U.S. Rare Earth Magnet Industry Surges? But Qualified 10% Global Share Likely Waits Until 2031?

Highlights

• China held ~94% of global sintered rare-earth magnet manufacturing in 2024, leaving the U.S. producing just ~300 tonnes against ~48,000 tonnes of domestic demand.
• REEx base case projects the U.S. reaching ~10% of global sintered NdFeB output around 2031–33, with 20% share pushed to 2035 or later.
• Heavy rare earths dysprosium and terbium remain the hardest constraint, as mixed carbonate from projects like Serra Verde must still be separated, refined, and qualified.
• A de facto Western NdPr price floor near $110/kg is forming through government contracts, equity investments, and strategic offtakes rather than conventional market discovery.
• Announced nameplate capacity is expanding far faster than qualified output—commissioning dates are the starting gun, not the finish line.

---

America is finally making rare-earth magnets again. Independence is another matter entirely. China accounted for roughly 94% of global sintered rare-earth magnet manufacturing in 2024, according to the International Energy Agency (IEA), alongside 91% of magnet-rare-earth refining. Rare Earth Exchanges® estimates the broader global rare-earth permanent-magnet market at approximately 385,000 tonnes in 2025. That broader consumption figure should not be confused with the narrower global production denominator for high-performance sintered NdFeB magnets.

The U.S., meanwhile, consumed approximately 48,000 tonnes of NdFeB magnets in 2025, against only around 300 tonnes of domestic production. REEx uses roughly 48,000 tonnes as its working 2026 demand baseline pending better full-year data. Measuring America's true exposure is difficult because magnets routinely enter embedded inside motors, vehicles, electronics, and machinery rather than as identifiable magnet imports.

REEx base case: sustained, qualified U.S. production approaching 10% of global sintered NdFeB output is most plausible around 2031–33. Twenty percent looks more like 2035+ and requires another generation of capacity beyond today's strongest announced pipeline.

REEx Insight: Stop Counting Magnet Factories

The West keeps counting furnaces. China counts industrial ecosystems. The delta between the two should concern anyone rooting for Western reindustrialization. A magnet plant does not become independent because somebody installs presses and sintering furnaces. A resilient supply chain requires:

ore/feedstock—including sustained heavy-REE access → separation → NdPr/Dy/Tb oxides → metals → alloys → sintered magnets → machining/coating → customer qualification → motors/components.

The weakest link determines actual production.

REEx has repeatedly examined this problem in The Midstream Trap, China's Magnet Empire and America Is Financing a Rare Earth Supply Chain Faster Than It Can Industrialize One. The bottleneck moves downstream as each preceding problem is addressed.

That is why REEx applies a 12–24 month—or potentially longer—analytical pad for commissioning, yield improvement, and customer qualification at major greenfield magnet projects. A commissioning date is the starting gun, not the finish line.

America's Magnet Demand Is Bigger Than Customs Data Suggest

REEx estimates approximately 48,000 tonnes of U.S. NdFeB consumption in 2025. Our modeling indicates tens of thousands of tonnes may reach America embedded inside imported products; approximately 30,000 tonnes is a working REEx estimate, not an observable customs statistic.

Consider an automobile. A Chinese magnet can be sold to a motor manufacturer in South Korea, Mexico, Canada, or Japan, incorporated into a traction motor, and eventually shipped to an American assembly plant. Customs sees a motor or vehicle component—not necessarily the Chinese magnet buried inside it. The real exposure sits several tiers down the bill of materials.

The Mine-to-Magnet Clock

Platform	Announced / Stated U.S. Capacity	REEx Qualified-Scale Window
MP Materials (NYSE: MP)	Independence ~3,000 tpa; 10X 10,000 tpa planned	2029–30+
USA Rare Earth (Nasdaq: USAR)	Stillwater ramp; broader U.S. platform targeting up to 10,000 tpa	2029–32
VAC	2,000 tpa initial; expansion contemplated	2027–29 / 2030+
Noveon Magnetics	~2,000 tpa stated capability	Ramping now
Evolution Metals & Technologies (Nasdaq: EMAT)	10,000 tpa stated future capacity; 6,000 tpa high-performance sintered	Unproven—watch commissioning/qualification

The table exposes the central problem: announced nameplate capacity is expanding considerably faster than qualified output.

MP Materials remains the strongest vertically integrated domestic candidate. Independence has restored commercial U.S. NdFeB production, while 10X is scheduled to begin commissioning in 2028. But qualification, yields, customer ramp, and heavy-rare-earth integration follow.

USA Rare Earth commissioned Stillwater in 2026, targeting 600 tpa by Q4 2026 and 1,200 tpa in Q1 2027, with additional U.S. expansion planned.

VAC's Sumter operation matters because VAC brings decades of magnet-manufacturing know-how. Energy Fuels (NYSE American: UUUU) potentially connects White Mesa separation, ASM's Korean metals/alloys capabilities, and VAC's magnets into another Western industrial network.

Lynas Rare Earths (ASX: LYC; OTC: LYSCF) remains heavily oriented toward Japan and Asian customers but is increasingly important to U.S. strategy. Its roughly $96 million, four-year U.S. arrangement covers light and heavy rare-earth oxide supply, while its planned Malaysian magnet venture with South Korea's JS Link illustrates why allied capacity—not simply U.S.-located factories—will matter.

Heavy Rare Earths Set the Speed Limit

The hardest constraint remains dysprosium and terbium. USA Rare Earth's acquisition of Serra Verde could be transformational because Brazil's Pela Ema operation already produces mixed rare-earth carbonate containing NdPr, Dy, and Tb. But mining is not separation. Pela Ema must still expand while maintaining recoveries, reagent and water management, and consistent Dy/Tb-bearing feedstock. And mixed carbonate does not magically become separated oxide.

Washington is putting extraordinary financial firepower behind this chain. The latest Serra Verde structure totals $1.55 billion—$750 million in government investment, at least $300 million in forward purchases, and up to $500 million in credit. Yet Pela Ema still produces mixed carbonate rather than separated oxides, metals, alloys, or magnets.

USA Rare Earth's Carester investment may help bridge that gap, but Caremag itself must commission and ramp.

Energy Fuels has made meaningful progress. White Mesa-produced Dy and Tb oxides have passed qualification with a major Japanese magnet manufacturer, while commercial circuits targeting approximately 120 tpa Dy and 20 tpa Tb are under construction. Feedstock availability and sustained commercial operation remain the test.

REEx Simulation: Beijing Does Not Need an Embargo

China has a subtler strategic option than simply turning exports off. It can permit enough material to prevent catastrophic shortages while maintaining licensing uncertainty—and retain the option to compress prices as Western competitors approach scale. Too little Chinese supply accelerates Western substitution; too much makes Western projects economically harder to justify.

Washington is responding by changing the payoff matrix through equity, loans, procurement mandates, price protection, and long-term offtakes. The Pentagon's $110/kg NdPr floor for MP Materials is the clearest example. But the pattern is broadening.

REEx cautions against calling $110/kg an industry-wide spot price—but the evidence is increasingly difficult to dismiss. What began as MP's company-specific contractual mechanism is appearing as a reference point in additional ex-China arrangements, including Serra Verde.

A market architecture appears to be forming around roughly $110/kg NdPr. This is not conventional commodity price discovery. Government guarantees, strategic offtakes, and national-security considerations are helping create the market itself. Continued clustering could nevertheless turn $110/kg into a de facto Western reference level or policy-supported floor.

China can also push the battlefield downstream. Restrict magnets and manufacturers may source Chinese motors, actuators, and assemblies instead. Control of the magnet can migrate into control of the component containing it.

Three Different Races

Investors should distinguish three very different benchmarks.

U.S. production versus U.S. consumption measures domestic self-sufficiency. Against approximately 48,000 tonnes of current U.S. demand, 10,000 tonnes of domestic output would already represent roughly 21% of today's U.S. consumption. U.S. production versus global sintered NdFeB production measures America's actual manufacturing share. Ex-China production versus global production measures allied resilience and includes established Japanese, South Korean, and European manufacturers.

REEx estimates the broader 2025 global rare-earth permanent-magnet market at approximately 385,000 tonnes. Ten percent of that broader market equals 38,500 tonnes and 20% equals 77,000 tonnes. But REEx does not automatically treat those figures as precise sintered-NdFeB production-share thresholds: the product definitions and denominators differ.

The conclusion is nevertheless clear. America must reach sustained qualified output measured in tens of thousands of tonnes—not merely announce tens of thousands of tonnes of factories.

And the goalposts are moving. EVs, robotics, industrial automation, defense systems, drones, and other magnet-intensive technologies should expand demand through the 2030s.

REEx Timeline: 2028 Is Not Independence

Optimistic case: approaching 10% of global sintered-NdFeB production around 2029–31, requiring unusually smooth ramps across MP, USAR, VAC, Noveon, and emerging producers plus reliable heavy-REE supply.

Base case: 2031–33 for approximately 10%; 2035+ for 20%.

Downside case: feedstock delays, metallurgical problems, qualification failures, financing constraints, Chinese price pressure, or Dy/Tb shortages push 10% toward 2034–36.

Japan, South Korea, and Germany already possess meaningful non-China magnet-manufacturing expertise. America does not need every magnet stamped Made in USA. It needs enough trusted, qualified allied capacity that no single government can switch the industrial system off.

Why REEx Keeps Moving the Goalposts Back to Reality

This is not an argument against America's industrial surge under Trump 2.0—it is an argument for sustaining it. Washington is moving with an urgency rarely seen in modern U.S. industrial policy: equity investments, loans, price floors, long-term offtakes, procurement restrictions, and billions of dollars directed toward MP Materials, USA Rare Earth, Serra Verde, Energy Fuels, VAC, and other supply-chain nodes. That is real progress.

Yet declaring victory because factories have been announced would repeat the complacency that helped create America's vulnerability. China still possesses formidable advantages across separation, metallization, alloy making, magnet manufacturing, technical know-how and—most dangerously—Dy/Tb processing. REEx's role is to keep pressure on government, industry, and investors until announced capacity becomes qualified, sustained industrial output. This effort must survive multiple election cycles, commodity cycles, and inevitable setbacks.

REEx is not forecasting failure. We are distinguishing industrial mobilization from industrial independence. Moving from roughly 300 tonnes of U.S. NdFeB production in 2025 to several thousand tonnes within only a few years would itself represent a remarkable industrial reversal. But a meaningful global manufacturing share requires tens of thousands of qualified tonnes with every link of the chain operating simultaneously.

The West will manufacture magnets years before it achieves magnet security. The danger is no longer that America is doing nothing. The danger is that America mistakes a remarkable beginning for a finished job.