Welcome back to CMJ,

Here are the 20-second highlights:

Semantics matter, especially when writing a Letter of Intent (or in this case, a ‘Report of the Agreement’ type of document).

The White House fact sheet, highlighting the post-Xi-Trump Summit, reads that China would “address” U.S. concerns about shortages of yttrium, scandium, neodymium, and indium, along with restrictions on rare earth processing technology.

The verb “address” is not equal to “eliminate”. It commits China to nothing on a timeline, names no verification mechanism, and defines no operational test for compliance.

Now compare that language to the one used six months earlier, after the Busan summit in October 2025, when the White House described the Chinese commitments to be “effectively eliminate” (including the proposed export controls).

The change / trajectory of the verbs tells most of the story.

In contrast, apparently (from non-official Chinese sources), China's Ministry of Commerce called its controls lawful and legitimate, and offered only to work with the U.S. on concerns it considers “reasonable”. 

China’s official communication regarding the summit never mentioned rare earths or critical minerals. It looks carefully written to be as vague as possible (link).

That matters because the two statements, issued only days apart, describe a new power dynamics (equilibrium): the U.S. ‘asks’, China ‘decides’ what counts as reasonable, and the licensing architecture stays fully intact.

What China bought with the word “address” is time and optionality, at no cost to the system it built in April 2025. A semantics masterpiece.

And speaking of the market, the flow data sets a strong narrative.

Yttrium oxide is one of the cleanest reads we have on how China's April 2025 rare earth controls actually behave once the licensing machinery is running. With the latest customs records, it’s inevitable not to recognize the supply shock impacts.

Note: the analysis considers China as the dominant player in Yttrium Oxide exports, meaning no alternative source.

Through 2024, China shipped 2,747 tonnes of yttrium oxide for roughly $21.0M, at a value-weighted realized price near $7.65/kg, across an average of nearly nine destinations a month. The trade was deep, boring, and predictable.

Then, in April 2025, the number of destination lines collapsed to three and 136 tonnes. May and June 2025 recorded zero exports. Those are the only two blank months in the entire 28-month series, and they sit exactly where China's controls on seven medium and heavy rare earths (yttrium among them) took hold.

Here is the summary:

Strip out seasonality by comparing like-for-like January–April windows, and the picture sharpens. Jan–Apr 2024 moved ~832 tonnes at $8.24/kg. The same window in 2025, still pre-control, moved ~887 tonnes at $7.91/kg. Jan–Apr 2026 moved 364 tonnes at $12.99/kg. That is roughly a 40% drop in volume and about 60% up in price.

Annualize the 2026 run-rate and recorded exports land near 1,090 tonnes, a 60% cut against 2024, while the realized price sits about 70% higher.

And that is only considering the Chinese price reference. If we look at the European (Rotterdam) reference, which as of today is floating around $350-450/kg (depending on the source), it represents a 30x in price gap.

The destination map is where the strategic content lives, and we’d like to bring your attention to it:

The Japan number is the one that should stop a reader. Japan was the gravitational center of this trade, averaging 131 tonnes a month in 2024. In 2026, it is averaging under two tonnes a month of direct Chinese yttrium oxide.

That matters because it tells you the controls are selective in practice, not just in statute. The volume that came back did not come back to the buyers or the ports it left from. Rotterdam and the Italian entry points, the European spine for this oxide, recorded zero direct flow from July 2025 onward, yet European industrial demand for yttrium (phosphors, refractories, YSZ coatings, garnet stabilizers) did not vanish. Something is arriving relabelled, re-routed, or drawn from inventory, the channel built ahead of the controls. And when new channels are built, various opportunities emerge.

In Greenland, Critical Metals Corp concluded a 15-year binding offtake agreement with REalloys for rare earth concentrate from its Tanbreez project in southern Greenland, one of the largest heavy rare earth-rich deposits in the Western world. The deal formalizes a Letter of Intent from October 2025 and follows Greenland's approval of Critical Metals' move to 92.5% ownership of the project.

The same day, Greenland Mines secured the Sarfartoq project through a US$35 million agreement with Neo Performance Materials (Neo), an asset-rich in neodymium (Nd) and praseodymium (Pr) (the base REE for the NdFeB permanent magnet, one of Neo’s core products). Neo keeps an equity stake and the right to buy up to 60% of future output.

Notice for a second who the buyers are. REalloys and Neo are not speculative juniors, they are both midstream companies stretching upstream to secure feed. That is the playbook we have covered for over 1.5 years:

Still on the offtake subject, Critical Metals Corp also closed its acquisition of European Lithium, consolidating Tanbreez and the fully permitted Wolfsberg lithium project in Austria, which already carries a BMW offtake, under a single owner.

There is a quieter risk inside the offtake optimism, though. A 15-year offtake transfers execution and timeline risk to the buyer. It does not remove it (even with covenants in place). And as more juniors sign offtake before securing separation, those contracts begin competing for processing capacity that still does not exist at scale outside China. Signed paper is not separated oxide.

The map of where this gets solved is short. Outside China, large-scale separation of the magnetic rare earths, especially the heavies, runs through a handful of nodes, with Lynas in Malaysia the most established (but ramping up) and a thin set of U.S. and European projects still scaling. That is why a separator securing its own feedstock is an optimistic sign. One note on securing feed: NdPr pipeline seems to be abundant, but we can’t say the same for DyTb or even Sm (for the SmCo permanent magnets), and here is where Tanbreez’s beauty shines.

The clearest expression of that logic is the ReElement Technologies and POSCO International US$200 million joint venture to build an integrated U.S. rare earth complex spanning separation, refining, and permanent magnet manufacturing, with a phased 3,000mtpa of separated oxides by 2028, and a second phase reaching 6,000mtpa by 2030 as major milestones. The two had been working toward this since a memorandum of understanding in September 2025 and an existing offtake relationship.

Interpret POSCO the same way you’d Neo (but on a larger scale). POSCO is a magnet and materials player moving upstream, not a miner moving down. The pattern we are seeing emerging is that it’s ‘easier’ for magnet makers to move upstream than for juniors to move downstream. And that makes sense, they are the ones feeling the shortage pain and have the most to lose if separation stays a ‘Chinese monopoly’. Partnerships and JVs are a great instrument for speed, much faster than a green or brownfield development.

But the same constraint bites here, too. A refining and magnet complex still needs feedstock to run, and ReElement's own scale is unproven, with its site still unselected and the U.S. reportedly reconsidering a US$80 million conditional loan. Not mentioning the actual feedstock of MREC that ReElements seems not to have secured yet.

Building economically viable ex-China separation at an industrial scale has repeatedly proven much harder than the announcements suggest. A magnet maker partnering with a refiner narrows the gap, but it does not close the part of the gap that is the ore itself.

Section 232 is the national security lever in U.S. trade law, and copper is the first investigation this administration carried to completion. The visible action is already in: a 50% tariff on semi-finished copper and copper-intensive derivatives since August 2025, restructured in April 2026 to apply to the full customs value of the product rather than its copper content alone.

But the part that moves a thesis has not happened yet. By June 30, Commerce must report on domestic refining capacity, and that report is the trigger for a phased universal duty on refined copper of 15% from January 2027, and 30% from January 2028. Refined copper, cathode, and concentrate are the inputs the regime has deliberately spared so far. Whether that exemption survives the report is the open question.

The sequence is worth holding in one frame:

When the U.S. left cathode out in 2025, the COMEX–LME arbitrage that opened on the announcement collapsed within days as the market realized that refined metal was untouched. The June 30 report reopens exactly that trade.

The useful observation is not that copper got tariffed. It is that this is the same managed-market architecture as rare earths, arriving on a base metal. China controls more than half of global smelting and four of the five largest refining facilities; the U.S. answer is not a wall but a structured attempt to build a domestic market by deadline and decree. The verb to watch on June 30 is whether "may impose" becomes "will", and that gap is where the copper trade lives, exactly as the rare earth trade lives in the gap between "address" and "eliminate".

But what nobody is remembering/modeling is the domestic-sale quota required under the same Section 232. A rising share of U.S. copper input materials is required to be sold domestically, starting at 25% in 2027, raising to 30% in 2028, and 40% by 2029, with a parallel 25% set-aside and an export licensing system on high-quality scrap.

Price only the tariff, and you miss the instrument that actually reshapes who gets the metal.

Thank you for reading this briefing and for being part of the CMJ community.

In markets driven by geopolitics, foresight is power. If you found this briefing valuable, share it with a peer who needs the same edge (or keep it close and use it to your advantage).

See you in the next issue of the Critical Minerals Journal.