Is the gallium ban leaking? Gallium after China's Dec-2024 US export ban

Trade-flow companion to the price wedge (R72). The wedge measures scarcity (what a controlled material costs ex-China). This case measures the other half: is the control actually holding, or is banned material re-entering the US through a laundered origin? This is a DUAL-SCORE / alternative-track signal — never folded into any Tier-1 exposure score. The divergence between "who China says it stopped selling to" and "who suddenly started selling to the US" IS the signal. Research, not investment advice; origin-relabelling is INFERRED from statistical implausibility plus public production records, never asserted as smuggling on any single shipment.

Live chart + method: [/situation-room → "Is the ban leaking?"](/situation-room).

Verdict — GATE 0 (the no-substitute twin of germanium; a distinct aggregate-asymmetry signature)

Gallium is the co-flagship of the whole MOFCOM control sequence: the first Chinese critical-mineral licensing regime (gallium + germanium, effective 1 August 2023, anchor sequence action), hardened into a full US export ban on 3 December 2024 (MOFCOM Announcement 46 — the same instrument that produced the clean antimony fingerprint and the mixed germanium signature). Gallium is also the most single-origin-dominant material in this entire corpus: China produces ~98% of the world's primary (low-purity) gallium (CSIS/USGS MCS 2024; some sources cite 99%) and refines ~50% of high-purity metal, and one Chinese firm (Sunresin) supplies ~90% of the gallium extraction resins the rest of the world would need to build capacity.

The germanium case ([germanium/gallium](2025-germanium-ban-circumvention-transshipment.md)) found a benign dominant reading: germanium oxide re-sourced cleanly to genuine allied refiners (Umicore in Belgium, Teck in Canada) at a much higher price — real Western capacity absorbed the ban. That escape hatch does not exist for gallium. The US has essentially no primary gallium production (CSIS: "one domestically based supplier" of high-purity metal, negligible primary output; USGS lists no US primary gallium), and there is no large Western primary gallium refiner equivalent to Umicore/Teck. The non-China "supply" the US leans on — Japan (Dowa's reprocessing), Germany — is small secondary/recycled capacity that is itself fed by Chinese-origin primary metal. So for gallium, the germanium sentence "the gap was filled by real allied refiners" is unavailable by construction.

That inverts the reading. Two facts, both traced, that cannot both be innocent:

Recomputed 2026-10-02 against UN Comtrade's own "all origins" total for the HS 811292 basket (the import-recovery claim tested directly, not inherited from the germanium case's country breakdown):

*US did not report net weight on this code in 2022 (field present, valued 0 on every row — a reporting gap, not a true zero; value is unaffected). †Jan-2023 is missing (one Comtrade rate-limit failure during this recompute, not a data gap) so the 2023 column understates by roughly one month.

The claim holds in dollars and complicates in kilograms. 2025's full-year value ($208.9M) is +30% over 2022 pre-control — the "equal or exceed" reading is correct on the money. But 2025's weight (1.91M kg) is −17% under 2022 (2.31M kg): the dollar recovery the case cites is partly a price effect, consistent with the Rotterdam wedge below, not a pure volume return. The freshest matched window sharpens this: Jan–Jul 2026 value is +13.5% year-on-year ($148.1M → $168.1M) while weight fell 44.2% ($1.53M → 856k kg) over the same two periods — value and volume now point in opposite directions, a divergence the original case did not have the data to see. Source: UN Comtrade public preview API, reporter USA (842), HS 811292, flow M, monthly/annual, single dimension combination (partner2=0, mot=0, customs=C00). This is the gallium/germanium/indium basket, all origins — not gallium alone and not the China/Taiwan/Malaysia breakdown the germanium case tabulates; it tests the "all origins, all partners" recovery claim directly instead.

This is the case's distinct signature — call it the import-recovery / export-shortfall asymmetry: the buyer's imports of the banned material climb back to pre-ban levels while the dominant seller's own recorded exports stay 36–66% down. When the buyer has no domestic mine and no genuine ally can produce the primary metal, a full import recovery cannot be genuine substitution the way germanium oxide was — the recovered tonnage has to be Chinese primary metal arriving under a relabelled origin (or through grey channels). The asymmetry is visible at the aggregate mirror level even though the per-shipment flow is invisible in free data (see method — HS 811292 is a basket).

Corroboration — China is policing its own leak

The strongest independent confirmation that this is trans-shipment, not honest diversification, is that China itself is chasing it. In May 2025 China's State Export Control Work Coordination Mechanism Office convened 10+ central ministries in a coordinated crackdown on the smuggling and trans-shipment of restricted critical minerals (gallium, germanium, antimony named). A dominant supplier does not mount an interagency enforcement push against a leak that isn't happening. Reuters likewise documents the grey channel qualitatively — one US buyer describes obtaining "about 200 kg of gallium a month from China" despite the ban, and ~1.1 t of China-origin gallium is estimated to have reached the US since the controls, most of it drawn down from pre-control European stocks rather than shipped direct.

The mechanism, seen cleanly on gallium's twin

The one place the relabel is legible in customs mirrors is gallium's ban-twin, germanium, and it shows exactly the shape the gallium aggregate implies: in 2024, China exported ~5,900 kg less germanium to the US and ~6,150 kg more to Belgium than in 2023 — while shipping the same combined total to the two destinations across both years (Stimson/mirror analysis). The molecules did not leave the system; the label moved from the US to a third country. For gallium, CSIS notes the US "sourced alternative supplies from Japan, Germany" after zeroing out China — Germany being a pre-control top-three destination for Chinese gallium and the natural relabel hub, exactly as Belgium is for germanium. Free HS-6 data cannot isolate the gallium-specific kilograms by country (the basket problem below), so this leg stays inferred from the twin, not independently gated.

The price wedge says the tap is real

Origin-laundering only pays when the material is genuinely scarce ex-China — the same fact the price wedge prices:

• Aug 2023: Rotterdam gallium +43% in one month on the licensing news.
• Early 2024: international gallium ~2× the Chinese domestic price.
• May 2025: Rotterdam $687/kg — +150% over pre-control.

A widening wedge and a flow recovery-with-asymmetry are two independent readings of one fact: ex-China gallium is genuinely scarce, so the volume the US is somehow still importing at these prices is worth relabelling.

Why it matters for the buyer

1. "Diversified on paper" is the trap here, not the comfort. Unlike antimony (where the US openly leaned on China), the US "never imported a majority of its low-purity gallium directly from China," so its books already looked diversified. But the Japanese/German supply it diversified into is secondary metal refined from Chinese primary — so a risk team reading a clean country-of-origin spread is mis-reading a chain that still bottoms out at ~98% China. The origin label is honest and the exposure is hidden underneath it. 2. A full import recovery is not an all-clear — for gallium it is the alarm. Because no genuine Western primary refiner exists, US gallium imports climbing back to pre-ban levels in 2025 is the opposite of the germanium-oxide story: there is nothing legitimate for the recovered tonnage to be. Recovery-to-pre-ban, for a material with a zero-capacity buyer and no ally producer, is the implausibility. 3. The control is binding — via both price and a policed leak. The 150% Rotterdam move prices the scarcity; China's own multi-ministry anti-smuggling push confirms the leak. Both say the tap is real even where the customs label reads Japan, Germany or "Other Asia, nes."

Method & honesty rails (why GATE 0, not GATE 1)

• The flow leg is basket-blind. Unwrought gallium trades under HS 811292,

which pools gallium with germanium, hafnium, indium, rhenium, thallium and their waste/scrap. The [germanium case](2025-germanium-ban-circumvention-transshipment.md) already tabulated this basket (US imports: China $32.8M→$53.3M→$15.8M; "Other Asia, nes"/Taiwan $3.7M→$11.5M→$20.9M, +466%, ~0% capacity; Malaysia new) and showed the surger unit values (~$375–400/kg) are consistent with gallium or indium but not germanium metal (~$1,500–3,000/kg). This case does not re-run that basket — it inherits it. The consequence: the Taiwan/Malaysia relabel tell is as much a gallium tell as a germanium one, and cannot be isolated to either at free HS-6 granularity. A US HTS-10 breakout (behind an API key) or a national customs portal isolating gallium would be needed to lift the flow leg toward GATE 1.

• The mirror figures are gallium-specific and traced (Chinese customs / GACC

via Reuters & Fastmarkets; CSIS; Reuters customs-record reporting). The germanium 5,900/6,150 kg discrepancy is labelled as germanium and used only to illustrate the mechanism gallium's aggregate implies — not asserted as a gallium figure.

• No common-ownership pipe traced (tell = N). The Youngsun → Thai Unipet →

Youngsun & Essen corporate pipe is antimony-specific; no equivalent gallium-specific namesake shell is in the public record here. The grey-channel "200 kg/month" example is a buyer account, not an ownership chain.

• Production shares: CSIS "Beyond Rare Earths" (China ~98% primary gallium,

~50% refined; Sunresin ~90% extraction resins) and USGS MCS 2024/2026 (no US primary gallium). Prices: Rotterdam via CSIS.

• Alternative-track only: this never touches buyerRelativeScore or the base

exposure — it sits next to them, exactly like the China–West price wedge.

• Inference, not accusation: the laundering read is inferred from a

zero-capacity buyer, an absent Western primary refiner, and an import-recovery that outruns any genuine substitute — reinforced by China's own enforcement. No individual shipment is asserted to be illegal.

• Generalises (R72): adds the import-recovery / export-shortfall asymmetry

variant to the detector — a signature readable at the aggregate mirror when the buyer has no domestic capacity and no genuine ally producer, even where the HS-6 flow is basket-obscured. Completes the Announcement-46 trilogy on the material where substitution is impossible: antimony (clean leak, GATE 1), germanium (mixed — genuine oxide substitution, GATE 0), gallium (no-substitute asymmetry, GATE 0).