01 THE SETUP

So far, tariff uncertainty has pulled around 1.1 million tonnes of copper into the United States. Of that, about 711,600t sits in COMEX warehouses, another 127,300t in US LME sheds, another 123,800t off-warrant, while the rest is off-market metal.

And all of it is privately owned.

This week, Reuters’ Andy Home asked whether the country can actually hold on to it.

02 WHY IT MATTERS

Pulling metal into the US has already changed the market. US refined imports nearly doubled in 2025, to about 1.70Mt, COMEX has traded at a premium to LME, and US mills have turned to cheaper LME-priced alternatives such as Congolese cathode. And that’s left the rest of the world living with an apparent deficit, despite the fact that we’re theoretically in surplus.

So where it goes next is worth paying attention to. If it’s exported, that supply goes back to the rest of the world. If it’s used at home, it displaces imports. And because it’s privately owned, either can happen without a decision from Washington.

03 THE DECODE

Home’s Reuters column sets out the broad picture — the metal is in the US because of market structure, it could leave if that structure shifts far enough, and rising cash premiums in London and Shanghai are already starting to pull against the CME premium.

What the column doesn’t get into is how far things would actually have to move. That’s where we want to pick up — by looking at three basic risks for the U.S. stockpile and attempting a rough estimate of some of the real costs to get a sense of what would need to change before the stockpile starts to shrink.

1. Exports

Exporting makes sense when the price an owner can get abroad, after freight, beats what the metal fetches in the US. But getting it out isn’t free.

At the Henry Bath warehouse in New Orleans, load-out, bracing and weighing come to about $79/t. Meanwhile, container freight to Rotterdam adds roughly $52-94/t, assuming a approximately 23 tonnes of copper per 20-foot container. Throw in insurance and port charges, and a rough estimate of the total comes to somewhere $165-300/t.

Exporting metal stored at the CME-approved sites in Glendale or Tucson, Arizona, may fall within a similar range. While it would require trucking to Los Angeles ($40-90/t), lower westbound shipping rates are available given the substantial imbalance in trans-Pacific container trade, potentially bringing the container freight portion down significantly.

So for exports to start, US prices have to fall below LME plus the local premium in Europe or China by more than those amounts.

That might come from firmer premiums overseas, a narrower US premium, or the market giving up on a US duty.

2. US buyers switch from imports

The stockpile can also shrink without leaving the country if US rod mills and fabricators buy it instead of importing. That comes down to delivered cost, which currently favors imported metal — Reuters reported in September that US mills were buying Congolese cathode that isn’t COMEX-approved, priced off LME and in some cases well below COMEX-priced metal. So the question is what would have to change for warehouse metal to become the cheaper choice.

We can illustrate with a simple example.

Suppose:

The delivered cost to a US mill would be:

Now change one thing at a time:

With these costs, warehouse metal wins once the COMEX premium, less any discount, is more than $25/t below the import’s extra costs.

The premium doesn’t have to go to zero, and a holder doesn’t have to wait for the futures market to move. They can simply offer a discount, giving up the COMEX-linked price in exchange for avoiding the roughly 0.5% a year it costs to keep holding (see below for estimation) if future tariff benefits look less attractive (or the economics of continued storage deteriorate).

With that said, the drawdown would be slow unless imports fall a long way. On a simplified balance using USGS figures, a return to 2024 import volumes, with production, consumption and exports otherwise unchanged, would imply a stock draw of only around 57,000 tonnes annually. Getting to 300,000 tonnes would require imports to fall to roughly 660,000 tonnes.

3. Holding stops paying

This is the one we can put the firmest number on. A hedged cash-and-carry holder earns the contango and pays for storage and financing.

On 12 October, December 2026 COMEX settled at $6.7470/lb ($14,875/t) and December 2027 at $7.1425/lb, a spread of $872/t, or 5.86% over the year. Storage at Henry Bath in New Orleans is $15.50 per short ton a month — about $205/t a year.

For funding, we’ll use Glencore as a reference — on 2 October it disclosed its marketing business pays roughly 5%, tied to one-month Term SOFR. On $14,875/t, that’s $744 a year.

That leaves net carry of about −$77/t a year before insurance and fees. Once storage is paid, the spread covers funding up to about 4.5%. Each percentage point on the funding rate is worth about $149/t, and at 5% the Dec 26-Dec 27 spread would need to widen to around $0.43/lb to break even.

That estimated $77/t annual shortfall is only about 0.5% of the metal’s value. A trader expecting to benefit from a future US tariff premium might consider that a reasonable price to pay for keeping copper in the country.

However, none of this is set in stone. In just the past year, financial conditions have shifted. The 10-year Treasury yield has risen from about 4% to over 5% in the last year, raising the opportunity cost of tying up capital in inventory. Of course, short-term rates, which set funding costs, have moved much less. But any rise there would squeeze the trade further unless contango widens to compensate.

There’s also the question of how long a hedged trader can can keep waiting. While they can lock in today’s futures spread for a specified period, maintaining the inventory beyond that means extending or rolling the hedge and financing at the terms available then.

So the holding calculation can change if funding becomes more expensive, the curve flattens, or the expected payoff from tariffs diminishes.

04 WHAT TO WATCH
 
Tariff policy: A confirmed duty could increase the value of copper already in the US. A definitive decision against refined copper tariffs would remove a major incentive for keeping it there.

International spreads: COMEX relative to LME and the physical premiums in Europe and China. Our illustrative export costs are roughly $165-300/t, depending on where the metal is stored and transport rates at the time of export.

Domestic flows: US refined copper imports and changes in both registered and off-warrant inventories. Falling imports alone won’t necessarily shrink the stockpile — the question is whether new supply continues to exceed domestic consumption.

Financing and carry: Short-term funding rates against COMEX futures spreads. Our October 12 calculation suggests that the Dec 2026-Dec 2027 spread covers financing at roughly 4.5% after storage. A higher funding rate or flatter curve makes continued holding more expensive.

/ European Commission Picks 46 New Strategic Critical Mineral Projects

Brussels designated 46 projects across 16 countries for accelerated permitting and financing assistance, including copper mining, processing and recycling developments. The projects represent an estimated €21.1 billion in investment, although strategic status comes with no guarantee of funding.

/ Striking Unions and Centinela Copper Mine Restart Negotiations

Negotiations have resumed after a week-long strike at Antofagasta’s Centinela mine in Chile. Unions say truck activity has halved and warn that production losses could become substantially more severe in November, while the company maintains that the strike affects only 22% of its workforce.

/ China’s Copper Inventories Rise, but Remain Below Last Year’s Levels

Chinese copper inventories rose 34,000 tonnes over the past week as arrivals outpaced purchases by downstream consumers. However, despite the post-holiday buildup, stocks remain 64,300 tonnes below the same period last year.

/ Glencore Sees Electrification Driving Demand for Australian Listing

Glencore is preparing to list on the Australian Securities Exchange on October 14. Analysts estimate copper could contribute around half of the company’s earnings by 2030, although coal remains a significant part of the business.

That’s the wire for today. Until next time, may your contango cover your carry.

Know someone who keeps asking you about copper?

They can subscribe free at copperwire.news.