Critical Minerals 2026: Copper, Lithium and Rare Earths as Geoeconomic Weapons
- Thierry Marquez
- 23 hours ago
- 25 min read
Updated: 9 hours ago
Contents
The Definitional Shift: From Commodity to Jurisdiction, and What 10 November Prices In
The Licencing Siege: Dysprosium Arithmetic and the Japan Precedent
The Acid Test: How Beijing Learned to Regulate the Input, Not the Metal
The Magnet Chokepoint: Gram-Level Dependencies, Decade-Level Timelines
The Jakarta Doctrine: Quota Mathematics and the Bid to Move from Price-Taker to Price-Maker
The Copperbelt Paradox: Record Prices Above, Structural Fatigue Below
The Subsidy Ceiling: What Project Vault Buys, and What It Cannot
Key Takeaways
The truce is a countdown, not a settlement. China's suspension of its October 2025 rare earth export controls expires on 10 November 2026; the suspension of the gallium, germanium and antimony bans lapses on 27 November. The military end-use prohibitions were never suspended at all. What markets are treating as normalisation is in fact a licence, renewable at Beijing's discretion — and the machinery of restriction (licensing, entity lists, export-retention rules, whistle-blower enforcement) has grown more sophisticated, not less, through the pause.
Refining capacity, not geology, is the binding constraint. Roughly 90% of rare earth refining and 96% of battery-grade graphite processing sit inside one jurisdiction. Heavy rare earths — dysprosium and terbium — will remain majority-controlled by Beijing into the mid-2030s even under the most optimistic Western build-out. The metal is in the ground on every continent; the chokepoint is the facility, and facilities take a decade.
Producers have learned the Chinese lesson and are applying it upstream. Jakarta cut nickel mining quotas by more than 30% and is building a state-run commodity exchange to become a price-maker. Kinshasa moved from an outright cobalt suspension to export quotas and, in August 2026, banned concentrate exports outright. The resource nationalism of the 2020s is no longer about owning mines; it is about managing scarcity deliberately — a strategy pioneered by Beijing and now operating against everyone's assumptions, including China's.
The West is buying time, not substitution. Project Vault's $12 billion architecture, the Pentagon's equity stakes and price floors, the EU's strategic projects pipeline — all are real money against real capacity gaps. European domestic extraction stands near 5% against a 2030 target of 10%; processing near 15% against 40%. Subsidies have changed the floor under Western processors; they have not yet changed the ceiling above Chinese ones.
Copper is the transmission mechanism to the non-mining balance sheet. A record rally above $13,000 a tonne, a Chilean output decline approaching 3%, an acid crisis engineered by a Chinese export ban on an industrial chemical — these are not mining-sector events. They propagate into grid capex, defence procurement, EV cost curves and inflation indices with a lag of quarters and a lag of years respectively.
Portfolio-level. The investable distinction over the next eighteen months is between instruments that price quota politics and instruments that price substitution capex. The first category — producer equities exposed to export-restriction regimes, refining-linked margins, royalty streams indexed to administered scarcity — carries a policy risk premium that is still under-priced because it has appeared to be episodic for fifteen years and is becoming structural. The second category — Western processor offtakes carrying sovereign backstops, price-floor-protected magnet capacity, ex-China material premia — is already partially priced. The exposure that is almost entirely unpriced is neither: it is the option value of 10 November 2026, a binary repricing event dressed as a calendar footnote.
1. The Definitional Shift: From Commodity to Jurisdiction, and What 10 November Prices In
Critical minerals in 2026 are no longer inputs. They are jurisdictions. The distinction sounds rhetorical; it is not. An input is procured on price, quality and logistics. A jurisdiction is entered with a compliance assessment, a sanctions screen and a cost-of-capital assumption. The entire vocabulary of procurement — spot price, approved vendor list, dual sourcing — presupposes that the seller is substitutable. On the evidence of the past eighteen months, in four families of materials, the seller is not.
The numbers on the demand side explain why this happened now. The U.S. Geological Survey placed net U.S. imports of processed metals and materials at $185 billion for 2025, against $77 billion the previous year — more than a doubling in twelve months (USGS Mineral Commodity Summaries, February 2026). The American critical minerals list broadened from fifty items in 2022 to sixty in the latest review, adding copper, uranium and lead. The European Commission's own modelling projects lithium demand multiplying twelvefold by 2030 and twentyfold by 2050, with permanent-magnet rare earths on a sixfold curve. Demand curves of this steepness convert supplier concentration into leverage arithmetically, without any strategic decision being taken at all.
The strategic decisions were taken anyway. In April 2025, China's Ministry of Commerce imposed mandatory licensing on seven medium and heavy rare earth elements. In October 2025, the framework was extended to twelve. In November 2025, following the Xi-Trump meeting in Korea and the trade truce it produced, MOFCOM suspended the October measures for one year — until 10 November 2026 — while the December 2024 bans on gallium, germanium and antimony exports to the United States were suspended until 27 November 2026. What the truce did not touch matters more than what it did: the prohibition on military end-uses remains fully in force, the licensing apparatus remains in place, and on 24 February 2026 forty Japanese companies were added to the dual-use entity list. A suspension that leaves the trigger mechanism armed is a negotiating instrument, not a disarmament.
It is therefore likely that the November expiries produce neither clean extension nor clean collapse, but a sequenced repricing: selective continuation of licensing for defence-linked materials, expansion of the general licences Washington purchased in the truce, and a further carve-out of Japanese and Taiwanese supply as the targeted channel. The pattern is already visible in the data. China's rare earth exports to the United States rose through mid-2026 — 647 tonnes of permanent magnets in July, the second-highest monthly volume since the controls began — while exports to Japan fell 51% year-on-year in the first half, collapsing 81% in June alone, with yttrium shipments at zero (Nikkei via Asia Times, 12 August 2026; Reuters, 20 August 2026). The weapon is no longer aimed at America. It is being calibrated, target by target, at everyone else.
For organisations with procurement exposure to Chinese-processed materials: model 10 November 2026 not as an expiry but as a repricing gate. Assume your current landed-cost baseline carries a 15-25% geopolitical premium that either vanishes on 11 November or doubles — and identify which of your tier-two suppliers' quotes silently embed an assumption one way or the other.
2. The Licencing Siege: Dysprosium Arithmetic and the Japan Precedent
The Japan campaign of 2026 is the most instructive controlled experiment in economic coercion since the 2010 Senkaku embargo — because unlike 2010, it is running against a prepared target, and it is working anyway.
The mechanics are surgical rather than spectacular. Beijing cited national security and Japan's "re-militarisation" in justifying the curbs, then layered controls on dozens of Japanese defence-linked firms, including the shipbuilding and aero-engine divisions of Mitsubishi Heavy Industries. Since January, exports of dual-use rare earth products to Japan have been tightened, and certain companies prohibited outright. July was the ninth consecutive month in which China sent Japan no dysprosium oxide, and the eighth for terbium oxide — two elements used in gram quantities in high-performance magnets, where they are irreplaceable. Rare earth magnet exports to Japan fell 52% in July to 111 tonnes, roughly half the January level (BigGo Finance, 20 August 2026; Reuters, 20 August 2026). The pressure has escalated in lockstep with Prime Minister Takaichi's parliamentary remarks on a Taiwan contingency — a reminder, as assessed in our China 2026: The Overstretch assessment, that Beijing's coercive toolkit is growing more precise precisely as its macroeconomic position weakens.
Taiwan is being squeezed through a different aperture. According to Nikkei Asia reporting (20 August 2026), Chinese customs have since 2025 been quietly delaying clearance for germanium-based and quartz-based materials and certain rare earth magnets bound for the island — no announced ban, no press release, just inspection regimes stretched to months, suppliers summoned for questioning about their customers. Taiwanese optics and semiconductor-equipment suppliers report delivery times of several months and lost orders. The lesson for any board watching the Taiwan file: the first phase of a Taiwan crisis will not arrive as a missile. It will arrive as a customs form.
Underneath the diplomacy sits an enforcement architecture that matured this year. On 24 June 2026, China introduced a whistle-blower system for dual-use violations involving strategic minerals, extending monitoring across the entire supply chain, effective 1 July. Add the February entity-list expansion, the export-VAT rebate reductions on battery materials, and the phased lithium-battery consumption tax taking effect in September 2026, and the picture is of a state that has industrialised the administration of scarcity. Gareth Hatch, managing director of Strategic Minerals Advisory, captured the unresolved question for industry: "Very few folks outside of China are talking yet about, 'how do we produce erbium, how do we produce yttrium?'... We're lagging" (IEEE Spectrum, 26 August 2026).
The scale of what is being administered, if fully applied, is quantified by the International Energy Agency: full implementation of the October 2025 controls would put roughly $6.5 trillion of annual downstream production outside China at risk (IEA estimate cited in Tocco, 24 August 2026). Against that ceiling, the actual coercive dose has been kept deliberately sub-lethal — enough to demonstrate optionality, not enough to trigger genuine Western mobilisation. It is highly likely that this calibration persists through the November gates, because both the demonstrative and the escalatory versions of the campaign serve Beijing's negotiating position, while the terminal version does not: as the Mexico 2026 assessment documented in the commercial sphere, China's leverage strategy depends on dependency remaining profitable enough to preserve.
For organisations with Japanese, Taiwanese or Korean supply-chain tiers: audit traceability to the smelter level, not the trader level, before 10 November. Where custom-clearance friction is already measurable — the Taiwan germanium channel is the template — treat a doubling of lead times as a baseline planning assumption for 2027 rather than a disruption scenario.
3. The Acid Test: How Beijing Learned to Regulate the Input, Not the Metal
The most elegant act of resource-statecraft in 2026 targeted no metal at all.
On 1 May 2026, China halted exports of sulphuric acid — the by-product of copper and zinc smelting that is the world's most produced industrial chemical and the unglamorous reagent on which roughly a fifth of Chilean copper output, and close to 45% of Congolese production, depends via acid-leach processing (Mining.com, 10 April and 4 May 2026; Mining.com, 16 July 2026). The proximate cause was the Iran war: the effective closure of Hormuz blocked Middle Eastern sulphur shipments, tightening the acid market worldwide, and Beijing converted a commercial squeeze into a policy exit — the ban covering smelter by-product acid exactly when that acid became scarce abroad. The result was arithmetic. China's acid exports collapsed from roughly 116,700 tonnes in May to 980 tonnes in June — a 99% single-month fall (Discovery Alert, 20 July 2026). Chile, which buys more than a million tonnes of Chinese acid a year and sourced about a third of its supply from the country in 2025, suddenly faced the conversion cost of its entire leaching segment.
The downstream chain is being repriced in sequence. Goldman Sachs analysts calculated that restrictions holding through year-end would put about 200,000 tonnes of acid-dependent Chilean output — near 1% of global supply — at risk (Goldman Sachs note, 21 April 2026). The IEA went further, warning that the copper supply outlook has "worsened considerably," that mines producing more than a seventh of world primary supply are hostage to the acid market, and that Chinese smelter capacity cuts of over 10% agreed this year are "not enough to meaningfully balance the market" (Mining.com, 16 July 2026). Meanwhile the same ban enriched Chinese smelters — earning above 5,000 yuan per tonne of copper from acid sales at record domestic prices, in Yang Changhua's account for Beijing Antaike (Mining.com, 21 April 2026) — a reminder that these instruments redistribute rents even as they restrict flows.
The copper price is the transmission belt. Three-month futures on the LME tore through their all-time records during the year, at one stage up as much as 10% intraday and topping $13,000 a tonne as supply disruption, trade policy and AI-driven grid demand converged (Yahoo Finance, 2026). The rally has carried the character that makes analysts uneasy — their own desk called it "unsustainable" even as it printed. Goldman maintained an average forecast of $12,650 for the year and a 490,000-tonne surplus, while simultaneously flagging the acid channel as capable of erasing it (Reuters, 21 April 2026). Daniel Yergin's framing — that copper is "either the enabler of the modern world and this age of electrification, or it's an obstacle to it" — has stopped being rhetorical.
Two second-order couplings deserve board attention. First, the interaction between the acid ban and the war's effects on diesel and logistics: Freeport lifted its 2026 cost estimates at Grasberg partly on volatile acid and diesel prices, and Zambia's Jubilee Metals explored pooling acid purchases with competitors — "sulfuric acid is a worry," its finance director Jonathan Morley-Kirk conceded on an earnings call (Mining.com, 24 April 2026). Robert Friedland, delivering Ivanhoe's quarterly results, warned bluntly of "a second-derivative effect... on global copper production due to the shortage of the world's most important industrial chemical" (Benzinga, 14 April 2026). Second, the geographic concentration of the vulnerability, which runs through exactly the theatres our Strait of Hormuz conflict analysis and Chile 2026 assessments mapped from the energy and sovereign sides.
Because the acid precedent — regulating the chemical rather than the metal — offers a deniability and reversibility that raw-material embargoes lack, it is now likely to become the template for future coercive episodes, and boards should read the next chemical shortage not as a market event but as a rehearsal.
For organisations with energy, grid or heavy-industry exposure: model your 2027 copper-cost assumptions on the 2025 realised price plus a disruption premium rather than on long-run mean reversion, and identify which capex programmes have break-evens that fail above $13,500 a tonne — they are the ones the November gates can injure without ever appearing in your supply-chain map.
4. The Magnet Chokepoint: Gram-Level Dependencies, Decade-Level Timelines
Permanent magnets are where geoeconomic leverage and industrial physics meet, and the physics are unforgiving. A NdFeB magnet contains perhaps a few hundred grams of dysprosium or terbium — but without those grams, the magnet demagnetises at operating temperature, and the traction motor, the missile fin actuator, the wind-turbine generator and the drone gimbal all become procurement problems instead of manufacturing ones.
The price data from 2026 describes a market that has already split in two. Terbium trades around $1,000-1,100 per kilogram inside China, while ex-China material clears closer to $3,600-4,000. Dysprosium runs roughly $240 per kilogram domestically against $800-900 outside (market commentary, Yahoo Finance, 2026). Those spreads — a factor of three and more — are not arbitrage opportunities; they are the visible surface of an export-licence queue. When a material sells at half price at home and a premium abroad, the difference is the political price of leaving.
The build-out math is what separates this domain from ordinary industrial-policy debates. Bloomberg synthesis of consultancy projections — McKinsey among them, corroborated by CRU and Benchmark Mineral Intelligence — concludes that for dysprosium and terbium, production outside China will still cover less than a fifth of ex-China demand by 2035 (Bloomberg, 2026). Not a fifth of global demand — a fifth of the demand the rest of the world will have. Lynas in Malaysia remains the only operator outside China that has mastered heavy rare earth refining at scale. That is thirteen years from now; the vehicles, ships and aircraft being designed in 2026 will still be in production then.
Against that timeline, the American response has been faster than conventional wisdom allows and slower than the deadline requires. MP Materials — in which the Department of Defense took a $400 million convertible position, roughly 15% of the company, alongside a $150 million loan, a ten-year offtake covering the entire planned output of its Fort Worth magnet facility, and a NdPr price floor of $110 per kilogram — reported Q2 2026 revenue of $108.5 million, up 89% year-on-year, NdPr production of 840 tonnes (+41%) and first magnetics revenue of $16.5 million, with $1.45 billion in liquidity (MP Materials Q2 2026 results). Apple signed a $500 million agreement. USA Rare Earth closed a $1.6 billion Commerce Department package in June and commissioned its Stillwater magnet line. These are real numbers against a real curve — and the curve's slope is the problem. The IEA's Global Critical Minerals Outlook 2026 documents capability gaps far beyond processing tonnes: in rare earth magnet production, grain-boundary diffusion technology is highly patented, with only one equipment supplier outside China, equipment costs more than ten times higher and lead times correspondingly longer (IEA, August 2026).
The second front is graphite, the quiet one. China controls roughly 96% of battery-grade graphite refining — the spherical and synthetic material in virtually every lithium-ion anode — with controls tightened in December 2023 and reinforced since. Qualifying a new anode supplier takes 18 to 36 months of cell-level testing before production entry (Tocco, 24 August 2026), which means graphite substitutions decided this quarter reach vehicles in 2028 at the earliest.
Given the capital intensity, the patent walls and the personnel constraints — and given Beijing's documented practice of crashing prices whenever a Western competitor approached viability, a pattern repeated for two decades — it is highly unlikely that the ex-China share of heavy rare earth supply moves materially above the consultancy projections before 2030. The honest conclusion is that magnet security is being purchased by decade, while the coercive risk against it is being repriced by quarter.
For organisations with defence, automotive or aerospace manufacturing exposure: the American deadline — new legislation prohibiting the Department of Defence from acquiring Chinese rare earth magnets, with compliance demanded from January 2027 — will cascade into prime-contractor flow-downs within months, not years. Inventory your magnetic bill of materials now, and assume any line that cannot certify non-Chinese content within four quarters will be re-engineered, voluntarily or otherwise.
5. The Jakarta Doctrine: Quota Mathematics and the Bid to Move from Price-Taker to Price-Maker
The most consequential shift of 2026 in this domain is not happening in Washington or Brussels. It is happening in producer capitals that have studied Beijing's playbook and are running it against their own buyers.
Indonesia is the vanguard. Since September 2025, Jakarta has cut nickel mining quotas — the RKAB approvals — by roughly a third: from 379 million tonnes approved in 2025 to a 2026 allocation of 260-270 million tonnes, with DBS tracking approved volumes as low as 205 million wet metric tonnes against Argus-estimated domestic ore consumption near 330 million tonnes — an ore gap measured in tens of millions of tonnes whatever the counting convention, the steepest policy tightening in the sector's history. The regulatory fine print matters as much as the headline number: from 2026, permits have been shortened from three years to one, making administered scarcity revisable annually — and therefore politically mobilisable at every budget cycle (Argus, 11 February 2026). PT Weda Bay Nickel, operator of the world's largest nickel mine, halted operations in May under the quota reductions (Yieh, 24 April 2026). The market effect was immediate and precisely attributable: LME nickel jumped roughly 15% to $17,860 a tonne by 22 January, surged 8.9% in a single session to $18,524 — its highest level since mid-2024, on BMO Capital Markets' reading — and touched $18,950 on 29 January, a 1½-year high, before consolidating near $18,460 in subsequent dealing (Manufacturing Asia/DBS, 22 January 2026; Northern Miner/BMO, February 2026). BMO calculates that the cuts could remove as much as 700,000 tonnes of nickel supply — enough to flip the market from surplus to deficit; BMI/Fitch dissents, forecasting a $15,500 average on a persistent surplus. The disagreement itself is the datum: Indonesia now sets the direction of travel for the global price through administrative documents, and analysts argue only about the amplitude.
The downstream logic inherited from the 2020 ore ban remains intact: national mineral exports grew from roughly $3.3 billion in 2017 to over $30 billion after raw-ore shipments were forced into domestic smelting, with Chinese state-linked companies investing an estimated $65 billion to control some 90% of the industry — a concentration documented extensively in our Indonesia 2026 assessment. What is new in 2026 is the direction of the pressure: Jakarta is now restricting ore to discipline its own Chinese-dominated smelter class and manage price.
President Prabowo's administration has added the institutional layer. A Strategic Mineral and Commodity Exchange — Icomex — is scheduled for launch by 1 January 2027, with implementing rules due by 17 September; a new regulation requires non-oil and gas exporters to retain earnings onshore for at least a year; and in April the Danantara sovereign-wealth vehicle committed a $20 billion Morowali joint venture alongside CATL, Hyundai and LG Energy Solution. The declared objective, in Prabowo's own framing, is to move the country from price-taker to price-maker. The local press is already asking the sceptic's question — the ICDX has operated for two decades without gaining global price influence — but the combination of quota control over 60% of world nickel supply, a state-run price venue and capital-retention rules is a coherent architecture for administered scarcity, not a rhetorical one.
Kinshasa has gone further, faster. The Democratic Republic of Congo — producer of some 70% of the world's cobalt — suspended cobalt exports in February 2025, then converted the suspension into export quotas of 96,000 tonnes a year for 2026-2027, paired with a strategic reserve of 9,600 tonnes. The market consequence was a price move from around $22,000 to the $54,000-55,000 range — Fastmarkets puts the cumulative lift near 160% — and Glencore's Congolese output fell 51% in the first half of 2026 under the quota regime (Fastmarkets via industry commentary, 20 August 2026; Minelistings, 23 August 2026). On 6 August 2026 came the next step: a ban on exports of copper and cobalt concentrates outright, with one-year exemptions available for strategically important flows and new by-product taxes following a three-month transition (Reuters, 6 August 2026). The official rationale is domestic value capture; the timing coincides with an investigation into uranium contamination in cobalt shipments — an alleged 2,000 to 5,000 tonnes of uranium — which could yet weaponise compliance screening against Congolese material. Christian-Geraud Neema of the China-Global South Project judged the concentrate ban's immediate impact limited, "as the bulk of Congo's copper and cobalt is already refined domestically," with Kamoa-Kakula the most exposed exemption case (Reuters, 6 August 2026). Our DR Congo 2026 assessment covers the underlying political economy.
The doctrinal point outruns both cases. Quota management, strategic reserves, concentrate-export bans and state marketing venues are converging into a standard producer toolkit — one visible from Harare's lithium restrictions to the quota arithmetic now travelling through Southern Africa's corridors. Given the demonstrated price elasticity — cobalt's near-doubling on a quota announcement, nickel's double-digit climb on a permit document — it is now likely that at least two additional significant producers adopt formal export-restriction regimes for battery or magnet materials before end-2027, because the demonstrated returns accrue to the state budget in year one while the investment-deterrence costs land in year ten, after the incumbent government's horizon.
For organisations with battery, stainless-steel or magnet supply-chain exposure: map your inputs against the quota architecture — RKAB numbers and their now annual review cycle, cobalt quotas, concentrate rules — and treat every producer-state election and budget between now and 2028 as a repricing event with published, dated mechanics.
6. The Copperbelt Paradox: Record Prices Above, Structural Fatigue Below
The paradox of the copper market in 2026 is that the price signal is screaming abundance of demand while the production base is signalling exhaustion of supply — and the political economy above the mines is discounting both.
Chile's year has been a study in systematic underdelivery. Cochilco, the state copper commission, cut its 2026 national output forecast for a second consecutive quarter, to 5.27 million tonnes, down 2.6% on 2025 and some 300,000 tonnes below its earlier projection, citing weak performance at Codelco, BHP's Escondida and Spence, plus structural constraints across operating sites (Bloomberg, 11 August 2026; SMM, 12 August 2026).
First-half production reached 2.481 million tonnes, down 6.6%, after April fell 13.8% year-on-year — a 399,954-tonne month, per the national statistics agency (Rio Times, 17 August 2026). Codelco's own output fell 11% to 564,000 tonnes in the half, even as higher prices quadrupled its pre-tax profit to $1.97 billion; Chairman Bernardo Fontaine conceded the company's 1.33-1.36 million tonne annual target was "difficult to achieve" (Reuters, 28 August 2026). Escondida is preparing to reduce production by 300,000 tonnes to 1.1 million tonnes in 2027 as concentrator feed grades fall from 1.02% to 0.90% (UPI, 24 August 2026). Severe winter storms cut further guidance. The mining sector dragged second-quarter national GDP to a 0.2% contraction — and this at a moment when the country holds roughly 18% of world reserves.
The lithium flank looks brighter on paper: SQM booked record quarterly sales above 84,000 tonnes of lithium carbonate equivalent in the second quarter, with the Codelco venture NovaAndino — formalised in December 2025 to run the Salar de Atacama through 2060 — producing 280,000-290,000 tonnes this year and approaching 300,000 tonnes of capacity next, anchoring a $3 billion 2026-2028 investment plan of which 60% flows to the venture, against a global demand line SQM expects to exceed 2.1 million tonnes (SQM Q2 2026 earnings call; BNamericas, 22 August 2026). Yet the lithium counter-example disciplines the entire analysis. Prices actually tumbled mid-year as traders braced for the restart of CATL's Jianxiawo mine — capable of roughly 46,000 tonnes of lithium carbonate annually, near 3% of global supply — with Citi judging the restart already priced in (Oilprice.com, citing Citi, 2026). The contrast with nickel and cobalt is the strategic lesson. Jakarta and Kinshasa cut quota into a concentrated processing and smelting base they control, and the price obeyed. Chilean lithium sits in a market where Chinese conversion capacity can expand and contract around it — the Jianxiawo restart moved the global price more than the Salar's entire year of policy. Owning the rock is leverage; owning the flows is power. Chile owns the rock and rents the flows, which is why its lithium trades at commodity prices while its copper, threading through Chinese acid and administered quotas, records geopolitical ones. The Argentina 2026 assessment tracks the same triangle from the RIGI incentive side.
Zambia presents the ambition ledger in mirror image. Output of 890,346 tonnes in 2025 missed a one-million-tonne target; the state nonetheless targets three million tonnes by 2031, courting investors — JCHX, Barrick, First Quantum, Vedanta, IRH, KoBold — with more than $10 billion committed since 2021, per Chamber of Mines president Anthony Malenga (Mining.com, 10 March 2026; Kitco, 4 August 2026). The August 2026 elections have made the trajectory explicitly political. But the binding constraints are physical: Trafigura withdrew from the proposed 2,000 MW transmission project meant to wheel surplus Angolan hydro into the DRC-Zambia mining belt (SMM, 22 July 2026); sulphuric acid pooling is now an operating concern for mid-tier producers; and First Quantum's Sentinel mine saw quarterly output slip on grades and recoveries even as the company raised full-year group guidance to 405,000-475,000 tonnes (Mining.com, 29 April 2026). KoBold's Mingomba — 300,000 tonnes annually in the early 2030s — captures the timeline problem in one project: fast by mining standards, irrelevant to a 2027 squeeze. Our Zambia 2026 assessment carries the political detail.
Across the belt, the pattern repeats: producer states are converting record prices into fiscal ambition rather than into throughput. That is a rational sovereign choice. It is also why the price is recording records. Given Chile's grade-decline profile, Codelco's multi-year restructuring and the three-to-seven-year cadence of the projects meant to fix both, it is unlikely that refined copper supply growth exceeds 2% annually through 2028 — well below the trajectory S&P Global says avoids a looming multi-million-tonne shortfall, with demand projected to grow from 28 million tonnes in 2025 toward 42 million by 2040 (S&P Global, cited by Yahoo Finance, 2026).
For organisations with grid, electrification or defence-infrastructure exposure: your procurement plans already assume inflation; re-run them assuming not 2026 prices, but 2026 prices plus quota behaviour from at least one major supplier — and mark the 2027 Escondida step-down as a dated input, not a scenario.
7. The Subsidy Ceiling: What Project Vault Buys, and What It Cannot
Washington's 2026 answer to all of the above is the largest industrial-minerals mobilisation since the Cold War — and it must be assessed against what it is actually for: time.
Project Vault, signed by executive order on 2 February 2026, combines a $10 billion Export-Import Bank loan — the largest in EXIM's 92-year history — with $1.67 billion of private capital to buy and stockpile critical minerals for industrial users, spanning the automotive, aerospace and energy sectors. EXIM has since issued $14.8 billion in letters of interest, bringing stated support above $30 billion within six months; the One Big Beautiful Bill Act allocates a further $7.5 billion, including $2 billion for the national stockpile by 2027 and a $500 million Pentagon credit programme; a bipartisan congressional bill would add a $2.5 billion Strategic Resilience Reserve with discretionary authority to buy above market (Mining.com, 2 February 2026; CNBC, 3 February 2026; Axios, 15 January 2026). In August the administration layered on a further $3 billion push for defence-linked minerals projects, hosting a mining-industry roundtable with the secretaries of Interior and State (CNBC, 8 August 2026). The U.S. convened the 2026 Critical Minerals Ministerial with 54 countries and the European Commission. The stockpile's scope covers any of the more than fifty minerals the USGS lists as critical.
Brussels has moved in parallel, with the standard European profile: regulation first, money later. The Critical Raw Materials Act sets 2030 benchmarks of 10% domestic extraction, 40% processing and 25% recycling, a 65% cap on single third-country supply, and 27-month permitting guarantees for strategic projects — 60 of which are now approved, 47 inside the bloc (European Commission; Munich Security Conference analysis, August 2026). Actual standing, as of August 2026: roughly 5% extraction, roughly 15% processing (industry tracking, August 2026). The European Court of Auditors' Special Report 04/2026 judged the CRMA targets "non-binding," covering only strategic raw materials and lacking justification, with funding "scattered" and results untracked — a verdict assessed in full in our May 2026 Critical Minerals Trilemma briefing. The RESourceEU plan of March 2026 added €3 billion alongside the EIB's €2 billion annual commitment.
The honest ledger for critical minerals in 2026 has two columns. On the asset side: the Pentagon's direct equity positions — MP Materials, Lithium Americas, Trilogy Metals, USA Rare Earth, Vulcan Elements, ReElement — constitute an industrial-policy posture without American precedent, and the price-floor architecture genuinely changes private capital's willingness to fund processing. The MP deal alone — equity, loan, decade-long offtake and a $110 per kilogram NdPr floor — converts a speculative magnet plant into a quasi-utility. On the liability side: none of this touches the mid-2030s math on heavy rare earths; the stockpile buys months of coverage, not years; and the Australian, Kazakh and African supply the Western architecture must scale through is being simultaneously courted by Beijing, a competition mapped in our Kazakhstan 2026 assessment of the C5+1 minerals gambit. Guancha's verdict on the January 2027 de-Chinese-magnet deadline — that stopping rare earth purchases from China within five months is "a mission impossible" for U.S. defence contractors — is propaganda-adjacent in source but sound in arithmetic (Asia Times, 12 August 2026).
The deeper constraint is doctrinal rather than fiscal. Beijing's control of these markets has always had a price-crash instrument at its disposal — flooding supply whenever a Western challenger approached viability, a pattern two decades deep — and Western policy has answered with floors rather than with anything that raises Beijing's cost of deploying it. Subsidies that make domestic processors whole when prices fall do not prevent the falls; they wait for them. Given the scale of announced capital and the demonstrated institutional commitment on both sides of the Atlantic, it is now almost certain that Western ex-China processing capacity grows materially through 2030. It is likely that Chinese share of global processing nevertheless declines by percentage points, not tens of points, over the same horizon. Both statements are true simultaneously; that is the subsidy ceiling.
For organisations with capital allocation across the sector: the investable question is not whether Western capacity gets built, but whether its economics were underwritten by policy that survives the first price crash — read every offtake, floor and credit for its fiscal-year dependency before you read its tonnage.
8. Critical Minerals 2026 — Three Scenarios
Scenario A — Managed Volatility: The Licence Economy Renegotiated (Probability: ~40-45%)
Beijing extends the November 2026 suspensions selectively — preserving general licences for U.S.-bound flow purchased in the truce, tightening Japanese and Taiwanese channels — while producer-quota regimes (RKAB, cobalt, concentrates) institutionalise into predictable, taxed, but stable scarcity. Prices stay elevated with periodic spikes; Western build-outs continue on schedule; the sector trades on fundamentals punctuated by named dates. This scenario holds unless one or more triggers fire: a kinetic or blockade incident in the Taiwan Strait converting customs friction into embargo; a second front in Beijing's coercive targeting (Korea, EU after a 21st-sanctions-package collision); a Congolese or Indonesian quota cut sharp enough to break cathode or smelter economics; or a Western price-floor programme that collapses politically on cost overruns during a copper correction. The probability is elevated by both governments' demonstrated preference for demonstrative rather than terminal coercion, and reduced only by the accumulation of armed mechanisms — entity lists, whistle-blower enforcement, customs slow-walking — that make incidental escalation easier than deliberate de-escalation.
Scenario B — The Two-Tier Regime: Formalised Bifurcation (Probability: ~30-35%)
The November expiries lapse without renewal in full, Beijing restores and extends the October 2025 control architecture, and the Western response — Project Vault stockpiling, DoD magnet prohibitions, price floors, friend-shoring qualification regimes — hardens into a permanent parallel market. Two price systems for the same molecule: administered domestic prices in the dominant producer, premium-bearing ex-China certification chains everywhere else. Producer states accelerate into the gap with quotas, bourses and reserves. The transition costs a recessionary impulse in Western manufacturing for two to four quarters, then stabilises — with defence and aerospace effectively detached from Chinese supply by 2028.
This scenario holds unless: a genuine trade settlement freezes the licensing architecture in exchange for tariff relief; the Chinese macro-crisis documented across our 2026 portfolio forces Beijing to court Western demand with supply, not withhold it; or producer-state quota regimes fracture under their own investment-deterrence costs before bifurcation can consolidate.
Scenario C — The Supply Shock: Cascade and Scramble (Probability: ~15-20%)
A trigger event — a Taiwan customs embargo expanding to a formal export ban, a Hormuz-Iran re-escalation compounding the acid and sulphur channels, a kinetic South China Sea incident, or a Congolese-Indonesian coordination of export cuts — converts administered scarcity into physical shortage. Spot markets for dysprosium, terbium, gallium and battery-grade graphite gap; magnet production lines idle within six to ten weeks of licence denial; strategic reserves deploy for the first time as operational rather than symbolic instruments.
The scenario has a lower individual probability than A or B but disproportionate consequences: a synchronised manufacturing contraction concentrated in exactly the defence-adjacent sectors the legislation of 2026 was written to protect. The probability is not negligible because each component risk is independently plausible and structurally interconnected. Convergence in this system is not random. It is coupled.
9. Implications
Procurement and supply chain. Complete a smelter-level traceability audit of all rare earth, gallium, germanium, graphite and battery-material inputs before 10 November 2026. Classify every input into three buckets — Chinese-sourced with no substitute inside 18 months, substitutable with requalification cost, and already diversified — and set inventory targets of at least four quarters of cover on the first bucket only, where carrying cost is justified by gap-risk rather than price-risk.
Treasury and cost of capital. Reprice supplier contracts on a dual-baseline: pre-truce and post-restriction landed costs, diverging 15-25%. Identify capex programmes with break-evens that fail at copper above $13,500 per tonne or dysprosium above $900 per kilogram ex-China, and mark them as deferrable in the annual review — not the crisis playbook.
Compliance and legal. Treat the January 2027 DoD prohibition on Chinese rare earth magnets as a flow-down certainty into prime-contractor and Tier-1 requirements within two to three quarters. Begin contractual re-papering of magnetic bill-of-material certifications now; the four-quarter re-engineering cycle means any line item not certified by mid-2027 becomes a forced redesign.
Corporate development. Screen acquisitions in the producer belt against quota architecture — RKAB allocations, cobalt quota eligibility, concentrate-export exemptions — as first-order valuation inputs, not appendices. Discount producer-state projects by the demonstrated correlation between election cycles and restriction regimes.
10. Core Analytical Judgment
The critical minerals system in 2026 is not a market undergoing a disruption. It is a jurisdiction acquiring a foreign policy. The variables are coupled in ways that defeat sequential analysis: the acid ban in Beijing sets the price of a transformer in Ohio; a quota decision in Jakarta reprices a stainless-steel margin in Bavaria; a customs officer's inspection queue in a Chinese port performs strategic effect that a fleet manoeuvre once had to deliver. Demand multiples make the leverage arithmetical; refining concentration makes it physical; producer-state imitation makes it systemic; and Western subsidy makes it durable — because every price floor dug on one side of the bifurcation guarantees that administered prices on the other side remain a going concern.
The system oscillates between Scenario A's licensed stability and Scenario B's formal bifurcation, with Scenario C as the tail risk that has quietly moved from theoretical to merely contingent. There is no stable equilibrium available in this configuration — only negotiation dates dressed as expiries, and stockpiles pretending to be strategy.
The countries that dug the metal were told for thirty years that the value was made in the factories. In 2026, the ground has called in the invoice — and it has discovered that it prefers being paid in leverage.
---
If your organisation is assessing exposure to critical minerals supply chains, quota and export-control regimes, or the investment implications of the November 2026 repricing gates, CES Intelligence maintains continuous situational awareness across these files and can provide bespoke risk assessments, scenario stress-testing, and board-level briefings. Enquiries by introduction or direct contact.
Want more analysis like this?
Free weekly digest. Full access and bespoke advisory available on request.
Thierry Marquez — Founder & Principal Advisor, CES Intelligence
+33 (0)9 55 16 54 98 →
DISCLAIMER
This analysis is provided for informational and strategic planning purposes only. It is not investment advice, financial advice, or legal advice, and it should not be treated as such. Probability assessments reflect the analyst's calibrated judgment based on available open-source intelligence as of the date of publication and are subject to revision as new information emerges. Some quantitative estimates and reported events are based on regional sourcing that may evolve as additional confirmation becomes available.
