The United States is writing large cheques. China is still running the plants.
That is the whole tension in one line. Reuters, citing the administration, reported on 14 September 2026 that Washington had signed or approved about 160 critical-minerals deals worth more than $40 billion since January 2025. Equity has gone into names such as MP Materials, Lithium Americas, and USA Rare Earth. Project Vault pairs a $10 billion Export-Import Bank loan with nearly $2 billion of private capital for a civilian stockpile. Australia has stacked billions more across gallium, magnet metals, graphite, tungsten, and nickel. Saudi processing talks sit in the same folder.
Then comes the IEA ledger. Rare-earth refining in China fell from more than 90% in 2023 to about 85% in 2025 after U.S. and Malaysian capacity inched on. That is progress. It is not a turning point. Strip rare earths out and China’s average share of other critical-minerals refining rose from 70% to 72%. The agency’s Global Critical Minerals Outlook 2026, published in July, said that even if every planned rare-earth refinery in the world starts on time, China’s share of that slice only falls to about 70%–73% by 2035.
How U.S. investment could challenge China is therefore not a mystery. It challenges China if oxide, metal, and magnets leave Western plants at a cost the buyer will pay. It does not challenge China if the cheque buys a groundbreaking and the concentrate still sails east.
This article is not a recommendation to buy rare earth stocks, critical minerals stocks, or any mining share. It is a map of U.S. critical minerals strategy against China mineral processing as it actually stands.
Where the Billions Went
U.S. critical minerals policy in this cycle is not one bill. It is a pile.
Defense and commerce credit. Conditional loans. Price-support talk, including a $110-per-kilogram floor cited in CSIS work on magnet-metal policy. Guaranteed offtake so a plant can raise the rest of the capital. Project Vault so industry has a buffer when an export licence stops. Bilateral packets with Australia, Japan, Malaysia, and Saudi Arabia so the midstream does not have to live only in Nevada.
That is serious industrial policy. Analysts at the Center for Climate and Energy Solutions still warn the obvious: mines and refineries take years. A grant from a prior administration that is not funded is a press clipping. Consistency across election cycles is the scarce input.
Industrial Info has put global critical-minerals capex under construction near $183 billion, with China about $57 billion and the United States about $25 billion. China is still building. The U.S. is still behind on the pour.
China Critical Minerals: The Midstream Is the Moat
China critical minerals power is not only pits. It is rare earth processing, graphite spherical plants, gallium, germanium, LFP cathode powder, and the skilled labour that runs them.
IEA-linked tallies for 2025 put China at 70%–95% of processing for lithium, cobalt, phosphate, manganese, and graphite. It made about 98% of LFP cathode material and about 80% of battery cells. Heavy rare-earth separation has sat near a monopoly. Permanent magnets have sat near 90%–plus. Mining is less concentrated — China is still about 60% of many critical-mineral mine tallies — but mining without refining is a raw-material colony with better branding.
Beijing has also shown it will use the licence. Export controls in 2025 taught buyers that a shipment can stop. Allies noticed. That is why the money moved. It is also why China added U.S. firms, including MP Materials, to control lists. A turning point that can be listed away is not a turning point.
Copper supply is the cousin story. China refines on the order of 40%–plus of world copper. The red metal’s tightness is mines and tariffs. The magnet-metal tightness is cookbooks and solvent extraction trains. Do not mix the two in one slogan.
Could This Be a Turning Point?
A turning point would look like three things at once.
First, Western separators that run at nameplate and sell oxide without a Chinese offtake. Second, magnet plants that take that oxide and sell a motor-grade product. Third, a price that does not collapse the day China dumps stock. The IEA 2035 path says the first condition, even in a bullish build-out, still leaves China with most rare-earth refining. That is a bend in the curve. It is not a break.
U.S. mine output is already number two in rare earths. For years that concentrate still went to China because the U.S. lacked the dirty, skilled middle. Building the middle is the whole U.S. critical minerals strategy. Recycling and substitution can leapfrog some tonnes. They cannot leapfrog a jet magnet this quarter.
Critical minerals 2026 is therefore a split screen. Capex announcements on one side. IEA shares on the other. Believe the second until the first ships.
What That Means for Listed Names
Rare earth mining stocks and rare earth mining files are not interchangeable with a magnet company. A pit in Australia or Canada is not a separator in Texas. Critical mineral companies that only mine will still feed someone else’s plant. Critical mineral processing is where the policy money is trying to go. That is also where losses hide when the flowsheet fails.
Canadian mining stocks show up as nickel, copper, uranium, and a thin rare-earth book. They are mining stocks to watch only as research files. A Canadian concentrate that still needs a Chinese or Malaysian plant is not a broken China chain. It is a longer truck ride.
Critical minerals investment will keep attracting speeches. It will keep punishing anyone who paid a mine multiple for a midstream hope. Do not confuse a Pentagon offtake with a finished supply chain.
Risks That Survive the Cheque
Permitting time. Community fights. Reagent costs. Radioactive residues in monazite circuits. Price crashes engineered by the incumbent. Allied projects that slip two years. A stockpile that sits unused because specs do not match the factory. LFP cathode rules that trip Western battery lines. A 2035 IEA case that assumes plants start on time — they do not.
China mineral processing was built over thirty years of cheap power, weak environmental enforcement, and a state that treated magnets as strategy. The U.S. is trying to compress that into five to seven years. Tu Le of Sino Auto Insights put the gap in one line: you need decades and tens or hundreds of billions for China’s scale, and you do not have decades.
Conclusion
The United States has invested billions. China still dominates the midstream. Supply chains could be at a turning point if oxide and magnets leave allied plants in volume before 2030. The IEA’s own math says China keeps most rare-earth refining into the next decade even then. Treat the $40 billion as a down payment. Treat 85% and 72% as the present. A turning point is a tonne on a dock, not a deal count in a briefing.
Disclaimer
Deal counts, dollar figures, and market-share estimates follow Reuters reporting dated 14 September 2026, the IEA Global Critical Minerals Outlook 2026, and related public analyses (CSIS, Industrial Info, and others). Shares and project lists change. Company names are examples, not recommendations. This article is not investment advice and not a recommendation to buy or sell any security. Speak with a licensed adviser. The author and publisher accept no liability for actions taken on this article.