Here's the thing nobody wants to admit: that 1.5 million tonne deficit everyone keeps throwing around for 2026? It's wrong. And the real numbers are messy enough without the hyperbole.
The actual 2026 copper deficit projections range from 150,000 to 600,000 tonnes, depending on who you ask. J.P. Morgan Global Research pegs it at roughly 330,000 tonnes. The International Copper Study Group goes conservative at 150,000 tonnes. ING takes the bearish side with 600,000 tonnes. Goldman Sachs flipped the script entirely and now projects a 300,000 tonne surplus, citing dampened Chinese demand and reduced U.S. strategic stockpiling.
That's a massive spread. But here's what every single one of these forecasters agrees on: new mines cannot close the gap in 2026.
Not even close.
The Timeline Problem Nobody Wants to Face
Copper mines don't appear overnight because a hedge fund decided the market looks attractive.
From discovery through permitting, construction, and ramp-up to full production, major copper projects require seven to ten years. That's not industry whining. That's geology, engineering, regulatory reality, and basic physics. You can't disrupt your way around bedrock.
Which means any project getting approved today: and we're talking full regulatory green lights, financing locked, equipment ordered: won't contribute meaningful tonnage until 2031 at the earliest. The 2026 supply equation was essentially locked in around 2016-2019.
The math gets uglier when you look at what's actually in the pipeline. New project approvals have averaged under 300,000 tonnes annually for the past three years. That's roughly half of what the industry needs just to maintain current supply levels, let alone grow into electrification demand.
Meanwhile, the industry shed 500,000 to 800,000 tonnes of capacity through disruptions in 2024 alone.
What's Actually Happening on the Ground
Let's talk about the projects that are supposed to help in 2026.
Freeport-McMoRan's Bagdad mine in Arizona: The site of a fatal accident that shut down operations. The company expects a phased restart in Q2 2026, but they're projecting 85% capacity initially. Not exactly closing gaps.
Ivanhoe's Kamoa-Kakula complex in the Democratic Republic of Congo: This was supposed to be the savior project, targeting 500,000 tonnes per year. They've pushed that milestone to 2027. Permitting delays, infrastructure challenges, and the usual complications of operating in the DRC.
Expansion projects in Chile: Codelco's aging infrastructure keeps requiring more capital and delivering less copper. Their production guidance keeps getting revised downward, not up.
These aren't exotic edge cases. These are the flagship projects that were supposed to ease supply constraints. And they're all running behind schedule or below capacity.
The Structural Crisis Beneath the Headlines
The real problem isn't the 2026 deficit: whether it's 150,000 tonnes or 600,000 tonnes. The real problem is that the entire copper supply chain is structurally broken for the next decade.
S&P Global estimates the industry needs to bring 80 new sizable copper mines online by 2040 to meet projected demand from electrification, data centers, grid infrastructure, and EV adoption. Not 80 exploration projects. Not 80 feasibility studies. Eighty actual producing mines.
At current approval and development rates, the industry might deliver 30.
The only supply source that can scale quickly enough to provide near-term relief is recycled copper scrap. And while that market is growing, it's not growing fast enough to offset the production shortfalls from primary mining.
This is where the AI boom becomes particularly ironic. The same technology sector that's driving unprecedented copper demand: data centers pulling 475 kilotons in 2026 alone, up 110 kilotons from 2025: is also the sector that supposedly thrives on disruption and rapid scaling.
But you can't disrupt geology. You can't machine-learn your way past a permitting process. And you can't download a copper mine.
The Demand Side Isn't Helping Either
Here's what makes this particularly uncomfortable: demand elasticity at elevated copper prices is proving far weaker than forecasters expected.
Industries that can theoretically defer copper-intensive purchases during price spikes: construction, consumer electronics, some manufacturing: are doing exactly that. But the sectors driving 2026 demand growth cannot defer.
Data centers need that copper now. Grid operators upgrading for renewable integration need it now. EV manufacturers competing for market share need it now. Defense contractors building next-generation systems need it now.
These aren't discretionary purchases that can wait for better pricing. They're strategic imperatives with their own timelines and capital commitments already made.
What the Market is Actually Pricing In
Goldman Sachs revised their 2026 outlook to a surplus based largely on two assumptions: Chinese demand cooling and U.S. strategic stockpiling scaling back.
Both assumptions are politically and economically shaky.
Chinese demand might soften in residential construction and some manufacturing, but Beijing is simultaneously pushing massive grid infrastructure upgrades and accelerating its own EV transition. Those two clocks don't sync cleanly.
U.S. strategic stockpiling under the White House's critical minerals initiative is driven by supply chain security concerns that haven't disappeared just because copper prices spiked. If anything, higher prices underscore the strategic vulnerability the stockpiling is meant to address.
The J.P. Morgan and ING projections: both showing meaningful deficits: assume that existing production faces continued disruptions and that Chinese demand holds relatively steady. That's probably closer to reality than the surplus scenario.
But the range of disagreement tells you everything about market uncertainty. When serious forecasters are 900,000 tonnes apart on a single year's balance, nobody really knows what's coming.
The Only Solutions That Actually Work in 2026
Since new mines can't help and existing operations are maxed out or struggling, what actually closes the gap?
Price rationing: Copper prices stay elevated until marginal demand gets priced out. Some construction projects get delayed. Some consumer electronics designs get adjusted. Some industrial applications substitute aluminum or other materials where technically feasible.
Scrap mobilization: Higher prices pull more copper scrap into circulation. Recyclers expand capacity. Demolition and industrial teardown accelerates. It's not glamorous, but it's the only supply lever that can move in a 12-month timeframe.
Demand destruction: The economy slows more than expected, killing projects before they consume copper. This is effectively what Goldman Sachs is betting on with their surplus call.
Inventory drawdown: Existing stockpiles at exchanges, manufacturers, and traders get depleted. This doesn't solve anything structurally, but it can smooth a single year's imbalance.
None of these solutions are what industry executives present at investor conferences. But they're what actually happens when supply can't meet demand in the near term.
The 2027-2030 Window Gets Even Tighter
If you think 2026 looks challenging, the out-years are genuinely grim.
Data center copper demand is projected to hit 550-600 kilotons by 2028. EV adoption continues accelerating. Grid infrastructure spending: particularly in the U.S. and Europe following renewable energy mandates: keeps ramping.
And the mine development timeline problem doesn't magically fix itself. Every project that should have been approved in 2020-2022 to hit production in 2027-2029 is a hole that can't be filled.
The industry is effectively running a three-to-five-year deficit in project approvals that compounds every year.
What This Actually Means
The 1.5 million tonne deficit for 2026 might be exaggerated, but the structural shortage is real and it's not getting better anytime soon.
New mines won't save 2026. They can't. The ones that might have helped are delayed, under capacity, or still in permitting. The ones that could help 2028-2030 haven't been approved yet.
Price will do most of the work in balancing the market, which means sustained elevated copper prices through the decade. Recycling will pick up some slack. Some demand will get deferred or substituted away.
But if you're waiting for a wave of new supply to flood the market and crash prices, you're going to be waiting a long time.
The chickens are coming home to roost on a decade of underinvestment in copper mining. And there's no quick fix when it takes ten years to build the fix.
That's not a forecast. That's just math.
