Mariia Volegova added up the copper. Canada's ten biggest listed copper miners, on her count, produced about 2.3 million tonnes in 2025. About 205,000 tonnes came from four Canadian operations. Highland Valley, Gibraltar, and Copper Mountain in British Columbia. Snow Lake in Manitoba. That is less than 9 percent, if her total is right. Zambia and the Democratic Republic of Congo supplied about 39 percent. Chile supplied about 31 percent. Those splits are her addition. The Toronto listing was the smallest fact on the page.
Here is the idea, and it is the only one. Buy the tonne you own, in the country that can stop it. A Canadian copper stock is not Canadian copper. The ticker is a mailing address. The ore is somewhere else, and the headline tonne is often a mine you only partly own. This is not a recommendation to buy or sell any miner.
One mine, twice the country
Kamoa-Kakula, in Congo, produced 388,838 tonnes of copper in concentrate in 2025. Ivanhoe Mines reported that figure. It landed inside a guidance range the company had already cut, after seismic shocks flooded part of the Kakula underground mine and halted work for weeks. That one complex still produced almost twice the copper of those four Canadian operations combined.
Ivanhoe owns 39.6 percent of it. Zijin Mining owns 39.6 percent. Crystal River owns 0.8 percent. The Congolese state owns 20 percent. Thirty-nine point six percent of 388,838 tonnes is about 154,000 tonnes. That is Ivanhoe's copper. It is less than Barrick's 2025 copper, which the company already states on an ownership basis, at 220,000 tonnes. A slide that prints 388,838 next to a Canadian ticker is printing the partners' copper too.
The flood is the second lesson inside the same number. The largest single mine behind this set sits in a country that is not Canada, and it can stop for a reason that has nothing to do with a hearing in Vancouver. The shares still trade in Toronto. The water was in Congo.
The flagship that produced nothing
First Quantum Minerals produced 395,772 tonnes of copper in 2025. None of it came from Cobre Panamá. The mine stayed on preservation and maintenance for the full year. Sentinel and Kansanshi, both in Zambia, produced 189,000 and 181,000 tonnes. Together, about 370,000. That is about 93 percent of the company total. The rest, about 26,000 tonnes, came from Çayeli in Turkey and Guelb Moghrein in Mauritania. Zambia carried the year. It did not produce every tonne.
Kansanshi is 80 percent owned, with 100 percent of the economic rights. The 181,000 tonnes are the tonnes the company claims. They are not a simple 80 percent cut, and they are not a Canadian tonne.
Panama has since allowed the company to process ore that was already stockpiled. The stockpile holds on the order of 70,000 tonnes of recoverable copper. First Quantum has been plain that this is not a mine reopening. A permit to mill rock that was dug before the shutdown is not a permit to dig again. An investor who reads "Panama" and hears "the mine is back" has not read the sentence the company wrote.
That is the cleanest case of the idea. First Quantum is a Canadian company. Its former flagship is not a Canadian mine. A government that is not Canada's turned the mill off. The listing did not turn it back on. If you own the stock, you own the political risk of the places that shipped metal, plus the unresolved risk of the place that did not.
The printed mine and the owned tonne
Teck Resources reported 453,500 tonnes of copper in 2025. Quebrada Blanca, in Chile, is in that table at 190,000 tonnes. Highland Valley, in British Columbia, is at 127,100. Carmen de Andacollo, also in Chile, is at 50,500. Antamina, in Peru, is at 85,900, and that line is already Teck's 22.5 percent. On the page, Chile is 240,500 tonnes, about 53 percent of the reported total. The post is right that the printed mine in Chile is larger than the printed mine in Canada.
The page is not the tonne Teck owns. Teck's own report says it books 100 percent of Quebrada Blanca and of Carmen de Andacollo, even though it does not own 100 percent of either. Teck owns 60 percent of Quebrada Blanca. Sixty percent of 190,000 is 114,000 tonnes. Teck owns 90 percent of Carmen de Andacollo. Ninety percent of 50,500 is 45,450 tonnes. Highland Valley is Teck's mine. All 127,100 tonnes are its tonnes. Antamina stays at 85,900. Add those and Teck's copper is about 372,000 tonnes, not 453,500.
On the tonne Teck owns, Highland Valley is larger than Quebrada Blanca. Chile is still the biggest country in the owned total, at about 160,000 tonnes, or about 43 percent. It is not more than half. The sentence "Teck's largest copper mine is in Chile" is true of the gross table. It is false of the copper the shareholder owns. That is the whole test.
Highland Valley is still a real Canadian mine, and a smaller one than the slogan needs. In July 2025 the board approved a life extension, after British Columbia issued the permits. The plan runs the operation from 2028 out to 2046, at an average of about 132,000 tonnes a year. That keeps a Canadian mine alive. It does not make Teck a Canadian orebody. Most of the owned copper is still outside Canada.
The merger makes the point sharper. Teck and Anglo American have cleared the deal with the regulators in the places where they operate, except China. Both sides have said they expect closing by March 2027. In October 2026, Reuters reported that China's antitrust regulator wants a commitment on copper concentrate supply before it says yes. Unnamed sources said it. The companies have not confirmed that demand. The listing will still say Canada. The approval that is left does not.
The name changed. The map did not.
In May 2025, Barrick Gold became Barrick Mining Corporation. Shareholders approved it on May 6. The New York ticker changed. Gold stayed in the business. The name was an admission that copper is now part of the story the company wants told.
The copper in the story is mostly not Canadian. Barrick produced 220,000 tonnes in 2025, up from 195,000, which is about 13 percent. That figure is already Barrick's share. Lumwana, in Zambia, produced 151,000 of those tonnes. The rest was Barrick's half of Zaldívar and its half of Jabal Sayid. Lumwana was about seven tenths of the copper, and it accounted for the year's growth. A name change in Toronto does not move a pit in Zambia.
Barrick is building Lumwana bigger. That may be a good mine. It is still a Zambian mine. The investor's question is not whether the word "gold" left the door. It is whether the model is using 151,000 tonnes in Zambia, on Barrick's books, and not a slogan about a broader metal.
Record years, and whose tonne
Capstone Copper reported a fourth record year in a row. Consolidated copper production was 224,764 tonnes, up 22 percent. Mantoverde produced 95,115 tonnes, up 65 percent. The company shows Mantoverde at 100 percent. Capstone owns 70 percent. Seventy percent of 95,115 is about 66,600 tonnes. The other 30 percent, about 28,500 tonnes, is not yours. And it is already inside the 224,764. Take it out, and if the other mines are wholly owned, the copper that belongs to Capstone's owners is about 196,000 tonnes. A record at the mine is not a record in your share count unless the release says whose tonnes they are. Read the footnote, or you will buy a partner's copper.
Hudbay Minerals met consolidated copper guidance for the eleventh year in a row. Full-year copper was about 118,188 tonnes. That sentence is true. It is not the whole map. Manitoba, which is the Snow Lake operation, produced about 9,249 tonnes of copper, inside a small range. Snow Lake is an underground mine, not a pit. British Columbia, which is Copper Mountain, produced about 23,784 tonnes. That was below the bottom of the site's own 2025 copper range, which started at 28,000. The company can meet the group target and still miss the Canadian mine you thought you were buying. "Met guidance" is a consolidated sentence. The miss was local. Both sentences belong in the note.
What the nine percent is, and is not
The four Canadian tonnes do add up. Highland Valley at 127,100, Gibraltar at 98 million pounds, which is about 44,450 tonnes, Copper Mountain at 23,784, and Snow Lake at 9,249 come to about 204,600. Her "around 205,000" holds. "Less than 9 percent" holds only if the 2.3 million total holds. That total, and the 39 percent and 31 percent splits, are her count of the ten. They were not rebuilt mine by mine here. They are plausible if Kamoa-Kakula is taken at 100 percent, which is what a headline slide does.
The nine percent is not a case that Canada should mine more, though the operations are few. It is not a case that Congo, Zambia, Chile, Peru, or Panama are uninvestable. Mines in those places produced the metal. Some of them had good years. Kamoa still hit a revised range after a flood. Kansanshi had its best copper year since 2021, helped by a new circuit. Lumwana rose. Mantoverde ramped.
The nine percent is a reading test. When a company is called a Canadian copper producer, ask three questions. Where was the tonne mined. What percent of that tonne do you own. Which government can turn the mill off. If the slide does not answer all three, the slide is a flag.
Some of these miners print a mine at 100 percent. Some print their share. Barrick's 220,000 is already its share. Ivanhoe's 388,838 is the whole complex. Capstone's 224,764 already includes 100 percent of Mantoverde. Teck's 453,500 includes 100 percent of two Chilean mines it does not wholly own, and only 22.5 percent of Antamina. Put those numbers in one column and you are mixing your copper with someone else's. The country error and the ownership error compound. You can think you own a Canadian tonne, and instead own two-fifths of a Congolese tonne. Or you can think Teck's biggest tonne is in Chile, when the tonne it owns is larger in British Columbia.
What you can underwrite
You can underwrite a mine you can find on a map, at the percent the annual report gives you, under the rules of the place that issues the permit. Highland Valley at 127,100 tonnes is that kind of tonne. It is Teck's. It is smaller than the Quebrada Blanca line on the gross table, and larger than Teck's 60 percent of that line. It is a tonne in British Columbia, with a life-extension plan and a permit in hand. Snow Lake's copper is a small tonne, from an underground mine. Do not let a gold story at the same camp inflate it. Copper Mountain's 2025 tonne missed its own range. That is information. It is not a reason to pretend the miss was Peru.
You can underwrite Zambia, Chile, or Congo if you have priced the country. Sentinel and Kansanshi produced most of First Quantum's copper. Lumwana produced most of Barrick's copper, and the growth. Kamoa produced the single biggest line in the set, and a Canadian shareholder owns less than two-fifths of it. Those can be rational holdings. They are not domestic holdings. The discount, or the premium, belongs to the country and the contract, not to the exchange logo.
You can underwrite the Panama stockpile as a stockpile. About 70,000 tonnes, if the permit is used, is a year of milling old rock. It is not a restarted mine. The company said so. Believe the narrower sentence.
What would make this reading wrong
The reading is wrong if you already buy on attributable tonnes and on country, and the word "Canadian" never enters the model. Then the nine percent is a lesson for the timeline, not for you. The error lives in the headline number, the flag in the title, and the slide that forgets the footnote. Teck's footnote is the proof. The gross table and the owned table do not pick the same mine.
The reading is wrong if the Canadian share of the tonne rises enough to change the map. Highland Valley's extension is a start. It averages about 132,000 tonnes a year, which is one large mine, not a national copper base. Four operations at about 205,000 tonnes do not become a new Chile because one of them gets eighteen more years. Watch the tonnes, not the ribbon-cutting.
The reading is wrong if a government you feared does not, in practice, touch the mine, and a government you trusted does. Panama already broke the comforting version of that sentence. A Canadian listing did not save Cobre Panamá. British Columbia can also delay, shrink, or refuse a mine. "Mined in Canada" is not a synonym for "safe." It is a synonym for a different rulebook. You still have to read it.
The reading is also wrong if you use it as a reason to ignore a good mine because the passport is wrong. The point is not to buy the maple leaf. The point is to stop paying maple-leaf prices for a tonne that is 39.6 percent yours, or 60 percent yours, and a long way from the exchange.
The idea, once
Less than nine percent of the copper from Canada's ten biggest listed copper miners, on this count, was mined in Canada in 2025. The four Canadian operations add up to about 205,000 tonnes. Kamoa-Kakula alone nearly doubled them, and the Canadian partner owns 39.6 percent. First Quantum shipped nothing from Cobre Panamá. Barrick's copper growth was mostly Lumwana. Capstone's record includes a partner's 30 percent of Mantoverde. Hudbay met the group target while Copper Mountain missed its own. Teck's printed copper is 453,500 tonnes, and the printed Chilean mine is larger. The copper Teck owns is about 372,000 tonnes, and the larger owned mine is Highland Valley. Buy the tonne you own, in the country that can stop it. The listing is not the mine.
A note on sources and limits
The tally of about 2.3 million tonnes, and the 39 percent and 31 percent splits, is Mariia Volegova's addition, posted on LinkedIn. It was not rebuilt company by company here. The Canadian tonnes were. Teck's Highland Valley was 127,100 tonnes in 2025. Taseko's Gibraltar was 98 million pounds, about 44,450 tonnes. Hudbay's British Columbia copper was about 23,784 tonnes, below a site range that started at 28,000. Hudbay's Manitoba copper was about 9,249 tonnes. Those four sum to about 204,600 tonnes.
Ivanhoe: Kamoa-Kakula 388,838 tonnes in 2025, Ivanhoe ownership 39.6 percent. First Quantum: 395,772 tonnes, Kansanshi 181,000, Sentinel 189,000, and no Cobre Panamá output. Kansanshi is 80 percent owned with 100 percent of the economic rights. Other copper that year was about 26,000 tonnes. The stockpile figure of about 70,000 tonnes, and the line that processing it is not a reopening, are the company's.
Teck, January 20, 2026, reported copper of 453,500 tonnes: Quebrada Blanca 190,000, Highland Valley 127,100, Carmen de Andacollo 50,500, Antamina 85,900 at 22.5 percent. Teck's later report says it includes 100 percent of Quebrada Blanca and Carmen de Andacollo even though it owns 60 percent and 90 percent. On that basis the owned tonnes are about 114,000, 127,100, 45,450, and 85,900. Barrick mine statistics: 220,000 tonnes of copper in 2025, of which Lumwana was 151,000, already on an ownership basis. The name change was approved May 6, 2025. Capstone: 224,764 tonnes, up 22 percent, a fourth record year. Mantoverde 95,115 tonnes on a 100 percent basis. Capstone owns 70 percent, and that 100 percent figure sits inside the consolidated total. Hudbay: about 118,188 tonnes, eleventh straight year of consolidated copper guidance. The Anglo Teck timing is from the companies. The report that China wants concentrate-supply assurances is Reuters, October 2, 2026, from unnamed sources. Nothing here is investment advice. Read the filings. The footnote is the mine.