the job. Private traders are financing what increasingly resembles a strategic US copper stockpile, while the administration retains the option to act later.

China is attracting copper too

There is another major difference between 2025 and 2026. Last year's dislocation was dominated by the pull of metal to the United States. This year the LME is being pulled in two direction.

SHFE copper stocks have fallen from ~433,000 mt in March 2026 to around ~70,000 mt today. The Yangshan copper premium - a key indicator of Chinese demand for imported cathode - rose from $59/mt in June to a four-year high of $115/mt in late July. China is therefore competing with the US for the same units of LME cathode.

This competition is putting upward pressure on LME physical cathode premiums with producers increasing their asking price above $400/mt. At the same time the huge US domestic stockpile is putting downward pressure on US physical premiums.

Copper Concentrates

The refined market is also being supported by continued stress in the concentrates market. Annual copper TC/RCs fell from $80/mt and 8 cents/lb in 2024 to $21.25/mt and 2.125 cents/lb in 2025. The 2026 benchmark is zero. Spot treatment charges have been negative for months, meaning smelters are effectively paying miners for the right to process their concentrate.

Negative TCs do not automatically mean that refined production will immediately fall. Smelters can continue operating because of free-metal gains, precious-metal credits, and revenues from sulphuric acid. But they are a clear signal that mine supply is struggling to keep pace with global smelting capacity.

Disruptions and constraints at major mines have tightened concentrate availability, while the recent DRC ban on copper and cobalt concentrates adds another source of uncertainty. The DRC has allowed waivers under previous bans and most of its copper is already refined domestically, so the direct global impact may be limited. But in a market this sensitive to supply risk, even a relatively small disruption can have an outsized price impact.

The arb is therefore no longer operating in a comfortably supplied global cathode market. Traders are competing for metal in a system where raw-material availability, exchange stocks, and regional premiums are already under strain.

What happens next?

There are three broad outcomes:

1) A 15% (or higher) cathode tariff is confirmed with a future effective date.

In this scenario the CME premium would likely widen immediately, particularly in the 2027 forward curve. Metal already cleared into the United States would become highly valuable. Traders with uncommitted cathode outside of the US would race to import before the effective date, further tightening LME and Chinese markets. The arb may not reach the full $2,100 duty value but the upside from current levels is still substantial.

2) Cathode remains exempt.

If cathode remains exempt for now, the CME premium over LME would likely compress sharply. US physical premiums would still face pressure from record inventory but traders would look for alternative homes for excess metal. Some units could move back toward Europe or Asia, putting downward pressure on those regional premiums. Traders that previously locked in arb would retain that profit but those who paid high physical premiums and waited to hedge the arb would be exposed.

3) The administration says nothing.

This may be the most interesting outcome because it is effectively what is happening today. The uncertainty keeps a premium embedded in the CME forwards, continues to attract copper into the US and leaves traders paying the carry while they wait. The US stockpile grows without the inflationary shock of an actual duty.

2026 vs 2025

In 2025, the levels of arbitrage achieved were deemed once in a lifetime, making the decision to execute arb and move metal to the US a lot easier. In 2026, LME premiums have risen more quickly, and there is not the same margin available by locking in arbitrage trades. Traders must choose when to lock in that arb, where to locate the physical metal, how long they can finance it, and how much backwardation they can absorb.

The CME/LME arbitrage has become more than a simple difference between two futures prices. It is now a market-implied probability of a tariff.

The great copper arbitrage is still alive, how well traders can take advantage of it remains to be seen.