03
The other five
Tungsten, gallium, germanium, antimony and graphite. Not rare earths, but the same shape of problem: concentrated supply, no practical substitutes, defense-critical end uses, and an export license regime pointed at all of them.
Last updated 2 August 2026 · Click any entry for detail
Why these five sit together
Three of them, gallium, germanium and graphite, are not mined for their own sake at all. Gallium comes out of bauxite refined for aluminum. Germanium comes out of zinc. That is the single most important fact about them: you cannot open a gallium mine. Supply is set by demand for a completely different metal, which makes it almost perfectly inelastic to price. A tenfold price rise produces very little new supply.
Tungsten and antimony are mined directly, but both are geologically concentrated and both carry defense end uses with no substitute. Carbide tooling and armor penetrators for tungsten, ammunition primers and flame retardants for antimony.
China restricted all five between July 2023 and September 2024, then partially suspended after the October 2025 Busan meeting. That suspension expires 27 November 2026, a different date from the rare earth suspension, which lapses on 10 November. Two separate instruments, seventeen days apart, and they are routinely conflated.
The military end-user prohibitions survived the suspension in every case, as did the licensing requirements themselves. That is the pattern worth learning: a trade deal relaxes commercial flows and leaves the defense restriction standing.
The control timeline
The same instrument, applied in sequence. Each step was a rehearsal for the next.
The byproduct problem deserves more attention than it gets. Most policy discussion treats these as mining problems with mining solutions. Permit faster, open more. For gallium and germanium that is close to a category error. The constraint is refining capacity for the host metal and the willingness to build a recovery circuit that only pays at elevated prices. Nyrstar's Tennessee work recovers gallium and germanium from existing zinc waste ponds, which is a far better model than a new mine.
Where this is thin. Production shares for byproduct metals are unusually unreliable. Much output is unreported or bundled into host-metal statistics. Figures here should be read as orders of magnitude. Project timelines are company guidance and have historically slipped.
Research and opinion. Not investment, financial, or legal advice. I hold positions in some companies named here. · Matt Stone · Thinking about selling? · matt@stonehammer.ai