Copper in 2026: the deficit is real — the question is who bears the cost

The balance debate: everyone is right, no one is certain

The most contested number in commodities this year is the 2026 copper market balance, and the range of estimates has rarely been wider. The divergence is not a sign that analysts are incompetent — it is a reflection of genuine structural uncertainty in a market that is simultaneously experiencing acute mine supply disruption, a demand composition shift, and the largest tariff-driven stockpiling event in history.

(FACT: ICSG Monthly Bulletin April 2026; ING Think Copper Forecast February 2026; Goldman Sachs Commodities Research; J.P. Morgan Metals Research)

The ICSG's April 2026 revision is the most striking. The industry body that had forecast a 150,000-tonne deficit in October 2025 — the first structural shortage since 2009 — flipped to a 96,000-tonne surplus. The explanation: weaker-than-expected refined copper usage growth (especially in China's property and grid sectors) and higher mobilization of scrap. (FACT: ICSG) This is the kind of revision that changes procurement strategy. If the consensus anchor itself shifts by a quarter of a million tonnes, every budget built on the deficit narrative needs re-examination.

But the ICSG surplus is contested. ING maintains a deep-deficit view of roughly 600,000 tonnes, anchored to the physical reality of concentrate supply losses. Goldman Sachs models a ~300,000-tonne surplus, betting that demand destruction and scrap substitution outpace mine losses. J.P. Morgan is in the deficit camp, noting that data center construction alone will consume roughly 475,000 tonnes of copper in 2026 — equivalent to nearly three Cobre Panamas. (FACT: J.P. Morgan)

The honest answer is that the real balance lies somewhere in the middle, and the margin of error is wider than any single estimate. But one thing is unambiguous: the market no longer has a meaningful supply buffer. If any single one of the major mine disruptions this year resolved positively, the balance could shift by 200,000-300,000 tonnes. That is precisely why lME copper is at $14,000/t — the market is pricing in the tail risk that the supply-side bets all go the wrong way.

TC/RCs at zero: the smelter crisis has arrived

If you want a single number that captures the state of the copper concentrate market, it is this: the 2026 annual TC/RC benchmark between Antofagasta and Chinese smelters settled at $0 per tonne, down from $21.25/t in 2025. (FACT: Benchmark Minerals, December 2025) Spot market TCs have fallen even further, to as low as -$90 per tonne — meaning smelters are literally paying miners to accept their concentrate. (FACT: IEA, March 2026; multiple market sources)

This is unprecedented. Treatment and refining charges are the fee miners pay smelters to process concentrate into refined copper. When TCs are high, concentrate is abundant and smelters have pricing power. When TCs are low or negative, concentrate is scarce and miners hold the leverage. At -$90/t spot, the relationship has inverted entirely: smelters are subsidizing miners to take product they cannot afford to refuse because shutting a smelter costs even more.

The root cause is a structural mismatch between mine supply and smelting capacity. Mine disruptions at Grasberg (force majeure from a mud-rush incident), Kamoa-Kakula (seismic event and mine redesign), and Cobre Panama (indefinite closure) have collectively removed roughly 1 million tonnes of planned concentrate supply from the market. Meanwhile, Chinese smelting capacity has continued to grow, creating a demand for concentrate that physically cannot be satisfied. (FACT: Wood Mackenzie; IEA)

The China Smelters Purchase Team (CSPT) has now refused to set quarterly TC guidance for four consecutive quarters — an unprecedented breakdown of the pricing mechanism that has governed the concentrate market for two decades. (FACT: Hellenic Shipping News, December 2025) Chinese smelters have agreed to cut 2026 output by more than 10%, and the government has halted roughly 2 million tonnes of planned new smelting capacity. (FACT: IEA) But these measures are cosmetic relative to the scale of the imbalance. Unless mine supply recovers meaningfully, the TC/RC crisis is structural, not cyclical.

The mine supply shock: Grasberg, Kamoa-Kakula, Cobre Panama

Three mine stories define the 2026 copper supply narrative, and none of them are bullish for availability.

Grasberg: the world's second-largest mine at 60%

Freeport-McMoRan's Grasberg complex in Indonesia — initially forecast to produce 1.7 billion pounds of copper in 2026 — is now expected to produce roughly 0.7 billion pounds, a ~60% production gap. (FACT: Freeport Q1 2026 earnings presentation; Fastmarkets; Investing.com) The cause was a September 2025 mud-rush incident at the Grasberg Block Cave that triggered force majeure and halted the operation. A phased restart began in Q2 2026, but Freeport has stated that infrastructure modifications will limit output to approximately 60% of capacity through mid-2027. (FACT: Freeport Q1 2026 slides)

The market had assumed Grasberg would be back at full rates by H2 2026. That assumption is now broken. In copper supply modeling, ~460,000 tonnes of expected production has been deferred by 18-24 months — a gap that no other mine can fill in the near term.

Kamoa-Kakula: 20% deferred

Ivanhoe Mines slashed 2026 production guidance for its flagship Kamoa-Kakula complex in the DRC from 380,000-420,000 tonnes to 290,000-330,000 tonnes — a roughly 20% reduction. (FACT: Ivanhoe Mines press release, March 31, 2026; Mining.com, April 2, 2026) The cause was a May 2025 seismic event that forced a mine plan redesign, lowering mineral reserves to ~60% of resources and requiring exclusion zones that deferred stoping at Kamoa to H2 2026 and at Kakula to H1 2027.

The key detail: this is described by management as a "one-year deferral" of the prior ramp-up trajectory, with 2027 guidance also cut to 380,000-420,000 tonnes from a prior 500,000-540,000 tonnes. (FACT: Yahoo Finance, April 2026) Full 500,000+ tonne run-rate is now pushed to 2028. For a market desperate for new supply, a one-year delay on one of the world's most important growth mines is material.

Cobre Panama: the silent wildcard

First Quantum's Cobre Panama mine remains closed with no restart timeline. Before its shutdown in late 2023, Cobre Panama produced roughly 350,000 tonnes of copper annually — about 1.5% of global mine supply. The political and legal situation in Panama remains unresolved, with the government's cancellation of the contract upheld and no viable path to restart visible. (FACT: Reuters; Columbia CGEP Copper Report)

If Cobre Panama were to restart tomorrow, it would add significant supply and shift the balance toward surplus. But the probability of restart within 2026 appears low. The mine's continued closure is a structural subtraction from global supply that the market has now largely priced in — making any restart announcement a potential bearish catalyst.

The COMEX-LME dislocation: tariff trade gone wild

One of the most anomalous conditions in the 2026 copper market is the divergence between COMEX and LME inventories. COMEX warehouses hold a record 548,000 tonnes of copper — metal that traders and consumers rushed into the United States to front-run potential import tariffs. (FACT: COMEX/CME warehouse reports, May 2026; Stockcrock Substack, January 2026) Meanwhile, LME on-warrant stocks outside China have fallen below 100,000 tonnes, the lowest in years. (FACT: LME Warehouse Reports, May 2026)

The mechanism is straightforward: with COMEX copper prices trading at a premium of roughly $500/mt over LME, traders can buy copper globally, ship it to the US, deliver it into COMEX, and lock in a risk-free arbitrage profit. Trafigura's May 2026 decision to withdraw 51,000+ tonnes from LME warehouses in New Orleans ahead of a US tariff ruling in late June shows the trade is still active. (FACT: Reuters, May 22, 2026; Sora Futures, May 24, 2026)

The question every copper buyer needs to answer: is the COMEX stockpile real supply or trapped supply? The answer depends entirely on US tariff policy. If import tariffs are imposed in June, that 548,000 tonnes becomes an increasingly expensive stranded asset inside the US — physically available for US buyers but disconnected from the global market. If tariffs are waived or delayed, the metal flows back out of COMEX, LME stocks surge, and the dislocation corrects sharply.

The critical insight: the COMEX stockpile is a financial buffer, not a physical one for most of the world. A buyer in Europe or Asia cannot access COMEX inventory without paying tariff-related economics. The 548,000 tonnes that looks like "record stockpiles" on a headline basis is largely irrelevant to non-US physical procurement. The real availability signal is LME on-warrant stocks outside China, and that signal is flashing red.

What $14,000 copper means for buyers

The transition from a $9,000/t copper world (2023-2024 average) to a $14,000/t world (May 2026) represents an enormous cost shift along the value chain. The question is no longer whether copper is expensive — it is who in the supply chain bears the burden.

At $14,000 per tonne, a buyer consuming 200 tonnes per month pays an additional $900,000 per month compared to a $9,000/t baseline — that is $10.8 million annualized for a single mid-sized fabricator. For a cable manufacturer using 500 tonnes/month, the annual impact exceeds $27 million. These are not theoretical numbers. They are showing up in earnings calls across the copper supply chain.

Chinese buyers are already adjusting. The Yangshan copper premium — the amount Chinese importers pay above LME cash to secure physical metal delivered into China — has recovered from a low of $20/t to $65/t. (FACT: SMM, May 2026) This signals that the China import arbitrage window has reopened, drawing metal toward Asia. It is the single best real-time indicator that physical availability outside the COMEC stockpile bubble is tightening.

The question of who bears the cost is cascading through the supply chain:

• Miners are capturing the most margin. At $14,000/t, even high-cost producers are deeply profitable. The C1 cash cost curve for the top decile of mines is roughly $2,500-3,500/t — meaning margins of $10,000-11,500/t.
• Smelters are squeezed. Zero or negative TC/RCs mean their processing revenue has collapsed. They survive on by-product credits and cathode premiums, but the business model is under severe pressure.
• Fabricators and end-users are the shock absorbers. Wire and cable makers, tube producers, and component manufacturers cannot pass through full price increases immediately. Their margins are being compressed, and some smaller fabricators with limited hedging programs face existential risk.

When will new mine supply arrive?

The honest answer is: not soon enough to matter for 2026-2027 procurement planning.

The copper industry's structural problem is that it underinvested in new mine development during the 2015-2020 price trough, and the projects that were sanctioned take 7-15 years to build. The pipeline of new supply through 2028 is remarkably thin:

• Grasberg recovery: Phased restart underway, but capped at ~60% through mid-2027. Full rates by 2028 at earliest. (FACT: Freeport Q1 2026)
• Kamoa-Kakula ramp-up: 500,000+ tonne run-rate pushed to 2028. (FACT: Ivanhoe Mines, March 2026)
• El Abra expansion (Chile): Targeted at 700+ million lbs/year incremental. Start-up in 2033. (FACT: Freeport Q1 2026)
• Safford/Lone Star (Arizona): 300-400 million lbs/year. Beginning in 2030s. (FACT: Freeport)
• Marimaca (Chile): FID targeted for H2 2026, construction in late 2026. Still years from production. (FACT: Crux Investor)

The only near-term relief valve is scrap and secondary production. The ICSG surplus revision was partly driven by higher scrap mobilization, and that trend can continue. But scrap is a supplementary source, not a replacement for primary mine supply. The copper market cannot scrap its way out of a 1-million-tonne concentrate deficit.

Chinese smelter output cuts of 10%+ could rebalance the concentrate market by reducing demand for raw material, potentially lifting TC/RCs from negative territory. But that requires Chinese policy to prioritize smelter margins over production volumes — a political calculation that is far from guaranteed.

What this means for buyers

High-volume buyer with LME-linked annual contracts: Budget for $12,500-15,000/t LME copper for H2 2026. Add a +15%/-10% price band on the copper component of all contracts. Include tariff contingency language: if US tariffs on copper are imposed, trigger a reopening clause. Negotiate Yangshan premium pass-through for China-delivered material. Monitor COMEX-LME spread weekly — a narrowing toward $200/t signals the dislocation is unwinding.

Mid-volume buyer with quarterly or spot exposure: Secure 50% of H2 volume via 3-month LME forwards at current levels ($13,500-14,500/t). Buy put options with a floor at $12,000/t to protect against a tariff-waiver crash. Keep 20% floating for opportunistic buying if the COMEX-LME dislocation corrects violently. If the ICSG surplus narrative gains traction and LME drops below $12,500, that is your buying opportunity — the structural deficit thesis hasn't changed, only the timing.

Multi-region buyer with US, European, and Asian exposure: The single biggest risk is that tariffs create a three-tier market: US (expensive, COMEC-stockpiled), Europe (tight, LME-dependent), China (Yangshan premium-driven, scrap-supplemented). Each market needs a separate pricing and sourcing strategy. Lock US supply now while COMEX stockpiles are plentiful — once tariffs are decided, that metal may be trapped or repriced. For Europe, build inventory above normal working capital levels — LME on-warrant stocks outside China below 100kt are a real liquidity risk. For China, accept that Yangshan premiums will remain elevated and include premium pass-through in your procurement contracts.