Three converging developments in May 2026 expose the same structural failure at the heart of Western critical mineral strategy. A 600,000-tonne copper deficit, a peer-reviewed diagnosis of U.S. processing incapacity, and a Pentagon scramble to stockpile defense-critical minerals all point to a single chokepoint: the midstream. Mining more is not the answer. The bottleneck is in the refineries, smelters, and separation plants the West has spent three decades quietly dismantling.
Introduction
Three stories dominated the critical minerals beat this week, and at first glance they look like separate problems. A copper market in structural deficit, a peer-reviewed analysis warning that U.S. processing capacity has hollowed out, and a Pentagon racing to stockpile tungsten and antimony before the shelves go bare. Read them together, and they are the same story told three ways.
The connecting thread is the midstream: the smelters, refineries, and chemical separation plants that sit between raw ore in the ground and usable metal on a factory floor. Building on my analysis of this structural gap in "The Midstream Gap" back in May, this week's evidence suggests the problem is not getting better. It is getting worse, faster, and with more immediate consequences than the policy conversation has yet absorbed.
The implications run from copper wire in a data center to primers in an artillery shell. That range is not rhetorical flourish. It is a precise description of what is now at risk.
A 600,000-Tonne Warning Signal
Start with copper, because the numbers are stark. ING projects a refined copper deficit of approximately 600,000 tonnes in 2026, up from around 200,000 tonnes in 2025. Morgan Stanley puts its own estimate at the same level, calling it the largest deficit in more than 20 years. The International Copper Study Group, which as recently as October 2025 was forecasting a 209,000-tonne surplus for 2026, has revised that figure to a 150,000-tonne deficit. The direction of travel is unambiguous.
The supply disruptions driving this are well documented: Freeport-McMoRan's force majeure at Grasberg in Indonesia, the world's second-largest copper mine, following a September 2025 mudslide that killed workers and buried 800,000 tonnes of material into the mine workings. Benchmark Mineral Intelligence estimates the total output loss from that single event at roughly 591,000 tonnes between late 2025 and the end of 2026. Ivanhoe's Kamoa-Kakula in the DRC is still recovering from flooding. Chile's copper output fell more than 9 percent year-on-year in March 2026, with Codelco, BHP's Escondida, and Collahuasi all posting double-digit declines. And China's May 2026 halt on sulfuric acid exports, as I covered in "The Architecture of Dependence," is now affecting roughly 15 percent of global copper production that depends on acid-based leaching.
Predict any one of those disruptions in isolation and the market adjusts. Stack them all in the same calendar year and you have a different category of problem. Wood Mackenzie has raised its supply disruption assumptions for 2026 through 2028. Benchmark has downgraded original 2026 production growth forecasts by around 700,000 tonnes. The buffer that global copper markets relied on is shrinking, and a June 2026 Section 232 tariff decision in Washington is making things measurably worse.
Economically Trapped: How U.S. Tariff Policy Is Fracturing the Global Copper Market
Here is where domestic policy becomes a supply chain problem for the rest of the world. Traders anticipating U.S. copper tariffs began pulling large volumes of refined metal into COMEX warehouses earlier this year, pushing inventories up more than 300 percent to over 400,000 tonnes. At the peak in late January 2026, roughly 515,000 tonnes of copper was sitting in U.S. warehouses, representing more than half of all visible global exchange inventories.
That copper is not available to manufacturers outside the United States. Benchmark analyst Mackenzie described it plainly: "We use the term 'economically trapped' to refer to that copper as the current arbitrage and premium environment means there is no incentive for that material to be removed from the US." ING commodities strategist Ewa Manthey put the systemic consequence in equally direct terms: "You see all this material piling up in the U.S. warehouses, but that means that supply outside the U.S. is really tight, and that's really leaving very little room to absorb the supply shocks for the market. It's like an artificial tightness right now."
The Commerce Department faces a June 2026 deadline to recommend whether to impose tariffs of 15 to 30 percent on refined copper imports. Goldman Sachs has flagged the risk of at least 25 percent tariffs under Section 232. Producers are already pricing in the distortion: European and Asian buyers are facing record premiums from suppliers compensating for the margins they could otherwise earn selling into the U.S. market. As Li Xuezhi of Chaos Ternary Futures told Bloomberg: "Inventories used to act as a buffer, but now they're locked in the US. So the buffer is gone and everyone will have to scramble."
This is not a temporary dislocation. Manthey noted that material already moved into the U.S. is unlikely to return to the global system because trade policy remains unpredictable and sector-specific tariffs are still in place. The June decision will not resolve that uncertainty. It will either confirm it or deepen it.
The Processing Gap Is Structural, Not Cyclical
The copper tariff story would be uncomfortable enough on its own. What makes it more serious is the underlying structural context that a University of Maryland and West Virginia University analysis, published this month, sets out in peer-reviewed detail. The argument is precise: the U.S. does not have a mining shortage. It has a midstream processing crisis, and fixing it will take longer than most policymakers are willing to acknowledge.
The numbers are almost paradoxical. Benchmark Mineral Intelligence calculates that the U.S. can meet 146 percent of its domestic copper demand through a combination of mine output and scrap recycling. Yet nearly 48 percent of U.S. mined copper concentrate is exported, largely to China, for smelting and refining, before cathode copper is shipped back for industrial use. The country operates just three major copper smelters, producing just over 400,000 tonnes of refined metal against industrial demand that requires 720,000 tonnes of imports annually. As the University of Maryland analysis concludes, a mine without refining does not constitute mineral security.
The IEA data reinforces this at the systemic level. For copper, lithium, nickel, cobalt, graphite, and rare earths combined, the average market share of the top three refining nations rose to 86 percent in 2024, up from 82 percent in 2020. China is the dominant refiner for 19 out of 20 energy-related strategic minerals, with an average market share of around 70 percent. That figure is moving in the wrong direction despite years of Western policy effort and billions in public investment. The IEA's own conclusion is sobering: market forces alone will not deliver diversification.
Rebuilding domestic processing capacity involves more than capital. A modern copper smelter costs between 1.8 and 2.5 billion dollars to build, requires acid-capture systems to meet environmental standards, and takes the better part of a decade from investment decision to full production. More constrainingly, it requires a workforce that no longer exists in sufficient numbers. Hydrometallurgists, solvent extraction engineers, rare earth separation chemists: these are skills that take years to develop through formal training and practical experience, and U.S. mining and mineral engineering programs currently produce only a few hundred graduates per year. The World Resources Institute has noted the absence of any coherent coordinating framework in the U.S., with responsibility for copper policy scattered across the Interior, Energy, and Commerce departments, the EPA, and in some cases the State Department, with minimal coordination between them.
When the Stockpile Is the Strategy
The defense dimension of this story sharpens the urgency considerably. S&P Global reporting from March 2026 documented what military supply chain specialists had been warning about privately for longer: shortfalls in tungsten, antimony, gallium, and germanium are becoming acute vulnerabilities for U.S. weapons replenishment capacity. Peter Clausi of the Critical Minerals Institute stated it without qualification at the Prospectors and Developers Association of Canada conference: "If the US is running out of tungsten and antimony and it doesn't have access to germanium and gallium supplies, it cannot build new weapons."
The Pentagon's response has been aggressive by historical standards but constrained by the same structural problems that afflict civilian supply chains. The Defense Logistics Agency is deploying approximately one billion dollars in stockpile purchases, including up to 500 million dollars in cobalt, 245 million dollars in antimony sourced from U.S. Antimony Corp.'s Alaska-to-Montana processing chain, 100 million dollars in tantalum, and 45 million dollars in scandium. The broader Trump administration has committed 7.5 billion dollars across production and stockpiling, taken equity stakes in MP Materials, Lithium Americas, and Trilogy Metals, and is developing an AI pricing model through DARPA's OPEN program to establish reference prices for germanium, gallium, antimony, and tungsten.
Congress is moving in parallel. A bipartisan bill from Senators Shaheen and Young, with House cosponsors, proposes a 2.5 billion dollar Strategic Resilience Reserve modeled explicitly on the Strategic Petroleum Reserve. Australia has committed AUD 1.2 billion (around 802 million U.S. dollars) to a national stockpile targeting rare earths, antimony, and gallium from domestic producers, with an operational target of year-end 2026.
The scale sounds impressive. The constraints are real. The U.S. imports roughly 85 percent of its antimony and is nearly 100 percent import-dependent for gallium, with China controlling approximately 85 percent of global antimony production and an estimated 95 percent of refined gallium production. Fastmarkets analyst Solomon Cefai has noted that some of the DLA's desired procurement volumes, including 222 tonnes of indium, would nearly equal total annual U.S. domestic consumption. Argus Media's Cristina Belda observed that the targets for certain materials may simply outstrip available non-China supply at any price.
China's export control record makes this procurement race feel genuinely urgent. Following restrictions on antimony in August 2024, Chinese shipments of antimony products dropped 97 percent. No shipments of wrought or unwrought germanium or gallium have been sent from China to the United States in the past year. China did suspend some dual-use export controls in late 2025, but the suspension runs only until November 2026, and the prohibition on exports to U.S. military users or for U.S. military end uses remains fully in effect. The underlying architecture of leverage has not changed.
The Convergence Problem: When Every Solution Runs Into the Same Wall
What links these three stories is not simply that they share a root cause. It is that the proposed solutions to each of them run into the same constraint from a different angle. The copper deficit can theoretically be eased by domestic refining capacity, but the U.S. lacks the smelters. The defense stockpile race depends on domestic processing chains, but for most of these minerals the processing infrastructure simply does not exist at the required scale. The midstream investment push requires a trained workforce, but the pipeline for that workforce takes years to build.
The time mismatch is the most acute version of this problem. The average time from mineral discovery to first production remains around 16.5 years. The June 2026 Section 232 tariff decision will be made in weeks. Military stockpiles are depleting in months. A new copper smelter takes a decade from permitting to production. A hydrometallurgist takes years to train. These timelines do not overlap in ways that are useful for near-term security planning.
There are glimmers of progress worth noting. Grupo México is evaluating the restart of its Hayden smelter in Arizona. Falcon Copper is assessing a two billion dollar greenfield smelter and refinery in the western U.S., with potential Japanese participation as off-taker. Australia's combination of stockpile purchasing, processing incentives, and workforce investment offers a model that connects these elements more coherently than most U.S. efforts to date. Canada's critical minerals strategy explicitly links mining projects to battery manufacturing through funded processing hubs and training programs tied to specific industries.
The IEA's projection that refining concentration will decline only marginally over the next decade, returning to roughly 2020 levels even under optimistic policy assumptions, suggests that these incremental efforts are not yet scaled to the problem. As the University of Maryland analysis observes, policy without actual investment is nothing more than a statement of intent.
What Comes Next
The June 2026 Section 232 copper tariff decision is the most immediate policy inflection point to watch. A tariff in the 25 to 30 percent range would lock additional copper into U.S. warehouses, tighten ex-U.S. markets further, and raise input costs for every downstream U.S. manufacturer that uses refined copper, from electric motor producers to defense contractors. It would also accelerate the COMEX/LME spread distortion that is already pricing European and Asian buyers out of accessible supply.
Beyond that, the SECURE Minerals Act's progress through Congress will test whether the bipartisan consensus on critical mineral vulnerability translates into actual appropriations or remains, as past efforts have, a statement of intent stalled in committee. Australia's strategic reserve becoming operational by year-end 2026 will matter more if it is accompanied by genuine off-take agreements with U.S. buyers under the October 2025 bilateral agreement, rather than simply building a national buffer.
The harder and slower question is workforce and infrastructure. The DOE's critical minerals workforce initiative is a necessary signal, but the gap between a few hundred annual mineral engineering graduates and the industrial base the U.S. wants to rebuild is not closable in a single budget cycle. The China-won-by-building-refineries insight that circulates in policy circles is correct and still underweighted in practice. The West now broadly understands that the bottleneck is in the middle. The distance between understanding that and actually fixing it remains the central challenge of the decade.