to reach approximately 50,000 tonnes per year by 2030. Current
non-Chinese production sits at roughly 120,000 tonnes of total
rare-earth oxide, and the magnet-grade share of that is substantially
smaller. The arithmetic of the coming five years is therefore unusually
tight — and the question is whether supply can scale fast enough.¹²
magnet applications — predominantly neodymium, praseodymium, dysprosium
and terbium — by 2030, based on the demand estimates in the Joint
projections aligned with Adamas Intelligence's magnet-market outlook.¹
are expected to consume around 120,000 tonnes of NdFeB magnets annually
by 2030 at projected production volumes; and wind turbines, which
together with industrial motors contribute another 30,000 tonnes. The
sum is approximately 150,000 tonnes of magnets, of which about a third —
the 50,000-tonne figure — is rare-earth content.¹
incremental demand but do not dominate the arithmetic. The
energy-transition applications — EVs and wind — are the structural
drivers.
mine production at 390,000 tonnes of rare-earth oxide in 2025. China
contributes approximately 270,000 tonnes, or roughly 69 percent of the
total. The remaining 120,000 tonnes is distributed across Australia
and a handful of smaller producers.²
magnet-grade. A substantial share is shipped to Chinese refineries for
separation and then re-exported — a structural reality that Western
policy-makers increasingly treat as equivalent to direct Chinese
dependence. Only Lynas in Malaysia, MP Materials in the United States
and a handful of smaller operators currently produce separated
rare-earth oxides at scale outside China.
magnet-grade rare-earth supply chain is significantly smaller than the
mine-output figure suggests. Estimates vary, but non-Chinese separated
supply available to Western magnet manufacturers is on the order of
20,000-25,000 tonnes per year of rare-earth content in 2025.
ionic-clay production is scaling: Serra Verde from 2,000 tonnes in 2025
toward 6,500 tonnes by end-2027, Aclara targeting commercial production
in 2028 at Carina, Meteoric advancing Caldeira through pre-feasibility.²
production by 2030 if all three operators execute on schedule.
capacity to 10,500 tonnes per year and adding the Seadrift Texas
facility with an additional 1,000-1,300 tonnes per year of NdPr plus
rare-earth refinery is commissioning in 2026. Additional Australian
ionic-clay exploration is underway.
will take combined magnet capacity to 10,000 tonnes per year by 2028,
and ReElement Technologies and Ucore Rare Metals are building additional
separation and metal-making capacity. The DoD-backed US$439 million in
federal commitments plus allied financing is supporting several smaller
projects.
but strategically important segment. Japanese and European partnerships
with African producers are adding a further supply leg outside Chinese
dependence.
current non-Chinese separated supply is 20,000-25,000 tonnes, the gap to
be closed is approximately 25,000-30,000 tonnes per year. The Brazilian
pipeline alone could contribute 10,000-15,000 tonnes; Australian
expansions another 5,000-8,000 tonnes; U.S. and African additions
perhaps another 3,000-5,000 tonnes combined.
their stated targets on schedule. That range straddles the 50,000-tonne
demand figure — possibly meeting it, possibly falling short by 10-20
percent. The margin is narrow.
sub-complex — the four elements neodymium, praseodymium, dysprosium and
terbium specifically — is less reassuring. Undersupply of dysprosium is
projected to reach 1,800 tonnes per year and terbium 450 tonnes per year
by 2040, with approximately US$7.3 billion of magnet rare-earth-oxide
demand going unmet by 2040 unless new supply grows beyond base-case
expectations.⁴
balance. Serra Verde must complete its ramp to 6,500 tonnes on schedule.
Aclara must secure permits and reach commercial production at Carina.
and into construction. The Viridis-Ionic Poços de Caldas separation hub
must commission and operate at design capacity. Lynas must execute at
capacity.
project execution is historically probabilistic, and any meaningful
delay at one or two of the major projects would open a real supply gap
at the 2030 horizon.
permitting approvals, U.S. federal support for domestic projects,
have to be taken repeatedly across the coming five years, not just once.
across the global non-Chinese pipeline matches current plans, the gap
closes. If execution lags by even 10-15 percent, the gap opens wide
enough that prices stay structurally elevated and Chinese market power
remains a meaningful leverage. Brazilian projects are among the most
important swing contributors. Every tonne of concentrate that Serra
gap. For investors, the lesson is that Brazilian rare-earth execution
risk is also global supp