Why Copper’s Supply Crisis Cannot Be Fixed by Higher Prices

Copper is trading near US$14,500 per tonne in September 2026, close to an all-time high. The instinct is to call it a commodity rally. It is not.

What the price is signalling is a structural break, one that took roughly three decades to build.

Most commodity cycles follow a familiar arc: demand surges, high prices pull new supply into the market, and the balance restores itself. This one behaves differently, because supply simply cannot respond at the pace the market demands. Grade decline, lengthening lead times, and doubling capital costs have effectively capped how fast new metal can arrive.

Meanwhile, demand is not one story but several running at once: electrification, electric vehicles, grid expansion, and now artificial intelligence data centres. A structurally constrained supply curve meeting a permanently elevated demand curve is what sets this copper thesis apart from any prior cycle.

This analysis gives U.S. investors a framework for understanding why the copper supply crisis cannot resolve through normal market mechanisms, why refining concentration adds a second layer of strategic urgency, and why Colombia’s 2026 political reset has moved that jurisdiction from overlooked to strategically relevant for large near-surface copper projects in the Americas.

Why this copper cycle is structurally different from any that came before

Every previous copper boom rested on a comforting assumption: that price would eventually fix scarcity. Pay enough, and the metal appears. That assumption is now breaking down, and it is breaking down slowly, through the accumulation of forces that no single price signal can reverse.

The first is grade. According to industry data, average copper ore grades have fallen roughly 40% since 1990, from around 1.6% to about 0.9%. Miners are processing more rock to extract less metal, which raises the cost and energy of every tonne produced.

The second is time. The average project lead time has stretched from 12.3 years to 16.3 years, as new deposits sit deeper and permitting grows more complex.

Industry research on copper grade decline and lead-time data cited by S&P Global finds that average global copper grades have fallen roughly 40% since 1991 and discovery-to-production timelines now average 17.5 years, figures that sit above even the conservative estimates used in most deficit modelling.

The third is capital. Capex per tonne of new capacity has doubled to US$15,000-20,000, meaning the cheque required to build the next mine is far larger than the one that built the last.

Taken together, these three forces are not cyclical headwinds that fade when prices rise. They are structural ceilings on how fast supply can grow at all.

• Grade decline: average ore grades down approximately 40% since 1990, from 1.6% to 0.9%.
• Lead time extension: average project timeline stretched from 12.3 years to 16.3 years.
• Capital intensity: capex per tonne of new capacity doubled to US$15,000-20,000.

What this tells you is blunt: even at today’s near-record prices, the supply response the market needs to close the gap cannot arrive in time to matter for the next decade.

From surplus to deficit: what the near-term data shows

The inflection is already visible in the balance data. The International Copper Study Group (ICSG) projected in late 2025 that the refined copper market would swing from a 178,000-tonne surplus in 2025 to a 150,000-tonne deficit in 2026, even with demand growing a modest 2.1%.

Short-term balances remain sensitive to disruptions, and later revisions have swung again, which illustrates volatility within the trend rather than a contradiction of it. The direction of travel is what matters.

The longer view is where the scale becomes clear. BloombergNEF’s December 2025 outlook holds that copper enters a long-term structural deficit from 2026.

BloombergNEF warns of a potential 19 million-tonne shortfall by 2050 if new mines and recycling infrastructure are not aggressively developed.

That figure is the outer bound of what inadequate supply response looks like across a full generation. It is the consequence of inaction, not a forecast of it.

The demand stack: why electrification, EVs, and AI data centres are not competing narratives but additive ones

Here is where analysts often get the copper story wrong. They treat electrification, EVs, and AI as competing claims on the same barrel, as though one theme must win at the expense of another. The reality is that these demand sources stack. Each one adds to the pile rather than replacing what came before.

Start with the established base. The International Energy Agency (IEA) projects copper demand from clean-energy technologies to climb from 6.3 million tonnes in 2023 to 12.0 million tonnes in 2030, and on to 16.3 million tonnes by 2040.

Sitting inside and alongside that is the grid itself. IEA data shows grid-specific copper demand growing from about 5 million tonnes in 2020 toward 7.5 million tonnes by 2040 under stated policies, approaching 10 million tonnes in more aggressive decarbonisation scenarios.

Then comes the newest layer, and the least fully priced in. S&P Global forecasts data centres will lift copper demand from 1.1 million tonnes in 2025 to 2.5 million tonnes by 2040. BHP Insights goes further, estimating data-centre copper usage grows six-fold globally, from roughly 0.5 million tonnes per year today to around 3 million tonnes per year by 2050.

That AI layer carries genuine uncertainty. Reuters noted in mid-2026 that data-centre demand projections range widely, from 1.7 to 2.7 million tonnes by 2040, warning that efficiency gains and alternative materials could temper the uplift.

Data-centre copper demand is particularly hard to model because it aggregates hardware procurement cycles, grid interconnection schedules, and cooling-infrastructure build-outs that each carry their own lead times and do not move in lockstep with headline AI investment announcements.

BloombergNEF projects that overall energy-transition copper demand will triple by 2045.

The read you should take is this. Even if data-centre demand lands at the bottom of its range and efficiency gains soften the AI contribution, the clean energy and grid layers on their own are enough to sustain demand above anything the supply side can deliver this decade. That is what makes the demand stack a structural argument rather than a thematic one.

The geopolitics of copper: why refining concentration makes this a national security question

Mining the ore is only half the problem. The other half is who turns that ore into usable metal, and on that question the answer is uncomfortably concentrated.

China’s share of global copper fabrication capacity rose from 32% to 57% between 2015 and 2025, according to research on the sector. Chinese smelters and fabricators have driven around 80% of global growth in copper and copper-alloy fabrication capacity since 2019.

That concentration is the strategic vulnerability. Even when mined supply is adequate, refined copper availability can bottleneck at the smelter, and most of the world’s smelting muscle now sits in one country.

The U.S. policy response

Washington has already reached its own conclusion on this, and the institutional footprints are stacking up in sequence.

1. In August 2023, the U.S. Department of Energy added copper to its critical materials list.
2. On 1 August 2025, the U.S. applied Section 232 tariffs to 50% of imports of semi-refined copper products, aimed at encouraging domestic smelting.
3. In November 2025, copper was formally added to the broader U.S. government critical minerals list.
4. The U.S. State Department announced Project Vault, a US$10 billion strategic reserve initiative to secure copper and cobalt flows for domestic manufacturers.

The critical minerals designation process itself carries weight beyond symbolism: it triggers eligibility for Department of Defense funding, Export-Import Bank preferential financing, and interagency permitting coordination under the Federal Permitting Improvement Steering Council, all of which tilt the capital cost calculus for Americas-based projects.

That escalating architecture tells you something plainly. Washington now treats copper as a strategic vulnerability, and that judgment will shape capital flows and jurisdictional preferences for the rest of the decade, regardless of where spot prices wander.

For a U.S.-based investor, this reframes copper. It is no longer only a commodity bet. The policy stack being built around supply security is itself a durable demand signal for copper produced in allied jurisdictions.

Friendshoring as the practical alternative

The obvious fix, building domestic smelting at scale, runs straight into permitting delays, social opposition, and environmental constraints that block rapid production ramps. Speed is exactly what the U.S. lacks at home.

Wood Mackenzie has concluded that entirely replacing China in the copper supply chain is unfeasible, given the prohibitive costs and delays of building alternative capacity.

If full decoupling is off the table, the operational consequence is friendshoring: securing supply from close allies in the Americas. And that is precisely where Colombia enters the picture.

Colombia’s 2026 reset: why sovereign and institutional investors are reassessing the jurisdiction now

Colombia has long been one of the few genuinely underexplored regions with the geology to host large, near-surface copper. According to interviews with Copper Giant chief executive Ian Harris and director Mark Gibson, the country’s Andean porphyry systems, the geological formations that host most of the world’s biggest copper deposits, remain relatively untested.

Colombia’s geological endowment in Andean porphyry systems is comparable in style to the belts that host Escondida and Collahuasi in Chile, yet copper resources there have seen a fraction of the systematic exploration that Chile’s established districts attracted over the past four decades.

The structural advantages extend beyond rock. Colombia is the United States’ longest-standing ally in South America, its oldest democracy, and it offers direct logistical access to both Atlantic and Pacific port infrastructure. UPME estimates the country’s copper reserves at between 7.7 and 9.7 million tonnes.

What changed in 2026 was the politics. In the presidential runoff on 21 June 2026, Abelardo de la Espriella narrowly won with 49.66% of the vote, and took office on 7 August 2026 for the 2026-2030 term. His administration moved fast on mining.

• Scrapped 10 mining restrictions by decree, including repealing Decree 044 and the APAS/ZAPAS environmental designations that had stalled projects.
• Allowed social and governance applications to run concurrently with the Environmental Impact Assessment, compressing approval timelines.
• Removed restrictions across 15 mining districts, unlocking 147 metals and minerals projects representing US$4.85 billion in potential investment.

Institutional capital has already responded. In August 2026, Denarius Metals made a CA$28.8 million strategic investment for a 15.6% stake in Copper Giant Resources, gaining exposure to the Mocoa project. That deposit has surpassed the one-billion-tonne resource threshold, with an inferred resource of 636 million tonnes at 0.45% copper equivalent, containing 4.6 billion pounds of copper and 511 million pounds of molybdenum.

What the risks still look like in late 2026

None of this makes Colombia a low-risk jurisdiction, and honest analysis has to size the residual risk rather than wave it away.

Mining Stock Report assigns Colombia a composite risk score of 12/25, rated “High Risk,” citing complex bureaucracy, tax uncertainty, security issues, and under-developed infrastructure.

Political transitions do not erase legal exposure either. Just before leaving office, the outgoing Petro administration signed a resolution putting 186,500 square miles of the Amazon off-limits to new mining, a reminder that prior decisions can outlast the government that made them.

The calibration reference sits next door. Panama’s Supreme Court ordered the shutdown of the US$10 billion Cobre Panama mine in November 2023, stranding a colossal amount of capital overnight. That is the downside scenario against which Colombia’s improving posture must be weighed.

For investors deciding where in the Americas to seek copper exposure, Colombia now offers a combination of world-class geology, improving regulation, and a U.S. alliance relationship that was not on the table at this price two years ago. The risks are real, but they can be sized rather than feared.

Americas supply competition has intensified precisely because Chile and Peru are operating closer to their geological limits, pushing attention toward frontier jurisdictions that were previously considered too risky or too early-stage to attract institutional capital at scale.

What the timeline reality means for near-surface projects in the Americas right now

The strongest argument for advanced, near-surface assets is not a preference. It is arithmetic. Once you accept the lead-time math, the conclusion writes itself.

Consider Quellaveco in Peru. Anglo American approved the project in 2018, saw total capex rise to US$5.5 billion including Covid-related costs, and expects 310,000-340,000 tonnes of production in 2026, with full capital payback around four years after commercial production began. Even a world-class, well-funded asset took years to move from approval to output.

Cobre Panama tells the other side. A US$10 billion project accounting for roughly 2% of global copper production in 2022, shut down abruptly in 2023, is now grinding through 31.6 million tonnes of stockpiled material. Scale offers no protection when the jurisdiction turns.

Then there is the greenfield benchmark. A representative U.S. copper project maps out an eight-year framework after federal approvals are even secured.

1. Infrastructure construction: 2026-2028.
2. Equipment installation: 2029-2030.
3. Production ramp-up: 2031-2033.
4. Full operations from 2034.

If the structural deficit begins biting meaningfully by the late 2020s, and new greenfield projects need eight-plus years from approval to full production, then the assets that will actually matter are the ones that exist today in jurisdictions moving toward permitting rather than away from it. For reference, a project like San Matias-Alacran in Colombia carries a capex of roughly US$473.4 million, and the sector expects Colombia to attract around US$4.7 billion in new mines over the next four years.

That narrows the investable field considerably, no matter how high copper prices climb.

Where the structural thesis points from here

Four threads now sit together. Supply is capped by grade, time, and capital. Demand is stacking across electrification, grids, and AI. Refining concentration has turned copper into a national security question for the United States. And Colombia has reset itself into a credible friendshoring destination almost overnight.

BloombergNEF’s structural deficit, entered from 2026, is the starting point. Its projection that energy-transition copper demand triples by 2045 is the floor beneath the whole thesis.

Rather than wait for spot prices to confirm what the structural data already implies, track these three variables over the next 12-24 months:

• U.S. friendshoring capital pace: how quickly Project Vault and allied investment actually deploy into Americas supply.
• Colombia regulatory durability: whether the US$4.85 billion in unlocked projects survives legal challenge and translates into permits.
• AI data-centre demand realisation: whether it lands in the upper or lower half of the 1.7-2.7 million tonne forecast range.

Watch those, and the thesis becomes something you monitor rather than something you gamble on.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results, and financial projections are subject to market conditions and various risk factors.