Policy · Critical Minerals

Why Copper Hit Record Prices While the Market Is in Surplus — the Midstream Squeeze Explained

📅 Sep 21, 2026 🏷️ Explainer · Evergreen 🕐 9 min read

By Luminesca · Updated Sep 21, 2026
Analysis compiled from public reporting with AI-assisted drafting. See our editorial policy.

Copper is the metal every electrification plan is drawn on, and in September 2026 it did something that should not be possible: it set a record price while the market it trades into was running a surplus. The London Metal Exchange touched roughly $14,875 a tonne, up about 49 per cent on a year earlier. Over the same half-year, the International Copper Study Group measured a refined copper surplus of about 131,000 tonnes, and exchange warehouses ended July holding more metal than at any time since August 2003. Both numbers are correct. The reason they coexist is that today's copper problem is not a shortage of copper. It is a chain of bottlenecks between the ore and the metal.

A price record next to a surplus

Start with the price. LME copper set an all-time high of about $14,875 a tonne in the second week of September 2026, having cleared $12,000 in December 2025 and $14,000 in May. It has since slipped a few per cent but sits roughly 49 per cent above the $9,910 it traded at on 9 September 2025 — a rally analysts call supply-led rather than demand-led, unusual for an industrial metal.

Now the market. Preliminary data from the International Copper Study Group's monthly data puts the refined copper balance in surplus of about 131,000 tonnes in the first half of 2026 — roughly 98,000 tonnes once estimated changes in Chinese bonded stocks are stripped out. Inventories across the LME, COMEX and the Shanghai Futures Exchange totalled 964,095 tonnes at the end of July, the highest since August 2003.

The reconciliation is that the surplus is measured in refined cathode, while the price is set by expectations about concentrate — the semi-processed rock that feeds smelters. Those two markets no longer tell the same story.

The squeeze is in concentrate, not in metal

World copper mine production fell 1.1 per cent in the first half of 2026. Concentrate output dropped 2.6 per cent, more than offsetting a 4.3 per cent rise in solvent extraction and electrowinning. The decline is concentrated: Chile, the largest producer, was down 6.6 per cent; Indonesian concentrate fell 32 per cent with Grasberg still constrained after the September 2025 mud rush; and the Democratic Republic of Congo's concentrate fell 34 per cent after the Kamoa seismic event, leaving national output flat once its SX-EW growth is counted. Peru rose 2 per cent as Antamina and Las Bambas expanded.

Refined output did the opposite, rising 2.4 per cent. Almost all of it came from China and the DRC — together about 59 per cent of world refined production — while output everywhere else fell 1.9 per cent.

Smelters are therefore adding metal faster than mines are adding ore, and competing for a shrinking pool of feedstock. The 2026 annual treatment and refining charge benchmark settled at zero dollars a tonne — the lowest ever agreed — and spot fees have been negative since 2024. When the processing fee is zero, the metal price can set records while the middle of the chain loses money.

The acid problem nobody priced in

The International Energy Agency's Global Critical Minerals Outlook 2026 says copper's short- and medium-term supply outlook has worsened considerably, even as its 2035 projection improves. The cause is chemical rather than geological.

More than 15 per cent of global primary copper is produced by leaching ore and recovering it through solvent extraction and electrowinning — a route that consumes sulphuric acid. Two events removed much of the world's supply of it. The effective closure of the Strait of Hormuz in February 2026 cut off roughly half of seaborne sulphur trade, the feedstock for acid; Gulf states and Iran together supply about a quarter of the world's sulphur. China then banned acid exports from May through year-end, removing close to a quarter of the acid available outside China.

Exposure is concentrated where leaching matters most: roughly 45 per cent of Congolese copper output depends on acid, about 1.5 million tonnes, alongside some 1.2 million tonnes in Chile. Acid is about 13 per cent of operating cost at an average SX-EW plant and closer to 20 per cent in the DRC, and some producers hold only 30 to 60 days of inventory.

Why policy moves the metal instead of making it

Trade policy now sets copper's price as much as geology does. The United States added copper to its Section 232 national-security tariffs in July 2025 at 50 per cent, then restructured the regime in April 2026 so duties apply to full customs value: 50 per cent on copper articles, 25 per cent on derivatives, and 15 per cent through 2027 for metal-intensive industrial and electrical grid equipment. Refined copper is the one file still open — as of mid-September no proclamation existed, though a Commerce review reportedly recommended 15 per cent from January 2027, rising to 30 per cent in 2028.

Nothing has been collected on cathode, but the expectation moved the metal anyway. Several hundred thousand tonnes have been shipped to the United States to capture the premium, lifting COMEX stocks to a record and draining LME warehouses. That is how global inventories can sit at a 23-year high while buyers outside the US complain of scarcity.

Producer policy cuts both ways. Zambia imposes a 10 per cent duty on concentrate exports precisely to force domestic smelting, and has suspended it — most recently until 30 September 2026 for 271,742 tonnes — because its own smelters were offline for extended repairs. A country that legislated beneficiation shipped raw rock because a furnace was down. The DRC's cobalt export quota lifted cobalt prices about 130 per cent.

The long game: grades, lead times and permits

The IEA did narrow copper's projected 2035 supply gap, from about 30 per cent to roughly 25 per cent. Look at where the improvement comes from and it is less reassuring: expansions and life extensions at mines that already exist — Kisanfu, Lumwana, Highland Valley, Antamina — rather than new deposits. Ore grades have fallen about 40 per cent since 1991, and the agency's numbers point to roughly $310 billion of copper investment inside a mining and refining bill above $750 billion to 2040.

Timelines are the harder constraint, with discovery to production now measured in decades rather than years. Resolution Copper in Arizona shows the shape of it. A land exchange mandated by Congress in 2014 completed only on 16 March 2026, after a Ninth Circuit ruling three days earlier. The project, one of the largest undeveloped copper deposits in the world, has committed about $500 million to enabling work and could eventually supply up to a quarter of US demand. There is still no decision to mine, and the San Carlos Apache Tribe continues to oppose it over Oak Flat.

What it means

Copper sits in every part of an electrified system: transformers, cables, busbars, motors, inverters. The IEA's ratios are blunt: renewable generation uses roughly 8 to 12 times more copper than the fossil plants it replaces, a wind turbine takes 2.5 to 6 tonnes per megawatt and solar about 4, while a battery-electric car carries about 80 kilograms against roughly 20 for a combustion model. Record copper prices therefore raise the input cost of the build-out meant to reduce energy insecurity, from transmission and grid expansion to data-centre power systems.

Three responses are visible. Aluminium substitution is accelerating in high-voltage transmission and being trialled in vehicle wiring, at a cost in volume and efficiency. Recycling is the fastest lever, with secondary refined production up 4.3 per cent in the first half, but scrap supply cannot be conjured. And governments increasingly treat the midstream as strategic: the EU's Critical Raw Materials Act, export duties and stockpiling all target refining capacity rather than mines.

The honest reading is that copper's price is doing its job, and the investment response is slow because the binding constraints are furnaces, acid plants, permits and time rather than price. Volatility is now structural.

Frequently asked questions

If the copper market is in surplus, why are prices at record highs?

Because the two numbers are measured in different markets. The surplus is in refined cathode; the price is being driven by the shortage of concentrate that feeds smelters, by tariff expectations that have relocated physical metal to the United States, and by the sulphuric acid squeeze on leach-based supply. Copper can be simultaneously abundant as finished metal and scarce as feedstock.

What is a treatment charge, and why has it fallen to zero?

A treatment and refining charge is the fee a smelter charges a miner to convert concentrate into refined metal. It is set annually in benchmark negotiations and floats in the spot market between them. The 2026 benchmark settled at $0 per tonne, the lowest ever recorded, and spot fees have been negative since 2024. The cause is capacity: China has accounted for more than 90 per cent of global smelter capacity growth since 2005 and now represents roughly half of world smelting, so smelters are bidding for far less concentrate than they can process. Zero fees mean the midstream absorbs the squeeze while metal prices rise.

Why does sulphuric acid matter to copper supply?

Because more than 15 per cent of the world's primary copper is produced by acid leaching rather than smelting. The effective closure of the Strait of Hormuz in February 2026 disrupted around half of global seaborne sulphur trade, and China's export ban from May removed close to a quarter of the acid available outside China. With roughly 1.5 million tonnes of Congolese output and 1.2 million tonnes of Chilean output dependent on leaching, and some producers holding only 30 to 60 days of stock, the IEA warns that a prolonged shortage would force production cuts into an already tight market.

Will high copper prices slow the energy transition?

They raise costs, but they are unlikely to stop the build-out. Copper is a modest share of most project budgets, and technical cost declines in solar, wind and batteries have historically outpaced input inflation. Analysts at S&P Global argue the practical impact will be smaller than headline prices suggest, while noting that substitution toward aluminium is likely in specific applications. The bigger risk is timing: higher and more volatile input costs make marginal projects harder to finance.

Can recycling close the gap?

Only partly. Secondary refined production, meaning copper recovered from scrap, grew 4.3 per cent in the first half of 2026, faster than primary output, and it responds to price more quickly than any mine can. But scrap availability is tied to past consumption, and the world's existing copper stock sits largely in buildings, grids and vehicles with service lives measured in decades. The IEA's projections assume recycling meets part of the demand increase, not all of it.