LME Warehouses Are Full of Aluminium No Western Buyer Can Use

The London Metal Exchange currently holds roughly 244,000 tonnes of aluminium available to Western markets, a level not seen in decades. The warehouses are not empty. They are full of metal that Western buyers will not touch.

That paradox was engineered on purpose. From April 2024 onward, a wave of US, UK, and EU sanctions barred newly mined Russian-origin aluminium from entering LME warehouse registration, converting the exchange’s substantial Russian holdings into stranded inventory that sits inside the system but outside the reach of Western demand.

The backdrop makes the mechanics matter. ING estimates a 2.1-million-tonne deficit for 2025, CRU projects a sustained shortfall through 2029, and the supply-demand baseline is thin enough that the composition of remaining stock now matters as much as its volume. After this piece, you will hold a structural map of the aluminium market: why the supply problem is not a demand spike or a single geopolitical event, but a multi-year architectural shift in where metal sits, who holds it, and why Western markets cannot easily reach it regardless of the headline inventory figure.

What the sanctions actually did to LME metal

The sanctions did not expel Russian metal from the exchange. They froze it in place.

The US and UK prohibition, enacted on 13 April 2024, blocked newly mined Russian-origin aluminium from entering LME warehouse registration. Pre-ban Russian holdings stayed eligible for trading, but with almost no Western buyer willing to take delivery, that metal simply stagnated inside the system.

The timing amplified the effect. At the moment the ban took hold, Russian metal already made up the overwhelming majority of exchange stock.

Russian dominance at the point of the ban: Russian-origin aluminium accounted for 91% of the 342,225 tonnes stored in LME warehouses in March 2024.

That single figure meant the prohibition instantly transformed the functional composition of the entire exchange. Almost everything on warrant was now metal Western buyers would avoid.

The stranding then deepened. Between January and May 2025, Russian aluminium holdings within the LME grew by 35%, not because more was flowing in through the front door, but because there was nowhere else for the displaced metal to go inside the exchange network.

The warrant data confirms the freeze. By June 2025, cancelled warrants on Russian aluminium (metal earmarked for physical withdrawal) had collapsed by roughly 97% relative to 2021 levels, while live warrants fell a less severe 64%. Nobody was pulling the Russian metal out.

The warrant data confirms the freeze, but reading it accurately requires understanding how LME shadow stocks, the off-warrant, in-shed metal that sits inside the warehouse network without appearing in headline figures, further obscures the true picture of accessible Western supply.

The sanctions timeline built the trap in stages:

• April 2024: US and UK ban newly mined Russian-origin metal from LME registration.
• 2025: The EU extends restrictions, barring Russian aluminium from warehouses listed with EU-affiliated exchanges.
• July 2026: European Aluminium flags that indirect routes and pre-sanctions stocks still let some Russian metal into the EU, calling for an “indirect ban” to close the loophole.

By mid-2026, the concentration had reached extremes. Russian-origin metal accounted for 93% of available on-warrant stocks in May 2026 and roughly 95% by June and July, with 234,025 tonnes of the 246,600 total on-warrant tonnes in June being Russian.

Analysis tracking LME aluminium stocks and Russian composition confirms that the concentration of sanctioned metal has pushed usable Western supply to levels that make the headline inventory figure a poor proxy for what buyers can actually access.

Here is what that concentration tells you. Every tonne of usable, non-Russian supply that leaves the system is effectively replaced in the headline figure by a tonne no Western buyer can use. The functional inventory available to Western markets is shrinking faster than the reported stock number suggests, which means anyone reading the LME headline figure is reading a number that increasingly misrepresents accessible supply.

Three years of inventory contraction in numbers

The compression did not happen all at once. It arrived as a sequence of accelerating consequences, and the numbers show exactly how the market reached today’s 244,000-tonne floor.

The clearest view comes from a three-year comparison of first-half LME holdings. The steepest break lands precisely in the year the sanctions took effect.

The 31.82% contraction between 2024 and 2025 is the sanctions-driven break point. The more moderate 12.88% decline into 2026 is the continuation phase, a market settling into its new, thinner shape.

The headline stock figure only tells part of the story. The on-warrant collapse, meaning the metal genuinely available to the market, moved faster.

The nine-month trajectory: LME stocks fell from 511,750 tonnes on 31 December 2025 to 244,525 tonnes on 9 September 2026, a decline of roughly 52% in under nine months.

On-warrant stocks hit record lows of 245,350 tonnes in late July 2026. That is the functional buffer the entire Western market has to work with.

Now place that buffer against the deficit forecasts, which do not agree on scale but agree firmly on direction:

• ING: A 2.1-million-tonne deficit for 2025, with global production of 71.8 million tonnes against consumption of 73.9 million tonnes, and deficits projected to continue at least through 2026.
• CRU: A market in deficit through 2029, with a 400,000-tonne shortfall estimated for 2026.
• Wood Mackenzie: A revised 2026 deficit of roughly 900,000 tonnes.

The steepest contraction happened in the year the sanctions bit, and functional supply has kept eroding since. What that means for your read on the market is straightforward: it entered each new disruption with a thinner buffer than the one before.

A market running below 250,000 tonnes of on-warrant stock against annual deficits measured in the hundreds of thousands to millions of tonnes has almost no room to absorb an unexpected demand surge or a fresh supply interruption. That is the baseline condition that makes 2026’s disruptions structurally different from prior years.

Where Russian metal went instead, and what that means for Western supply

The Russian metal is not sitting idle, waiting for the political weather to change so it can flow back to the West. It has been absorbed into a supply relationship that is deepening quarter by quarter.

China is the destination. The redirection shows up as a steep, sustained climb in the Russian-origin share of Chinese imports.

The single-month peak captures how complete the shift became.

A one-month snapshot: In April 2025, China sourced roughly 91% of its primary aluminium imports from Russia.

The July 2025 interlude, when Indian-origin metal reached 34% of Chinese imports, is worth reading carefully. It represents a structural ceiling, not a scalable alternative. By May 2026 the pattern had reverted hard, with Russia supplying 189,600 tonnes of China’s 208,200 tonnes of primary aluminium imports, or 91.1%.

Across the first half of 2026, Chinese imports of unwrought Russian aluminium reached US$3.76 billion, up 24.1% on the same period in 2025. That number tells you the redirection is structural, not cyclical. The probability of this metal re-entering LME-eligible Western supply within any near-term investment horizon is very low.

The consequence is already priced into physical markets on the other side of the split:

• Rotterdam P1020A premium: rose from US$190-215 to US$240-250 per tonne in early 2024.
• US Midwest all-in cost: near a record high of roughly US$4,792 per tonne, with the US tariff on Russian aluminium doubled to 50%.

The Rotterdam P1020A premium data cited here captures only the early-2024 inflection; European aluminium premiums have since continued climbing, reaching a four-year peak of US$587 per tonne in 2026 as the composition problem in LME warehouses translated directly into physical scarcity pricing.

For Western buyers and investors, the takeaway is uncomfortable but clear. You cannot plan around a Russian supply return. The premium inflation visible in Rotterdam and the US Midwest is not a temporary disruption charge; it reflects a permanent reallocation of the metal that once underpinned LME liquidity.

Why Southeast Asia’s capacity build cannot close the gap

The obvious counterargument is Asian supply. Billions of dollars are flowing into new smelting capacity across Southeast Asia, and the headline numbers sound large. The problem is that the investment is optimised for the wrong problem, on the wrong timeline, for the wrong markets.

Indonesia is running a US$7 billion downstream programme through its sovereign wealth fund Danantara, targeting national self-reliance by 2030. Vietnam is building its first domestic primary production at the Dak Nong smelter. Both are genuine long-cycle investments. Neither rescues the Western deficit.

Indonesia broke ground on six Phase-I projects on 6 February 2026, with roughly US$3 billion allocated to an integrated Mempawah complex featuring a 600,000-tonne-per-year smelter and a 1-million-tonne-per-year smelter-grade alumina refinery. Its primary aluminium exports have already climbed from 155,000 tonnes in 2024 to 511,000 tonnes in 2025.

But the feedstock ceiling is real. The Centre for Research on Energy and Clean Air warns that bauxite demand could surge from roughly 14 million to between 65 and 94 million tonnes per year.

The feedstock ceiling: At projected demand rates, Indonesia’s proven domestic bauxite reserves could be depleted in under 12 years.

Vietnam’s Dak Nong smelter targets 150,000 tonnes per year in 2026, rising to 300,000 tonnes in early 2027 and full design capacity of 450,000 tonnes by late 2027 or 2028. Even combined, ASEAN output of roughly 2.7 million tonnes in 2026 is modest against global needs.

The 77.16 Mt trajectory for global aluminium capacity growth in 2026 places the Indonesian and Vietnamese investment programmes in a broader context: even with ASEAN additions, the distribution of new capacity is overwhelmingly tilted toward integrated Asian value chains rather than the LME-warrantable grades that Western markets require.

Three structural mismatches keep this capacity irrelevant to the Western deficit:

1. Timeline: full commissioning largely lands beyond 2026, too late for the current shortfall.
2. Feedstock: the bauxite reserve risk caps how fast Indonesian expansion can scale.
3. Market orientation: these facilities are built for domestic and integrated regional value chains, not the LME-warrantable grades Russia historically supplied to the West.

Even if every announced Indonesian and Vietnamese tonne arrives on schedule, the metal is not destined for LME warehouses or Western end-users. You should not read Asian capacity announcements as a demand-side offset to the Western premium. The two markets are running on separate supply tracks.

What the structural deficit means for aluminium markets through 2027 and beyond

Pull the four threads together and a single forward view emerges. This is a framework for reading what comes next, not a prediction.

Three structural conditions will define the aluminium market through at least 2027:

• Stranded LME inventory composition: the headline stock figure is dominated by Russian metal Western buyers will not touch, so usable supply is thinner than it appears.
• The China-Russia supply lock-in: a trade relationship deepening past US$3.76 billion in a single half-year is not one that reverses on a near-term horizon.
• The Asian capacity timeline mismatch: new smelting is real but too late, too feedstock-constrained, and aimed at the wrong markets.

History frames what these sanctions have and have not resolved. The 2018 Rusal sanctions triggered sharp price spikes, pushed LME spreads into backwardation, and forced auto, aerospace, and packaging users to reconfigure supply chains at speed. The 2022 nickel squeeze showed how concentrating restrictions on high-purity metal can nearly destroy market liquidity.

The current sanctions were designed to squeeze Russian revenue while avoiding those uncontrolled price shocks. They largely succeeded on the price-spike front, but produced structural bifurcation instead: tight London cash premiums coexisting with a relative physical glut inside China, and rising compliance costs as banks, logistics providers, and shipping insurers limit exposure to Russian metal even where no formal ban applies.

The longest-horizon call: CRU projects the global aluminium market will remain in deficit through 2029.

Three variables are worth monitoring more closely than headline price:

1. The EU indirect-route loophole: any move to close it, following European Aluminium’s July 2026 call, would tighten usable supply further.
2. Asian commissioning versus feedstock: whether Indonesian and Vietnamese tonnes arrive on schedule and whether bauxite availability keeps pace.
3. LME on-warrant stock erosion: the clearest leading indicator of acute tightness.

The bifurcation between tight London premiums and Chinese abundance is not a pricing anomaly waiting to be arbitraged away. It is the market’s structural response to a supply architecture that has been permanently altered, and that permanence belongs in any medium-term position or procurement plan.

The bifurcation between tight London premiums and Chinese abundance is one of several cases in 2026 where geopolitical risks in commodity markets have produced structural splits rather than uniform price shocks, a pattern with implications across copper, nickel, and energy as well as aluminium.

Building an aluminium view when the architecture has already shifted

The central insight is not a news event. It is a durable shift in how the market is built.

The LME aluminium market is now defined by a composition problem, not a volume problem. That distinction should govern how you interpret every future inventory or price figure, because the headline number no longer describes what Western markets can actually access.

The question is no longer whether the supply problem persists. The data confirms it will, through at least 2027 by the most conservative forecasts. The real question is which of the three monitoring variables moves first, and what that movement signals about whether the deficit narrows or deepens.

For procurement and investment positioning, the implication is direct. Premium inflation in Rotterdam and the US Midwest is not temporary, Asian capacity is not a near-term offset, and the strategic response is to build for a structurally tighter Western aluminium market rather than wait for a correction the architecture does not support.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results. Financial projections are subject to market conditions and various risk factors.