Where did the banned metal go? Russian aluminium after the 2024 LME/US-UK ban

Trade-flow companion to the price wedge (R72), second non-China entry. The Russian-crude case showed the consumer-embargo shape: a Western ban on a sanctioned origin does not delete the molecules — it relocates the buyer, and the material can re-enter the West after a genuine industrial transformation (crude → Indian/Turkish diesel → EU) that legally re-flags its origin. This case is the metals analogue: a Western ban on Russian aluminium that did not capture the metal at all — it re-routed ~1.5 Mt/yr of it into China, the one large buyer still taking it. This is a DUAL-SCORE / alternative-track signal — never folded into any Tier-1 exposure score. The divergence between "the West stopped buying Russian aluminium" and "the underlying Russian units simply changed buyer, and now travel embedded in Chinese-origin product" IS the signal. Research, not investment advice; the re-entry leg is INFERRED from fungibility, not asserted on any single shipment.

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

Verdict

The West did not ban Russian aluminium in one instrument; it de-risked over three years and then formalised it:

• 12 Apr 2024 (US + UK). OFAC issued a determination under E.O. 14068

prohibiting US import of aluminium, copper and nickel of Russian origin produced on or after 13 Apr 2024, and a complementary E.O. 14071 determination barring the LME and CME from warranting or physically settling new Russian metal. (US Treasury press release JY2249.)

• 24 Feb 2025 (EU). Council Regulation (EU) 2025/395 (16th package) imposed a

direct import ban on Russian primary aluminium (CN 7601) under a one-year 275 kt transition quota — roughly 80% of 2024 EU imports (~130 kt Jan–Oct 2024, ~6% of EU aluminium imports) — before a full ban.

The metal did not disappear from the market. UN Comtrade (reporter China, HS 7601 unwrought aluminium, imports) shows exactly where it went — and it is not a zero-capacity phantom flag, it is the diversion fingerprint: as Western buyers and exchanges backed away, China absorbed the displaced Russian metal wholesale.

Source: UN Comtrade, reporter China (156), HS 7601, flow = imports, annual, pulled via the free public preview API (2026-07-18). Full annual series below. Rusal capacity / world-share context: USGS Mineral Commodity Summaries (aluminium) + Rusal public filings.

Full series (Russia → China, HS 7601, gross weight): 2020 414,011 t (18.0% share) · 2021 324,135 t (11.9%) · 2022 519,178 t (26.7%) · 2023 1,239,192 t (46.3%) · 2024 1,490,928 t (44.5%). The step change lands in 2023–2024 — the window in which Western traders, automakers and the exchanges progressively refused Russian units, culminating in the 12 Apr 2024 LME/CME cut. Russia went from ~1/8 of China's primary-aluminium imports to nearly half.

Why this is the diversion shape, not the antimony trans-shipment shape: China is the world's largest aluminium producer and a legitimate buyer of primary metal — a Russia→China surge is not physically implausible the way a Thailand antimony surge is. The implausibility is not in the flow; it is in the claim the ban rests on. A Western risk team reading "we no longer touch Russian aluminium" has severed a label, not a dependency: the same Rusal units are now inside the Chinese aluminium pool that the West buys semis and articles from.

The corporate entanglement (why the ownership tell points into China, not back out)

The antimony case had a clean namesake shell routing metal back to the buyer (Youngsun → Thai Unipet → Youngsun & Essen). Russian aluminium has a real corporate hop, but it runs the other direction — Rusal integrating into China for feedstock, not a re-export pipe:

> UC Rusal (Russia — world's largest ex-China aluminium producer, ~5.9 Mt/yr) > → 30% equity stake in Hebei Wenfeng New Materials (HWNM) (China — alumina > refiner), consideration ~$262 M rising to ~$316 M, rights transferred April > 2024, securing ~1.4 Mt/yr of alumina and lifting Rusal's alumina output +25% > to 6.43 Mt in 2024.

Rusal bought into China because it lost its Western feedstock — the Nikolaev (Ukraine) alumina refinery and Australian alumina supply were cut off in 2022. So the ownership tell is genuine, but it evidences deepening Russia–China supply integration (upstream), not a shell laundering metal back to the West. That is why this case carries a strong, hard-numbered diversion leg but no traced re-export pipe — the corporate structure points the wrong way for laundering.

Ownership hops: Mining Weekly (2023-10-24), Bloomberg (2023-10-24), Interfax, Global Times, and Rusal's 2024 interim results. Presented as the publicly-reported integration — NOT a per-shipment accusation.

The re-entry leg (mode C — labelled hypothesis, not quantified here)

The half of the question a buyer actually cares about — does the banned metal come back? — is the transformation-washing leg, and it is where this case is honest about its gap. Russian primary aluminium melted, cast or extruded in China becomes Chinese-origin semi-fabricated product (HS 7604 bars/profiles, 7606 sheet/plate, 7607 foil, 7616 articles) under substantial-transformation origin rules — the exact mechanism by which Russian crude becomes Indian diesel. China is a dominant exporter of aluminium semis to the West, so at the embedded-content level the West's "no Russian aluminium" claim is undermined: fungible Russian units now sit inside the Chinese aluminium pool the West still buys from.

Why this stays GATE 0 / a labelled gap: free trade data cannot attribute the Russian-origin content of a Chinese-origin extrusion — origin rules erase it by design, exactly as they do for refined fuel. And China's aluminium-product exports to the US/EU already run into a thicket of anti-dumping and Section 232 duties, so the re-entry channel is partly throttled and partly opaque. The leg is inferred from fungibility, not measured — no figure is asserted for it.

Why it matters for the buyer

1. A ban on an origin is not a ban on a dependency. The LME/US/UK action severed Russia's direct line to Western exchanges and importers, but the metal re-priced to China and re-entered the global pool. An exposure model that scores "Russian aluminium = 0" after April 2024 is reading a customs label, not a supply reality. 2. The FEOC / origin-rule blind spot, in reverse. The nickel/bauxite cases showed Chinese-owned, Indonesian-origin metal slipping past origin screens. Here it is Russian-origin metal slipping past inside Chinese-origin product. Both defeat frameworks that screen the customs flag rather than the economic origin. 3. It confirms the control is a revenue tool, not a supply cut. The point of the LME/CME action was to reduce Russian metals revenue; diversion to China at a discount does dent Rusal's margins (the price channel), but it leaves physical availability — and Western embedded exposure — largely intact. Two independent readings of one fact: the tap is real, but it drains east.

Method & honesty rails

• Trade data: UN Comtrade, reporter China, HS 7601 (unwrought,

non-alloyed + alloyed aluminium), imports, annual. Gross weight. 2025 annual not yet published in the free preview (China's country-level detail lags/aggregates); the series ends at 2024. Committed nowhere as a per-shipment claim.

• Diversion, not trans-shipment: unlike antimony/Thailand, the surging origin

(China) can legitimately buy and consume this metal — so this case does not rest on a zero-capacity implausibility. The signal is the re-routing of a sanctioned origin's output to the one buyer still taking it, plus the mode-C re-entry that fungibility opens.

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

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

• Inference, not accusation: the re-entry leg is inferred from fungibility and

substantial-transformation origin rules, not measured; no individual Chinese aluminium-product shipment is asserted to contain Russian metal beyond what the public record states. The Rusal–HWNM hop is upstream integration, publicly reported, not a re-export shell.

• Second non-China entry (R72): extends the mode-C / consumer-embargo track

from energy (Russian crude → Indian/Turkish diesel → EU) to metals, and adds the diversion shape — a Western origin-ban that relocates the buyer rather than removing the material.