China's export controls on core rare earth elements (REE) has led to scarcity across the globe and put the spotlight on REE supply chains, especially those which exclude China which is currently the world's biggest source of raw and refined rare earths, which affords it a unique advantage in the ongoing global trade wars.

In a recent must read report (available to pro subscribers), Morgan Stanley's metal and mining team published an in-depth look at a number of early-stage REE and magnet projects to map out where Western countries may be able to source these critical minerals, and to find the biggest potential winners from the biggest clash in the ongoing trade war between the US and China.

We excerpt the key highlights below.

1. Mapping a Potential ex-China Rare Earths Supply Chain

Key Takeaways

• REEs and permanent magnets are needed to build cars today and humanoids/robots in the future. Chinese export controls have pressured these supply chains.

• The US has a robust pipeline of ex-China upstream projects, but most are in the engineering and permitting phase, with a limited number under construction.

• Few ex-China permanent magnet projects in pipeline, and capacity of new facilities is <10% of Chinese facilities, in part due to lack of ex-China REE feedstock. 

• REE prices to rise as customers & governments diversify ex-China supply. This should lead to premium pricing for existing players and incentivize new supply.

• MP Materials, Lynas, and ILU are best positioned to capitalize as REE trade flows and pricing adjust.

US reliance on imported critical minerals has increased significantly, with growing implications for national security...   In the past 35 years, the US has seen significant growth in both the variety of imported minerals and its dependence on these imports. In 1990, the US was fully reliant on imports for the supply of 9 minerals, and imports exceeded 50% of consumption for 27 minerals (all minerals, not only critical ones). These figures have risen to 15 and 46, respectively, as of 2024. For US policymakers, recent trade tensions highlight the growing vulnerability of the US rare earth supply chain and the urgent need to strengthen domestic supply chains. As the risk of continued export restrictions and geopolitical uncertainty rises, the US and other nations will increasingly prioritize long-term national security over economic concerns.

… and the EU is finding itself in a similar position as it looks to initiate projects both inside and outside the EU...   The EU is placing greater focus on incentivizing an ex-China supply of critical minerals. Under the Critical Raw Materials Act (CRMA), the EU has identified 60 projects, 13 of them outside of the EU and 47 inside the EU, to secure critical mineral supply chains for the region. The projects span the full list of critical raw materials and include two rare earth element (REE) extraction projects outside of the EU — Songwe Hill in Malawi and Zandkopsdrift in South Africa — as well as three REE processing projects and two REE recycling projects inside the EU.

… leaving China in a powerful position to influence global rare earth markets, primarily via mineral export restrictions.  China dominates the global REE processing/refining industry, accounting for an estimated 65% of mined NdPr supply, 88% of refined NdPr supply, and >90% of downstream NdFeB permanent magnet supply in 2025, according to Woodmac. China has not hesitated to use its dominance of this critical supply chain as a geopolitical tool, dating back to 2010, when the country halted exports to Japan amid an escalating diplomatic crisis, a move that sent REE prices soaring. In 2019 and 2021 (here and here), China floated REE export controls amid rising trade tensions, which, according to The Financial Times, aimed to test how much defense contractors in the US and Europe would be affected. China has since banned the export of rare earth processing technology; most recently, in early April, it placed export controls on the export of heavy rare earth elements (HREE) and permanent magnets to all countries (i.e., not just the US). Because virtually all separated HREE production comes from China, this has led to distortions in market prices outside of China as companies scramble to secure supply (Exhibit 5 and Exhibit 6). 

Rare earths are critical to many current applications…  The US Department of Defense (DoD) deems rare earths as critical minerals because they are required (and difficult to substitute) in national security applications. While the DoD is estimated to be only ~5% of US demand, rare earths are used in essential military equipment such as motors in aircraft and tanks, missile guidance systems, and radar/sonar for submarines and surface ships (Exhibit 1 and Exhibit 2). The remaining ~95% of demand is primarily used in permanent magnets that go into autos (~40%); electronics & mechanical components, including speakers (~20%); HVAC systems (8%); wind energy (~8%); and others (~19%).

… and are key to future humanoid development.  As governments vie for the pole position in humanoid development, China is off to a strong start given its advanced rare earth and permanent magnet supply chains. Morgan Stanley's Metals and Mining team's work suggests NdPr will enter deficit by 2037 as a result of incremental demand from humanoids; however, this analysis includes Chinese supply of NdPr oxides, which we know has been exploited in the past to adversely affect foreign markets. Put plainly, the world is quickly realizing it is reliant on China's REE and permanent magnet supply chain to build cars today and humanoids/robots tomorrow. This uncomfortable realization may drive a meaningful uptick in mining investment in the US and other developed countries after decades of underinvestment. 

So, where will the ex-China supply come from in a multipolar world?  While MP Materials (MP) and Lynas (LYC.AU) have championed non-Chinese rare earth and permanent magnet supply chains (with ILU.AU currently building a refinery, with first production expected in CY27), China remains in control with ~88% of global refined NdPr supply, an even higher portion of permanent magnet capacity, and >99% of refined HREE supply. As a result, the rest of the world will need to develop not only upstream REE deposits but midstream separating/refining and downstream permanent magnet capabilities. This report highlights over 30 projects across the supply chain, many of which have received some form of government assistance. In the case of upstream and midstream projects, these are categorized based on their exposure to either light or heavy rare earths (Exhibit 7 and Exhibit 8). Based on the data provided by companies regarding their individual projects, the identified projects (mining, refining/separating, and magnet-making) will require $10-12 billion of initial capital investment. However, there are currently very few permanent magnet projects outside China and true level of capital investment needed to break China's grip on the permanent magnet market is likely much higher.

2. Story in Charts

Exhibit 1: Rare Earths and permanent magnets are required for key defense applications...

Exhibit 2: ...which can use 920-9,200lbs of rare earths, raising national security concerns given the world's reliance on imports from China for these materials

Exhibit 3: Moreover, MS forecasts the growth in humanoids will push NdPr into deficit by 2037…

Exhibit 4:… however, this is inclusive of China supply, which is estimated to account for >70% refined NdPr through the forecast period...

Exhibit 5: …but China's weaponization of REEs via export restrictions have led to scarcity in Dysprosium...

Exhibit 6: …and Terbium outside of China

Exhibit 7: … suggesting supply chains will need to look elsewhere and incentivize rest of world upstream REE...

Exhibit 8: …and downstream permanent magnet projects to come online...

Exhibit 9: …which will require higher incentive prices according to Woodmac...

Exhibit 10: …underpinning forecasts for increasing NdPr prices through the decade

Exhibit 11: Projects across the Rare Earth + Permanent Magnet Supply Chain located outside China

3. The Upstream: Where Will ex-China Mined Rare Earths Come From?

The pipeline of ex-China projects seems robust, but most projects remain in trial phases, with few companies having published preliminary economic assessments (PEAs) for their respective projects. Moreover, some of the most advanced projects, such as VHM's Goschen and Lindian's Kangankunde, both targeting initial production by YE26, are not currently looking to move downstream into separated oxide production. China controls 85-90% of refined production, and essentially all refined heavy rare earth element (HREE) production. Therefore, these projects are unlikely to make a dent in China's dominance as they will need to ship the concentrate they produce to the customers that can currently process it: Chinese REE refiners.

This list of projects is by no means intended to be exhaustive, and we will continue to monitor for additional developments.

Light REE Upstream Projects

Eneabba Project — Iluka Resources (ILU.AX; Market Cap: US$974m): Iluka Resources is advancing on a rare earth refinery to be located in Western Australia. The facility will have capacity to take ~55kt of concentrate and is expected to produce ~5.5ktpa of NdPr and ~0.75ktpa of DyTb. The project is expected to cost a total of $1.7-1.8bn, of which $680m has been sunk/committed to date, with commissioning of the site expected to begin in 2027. Once complete, Eneabba will represent Australia's first fully integrated refinery for the production of separated light and heavy rare earths oxides. The refinery will use monazite stockpiles at Eneabba as initial feedstock, along with material from the company's Balranald mine that will be commissioned in 2H25. Moreover, the facility will have the ability to take 3rd party concentrate as feedstock, though management will pursue these opportunities once closer to start-up.

Phalaborwa Project — Rainbow Rare Earths (RBW.L | Market Cap: US$99m): The Phalaborwa project is located in South Africa and is 85% owned by Rainbow Rare Earths, with option to acquire remaining 15% from Bosveld. The company also sold a 0.85% revenue royalty to Ecora Resources for $8.5m in cash in 2024. The project has a mineral resource of 35.0Mt at 0.44% TREO contained within two phosphogypsum stacks derived from historic phosphate hard rock mining (Exhibit 13). Phalaborwa’s operating cost is expected to be considerably lower than traditional rare earth projects as the phosphogypsum material is already sitting at surface in a chemically cracked form, which eliminates the cost and risk of mining, hauling, crushing, grinding, flotation, and cracking; thus allowing the company to overcome the relatively low grades at Phalaborwa.

The project has a 16-year mine life and expects to process 2.2Mtpa (of the 35.0Mt total resource). Using the 0.44% grades and ~65% recoveries in Phalaborwa's economic assessment, the company expects to produce ~1,900tpa of rare earth oxides, which includes ~1,820t of NdPr, ~60t of Dysprosium (Dy) and ~20t of Terbium (Tb). The company expects operating costs of $40.83/kg of magnet REO produced, which, based on spot prices for NdPr ($60/kg), Dy ($225/kg) and Tb ($1,000/kg), suggests operating margin of ~46%. Rainbow Rare Earths is targeting initial production in 2027 and is currently in process of preparing a definitive feasibility study for the project, after which financing and construction will begin. The project is expected to integrate midstream operations producing separated oxides of >99% purity.

Bear Lodge — Rare Element Resources (REEMF.OTC; Market Cap: US$374m): Rare Element Resources (RER) is a Colorado based company with 100% ownership of the Bear Lodge asset located in northeastern Wyoming as well as a midstream processing facility located nearby in Upton, Wyoming, where the company is currently finishing construction of a Demonstration Plant that has received ~$24m in funding from the DoE. The company is ~65% owned by a subsidiary of General Atomics who is working with RER to advance the company's proprietary extraction/separation technology. The Demonstration Plant, which will be using the aforementioned proprietary tech, is expected to begin operations in late 2025 and operate for up to 10 months and produce 10 tons of separated NdPr oxide, which will inform later feasibility studies. The Demonstration Plant will use stockpiled sample materials from the Bear Lodge REE project.

Bear Lodge has a total resource (measured, indicated and inferred) of 7.92Mt at a TREO grade of 3.97%, which implies total contained RE 314kt. Based on a 2014 technical report, the TREO grade of 3.97% has ~23% NdPr content, 0.5% Dysprosium and 0.1% Terbium. The company plans to conduct further economic assessments following completion of the 10 month production period at the Demonstration Plant.

Longonjo — Pensana (PRE.L; Market Cap: US$259m): The Longonjo project is located in Angola and will produce a mixed RE concentrate to export via the Lobito Corridor to a downstream separation plant in the UK that would produce separated oxides. The site has a total Mineral Reserves of 21.5Mt at a TREO grade of 3.04%, which includes ~21.5% NdPr. The current mine plan, based on Mineral Reserves only, has a 20+ year life of mine and expects to produced 20,000t/yr of a mixed rare earth concentrate (MREC). This would be sent to the downstream separation facility that is expected to produce 12,500tpa of separated rare earth oxides, including 4,400tpa of NdPr. The company announced that construction at the site began as of May 2025 and expects the construction and commissioning process to take ~22 months.

Songwe Hill Mine + Pulawy REE Separation Plant — Mkango Resources (LSE: MKA.L; Market Cap: US$81m): Mkango Resources is currently pursuing the development of an REE mine in the East African country of Malawi, the Songwe Hill REE project, that will produce a mixed rare earth carbonate (MREC) to feed a processing/separation facility to be built in Poland. Both the mine and separation facility have been deemed startegic projects by the EU's Critical Raw Materials Act. Songwe Hill is one of the few rare earth projects to have completed a definitive feasibility study and the mine is expected to have an 18 year life of mine, averaging 5,964tpa of TREO, which includes 1,953tpa of NdPr and 56tpa of Dy and Tb oxide. The MREC will be shipped to the separation facility in Poland, which is currently advancing a feasibility study, that is expected to produce 1,018tpa of NdPr oxide. Both sites have an estimated starting date of production in 2027.

Zandkopsdrift — Frontier Rare Earths (private): The Zandkopsdrift deposit is located in South Africa and has been deemed a strategic project by the EU's Critical Raw Materials Act (CRMA). Frontier Rare Earths completed a prefeasibility study and began the definitive feasibility study (expected to last 18-24 months) in 2H2023. Upon completion of the DFS, the company will complete front end engineering and construction financing works before commencing a 24 month construction period with first production expected in 1H28. The site plans to process 1Mt of ore each year, which would produce ~4,000tpa of magnet rare earths over a 45+ year life of mine based on current proven and probable reserves. The production will be heavily weighted toward NdPr, which will account for 3,800 of the 4,000t, with Dy at ~143tpa and Tb at 32tpa. Zandopsdrift is currently only focused on extraction of these minerals and producing a TREO concentrate.

Kangankunde REE Project — Lindian Resources (LIN.AX; Market Cap: US$80m): The Kangankunde project is located in Malawi and is fully permitted to commence construction and operations. The company completed a feasibility study in June 2024 that covers "Stage 1" operations, which includes mining operations and a mineral processing plant to produce monazite concentrate. The current operational plan is focused on concentrate production and there are currently no plans to produce separated rare earths. Lindian resources has begun preconstruction activities and is nearing completion of a further optimized feasibility study. The company is targeting initial production in 2026. The asset is expected to produce ~8,250t of TREO each year, which includes ~1,600tpa of NdPr.

Halleck Creek — American Rare Earths (ARR.AX; Market Cap: US$90m): The Halleck Creek project has one of the largest REE resources in North America at 2.63 billion tonnes, however the site has relatively low TREO grades of ~0.33%. The project is located on state and federal land and the company plans to move forward with the Cowboy State Mine, located within Halleck Creek, as it is entirely on state land which can speed up the permitting process by 5-10 years. Based on an updated scoping study in Feb 2025, the company expects to mine 3Mtpa over a ~20 year life of mine and plans to have onsite mineral processing and separation facilities. The company's scoping study suggests ~1,800tpa of NdPr oxide, 24tpa of Tb oxide and 98tpa of Dysprosium; a pre-feasibility study is expected to be published in 4Q25. In addition, the company is progressing on state permitting and environmental baseline studies, and ARR's scoping study envisions a 2.5-year construction period. As a result, initial production is not expected until 2029/2030.

Sheep Creek Deposit — US Critical Materials Corp (private): The Sheep Creek Deposit is located in southwest Montana and the land claims owned by US Critical Materials are located on multi-use land administered by the US Forest Service. The site was previously mined in the 1960s, which has allowed US Critical Materials the ability to access significant mineralization and underground samples via pre-existing mine workings that were reopened in 2023 providing access to the property 150-400 feet below the surface. Early stage exploration efforts continue at the site and initial results suggest potential for one of the highest grade deposits in North America at ~9% TREO, of which ~27% is NdPr. The company is partnering with Idaho National Laboratory (INL) on developing a process to handle, separate, and extract metals from the carbonatite ore located at Sheep Creek. The project is still in the very early stages and US Critical Materials in conjunction with INL expect to provide updates on progress made with extraction and purification testing along with a draft flowsheet update in summer 2025.

Lulea Industrial Park + Malmberget and Per Geijer Mines — LKAB (private): LKAB is advancing on development of two iron ore mines with REE byproducts and an industrial processing facility all located in Sweden. The project has been deemed a strategic project by the EU's Critical Raw Materials Act (CRMA). The company is currently constructing a demonstration plant at Lulea to verify the process to produce critical minerals. The company is still in the permitting phase for the 3 aspects of this strategic project and does not expect to reach full-scale commercial operations at the industrial park until the 2030s.

Brook Mine — Ramaco Resources (METC.O; Market Cap: US$616m): The Brook Mine is located in the Sheridan Coal Field in northern Wyoming and consists of approximately 15,800 acres, of which 4,600 acres are already permitted. The land was originally acquired to be converted into a coal mine in 2012, but Ramaco pivoted and began REE exploration in 2021 and 2022. TREO content is estimated at 1.1-1.4mt at a grade of between 0.0375% and 0.0469%. That said, the companies current tonnage estimates are not considered mineral resources nor mineral reserves as no cut-off grade was applied in preparing the estimated tonnage and grades (i.e., it represents a comprehensive estimate of all in-place minerals, regardless of grade). The company, in conjunction with the Flour corporation, which has been assisting with testing and engineering for the project, plans to release a preliminary economic assessment by the end of 2Q25. Under Ramaco's commercial development timeframe, the company plans to have pilot operations producing rare earth concentrate in 2026 and also begin production on a full-scale plant to produce commercial oxides by 2028.

Woods Creek — Integral Metals (INTG.CN; Market Cap: US$20m): Located near the Sheep Creek Deposit, in southwest Montana, the site is part of a similar mineralization and targets REE within carbonatite ore. The site is located on federal lands, which face longer permitting timelines, and is currently in early stage exploration/drilling. Initial grab samples of carbonatite collected in 2024 returned a total rare earth oxide (TREO) of 7.08% containing ~14% NdPr. Given the need for further permitting and the long-lead times for permitting on federal land (10+ years in some cases) this project is unlikely to begin production until mid-next decade, barring any changes to the Federal permitting process.

INSPIREE Project — Itelyum (private): The INSPIREE project is located in northern Italy and has been deemed a strategic project by the EU's Critical Raw Materials Act (CRMA). The site will be a small-scale magnet recycling facility that produces REE oxalates from end of life magnets. The site will have a two-step process with Level I involving the disassembly of magnets and Level II consisting of the recovery of REE oxalates from said magnets at Itelyum's hydrometallurgical plant. The Level I plant will have dismantling capacity of 1,000 tons/year and the Level II plant will treat 2,000 tons/year of permanent magnets, resulting in the recovery of 700 tons/year of REE oxalates.

Upstream — Heavy REE Projects

Lynas HREE Separation Facility — Lynas Rare Earths (LYC.AX; Market Cap: US$5.6bn): Lynas is currently ramping up operations at its Malaysia separation facility, which according to the company will have separating capacity of 1,500 tonnes of heavy rare earths per year. The company installed the Dy and Tb separation circuits in February of this year and announced first production dysprosium oxides in May, 2025 followed by first production of terbium oxides in June, 2025. Achieving production of separated Dy and Tb has made Lynas the only commercial producer of HREE products outside of China, at a crucial time given China's recent placement of export controls on these materials. Plans are also underway to construct Light Rare Earth (1.25ktpa NdPr) and Heavy Rare Earth (2.5-3ktpa HRE) separation plants in the US, in partnership with the US Government through the US Department of Defense which were previously expected online in FY26. However there have been some project delays due to wastewater management plan challenges, with LYC in discussions with the US DoD around extra capex required for plan changes underway currently.

Mountain Pass HREE Separation Facility — MP Materials (MP.N; Market Cap: US$7.1bn): MP was awarded $35m in 2022 from the DoD to design and build a facility to process HREE at its Mountain Pass mine. The company is currently mining the Mountain Pass deposit producing ~45kt of TREO in 2024 and is ramping production of its light REE separation facility where it's producing ~550t per quarter (vs. nameplate capacity of ~1,500t per quarter). MP is currently advancing detailed engineering work and equipment procurement for the project. Based on recent commentary from the company, they expect to bring heavy rare earth separation production at scale starting in 2026. The HREE facility is expected to produce separated Dy and Tb to be used as feedstock for MP's magnetics facility in Texas.

White Mesa Mill — Energy Fuels (UUUU.A; Market Cap: US$1.2bn): Energy Fuels, traditionally focused on uranium, has been operating the White Mesa processing facility in Utah, the largest uranium processing facility in the US, for 40+ years. Given the sites history processing uranium, the company has expanded its capabilities to process monazite ores, traditionally high in radioactive content, to produce high-purity REE oxides and achieved initial NdPr oxide production in 2024. The material is being validated by customers for potential offtake. The company currently processes monazite sand byproducts from Chemours heavy mineral sands (HMS) mines in the Southeast US. The company plans to import monazite sands from other mining projects they are developing in Australia, Madagascar, and Brazil in the coming years.

The White Mesa mill has installed capacity to process up to 10ktpa of monazite and produce up to 1ktpa of NdPr oxide. However, the current offtake agreement is for 1-2ktpa of monazite sands from Chemours. The company is currently piloting REE oxide separations of heavy REEs such as Dy and Tb while also advancing engineering work on a Phase 2 REE expansion to bring total capacity to 6ktpa of NdPr, 225tpa of Dy and 75tpa of Tb oxides, the Phase 2 expansion is not expected to come online until 2028. The company is simultaneously advancing on HMS mining projects to increase monazite sands production, the mining projects are not expected to come online until 2H27.

Caremag Processing Facility — Carester (private): The Caremag facility, which was deemed a strategic project by the EU's Critical Raw Materials Act (CRMA), will be based in Lacq, in southern France, and will be a large-scale REE recycling and refining facility. The plant is expected to be operational by late 2026 and will recycle 2,000 tonnes of rare earth magnets and refine 5,000 tonnes of mining concentrate annually. The plant is expected to produce 600 tonnes per year of dysprosium and terbium oxides, as well as 800 tonnes per year of Nd and Pr oxides. The company has raised €216m to build the facility, which includes €110m from Japan Organization for Metals and Energy Security (JOGMEC) and private-sector Iwatani corporation as well as €106m from from various subsidies and advances provided by the French government. The company has already entered into a long-term agreement with Stellantis in September 2024 to supply over 3,400 tonnes of Nd, Pr, Dy and Tb oxides over the next 10 years.

Carina Project — Aclara Resources (ARA.TO; Market Cap: US$157m): The Carina Project is Aclara's flagship REE project. It is an ionic clay deposit, historically the source for most of the world's supply of HREE, located in Goias, Brazil. The company is testing operations from a pilot plant to process ionic clays and recently announced the successful production of HREE concentrate (here). Moreover, the company recently submitted its Environmental Impact Assessment to the Brazilian government (here) and is simultaneously advancing on a pre-feasibility study and feasibility study, which are expected to be released in 3Q25 and 1Q26, respectively.

The site has a total Mineral Resource of 297.6Mt at a TREO grade of 1.45% (~19.5% NdPr, 2.7% Dy and 0.4% Tb). Annual production is expected to be ~1,350t of NdPr, 163t of Dy and 28t of Tb. Mining operations are expected to begin in 2028 and concentrate will be sent to a separation facility to produce oxides. The separation facility will be located in the US and Aclara is working with the US Department of Commerce to identify the most strategic site for the facility. The company expects to progress on laboratory test work and integrated pilot scale testing for its separation plant in 2025.

Penco Module — Aclara Resources (ARA.TO; Market Cap: US$157m): The Penco Module is Aclara's second REE project, is located near Concepcion, Chile, and is expected to start operations in 2027. The site has a relatively small Mineral Resource of 29.2Mt at a TREO grade of 2.28% (19.4% NdPr, 2.9% Dy and 0.4% Tb). Annual production is expected to be ~126t of NdPr, 45t of Dy and 6t of Tb. The company's pilot plant ran in 2023 and confirmed the processing parameters and process flowsheet design for a future full-scale plant design. In addition, it produced high purity HREE concentrate to feed separation trials. The company applied for its environmental permits in 2024, received a response from the relevant Chilean Environmental agency in September 2024, and submitted a comprehensive report in March 2025 addressing technical observations that were raised.

Goschen Rare Earth Project — VHM (VHM.AX; Market Cap US$34m): The Goschen project is located in Victoria, Australia. The project is one of the only REE projects in development with proven and probable reserves along with its larger mineral resource. The project has total reserves of 98.8Mt at TREO grades of 2.45% with NdPr as ~18.5%, Dy at 2.5% and Tb at 0.4%. The total Mineral Resource is 491.8Mt at TREO grades of 2.14% with NdPr at 19.2%, Dy at 2.3% and Tb at 0.5%. The company has already received its mining license and expects to receive its environmental and workplan approvals within the 2Q25-3Q25 timeframe, with a final investment decision expected to be announced by YE25. Construction and commissioning are expected to begin soon thereafter with initial production slated for 4Q26, pending permit approvals.

While this project seems like it will be one of the first to market, it is important to note that the current definitive feasibility study only covers the sale of concentrate and downstream separation into REE oxides is not yet being considered by the company. The company is planning a staged development of the deposit to limit upfront capital costs with Stage 1 having an expected throughput of 1.5Mtpa for years 1-3 (4Q26-4Q29) before expanding to 5Mtpa beginning in 4Q29. Stage 1 will produce ~4,300tpa of REE concentrate and Stage 2 would increase to 9,000tpa. Using the Mineral Reserves grades this would imply annual production of ~800tpa of NdPr, 108tpa of Dy and ~17tpa of Tb for stage 1 with production of each mineral roughly doubling in stage 2.

Estonia HREE Separation Plant — Neo Performance Materials (NEO.TO; Market Cap: US$421m): The facility is located in Sillmae, Estonia near the company's permanent magnet facility in Narva, Estonia. Engineering and procurement have already begun with equipment delivery expected in 3Q25 and initial commissioning target by YE25. The company has a 30+ year history in REE separation via its ownership in Chinese separation facilities, which it divested earlier this year (here), with plans to leverage this expertise through its pilot scale plant.

Round Top Deposit — USA Rare Earth (80%) (USAR.O; Market Cap: US$421m) and Texas Mineral Resource Corp. (20%) (TMRC.OTC; Market Cap: US$57m): The Round Top Deposit is located in Hudspeth County, Texas, approximately 85 miles southeast of El Paso. It is an extremely low grade mine at 0.063% contained TREO; however, it is an Ionic Clay deposit whose TREO content is more skewed to heavy rare earths. The company expects to overcome the low-grades using a heap leaching process that is seeing up to 80% recoveries using dilute mineral acid. That said, USAR acknowledges that there are a number of risks along the way to bringing Round Top into production and the company does not expect the mine to enter full production for at least another 5+ years. The companies are currently developing a flowsheet, which would lead to a preliminary feasibility study. Upon completion of these studies, the company would then pursue construction of a pilot plant to test processing of stockpiled material from Round Top before moving to a definitive feasibility study and eventual mine construction. The key goal of the mine is to provide long-term feedstock for USA Rare Earth's magnet plant in Stillwater, Oklahoma, however, at this point, it is unlikely that the company will begin full-scale mining/processing operations this decade.

Monte Alto — Brazilian Rare Earths (BRE.AX; Market Cap: US$364m): The Monte Alto project is BRE's most advanced project located in Bahia, Brazil. The deposit sits within the larger Rocha da Rocha mineral district where the company is doing exploratory drilling on further deposits including the Sulista and Pele project. The Monte Alto ore body constitutes monazite sands at surface level and a hard-rock deposit beneath. The surface level monazite sands have a JORC-compliant resources of 25.2Mt at 1% TREO while initial drill results of the hard-rock deposit have intersected ultra-high-grade mineralizations with up to 45.7% TREO. The hard rock mineralization remains open in all directions and drilling is ongoing. The company expects to release a scoping study by year-end 2025. Notably, the current project is focused on producing concentrate to export and sell to 3rd parties for separation/refining with the goal of using the proceeds from these initial concentrate sales to fund a potential move into downstream refining

4. The Downstream: Lack of ex-China HREE Supply Limits Magnet Projects on Uncertain Feedstock

Downstream Permanent Magnet Projects

China's dominance of the REE / permanent magnet supply chain is most evident when it comes to downstream magnet production. For context, in 2024 China produced ~270kt of NdFeB magnets according to Argus. This compares to the 3 US facilities — MP Materials' Ft. Worth, Texas; USA Rare Earth's Stillwater, Oklahoma; and Noveon's Texas facility — that are expected to all be in commercial production by 1H26 and will have an initial combined capacity of 4-5kt, or just ~2% of total Chinese production. Moreover, the major Chinese permanent magnet facilities have capacity to produce 20kt+, with some as high as 40kt, whereas most ex-China facilities are targeting capacity of 1-2kt. This leads to higher costs from, amongst other reasons, a lack of economies of scale and limits the impact on China's hold over this supply chain given the difference in scale.

Importantly, the bottleneck for many of these magnet projects will be sourcing quality separated heavy rare earths. In our view, the lack of ex-China supply for Dy and Tb, coupled with uncertainty on China's willingness to export these materials, results in the relatively low number of ex-China permanent magnet projects. Indeed, China's decision to place export restrictions on Dy and Tb has led to a major decoupling in Chinese vs. European prices for Dy and Tb (Exhibit 14 and Exhibit 15). Given this uncertainty on feedstock, there are also a number of ex-China magnet projects being considered that focus on recycling magnets from, for example, end-of-life vehicles.

Independence Magnet Facility — MP Materials (MP.N;  Market Cap: US$7.1n): The Independence Magnetics facility is located in Ft. Worth, Texas, and will have 1,000t of permanent magnet nameplate capacity. The company has already begun producing magnets from a pilot plant located at the facility and plans to begin commercial magnet production by YE25. The company is currently stockpiling separated HREE, to use as feedstock for the initial ramp up of production with the goal of transitioning to using separated heavies from the HREE separation facility beginning in 2026. The company has signed a long-term contract with GM to sell magnets produced at the facility.

Estonia Magnet Facility — Neo Performance Materials (NEO.TO; Market Cap: US$421m): Neo's permanent magnet facility located in Narva, Estonia is projected to have an initial production capacity of 2,000t with plans to scale production to 5,000t annually. Construction of the facility was completed in 2025 and the company announced in April 2025, that it had produced 18,000 EV traction motor grade magnets. The magnets were shipped to a Tier 1 traction motor customer where they will be assembled into traction motors for performance testing by the Tier 1 customer and OEM. The approval process is scheduled to be complete in 1H26, with mass production expected to begin later that year.

Stillwater Magnet Facility — USA Rare Earths (USAR.O; Market Cap: US$421m): The Stillwater, Oklahoma magnet facility is currently planning to begin commercial production in 1H26. The facility will have an initial annual capacity of 1,200t with the potential to expand the facility to 4,800t of magnets. The company's Innovation Lab began commissioning in 1Q25 and is expected to begin prototype shipments to customers in 2Q25. The company has an initial supply chain for feedstock in place for REE metal with ex-China suppliers in South Korea and the US to support manufacturing through 2027. The facilities initial focus will be on a more diverse set of non-EV customers to accelerate path to revenue.

MagFactory Project — MagREEsource (private): The MagFactory project is located in Noyarey, France and has been deemed a strategic project by the EU's Critical Raw Materials Act (CRMA). The plant uses hydrogen to recover a powder from end-of-life magnets that can then be directly reused to manufacture new magnets with a 91% reduction in carbon footprint as compared to those produced by Chinese mining extraction. The company has developed a pilot plant that began operations in September 2024 and is expected to produce 50 tonnes per annum of high-performance magnets. The company is also advancing on a commercial scale plant with design and construction having already begun. Production at the full-scale plant is expected to begin in 2027 with initial production capacity of 500 tonnes scaling to 1,000 tonnes by 2030.

HyProMag Recycling Facilities — Mkango Resources (MKA.L;  Market Cap: US$81m): HyProMag is a subsidiary of Mkango who is developing a number of recycling facilities, namely in the UK, Germany and US, using their patented Hydrogen Processing of Magnet Scrap (HPMS). The UK facility has already achieved pilot production and commercial production is targeted by the end of 2Q25, with initial production of 25-30tpa scaling up to targeted 100-330tpa. The facility in Germany is also targeting 100-330tpa of annual production and commercial production is targeted for 2025. Lastly, the US facility would be larger with targeted production of 1,000tpa starting in 1H27, detailed engineering for the production is expected to begin in mid-2025.

San Marcos, Texas Magnet Facility — Noveon(private): Noveon produces permanent magnets in Texas from its magnet facility that was commissioned in 2021, following successful pilot production in 2016 and equipment installation in 2020. The facility has ~2,000t of annual magnet making capacity and produces its patented EcoFlux magnets that the company claims can be produced with everything from virgin materials to 100% recycled, end-of-life feedstock. The company's patented tech requires 20% less HREE than traditionally produced magnetic materials and despite the lower HREE content is able to perform at higher temperatures allowing the magnets to maintain their properties — even at temperatures above 100 degrees Celsius.

Yesan, Korea Magnet Facility — JS Link (private): JS Link is establishing a permanent magnet plant in Yesan, Chungnam with projected annual capacity of 1,000t. The company expects to be a manufacturer of high-performance permanent magnets utilizing rare earth materials sourced and processed by partners in Western and allied countries, including the US, Australia, and the UK. The company is expected to increase capacity to 5,000t at a later date and recently announced they are in conversations with Lynas to source REE materials for trial production.

5. Shift in China REE Policy and Potential Responses

China's control over rare earth supply has become a calibrated yet assertive tool for strategic influence. Its near-monopoly of the supply chain means rare earths will remain a significant bargaining chip in trade negotiations.

China is reshaping its rare earth export controls. Among 17 rare earth elements, China has imposed export controls on seven heavier ones (Dy, Tb, Sm, Gd, Lu, Sc, Y, Exhibit 17) and their processed products (including alloys and magnets) since April 4, 2025. China is increasingly leveraging its rare earth supply chain dominance. Recent actions point to systematic efforts to refine mechanisms so that it could become a more effective strategic lever.

1. Tightening of export licenses. Key rare earth and magnet-related materials are subject to export license requirements. This enables China to regulate volume, destination, end-use, and recipient companies, thus enhancing its ability to exert targeted influence on certain entities.

2. Export tracking system to strengthen oversight of finished magnet exports, which includes end-user declarations, customs harmonization codes, and re-export risk reporting. This infrastructure could be improved over time to monitor, trace and, if needed, restrict outbound rare earth flows.

3. Warnings to trading partners:China has formally warned certain trading partners to avoid facilitating indirect exports of rare earths or magnets to third-party countries, signaling a preemptive stance to prevent circumvention of its controls.

Exhibit 17: Rare Earth Elements

Why the shift in strategy?China had largely refrained from using rare earths as a lever (even during the 2018-19 trade tensions) due to concerns over: (1) global confidence in Chinese supply chains and (2) whether rare earth leverage would endure. We believe the reduced hesitation in leveraging rare earths stems from two key developments:

• Rising trade tensions and tech restrictions: Ongoing tech sanctions, export controls, and tariff volatility have diminished previous concerns about undermining foreign investment in China. The overall global trade environment reduced the risk of deploying rare earth leverage.

• Strengthened supply chain dominance: Alternative rare earth refining and magnet production capacity buildup remained slow in the past 5 years due to economic, technical, and environmental hurdles. China now controls over 85% of global rare earth refining and 90% of NdFeB magnet production, critical for national defense, wind turbines, and electric vehicle motors.

Can China effectively wield this lever? While China's export control regime is advancing, it remains in its early stages. Unlike the US's refined and time-tested "small yard, high fence" strategy, China's system still lacks institutional depth and international compliance frameworks like the Wassenaar Arrangement. Its export control institutional framework, taking shape only since 2021, is still evolving. China is developing sophisticated tracking and licensing systems, which, while they cannot yet match US-style precision, suggest rapid progress (Exhibit 18).

What is the end game? “Leverage for tech and geopolitical reciprocity.” The current rare earth controls are as much about testing mechanisms as they are about immediate economic impact. Yet ultimately, China appears to be establishing a calibrated system of deterrence, where strategic elements (such as rare earths) are used to reshape the cost-benefit calculus of siding with the US export control regime. It will likely respond selectively to tech restrictions imposed by the US and its allies, mirroring the “small yard, high fence” logic. For example, if a US ally were to block chips and lithography tools, China may target critical inputs like rare earths to that country. This reciprocal escalation seeks to deter full alignment with US policies and reshape global tech and trade dynamics.

What steps can China take to maintain its REE dominance?

Rare earths and permanent magnets are China's strongest bargaining chip in an increasingly multipolar world. In our view, China has a strong incentive to continue controlling this supply chain and take actions to mitigate likelihood of future ex-China supply coming to market.

Upstream (Mining & Refining) — China has already taken a number of steps to reduce the likelihood of ex-China supply coming to market, the most notable of which was a 2023 ban on the export of technology and equipment used in REE separation (part of refining process), REE metal production, and rare earth permanent magnet production (here). Despite these bans, companies like MP and Lynas have been able to advance their respective refining capabilities suggesting the know-how still exists/can be developed outside China, despite the tech/equipment ban.

As noted above, there are a robust number of projects outside of China in the pipeline. That said, most are targeting initial production for 3-5 years from now and virtually none of these projects have been sanctioned by their respective boards, meaning construction, which itself is a roughly 2-year process, is yet to begin. China's upstream production of refined rare earth products is predominantly via state owned enterprises and is based on a quota system decided by the Chinese government twice a year. The government is using this quota system to limit illegal production which used to be 50% of China supply and control the environmental impacts.

Downstream (Magnets) — China's dominance over the magnet portion of supply chain is likely harder to crack, or at the very least will take longer. Not only has China advanced on the technological know-how, but given the lack of ex-China supply of heavy rare earths there are very few magnet projects in the pipeline due to the uncertainty on supply of raw materials. As a result, HREE refining projects will need to come to market first before the pipeline of magnet projects expands, and with limited ex-China capacity coming online this decade, China's dominance in magnets should persist.

While the technological advancements and ban on export of magnet making tech/equipment provides a further barrier to entry, we believe this is more manageable as ex-China producers, such as MP and Neo Performance, have begun trial production of automotive grade permanent magnets. That said, these projects are still moving toward commercial production later this year, are significantly smaller than Chinese facilities in scale, and will need to prove out their production capabilities at scale.

7. Industry Impacts

Automotive

Magnets are relevant for the Auto sector, particularly EVs. Autos account for ~38% of NdFeB magnets, used in automatic transmissions, throttle bodies, alternators, sensors, seat belts, lights, power steering, cameras, audio speakers, braking systems, and mostly on EVs motors, allowing manufacturers to reduce size and improve efficiency. Some Indian OEMs have shared that REEs usage per EV is 3 kg/per car while it is 100 gms/per car in ICE.

Shortage starting to hit the Auto value chain. Implemented in April 2025, the restrictions are starting to have an impact on global automotive supply chains. OEMs often hold 2-4 months of magnets in inventory, but we suspect that they were running with lower levels in April as the margin hit pushed for leaner operations. Several OEMs plants and production lines have already been shut down because of the shortage, including Ford's production of its Explorer SUV at its Chicago plant for a week in May, Suzuki's production of Swift for a few weeks and several European auto parts suppliers plants, with further impacts expected in the coming weeks as inventories deplete.

Likely no winner in the West. The export control is only on REEs and magnets, not on motors that use those magnets, so an alternative solution is to produce motors in China (which would take some time) and then export them. If the REE restrictions remain in place, Western suppliers would have an urgent need to diversify their sources. China has warned Japan and Korea against re-routing to avoid indirect exports to third-party countries. According to several reports, hundreds of export license applications have been submitted to Chinese authorities since early April, yet only about one quarter appear to have been approved. German carmaker Volkswagen has previously told CNN that its suppliers have been granted “a limited number of export licenses“ and some US OEMs seem to be getting 6-month temporary licenses. But other applications are also being rejected, impacting Indian, US, Japan, and EU OEMs supply.

No quick fix. The implementation of alternative old motor technology would take time and is not energy efficient. The US and Europe are working to reduce their dependence on REEs, the same applies to India as commented by the Commerce and Industry minister in an interview: "India is actively working with stakeholders to develop domestic capacity and reduce dependence on China." But finding and validating alternate sources will take time and will be expensive. The average lead times for new mines continues to rise, ~18 years according to S&P Global. It would be faster to expand capacity in current mines, but today only China, Burma, the US and Australia are mining REEs at scale. The refining side would take less time to be reallocated, but it produces pollution, requiring complex environmental processes in DM, explaining the current location of most processing plants in China and Malaysia. The magnet part of the value chain is the most difficult to replace, given China's know-how, and currently only China and Japan are capable of producing magnets at scale. Europe has reduced its reliance on Chinese REEs from 98% to 46%, according to the European Commission, but current stockpiles may last only until mid-2025. Carmarkers such as BMW and VW are reviewing long-term supply arrangements.

A deja vu of semiconductor shortage in 2021? If China's export controls persist, global light vehicle production forecasts could face downward revisions, similar to the semiconductor shortage that disrupted the automotive industry, which led to 12+ million vehicles being removed from production globally in 2021 (-12% vs. 2019). However, the impact of REEs shortages could be more profound given the lack of quick solutions — rare-earth supply chains are far more concentrated and difficult to diversify.

Increasing risk of a guidance hit in 2Q. It is going to be difficult to avoid guiding for FY25 when 2Q results come in at the end of July. The negative impact from tariffs started in May, will increase in June-July and OEMs may provide a range, assuming different tariffs and REEs scenarios. In any case, Morgan Stanley expects guidance to be below consensus, probably a middle point between the base case (10% auto tariffs) and a "25% tariff" scenario.

Risks and opportunities. A value chain disruption is a risk to consider in a fixed cost business, as the whole Auto value chain would be impacted. On the flip side, China's minister of Commerce said that China is willing to accelerate the examination and approval of REEs exports to EU firms if their applications meet requirements and could be linked to some progress in negotiations on EU tariffs on China EV exports. And this restriction could be a catalyst for a faster resolution of US-China trade tensions (e.g., President Trump and President Xi's mutual visit invitations), with room to impact tariff risk perception.

Humanoids

The Embodied AI/Humanoid Robot revolution further highlights the S/D risks. The average humanoid has around 0.9kg of rare earth metals. Humanoids likely just scratch the surface of ultimate magnet demand. Humanoids could create a significant uplift in demand for critical minerals — especially rare earths — and cumulatively add up to US$800bn of incremental demand across covered critical minerals by 2050. The take-up of humanoids will start to gather pace from 2035 onward (MSe) and as a proportion of 2030 demand for rare earths (specifically NdPr), humanoids could add an additional 40%/110%/167% in 2040/2045/2050.

What's the US response? While the situation remains fluid, we understand there have been stoppages and other disruptions from Chinese supply. Auto companies source permanent magnets up to Tier 4/Tier 5 suppliers. Many are scrambling to gather intelligence on the volatile supply situation and seeking possible alternatives and contingencies - both near term and longer term. One expert from a US-based OEM says the US Inflation Reduction Act was a "huge help" to re-industrialize battery making in the US. Overall, however, there has been a lot of talking about the issue for the past 3 or 4 years, with very little action, until recently.

Few near-term alternatives. From our discussions, it appears the US will likely remain dependent on Chinese rare earths for the foreseeable future. The average lead times for new mines continues to rise, which could pose challenges for meeting the rising demand for critical minerals. According to S&P Global, the average lead time for mines operational between 2020 and 2024 reached ~17.8 years.

National security concerns. In addition, as alternative sources of permanent magnets and critical refined rare earth materials ramp up, there will be issues of quality, economics and environmental sensitivity that must be considered. In other words, there is no quick fix here. This is likely a multi-decade problem. "There has to be a start somewhere and it has to be in the US and if we don't do it now we never will," according to a senior supply chain executive of a major US auto company we recently spoke with.

Aerospace and Defense

REs are Critical Inputs to Many Defense Products

The defense industry leverages rare earths for a wide range of products, including fighter jets, surface ships, submarines, radar systems, missiles, lasers, satellites, and UAVs. A single F-35 fighter jet contains ~920 pounds of rare earths; an Arleigh Burke-class DDG-51 destroyer contains ~5,200 pounds; a Virginia-class submarine contains ~9,200 pounds.

A Range of Defense Applications

Rare earths feature properties that are well-suited for a range of defense applications. Key use cases include powerful permanent magnets, energy amplification, and energy storage. Neodymium iron boron (NdFeB) magnets, for example, are considered the world’s strongest permanent magnets and are leveraged broadly in Defense systems. Samarium cobalt (SmCo) retains its magnetic strength at high temperatures, making it particularly attractive for missiles / munitions and aircraft systems.

Still, DoD Demand for REs is <0.1% of Global Demand

Despite significant use of rare earths across US defense products, US Defense Department (DoD) demand for rare earths represents <0.1% of global demand, according to the US Government Accountability Office (link). As a result, DoD has more limited market influence, including over the supply chain for rare earths.

DoD is Developing a Domestic Mine-to-Magnet Supply Chain

The Defense Department is working to foster US-based rare earth mining and processing capacity. Per the 2024 National Defense Industrial Strategy, DoD established a target to develop a ‘mine-to-magnet’ rare earth supply chain to satisfy US defense demand by 2027 (link). Since 2020, DoD has awarded >$400mn to help companies build a domestic rare earth supply chain. Meanwhile, DoD’s Industrial Base Policy office is reportedly overseeing a rare earth investment strategy with the goal of developing sourcing, separation, processing, and manufacturing capacity in the US.

LMT Recent Commentary on Rare Earth Supply

Lockheed Martin (LMT), on its 1Q25 earnings call, noted that a potential disruption in rare earth supply should not impact delivery commitments in 2025 given the quantity of rare earths already integrated into the company’s value chain. LMT also highlighted there are stockpiles available, and that the company – and its supply chain – are “constrained from using Chinese inputs of any kind from any source into our products and services.” That said, the company in its 1Q25 10-Q flagged that recent government actions could constrain material availability over time.

8. Rare Earths in Humanoids

Significantly more NdPr will be required to meet humanoid demand, potentially increasing deficits in a market where supply security remains a key issue for Western countries (China supplies >74% of global refined supply by 2050e).

9. Rare Earths Overview

Rare earths are a group of 17 elements that are used in various clean energy, high-tech, and defense applications. Though generally as abundant as base and precious metals, these metals are not easily found in large, high-concentration mineable deposits. They have unique chemical, optical, electrical, magnetic, and metallurgical properties. As these metals are found together in nature, and their properties are similar, separation into oxides is technically challenging.

Light vs. heavy rare earths. Rare Earth Oxides (REO) are commonly divided into two categories — light and heavy REO. Light REOs (La-Sm) are found more strongly concentrated in nature and face fewer shortages. Example of light REOs are cerium, lanthanum and neodymium, whereas heavy REOs (Eu-Lu) include europium, terbium, and dysprosium. Industry experts believe Chinese deposits contain about 80% of global heavy earth reserves and account for an even larger portion of separated HREE production, with non-Chinese producers like Lynas having only recently produced their first separated HREE oxides in 1H25 (here).

Types of REO ores: REO occur primarily as two types of ores, Bastnasite and Monazite. Bastnasite deposits in China and the United States constitute the largest percentage of the world's rare earth resources, MP's Mountain Pass mine is a Bastenite ore body. Monazite deposits in Australia, Brazil, China, India, Malaysia, South Africa, Sri Lanka, Thailand, and the United States constitute the second-largest segment. Monazite ore often has a higher concentration of radioactive thorium in it, and therefore needs more processing. However, monazite ore bodies generally include higher proportion of heavy rare earth elements.

Exhibit 28: The NdFeB magnet supply chain begins with mined material that is separated into individual REE oxides before being reduced into metal which is ultimately magnetized for end-use applications

Major Use Cases for Rare Earths

In the past decade, applications have grown significantly, driven by use of rare earth metals in miniaturization of electronics. Rare earth elements are critical to new-age electronics, clean energy technologies, high-tech, and defense applications. Substitutes are available, but are generally less effective. However, the single fastest growing use for rare earth elements is in permanent magnets, which contain light rare earths like Nd and Pr (generally in a 75/25 ratio for Nd and Pr, respectively), as well as heavy rare earths like Dy and Tb (Exhibit 30). Permanent magnets are used in green energy applications such as electric vehicle motors and wind turbines as well as in nascent technologies like humanoids and physical AI.

Exhibit 29: Typical NdFeB magnet compositions require the use of Nd, Pr, Dy and Tb

Total rare earth demand was estimated at ~211.5kt in 2024, according to Argus, and magnet materials (Nd, Pr, Dy and Tb) accounted for ~37% of rare earth demand by volume while relatively low value REEs, like Cerium and Lanthanum, still account for the lion's share of demand by volume at ~55%.

Global Actions to Mitigate China Dominance

Outside of China, governments have become increasingly aware of China's influence over the REE market (exhibit 34) and have begun implementing various programs to incentivize projects and speed up time to market — primarily in the form of government grants and critical mineral funds.

US is leading with several initiatives:

• Defense Production Act (DPA) investments, allocated $750mn ($439m for just REE projects since 2020) to support domestic production of critical minerals via grants provided by the Department of Defense (DoD).

• Inflation Reduction Act (IRA) offers 10% production tax credits (45X) and funding for up to 30% of capex for critical material projects (48C). US REE producers, such as MP Materials are currently receiving 10% tax credits to offset mining and refining costs of NdPr oxide and received a ~$60m grant for construction of their Ft. Worth, TX magnetics facility.

• Department of Energy's (DOE) loan program office (LPO) and critical materials institute provides loans and guarantees to critical mineral processing and extraction projects under its til 17 Clean Energy Financing Program. The CMI is a research hub funded by the US government to advance technologies for rare earth recovery, substitution and recycling

The EU as whole, as well as several member states, have also established national funds to bolster their critical mineral projects.

• €100 million European Bank Fund: The European Bank for Reconstruction and Development (EBRD) and the European Union (EU) have mobilized up to €100 million in investments toward critical and strategic raw materials. More details here.

• €500 million France Mineral Fund: In May 2023, France committed €500 million for critical minerals and metals as part of a public-private investment fund. The fund seeks to raise an additional €1.5bn from private investors and will be managed by InfraVia. The fund is also expected to work in coordination with the German and Italian initiatives below. More details here.

• €1 billion German Fund: The raw materials fund will be managed by KfW development bank with a focus on projects in extraction, processing and recycling of critical minerals. The fund was created in 2024 and is set to make investments over the next four years.

• €1 billion "Made in Italy" fund: The goal of the "Made in Italy" initiative is to support specific projects and to encourage the creation of national champions within the critical minerals space. The fund, along with others, has the goal to increase the EU's capacity to meet 10% of annual consumption from locally extracted materials.

Substitution Risks

The long-term outlook for REE demand is expected to grow driven by the use of permanent magnets in growing applications such as EVs, humanoids/robotics, and clean energy. However, emerging technologies present credible substitution risks. Advances in material science and motor design are enabling alternatives that could reduce reliance on REE-based magnets over time.

Iron nitride (FeN) offers high remanence (magnetization left behind in a ferromagnetic material after an external magnetic field is removed) similar to REE magnets but suffers from low coercivity (measure of the ability of a ferromagnetic material to withstand an external magnetic field without becoming demagnetized), making them prone to demagnetization. They require new motor designs, but collaborations with major OEMs like General Motors suggest strong commercial interest. FeN could become a viable REE-free alternative if its engineering limitation is resolved.

Manganese bismuth (MnBi) magnets deliver torque comparable to NdFeB but at a 60% increase in volume and 65% more weight. However, they can potentially reduce motor cost by 32%, making them attractive for low-cost or space-flexible applications. MnBi may replace REEs in budget EVs or segments where cost outweighs efficiency.

NdFeB without heavy REEs (e.g., Dysprosium) eliminating heavy REEs like dysprosium reduces high-temp coercivity. However, this can be mitigated through design workarounds like thin segmented magnets, carbon-fiber rotor reinforcement, and active cooling as demonstrated in a 100 kW prototype by Oak Ridge National Lab. Even partial substitution through smart engineering can significantly cut REE use, particularly for HREE, which is the current choke point in the greater NdFeB magnet supply chain.

FeN, MnBi, and heavy REE free NdFeB highlight credible substitution paths, with growing OEM interest despite certain trade-offs. In addition, options like Ferrite and SmCO magnets, magnet-free switched reluctance motors and axial flux designs further expand the range of REEs alternatives. Together, these developments present a meaningful risk to sustained rare earth demand in EV motors.