Tearing apart the antimony trioxide refining chokepoint

Antimony.

The spring headlines all read the same way. China suspended the export ban in November. Prices rolled off the summer record. The 2026 supply outlook came back "adequate." Crisis over. Everybody go home.

If you've read this letter for more than a week, you know what comes next. They're watching the spot price of the metal and pronouncing the whole machine healthy. The metal was never the machine.

Antimony is two products sharing one name. There's the metal — ingots and ore, the thing with a price you can quote on a screen. And there's antimony trioxide — the white powder that actually goes into the flame retardant in your laptop shell, the clarifier in solar glass, the primer in a rifle cartridge. Industry doesn't buy the metal. It buys the powder.

The market declared victory on the part it can see. The part it can't — the refining step that turns ore into usable oxide — never loosened its grip. I've watched this exact misdirection in other metals. The spot price calms, everyone exhales, and the bottleneck quietly relocates one step downstream where nobody's looking.

Story off. Numbers on.

China started this in August 2024 — a licensing requirement on antimony's dual-use exports, effective that September. By December it had banned shipments to U.S. military end-users outright. (Quest Metals)

The reaction wasn't subtle. Chinese antimony exports fell roughly 97% after the August restrictions, and global prices roughly tripled — about a 200% move. (Andersen Institute) The metal hit a record $59,750 a tonne on July 4, 2025, then rolled over once Beijing suspended the broad ban in November and the 2026 outlook turned to "adequate." (Fastmarkets) That's not a commodity cycle. That's a valve being shut by hand, then cracked back open by the same hand.

Here's the number the all-clear crowd skips. China's own mine output has been falling for two decades — from 100,000 tonnes in 2000 to about 40,000 tonnes in 2024, out of a world total near 83,000. (Quest Metals) The country that owns the chokepoint is running short of its own ore. That doesn't make the chokepoint looser. It makes Beijing more protective of the part it still controls: the refining.

So Southeast Asia and Central Asia are pulling more antimony out of the ground. Good. That solves the metal.

It does nothing for the oxide. Antimony trioxide refining — the conversion that makes the actual industrial product — still runs through a short list of chokepoints: China, Belgium, France, and a handful of Southeast Asian plants. (Skillings) You can now buy raw metal from more places. You still have to send it through the same few refineries to get the powder anyone can use.

And the licensing architecture didn't disappear in November — it got formalized. The new 2026 Chinese export rules tighten eligibility for exporters: creditworthiness, compliance history, the kind of paperwork that's easy to grant and easy to quietly deny. (Strategic Metals Invest) The suspension loosened the spigot. It didn't take Beijing's hand off it.

The defense buyers understand this better than the commodity desks do. A primer needs antimony trisulfide. A flame retardant needs the trioxide synergist or it doesn't pass code. Neither one accepts "we've got plenty of ore in Tajikistan" as an answer. They need the converted product, and the conversion is the bottleneck nobody re-shored.

You already know how this resolves, because you watched gallium and germanium run the same play.

Something reminds the market the suspension was temporary. A 2026 license gets denied on "compliance" grounds. A defense audit finds a prime is one supplier deep on trioxide. PV glass makers in a tariff fight discover their oxide all traces back to one province. First move: the spot metal everyone watches barely twitches, because the metal isn't the problem. Second move: trioxide premiums blow out while metal sits flat — the same paper-versus-physical split this letter keeps finding.

Third move: anyone who can convert ore to oxide outside China gets repriced as strategic infrastructure instead of a chemicals footnote.

Where does the capital go? Not into the pure-play ore miners — they still have to ship concentrate somewhere to convert it, and "somewhere" is mostly the problem. The edge sits with producers who own the full metal-to-oxide-to-product chain on allied soil. The U.S. has one obvious candidate scaling for it: United States Antimony completed a six-fold expansion of its Montana smelter capacity in January 2026. (United States Antimony Corp) Whether the economics hold at lower metal prices is the open question — but if the oxide chokepoint bites, the people who own conversion get paid first.

The spot screen says the antimony crisis ended in November. The refinery map says it just moved one step down the chain, where it's harder to see and harder to fix. When those two disagree, I've learned to trust the map.