Highlights

• Official trade data captures only 58,000 tonnes of direct Chinese magnet exports.
• An additional 20,000-40,000 tonnes are embedded in EVs, wind turbines, and electronics, revealing a true global dependence of 80,000-100,000 tonnes annually.
• Ex-China magnet production represents just 16% of global output at 22,800 tpa.
• Even if all announced projects reach 75,100 tpa capacity, they require 22,300 tpa NdPr oxide that non-Chinese separation facilities cannot supply.
• The critical bottleneck is rare-earth oxide separation—especially heavy rare earths (Dy/Tb)—not mining or magnet manufacturing, meaning China's strategic control simply shifts upstream as downstream capacity relocates.

---

For years, governments and markets have relied on a deceptively simple statistic to assess rare-earth risk: China exports roughly 58,000 tonnes of NdFeB magnets per year. This figure is routinely cited as a proxy for global dependence on Chinese magnet supply.

The assumption is wrong.

Analysis from REEx Insights™ Magnet Rankings and Capacity Database shows that magnet trade data captures only a visible fraction of China’s true control over global magnet supply. The larger dependency sits upstream and inside finished goods — hidden from trade statistics, yet central to the energy transition.

Two Magnet Supply Chains — Only One Is Counted

There are two distinct magnet supply channels:

1. Direct magnet exports

Finished NdFeB magnets shipped overseas

→ ~58,000 tonnes per year (visible in customs data)

2. Embedded magnets in exported products

Magnets manufactured in China, integrated into EVs, wind turbines, motors, electronics, and industrial equipment — then shipped overseas as finished goods

Only the first category appears in official trade statistics.

The second isstructurally invisible.

This mirrors earlier blind spots in semiconductors, batteries, and solar polysilicon, where dependence was underestimated because critical components were embedded in finished products and crossed borders.

The Hidden Volume: Embedded Magnets

Using capacity-derived modeling that links downstream manufacturing output to magnet intensity, REEx estimates that 20,000–40,000 tonnes of magnets per year are embedded in Chinese exports.

This range reflects uncertainty in product mix and magnet loading, but the order of magnitude is robust:

• EVs alone typically contain 2–5 kg of NdFeB magnets per vehicle
• China exports millions of EVs annually
• Wind turbines, industrial motors, HVAC systems, robotics, and consumer electronics add substantial additional demand

Independent academic and industry analyses increasingly converge on tens of thousands of tonnes per year of embedded magnet flows.

Key point: this is existing demand, not speculative growth.

The Real Number That Matters

When direct exports and embedded magnets arecombined, the true volume of magnets already consumed outsideChina is closer to:

Approximately 80,000–100,000 tonnes per year. This demand does not disappear if manufacturing relocates. It simply reappears as ex-China magnet demand.

Every EV, turbine, or motor no longer assembled in China requires its magnets to be sourced elsewhere — immediately and at scale.

Current Production Reality

According to the REEx Magnet Rankings:

• China magnet production: ~120,500 tpa
• Ex-China magnet production: ~22,800 tpa

Ex-China accounts for ~16% of global NdFeB magnet output.

This is a functioning industry — but a small one, sized for niche programs, not global electrification.

The Chemistry Behind the Bottleneck

REEx links magnet capacity directly to rare-earth inputs using industry-standard assumptions:

• ~30% of magnet mass is rare earths
• ~98–99.5% NdPr
• ~0.5–2.0% Dy/Tb

At current ex-China magnet output levels, this implies annual consumption of roughly:

• **~**6,800 tonnes total REO
• ~6,770 tpa NdPr
• ~68 tpa Dy/Tb

Enough for limited EV programs and defense supply chains — not for mass deployment.

Planned Expansion — and the Problem It Exposes

If all announced ex-China magnet projects were fully funded, permitted, and vertically integrated, total ex-China capacity could reach ~75,100 tpa.

That would imply oxide demand of approximately:

• ~22,300 tpa NdPr oxide
• ~225 tpa Dy/Tb oxide

On paper, this could replace direct Chinese magnet exports.

In practice, it exposes the real constraint.

Processing, Not Mining, Is the Choke Point

Mining is not the bottleneck.

Magnets are not the bottleneck.

Separation is.

Rare-earth concentrates must be refined into ultra-high-purity oxides before they can become metals, alloys, and magnets. This stage is:

• capital-intensive
• chemically complex
• environmentally sensitive
• slow to scale

REEx tracking shows:

• Very few ex-China separation plants operate at a meaningful scale
• Heavy rare-earth (Dy/Tb) separation capacity outside China is minimal
• No public disclosures indicate ex-China HREE processing volumes anywhere near those implied by magnet localization ambitions

The bottleneck simply moves upstream — preserving China’s leverage even as downstream capacity grows.

What the REEx Insights™ Rankings Ultimately Show

• Magnet export data dramatically understates real dependence
• 80,000–100,000 tonnes of magnets are already consumed outside China each year
• Ex-China magnet manufacturing can expand — oxide processing cannot keep pace
• HREE separation is the critical strategic bottleneck, and China has effectively monopolized
• Supply-chain risk shifts upstream, not away

The magnet supply gap is not disappearing.

It is being hidden — and postponed.

Why This Matters Now

As governments mandate EV reshoring, wind localization, and defense supply-chain security, today’s embedded dependence becomes tomorrow’s visible shortage.

The policy failure is not a lack of intent.

It is a misunderstanding of where control actually resides.

REEx data makes that reality unavoidable.