RESOURCES & STATECRAFT DESK · COLUMN
The claim is simple and, until you lay the deals side by side, easy to miss: the non-Chinese rare-earth and magnet supply chain is not being built by one state with a grand plan. It is being built by five treasuries, separately, each reaching for the same chokepoint within the same eight weeks. The pattern is convergence, and convergence is harder to reverse than coordination — because there is no single agreement to walk back.
Each deal is best read in its own instrument, because the instrument is the act.
The United States went first and went biggest in form: on 20 April 2026, White House-backed USA Rare Earth agreed to acquire Brazil's Serra Verde at a roughly US$2.8 billion implied equity value — US$300 million in cash, the rest in newly issued stock, closing expected in the third quarter — securing all four magnetic rare earths outside China and bundling in a fifteen-year offtake to a vehicle capitalized by U.S. government parties, with price floors attached. That is acquisition plus demand-side underwriting — the state as buyer.
Australia acted through finance and statute. On 21 May 2026, Arafura Rare Earths reached a Final Investment Decision on its Nolans project, backed by Export Finance Australia (EFA) and the Northern Australia Infrastructure Facility (NAIF) — the first operational use of Australia's new Strategic Reserve powers in a non-fuel commodity, locking in non-China light rare-earth (neodymium-praseodymium) supply. That is an export credit agency and a concessional facility, plus a reserve power reclassifying magnets as strategically equivalent to fuel.
Germany bought in — literally. In April 2026, KfW, the German state development bank, took A$84 million in cornerstone equity in Arafura on behalf of the German Raw Materials Fund — with a permanent board seat attached — Berlin's most concrete deployment of state capital into AUKUS-adjacent strategic minerals. Equity, not a loan: a sovereign sharing the downside.
Australia and Japan then anchored the chain at scale. On 1 May 2026, Australia committed A$1.3 billion for Australia–Japan critical-minerals cooperation — sovereign co-investment behind a supply chain deliberately not concentrated in China.
And the United Kingdom added the smallest check with the clearest framing. On 22 June 2026, Britain announced a £50 million package to expand domestic extraction, processing and recycling of critical minerals — a move set against China's roughly 70 percent of rare-earth mining and roughly 90 percent of processing, on the standard CSIS estimate; the UK announcement itself names no country.
Now read the pattern. Five states. Five different instruments — acquisition-plus-offtake, export credit, cornerstone equity, large-scale co-investment, a domestic program. One target. Note that they do not add up to a single number, and must not: a roughly US$2.8 billion stock-weighted acquisition value, a Final Investment Decision against a reported A$1.6 billion capital requirement, A$84 million of equity, A$1.3 billion of co-investment, and £50 million of program funding are five different kinds of object. Summing them would be the analytical error the whole desk exists to avoid. The story is not a total; it is a shape.
The shape is this: each state is reaching past the mine toward the same scarce step. The UK package puts its weight there — a £25 million accelerator spanning extraction, processing and recycling, and a £20 million magnet hub — against the step where China holds roughly 90 percent. Processing is the bottleneck, not extraction, where its share is "only" 70. Germany's equity goes into a company built to produce magnet feedstock. Australia's reserve commitment targets neodymium-praseodymium, the magnet metal. Even the U.S. offtake is, at bottom, a guarantee that magnets will have a market. Five treasuries, converging on the midstream.
What does the balance-sheet behavior reveal? That magnets have crossed a threshold in allied capitals from commercial input to security input — and that the response is being made in capital, not communiqués. There is no treaty here, no signed alliance pledge. There is a Final Investment Decision, an equity check, an offtake, a grant. The commitments are harder than a framework precisely because they are already disbursing or contracted; you cannot un-sign a Final Investment Decision the way you can let a memorandum lapse.
The contrary read deserves a hearing: convergence without coordination can also mean duplication and gaps. Five states each building a partial chain may end up with five sub-scale efforts and no one owning the full midstream. The Australia–Japan and KfW-Arafura links suggest some stitching is happening; the rest is, for now, parallel rather than integrated. That is the risk the next year will test.
But the direction is unmistakable. Read as capital statecraft, these are not five mining stories. They are one positioning move, executed by separate sovereigns who have independently concluded that the magnet is the new fuel — and that the place to fight for it is the refinery, not the pit.
When five treasuries reach for the same chokepoint without a meeting, you are not watching a policy. You are watching a consensus pay for itself.
— Capital Statecraft Intelligence · Resources & Statecraft Desk
Primary source(s): Fortune; ABC News; Export Finance Australia; MinterEllison; Australian Dept of PM and Cabinet; UK Dept for Business and Trade (gov.uk)