AI's Next Trade? Goldman Sees a 'Mine-To-Magnet' Boom-But January 2027 Is the Real Test

Goldman Sachs says AI demand is reshaping the critical-minerals trade

The debate is no longer whether these minerals matter. Goldman Sachs has already said dealmaking is rising in critical minerals and rare earths as demand from AI, EVs, and energy infrastructure accelerates. The harder question is supply-chain function: the U.S. can mine some of the material, but it still struggles to turn those ores into the finished magnets used in AI hardware, EVs, robotics, and defense systems.

Reuters says the January 1, 2027 deadline for defense contractors to stop buying certain Chinese-linked rare earth materials is now less than five months away, and the gap remains large: in 2025, U.S. demand for the most common rare earth magnets was roughly 48,000 metric tons, while domestic sources supplied only 300 metric tons.

Why "mine-to-magnet" matters more than ore scarcity

That is where the mine-to-magnet upside comes from. When the bottleneck shifts from raw ore to usable supply, companies closest to closing that gap can trade less like simple commodity miners and more like strategic infrastructure. Goldman says companies are adopting "mine-to-magnet" strategies to integrate mineral production and reduce supply-chain risk, and market attention is concentrating on MP Materials, Critical Metals Corp., TMC, and U.S. Antimony.

MP looks like the most developed proxy in the group because it is widely seen as building an integrated mine-to-magnet supply chain in the U.S. CRML, TMC, and UAMY are smaller, more speculative plays on rebuilding parts of that chain at home. Bulls see a national-security-backed shortage that could reward even early progress. Bears argue January 2027 may arrive before domestic supply is real, with waivers still likely because homegrown output remains too small.

MP Materials has operating proof; the smaller names are still earlier-stage bets

MP stands apart because it is not just a future-chain story. It is an operating company with an active piece of that chain already producing and shipping. Reuters says MP posted $9.4 million net income for the quarter ended December 31, helped by a price support agreement with the U.S. government and magnetic-material sales. MP also controls the only rare earths mine in North America, processes material in California, and has built a magnet facility in Texas. That operating footprint is the key distinction: it is the only name in this group that currently combines mining, processing, and magnet manufacturing under one U.S. footprint.

Why operating progress gets a premium

Markets usually pay more for proof than for promise. When a company can show revenue, shipped product, and a government backstop that reduces early-ramp risk, the story starts to look more like infrastructure and less like a concept. MP's $110 per kilogram price floor is not just a contract detail; it lowers some of the downside risk while domestic demand continues to run far ahead of homegrown supply.

Why CRML, TMC, and UAMY still have a case

That does not make CRML, TMC, or UAMY irrelevant. Their case is simply earlier stage. If government support keeps shifting from rhetoric into grants, equity, and off-take commitments, these projects could advance faster than the market expects. The risk is the same either way: none of them currently has MP's operating proof, and none should be priced as if the January 2027 bottleneck is already solved.

The China export-control test

There is also a live stress test in the background. China recently put MP Materials and USA Rare Earth on an export control list that bars Chinese exporters from selling dual-use items to them. In the short term, that is a supply-chain headwind. In the longer term, it reinforces the argument that the U.S. cannot outsource this supply chain if it wants strategic independence.

What decides the trade before January 2027

Investors are not waiting for a fully completed U.S. supply chain. They are trying to price, today, who is best placed to capture value before the January 1, 2027 deadline.

The market is still pricing a major shortfall: about 48,000 metric tons of demand versus roughly 300 metric tons from domestic sources. In practical terms, investors are buying the possibility that some part of the missing middle gets built soon-not a clean finish line.

What would validate the thesis

The rerating path is straightforward: usable capacity has to appear before the calendar does. If reported U.S. progress starts to turn into real output, the market gets evidence that the bottleneck is easing.

MP remains the cleaner single-stock proxy because it already has operating proof, including an earlier quarterly profit and progress toward domestic magnet production. CRML, TMC, and UAMY are higher-beta options and only make sense if policy support keeps converting plans into real assets.

China's decision to place MP Materials on an export control list is another reminder that this supply chain is now a strategic pressure point, not just a commodity story.

The thesis weakens quickly if headlines keep improving while actual U.S. output remains too small to matter.