Magnet Rare Earths: Neodymium-Praseodymium Supply Risk and Permanent Magnet Alternatives

Executive Summary

China controls approximately 92% of refined NdPr supply and between 98-99% of separated heavy rare earths including dysprosium and terbium, and Beijing's April 2025 export licensing regime has converted that structural fact into an active geopolitical instrument, with direct consequences for EV production timelines and defense procurement security. The controls introduced a permanent licensing architecture that has survived two diplomatic cycles intact: the evidence strongly suggests that the core elements of China's export control regime are permanent features of global trade rather than temporary measures, with the progression from reactive trade countermeasure to institutionalized licensing architecture indicating that Western manufacturers should plan on the basis that these controls will remain in place. The supply chain response is real but delayed, with allied capacity low confidence to close the processing gap before 2028 at the earliest.

• Supply-chain/operations: Map your tier-2 magnet dependencies by element now; heavy rare earths (dysprosium, terbium) face structural supply denial to defense end-users, while NdPr faces cyclical pricing risk pending China's H2 quota announcement.
• Risk officers/investors: The November 2026 suspension expiry is your primary risk horizon. If the agreement lapses without an extension, or if geopolitical conditions deteriorate before then, a return to full October 2025 controls would trigger acute shortages across automotive and electronics supply chains.
• Defense/OEM procurement: Lockheed Martin and peers face a 2027 DFARS deadline requiring full traceability to the mining level; qualifying domestic sources now, before the mandate takes effect, avoids a compliance cliff.

China's processing lock on NdPr and heavy rare earths will not be materially loosened before 2028, making the November 2026 expiry of the October 2025 suspension the single most consequential near-term inflection for both the EV and defense industries.

What Changed

On April 4, 2025, China's Ministry of Commerce introduced export licensing requirements on seven rare earth elements in retaliation for US tariff announcements, permanently institutionalizing what had previously been an implicit coercive capacity. The restrictions announced in late 2025 represent China's most consequential measures to date targeting the defense sector; under the new rules starting December 1, 2025, companies with any affiliation to foreign militaries will be largely denied export licenses. A partial suspension was agreed at the October 2025 Busan summit, but the suspension applies only to the controls announced on October 9, not the previous round enforced in April 2025, meaning Trump's claim that the rare earth issue has been "settled" is not entirely accurate.

The Chokepoint Is Processing, Not Mining

The common analytical error when discussing rare earth concentration is to anchor on mining geography, where diversification is proceeding at a visible pace. The more important variable is the processing stack, and here the picture is materially different.

Despite relatively modest global volumes, the rare earths market remains acutely vulnerable due to extreme supply concentration, particularly in processing and refining. While mining is somewhat diversified geographically, China's dominance becomes overwhelming further down the value chain. This processing concentration constrains the entire non-Chinese value chain by a specific mechanism: even miners operating outside China, such as MP Materials in California and Lynas in Australia, must either route concentrates through Chinese facilities or invest billions in entirely new refining infrastructure.

What is not being reported: Western supply chain press coverage focuses heavily on mine announcements and government loan packages. The bottleneck that receives far less attention is metallization, the step between separated oxide and usable alloy. Metallization is the least developed part of the value chain outside China. It requires deep, accumulated operating expertise and process control systems capable of managing complex variables in continuous production. Even with capital and strong execution, replicating that capability typically takes three to seven years or more, with meaningful technical and qualification risk, according to REalloys co-founder Tim Johnston.

China controls more than 90% of the world's rare earth refining and separation capacity. Its light rare earth magnetic separation capacity alone reaches 350,000 tons per year, while the combined capacity of all other regions is less than 50,000 tons per year, a substantial gap that cannot be easily closed. This capacity gap translates directly into commercial terms: rare earths processed outside China are now commanding premiums of four to six times domestic Chinese prices, and this spread is not merely a market anomaly, it is a direct consequence of policy-driven fragmentation and supply chain realignment.

This processing bottleneck spills into the defense domain through a specific mechanism: light rare earth permanent magnets produced outside China currently cost 30-50% more than comparable products made in China, severely limiting their market competitiveness, which in turn constrains the economic viability of domestic defense-oriented production at scale without sustained government price supports.

The Ex-China Build-Out: What Is Real And What Is Not Yet

Two non-Chinese players anchor the emerging ex-China production base: Lynas Rare Earths and MP Materials. Lynas remains the largest non-Chinese separator, at 7 ktpa of NdPr-equivalent, now pushing toward 10-12 ktpa NdPr-equivalent capacity, while MP Materials is ramping NdPr separation at Mountain Pass and building a "10X" magnet facility backed by a 10-year Pentagon price floor and offtake for 100% of output.

The Pentagon is financing this build-out through direct equity and debt. The DoD has taken a $400 million equity stake giving it approximately 15% ownership of MP Materials, making MP a declared national-security asset, alongside a $1 billion loan commitment from JPMorgan and Goldman Sachs for the 10X magnet facility, a $150 million DoD loan for heavy REE separation at Mountain Pass, and a $500 million Apple magnet and recycling partnership.

On the heavy rare earth side, the gap is more acute. Until recently, ex-China supply of separated dysprosium and terbium was negligible at commercial scale. MP Materials is commissioning its own dysprosium and terbium separation capability at Mountain Pass, on track for mid-2026, which would add a second non-Chinese source of separated heavy rare earth product to the market.

Short-term gain, long-term cost: The Pentagon's price floor of $110/kg for NdPr, set approximately $50/kg above market rates as reported by Resources for the Future, guarantees domestic supply but at a cost structure that would not survive open-market competition against Chinese pricing. US public investment, spearheaded by the administration, could stimulate 29,600 metric tons of new NdPr supply outside China this decade, but this output will remain price-uncompetitive for much of that window without continued subsidy.

Australia's CSIS-confirmed role as the most important US partner is reinforced by its investment posture: Australia attracted $64 million, about 45% of global rare earth exploration investment in 2024, five times more than Brazil, and hosts 89 active projects, far ahead of Canada, Brazil, and the United States. The US-Australia Critical Minerals Framework, signed in October 2025, provides the bilateral scaffolding for integrating Australian mining with US and Japanese processing and downstream magnet production.

Alternatives: The Motor Technology Hedge

The EV and robotics industries are pursuing two parallel hedging strategies: securing NdFeB supply through allied channels, and developing motor designs that reduce or eliminate rare earth content.

Niron Magnetics is the most advanced public effort on the materials substitution side. At CES 2026, Niron Magnetics and Indian EV firm MATTER revealed the first high-performance Variable Flux Motor prototype for light mobility using iron nitride magnets and a specialized controller to dynamically adjust flux, optimizing efficiency across both high-torque launches and high-speed cruising, without using rare-earth materials. Niron counts Volvo, Stellantis, and GM among its investors, according to IDTechEx.

The performance gap, however, remains a constraint. IDTechEx reports that due to their poorer magnetic performance, alternative magnets need to be larger. A typical rare earth motor will use 1-2 kg of magnets; for a ferrite motor this figure would moderate-to-high confidence double or even triple, creating a challenge for rotor structural integrity as the magnets make up a much larger proportion of the rotor.

Motor-level alternatives are also advancing. Honda became the majority shareholder (61%) of Astemo in late 2025, accelerating integration of Astemo's rare-earth-free e-Axles into Honda's "0 Series" EVs. Renault and Valeo are jointly developing a completely rare-earth-free EV motor, and ZF Friedrichshafen's experimental in-rotor inductive-excited synchronous motor dispenses with rare earth elements by using electromagnets, according to IEEE Spectrum.

The recycling avenue is real but pre-commercial at scale. IDTechEx predicts rare earth magnet recycling will increase 6.5 times over the next decade and represent up to 10% of global supply by 2036. Recovering neodymium and dysprosium from end-of-life EV motors and hard disk drives has been demonstrated at pilot scale by companies including Cyclic Materials in Canada and REEtec in Norway. As of mid-2026, recycled rare earth material is not cost-competitive with Chinese-produced virgin material at current market prices.

The supply risk framing that matters for decision-makers: Rare Earth Exchanges correctly frames the automotive sector's exposure not as "absence" but as "allocation." Supply may exist, yet access can still tighten through licensing, geopolitics, or preferential allocation to defense and the largest OEMs. The automotive sector would be the most exposed downstream industry if rare-earth controls were fully enforced by China.

Counterarguments

1. The processing gap overstates China's near-term leverage because buyers are stockpiling: A competent critic would note that the April 2025 controls accelerated strategic stockpiling by US and Japanese OEMs, compressing the practical impact window. S&P Global reported that China's export volumes fell to 4,392 mt in December 2025, 15.8% below the monthly average, yet major production shutdowns outside China did not materialize at scale through mid-2026. If inventory buffers are larger than publicly disclosed, the window during which China can cause acute supply disruption may be narrower than this assessment implies. What would lower confidence: confirmed inventory data from major defense contractors showing 12+ months of buffer stock on dysprosium and terbium.

2. The "permanent controls" thesis may project permanence onto what is calibrated strategic ambiguity: Resources for the Future's analysis raises the possibility that China views export restrictions as temporary instruments, withdrawn precisely when they begin accelerating Western investment. Damage felt by the rest of the world pushes it to begin investing in substitution measures, at which point China withdraws the restrictions, neutralizing the diversification investment before it becomes competitive. If Beijing's optimal strategy is cycling controls on and off to prevent Western self-sufficiency from reaching commercial viability, then treating the current regime as "permanent" could lead to over-investment in high-cost allied capacity that faces Chinese price competition precisely when it reaches scale.

3. The defense sector's formal supply denial may be less operationally binding than it appears: CSIS confirmed that as of the October 2025 controls, the December 2025 rules deny export licenses to military-affiliated end-users. However, Noveon Magnetics is currently the only manufacturer of rare earth magnets in the United States, and Noveon and Lynas announced an MOU focused on building a scalable domestic supply chain. If DoD-backed facilities such as the MP Materials 10X campus, eVAC Magnetics in South Carolina, and USA Rare Earth's Texas plant collectively qualify and deliver before the 2027 DFARS deadline, the formal denial becomes commercially moot for the US defense base, even if the Chinese licensing architecture remains on paper. The open question is qualification speed, not plant construction.

Indicators To Watch

Near-term watch list: (1) China MOFCOM H2 rare earth mining quota announcement (July-August 2026), which will set the physical supply direction for the rest of the year and signal whether Beijing intends to tighten or ease pressure ahead of the November expiry; (2) MP Materials Q2 2026 earnings call (August 2026), which will disclose Fort Worth magnet campus qualification progress and whether the DoD price floor offtake begins generating qualified output; (3) US-China November 2026 suspension renewal negotiations (outcome moderate-to-high confidence visible by October 2026), which is the single highest-impact binary event for OEM procurement planning through 2027.

Decision Relevance

Scenario A (~55%): Controlled tension, licensing regime stable through 2027 with periodic tightening. China maintains the April 2025 architecture, grants general licenses for civilian commercial buyers at reduced throughput, extends the October 2025 suspension for another year in November 2026, and continues blocking defense end-users. If you manage supply chain procurement for an EV OEM or Tier 1 supplier, the correct posture is to lock in multi-year offtake agreements with ex-China producers now, even at the 30-50% premium, treating the premium as insurance rather than inefficiency. If you lack direct rare earth exposure, monitor the NdPr FOB price spread monthly as a leading indicator of supply stress migrating upstream into motor production schedules.

Scenario B (~30%): November 2026 suspension lapses, China reimplements October 2025 controls on five additional elements. This would extend the formal supply denial to elements not currently restricted and reactivate extraterritorial provisions affecting products made with Chinese-origin content outside China. If you have rare earth magnet dependencies in aerospace or defense supply chains, this scenario requires immediate qualification of domestic alternatives and pre-positioning of strategic inventory. Defense contractors who have not begun DoD qualification processes for domestic magnet sources will face a compliance cliff against the 2027 DFARS deadline. If you are an investor, watch for aerospace and defense sector rotation into MP Materials, USA Rare Earth, and Lynas as the first market signal that Scenario B is pricing in.

Scenario C (~15%): Diplomatic resolution produces genuine licensing rollback, Chinese exports return to pre-April 2025 volumes. The Resources for the Future strategic game theory framework suggests China might withdraw controls precisely to prevent Western investment from reaching commercial viability. If you have committed capital to ex-China processing capacity, Scenario C is the scenario where that investment faces the most acute commercial pressure, as Chinese producers would respond with price competition targeted at the cost-of-production threshold for allied facilities. Do not reduce diversification investment on the basis of any diplomatic statement unless confirmed by published MOFCOM regulatory text; two consecutive summits have produced White House announcements that China's official statements did not corroborate.

Expert Integration

Government, academic, and industry analysts broadly agree that Chinese processing dominance cannot be materially reduced before 2028, that the April 2025 licensing architecture is now a permanent feature of the regulatory landscape, and that the defense sector faces a distinct and more acute access-denial risk than commercial EV producers. The CSIS May 2026 retrospective, Bloomberg Intelligence's March 2026 outlook, the Anderssen Institute's legal analysis, and the IDTechEx supply chain report reach consistent conclusions on these structural points.

Expert Disagreement Areas:

• Permanence vs. strategic ambiguity of controls: Resources for the Future argues China's optimal play is cycling controls on and off to prevent Western capacity from reaching competitive viability, implying the current architecture is a tool, not a permanent state. CSIS and the Anderssen Institute treat the licensing system as a permanent instrument. Both positions are logically consistent with available evidence.
• Timeline to meaningful processing independence: Bloomberg Intelligence projects China's NdPr share falling to 69% by 2030; industry sources closer to production schedules (Rare Earth Exchanges, neodymiummagneti.com) assess ex-China output remaining below 10% of China's capacity through 2026 and scaling slowly thereafter, with 2028 the more realistic threshold for any material shift.
• Alternative motor technology timing: Academic and patent analysis sources (Patsnap, IEEE Spectrum) are more optimistic about iron nitride and manganese bismuth commercialization timelines; industry-focused sources (Rare Earth Exchanges, IDTechEx) consistently cite 4-7 years to mass automotive deployment.

Systematic-Expert Alignment: MIXED

This assessment aligns with expert consensus on the structural permanence of Chinese processing dominance and the defense-sector access-denial risk. It diverges from the more optimistic Bloomberg Intelligence share-shift projections by weighting the metallization bottleneck more heavily, on the basis that the processing capacity gap is larger than oxide production figures suggest, and that qualification timelines for defense-grade material add further delay that aggregate market share projections do not capture.

• Actual DoD and allied defense contractor stockpile levels for dysprosium and terbium are not publicly disclosed; if inventories are substantially larger than market pricing implies, the practical impact of the defense-sector supply denial may be compressed relative to this assessment.
• Production ramp data for MP Materials' Fort Worth magnet campus and Lynas's Texas Seadrift facility is reported quarterly but with commercial sensitivity; qualification milestones for specific defense programs are not publicly confirmed, introducing timing uncertainty into the 2027 DFARS compliance picture.
• The assessment that China's licensing architecture is "permanent" rests on behavioral evidence across two diplomatic cycles, not on any Chinese official statement of intent; Beijing could alter this calculus for reasons not visible in open-source material, including domestic economic pressure or strategic recalibration.
• Rare-earth-free motor technology readiness ratings draw on OEM press releases and trade publication summaries; actual production qualification data at Tier 1 OEM scale is not independently verifiable, and commercial launch timelines in this sector have historically slipped by 12-24 months relative to initial announcements.
• The cost premium data for ex-China rare earth processing (30-50% above Chinese benchmarks) reflects 2026 market conditions; this spread could narrow materially if Chinese domestic costs rise due to environmental enforcement or quota tightening, or widen if allied capacity scales faster than demand growth.