This article is for information only. It is not investment advice. It is not a recommendation to buy or sell any security, metal, or futures contract. Copper prices, warehouse stocks, and tariff rules can change fast. Do your own work. Speak with a licensed adviser if you need one.
Copper has two stories at once. One story is surplus. The other is shortage.
The surplus story is easy to print. Combined stocks on the London Metal Exchange, COMEX, and the Shanghai Futures Exchange reached about 997,600 tonnes at the end of August 2026. That was the highest level since June 2003. The International Copper Study Group said the world refined market showed a preliminary surplus of about 32,000 tonnes in the first seven months of 2026.
The shortage story is harder. It lives in the mine, not the warehouse. World mine production fell about 0.8% in those same seven months. Treatment charges for concentrate hit a zero annual benchmark. Spot charges went deeply negative. China warehouse stocks fell. LME metal available for prompt delivery shrank. Cash copper traded above three-month metal. That is backwardation. It means buyers will pay more for metal now than later.
Then came the tariff headline. Semi-finished copper already faces heavy Section 232 duties. Refined copper cathodes do not. Commerce was asked to update the President by June 30, 2026, on a possible phased duty on refined metal. As of late September, no new proclamation had imposed that duty. A mid-September report that the White House plan had stalled knocked more than 5% off the price from the early-September peak.
That is the mirage. Metal piled up in U.S. sheds ahead of a tariff that has not arrived on refined copper. Headline stocks rose. Prices sold off. The mine market did not suddenly become loose.
This piece asks one question. If the surplus is mostly location and accounting, does a tariff delay create a research window in copper mining stocks? It does not tell anyone to buy the dip. It maps the data so readers can decide.
What the latest ICSG numbers actually say
The International Copper Study Group released its September 24, 2026 monthly bulletin. It is the cleanest official scorecard. It is also lagged. July is the latest full month in that cut.
World refined copper usage rose about 2.5% in January–July 2026 versus the same period in 2025. Refined output also rose. Secondary metal from scrap did a lot of the work. Mine output did not.
World copper mine production was about 13.32 million tonnes in the first seven months. That was down 0.8% from a year earlier. Concentrate output fell more than SX-EW cathode in several reports. That split matters. Smelters eat concentrate. When concentrate is scarce, treatment charges collapse. That is what happened.
The headline refined balance was a 32,000-tonne surplus for January–July. That sounds like oversupply. Compare it with 2025. The same period last year showed a surplus near 157,000 tonnes. The surplus shrank hard.
ICSG then does a second calculation. It adjusts for estimated changes in Chinese bonded stocks. Those stocks are thought to have fallen by about 37,000 tonnes in the first seven months. After that adjustment, the world balance flips. ICSG put the adjusted figure at a deficit of about 5,000 tonnes.
A 5,000-tonne deficit on a 16-million-tonne seven-month market is a rounding error. It is not a crisis. It is also not a glut. It is close to balance after you look behind China’s bonded warehouses.
ICSG itself warns readers. Its China demand figure is apparent usage. It does not fully capture unreported stocks at the State Reserve Bureau, producers, consumers, or merchants. Those hidden piles can swing the global balance. They have done so before. They can do so again.
In April 2026, ICSG projected a full-year 2026 surplus of about 96,000 tonnes and a 2027 surplus of about 377,000 tonnes. That April note already admitted the October 2025 forecast had been a 150,000-tonne deficit. The group flipped because usage came in softer than hoped and secondary refined output rose. It also said actual balances often miss the forecast when mines break or wars disrupt acid and freight.
Hold both facts. The official full-year model still prints surplus. The latest year-to-date print, after China bonded stocks, prints a tiny deficit. Prices near record highs voted with the tight version of the story for most of September.
Why warehouse stocks can lie about tightness
Visible exchange stocks rose 253,480 tonnes, or 34%, from the end of December 2025 to the end of August 2026. COMEX did most of the heavy lifting. ICSG said COMEX stocks rose about 238,223 tonnes. LME stocks rose about 88,175 tonnes. SHFE stocks fell about 72,914 tonnes.
That is not a global flood. That is a transfer.
Metal moved toward U.S. delivery points. Importers and traders pulled cathodes into COMEX warehouses while Washington debated refined-copper tariffs. By late September, COMEX held the large majority of exchange-monitored copper. One widely cited mid-September snapshot put COMEX near 696,000 tonnes, or about 69% of tracked exchange metal. Other late-September warehouse tallies put total COMEX registered plus eligible metal even higher, above 770,000 short tons depending on the cut.
London told a different story. Headline LME warehouse stock sat near 250,000 to 256,000 tonnes in late September. Cancelled warrants took a huge slice. Cancelled warrants are metal already booked to leave the shed. Several reports put cancelled warrants near 45% to 48% of LME stock. Metal freely available on warrant fell toward about 133,725 tonnes.
Shanghai told a third story. SHFE warehouse stocks dropped about 70% from early June levels in some weekly briefs. Shanghai cathode inventories in the main consumption hub fell toward multi-year lows near 44,000 tonnes in one mid-September reading. The Yangshan import premium sat near four-year highs. Incoming metal was going to fabricators, not to sheds.
LME cash copper moved from an $86-per-tonne discount to three-month metal in mid-September to a premium days later. Backwardation is not a surplus signal. It is a “we need it now” signal.
COMEX stocks also stopped growing. After months of inflows, weekly figures slipped. Reports said New Orleans, a key COMEX hub, was largely full. More African and South American metal was still scheduled to arrive. The United States was running out of cheap warehouse space. That is not the same as the world having too much copper. It is the same as the world putting too much copper in one country for a policy that never fully landed on refined metal.
This is the copper surplus mirage. Headline tonnes are high. Location is wrong. Availability where fabricators buy is tighter than the global sum.
What “tariff rejection” actually means in 2026
The phrase needs care. The United States did not reject copper tariffs as a whole.
Section 232 copper duties already exist. President Trump first imposed 50% tariffs on certain copper imports in 2025. Those duties hit semi-finished products such as pipes, wires, rods, sheets, and tubes. They also hit many copper-intensive derivatives. They did not hit copper ores, concentrates, mattes, anodes, cathodes, or scrap in that first wave.
In April 2026, Proclamation 11021 changed how the duties are calculated. Many rates now apply to the full customs value of the article, not only the metal content. Annex lists split products into 50% and 25% buckets. Some industrial equipment got a temporary 15% rate through 2027. Products with high U.S.-smelted and cast copper content can pay 10%. A June 1 proclamation lowered the U.S.-content threshold from 95% to 85% for that relief path and tweaked more derivative lists.
Refined copper remained the open question. The Commerce Secretary was to update the President by June 30, 2026. The public discussion included a possible 15% duty on refined copper in 2027 and 30% in 2028, plus domestic sales ideas for input materials and scrap. Trade lawyers tracking the file in late September said they had found no proclamation that actually imposed those refined-cathode duties.
That delay is the “rejection” in the market’s shorthand. It is more accurate to call it a stall.
On September 10, Reuters reported the White House plan had stalled. COMEX copper had just printed a record settlement near $6.8885 a pound. The report knocked more than 5% off the price. The metal later climbed back toward those highs, then faded again when the dollar firmed and the Federal Reserve sounded hawkish.
A tariff stall can create a buying opportunity in two ways. First, it can dump the price without dumping the mine shortage. Second, it can strand metal in U.S. warehouses that the rest of the world still needs. If those tonnes later leave COMEX for China or Europe, the headline surplus shrinks in the places that set the LME price.
A tariff stall can also create a trap. If Washington later slaps 15% or 30% on refined copper, more metal may rush into the United States again. COMEX premiums can blow out. LME tightness can deepen. Or U.S. manufacturers can scream about cost and force another delay. Policy risk cuts both ways. It is not a free gift to bulls.
This article treats the stall as a catalyst for a selloff, not as proof that tariffs are dead forever.
Mine supply is the part that is not a mirage
Refined surplus can appear while mines shrink. That sounds impossible. It is not. Scrap and existing concentrate stocks can feed smelters for a while. SX-EW plants can keep leaching oxide ore. Then the cupboard shows.
Sprott and several mine-side notes said global mined output could fall in 2026 for the first time since 2017. First-half mine production was already down about 1.1% in some cuts. ICSG’s seven-month figure was down 0.8%.
Two names dominate the disruption math. Freeport-McMoRan’s Grasberg complex in Indonesia suffered a major 2025 incident. Concentrate output fell sharply in 2026. Guidance was cut by a large slice versus earlier plans. Full recovery talk now points toward late 2027 in several company and analyst notes, not this year.
Ivanhoe Mines’ Kamoa-Kakula complex in the Democratic Republic of Congo also lost expected tonnes. Flooding, seismic issues, and later acid constraints all featured in 2025–2026 coverage. Combined, Grasberg and Kamoa-related hits have been estimated near 600,000 tonnes of expected 2026 mine supply. That is about 2.5% of world mine output. In copper, 2.5% is a lot.
Chile added its own drag. The world’s largest copper country saw weak first-half output. Storms and power cuts hit operations. Antofagasta cut 2026 guidance after weather shut Los Pelambres. Lundin Mining cut Caserones guidance after an Atacama winter storm. National forecasts were reduced. Chile is no longer a quiet baseload.
First Quantum’s Cobre Panamá remains a political asset, not a running engine. Stockpile processing is not a restart. Some bearish bank models still assume a 2027 restart and a ramp toward prior scale. That assumption is a policy bet. It is not metal in a ship today.
Treatment and refining charges tell the concentrate story in one number. The 2026 annual TC benchmark settled at $0 per tonne. That is the lowest on record in the modern Antofagasta–China smelter negotiation series. Spot TCs went negative years ago and stayed there. Some late-2026 prints went deeper than minus $200 a tonne in aggressive spot talk. Negative TCs mean smelters pay for feed. They do that when they fear idle lines more than they fear bad margins.
China still dominates smelting. It has more furnace capacity than free concentrate. Beijing has talked about curbing new smelter builds and trimming run rates. Those cuts have not restored a normal TC. Custom smelters outside China feel it first.
Sulphuric acid is another hidden pipe. Many SX-EW and concentrate processes need acid. China has restricted acid exports at times in 2026. Middle East conflict talk has added freight and sulphur risk in other notes. Acid is not copper. Without it, some copper never gets made.
This is why the surplus can be a mirage. Refined metal can look ample in U.S. sheds while the next tonne of mine feed is missing.
Demand is not a cartoon
Bulls talk about grids, EVs, and data centers as if demand only rises. That is lazy. High prices destroy some use. China property is not the copper engine it was. Substitution into aluminum happens at the margin in some cables and heat exchangers. ICSG already cut its 2026 usage growth view once. April’s forecast put world apparent refined usage growth near 1.6% for 2026, down from an earlier 2.1% idea.
Still, the demand side is not empty.
China remains about 59% of world refined copper use in ICSG’s recent framing. Chinese net refined imports fell 10% in the January–July snapshot. That sounds weak. Bonded stocks also fell. Fabricators were drawing metal, not building a fat cushion. Holiday restocking into late September and early October added a seasonal bid. Markets closed around the Mid-Autumn window and again around National Day.
Power grids still eat copper. Electrification is slow and real. Data centers add a new slice. Bank notes have thrown around figures of several hundred thousand tonnes a year for hyperscale power and cooling gear. Those estimates vary. They should be treated as scenarios, not census data. The direction is not a mystery. More compute needs more metal and more power cable.
The rest of the world grew usage too, just slower than China in the official tables. That is enough to keep the market near balance when mines shrink.
Macquarie has been the loud skeptic. Mid-year it argued the rally ran ahead of the physical market. It pointed to large visible stock builds since 2025. It sketched 2026 and 2027 surpluses if Cobre Panamá returns and Grasberg heals on a friendly timetable. That is a coherent bear case. It depends on mines behaving. 2026 mines have not behaved.
JPMorgan, Morgan Stanley, UBS, and others have published deficit-style 2026 or 2027 views at various points this year. The range is wide. One house says surplus. Another says 300,000 to 600,000 tonnes short. Readers should not pretend the Street agrees. They should watch the monthly ICSG print, cancelled warrants, SHFE stocks, and TCs. Those four beat any annual slide deck.
Price action around the tariff headline
LME three-month copper printed a record near $14,875 a tonne on September 10, 2026. COMEX settlement records sat near $6.8885 a pound the day before the stall report. The selloff that followed was sharp. LME traded down toward about $14,000 and briefly under that in mid-month session lows. That was roughly 6% off the high.
The metal then recovered. By the week ending September 25, copper had again tested the mid-$14,000s and threatened the old high. COMEX December traded above $6.80 a pound on several sessions. A firmer dollar and Fed hike talk capped the rebound. Cash-to-three-month LME spreads flipped to backwardation even as the headline price chopped.
That pattern is the investor puzzle. A policy headline created a gap. Physical tightness tried to close the gap. Rates and the dollar fought back. None of those three forces has left the tape.
For copper mining stocks, the same week can look like a gift and a trap. Equities often fall harder than the metal on a tariff scare. They can also bounce harder if the scare fades and TCs stay negative. Beta cuts both ways. Position size matters more than slogans.
Canadian copper stocks in this tape
Canada does not set the LME price. It does list many of the names that live on that price.
Teck Resources remains a diversified producer with a large copper growth story at Quebrada Blanca and other assets. A high copper price lifts the copper division. A tariff maze on fabricated metal is less direct than a cathode duty would be. Teck is still a stock. It trades with Canada’s dollar, coal sentiment, and project execution. It is not a pure copper ticket.
First Quantum Minerals is the Cobre Panamá story whether investors like it or not. Kansanshi and other operating mines still ship metal. The market prices the company as a balance-sheet and jurisdiction puzzle as much as a copper puzzle. A refined-copper tariff stall does not reopen a closed Panamanian pit. Readers who study First Quantum should start with law and cash, then add the copper tape.
Lundin Mining has Chile and other Americas exposure. Weather already cut Caserones guidance. That is the mine-side tightness showing up in a Canadian name. Higher prices help. Missed tonnes hurt. Both can be true in one quarter.
Hudbay Minerals is smaller and more torque-heavy. Copper and gold both move the model. A copper surplus scare hits the multiple. A confirmed mine deficit can expand it. Liquidity is thinner than in the senior names. Drawdowns can be rude.
Ivanhoe Mines lists in Toronto. Kamoa-Kakula is one of the disruption sources in the 2026 maths. The stock is a bet on the asset healing and on DRC risk. It is not a quiet utility. Treat it as a high-beta research file.
Capstone Copper, Foran Mining, and other developers and junior producers sit further out on the risk curve. They can move more than the metal on any surplus-versus-deficit argument. They can also fail to finance if the tape stays choppy. A “buying opportunity” in a developer is often just a cheaper option on a permit and a raise.
None of these names is a recommendation. They are the liquid Canadian ways to express a view if a reader already has one. Filings, guidance, AISC, and jurisdiction come first. The ICSG table comes second.
How to test whether the surplus is real
Investors do not need a crystal ball. They need a checklist.
First, watch the next ICSG monthly. If the unadjusted surplus stays near 30,000 tonnes or less for another two months, the glut story weakens. If it jumps back above 100,000 tonnes a month, Macquarie’s warning gains weight.
Second, watch the China adjustment. Bonded-stock draws flipped the seven-month balance to a small deficit. If bonded stocks start rising again, the apparent tightness was destocking. If they keep falling while SHFE sheds stay lean, China is consuming more than it is showing.
Third, watch LME cancelled warrants and the cash spread. High cancellations plus backwardation mean prompt tightness. A swing back to fat contango and rising on-warrant stocks would support the surplus camp.
Fourth, watch COMEX. If U.S. warehouse stocks start a sustained decline after the tariff stall, metal may be leaving for the rest of the world. That would tighten LME. If COMEX starts filling again, traders still believe a refined duty is coming.
Fifth, watch treatment charges. Zero annual TCs and deep negative spots are the opposite of a mine glut. A sharp TC recovery toward $20 or $50 would mean concentrate got easier. That would be the first honest surplus signal on the supply side.
Sixth, watch Grasberg, Kamoa-Kakula, and Chile monthly output. The 600,000-tonne hole is the core of the 2026 bull case. If those assets surprise to the upside, the hole fills. If they slip again, the refined surplus can vanish even with strong scrap.
Seventh, watch the Federal Reserve and the dollar. Copper is a metal and a risk asset. A stronger dollar and higher real yields can cap price even when sheds in Shanghai are empty. That is how you get a tight physical market and a soft tape at the same time.
Eighth, watch the actual proclamation file. A refined-copper duty of 15% in 2027 would reprice U.S. premiums. A clear decision to leave cathodes untaxed would let some COMEX metal leave. Either path ends the current fog. Fog is what created the mid-September air pocket.
People also asked
Why copper surplus may not last
Because much of the visible stock sits in U.S. warehouses after tariff front-running. Mine output is down. Concentrate is tight. China bonded stocks fell enough to turn ICSG’s adjusted January–July balance into a small deficit. Scrap and smelter runs can hide a mine shortfall for a few quarters. They cannot hide it forever if Grasberg and Kamoa stay wounded.
Could copper prices rise after the selloff
They already did, in part. The mid-September drop from the September 10 area highs was followed by a rebound toward those highs before the dollar and the Fed capped it. A further rise would need the physical tightness to beat rate fear. That is possible. It is not promised. High prices also risk demand destruction.
Is the copper surplus a mirage
Parts of it are. The 997,600-tonne exchange total is real metal. The location of that metal is distorted. The 32,000-tonne refined surplus is small next to last year. The bonded-stock adjustment erases it. TCs near zero are not a surplus signal. Call it a split market, not a cartoon glut.
What a research process looks like after a tariff scare
A scare is not a strategy. It is a calendar event.
Start with the metal, not the tweet. Read the latest ICSG table. Read the LME warehouse report. Read COMEX registered versus eligible stock. Read one China physical note on premiums and SHFE stocks. That is four pages, not forty.
Then read one producer. Pick Teck or Lundin if you want operating cash flow. Pick Hudbay if you want more torque and can stand the swing. Pick First Quantum only if you can explain Cobre Panamá in two sentences without waving your hands. Pick a junior only if you can fund a 50% drawdown without selling the family house.
Then write the bear case in your own words. The honest bear case is this. Secondary supply keeps rising. Usage stays soft. Cobre Panamá returns. Grasberg heals. ICSG’s 377,000-tonne 2027 surplus arrives. COMEX metal never leaves. Prices mean-revert toward $11,000 a tonne, the kind of floor some banks still sketch for later years. Equities that rallied 90% with the metal can give a lot of that back.
Then write the bull case in your own words. The honest bull case is this. The surplus is a U.S. warehouse story. Mines cannot grow fast. TCs stay negative. China keeps drawing stock. Data centers and grids eat more cathode. A tariff stall is a discount, not a new supply source. Record prices hold because replacement mines take a decade.
If you cannot write both cases, you do not have a view. You have a mood.
Position size is the only risk tool most people will actually use. Copper stocks are stocks first. In a broad equity shock they fall with the S&P, then they remember they are miners. Stink bids and staged entries are how some investors handle that. They are not magic. They are humility with a limit order.
Risks that can make the mirage real
A deep U.S. or China slump would cut usage faster than mines can cut output. Then the surplus stops being a mirage.
A peace-and-restart surprise at Cobre Panamá plus a clean Grasberg ramp would add hundreds of thousands of tonnes. Models that look tight in 2026 look loose in 2027 under that path.
A refined-copper tariff that finally arrives could lock even more metal inside the United States. Global fabricators would scramble. U.S. users would pay more. The political blowback could be ugly. Markets hate ugly.
Substitution is slow, then sudden. Aluminum, new cable designs, and thrifting show up after a year of $14,000 copper, not after a week.
Equity valuations can already discount $15,000 copper. If they do, a tight market can still produce a dull stock. That is why this article stays with research language. Price is not the same as a ticker.
Jurisdiction risk in Panama, the DRC, Chile, and Peru can cancel a thesis overnight. Canadian listings do not cancel that risk. They package it.
The investor’s one-theme summary
The copper market is not simply oversupplied. It is mislocated and mine-constrained.
ICSG’s raw seven-month surplus is 32,000 tonnes. Last year’s comparable surplus was about 157,000 tonnes. After Chinese bonded-stock changes, the 2026 year-to-date print is a 5,000-tonne deficit. Exchange stocks at a 23-year high are mostly a COMEX story tied to tariff front-running. SHFE is lean. LME available metal is lean. Treatment charges are at historic lows. Grasberg and Kamoa took a hole out of 2026 mine plans.
Washington taxed pipes and wire. It has not yet taxed refined cathodes. The stall on that next step dumped the price in mid-September. That dump can look like a buying opportunity if the mine story stays tight. It can look like a warning if 2027’s official surplus arrives on time.
Canadian copper stocks give investors a way to study that split. They do not give a free lunch. Study the filings. Study the sheds. Study the proclamation page. Then decide whether the surplus is metal the world can use, or metal sitting in the wrong country waiting for a tariff that did not come.
This article is not a call to buy copper, copper futures, or any mining share. It is a map of why the surplus headline and the price tape disagree. That disagreement is the story. It may not last. While it lasts, it is the only story that matters.
Disclaimer. Canadian Mining Report and the author are not advising you to purchase or sell any security. All companies named are for illustration and research context only. Copper, tariffs, and mining equities are volatile. Past prices do not predict future prices. Verify every figure against primary sources, including ICSG releases, LME and COMEX warehouse reports, company filings, and official U.S. tariff proclamations, before you act.