The Rare Earth Map

Social Capital published a deep dive on critical minerals arguing that the United States is still badly underestimating how short it is. We agree with the framing. What the piece does not do — because it sits behind a paywall and is written for a different purpose — is tell you which listed companies are actually standing in that gap, what they earn, and which of them own something durable versus something that is currently just a press release with a stock ticker attached.

So we built the map. Below is the full public-market universe of rare earth and adjacent critical minerals companies we consider relevant, sorted by how close each one is to producing a real, sellable kilogram. For each: the ticker, the last two fiscal years of reported results, the most recent quarter, and our view.

One picture, one conclusion. China Northern Rare Earth alone is worth more than Lynas and MP Materials combined — the two largest Western producers — and the three Chinese listings together outweigh all ten American ones. The West's response is not one big company. It is a long tail of small ones, nine of which are worth under $1.2bn each.

What the argument actually is

The public portion of the Social Capital piece makes four claims, and all four are checkable:

1. The U.S. critical minerals list has grown from 35 minerals in 2018 to 60 in 2025, driven by AI, defense, and industrial demand. The USGS modeled more than 1,200 supply-disruption scenarios to produce the current ranking.
2. Seven industrial baskets depend on these materials: batteries, the power grid, magnets, chips and displays, defense, agriculture, and general industrial.
3. China controls roughly 85% of global rare earth processing — a position built deliberately after Deng Xiaoping's 1992 declaration that "the Middle East has oil, China has rare earths."
4. The United States has the geology. It does not have the processing. The bottleneck is midstream — separation, metallization, alloying, magnet-making — not in the ground.

The investable version of this thesis is narrower than it sounds. If the constraint is separation and magnets rather than mining, then the companies that matter are the ones who can turn concentrate into a qualified magnet — not the ones who can point at a deposit. That distinction organizes everything below.

Why 2025–26 broke the old pattern

Rare earths have had false dawns before — 2010–11 most memorably, when prices spiked after a Chinese export squeeze and then collapsed, taking a generation of Western juniors with them. Four things are structurally different this cycle.

1. China weaponized the midstream, then partially stood down

On 9 October 2025, China's MOFCOM introduced a 0.1% de minimis rule: any product made anywhere in the world containing Chinese-origin rare earths at 0.1% or more of its value would need a Chinese export licence. That is extraterritorial control over essentially every magnet-containing device on earth. Following the Xi–Trump meeting, China suspended those measures for one year, until 10 November 2026.

Critically, the earlier April 2025 controls were never suspended. Seven elements — samarium, gadolinium, terbium, dysprosium, lutetium, scandium and yttrium — still require case-by-case MOFCOM approval for every export. Those are the heavies. They are also the ones with no meaningful Western supply.

2. Washington started setting prices, not just writing cheques

The July 2025 Department of Defense partnership with MP Materials is the template: a $110/kg floor price on NdPr for ten years, a ten-year offtake of 100% of a new magnet plant's output, $400M of preferred stock, and warrants that make the Pentagon the company's largest shareholder at roughly 15%. That is not a subsidy. It is a state-underwritten commodity price.

On 14 January 2026, Proclamation 11001 extended Section 232 to processed critical minerals and their derivative products, with Commerce and USTR directed to negotiate price floors and trade measures with allies. The stated rationale is explicit: price volatility, not scarcity, is what deters Western capital.

3. The U.S. government became an equity holder

Across 2025–26 the federal government took direct stakes rather than issuing grants: MP Materials (~15% for $400M), Trilogy Metals (10% for $35.6M), a warrant-based ~5% in Lithium Americas alongside the restructured $2.23B DOE loan, and equity or equity-like positions in Vulcan Elements, ReElement Technologies, USA Rare Earth and Korea Zinc. Whatever you think of the policy, it changes the cost of capital for every name below.

4. Prices moved, and then settled

NdPr oxide ran from roughly $53/kg in January 2026 to about $126/kg by April — a 138% move — before easing back to around $96.69/kg in September 2026. Dysprosium sat near $239/kg in early September. Terbium traded in a wide $840–970/kg band across the year depending on quote and purity.

How to read the map

We sort the universe into four tiers by distance from a sellable kilogram. The tier matters more than the market cap:

• Tier 1 — Producing and separating. Revenue exists. Oxide ships today.
• Tier 2 — Metal, alloy and magnet midstream. The actual chokepoint. Mostly pre-scale, some private.
• Tier 3 — Developers. Permits, pilots, and financing. Years from cash flow.
• Tier 4 — Adjacent critical minerals. Tungsten, antimony, graphite, lithium, copper — same policy tailwind, different chemistry.

The universe at a glance

Prices as of 25 September 2026. Market caps shown only where independently verified; "—" means not confirmed at time of writing, not zero.

Where they are listed

Listing venue is not a cosmetic detail. It decides who is allowed to own the shares, which indices the company can enter, what it has to disclose and how often, and what currency the earnings arrive in. For a sector this small and this politically charged, it also decides how much of the story a Western investor can actually buy.

By region of primary listing

Ten American listings are worth less than three Chinese ones. The U.S. has the most companies and the least value per company — the median American listing here is worth about $2.0bn against roughly $5.6bn for a Chinese one. That is what a build-out looks like while it is still being built: many small bets, none of them yet at incumbent scale.

By exchange

What the venue split actually tells you:

• Nasdaq is where the new American money went. Seven of the ten U.S. listings sit there, and most arrived recently — USA Rare Earth, REalloys and Critical Metals all came to market through SPAC or reverse-merger routes. The NYSE proper has exactly one name in this report: MP Materials.
• The ASX cohort earns its keep. Three listings, $12.65bn, and the only one of the four regions where the largest company is actually shipping separated oxide at scale and reporting a profit. Australia is the closest thing to a functioning Western rare earth industry rather than a plan for one.
• Canada trades at a discount that is hard to defend. Three listings worth $1.97bn combined — less than 3% of the group — despite Neo Performance operating the most productive non-Chinese midstream on this page. Same asset, different exchange, a fraction of the multiple.
• The Chinese names are mostly unbuyable. All three list domestically in Shanghai or Shenzhen, reachable only through Stock Connect, QFII or an ETF wrapper. Only JL MAG carries a Hong Kong line. For most Western investors the incumbent — 42% of the sector's value — is simply not investable directly.
• Dual listings hide duplicates. Nine of the nineteen carry a second line — Lynas, Energy Fuels, NioCorp, Ucore, Ramaco, Aclara, Almonty, Perpetua and JL MAG. Screening by exchange without deduplicating will double-count nearly half of this universe.

Tier 1 — Producing and separating

Four Western companies currently ship separated rare earth oxide at scale. Everyone else in this report is selling a plan.

Tier 2 — Metal, alloy and magnets

This is where the actual chokepoint lives. The U.S. can mine. It largely cannot turn oxide into metal, metal into alloy, and alloy into a qualified magnet. Note how much of this tier is private — the public market gives you only partial access to the bottleneck.

The private companies that matter

You cannot buy these, but ignoring them will cause you to misprice the listed names, because they are competing for the same customers and the same federal dollars:

• Vulcan Elements — a $1.4B partnership with the U.S. government and ReElement: $620M in direct DoD Office of Strategic Capital loans, $50M in CHIPS Act incentives, $550M private capital. Building a 10,000-tonne magnet facility — the same scale as MP's 10X.
• ReElement Technologies — $80M OSC direct loan; recycling and refining oxides from end-of-life magnets and e-waste. Also holds a federal equity stake.
• Noveon Magnetics — $215M Series C in early 2026, led by OneIM (Rajeev Misra) with a $200M cheque. Already producing sintered magnets in Texas.
• Phoenix Tailings — $40.2M B-3 round (Series B total $116.6M), a $66M DOE grant inside a $147.8M project, and a $500M conditional OSC debt commitment for its "Freedom Facility."

The implication is uncomfortable for public shareholders. Add Vulcan's 10,000 tonnes to MP's 10X 10,000 tonnes, plus USAR, Noveon and Neo's Estonian line, and the announced Western magnet capacity for 2027–29 is large relative to any realistic non-Chinese demand in that window. Not every one of these plants earns a return. Policy is currently funding all of them simultaneously.

Tier 3 — Developers

Everything here is a financing story. The question for each is identical: who writes the cheque, and what does the equity look like after they do.

The Australian and Canadian developers

• Iluka Resources ASX: ILU — the Eneabba refinery will be Australia's first fully integrated rare earth refinery. Capital cost has escalated from ~A$1.2B to A$1.7–1.8B and commissioning has slipped to 2027. The first A$1.25B government tranche is expected fully drawn by end-2026 at 75% completion. Its first binding offtake — roughly 1,200 tpa of magnet rare earth oxides to an undisclosed global automaker, take-or-pay from 2028 — is the single most important commercial validation in the Australian sector. Outlook: state-financed, schedule-slipping, but with a real customer. The capex overrun is the risk to watch.
• Arafura Rare Earths ASX: ARU — Nolans (NdPr, Northern Territory) is at the finish line of a decade-long financing. More than A$930M of equity commitments are banked, including a $375M raise in May 2026 with Hancock Prospecting lifting to ~17.5%. Contractual close was slated for October 2026 pending final lender credit approvals. Outlook: binary and near-term. If October closes, Arafura becomes a funded developer; if it slips again, the board pressure that has already surfaced gets worse.
• Aclara Resources TSX: ARA — ionic clay heavy rare earths at Penco (Chile) and Carina (Brazil), which is a genuinely different and lower-cost deposit type than hard rock. Development phase across upstream, separation and metallization completes end-2026, construction starts subject to financing, commercial production targeted 2028. Outlook: the most credible heavy-REE geology outside China and Greenland, gated entirely on Chilean permitting and a construction cheque that has not been written.

Tier 4 — Adjacent critical minerals

The same policy machinery is being pointed at tungsten, antimony, graphite, gallium and germanium — all of which China restricted before it touched rare earths.

Also in the frame

• Westwater Resources NYSE American: WWR — graphite anode material, Kellyton, Alabama.
• Syrah Resources ASX: SYR — Balama graphite (Mozambique) plus U.S. anode capacity; persistent operating and country risk.
• Nouveau Monde Graphite TSX: NOU · NYSE: NMG — Quebec graphite, mine-to-anode.
• Lithium Americas NYSE: LAC — Thacker Pass; the DOE holds a warrant-based ~5% stake alongside the restructured $2.23B loan.
• Trilogy Metals NYSE American: TMQ — Alaskan copper-cobalt; the U.S. government took 10% for $35.6M.

The incumbent: China's listed complex

Any honest map has to include the companies actually setting the price. They are also the only ones in this report with the kind of margins the Western build-out is implicitly promising.

The numbers to sit with. China Northern alone did ¥42.6B — roughly $6B — of revenue in 2025 with profit up 124%. MP Materials, the American champion, did $224M. That is a 25-to-1 gap at the top of the market. Shenghe's ~700–780% profit surge is what happens when you control the midstream during a price spike. Western investors are not buying into a fair fight; they are buying into a state-funded attempt to build a second, higher-cost supply chain in parallel. Whether that ever earns a normal return is the open question in every valuation above.

What we are watching

1. 10 November 2026. China's suspension of the 0.1% de minimis rule expires. Whether it is extended, allowed to lapse, or reimposed is the single largest near-term swing factor for every name on this page.
2. The Section 232 negotiation outcome. Commerce and USTR were directed to report on allied price floors and trade measures by 13 July 2026. If a Western price floor generalises beyond MP's bilateral deal, the whole Tier 1 and Tier 3 cohort re-rates. If it does not, MP's floor stays a unique asset.
3. NdPr versus $110. The market price is ~$97. The U.S. government has declared $110 the level at which domestic production makes sense. Everything about Western project economics sits in that $13 gap.
4. Magnet capacity versus magnet demand. MP's 10X, Vulcan's 10,000 tonnes, USAR's 1,200, Neo's Narva line, Noveon's Texas plant — all landing 2027–29. Somebody's plant runs at low utilisation.
5. Heavies, not lights. Dysprosium and terbium never came off the Chinese control list. Aclara, Ucore, Critical Metals and Lynas's DyTb line are where genuine scarcity actually sits — and where the least Western capacity exists.

How we frame it

We are not commodity traders, and this is not a sector where we would pretend to have an edge on quarterly price prints. But it is exactly the kind of place where a long time horizon is an advantage, because the entire thesis resolves over a five-to-ten year build-out and almost nobody holding these stocks intends to be there for it.

Three principles we would apply to our own exposure:

1. Pay for tonnes, not for announcements. The gap between MP Materials at ~20× trailing revenue and Neo Performance at ~1.5× is not explained by asset quality alone. It is explained by narrative proximity to Washington.
2. Prefer the balance sheet that does not need the market. Lynas (A$1.2B cash), USAR ($1.53B), NioCorp ($419M net cash) and Perpetua ($2.9B committed) can execute without asking permission. CRML, Ucore and Arafura cannot. In a sector where dilution is the primary way retail investors lose money, that distinction outranks geology.
3. Assume the policy is temporary and the plant is permanent. Price floors, equity stakes and Section 232 authority are all reversible by an election. A commissioned separation plant with qualified customers is not. Underwrite the second thing.

Where that leaves us. The two names that survive all three filters today are Lynas — real tonnes, real cash, obvious unused capacity — and Neo Performance Materials — real midstream, real customers, a valuation that has not been bid up, with China exposure as the acknowledged offset. MP Materials owns the best strategic position in the West and we would expect to own it eventually, at a price that reflects execution rather than expectation. Everything else in Tiers 2 and 3 we would treat as venture exposure inside a public wrapper, and size it that way.

Sources

Company filings and releases: MP Materials Q4/FY2025 and Q2 2026 8-Ks; Lynas FY2025 and FY2026 results and the June 2026 quarterly report; Energy Fuels 2025 results and Q2 2026 release; USA Rare Earth FY2025 10-K and Q2 2026 release; Neo Performance Materials Q4 2025 and Q1 2026 releases; Ramaco Resources FY2025 and 2026 quarterly releases; NioCorp 10-Q (Mar 2026); Critical Metals Corp project updates; Ucore engineering report (May 2026); Idaho Strategic Resources 2025 results; Almonty, Perpetua and REalloys company disclosures. Policy: White House Proclamation 11001 (14 January 2026); DoD–MP Materials partnership announcement (July 2025); China MOFCOM announcements of April and October 2025 and the subsequent one-year suspension. Market data via public quote services as of 25 September 2026. Thesis framing from the public section of Social Capital's critical minerals deep dive.