China's rare-earth magnet export pause is expiring November 10

China's rare-earth magnet export pause is expiring November 10

John KoetsierNovember 7, 2026
ShareXLinkedInFacebook
Each humanoid robot contains 30+ electric motors, each requiring neodymium-iron-boron permanent magnets, and China controls 94% of global production. With China's export control pause expiring November 10, this is no longer an abstract supply chain risk.

So many motors, so many magnets

Every humanoid robot has approximately 30 electric motors: one per joint, one for each degree of freedom that lets it walk, reach, and manipulate. Every one of those motors requires a neodymium-iron-boron permanent magnet. Add it up: 2 to 4.5 kilograms of rare-earth magnets per robot, roughly 8 to 10 pounds.

China makes 94% of them.

That's not a hypothetical supply chain risk. It's the current state of the industry, as humanoid robots begin shipping at scale and the investment thesis around magnets starts to materialize.

The magnet problem in numbers

Sintered neodymium-iron-boron (NdFeB) permanent magnets are not interchangeable with cheaper alternatives for high-performance joint actuators. The torque density and power-to-weight ratios of rare-earth magnets are what make it possible to put 30 motors into a robot that weighs 50 to 90 kg and still moves.

The breakdown per robot, per MP Materials CEO James Litinsky:

  • approximately 1.3 kg of NdPr (neodymium-praseodymium)
  • 35 to 79 grams of dysprosium required for heat resistance in high-load joints like hips and shoulders
  • terbium as an additional additive

Shockingly, that's more rare-earth content per unit than a typical electric vehicle.

At projected humanoid scale, this becomes an industrial problem.

Morgan Stanley projects we'll have over 1 billion humanoid robots by 2050, implying $800 billion in cumulative critical mineral demand, with rare-earth magnets as the central input.

China's position in the stack

The numbers are stark.

China controls approximately 60 to 70% of global rare-earth mining. It runs approximately 90% of rare-earth processing capacity. It produces approximately 94% of sintered NdFeB permanent magnets globally.

The integration goes much deeper than just mining. China has built a fully stacked industrial chain:

  • rare-earth separation
  • alloy production
  • magnet manufacturing
  • motor production
  • precision robotic components
  • robot assembly

And all of that matters.

A humanoid robot assembled in Japan, Germany, or the US may still contain Chinese rare-earth materials when traced at the magnet and motor level. China Rare Earth Group held a major industry event in May 2026 specifically to integrate state-backed supply chains directly into humanoid robotics and embodied AI ecosystems: a deliberate policy move to lock Western robot makers into Chinese supply dependencies.

And there's an immediate element: China's rare-earth export control pause, implemented under a bilateral trade truce with the US in November 2025, expires November 10, 2026.

As of this right now, China has not published a formal renewal notice. An informal extension to January 2027 has been reported but not confirmed. Three days from now, that pause could lapse.

The domestic alternative

The US is building alternatives, but not fast enough for the current deployment wave.

MP Materials (NYSE: MP) is the most visible play. The company operates the Mountain Pass mine in California, the only scaled rare-earth mining and processing facility in North America, running a vertically integrated mine-to-magnet model. The Department of Defense has put $400 million in preferred stock into the company and committed to purchasing 100% of output from an upcoming facility at a minimum price floor of $110/kg for NdPr oxide: a 10-year government backstop.

MP's Q3 2026 numbers were strong: NdPr production of 1,018 metric tons, up 41% year-over-year. Revenue of $155.8 million, up 176%, cash position of $1.34 billion, and first commercial magnet shipments to GM began in Q4 2026.

Apple deliveries start H2 2027.

The problem is timing.

MP's larger 10X magnet facility won't be commissioned until 2028. USA Rare Earth, which received up to $1.6 billion in CHIPS Act funding, is in a similar position. Noveon Magnetics in Texas is operational but small-scale. So western domestic NdFeB production at humanoid-relevant scale doesn't arrive until 2028 at the earliest.

The gap between thesis and reality

This timing mismatch explains the puzzle in this week's Motley Fool piece: why MP Materials stock is down 62% from its all-time high even as the secular thesis strengthens.

The investment thesis is sound, but the revenue isn't here yet. First commercial shipments to GM just started, Apple deliveries haven't begun, and humanoid robot supply contracts haven't been announced.

The demand surge that justifies the valuation is priced in; the actual purchase orders aren't signed. Three analyst cuts in October 2026, from Baird, Morgan Stanley, and Goldman Sachs, all maintained buy ratings while reducing targets.

The bull case isn't dead. It's waiting on a signed supply agreement with Tesla, Boston Dynamics, Figure AI, or any of the major humanoid makers currently scaling production.

What to watch

November 10 is the immediate date.

Whether China formally renews its rare-earth export control pause matters for near-term supply chain stability for every non-Chinese robot manufacturer currently operating.

Beyond that, the structural reality is straightforward. The humanoid robot industry is entirely dependent on rare-earth magnets it currently cannot source domestically at scale. The US is building the infrastructure to change that, but the gap between today's deployment wave and 2028's domestic capacity is real, and it's not shrinking on its own.

Every humanoid robot shipping before 2028 is, at the magnet level, a Chinese supply chain story.

Sources

Share this article

ShareXLinkedInFacebook

Written by

John Koetsier
John Koetsier

I tell stories with data and build startups. I've raised angel and VC funds, led software companies and startups, and chronicled the rise of the mobile economy. Currently I write for Forbes, host several podcasts, and am starting up HumanoidMegaHub. I was TUNE's mobile economist and Singular's VP of Insights, built the research team at VentureBeat, and managed teams creating software for partners like Intel and Disney. In past lives I built websites, e-learning platforms, and mobile apps, and, a long time ago, was named to Folio's top 100 of the media industry's "most innovative entrepreneurs and market shaker-uppers.”