|
Part I
The Mechanism
|
Copper.
LME warehouse stocks just hit a four-month high. Bloomberg ran the number without blinking. The sell-side read it as easing. "Copper inventories swell," said the chyron. A fund manager on CNBC used the phrase "building surplus" with a straight face. If you are looking at total stocks and concluding that global copper supply is loosening, you are reading a map upside down.
Here is what no one on the terminal seems to have done: check which warehouses the metal went into.
Seventy percent of the world's visible exchange copper is sitting inside the United States. Not because Americans suddenly need more copper — because traders are front-running a tariff. Section 232 hit semi-finished copper at 50% in August 2025. The Commerce Department has proposed an additional 15% on refined copper starting 2027, scaling to 30% by 2028. So metal is flooding into US bonded LME warehouses where it can sit duty-free until someone decides what to clear it through customs as.
Meanwhile, on the other side of the Pacific, copper inventories across major Chinese markets have collapsed to roughly 62,600 tonnes — down 53% from the same week last year. Shanghai spot premiums just hit 2,615 yuan per tonne, the highest level ever recorded. The Yangshan import premium — what it costs to actually land copper in China — reached $195–210 a tonne, a record in Fastmarkets data going back to 2017.
The headline says surplus. The loading dock says famine. The difference is a hemisphere and a tariff wall nobody wants to talk about.
I’m about to get into this Cybercab… and it’s going to take me to what Elon Musk called…
“The most epic AI project ever.”
Because this is all part of Elon Musk’s $320 trillion masterplan.
|
Part II
The Diagram
|
Story off. Wiring diagram on.
On October 30, LME copper stocks stood at 274,000 tonnes. By November 2, they were 293,775 — a jump of 19,775 tonnes, 7.2%, in a single session. By November 3, Reuters reported the total had hit 309,650, a four-month high. Almost all of the inflow landed in New Orleans and Baltimore. As of November 5, US warehouses held 223,300 tonnes — 70.1% of total LME stocks, the highest US share since March 2019.
Now the part that should stop you cold. On November 2, LME cash copper settled at $14,380 against $14,242 for three-month delivery. That's a $138 backwardation — the market paying up for the privilege of having metal now rather than in ninety days. A build of twenty thousand tonnes in a single session is precisely the event that should kill a backwardation. It didn't. The metal arrived, and the prompt market still charged for immediacy. When supply rises and the price of urgency doesn't fall, the supply is not reaching the people who are urgent.
The mechanism is straightforward. Section 232 imposed 50% tariffs on semi-finished copper from August 2025. Commerce proposed an additional 15% on refined copper from 2027, rising to 30% from 2028. Traders responded rationally: move metal into US bonded LME warehouses, where it sits duty-free and can transfer to COMEX once cleared. The COMEX-LME spread narrowed to effectively nothing — on September 29 the October COMEX contract traded at a $46.94 discount to LME three-month. That spread doesn't cover freight, financing, storage, or insurance. The metal went in. It is not coming out.
The ICSG's own numbers make the mirage plain: the global refined copper market has swung from a projected 150,000-tonne deficit to a surplus of 104,000 tonnes for 2026 — on paper, a market in balance. But the surplus sits in American warehouses while the July monthly balance alone showed a 51,000-tonne shortfall. This is not a market loosening. It is a market fracturing along a tariff line.
A tonne of copper in Baltimore is not a tonne of copper in Ningbo. That sentence is the whole market right now, and the stock total cannot say it.
|
Part III
The Weak Link
|
Kamoa-Kakula. Largest new copper mine on the planet. The one that was supposed to fix the supply gap.
In December 2025, Ivanhoe told the market to expect 500,000 to 540,000 tonnes for 2027. In March 2026, they walked it back to 380,000 to 420,000 — seismic disruption, hydrology problems, a more cautious mine plan. Last week, in their Q3 results, they cut it again: 300,000 to 350,000 tonnes. Cash cost guidance went from $2.10–$2.50 per pound to $2.50–$3.00. Two guidance cuts in one calendar year from the mine the bulls spent five years pointing at as the supply cavalry. The cavalry arrived, dismounted, and asked for more diesel money.
That is not the weak link. The weak link is what happens when this mine's shortfall collides with the tariff trap.
This is the configuration I've learned to pay close attention to — and I've been burned often enough by commodity positioning data to say that with appropriate humility. When financial players lighten up on a commodity because the total inventory number looks soft, while the physical buyers in Shanghai are paying the highest premiums in recorded history to get their hands on actual metal, somebody is wrong. One side is reading a spreadsheet. The other is trying to keep a wire plant running.
The COMEX-LME spread has compressed to the point where arbitrage no longer works. On September 29, October COMEX traded at a $46.94 discount to LME three-month. That doesn't cover freight, insurance, financing, or warehousing. So the metal that went into the US is effectively stuck there — marooned behind a tariff wall that only goes higher. You cannot redirect a Baltimore warehouse receipt to solve a shortage at a smelter outside Ningbo. The logistics don't pencil. The tariff structure actively prevents it.
And now Kamoa-Kakula's 150,000 to 190,000 tonnes of missing supply versus its original December 2025 guidance walks straight into a market where the ICSG's 104,000-tonne headline surplus is trapped in US warehouses while Chinese stocks collapse by half. If you are running a cable factory in Guangdong right now, your procurement team is not looking at LME warehouse totals. They are looking at the Yangshan premium and doing math that does not make them feel calm.
|
Part IV
The Chain Reaction
|
Two scenarios, same plumbing, different triggers.
The second scenario is subtler and frankly more likely. Commerce delays. The tariff proposal lingers into Q1 2027. Metal stays parked in US warehouses on optionality — traders can't afford to ship it out because the tariff might drop any day, and they can't afford to clear it through customs because they'd eat the duty. The copper sits. China's deficit deepens. The rest of the world competes for a shrinking pool of non-US metal while the ICSG's headline surplus sits locked in American warehouses.
I watched exactly this dynamic play out with aluminium premiums in 2018 after the Section 232 steel and aluminium tariffs. The US became a roach motel for metal — it went in, it didn't come out, and the rest of the world tightened while headline inventory looked comfortable. Copper is following the same blueprint, except the underlying market is already fractured along a tariff line and the largest new mine just cut its output guidance by roughly 37%.
Where does the edge sit? Not in the copper price itself — that's a blended number averaging two fundamentally different markets. The edge is in the spread. Producers who sell into the Asian physical market at Yangshan-referenced premiums are capturing a $200+ per tonne bonus that doesn't show up in their forward curves or in the models the sell-side runs off LME flat price. Their realized revenue is diverging from the price the analysts use to value them. That divergence is the trade.
The inventory build is real. The surplus is not. One is a number on a warehouse receipt in Louisiana. The other is a cable plant in Guangdong that can't get metal at any price its margin can absorb. The headline will keep saying one thing. The premium will keep saying the other. In my experience — and I have been wrong about copper more than once, so take this for what it is — the premium eventually wins. It just takes longer than the model expects.
Sources: LME, COMEX, SMM, Fastmarkets, Reuters, ICSG, CFTC, Ivanhoe Mines Q3 2026, U.S. Dept. of Commerce
