The 2026 annual copper smelter treatment and refining charge benchmark settled at exactly $0 per tonne, the lowest ever agreed in annual negotiations, as a structural concentrate squeeze forces smelters to process for free just to keep furnaces lit. Copper itself hit an all-time high of $14,875 per metric ton on the LME on September 10, up roughly 48% year-on-year, making it the standout performer among base metals. The zero-fee settlement is not a negotiating anomaly; it is a precise signal of where value has migrated in the copper supply chain, and it is reshaping the investment case for miners versus smelters in real time.
Introduction
The number that defines the copper market in 2026 is not $14,875, the LME all-time high printed on September 10. It is zero. Specifically, $0 per dry metric tonne, the treatment charge that Antofagasta agreed with Chinese smelter counterparties for the 2026 annual benchmark, a figure the International Energy Agency confirmed as the lowest annual TC/RC settlement ever recorded in the history of the benchmark system.
To put that in context: the 2025 benchmark was $21.25 per tonne and 2.125 cents per pound. The 2024 and 2023 benchmarks were both $80 per tonne. The 2022 benchmark was $65 per tonne. The collapse from $80 to $0 in three years is not a cyclical fluctuation; it is a structural breakdown in the economics of copper conversion, and it is happening simultaneously with copper prices trading at all-time nominal highs.
For investors, the zero benchmark is the single most actionable data point in the copper complex right now. It tells you exactly where the pricing power sits (with miners, not smelters), which corporate structures are winning (integrated producers with captive mine supply), and how severe the physical concentrate squeeze has become across the global supply chain. The price rally gets the headlines. The TC/RC collapse is the mechanism explaining why the rally has legs.
Price Action: LME Records, Backwardation, and the 48% YoY Move
LME three-month copper hit $14,875 per metric ton on September 10, 2026, an all-time nominal high, before selling off roughly 3% in a single session on tariff hesitation reports. The metal ended Q3 near $14,416 to $14,447 per tonne, booking a 7.9% quarterly gain. On Comex, the record was $6.83 per pound set on September 22, with the LME following the next session at $14,833 per tonne. The year-on-year gain as of late September sits at approximately 48%, following a 42% rally in 2025 that ING analysts Ewa Manthey and Warren Patterson described as copper's best annual performance since 2009.
The structure of the LME curve is reinforcing the physical story. Cash copper settled at $14,424.50 per tonne in mid-August at a $207.50 premium over the three-month contract, the widest backwardation of 2026. Backwardation of that magnitude signals that nearby physical demand is overwhelming available supply, not a paper trade. LME warehouse stocks fell more than 35,000 tonnes, or 14%, from end-July levels to 214,550 tonnes by mid-August, while Shanghai Futures Exchange stocks remained below 70,000 tonnes. The market is drawing down inventory across both major exchange networks simultaneously.
The January 29 session deserves a separate mention. Three-month LME copper hit $14,527.50 per tonne that day, breaking the previous nominal peak from March 2022, in the largest single-day price move since 2008, nearly touching the 12% daily limit set by the exchange. That kind of price action, driven by physical scarcity rather than speculative excess, is what differentiates this rally from the 2021 to 2022 cycle. The zero TC/RC benchmark settled just days earlier in January was the first signal that this move had structural underpinning.
Supply and Demand Dynamics: The Concentrate Squeeze in Granular Detail
Global copper mine production fell 1.1% in the first half of 2026, while concentrate production specifically declined 2.6%, according to preliminary International Copper Study Group data. That divergence matters: the concentrate market is tightening faster than headline mine output suggests, because ore grades are falling and processing recoveries are under pressure at multiple major operations simultaneously.
Grasberg is the largest single source of disruption. Freeport-McMoRan declared force majeure at the world's second-largest copper mine following a September 2025 mudflow incident, and the company has since pushed the full restart of the Grasberg Block Cave into early 2028. Benchmark Mineral Intelligence estimates the disruption will remove 591,000 tonnes of copper output between September 2025 and year-end 2026. As of April 2026, undamaged areas were operating at only 40 to 50% of capacity due to groundwater issues and damage to ore-handling systems. Freeport cut its 2026 production guidance by 35%, with Grasberg capacity running at roughly 65% through H2.
Chile compounded the picture. National copper output fell 9.04% year-on-year in March 2026 to 434,314 tonnes per Cochilco data. Codelco dropped 10%, BHP's Escondida fell 15.75%, and the Glencore-Anglo American Collahuasi joint venture declined 10.80%. The Democratic Republic of Congo added another layer: Ivanhoe Mines' Kamoa-Kakula operation, which was expected to offset some of the Grasberg shortfall, is being constrained by sulfuric acid shortages critical to processing operations. Separately, the DRC government enacted an immediate ban on copper and cobalt concentrate exports on June 29, 2026, although the impact on global concentrate supply is somewhat limited given that the country ships predominantly refined cathode (696,725 tonnes of cathode versus 53,926 tonnes of concentrate in Q1 2026 alone).
On the demand side, the energy transition and AI infrastructure buildout are consuming copper at an accelerating rate. Goldman Sachs estimates that copper demand from the energy transition will grow nearly 600% to 5.4 million tonnes by 2030 in its base case, and as high as 8.7 million tonnes under a hyper-adoption scenario. Data centers alone are projected to consume 475,000 metric tonnes of copper in 2026, up 110,000 tonnes year-on-year per J.P. Morgan Global Research, with each megawatt of data center capacity requiring 20 to 30 tonnes of copper. The ICSG has revised its 2026 refined copper balance from a 209,000-tonne surplus in October 2025 to a 150,000-tonne deficit in May 2026. ING's model shows a 600,000-tonne deficit. J.P. Morgan sits at 330,000 tonnes short. The range of estimates is wide, but the direction is not in dispute.
Institutional Activity: Chinese Smelters, CSPT Cuts, and the Byproduct Lifeline
The structural backdrop to the zero benchmark is a decade of Chinese smelter expansion that dramatically outpaced mine supply growth. Since 2005, China has accounted for over 90% of the growth in global copper smelting output, lifting its share from roughly 15% to approximately half of global supply by 2025. The 1.62 million-tonne gap between installed smelting capacity (16.2 million tonnes) and actual output (13.64 million tonnes) is precisely the overcapacity that drove TC/RCs to collapse between 2024 and 2026. Too many furnaces chasing too little concentrate is the arithmetic of a zero fee.
The China Smelter Purchasing Team, the coordinating body for China's major copper smelters, announced cuts of more than 10% to 2026 output and the suspension of roughly 2 million tonnes of planned new capacity. The China Nonferrous Metals Industry Association publicly stated that it "firmly opposed" zero and negative processing charges. In practice, however, official data from China's National Bureau of Statistics showed refined copper output rising 7.4% year-on-year between January and April 2026. Jiangxi Copper and Yunnan Copper both raised their 2026 production targets after the cut was announced. The gap between pledged discipline and actual output is stark, and the CSPT has withheld quarterly TC/RC guidance for six consecutive quarters.
The smelters are surviving on byproduct revenues, not processing margins. Sulfuric acid prices surged 73.6% between January and April 2026, rising from 890 RMB per tonne at the start of the year to 1,660.5 RMB per tonne in April. Given that copper smelting co-produces 3 to 4.5 tonnes of sulfuric acid per tonne of electrolytic copper, acid revenues are now covering the concentrate procurement cost and part of the processing cost for better-positioned operations. Gold and silver byproduct credits are providing a similar offset for smelters processing precious-metal-rich concentrates. A North American trader quoted at Fastmarkets CESCO 2026 Week put it plainly: "The environment we're in right now is one where by-products and free metal help smelters absorb the shock. Smelters are fragile right now." One Chinese smelter operator was more direct still: "As long as sulfuric acid prices are high, smelters maintain their intake and production levels. If they keep capacity full, they can produce and sell cathodes with premiums."
The margin data tells the story in RMB terms. Top-tier integrated smelters such as Jiangxi Copper, Tongling Nonferrous, and Zijin, which have captive mine supply, maintained per-tonne blended cathode margins of 1,500 to 3,000 yuan in H1 2026. Mid-tier smelters relying on spot concentrate purchases saw margins compress to negative 500 to positive 500 yuan per tonne, surviving almost entirely on byproduct income. The divergence between integrated and custom smelters is the defining equity theme in Chinese non-ferrous metals this year.
The Trade: Where Value Has Migrated in the Copper Supply Chain
The investment thesis here is straightforward, and the TC/RC data makes it quantifiable rather than directional. At $21.25 per tonne (the 2025 benchmark), a miner delivering concentrate paid that fee to have its ore converted into refined cathode. At zero, the miner retains that $21.25 per tonne. For Antofagasta, which has since reached agreement with all its Chinese clients at zero, that represents roughly $85 million annually based on its concentrate production capacity. The value has not disappeared; it has transferred from the smelter's P&L to the miner's.
The spot TC/RC market has gone significantly further than the benchmark. Argus assessed spot charges at negative $126.80 per tonne in late June 2026, compared to negative $119.30 the prior week and negative $44.80 one year earlier. Platts assessed spot at negative $78.50 per metric ton and refining charges at negative 7.85 cents per pound on April 9. An extreme transaction was reported at negative $220 per tonne, a figure that would have been dismissed as a data error five years ago. At negative TC/RC, the economics are inverted: the smelter is paying the miner for the right to process the material. Most modern custom smelters require a TC/RC of at least $50 to $60 per dry metric tonne to break even on energy and labor costs alone. Spot conditions are more than $180 per tonne below that floor.
For equity investors, the screen is simple: long miners with large, consistent, high-quality concentrate streams; short or underweight custom smelters with minimal byproduct exposure and high spot concentrate dependence. Companies with integrated smelting capacity tied to captive mines are insulated from spot TC/RC volatility in a way that pure custom operators are not. The non-Chinese smelter closure risk is real and rising; Mitsubishi Materials has already announced plans to reduce primary copper smelting output by up to 40% over the next decade, with the Onahama facility targeted for partial suspension by early 2027.
Antofagasta's proposal to shift concentrate sales to spot-indexed contract pricing for H2 2026 through H1 2027, replacing fixed annual benchmarks with floating averages derived from published TC/RC indices, is the most significant structural shift in copper concentrate contracting since the benchmark system was established. If major pairs adopt that structure, Fastmarkets and Platts spot assessments become the de facto pricing basis for a substantial share of global seaborne concentrate trade. Japanese smelters, through JMIA chairman Tetsuya Tanaka, are already trying to negotiate terms individually rather than tracking the benchmark, acknowledging that the traditional annual settlement framework has lost its anchor function. The 2027 benchmark negotiation, scheduled to begin in Q4 2026, is effectively a referendum on whether the benchmark system itself survives in its current form.
Key Levels to Watch
The immediate technical level on LME copper is $14,416 per tonne, where the metal ended Q3. The all-time high at $14,875 is the ceiling that needs to break to confirm the next leg of the structural bull market. A close above $15,000 per tonne, which Goldman Sachs cited as its long-term target for 2035 but which the current supply-demand configuration may force earlier, would likely trigger a fresh wave of institutional positioning and mining equity upgrades.
On the TC/RC side, the figure to track is the Argus spot assessment, which stood at negative $126.80 per tonne at end-June and has been the most reliable leading indicator of concentrate tightness. A move further toward negative $150 per tonne in the spot market before year-end would signal that the 2027 benchmark negotiation is headed toward a sub-zero fixed print or a full collapse into index-linked pricing. Either outcome accelerates non-Chinese smelter closures and further concentrates processing power in integrated producers with captive mine supply.
The ICSG deficit forecast of 150,000 tonnes for 2026, conservative relative to ING's 600,000-tonne estimate and J.P. Morgan's 330,000 tonnes, is the number to revisit after Q3 mine production data lands in November. If Grasberg production remains at 40 to 50% of capacity through year-end rather than recovering toward the 65% H2 guidance, and if Kamoa-Kakula's acid supply constraint persists, the ICSG number moves materially toward the higher-end estimates. Every 100,000-tonne revision to the deficit in that direction is additional fuel for the concentrate squeeze and additional downward pressure on TC/RCs heading into 2027 negotiations.
The structural read is this: mine supply is the binding constraint, smelting capacity is the excess variable, and the zero benchmark is the price signal confirming exactly that. Goldman Sachs describes copper as "a major beneficiary of investments in grid and power infrastructure globally, as AI and defence heighten the need for robust and secure energy networks," and their $15,000 to $25,000 range for 2027 to 2030 captures the medium-term scenario if the mine supply gap persists. Robert Friedland's warning that "prices will have to go much higher to ration demand" looks less like hyperbole and more like arithmetic at this point in the cycle.