Is the ban leaking? Superhard materials after China's Dec-2024 US export ban

Trade-flow companion to the price wedge (R72). This case completes the Announcement-46 set. The same 3 December 2024 MOFCOM instrument that banned antimony, gallium and germanium to the US also banned "superhard materials including diamond and cubic boron nitride." The antimony leg leaked hard (GATE 1); the germanium leg split by product form; here we test the superhard leg — and find the opposite of antimony: a nominally-hard "ban" that behaves like a licence. China's direct line to the US did NOT collapse. This is a DUAL-SCORE / alternative-track signal — never folded into any Tier-1 exposure score. Research, not investment advice; origin-relabelling is INFERRED from statistical implausibility, never asserted as smuggling on any single shipment.

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

Verdict — GATE 0 (detector reads muted; the control did not sever the line)

Announcement 46 targeted dual-use superhard exports with military end-use review. The tradeable superhard line is HS 710510 — diamond dust and powder, the industrial superabrasive grit that is overwhelmingly synthetic diamond, of which China makes >95% of world output (>20 bn carats/yr; USGS lists China as the leading producer of synthetic industrial diamond). On paper this is a better single-origin choke than antimony. Yet US customs data shows the line held, not collapsed:

Source: UN Comtrade (reporter USA, monthly, HS 710510), pulled via the free public preview API; committed artifact `data/intelligence/superhard-circumvention.json`. Production context: USGS Mineral Commodity Summaries — Diamond (Industrial).

China's line dipped in individual months (Apr-2025 to $0.66 M, ~21% share) but then spiked back — Oct-2025 and Nov-2025 ran $3.5 M each at 79% and 69% share. Over the full 12-month window China stayed the dominant supplier at 53% share. There is no collapse here; the banned origin kept shipping.

Why the ban didn't bind — two structural reasons

1. It is end-use-scoped, not a total cut. Announcement 46 bans dual-use superhard exports and imposes military-end-use review — but civilian abrasive/tooling grit remains licensable. Unlike antimony (a blanket bilateral cut with no Western refiner in-window), the commercial diamond-grit channel stayed open, so the direct China line persisted. A "ban" that licenses the civilian form behaves, in the flow data, like a licence. 2. Genuine non-Chinese producers exist. Antimony had no Western refiner inside the ban window; superhard grit does — Element Six (De Beers, Shannon, Ireland), Iljin Diamond (Korea), and small Japanese makers. The ~7-point China share that did move went mostly to Ireland (+46%, 11%→17%) and Japan (+38%) — the bismuth-style genuine ally-substitution pattern, not a laundered origin. (Korea/Iljin actually fell −68%, so not every genuine producer gained — the substitution is lumpy, not uniform.)

The one implausible-origin tell — Thailand (weak, transient)

The single relabel-shaped signal is Thailand: a +239% value surge (share 1.2%→4.3%) from a country that appears in no synthetic-diamond-grit production table. But three things hold it at GATE 0:

• It is small and transient. The surge is concentrated in Q1-2025

(Feb-2025 $0.42 M, Mar-2025 $0.27 M) right after the ban, then fades — H2-2025 months run ~$25 k and 2026 months ~$0–33 k. A durable laundering corridor does not evaporate after one quarter.

• It is a value spike, not a volume spike. Thailand's tonnage rose only

+21% (1,333→1,608 kg) while value rose +239% — the surge is unit-value (Feb-2025 ran ~$1,860/kg vs China's ~$800/kg), consistent with a small batch of high-grade material or a price effect, not a flood of relabelled grit.

• No ownership pipe traced. Unlike the antimony case (Youngsun → Thai Unipet

→ Youngsun & Essen), no common-ownership chain between a Chinese grit producer and the Thai exporter of record is established in any public registry/filing. Hong Kong also appears as a tiny new entrepot line ($0.25 M post, 0.6% share) — an entrepot flag worth watching but immaterial in size.

So Thailand is a latent corridor to watch, not a confirmed leak.

Why it matters for the buyer

1. A hard-sounding "ban" can leave the dependency fully intact. A risk team reading "China banned superhard-material exports to the US" would assume the grit dependency is severed. The flow data says China still supplies the majority (53%) — because the ban is scoped to military end-use and licenses the civilian form. Read the instrument's scope, not the headline verb. 2. The structural choke is upstream of the flow. China restricted the six-sided-press equipment that makes synthetic diamond grit (Aug-2024, four months before the export ban). Element Six/Iljin can absorb only at the margin (world imports still fell −7% and China held the majority) precisely because ex-China capacity cannot scale quickly without the presses. The dependency is intact even though the flow looks calm. 3. Watch the Thailand/HK entrepot lines for a durable turn. If the transient Q1-2025 Thai surge returns and persists — especially on rising tonnage rather than unit value, or with an ownership pipe — this reclassifies toward the antimony pattern. Today it does not.

Method & honesty rails

• Trade data: UN Comtrade, US imports, monthly, HS 710510 (diamond dust and

powder). netWgt is gross product weight, not contained carats; late months lag ~2–3 months and are omitted.

• Product-form caveat: HS 710510 pools natural and synthetic diamond dust,

but industrial diamond powder is overwhelmingly synthetic and China's dominance of the line is synthetic grit — so the line is a clean proxy for the controlled superhard-abrasive form. Cubic boron nitride (the second named superhard material) has no isolated HS6 line (pooled in HS 2850 / HS 6804 abrasive articles) and is covered qualitatively only — a labelled data gap.

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

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

• Inference, not accusation: the Thailand read is inferred from ~zero USGS

synthetic-grit capacity plus an implausible value surge. No individual shipment is asserted illegal; no ownership pipe is claimed because none is publicly traced.

• New shape for the corpus: antimony = hard cut → leak (GATE 1); bismuth =

hard cut → genuine ally-substitution; tellurium/molybdenum = collapse or silence with no leak. Superhard adds a distinct shape: an end-use-scoped "ban" that never severs the dominant line at all — the civilian form keeps flowing, allies pick up the margin, and only a small transient relabel tell (Thailand) appears. The lesson: classify the instrument's legal scope before reading the flow, or you will mistake a licence-in-ban's-clothing for a cut.