Is the ban leaking? Molybdenum after China's Feb-2025 export-licence controls

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 controlled 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 ownership records, never asserted as smuggling on any single shipment.

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

Verdict — detector-negative despite a China-line collapse (a fifth distinct shape)

On 4 February 2025 China's MOFCOM and Customs (Announcement No. 10 [2025]) imposed dual-use export-licence controls on molybdenum items alongside tungsten, tellurium, bismuth and indium. Molybdenum is the structural outlier of that five-metal set, and it breaks the two-layer detector in an instructive way — the one Chinese line that was dominant collapsed cleanly without any laundered origin appearing to fill it.

Two facts do the work:

1. The control is narrow. Announcement 10 covers only Mo powder ≥97% purity, ≤50 µm particle size (a missile-component / electron-tube spec), plus Mo alloy granules and Mo-Re alloys. The bulk of world Mo trade — roasted molybdic oxide (MoO₃), ferromolybdenum, molybdates — is uncontrolled. 2. China is not the US's structural Mo supplier, and the US makes its own. Molybdenum is a copper-mine by-product plus two US primary mines. Freeport- McMoRan runs Climax and Henderson (Colorado) — the only two primary Mo mines in North America — which produced 37 million lb Mo in 2025 (≈16.8 kt, up from 30 M lb in 2024), and roasts/makes Mo powder and chemicals domestically (Fort Madison, Iowa). The US is historically a net exporter of molybdenum.

The controlled form (Mo powder, HS 810210): China collapsed, no launderer replaced it

Source: UN Comtrade free public preview API (reporter USA 842, flow imports, monthly, HS 810210), committed artifact `data/intelligence/molybdenum-circumvention.json`. Production/refining structure: USGS Mineral Commodity Summaries 2026 (molybdenum); Freeport-McMoRan FY2025 10-K (Climax/Henderson output).

China's direct Mo-powder line to the US near-vanished (69.8 t → 0.86 t, −99%), and — unlike antimony — nothing laundered stepped in. The replacements are exactly the genuine non-Chinese powder houses you would expect: Germany (H.C. Starck), Austria (Plansee, the world-leading Mo/W powder-metallurgy firm), plus new Canada and Korea lines. There is no surge from a ~0%-capacity origin (no Taiwan/Vietnam/Thailand/Malaysia spike). And the whole line shrank 85% — the volume did not reappear under a new flag, it simply left the import channel, consistent with the US covering the controlled fine-powder form from domestic Climax output rather than replacing Chinese imports at all.

The bulk form (Mo oxide, HS 282570): China was never the source

The real Mo trade — roasted oxide, the ~4,000-t/yr line — is uncontrolled and runs through Chile, not China:

Source: UN Comtrade (reporter USA, HS 282570), same artifact. China supplies under 1% of US bulk Mo oxide; the year-on-year decline is a Chile/demand story (Molymet volumes, moly-price softness), not a control story — there is no Chinese line here to collapse or reroute.

The unwrought (HS 810294) and scrap (HS 810297) lines tell the same story: China fell (−69% and −84% respectively — these forms are not the controlled powder spec, so the fall reflects broader Feb-2025 licence friction / Chinese demand), and the replacements are genuine — Canada, Uzbekistan (Zangezur Cu-Mo), the UK, Austria, Korea — with only a modest "Other Asia, nes" rise on the scrap line and no ownership pipe.

Why there is no common-ownership tell here

Origin-laundering only pays when a material is genuinely scarce ex-China and no legitimate non-Chinese source of the controlled form exists — the antimony and graphite conditions. Molybdenum fails both tests: high-purity Mo powder is made by Plansee (AT), H.C. Starck (DE), Molymet (CL) and Climax (US, domestic), so a US buyer losing the Chinese line substitutes to a real supplier or to its own mines rather than paying a relabelled origin. Common-ownership tell = N — and here that "N" is a structural N, not an untraced one.

Why it matters for the buyer

1. A China-line collapse is not automatically a leak. On the controlled powder line China fell −99% — as hard as antimony — yet the exposure did not re- route through a laundered flag. A risk model that flags every collapsed Chinese line as "probably leaking elsewhere" would raise a false alarm here. The implausible-origin cross-check is what tells the two apart: antimony surged into Thailand/Mexico; molybdenum surged into Germany, Austria, Canada — genuine producers — and into domestic supply. 2. The controlled form ≠ the material's real supply chain. The narrow spec (fine ≥97%/≤50 µm powder) is a rounding error against US Mo consumption, which is fed by Chilean oxide (Molymet) and two US primary mines. A buyer reading "China controlled molybdenum" should scope the exposure to the controlled form and the defense primes that need it — not to bulk Mo demand, which China never supplied. 3. Absence of a fingerprint = prior resilience. Like tellurium (already ally- diversified) and unlike antimony (no Western refiner in the ban window), the silent origin channel is the good outcome: it means genuine allied capacity and domestic production absorbed the control. Molybdenum is the cleanest case in the set of a control that bites the controlled origin without creating a circumvention corridor.

Method & honesty rails

• Trade data: UN Comtrade, US imports, monthly, HS 810210 (Mo powder — the

controlled-spec proxy), 810294 (unwrought), 810297 (waste/scrap), 282570 (oxide — the bulk uncontrolled line). netWgt is gross product weight, not contained Mo. Late months lag ~2–3 months; 2024 = clean pre-control, 2025 = post-control calendar year (includes one pre-control month, Jan-2025 — standard caveat).

• Powder volumes are small. The −99% China / −85% world moves sit on a modest

base (tens of tonnes), so read them as a directional collapse of the controlled line, not a large-tonnage event. The bulk exposure lives on the oxide line, where China is ~0%.

• Domestic-substitution inference: the −85% world powder collapse is read as

the US covering the controlled form from domestic Climax output. This is inferred from USGS/Freeport production structure (US net exporter; Fort Madison powder plant), not from a measured domestic-output-vs-imports series in the trade data. Labelled as inference.

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

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

• Inference, not accusation: no origin is asserted to be a launderer; the

finding is the absence of an implausible-origin surge and the presence of genuine producers.

• Completes the Feb-2025 Announcement-10 five-metal set (tungsten, tellurium,

bismuth, indium, molybdenum) — molybdenum is the outlier where the controlled origin was dominant, collapsed, and left no corridor.