The 13 Atoms
13 chokepoints. 8 controlled by China. $55 trillion in downstream exposure. The truce expires in November.
Ahmed Mir
Confidence:
Framework: Meta-casefile
Horizon:
Press: [email protected]
Cite this analysis →
Executive Summary
Western spot germanium trades at $5,475/kg against $2,049/kg on the Chinese domestic market — a 167% premium (Introl/Fastmarkets, April 2026) that widened from near-zero over the 32 months since China’s August 2023 export-licensing regime took effect.[1] That is not a tariff. It is a two-tier pricing regime forming in real time, structurally identical to the Urals–Brent spread that emerged after the 2022 sanctions. When persistent price bifurcation appears in a commodity, a new collateral regime is emerging — and any downstream manufacturer whose bill of materials crosses that bifurcation faces a structural cost penalty that only relocation of processing can resolve.
Thirteen materials appear as chokepoints in two or more of AI, defense, clean energy, and robotics simultaneously. Eight are predominantly controlled (>60% of processing or refining capacity) by China; five have active export controls as of April 2026.[2] Beneath this concentration sits a statutory demand floor: the annual critical minerals procurement covered by NDAA Section 854 — estimated at $2–5 billion in materials value within $350–400 billion of total allied defense procurement — is legally prohibited from Chinese-origin sourcing after January 1, 2027.[3] That compliance mandate is non-discretionary and price-inelastic — it flows into the constrained supply chain regardless of spot dynamics.
The thirteen decompose into three distinct sub-theses — rare earth permanent magnets (Western response: 5–8 years), compound semiconductor precursors (18–36 months), and industrial chemical precursors (3–7 years) — each on a structurally different timeline. The six rare earth entries are not one chokepoint counted six times; they are six stages in a physical processing chain, from ore to alloyed magnet, with distinct exposure at each.
The counter-thesis is real — substitution is advancing, recycling is scaling, and Chinese controls may be leverage rather than permanent architecture — but the compliance deadline is fixed and the qualification clock cannot be compressed. The architecture is the thesis. The monitoring is the edge.
Sub-Thesis Navigator
The thirteen materials share a common upstream driver — Chinese processing concentration and export control leverage — but diverge on Western response timeline, funding mechanism, and investable equity set. Treating them as a single undifferentiated thesis obscures three structurally distinct investment sub-theses. The table below maps the decomposition. Each sub-thesis is addressed in detail in the sections that follow; this navigator is the structural map.
| Dimension | Sub-Thesis A — RE Permanent Magnets | Sub-Thesis B — Compound Semiconductor Precursors | Sub-Thesis C — Industrial Chemical Precursors |
|---|---|---|---|
| Materials | Nd, Pr, Dy, Tb (+ 6 RE processing chain stages) | Ga, Ge, In, InP substrates | Fluorite (Tier 2 watch), tungsten/APT, tantalum, phosphorus |
| China control stage | Separation + metal/alloy (sev 4–5). Ionic clay extraction in Kachin/Wa State feeds Chinese refineries exclusively. | Refining (sev 3–4). Ga is ~80% Chinese; Ge ~60%. Export licensing regime active since August 2023. | Processing stage (sev 3–4). Varies by material; tungsten APT most exposed. |
| Western response timeline | 5–8 years (mine development + separation + sintering qualification) | 18–36 months (Ga recycling from GaN fab waste; Ge from Teck coal ash/smelter byproduct) | 3–7 years (fluorite mine development; APT processing alternatives) |
| Primary funding mechanism | DPA Title III (Defense Production Act Title III — US, Canada, UK, Australia eligible per FY2024 NDAA Section 1080); NDAA Section 854 forced-buyer demand floor | CHIPS Act supply chain resilience provisions; commercial qualification by semiconductor primes | IRA domestic content; EU CRMA strategic project designation; limited DPA Title III |
| Key monitoring trigger | GASC oxide exports <60kt; Myanmar Kachin/Wa State border status; Tb price tier breach | MOFCOM Ga/Ge license approval rates (ACREI — Association of China Rare Earth Industry / Reuters); Rotterdam spot premium vs June 2023 baseline | APT Rotterdam spot premium; fluorspar CIF Europe vs China ex-works spread (SMM) |
| Scenario matrix sensitivity | Strongest in Legs 1–2 (structural scarcity). Full thesis even under Busan renewal (absorption alone tightens). | Responsive to partial carve-out — if Ga/Ge get item-specific deal, Sub-Thesis B moves toward Legs 4–5 while Sub-Thesis A remains in Legs 1–2. | Longest tail; weakest near-term catalyst. Thesis depends on sustained price bifurcation, not policy cliff. |
| Dimension | Sub-Thesis A | Sub-Thesis B | Sub-Thesis C |
|---|---|---|---|
| Named companies by sub-thesis | MP Materials (MP), Lynas (LYC), Energy Fuels (UUUU), Serra Verde, Neo Performance Materials (NEO) | 5N Plus (VNP), AXT Inc (AXTI), Umicore (UMI) | AMG Advanced Metallurgy, Global Advanced Metals, Italmatch Chemicals (private) |
Named Companies by Sub-Thesis
Specific tickers mapped to each structural leg — rare earth magnets, compound semiconductors, and industrial chemical precursors — showing which names carry exposure to which chokepoint.
Unlock Full AnalysisFor the analytical basis of this decomposition, see the counter-thesis section on shared-atom dependency structures. The scenario matrix governs all three sub-theses simultaneously — Busan and absorption outcomes affect each cluster; the response timeline and beneficiary set diverge by cluster.
Why Now
Operational proof that controls are binding
- Germanium price divergence, April 2026: Western spot price $5,475/kg vs. Chinese domestic price $2,049/kg — a 167% premium gap that did not exist in 2022.[1] This is the operational proof that export controls are binding, not theoretical.
- Gallium exports to US/Japan near zero: Following August 2023 MOFCOM (China Ministry of Commerce) licensing requirements and January 2025 expansion to chelating resins critical for gallium recovery, exports to Western destinations have been effectively suspended.[4]
- Brazil tantalum acquisition, November 2024: China’s CNMC acquired Mineracao Taboca, which accounts for an estimated 10% of global tantalum mine production — moving a previously Western-accessible supply node behind the same licensing apparatus.[5]
- MOFCOM Announcement 18/2025 (active, April 2025): Requires export licenses for medium and heavy rare earth metals, alloys, and compounds. A broader October 2025 package (Announcements 55–58, 61, 62 — covering extraterritorial controls and technology transfer bans) is suspended through November 10, 2026 under the Busan truce. If reactivated after Busan expiry, the control regime escalates significantly.[2]
- Busan ceasefire truce expires November 10, 2026: The informal minerals dialogue facilitated at Busan has a hard expiry. If not renewed, the diplomatic backstop to Western buyers seeking exemptions disappears.
- Trump–Xi Beijing summit, May 14–15, 2026:[6] The only scheduled bilateral checkpoint before the Busan expiry. Summit tone will shift renewal probability materially — failure reprices harder than success because the October 2025 package reactivates automatically on lapse. → Conditional probability framework and monitoring tells: Appendix G
- NDAA Section 854 effective January 1, 2027: US defense contractors will be prohibited from using covered critical minerals — including gallium, germanium, tantalum, and rare earth elements — sourced from China. The DFARS (Defense Federal Acquisition Regulation Supplement) compliance deadline creates a forced-buyer dynamic for Western-sourced alternatives regardless of spot price.[3]
- MIIT production-control tightening, April 29, 2026: China’s Ministry of Industry and Information Technology proposed fines of up to 5× illegal gains for minor quota breaches, with business-license revocation for violations exceeding 30% above quota, plus inspections targeting flow-record violations. Public comment runs through end of May 2026. Major Chinese rare-earth producers closed limit-up the same session — markets read tightening as supply-supportive, not punitive. The proposal precedes the May 14–15 Trump–Xi summit and tightens the same MOFCOM-licensed control regime documented above.[25]
- MOFCOM blocking statute invoked for the first time, May 2, 2026: MOFCOM Announcement 2026 No. 21 orders all Chinese firms not to recognise, enforce, or comply with US sanctions (EO 13902 & EO 13846) targeting five teapot refineries for Iranian oil transactions. First-ever use of China’s 2021 blocking statute. Creates a compliance bifurcation structurally identical to the gallium/germanium price divergence documented above: Western firms face sanctions risk on one side, Chinese firms face domestic legal liability on the other. The same MOFCOM apparatus that controls rare earth and critical mineral exports is now being used offensively to shield sanctioned entities.[26]
An energy underlayer — LNG supply concentration and Hormuz transit dependency — compounds these material chokepoints. → Appendix H: Energy Underlayer
The Operator’s View
The rest of this casefile reads the constraint through an investor lens — price, policy, positioning. An operator reads the same constraint through a different vocabulary: lead times, qualification windows, spec rigidity, line-stoppage risk. Both lenses describe the same underlying failure mode. The operator’s version surfaces first in the cascade, often months before price reflects it.
| Failure mode | What it looks like operationally | Breakpoint threshold | Atoms where this dominates |
|---|---|---|---|
| Spec lock | Qualified supplier cannot be substituted without requalification. Requalification cycle is 9–36 months for defense-grade materials, 6–18 months for semiconductor-grade. | No drop-in replacement exists at current purity or isotope spec. | InP wafers, HALEU (High-Assay Low-Enriched Uranium), NdFeB magnets (for motor-grade applications), hi-purity phosphorus |
| Yield collapse | Off-spec material reaches the line; defect rates spike; output drops even though nominal input volume is unchanged. | Incoming purity trends out of specification; defect rate crosses contractual penalty threshold. | Gallium, germanium (both gated by Chinese export controls on higher-purity grades) |
| Lead time explosion | Contracted delivery windows expand from weeks to quarters. Backlog builds behind the gate even as price indicators look stable. | Lead time exceeds buffer inventory ÷ monthly consumption rate. | APT-derived tungsten products, dysprosium and terbium for high-temperature magnets, tantalum capacitor-grade |
| Allocation rationing | Supplier honors strategic-customer contracts first. Spot buyers and smaller OEMs are cut back or put on waitlists. | Supplier issues force-majeure or allocation notice; pricing leaves quotation and becomes negotiated per-shipment. | All rare earth oxide and metal grades under MOFCOM Announcement 18/2025 (active licensing); gallium for defense end-uses |
Operators and investors see different early indicators of the same event. Procurement teams notice lead time drift and allocation language in supplier communications 3–9 months before the price movement propagates to traded commodity benchmarks. A casefile that triangulates both views will see the cascade forming earlier than one that reads only spot prices and policy text.
For each of the thirteen atoms, the operational tell is documented in the monitoring framework below. The lead times and spec-lock windows are bucketed by material class (defense qualification, semiconductor grade, EV traction motor). Operators reading this casefile are invited to match their own bill of materials against the atom list and self-assess which failure mode is nearest.
Who Breaks First — Weakest Buyer in Each Cascade
The shared-atom map identifies chokepoints. The sub-casefiles identify who cannot substitute, cannot pay, and cannot stockpile. Four downstream cascades have been individually worked — each one surfaces a different “weakest buyer” whose operational failure is the first visible break in the chain.
Each cascade is a standalone casefile with its own hypothesis registry, monitoring framework, and tracked positions. The sub-casefile map below shows the material and policy linkages back to the shared 13-atom architecture.
The Counter-Thesis
Shared atom does not mean identical dependency
The shared-atom map identifies where sectors share a gate — it does not imply identical vulnerability. A fabless chip designer can engineer around a gallium constraint in 12–24 months; a defense prime under DFARS Section 854 cannot. Read this map as a topology, not a severity rank.
Material substitution is real and moving
For molybdenum in semiconductor metallization, Samsung and Kioxia are qualifying fluorine-free processes — if V10 NAND qualifies on Mo at volume, gallium demand from that node declines.[7] For permanent magnets, ferrite alternatives exist for low-torque applications but cannot match NdFeB power density in traction motors, wind turbines, or defense actuators. Substitution is real but bounded by requalification timelines and performance ceilings.
Recycling closes the gap faster than expected
NdFeB magnet recycling rates are currently 1–3% of total demand but could reach 30–35% by 2035.[8] If recycling scales faster than expected, the 8–12 year Western mining timeline becomes less critical. Monitor: capacity announcements from Urban Mining Company, REEtec, and Cyclic Materials.
US-China deal probability is non-trivial
A negotiated minerals framework — analogous to the Phase 1 trade deal structure — could release exemptions for specific buyers or specific materials with 6–12 months notice. The probability is not zero, particularly in an election cycle or in response to reciprocal tariff pressure. The Busan expiry date (November 10, 2026) is the monitoring trigger: if diplomatic engagement intensifies before that date, the thesis weakens. However, the scale of bilateral deal-making since 2023 — six completed agreements and a 54-country preferential trading zone announced in February 2026 — suggests Washington is building an alternative minerals architecture rather than negotiating re-entry to the Chinese-dominated one, which lowers the probability of a comprehensive reversal.[23]
Fiscal austerity compresses defense demand
NDAA Section 854 creates forced-buyer dynamics, but defense procurement timelines run 5–7 years. A defense budget under fiscal stress may stretch procurements, reducing near-term demand even while the structural dependency persists.
The energy pillar is thinner than AI and defense
The energy pillar in the 13-atom map carries lower confidence than the defense and AI pillars. The phosphorus and indium connections to solar are thinner than the magnet chain to wind and EVs. Weight the pillars unequally.
Demand destruction through motor and magnet redesign
OEMs are engineering around the constraint on two tracks: material-level (Toyota La/Ce magnets reducing Nd content ~50%,[9] GBD technology cutting Dy/Tb requirements ~50%[10]) and architecture-level (Renault, BMW, and VW qualifying RE-free motor designs[11]). Material substitution is already in production; architecture substitution lags by 5–8 years. If enough OEMs shift to RE-free motors, the NdFeB demand curve flattens and this casefile’s severity ratings are too high. → Full OEM program detail: Appendix K
What is worse than the thesis
- The counter-thesis above assumes China’s export controls are a political constraint that could be lifted by negotiation. There is a scenario in which they become irrelevant — not because they are removed, but because Chinese domestic demand absorbs the exportable surplus without them.
- The numbers: Revised estimates place Chinese domestic NdFeB consumption at 170,000–195,000 tonnes against ~250,000t production, leaving only 55,000–80,000t available for export[12] — thinner than headline production figures suggest. The NEV segment introduces the most variance. → Full wind/NEV disaggregation math: Appendix I
- The trajectory: If Chinese EV production grows 15% per year and wind installation holds at 80+ GW[12], domestic demand for NdFeB magnets could absorb the entire exportable surplus within 2–3 years — with or without export controls. In this scenario, the Busan truce is irrelevant and a US-China deal does not help: you cannot negotiate with thermodynamics.
- This is the scenario in which the structural thesis is stronger than the casefile states, because the scarcity persists regardless of geopolitics. It is also the scenario in which Western processing capacity becomes existentially necessary rather than strategically preferable. Monitor Chinese domestic NdFeB consumption data (available quarterly via Adamas Intelligence and the Chinese Rare Earth Industry Association) as the leading indicator.
- LREE vs. HREE precision: The absorption argument applies primarily to heavy rare earths (Dy, Tb) — geologically constrained to ionic clay deposits with no scalable non-Chinese alternative. Light rare earths (NdPr) face commercial, not geological, competition. The positioning framework distinguishes HREE-exposed names from LREE volume producers accordingly.
- Current baseline & falsification threshold: Chinese rare earth exports currently run at an estimated 50,000–70,000 tonnes per year (Adamas Intelligence, cross-referenced against GASC and REIA).[12] The kill number is 60,000 tonnes: if exports sustain above 60kt for two consecutive quarters, the scarcity premise weakens. Adamas quarterly reports are the preferred leading indicator given 60–90 day lag in official customs data. → Source hierarchy and HS code methodology: Appendix J
- There is a structural fragility in the bilateral deal network itself: the “poison pill” provisions in the six critical minerals agreements restrict partner countries from engaging with Chinese investors or processors. For LMICs (low and middle-income countries) like the DRC, Malaysia, and Thailand, this narrows their strategic options and may cap their willingness to scale production, creating a paradox where the policy architecture designed to diversify supply simultaneously constrains the partners it depends on.[23]
The Chokepoint Map
The map below shows the 13 atoms positioned by pillar coverage and China control stage. Atoms appearing in more than two pillars simultaneously are the highest-priority monitoring nodes — a disruption at those nodes propagates across multiple downstream themes at once. The cross-pillar chokepoints are gallium, germanium, tantalum, indium, high-purity phosphorus, indium phosphide wafers, and tungsten; the rare earth chain contributes six additional stages from ore through finished magnet.
The 13 Atoms: 8 Cross-Pillar Chokepoints + 6-Stage Rare Earth Chain
The thirteen entries include seven distinct material types plus six stages of the rare earth processing chain. The rare earth chain stages are listed separately because exposure at each stage is structurally distinct: a company mining rare earth ore but not separating it has different risk than a company refining dysprosium metal. The pillar columns use the following abbreviations: AI (artificial intelligence semiconductor and data center), D (defense), E (energy — wind, solar, EV), R (robotics).
| Atom | Pillars | China Stage Control | Export Status | Western Beneficiaries | Sub-Casefile |
|---|---|---|---|---|---|
| Gallium sev 4 | AID | ~98% refining[4] | Active (Aug 2023) | 5N Plus (VNP), Vital Materials | — |
| Germanium sev 4 | AIDE | ~80% refining[1] | Active (Aug 2023) | Umicore, Indium Corporation | — |
| Indium sev 3 | AIE | ~70% refined indium[13] | No controls yet | AXT (AXTI) — InP (indium phosphide) wafer substrate chain | — |
| InP Wafers sev 4 | AID | Substrate: AXT (US); epi: multiple | No direct controls | AXT (AXTI), Wafer Technology | — |
| Phosphorus (high-purity) sev 3 | AIE | ~85% yellow phosphorus[14] | No controls yet | Italmatch, Thermphos (limited) | — |
| Tantalum sev 3 | AIDE | Processing: DRC/Brazil now CNMC-linked[5] | Acquisition risk | AMG Advanced Metallurgy, Global Advanced Metals | — |
| Rare Earth Ore sev 3 | DER | ~60% mine output[15] | Indirect (quota) | MP Materials (MP), Lynas (LYC), Energy Fuels (UUUU) | NDAA Rare Earth Cascade |
| Mixed Rare Earth Carbonate sev 3 | DER | ~85% separation[15] | Active (Apr 2025)[2] | MP Materials, Lynas LAMP facility (Malaysia) | NDAA Rare Earth Cascade |
| RE Oxide / Separation sev 4 | DER | >90% individual oxides[15] | Active (Apr 2025)[2] | Lynas (NdPr — neodymium-praseodymium — oxide); Energy Fuels (Nd/Pr) | NDAA Rare Earth Cascade |
| Neodymium Metal sev 4 | DER | ~90% metal production[15] | Active (Apr 2025)[2] | Neo Performance Materials (NEO), Shin-Etsu Magnetics, TDK | NDAA Rare Earth Cascade |
| NdFeB Magnets sev 4 | DER | ~92% sintered magnet production[16] | Active (Apr 2025)[2] | Neo Performance (NEO), Arnold Magnetic, Shin-Etsu Magnetics, TDK | NDAA Rare Earth Cascade |
| Dysprosium sev 5 | DER | >90% mining (ionic clay); ~95% separation[2] | Active (Apr 2025)[2] | Lynas (first non-Chinese Dy oxide, May 2025); no full alternative chain yet | NDAA Rare Earth Cascade |
| Uranium / HALEU sev 3 | ED | Russia: ~44% enrichment capacity (pre-2024)[17] | Sanctions-gated | Cameco (CCJ), Centrus Energy (LEU), Uranium Energy (UEC) | — |
NdFeB = neodymium-iron-boron, the dominant commercial permanent magnet alloy. HALEU = High-Assay Low-Enriched Uranium (enriched to 5–20% U-235), required for advanced reactors and some naval propulsion systems. LAMP = Lynas Advanced Materials Plant, Malaysia.
Several of the 13 atoms are not mined directly — they are recovered as byproducts of other industrial processes. Gallium is a byproduct of alumina refining (the aluminum production chain). Germanium is a byproduct of zinc smelting. Indium is a byproduct of zinc and lead processing. This means their supply is structurally inelastic to their own price: if gallium rises from $300/kg to $3,000/kg, no alumina producer will build a new refinery to capture the gallium revenue — gallium is roughly 0.1% of the revenue from a bauxite-to-alumina operation. Supply scales only when the host commodity scales, not when the byproduct price spikes. This byproduct dynamic is one of the most important structural features of the critical minerals market and explains why price dynamics alone cannot resolve the concentration problem on any relevant timeline.
Cross-Pillar Overlap Logic
The investment relevance of the 13-atom map is not that each material is scarce — it is that the same upstream processing bottleneck simultaneously constrains demand from unrelated industrial sectors. When a gallium refinery goes offline, it does not affect one market: it constrains GaN (gallium nitride) power semiconductors for AI data centers, night-vision-device substrates for defense, and 5G base station radio frequency chips simultaneously. The demand curves do not correlate, but the supply constraint does. This is the structural shared upstream chokepoint logic.
The cross-pillar matrix below maps each atom to its downstream pillars and the specific product or application that creates the dependency. The diagonal entries — atoms that appear in all four pillars — are the highest-priority monitoring nodes.
The rare earth chain (atoms 7–12 in the table) is the clearest example: the same processing infrastructure that produces neodymium oxide for wind turbine magnets also produces neodymium for EV traction motors, defense missile guidance actuators, and robotic joint motors. The molecule is the same. The supply chain is the same. The buyers compete for the same output from the same facilities.
Japan: The Second Chokepoint Layer
China gates the raw material stage. Japan gates the next stage — converting those inputs into the precision industrial materials that fabs, defense primes, and robotics assemblers actually consume. Same architectural mechanism (qualification walls, embedded tacit knowledge, no economic incentive for incumbents to facilitate entry), applied one step downstream.
The risk archetypes differ: China’s concentration is active and adversarial; Japan’s is passive and fragile — decades of precision manufacturing investment, no hostile intent, but identical effect on supply chain resilience if a seismic or geopolitical shock hits the Tokai or Kanto corridors. The US-Japan Critical Minerals Agreement (March 2023) was the first sector-specific bilateral deal of its kind, and a subsequent 2025 Strategic Trade and Investment Agreement commits $5 billion in Japanese investment toward US critical minerals processing facilities — making Japan both a chokepoint and a direct funding source for the Western alternative supply chain.[23]
That bilateral architecture is now being tested in real time. In March 2026, China’s exports of rare earth magnets to Japan fell 17% month-on-month to approximately 184 tonnes — a nine-month low — while volumes of intermediate materials including oxides slumped by nearly 90%.[24] The asymmetry is the signal: finished magnets are substitutable over 18–36 months (Vietnam, Estonia, domestic Shin-Etsu/TDK capacity). Oxides and intermediates target the processing layer Japan cannot replace — heavy rare earth separation at commercial scale does not exist outside China. A 90% monthly collapse in contracted oxide flows is not commercial volatility; it is a policy signal dressed as trade data. The timing is not coincidental: Japan’s decision to lift its post-war arms export restrictions — opening defense equipment sales to Philippines, Australia, and its GCAP fighter partners — requires sovereign control of defense-grade materials. China is starving Japan’s capacity to build that sovereignty.
| Japanese Material | Japan Share | 13-Atom Linkage |
|---|---|---|
| GaN substrates (Sumitomo Electric) | ~60%[18] | Direct downstream consumer of gallium (atom 1); same gallium pool that Chinese export controls are already constricting |
| Silicon wafers / Shin-Etsu Handotai | ~51%[18] | Ga/Ge-doped wafer variants use 13-atom feedstocks. Shin-Etsu Chemical also operates Shin-Etsu Magnetics (leading NdFeB producer) — the group straddles both the semiconductor input layer and the rare earth magnet output layer |
| Strain wave gears (Harmonic Drive Systems) | ~50%[18] | The gear itself has no RE content — the dependency is in the servo motor driving it (NdFeB magnets). A robotic joint assembly sources strain wave gears (Japan chokepoint) and NdFeB magnets (China chokepoint) simultaneously — two independent chokepoints converge at one assembly step |
15 controlled Japanese materials analyzed in full: Japan Supply Chain Dependency Analysis →
Full Japan Supply Chain Dependency Analysis → — 15 controlled materials, domain breakdown (semiconductors, robotics, energy, defense, space, quantum), and Nankai Trough seismic risk quantification ($23.8T downstream exposure).
Cascade Math: $55T Four-Hop Reach
The graph traversal methodology works outward from each of the 13 atoms by hop count — a “hop” being one supply chain relationship step. Hop 1 captures the direct producers and immediate processors. Hop 4 reaches the final consumer companies whose revenue depends on the atom — not directly, but through the upstream chain of suppliers and manufacturers. The aggregate market capitalization of companies reachable at each hop is:
| Hop 1 | 83 companies · $12.5T aggregate market cap — direct producers and first-tier processors |
| Hop 2 | 454 companies · $48.8T aggregate — component manufacturers, alloy producers, substrate makers |
| Hop 3 | 712 companies · $54.2T aggregate — system integrators, OEM manufacturers |
| Hop 4 | 781 companies · $55.2T aggregate — end-market companies with ultimate dependency |
| Top 5 concentration | NVDA ($4.8T), GOOG ($4.0T), AAPL ($3.6T), MSFT ($2.8T), AMZN ($2.7T) — 33% of 4-hop aggregate |
Note: 324 of 781 four-hop companies are private or carry null market cap values in the current graph and are excluded from the $55.2T aggregate. That the four-hop reach covers nearly the entire technology and defense value chain is not an artifact of a loosely connected graph — it is the thesis: these 13 materials sit upstream of almost everything the modern industrial economy builds, powers, and defends. Top-5 concentration (33%) reflects the outsized role of hyperscalers and chipmakers in AI and semiconductor demand, not disproportionate exposure at the supply end.
The topology above traces BFS paths from six chokepoint materials through processing stages to end-use industries. Each branch shows how a single material constraint propagates to multiple downstream sectors simultaneously. From the November Cliff Cascade.
Policy Floor: Three Tiers
The structural demand floor created by legislation and procurement policy that requires Western-sourced alternatives regardless of market price dynamics. This floor has three observable tiers:
Tier 1 — Forced Buyer ($350–400B estimated procurement floor)[3]
US, UK, Australian, and Japanese defense procurement that, after January 1, 2027, is legally prohibited from using covered critical minerals of Chinese origin. The DFARS compliance requirement does not reduce demand — it redirects it. This tier is the most durable floor because it is statutory, not subject to market price sensitivity, and the qualification timelines for defense-grade materials run 3–7 years. The buyer cannot simply switch suppliers in a quarter.
DPA Title III eligibility (50 USC §4533): Title III funding authority extends to facilities in the US, Canada, and — as of the FY2024 NDAA Section 1080 — the UK and Australia. European and Southeast Asian allied-nation facilities remain ineligible. This creates a two-tier addressable market: domestic + Title III-eligible facilities command broader demand access than allied-nation facilities that satisfy NDAA Section 854 but cannot access Title III capital. Appropriation caveat: Title III requires annual defence appropriations, not a multi-year entitlement. → Full facility-by-facility eligibility mapping: Appendix B (Pro)
Tariff architecture confirms the thesis: 56 of 60 critical minerals exempted
Of the 60 USGS-designated critical minerals, 56 were exempted from the April 2025 reciprocal tariffs (EO 14257, Annex II), overwhelmingly covering midstream processing segments.[23] The same administration imposing the broadest tariff regime in a century explicitly carved out the materials this casefile covers. Washington treats critical mineral supply as a national security exception to trade policy — even the most aggressively protectionist trade posture in modern US history chose to protect, not tax, these supply chains.
Tier 2 — Policy-Incentivized ($2T+ over 10 years in committed programs)
IRA (Inflation Reduction Act) domestic content requirements for EV battery minerals, EU Critical Raw Materials Act (CRMA) strategic project designation for rare earth separation, and CHIPS Act supply chain resilience requirements for semiconductor-grade gallium and germanium. These create demand support through subsidy and qualification preference, not prohibition. They are softer than Tier 1 but represent large committed capital pools.[19]
Tier 3 — Commercial Qualification (Timing-Uncertain)
Hyperscaler and OEM qualification of non-Chinese gallium, germanium, and rare earth supply chains for non-defense applications. Intel, TSMC, and Samsung all have stated supply chain diversification goals but no statutory compliance deadline. This tier converts to Tier 2 or Tier 1 over time as geopolitical risk becomes balance-sheet risk. Timeline uncertain; monitoring triggers are qualification announcements in earnings calls and supplier approval notifications.
Geopolitical Landscape
Precision on the Chinese export control timeline
Chinese rare earth export controls are frequently conflated in press coverage. There is no “MOFCOM Regulation 61” from December 2023 — that designation does not exist in the official record. The actual regulatory timeline, with current operative status as of April 2026:
August 2023 — Announcement 23/2023: Gallium and germanium export licensing. Currently active.[4]
December 2023 — Announcement 57/2023: Revised technology catalogue restricting export of rare earth extraction, processing, and refining technology. Currently active. This is the action most often mislabeled as “Regulation 61.”
April 4, 2025 — Announcement 18/2025: Medium and heavy rare earth licensing — samarium, gadolinium, terbium, dysprosium, lutetium, scandium, yttrium and their oxides, alloys, and compounds. Currently active and NOT suspended. This is the operative gate on the rare earth processing chain as of this casefile’s date.[2]
October 9, 2025 — Announcements 55–58, 61, and 62: A six-announcement package adding extraterritorial reach (Announcement 61 — China’s equivalent of the US Foreign Direct Product Rule) and technology transfer bans (Announcement 62) to the existing licensing regime. Suspended November 7, 2025 through November 10, 2026 (Announcement 70/2025).[2] That suspension expires the same day as the Busan truce — a double cliff where both diplomatic and regulatory restraints lift simultaneously.
Enforcement caveat: Announcement 61’s extraterritorial provisions face material enforcement gaps — chain-of-custody tracking across fragmented downstream supply chains does not yet exist at scale. The practical effect is asymmetric: it constrains named Tier 1 magnet producers (who will self-comply) far more than the long tail of smaller assemblers. → Full enforcement analysis: Appendix L
Busan Truce Expiry — November 10, 2026
The minerals dialogue facilitated at Busan expires November 10, 2026. Neither party has publicly opened renewal talks as of the date of this casefile. If the dialogue lapses without renewal, Western buyers currently operating under informal exemption arrangements lose the diplomatic backstop that has allowed limited continued access to certain processed materials. The Busan expiry is a dated falsification trigger for the “diplomatic de-escalation” counter-thesis.
Bilateral Critical Minerals Agreements (2023–2025)
Since 2023, the United States has executed six bilateral critical minerals agreements, each operating outside WTO and FTA frameworks as sector-specific instruments:[23]
| Partner | Date | Instrument |
|---|---|---|
| Japan | Mar 2023 (CMA), Oct 2025 (framework) | Critical Minerals Agreement + Strategic Trade and Investment Agreement |
| Ukraine | Apr 2025 | Binding reconstruction agreement |
| DRC | Jul 2025 | Strategic partnership (likely binding) |
| Australia | Oct 2025 | Framework agreement |
| Malaysia | Oct 2025 | Memorandum of Understanding |
| Thailand | Oct 2025 | Memorandum of Understanding |
Three novel features distinguish these agreements from conventional trade diplomacy. First, they are coupled with parallel reciprocal trade negotiations. Second, each embeds “America First” reciprocity requiring partners to prioritize US entities through offtake rights and fiscal preferences. Third, several contain “poison pill” provisions restricting third-party participation in covered supply chains — excluding Chinese state-backed processors from partner-country mineral flows.[23]
These deals function as the diplomatic scaffolding for the Policy Floor. Tier 1 demand created by NDAA Section 854 requires physical supply from non-adversarial sources; the bilateral agreements are the mechanism designed to create it. No single deal has yet triggered material diversification of actual supply — their primary value is in communicating strategic priority and shaping the institutional environment for future procurement.[23]
Preferential Minerals Trading Zone (February 2026)
Vice President Vance announced in February 2026 a preferential minerals trading zone, inviting 54 countries to participate. The proposed zone would sustain price floors for critical minerals while excluding China and non-members.[23]
If implemented, this creates a structural price floor independent of NDAA Section 854 — operating through trade architecture rather than procurement law. Where 854 creates demand by mandating allied-source purchasing, the trading zone would create a floor by restricting market access. The two mechanisms are complementary: one pulls demand toward non-Chinese supply, the other sustains prices to make that supply economically viable.
NDAA Section 854 DFARS Effective Date — January 1, 2027
Public Law 118-31 Section 854, effective January 1, 2027, prohibits US defense contractors from procuring covered critical minerals — gallium, germanium, tantalum, tungsten, and rare earth elements — from covered entities.[3] January 1 functions as a procurement urgency accelerant rather than a hard binary cutoff: programs unable to source compliant materials by that date operate on rolling exceptions while pursuing qualification. The practical effect is that qualification processes must begin now — creating durable demand pull regardless of whether full DFARS-compliant supply is available by January 1. → Exception mechanism detail: Appendix L
The 52-Day Gap — November 10 to January 1 Derived
The Busan truce expires November 10, 2026 — the same date Announcement 70/2025 expires, meaning both the diplomatic backstop and the MOFCOM extraterritorial control suspension lapse simultaneously. The NDAA Section 854 compliance deadline falls January 1, 2027. These dates create a 52-day window in which China holds maximum leverage on two concurrent legal instruments and the United States faces a statutory mandate to procure from non-Chinese sources that do not yet exist at scale.
DFARS qualification runs 24–30 months per program (2026 campaigns resolve 2028–29). The thesis is not Western supply replacing China by Jan 2027 — it is statutory urgency forcing qualification NOW. → Phase breakdown: Appendix E. If the suspended October 2025 package (Announcements 61/62) reactivates on or shortly after the Busan expiry, the US enters January 1 with a legal obligation to buy and a constricted supply of what it must buy. This is not a supply shock in the traditional sense — it is a price discovery event for materials that have never had to clear a market without Chinese marginal pricing. The 52-day gap is a monitoring priority; absence of renewal talks by August 2026 is the early warning.
Scenario Interaction Matrix
The four thesis vectors — Busan expiry, domestic absorption, the 52-day gap, and NDAA 854 — interact. The matrix below shows which combinations produce which price and positioning regimes, ranked by probability given current data (April 2026).
| Busan outcome | Absorption on track? | Price regime | Positioning implication | Est. probability |
|---|---|---|---|---|
| Lapses + Announcements 61/62 reactivate | Yes (<60kt by end-2027) | Extreme squeeze: dual constraint (export controls + structural scarcity). No Chinese marginal price. Qualification premium dominates. | Favours Western integrated producers and DFARS-compliant processors. Consumer electronics exposed to Chinese-origin supply face headwinds. | ~15% |
| Lapses, Announcements 61/62 remain suspended | Yes | Strong squeeze: structural scarcity but no extraterritorial enforcement. Licensing controls (Announcement 18/2025) remain active. 52-day gap is real but not amplified. | Favours Western producers at moderate conviction. NDAA 854 compliance premium intact. | ~25% |
| Lapses, Announcements 61/62 remain suspended | No (>60kt in 2027) | Policy-dependent: controls create tightness, but exportable surplus persists. Western premium range: ~20–40% above Chinese domestic price (moderate backwardation). Price spikes on enforcement events, reverts on diplomatic cues. Thesis is real but fragile; no structural scarcity underpins it. | Favours Western producers with caution. Monitor absorption data quarterly. | ~20% |
| Renewed (truce extended) | Yes | Structural-only: absorption tightens supply over 2–3 years regardless of politics. Slower price ramp, durable. Molycorp risk — China can crater price if it chooses to export before absorption completes. | Favours Western processors on a 2–3 year horizon. Fundamentals-driven names over speculative positions. | ~30% |
| Renewed | No | Thesis weakest: diplomatic backstop holds, Chinese exports remain elevated. Near-term pricing pressure on Western producers. NDAA 854 demand floor provides a floor but not momentum. | Reduce exposure. Wait for absorption data to inflect or truce expiry in late 2027 cycle. | ~10% |
Probabilities are author estimates as of April 2026 based on current trajectory of Busan renewal talks (absent), absorption data (on-track), and MOFCOM enforcement posture (licenses active, extraterritorial suspended). They should be updated as monitoring triggers fire.
Sub-Casefile Linkage Map
The 13-atom map connects to a set of more granular casefiles, each investigating a specific material or cascade path in depth. The sub-casefile map below shows the relationship between this meta-casefile and its constituent analyses. Each sub-casefile has its own hypothesis registry, monitoring framework, and falsification thresholds.
Current active sub-casefiles with published hypothesis registries:
- The Tungsten Squeeze — APT (ammonium paratungstate) export freeze, three downstream industrial paths, NDAA 854 catalyst
- NDAA Rare Earth Cascade — NdFeB magnet chain from mining through sintered magnets, defense forced-buyer analysis
- The Sulfuric Acid Cascade — three simultaneous geopolitical shocks converging on a single industrial chemical
Who Benefits, Who Pays
The 13-atom map creates three structural groups: companies positioned to benefit from scarcity and the policy floor (producers), companies whose input costs rise or whose supply chains become constrained (consumers), and companies that sit between the two and can pass through price increases (intermediaries). The three groups are presented in order of investment relevance, not alphabetically.
Producers — Supply Scarcity Tailwind
These companies produce, refine, or hold Western-accessible reserves of one or more of the 13 atoms. They benefit from both the price effect of constrained Chinese supply and the policy-floor demand effect of NDAA Section 854 and allied legislation. The tailwind is structural, not cyclical: even if spot prices moderate, the qualification premium for DFARS-compliant supply persists.
Operates Mountain Pass, California — the only rare earth mining and processing facility at commercial scale outside China. Produces mixed rare earth concentrate and is actively commissioning neodymium-praseodymium oxide separation and metal production. The NDAA compliance deadline is a direct demand catalyst for DFARS-certified NdPr oxide.[20]
The only non-Chinese company operating commercial-scale rare earth separation, processing approximately 10,500 tonnes of neodymium-praseodymium oxide per year at its Malaysian LAMP facility[15] from Mt Weld xenotime-enriched ore in Australia. In May 2025 it became the first non-Chinese producer to deliver commercial dysprosium oxide — closing the single most critical gap in the Western heavy rare earth chain[15].
The only Western company simultaneously producing ultra-high-purity electronic-grade gallium (UltraGall), processed germanium, and refined indium at 3N–7N purity — touching three of the 13 atoms in a single corporate entity. Its gallium recovery partnership with Rio Tinto and long-term indium supply agreement with Teck Metals (since 2013) give it byproduct-sourced feedstock that is structurally independent of Chinese export licensing.
The only non-Chinese producer operating from ionic-adsorption clay geology — the formation type that naturally concentrates heavy rare earths including dysprosium and terbium. Serra Verde’s Pela Ema deposit in Goias, Brazil entered commercial production in 2024 following a $1.1B development investment, targeting 6,400 tonnes of total rare earth oxide per year by end 2027 — projected to represent over 50% of non-China heavy rare earth supply at that scale. USA Rare Earth (Nasdaq: USAR) is acquiring Serra Verde for $2.8B ($300M cash plus stock), with the transaction expected to close in Q3 2026. A 15-year, 100% offtake agreement with US Government entities, with guaranteed floor prices for neodymium, praseodymium, dysprosium, and terbium, underpins the commercial thesis. Ionic clay is the source geology for the atoms that determine whether a permanent magnet survives at motor operating temperatures.
The world’s largest single-site indium producer — not because indium is its business, but because it is a byproduct of the Onsan zinc smelting complex, the world’s largest zinc smelter. Korea Zinc’s indium supply is structurally de-linked from indium pricing and scales with global zinc demand, making the company an indium volume anchor even when indium markets are constrained.
The world’s largest publicly traded uranium producer, operating McArthur River and Cigar Lake in the Athabasca Basin. The Russian uranium import ban creates a structural re-sourcing event that runs through Cameco’s contracted book; the large majority of its sales are under long-term supply agreements, insulating realized prices from spot volatility while the Western enrichment chain rebuilds.
Producer Profiles
The remaining producers across all 13 atoms, each with chokepoint role, atom-level exposure, and Western qualification status.
Unlock Full AnalysisConsumers — Input Cost Headwind
These companies depend on one or more of the 13 atoms in their manufacturing process and cannot easily substitute or pass through cost increases within the 12–24 month thesis window. The headwind is most acute where the company (1) cannot qualify alternative sources quickly, (2) sells into competitive markets where cost pass-through is limited, and (3) is not itself a defense contractor eligible for DFARS-compliant supply preference.
US defense prime contractors are simultaneously the most structurally exposed (they cannot use Chinese-origin covered minerals after January 1, 2027) and the most protected (cost-plus contracts allow pass-through, and DFARS-compliant supply premiums are reimbursable). The exposure is transitional: during the qualification period before DFARS-compliant alternatives scale, lead times extend and program delivery schedules are at risk. The concern is delivery reliability, not margin compression.
The least-discussed atom exposure in the semiconductor chain: TSMC’s doping process depends on high-purity phosphorus, and approximately 85% of global yellow phosphorus comes from China. Unlike gallium or germanium — which have attracted intense policy attention — phosphorus has not yet been added to China’s export control list, meaning the market has not priced this dependency. If it joins the list, the exposure is at the foundry layer that fabricates every advanced chip in the world.
Produces tungsten hexafluoride (WF6), the deposition gas used in every NAND flash memory fab worldwide — and its feedstock is Chinese tungsten metal powder. Not a miner, not an equipment maker, but the single-step chemical conversion that sits between Chinese raw metal and the global memory supply chain. The company has disclosed it is preparing production cuts in the second half of 2026 due to supply constraints from February 2025 Chinese export controls.
Dual-atom exposure across the two most sensitive defense applications: tungsten for kinetic energy penetrators (armor-piercing rounds) and NdFeB magnets for electric drive motors in military vehicles. Both atoms are China-gated. Tungsten and rare earth magnets are core to Rheinmetall’s products, not commodity inputs to a diversified business. European rearmament is the demand tailwind, but atom availability is the production ceiling.
The world’s largest motor manufacturer, with every major product line — EV traction motors, industrial robot joints, hard drives — built around NdFeB permanent magnets. The demand destruction bear case in this casefile (externally excited synchronous motors eliminating NdFeB) is existential specifically for Nidec’s magnet-motor business model. Nidec is the right proxy for monitoring how fast that substitution actually develops.
Consumer Profiles
Companies with hard dependencies on one or more of the 13 atoms — specific input exposure, substitution difficulty, and lead-time risk.
Unlock Full AnalysisIntermediaries — Pricing Power Rate-of-Change
These companies sit between the raw material and the end product: specialty chemical distributors, gas suppliers, and component traders. Their exposure is time-bounded — they benefit when the spread between constrained input prices and negotiated customer prices widens, but the benefit reverses if Western supply chains develop faster than expected.
Specialty gas and chemical intermediaries for gallium compounds, germanium tetrachloride, and phosphorus trichloride supply chains. Their pricing power is a function of scarcity duration: in a 12–24 month constraint window, long-term contracts with Western producers command premium pricing. If the constraint resolves faster, the pricing window closes.
One of the only non-Chinese germanium processors and recyclers, recovering germanium from end-of-life infrared optics and fiber optic scrap. The directional read is arithmetic: the current bifurcation between Chinese domestic germanium pricing and the Western price directly expands the margin on Umicore’s recycled output. Recycled germanium is priced off the Western market, not the Chinese one — the wider the bifurcation holds, the more valuable feedstock-independent supply becomes.
The second WF6 producer alongside Kanto Denka — meaning the deposition gas that every NAND memory fab depends on runs through a two-company intermediary layer, both Japanese, both sourcing Chinese tungsten powder disrupted by the February 2025 export controls. The two-company structure with shared feedstock concentration is the chokepoint geometry the graph makes legible that company-level analysis misses.
Operates the Silmet separation facility in Estonia — one of only two non-Chinese rare earth separators — and produces Magnequench bonded NdFeB magnet powder, straddling the producer/intermediary divide. The graph identifies the feedstock risk that the headline narrative misses: Neo processes Chinese rare earth intermediates, meaning it inherits upstream vulnerability even as it provides Western-qualified downstream output. It is a bridge node — the supply chain runs through it, not around it. The March 2026 data point sharpens both sides of that position: China’s near-90% cut in oxide and intermediate exports to Japan demonstrates that the feedstock risk is active and directional, while simultaneously elevating Neo’s strategic value as the only Western-allied separator not dependent on Japanese processing infrastructure.[24]
Intermediary Profiles
The distributors and component suppliers sitting at the narrowest points between constrained inputs and end markets — where pricing power concentrates.
Unlock Full AnalysisHistorical Analog Warning
The Rare Earth 2010–2011 analog — read this before sizing any position
- The setup was structurally identical: China reduced rare earth export quotas in 2010, prices spiked 3–10x, Western producers attracted enormous capital. Molycorp went from ~$14 to $74 per share, then filed for bankruptcy in 2015.[21] Equity went to zero while the structural thesis remained partially valid.
- The 2026 difference: NDAA Section 854 creates a statutory demand floor that did not exist in 2011.[3] EU CRMA provides capital support.[19] The policy architecture is more durable. But the equity pattern of Phase 2 — narrative-driven re-rating outpacing fundamentals — remains a risk.
- What the NDAA floor does not protect: Section 854 guarantees Western demand but not Western price floors. If China floods the market as it did in 2012–2013, producers face the Molycorp problem: technically operational, economically unviable.
- Processing gap is closing but not closed: MP Materials produced 2,599t separated NdPr oxide (FY2025, 101% YoY increase) and first NdFeB magnets at Fort Worth (Q4 2025).[20] Western separation is no longer vaporware but not yet at replacement scale. → Full 4-phase narrative and comparison table: Appendix F
Positioning Framework
The scenario matrix above maps five distinct price regimes. Each regime has a different set of names that benefit, a different time horizon for the thesis to resolve, and different monitoring indicators that tell you which regime is unfolding. The positioning framework below maps those connections explicitly.
Scenario Leg 1 — Extreme Squeeze (Busan lapses + Announcements 61/62 reactivate + absorption on track)
- Primary names: MP Materials (MP), Lynas (LYC) — NdPr producers with DFARS-compliant qualification. Energy Fuels (UUUU) and Serra Verde — the primary Western-accessible Dy/Tb plays; Energy Fuels producing from monazite (emerging scale, NDAA-qualified trajectory), Serra Verde the only ionic clay deposit with Western-listed equity at commercial development stage. 5N Plus (VNP) — multi-atom refiner (Ga, Ge, In) insulated from Chinese licensing.
- Time horizon: 6–18 months from Busan expiry (November 2026) for initial price discovery event; 18–36 months for equity re-rating to mature.
- Add: Absence of Busan renewal talks by August 2026 + GASC monthly export data showing sub-60kt trajectory. Both conditions must be present. Observable definitions, MOFCOM quota monitoring notes, and communiqué language guidance: Appendix G.
- Reduce: Any confirmed bilateral minerals framework announcement at or after the Trump-Xi summit (May 2026) that extends a suspension window. Historical precedent favors item-specific carve-outs over blanket outcomes — a partial deal would require a split read across scenario legs. Monitor for item-specific language in any MOFCOM announcement.
Full Positioning Framework
All 5 scenario legs with entry criteria, exit conditions, and time horizons — plus the HREE edge analysis and Molycorp-informed sizing discipline.
Unlock Full AnalysisScenario Leg 2 — Strong Squeeze (Busan lapses + Announcements 61/62 remain suspended + absorption on track)
- Primary names: Same producer set as Leg 1 but at lower conviction. NdPr premium narrows (no extraterritorial enforcement). Focus on companies whose revenue is most directly tied to DFARS-compliant qualification premiums rather than spot commodity price.
- Time horizon: 12–24 months post-Busan for premium to embed in contract prices.
- Add trigger: NDAA FY2027 conference report expanding Section 854 scope. Lynas Dy oxide production confirmation above 50t/yr.
- Reduce trigger: MP Materials spot NdPr realization falling below $60/kg for two consecutive quarters (indicates commodity pricing regime, not qualification premium).
Scenario Leg 3 — Policy-Dependent (Busan lapses + no structural absorption)
- Primary names: Reduce speculative positions. Retain only companies with direct DFARS demand floors (MP, Lynas at reduced size). Avoid junior miners entirely — this is the Molycorp regime: commodity prices vulnerable to Chinese export normalization at any time.
- Time horizon: Thesis resolves within 6–12 months of first GASC data showing exports persistently above 60kt through 2027. Exit before the data confirms; position for it.
- Add trigger: None — this regime is not a high-conviction entry point. It is a monitoring-and-wait regime.
- Exit trigger: Two consecutive quarters of Chinese rare earth exports above 65kt. This is the falsification of the absorption thesis.
Scenario Legs 4 & 5 — Busan Renewed
- Leg 4 (renewed + absorption on track): Long Western processors on a 24–36 month horizon. Reduce event-driven speculative positioning. Increase in names with recurring DFARS contract revenue and low project-risk profiles. Absorption still tightens supply regardless of Busan — this is a slower, more durable thesis. Molycorp risk is lower here because the price ramp is gradual, not shock-driven.
- Leg 5 (renewed + no absorption): Full reduction. NDAA Section 854 demand floor provides a floor on qualified names but not momentum. Wait for a new entry point: either Busan expiry in a future renewal cycle, or absorption data that inflects toward the 60kt threshold.
- Time horizon: Leg 4 — 24–36 months. Leg 5 — wait for cycle reset.
Sizing discipline — the correlation caveat
- The scenario matrix assigns independent probabilities to five outcomes. In a Taiwan contingency, the China and Japan risk factors fire together — moving from any of rows 1–3 directly to a correlated extreme outcome not captured in the matrix. Any portfolio sizing that treats the five scenarios as independent overstates diversification. Sizing should assume, at minimum, a 10–15% tail allocation for the correlated-shock outcome.
- The Molycorp warning, restated: This is a Phase 2 equity market setup. The structural argument is valid. The risk is that equity prices run to Phase 3 before Western supply chains are genuinely operational. Proof points to watch:
- MP Materials Fort Worth — H2 2026 commercial magnet revenue start is the first testable proof point; Northlake expansion (~10,000 MT/yr) targeted ~2028[20]
- Lynas LAMP throughput vs. nameplate (quarterly)
- Energy Fuels Dy/Tb production confirmation (semi-annual)
The structural edge: heavy rare earths over light
The rare earth thesis is not monolithic. NdPr (neodymium-praseodymium, the “light” rare earths used in magnet body composition) is a capacity buildout story — MP Materials and Lynas are scaling Western separation, and the timeline is measurable. Dy and Tb (dysprosium and terbium, the “heavy” rare earths that determine whether a permanent magnet survives at motor operating temperatures) are a geological hostage situation.
- Why HREE is structurally different: Global Dy/Tb production depends overwhelmingly on ionic clay deposits in Myanmar’s Kachin and Wa States (~57% of global supply).[12]. These deposits feed exclusively into Chinese separation facilities in Yunnan. There is no non-Chinese ionic clay source at commercial scale, no proven Western heavy rare earth separation pathway independent of Chinese feedstock, and no substitute element that provides equivalent high-temperature coercivity in sintered NdFeB magnets. The constraint is geological and metallurgical, not political — diplomacy can ease export licensing but cannot create deposits that do not exist.
- The 2026–2027 window: Western NdPr separation is scaling (MP produced 2,599t in FY2025). Western Dy/Tb separation is not. Energy Fuels is producing heavy rare earth concentrate from monazite at White Mesa, but volumes remain early-stage. Serra Verde holds the only ionic clay deposit with Western-listed equity at commercial development stage. No amount of NDAA Section 854 compliance urgency changes the fact that the HREE supply chain has fewer alternative paths than LREE.
- Implication for the thesis: In every scenario leg of the matrix, the HREE constraint binds tighter and longer than the LREE constraint. NdPr pricing is a spread trade against Chinese capacity decisions. Dy/Tb pricing reflects a geological scarcity that persists regardless of Busan renewal, export licensing changes, or Chinese production decisions — because the bottleneck is the deposit, not the policy.
Monitoring Dashboard — Positioning Triggers
The trigger conditions that indicate which regime is unfolding and when each scenario leg warrants adjustment.
Unlock Full AnalysisCoverage Gaps
Known gaps in the current model affect confidence levels and are documented below.
| Gap | Impact | Status |
|---|---|---|
| China-Japan risk correlation in Taiwan contingency | China and Japan supply chain risks are treated as independent in the current matrix, but a Taiwan contingency would correlate them simultaneously — China via export controls and Japan via alliance obligations and shipping lane disruption. This underweights extreme-outcome probability by an estimated 5–10%. → Correlated shock annotation and escalation index methodology: Appendix D | Partially modeled — correlated shock annotation added; full scenario matrix row deferred to Phase 2 |
| Mid-stream refining layer thin | The mid-stream rare earth conversion layer (oxide-to-metal, metal-to-alloy, alloy-to-magnet) is underweighted in the graph. Five Western-allied facilities have been identified, but throughput constraints and upstream feedstock dependencies are not yet mapped as graph edges. → Facility eligibility: Appendix B (Pro); ITAR/EAR compliance: Appendix C | Partially addressed — nodes identified, edges not mapped |
| Labor not modeled | The mining and processing chain for rare earths and gallium is labor-intensive at the artisanal end (Myanmar heavy rare earth ore, DRC coltan). Labor disruptions and militia-controlled supply chains are not captured as graph edges. | Phase 2 build |
| Water dependency narrow | Rare earth separation is water-intensive. Mountain Pass recycles process water but Jiangxi ionic clay extraction has significant water runoff impact. The graph does not model water stress as a constraint on production throughput. | Not modeled |
| DPA Title III appropriation conditionality | Title III requires annual Congressional appropriations — it is not a multi-year entitlement. If appropriations lapse or are reduced, facility-level eligibility assumptions change. → Eligibility mapping by facility: Appendix B (Pro) | Partially addressed — funding mechanism and eligibility now mapped; appropriation risk not modelled | Financial plumbing sparse | Trade finance, letters of credit, and correspondent banking constraints on sanctioned supply chains are partially modeled (6 financing nodes) but the coverage is incomplete. A financing disruption can be as effective as a physical export control. | Partial (6 nodes) |
| MOFCOM regulatory timeline | Resolved. Previously conflated MOFCOM timeline now correctly scoped to three separate graph nodes (Announcements 18, 61, 62). | Resolved |
| Policy node coverage | Resolved. All 27 policy nodes populated with verified names, dates, severities, and source URLs. | Resolved |
| Rival minerals bloc formation | The US bilateral deal architecture (6 agreements + 54-country preferential trading zone) may trigger a Chinese counter-architecture of mineral partnerships across Asia, Africa, and Latin America.[23] If rival blocs formalize, the 13-atom supply chains bifurcate permanently rather than temporarily — strengthening the structural thesis but raising the risk that Western producers face permanently smaller addressable markets. | Not modeled |
Monitoring Framework
Monitoring is organized by trigger type. Escalation triggers strengthen the structural thesis. De-escalation triggers weaken it. Each entry includes a named source and cadence.
Escalation Triggers
| Indicator | Source | Cadence | What to Watch |
|---|---|---|---|
| Myanmar heavy rare earth exports (Tb/Dy) | Adamas Intelligence quarterly tracker; Heinrich Böll Stiftung / JOGMEC annual reports; Myanmar Mines Ministry announcements | Quarterly | Myanmar (Kachin + Wa States) supplies ~57% of global Dy/Tb (Adamas 2023). Two control zones with distinct risk drivers: Kachin (KIA, active conflict) and Wa State (UWSA, border-revenue dependent). Both zones feed border crossings into Yunnan. Tb spot price tiers: $2,000/kg early warning, $3,000/kg full disruption (currently ~$1,400/kg). → See Appendix A: Myanmar Production Zone Detail |
| MOFCOM press conferences | mofcom.gov.cn (in Chinese) | Weekly (Thursdays) | Any statements on rare earth export licensing, new material categories added |
| Chinese customs critical mineral export data | SMM Metal Intelligence / metal.com | Monthly (mid-month) | Gallium, germanium, RE oxide export volumes; any further reduction indicates escalation |
| Samsung V10 NAND molybdenum qualification | Samsung IR; The Elec (Korean tech press) | Quarterly | If V10 qualifies on Mo, the semiconductor gallium demand leg weakens materially |
| Busan truce renewal negotiations | Trade press / diplomatic cables | Leading to Nov 10, 2026 | Any public indication of renewal talks; absence of talks by August 2026 is a warning |
| Trump–Xi Beijing summit readout[6] | White House / Xinhua joint statements; USTR post-summit briefing | May 14–15, 2026 | Tone and deliverables on trade stability, technology controls, and minerals framework. A constructive readout increases Busan renewal probability; visible tension on Taiwan or Entity List decreases it. Watch for any mention of a bilateral minerals working group or ministerial follow-up mechanism |
| NDAA FY2027 conference report | congress.gov | As published (est. Dec 2026) | Whether Section 854 scope is expanded to additional materials or contractors |
| Lynas separation volumes (NdPr, Dy/Tb) | Lynas quarterly reports | Quarterly | LAMP throughput vs. nameplate; Dy oxide production confirmation |
| ITIA annual rare earth supply report | ITIA (International Tungsten Industry Association) — separate RE annual report | Annual (Q1) | Recycling rate trend; NdFeB magnet demand by end market |
| EU CRMA strategic project approvals | European Commission | As announced | Approved project list expansion; which rare earth separation projects receive designation |
| Announcement 61 reactivation (extraterritorial controls) | MOFCOM; CSET Georgetown; Pillsbury; company filings | Post-Nov 10, 2026 | If reactivated, the 0.1% value threshold on foreign-manufactured goods containing Chinese-origin rare earths creates extraterritorial jurisdiction over leading non-Chinese magnet producers (Shin-Etsu, TDK, Vacuumschmelze) and semiconductor equipment makers (ASML) — potentially subjecting the same equipment to contradictory US and Chinese export control regimes simultaneously. Note: Vacuumschmelze (private, PE-owned) is named in industry analyses but is not publicly listed; focus monitoring on listed producers (Shin-Etsu Magnetics via parent Shin-Etsu Chemical, TDK, VAC Group filings where available) |
| Bilateral minerals deal accessions | USTR; US Department of State; partner country trade ministries | As announced | New country accessions to bilateral minerals agreements; formalization of the 54-country preferential trading zone (announced February 2026); enforcement actions under third-party exclusion provisions. Expansion beyond 6 current deals or zone formalization shifts the Policy Floor from statutory (Tier 1) to structural (trade architecture).[23] |
| China rare earth exports to Japan (magnets + oxides) | China customs administration (monthly); Bloomberg; SMM Metal Intelligence | Monthly (mid-month) | In March 2026, finished magnet volumes fell 17% MoM to ~184 tonnes (nine-month low); oxide and intermediate material volumes fell ~90% MoM.[24] The oxide cut is the signal: contracted intermediate flows do not collapse 90% on commercial grounds. Watch whether this persists into Q2 2026 — a sustained cut confirms China is targeting Japan’s processing layer, not just finished goods. Reflexive trigger: Japan’s defense export liberalization (arms sales to Philippines, Australia, GCAP fighter partners) removes the passive-fragile framing — China’s chokepoint posture is now directly responsive to Japan’s security posture. A sustained squeeze accelerates Japanese JOGMEC offtake expansions with Lynas and MP Materials. |
De-escalation Triggers
| Indicator | Source | Cadence | What to Watch |
|---|---|---|---|
| MOFCOM export license approvals (any Western buyer) | Industry sources / company announcements | Monthly | Any confirmed license issuance to US/EU buyers indicates de facto ban softening |
| New Western rare earth separation commissioning | Company announcements (MP, Lynas, Energy Fuels) | Quarterly | Capacity exceeding 5,000 tonnes NdPr oxide per year outside China would materially shift the thesis |
| NdFeB magnet recycling capacity announcements | Urban Mining Co., REEtec, Cyclic Materials press releases | As announced | Recycling capacity announcements above 1,000 tonnes per year represent a structural demand offset |
| US-China minerals framework negotiations | USTR / State Department announcements | As announced | Any announced bilateral minerals working group or Phase 2 trade deal framework |
| Gallium or germanium price convergence | SMM Metal Intelligence; Fastmarkets Rotterdam spot assessments | Weekly | Western spot converging within 30% of Chinese domestic price indicates supply normalization. Current premiums remain elevated (gallium ~150% above June 2023 baseline,[4] germanium ~56%[1]) despite partial US suspension through November 2026. Monitor Rotterdam premium compression vs. Chinese domestic price as the de-escalation tell. |
This casefile is tied to a cryptographic graph snapshot. The current snapshot is —, taken —. It captures the graph state at that moment — not the casefile prose, which updates as new evidence arrives. Graph snapshot at time of publication: 365e63ac0de9 — 2,901 nodes, 10,447 edges, 1,109 companies. Use the share button above to generate a pinned link with a ?v= parameter; anyone opening that link will see a banner identifying the exact graph state you shared.
Appendix
Detailed reference material for sections condensed in the main body. Each appendix is self-contained and sourced independently.
Appendix A: Myanmar Production Zone Detail
Myanmar's ionic clay deposits supply approximately 57% of global dysprosium and terbium output.[A1] The Heinrich Böll Stiftung characterizes Myanmar as providing "almost two-thirds" of global heavy rare earth supply.[A2] Production concentrates in two politically distinct zones, each with separate risk profiles.
| Factor | Kachin State | Wa State |
|---|---|---|
| Key Districts | Chipwi, Pangwa | Mong Yun, Mong Pawk, Mong Hsat |
| Controlling Authority | Kachin Independence Army (KIA) | United Wa State Army (UWSA) |
| Security Status | Active conflict since Oct 2024 | Stable, but border-revenue dependent |
| Yunnan Crossings | Northern corridor | Southern checkpoints |
| Primary Risk | KIA territorial control disrupts mine access and transport routes to Chinese processors | UWSA fiscal model depends on cross-border trade revenue; any Chinese border tightening compresses operating margins |
Current terbium oxide price: approximately $1,400/kg (3-year range: $900 – $1,800/kg).
| Tier | Price Level | Above Current | Trigger |
|---|---|---|---|
| Early Warning | $2,000/kg | ~43% | Sustained disruption in one zone, partial flow recovery via alternate routes |
| Full Disruption | $3,000/kg | ~114% | Simultaneous disruption in both zones or Chinese border closure |
Monitoring Sources
| Source | Coverage | Frequency |
|---|---|---|
| Adamas Intelligence | Rare earth production volumes, pricing, and trade flows | Quarterly tracker |
| The Irrawaddy / Myanmar Now | KIA-Tatmadaw ceasefire status, territorial changes | Ongoing reporting |
| UWSA border checkpoint reports | Cross-border trade activity, checkpoint closures | As available |
[A1] Adamas Intelligence, Rare Earth Market Outlook Q4 2023.
[A2] Heinrich Böll Stiftung, "Myanmar's Rare Earths and the Global Green Transition," 2025.
Appendix C: ITAR/EAR Compliance Path for LCM Birkenhead Alloy
LCM Birkenhead (UK), acquired by USA Rare Earth for approximately $100M in 2025, produces strip-cast NdFeB alloys for permanent magnet applications. As a UK-based facility, exports into US defence supply chains must navigate one of three regulatory frameworks.[C1]
| Path | Regime | Mechanism | Timeline |
|---|---|---|---|
| 1. EAR | Commerce Control List (CCL) | License Exception STA, 15 CFR 740.20; UK qualifies as Country Group A:5 | Days (self-classification) |
| 2. ITAR | US Munitions List (USML) | DSP-5 permanent export licence; DSP-61 temporary import authorisation | 30–90 days processing |
| 3. Treaty | US-UK Defence Trade Treaty (signed June 2007, force April 2012) | Licence-free transfer for approved articles between approved communities | Days (pre-approved)[C2] |
Congressional notification thresholds
Under AECA (Arms Export Control Act) Section 36(b) (22 USC 2776), congressional notification is triggered at $25M for major defence equipment and $100M for defence articles or services to NATO allies.[C3]
NDAA Section 854 compliance
The UK is not a covered country under Section 854. LCM Birkenhead alloy is compliant with sourcing requirements by default, provided feedstock traceability excludes covered-country origin.[C4]
Structural premium implication. Regardless of path, ITAR/EAR compliance adds 1–3 months of administrative lead time compared to a domestic US source. This delay premium is structural and cannot be shortened. It is the primary reason domestic rare earth producers command a qualification premium over foreign-sourced equivalents.
[C1] USA Rare Earth LLC, "Acquisition of Less Common Metals," 2025.
[C2] US-UK Defence Trade Cooperation Treaty, Treaty Doc. 110-7.
[C3] Arms Export Control Act, Section 36(b), 22 USC 2776.
[C4] NDAA FY2024, Section 854.
Appendix D: Taiwan Escalation Index — Full Indicator Methodology
The Taiwan escalation index tracks three publicly available indicators that, in combination, have preceded every major cross-strait military escalation since 2022. All three are fully replicable from open sources.
Indicator A: PLA Exercise Footprint
Source: Taiwan Ministry of National Defence daily releases (@MoNDefense on X); Focus Taiwan (focustaiwan.tw).[D1]
| Year | Approx. Crossings | Key Events |
|---|---|---|
| 2021 | 953 | Baseline year; ADIZ reporting formalised |
| 2022 | ~1,727 | Pelosi visit (August); sustained elevated tempo |
| 2023 | ~1,709 | Plateau at elevated baseline |
| 2024 | 3,070 | Joint Sword-2024A (May), Joint Sword-2024B (October)[D2] |
Trigger: Sustained step-change above the 90-day rolling rate, maintained for a minimum of 14 consecutive days.
Indicator B: US Arms Sale Notifications
Source: Defense Security Cooperation Agency (DSCA), dsca.mil. Published same-day, 30-day Congressional review.[D3]
Recent baseline: December 2025 — 8 packages totalling ~$11.1B.[D4]
Limitation: Lagging confirmation of policy direction, not a leading indicator. Confirms escalation posture but does not predict it.
Indicator C: Presidential Framing Shifts
Source: Official transcripts at president.gov.tw; Global Taiwan Institute tracks Beijing reaction cadence.[D5]
| Marker | Weight | Description | Precedent |
|---|---|---|---|
| i. Mutual non-subordination | Highest | Explicit statement that ROC and PRC are “not subordinate to each other” | Beijing cited as trigger for Joint Sword-2024A and 2024B[D6] |
| ii. Framework omission | High (negative) | Deliberate absence of cross-strait framework language in major addresses | Lai May 2024 inaugural — first since Lee era to omit (CFR analysis)[D7] |
| iii. Active-defensive framing | Moderate | Shift from passive (“maintaining status quo”) to active (“will absolutely not back down”) | Tsai Aug 2022, post-Pelosi visit[D8] |
Not markers: Generic sovereignty language, “democracy under threat” framing, routine diplomatic boilerplate. These carry no discriminatory value.
Probability rule
When 2 of 3 indicators are active simultaneously: +5–10 percentage points toward scenario Legs 1–3. When all 3 are active with kinetic engagement: full strait closure probability applies.
[D1] Taiwan MND daily PLA activity reports; Focus Taiwan annual compilations.
[D2] Focus Taiwan, "Record 3,070 PLA crossings in 2024," January 2025.
[D3] DSCA, dsca.mil/press-media/major-arms-sales; AECA Section 36(b).
[D4] DSCA press releases, December 2025.
[D5] Office of the President ROC; Global Taiwan Institute.
[D6] PRC MND statements following Joint Sword-2024A (May 2024) and 2024B (October 2024).
[D7] Council on Foreign Relations, Lai Ching-te inaugural analysis, May 2024.
[D8] Office of the President ROC, Tsai Ing-wen statement, August 2022.
Appendix E: DFARS Qualification Phase Breakdown
DFARS-compliant qualification for a new NdFeB magnet supplier runs 24 to 30 months per program, with each phase formally gated before the next can begin.
| Phase | Stage | Timeline | Scope |
|---|---|---|---|
| 1 | Supplier Audit | Months 1–3 | DCMA pre-award survey, facility security clearance (NISP), AS9100 Rev D quality management certification |
| 2 | Magnetic Property Testing | Months 3–9 | BH curve characterization per MIL-STD, temperature coefficient (−40°C to +150°C), salt spray corrosion per ASTM B117 |
| 3 | First-Article Inspection | Months 9–14 | Production-representative samples, dimensional and magnetic verification, full material traceability documentation |
| 4 | Design Verification | Months 14–20 | Integration into program assembly, MIL-STD-810H environmental testing (vibration, thermal shock, humidity), EMC testing |
| 5 | Production Validation | Months 20–26 | Pilot production lot, statistical process control demonstration (Cpk ≥ 1.33), yield confirmation at production rates |
| 6 | AVL Addition | Months 24–30 | Formal Approved Vendor List addition, contract modification, Engineering Change Proposal approval, initial production order (60–90 days after AVL) |
Qualification campaigns initiated in 2026 resolve in 2028–2029. The statutory urgency created by NDAA Section 854 forces qualification to begin now — waiting pushes first deliveries past compliance deadlines. Every month of delay is a month added to the far end.
The conversion of strip-cast alloy to sinterable powder requires hydrogen decrepitation (HD) and jet milling under inert atmosphere. This is the least documented Western mid-stream step. Commercial-scale HD capability outside integrated Chinese facilities has not been publicly confirmed. Until this step is demonstrated, the Western chain has a qualification-blocking gap between oxide separation and finished magnet sintering.[E7]
[E1] DFARS 252.225-7052 and DoD Instruction 5000.02.
[E2] DCMA Manual 2301-01; SAE AS9100D:2016.
[E3] MIL-STD-810H (Jan 2019); MIL-STD-461G.
[E4] ASTM B117-19.
[E5] AS9103, Variation Management of Key Characteristics.
[E6] DLA QPL procedures; DFARS 242.302.
[E7] Based on public disclosures from MP Materials (10-K FY2025), Lynas (annual reports), USA Rare Earth (press releases Q1 2026).
Appendix B: DPA Title III Facility Eligibility Table
The Defense Production Act Title III (50 USC §4533) is the primary United States government funding authority for critical minerals processing and production capacity.
Baseline: United States and Canada (bilateral agreement, 1956).
FY2024 NDAA Section 1080 extension: United Kingdom and Australia now eligible.
Not eligible: France and other European allies remain outside the statutory scope.
| Facility | Location | Title III Eligible | NDAA Section 854 Compliant |
|---|---|---|---|
| MP Materials — Mountain Pass | CA, USA | Yes | Yes |
| MP Materials — Fort Worth | TX, USA | Yes | Yes |
| Noveon Magnetics | TX, USA | Yes | Yes |
| Electron Energy Corporation | PA, USA | Yes | Yes |
| Energy Fuels — White Mesa | UT, USA | Yes | Yes |
| LCM Birkenhead | Merseyside, UK | Yes (via Section 1080) | Yes |
| Lynas — Kalgoorlie | WA, Australia | Yes (via Section 1080) | Yes |
| Lynas — LAMP | Kuantan, Malaysia | No | Yes |
| Solvay — La Rochelle | La Rochelle, France | No | Yes |
MP Materials' Fort Worth facility commands the broadest addressable market among Title III-eligible sites: domestic US siting, full Title III eligibility without statutory extension, and NDAA Section 854 compliance.
Title III funding is subject to annual defence appropriations, not a multi-year entitlement. A change in defence spending priorities can delay or reduce disbursements regardless of eligibility.
[B1] 50 USC §4533, Defense Production Act of 1950, Title III.
[B2] US-Canada bilateral defence production sharing agreement, 1956.
[B3] National Defense Authorization Act for Fiscal Year 2024, Section 1080.
Appendix F: The Molycorp Analog — Full Narrative
The 2010–2015 rare earth cycle is the closest analog to the current NdFeB structural thesis. It ended in bankruptcy. Understanding exactly why — and what has changed — is the most important risk assessment for any position in this space.
Phase 1: Correct Diagnosis (2009–2010)
China cut rare earth export quotas by roughly 40%. Neodymium oxide rose from approximately $19/kg to $80/kg.[F1] The diagnosis was correct: Western industry had ceded critical material supply to a single sovereign producer. The directional read was right. The problem was not the thesis. It was what came next.
Phase 2: Narrative Overshoot (2011)
Molycorp ran from $14 to $74. Neodymium oxide peaked at approximately $340/kg — an 18x move.[F2] Junior miners IPO'd. The prevailing narrative: “The West will build its own supply chain.” The warning: equity re-rating was outpacing industrial reality by years. The stock priced in a chain that did not yet exist.
Phase 3: The Collapse (2012–2015)
China expanded output. Neodymium oxide collapsed to roughly $40/kg.[F3] Molycorp filed Chapter 11 in June 2015. Equity went to zero.[F4] The WTO ruled against China in 2014.[F5] But: Mountain Pass was still running when Molycorp went bankrupt. The mine operated. The problem was never geology. It was margin.
Section 854 guarantees demand, not price floors. If China floods the global market with below-cost NdFeB, the statutory mandate holds — but commercial prices collapse. The demand floor prevents a zero-revenue outcome. It does not prevent a negative-margin outcome. Mountain Pass was producing ore when Molycorp filed for bankruptcy. Revenue existed. Margin did not.[F6]
What has changed between 2011 and 2026:
| Factor | 2011 | 2026 |
|---|---|---|
| Statutory demand floor | None. Demand was purely commercial and price-elastic. | NDAA Section 854 prohibits non-allied magnet sourcing for defense programs from 2027.[F6] |
| Western separation capacity | Effectively zero. Molycorp building from scratch. | MP Materials: 2,599t NdPr in FY2025, +101% YoY.[F7] Lynas LAMP operational. |
| Western magnet manufacturing | No Western sintered NdFeB at scale. | MP Fort Worth commenced magnet production Q4 2025.[F8] |
| Chinese price weapon | Available and used. Output flood collapsed prices. | Still available. China retains ~60% mining, ~90% processing.[F9] |
| Equity pattern risk | Narrative re-rating preceded capability by years. | Persists. Same dynamic structurally possible. |
The policy framework is more durable — statutory mandates are harder to unwind than WTO-challengeable quotas. Western processing is more advanced, with real tonnage rather than pilot plants. But China retains the ability to flood the market. The Molycorp lesson: distinguish between demand mandates (durable) and price assumptions (fragile). Any position sized on price recovery alone carries Molycorp risk.
[F1] USGS Mineral Commodity Summaries 2011; NdO spot via Asian Metal Inc.
[F2] Molycorp (NYSE: MCP) historical price; NdO peak ~$340/kg, Asian Metal Inc.
[F3] NdO collapse: Asian Metal Inc, 2012–2015. USGS Mineral Commodity Summaries.
[F4] Molycorp Chapter 11 filing, Delaware Bankruptcy Court, Case No. 15-11357, June 2015.
[F5] WTO DS431/432/433, Appellate Body Report adopted August 2014.
[F6] NDAA FY2024, Section 854.
[F7] MP Materials 10-K FY2025, SEC filing Feb 2026.
[F8] MP Materials Q4 2025 earnings call.
[F9] USGS Mineral Commodity Summaries 2025; IEA Critical Minerals Market Review 2025.
Facility Eligibility Map + Molycorp Analog
Title III eligibility assessed facility-by-facility across 9 sites, plus the complete Molycorp 2010–2015 analog with a 2026 comparison table.
Unlock Full AnalysisAppendix G: Busan Renewal — Conditional Probability Framework
The Busan Framework expires November 10, 2026. Whether it renews depends heavily on the Trump-Xi Beijing summit scheduled for May 14-15, 2026 — a mid-truce checkpoint that will set the diplomatic tone for the remaining six months.
USTR Jamieson Greer stated on April 8 that rare earth issues are "preferable to resolve at ministerial and staff level," suggesting the administration views mineral supply as a technical track rather than a summit-level bargaining chip.
| Summit Outcome | P(Busan Renewal) | Key Indicators |
|---|---|---|
| Visible goodwill or minerals working group announced | 55-65% | Joint statement references "critical mineral cooperation" or "supply chain resilience working group" |
| Neutral — no deliverables, no visible friction | 35-45% | Generic communiqué language; no Entity List changes within 72 hours |
| Visible tension on Taiwan, Entity List, or Hormuz | 15-25% | Hardened Taiwan language; Entity List additions within 72 hours; Hormuz escalation |
The blended ~40% renewal probability in the scenario matrix assumes roughly equal weight across these outcomes.
Bearish: Joint statement hardening on Taiwan; Entity List announcement within 72 hours of summit close.
Bullish: Bilateral minerals working group mentioned; confirmed USTR/MOFCOM ministerial sessions before October 1, 2026.
Trigger invalidation: If State Council readout language shifts toward "supply security" or "critical mineral cooperation" by September 1, 2026, OR confirmed ministerial sessions are announced before October 1, the "talks absent" trigger has not fired.
On MOFCOM quota data. China stopped publishing annual rare earth export quota volumes in 2021. The best available proxy is MOFCOM license application processing rates tracked via ACREI, covered by Reuters and Bloomberg. Processing delays or elevated denial rates are leading indicators of tightening.
Appendix H: Energy Underlayer — LNG and Hormuz as Compounding Risk
The atom-level chokepoint is one layer. Beneath it: an energy-cost layer operating on a different timeline but compounding the same structural thesis.
Frontier AI compute is constrained by cheap, reliable electricity — in the near term, natural-gas-fired generation at scale. The US-gas-to-seaborne-LNG price spread is an AI valuation input most sell-side models do not capture.
| Layer | Scarcity | Question |
|---|---|---|
| Atom layer (13 critical materials) | Material scarcity | Who can physically build AI/defense/energy infrastructure? |
| Energy layer (gas-to-LNG + Hormuz) | Energy scarcity | Who can afford to run it? |
Europe and Asia are building digital sovereignty on top of a maritime chokepoint they do not control (Hormuz, ~20% of global LNG trade). The US sits on abundant domestic gas, deep capital markets, and the dominant AI ecosystem. Material scarcity and energy scarcity compound: a nation unable to source critical materials and unable to power its fabs faces a qualitatively different constraint than either shock alone.
Framing adapted from James E. Thorne: "the AI arms race is being won in the wellheads and power plants of the United States." The Hormuz exposure thread is developed in the November Cliff Cascade.
Appendix I: China NdFeB Demand Disaggregation
China installed approximately 130 GW of wind capacity in 2025, but magnet demand requires disaggregation by turbine type. Applying a single NdFeB intensity rate across all wind additions overstates demand by 40–50%.
| Segment | GW Installed | NdFeB Intensity (kg/MW) | Total NdFeB (tonnes) |
|---|---|---|---|
| Offshore direct-drive (Goldwind, MingYang) | ~18 | ~650 | ~11,700 |
| Onshore PMDD (permanent magnet direct drive) variants (~15–20% of ~110 GW) | ~17–22 | ~80 | ~1,500–2,000 |
| Onshore DFIG (doubly-fed induction generator) — majority | ~88–93 | ~0 | negligible |
| Total wind | ~130 | ~13,000–14,000 |
Full Chinese NdFeB Demand Stack (2025 Estimate)
| Demand Segment | Volume Driver | NdFeB Estimate (tonnes) |
|---|---|---|
| Wind (offshore + onshore PMDD) | ~130 GW, disaggregated above | 13,000–14,000 |
| New energy vehicles | ~15M units × 2–5 kg | 30,000–75,000 |
| Industrial motors, robotics, military | Broad electrification, defense procurement | 80,000–105,000 |
| Total domestic consumption | ~170,000–195,000 |
Export surplus: Against ~250,000t total production, revised domestic consumption of 170,000–195,000t leaves roughly 55,000–80,000t available for export — wider than the prior 50,000–70,000t estimate. The NEV segment introduces the most variance: at 2 kg/unit, modest; at 5 kg/unit, it alone approaches wind-equivalent volumes.
Appendix J: Export Monitoring Methodology — Source Hierarchy and Falsification
The structural scarcity thesis depends on Chinese exports declining as domestic consumption absorbs production capacity. Monitoring requires precise source selection and correct HS code coverage.
| Source | Frequency | Lag | Role |
|---|---|---|---|
| Adamas Intelligence quarterly tracker | Quarterly | ~30 days | Leading indicator — preferred for trend detection |
| GASC monthly customs data | Monthly | 60–90 days | Confirmation indicator — highest HS code granularity |
| REIA (Chinese Rare Earth Industry Association) | Periodic | Variable | Production and policy context |
| JOGMEC annual reports | Annual | ~6 months | Structural baseline, Japanese import perspective |
Kill number: 60,000 tonnes of NdFeB-equivalent rare earth oxide exports per year.
If Chinese exports remain consistently above 60,000 tonnes through end of 2027, the absorption thesis is not on schedule and scarcity weakens to primarily policy-dependent. At current growth rates (~15% EV, direct-drive wind), exports should fall below that threshold by mid-2027 without additional controls.
Appendix K: OEM Demand Destruction — Motor and Magnet Redesign Programs
Two parallel tracks are reducing rare earth intensity in the automotive and industrial motor sector. Material-level substitution is already in production; architecture-level substitution is 5–8 years from volume deployment.
| Track | Program | Effect | Timeline | Source |
|---|---|---|---|---|
| Material-level | Toyota La/Ce magnet substitution | Reduces neodymium content by ~50% by substituting abundant lanthanum and cerium in traction motor magnets | Mid-2020s deployment targeted | Toyota Environmental Report 2024; Nikkei Asia[9] |
| Grain boundary diffusion (GBD) | Reduces Dy/Tb requirements by 50%+ per sintered NdFeB magnet while maintaining high-temperature coercivity. Now industry-standard across Shin-Etsu, TDK, Vacuumschmelze, Arnold Magnetic | In production | Industry technical literature[10] | |
| Architecture-level | Renault EESM (electrically excited synchronous motor) | Eliminates NdFeB magnets entirely; confirmed RE-free in Mégane E-Tech drivetrain | In production (2023) | Renault Group press release[11] |
| BMW wound-rotor synchronous motor | RE-free motor design for iX3 and successor platforms | In production (2024) | BMW Group technology briefing[11] | |
| Volkswagen Group wound-rotor exploration | Evaluating RE-free designs for MEB+ platform | Design phase (~2027–2028 volume) | VW investor day 2024[11] |
Appendix L: Announcement 61 Enforcement Analysis + NDAA 854 Exception Mechanism
Announcement 61 enforcement. The extraterritorial mechanism requires tracing Chinese-origin rare earth content through assembled goods crossing third-country borders — a chain-of-custody tracking system comparable to the US FFDCA lot certification regime, applied to tens of thousands of magnet-containing product lines. China does not presently have a customs technology layer capable of distinguishing a Japanese-manufactured permanent magnet containing Chinese-origin NdPr from one using non-Chinese NdPr at the point of third-country export.
This does not make the extraterritorial claim toothless. Major listed magnet producers (Shin-Etsu Magnetics, TDK, and other Tier 1 suppliers with public compliance obligations) will self-comply to avoid regulatory risk. But voluntary compliance by named producers is structurally different from enforcement against the long tail of component assemblers in Southeast Asian free-trade zones. The practical effect is asymmetric: it constrains Tier 1 producers immediately while smaller assemblers operate in a compliance grey zone until enforcement infrastructure is built.
Monitor: MOFCOM guidance documents post-reactivation for whether a certificate-of-origin tracking framework is announced alongside the controls.
NDAA Section 854 exception mechanism. The statute includes a waiver path: the Head of Contracting Activity may grant nonavailability exceptions for individual contracts, and the USD(A&S) may issue class determinations exempting entire programme categories, subject to 30-day Congressional public notice.[3] This means January 1, 2027 is not a hard cutoff — programmes unable to source compliant materials operate on rolling exceptions while pursuing qualification. The exception mechanism extends the demand runway for Western supply chain build-out, converting the statutory deadline into a sustained qualification urgency rather than a single compliance cliff.
Methodology
| Nodes | 2,901+ (companies, materials, policy actors, facilities, mechanisms) |
| Edges | 10,447+ supply chain relationships |
| Graph algorithms | Directed cascade traversal (graph traversal forward from material nodes), bridge node detection, single point of failure analysis, betweenness centrality (a measure of how often a node sits on the shortest path between others in the network), hop-level market cap aggregation |
| Severity index | 1 (commodity, many suppliers) to 5 (structurally binding at current margin). Derived from supplier concentration, qualification timelines, and cascade exposure counts. |
| Cross-pillar identification | Atoms identified by running directed traversal from each of 4 pillar seed sets (AI semiconductor, defense, clean energy, robotics) and finding material nodes appearing in 2+ traversal outputs simultaneously |
| Policy Phase 0 expansion | 73 policy and regulatory nodes added across two expansion passes, adding 72 edges. Increased MOFCOM, NDAA, CRMA, IRA, and CHIPS Act coverage to 29% of policy-relevant material nodes |
| Material selection criteria | Inclusion requires all three conditions: (1) appears in ≥2 of the four pillar traversals (AI, defense, clean energy, robotics); (2) has a China-gated processing or export-control constraint; (3) is not substitutable within 12–24 months without requalification. Cobalt excluded: substitution is advancing rapidly (LFP cathodes, cobalt-free chemistries). Fluorite reclassified as a Tier 2 watch item: does not meet all three criteria, but acidspar price bifurcation (30–40% premium over metspar at peak) warrants monitoring. |
| Sourcing standard | Every factual claim is tied to a primary or tier-1 secondary source (company filings, USGS, MOFCOM notices, SEMI, SMM, Fastmarkets, NDAA public law text). Severity 4–5 ratings on supply chain nodes require named competitor evidence and market share sourcing. Claims that cannot be verified to this standard are either removed or explicitly qualified. |
| Data sources | USGS Mineral Commodity Summaries (annual), MOFCOM official announcements, CSET Georgetown critical materials database, Future Market Insights market reports, ITIA, SMM Metal Intelligence, company filings (Lynas, MP Materials, Energy Fuels, Cameco, Centrus Energy), NDAA Public Law 118-31, EU CRMA implementation guidance |
Built on the ForcedAlpha Supply Chain Intelligence Graph — a directed graph mapping upstream material dependencies to downstream consumers. Every factual claim is sourced, every severity rating is competitor-verified, and the graph snapshot is cryptographically hashed via the ?v= URL parameter so any reader can audit the exact dataset that produced it.
Multi-pass research and critique process. Framing evolved from “China rare earth risk” to “shared upstream chokepoints across four uncorrelated demand pools.”
- Supply-side sourcing: 13 atom data packs verified against USGS, MOFCOM, CSET, company filings. China control percentages sourced with publication dates. UNVERIFIED fields flagged explicitly in data packs.
- Demand-side sourcing: Pillar assignments verified — specifically testing that each material is a genuine input to the pillar application, not a substitutable commodity input. InP wafers confirmed as non-substitutable for III-V semiconductor applications.
- Internal critique pass: Molybdenum substitution timeline quantified and elevated to counter-thesis. Rare earth 2010–2011 analog added. Coverage gaps section promoted to named section.
- Graph topology review: Cross-pillar matrix confirmed via directed traversal from 4 pillar seed sets. Policy Phase 0 expansion validated: 72 new policy edges confirmed plausible, 4 policy nodes flagged as unclassified.
- Domain sanity check: material science claims, production processes, and geopolitical timelines verified against primary sources. Corrections applied where needed; resolved items documented in Coverage Gaps above.
Related Analysis
How to Cite This Analysis
Pre-formatted citations by topic:
For critical minerals / rare earth coverage:
For semiconductor / WF6 tungsten coverage:
For defense policy / NDAA coverage:
For germanium / price divergence coverage:
For supply chain / downstream exposure:
For research inquiries, data access, or interview requests: [email protected]
Sources
- Germanium price divergence (Western vs. China domestic), April 2026. introl.com. Western spot $5,475/kg vs. Chinese domestic $2,049/kg cited. USGS MCS 2026 primary production share confirmation: pubs.usgs.gov.
- MOFCOM export control timeline. Announcement 18/2025 (April 4, 2025): medium/heavy rare earth licensing (Sm, Gd, Tb, Dy, Lu, Sc, Y) — currently active. mofcom.gov.cn. Announcements 61 and 62 (October 9, 2025): extraterritorial goods controls and technology transfer ban — suspended November 7, 2025 through November 10, 2026 (Announcement 70/2025). Analysis: CSET Georgetown. cset.georgetown.edu. Suspension analysis: Pillsbury.
- NDAA FY2024 Section 854, Public Law 118-31. congress.gov. Effective January 1, 2027. Full scope analysis: Crowell & Moring. crowell.com.
- USGS Mineral Commodity Summaries 2024: Gallium. pubs.usgs.gov. China ~98% of global gallium recovery from alumina. August 2023 MOFCOM export licensing and January 2025 chelating resin expansion confirmed.
- CNMC acquisition of Mineracao Taboca (Brazil tantalum), November 2024. mining.com. Taboca accounts for an estimated 10% of global tantalum mine production from Pitinga mine, Amazonas state (Roskill estimate; some sources cite higher figures).
- Trump–Xi Beijing summit confirmed May 14–15, 2026. White House announcement March 25, 2026: CNBC. USTR Greer statement on rare earth ministerial preference, April 7, 2026: Al Jazeera.
- Lam Research ALTUS Halo for molybdenum atomic layer deposition. lamresearch.com, February 19, 2025. Kioxia fluorine-free molybdenum word line process (BiCS10): kioxia.com.
- NdFeB magnet recycling rates. Estimate of 1–3% current recycling rate from end-of-life motors: ITIA and industry estimates. 30–35% by 2035 projection from academic and industry forecasts cited in Critical Materials Institute reports. Primary source for NdFeB demand by end market: ITIA annual report.
- Toyota rare earth magnet redesign. Toyota R&D press release: magnets substituting lanthanum and cerium for neodymium, reducing Nd content by approximately 50% in traction motor applications. Confirmed in Toyota Environmental Report 2024 and Nikkei Asia coverage.
- Grain boundary diffusion (GBD) technology. Industry consensus: GBD reduces Dy/Tb requirements by 50%+ per sintered NdFeB magnet while maintaining high-temperature coercivity. Adopted across Shin-Etsu Magnetics, TDK, Vacuumschmelze, and Arnold Magnetic. Arnold Magnetic Technologies technical white paper; Shin-Etsu Chemical annual report 2024.
- European OEM rare earth elimination programs. Renault Mégane E-Tech electrically excited synchronous motor (EESM) confirmed RE-free, Renault Group press release 2023. BMW iX3 wound-rotor synchronous motor, BMW Group technology briefing 2024. Volkswagen Group exploring wound-rotor designs for MEB+ platform per investor day presentation 2024.
- Chinese NdFeB production, domestic consumption, and export estimates. Adamas Intelligence quarterly rare earth market tracker, Q4 2025 and Q1 2026 editions. Cross-referenced against GASC monthly customs data and Chinese Rare Earth Industry Association (REIA) periodic reports. EV production figures: CAAM (China Association of Automobile Manufacturers) monthly sales data. Wind installation: CWEA (Chinese Wind Energy Association) annual statistics.
- USGS Mineral Commodity Summaries: Indium. Annual. China ~70% of global refined indium output, recovered as byproduct of zinc and lead smelting. No MOFCOM controls as of April 2026.
- High-purity phosphorus (yellow phosphorus) China production share approximately 85%: inferred from China’s dominance in yellow phosphorus smelting capacity. Primary confirmation: USGS Mineral Commodity Summaries: Phosphate Rock (annual). Specific yellow phosphorus production share noted as UNVERIFIED in atom data pack — figure is industry consensus estimate, not a primary source number.
- Rare earth mine output, separation, and metal production China shares. USGS Mineral Commodity Summaries: Rare Earths. pubs.usgs.gov. 60% mine output, 85%+ separation, 90%+ metal production figures. Lynas first non-Chinese dysprosium oxide production (May 2025): Lynas Rare Earths investor announcements.
- NdFeB sintered magnet production China share approximately 92%. Source: Adamas Intelligence market report cited in industry coverage. Cross-referenced against USGS RE summary. Individual magnet producer shares not publicly audited at this precision — treat as directional estimate.
- Uranium enrichment capacity. Pre-2024 Rosatom share of approximately 44% of global enrichment capacity (Separative Work Unit): World Nuclear Association. world-nuclear.org. HALEU definition: US NRC (Nuclear Regulatory Commission).
- Japanese material market shares. GaN substrates (~60%): Yole Développement compound semiconductor substrate report 2025. Silicon wafers (~51%): SEMI silicon wafer shipment statistics 2025; Shin-Etsu Chemical investor materials. Strain wave gears (~50%): Harmonic Drive Systems annual report FY2025; confirmed in JARA (Japan Robot Association) industry survey.
- IRA domestic content requirements, EU Critical Raw Materials Act strategic project framework, CHIPS Act supply chain provisions. EU CRMA: eur-lex.europa.eu. Combined $2T+ program estimate is a ForcedAlpha aggregation across IRA clean energy credits, CHIPS manufacturing incentives, and EU CRMA strategic project capital envelope; treat as order-of-magnitude, not precise figure.
- MP Materials Mountain Pass operations and NdPr commissioning. MP Materials investor relations. mpmaterials.com. DFARS compliance relevance per Section 854 confirmed against Crowell & Moring analysis (see footnote 5).
- Molycorp historical share price and bankruptcy. NYSE historical data (ticker MCP): approximately $14 (late 2010) to $74 (May 2011). Chapter 11 filing: U.S. Bankruptcy Court, District of Delaware, Case No. 15-11357, June 25, 2015.
- China rare earth export quota transparency. China stopped publishing annual rare earth export quota volumes in 2021 after maintaining them for 30 years. Analysis: Shen, L. et al., “China’s rare earth export controls: a review,” Mineral Economics (Springer), 2025. Historical quota data: USGS Mineral Commodity Summaries, Rare Earths, 2010–2021 editions.
- Usman, Z. (2026), “The International Trade Dimensions of the United States Critical Minerals Security Strategy,” UNU-WIDER Working Paper 2026/35, April 2026. Center on Global Energy Policy, Columbia University. doi.org/10.35188/UNU-WIDER/2026/710-7. Analysis of 6 bilateral critical minerals agreements (Australia, DRC, Japan, Malaysia, Thailand, Ukraine), tariff exemptions, and preferential trading zone proposal.
- Bloomberg News, “China’s Rare Earth Exports to Japan Fell to Nine-Month Low in March,” April 20, 2026. China customs administration data: magnet volumes -17% MoM to ~184 tonnes; intermediate materials including oxides -~90% MoM. Attributed to souring China-Japan relations.
- Qu, Tracy. “Rare-Earth Stocks Rise as China Looks to Tighten Production Oversight,” The Wall Street Journal, April 29, 2026. China Ministry of Industry and Information Technology proposal: fines up to 5× illegal gains for minor quota breaches; business-license revocation for violations exceeding 30% above quota; scrutiny on flow records and inspection obstruction. Public comment period through end of May 2026. Cross-source verification: Mining.com summary of WSJ reporting. Same-session price action: China Rare Earth Nonferrous Metals +8.9%, JL Mag +8.1%, China Rare Earth Holdings and Shenghe Resources both at 10% daily limit.
- MOFCOM Announcement 2026 No. 21 (May 2, 2026). First-ever invocation of China’s blocking statute (Rules on Counteracting Unjustified Extra-territorial Application of Foreign Legislation and Other Measures, 2021). Orders all Chinese entities not to recognise, enforce, or comply with US sanctions under EO 13902 and EO 13846 targeting five Chinese teapot refineries. MOFCOM official announcement (Chinese). Analysis: Geopolitechs, May 2, 2026.