What the Aluminium Selloff Signals for Q4 2026 Prices

On 29 September 2026, LME three-month aluminium settled at US$3,213.5/tonne, down US$34.5 on the session and roughly US$147/tonne below its early-month peak. The death cross on the MACD daily chart was not a warning that trouble might be coming. It was a confirmation that it had already arrived.

The selloff did not come from one direction. Bearish technical momentum on both the LME and the Shanghai Futures Exchange (SHFE) converged with pre-National Day position-squaring in China, a US consumer confidence print that undershot every Bloomberg economist estimate, and near-even Fed rate-hike odds heading into October.

Any one of these forces would matter in isolation. Together they build a pressure environment that anyone sizing positions through the holiday needs to read carefully. This aluminium market outlook maps each force with precision, examines where the physical market is genuinely tightening versus where the bearish case holds, and frames the specific variables that will decide whether the post-holiday period brings stabilisation or another leg lower. By the time you finish, you will have a clear picture of both the bearish case and the structural limits on how far it can run.

What the technicals are telling you about the September selloff

The September breakdown was not a single bad day. It was a trajectory, and the shape of that trajectory is where the signal lives.

LME aluminium touched a roughly two-month high near US$3,360/tonne on 10 September 2026 before the descent began. The path from there ran steadily downhill:

• 10 September 2026: peak of approximately US$3,360/tonne, the high-water mark for the month
• 14 September 2026: close near US$3,236/tonne, a decline MetalsCost tied directly to rising Fed rate-hike bets
• 22 September 2026: trading in the US$3,242-3,264/tonne range, with LME exchange stocks near 242,125 tonnes
• 29 September 2026: settlement at US$3,213.5/tonne, the session that confirmed the breakdown

On that final session, price sliced through the 5-, 10-, 20-, and 40-day moving averages and pushed toward the 60-day line. Breaching four moving averages in a single day is a different animal from a routine pullback. It signals the short-term trend structure itself has broken, not merely paused.

LME aluminium price volatility is not a single mechanism: it reflects the interaction of exchange positioning, physical inventory signals, and macro rate expectations across different time horizons, and the September breakdown is a case study in all three channels activating simultaneously rather than sequentially.

LME three-month aluminium, 29 September 2026: settled at US$3,213.5/tonne, down US$34.5 or 1.06%, with an intraday range of US$3,205.5-3,268.5/tonne.

The MACD daily chart printed a death cross: the DIFF line crossed beneath the DEA line, and the green histogram widened, a sign of bearish momentum still building rather than exhausting. Open interest rose to 577,000 lots, up 2,441 lots on the day, driven predominantly by new short position building.

That last detail is the one that matters most for your read. Open interest expanding on fresh shorts, rather than shrinking on long liquidation, tells you professional money is actively betting against aluminium here. It is not mechanically clearing stale positions. Fresh bearish conviction requires a catalyst to reverse; a stop-loss cascade often resolves on its own. This is the former.

The SMM aluminium morning meeting summary for 30 September 2026 provides the underlying open interest, moving average, and MACD configurations across both exchanges that the technical picture above is built from, with LME open interest rising to 577,000 lots confirming fresh short accumulation rather than long liquidation.

SHFE confirms the pattern

Shanghai told the same story on the same night, which is what turns two separate signals into one conclusion.

During the night session on 29 September 2026, SHFE aluminium settled at RMB 23,910/tonne, down RMB 110 or 0.46%, trading between RMB 23,845 and RMB 23,960/tonne. Every moving average across timeframes turned lower.

The 4-hour MACD death cross persisted, with DIFF holding well under DEA and the green histogram continuing to widen, mirroring the LME daily configuration almost exactly. Open interest climbed to 267,000 lots, up 1,187 lots, and as on the LME, the growth came from new short accumulation.

Two exchanges, two continents, the same technical alignment driven by the same source of positioning. When the bearish signal appears independently in both venues, it stops looking like a local quirk and starts looking like a market-wide verdict.

China’s physical markets are splitting, and where you sit on the value chain matters

The comforting story is that low Chinese inventories will floor the price. That story is only half true, and the half that is false could cost you.

Primary ingot is genuinely tight. Social inventories fell to 638,000 tonnes by 30 September 2026, down from a July peak of roughly 1.13 million tonnes. The pre-holiday draw accelerated across three consecutive weeks:

• Week to 17 September 2026: down 63,000 tonnes
• Week to 24 September 2026: down 55,000 tonnes
• Week to 30 September 2026: down another 55,000 tonnes

That is a 158,000-tonne cumulative three-week draw, which SMM flags as the steepest comparable pre-National Day decline in its 2020-2026 data set. Bloodstone Capital corroborates the pattern. For anyone exposed to primary ingot, this is real support that limits downside.

The steepest pre-holiday aluminium inventory cycle on record frames the 638,000-tonne ingot figure as more than a seasonal artefact; it represents a structural draw that has run consistently deeper than comparable periods back to 2020, giving primary ingot holders a genuine cost floor that billet and scrap segments simply do not share.

Billet tells the opposite story. Inventories sat at roughly 149,000 tonnes in late September, approximately 26,000 tonnes above the prior year and the highest for this point in the year in three years, according to Bloodstone Capital. The macro selloff has far less of a physical floor to land on in the billet segment.

Cast alloy sits somewhere in between: 31,300 tonnes as of 30 September 2026, down 500 tonnes week-on-week and destocking for a second straight week, with SMM describing short-term cost support as strong.

The read here is simple but easy to get wrong. The physical market is not a single bullish offset to weak technicals. Treat it as one and you build a systematically wrong risk assessment. Your segment exposure should be the first filter when you size pre-holiday risk.

Spot market behaviour ahead of the National Day break

The spot market showed exactly what a race to the exits looks like. Across both Central and South China on and around 29-30 September 2026, suppliers cut prices and liquidated holdings rather than carry stock through the break.

In Central China, downstream processors had already bought volumes at earlier, lower prices and were reluctant to add. Spot transactions changed hands at a discount of roughly RMB 40-80/tonne to the SHFE 2610 contract, which itself fell about RMB 100/tonne on the session.

South China saw brief attempts to hold prices collapse into discounting, with cargo widely available and buyers expecting heavier arrivals over the holiday. A00 spot settled at a premium of RMB 245-285/tonne to the SHFE 2610 contract, while SMM A00 aluminium closed at RMB 24,110/tonne, down RMB 60 on the session.

Scrap moved lower too, with quotes off RMB 50-100/tonne on the session and cumulative weekly declines near RMB 200/tonne. The interesting split is regulatory: deepening tax inspections across Henan, Hunan and other regions have tightened the availability of fully compliant, invoiced scrap material and kept its pricing firm, while traders holding non-invoiced stock find their sales options shrinking under compliance scrutiny and are cutting prices to clear. That bifurcation is a structural cost floor for secondary alloy, not a passing one.

How the macro picture shifts the probability distribution for aluminium prices

Two independent macro data streams turned against aluminium in late September, and the danger is that they do not simply add. They compound.

The first is US consumer confidence. The Conference Board index dropped 6.7 points to 81.9 in September 2026, down from 88.6 in August, its weakest reading since April 2014 and, per Reuters, near a 12.5-year low. The Expectations Index fell to 63.6, a third straight monthly decline, with weakness broad-based across every demographic, income and geographic group.

US Conference Board Consumer Confidence, September 2026: 81.9, the lowest since April 2014 and below every economist estimate surveyed by Bloomberg.

A print that undershoots every forecast is the kind of downside surprise that moves expectations rather than confirming them. The second stream is where that surprise gets transmitted.

Fed statement signalling for metals operates through a faster and more durable channel than the rate decision itself: the language around future hiking intentions re-prices the dollar and Treasury yields ahead of any actual policy change, which is why the near-even October split carries asymmetric risk for aluminium longs even before the meeting outcome is known.

CME FedWatch put the October meeting at almost a coin flip in late September: 50.4% for a 25bp hike against 49.6% for a hold. Looking to December, the distribution assigns just 8.5% probability to no net change, 49.8% to a cumulative 25bp hike, and 41.8% to a cumulative 50bp hike.

The transmission chain runs in three steps:

1. Deteriorating consumer confidence signals weaker growth expectations, denting discretionary metals demand.
2. Elevated rate-hike probability strengthens the US dollar and lifts Treasury yields.
3. Dollar strength applies direct downward pressure on US dollar-denominated commodity prices, aluminium included.

MetalsCost frames this as a cyclical macro headwind rather than a structural change in aluminium demand, which is the honest read. A near-even October split is not a neutral backdrop for longs. It means the next rate-sensitive data release will immediately re-price aluminium, and with fresh shorts already dominant, the asymmetry favours the downside. China’s own easing partly offsets this abroad: the PBoC cut its PSL rate 25bp to 1.5% and expanded relending quotas for technology and innovation by RMB 200 billion to RMB 1.4 trillion and for agriculture and small business by RMB 500 billion to RMB 4.85 trillion around 29-30 September 2026, though no clear aluminium demand transmission has yet been quantified.

Processing PMI recovery: what the September data actually means for near-term demand

Here is where a headline can mislead you if you stop reading at the first number.

China’s aluminium processing composite PMI recovered to 53.9% in September 2026, back above the 50-point expansion threshold for the first time in two months after two consecutive contractions. That is a genuine signal that the traditional September-October peak season provided real support to operating rates and order intake.

Then the internal indices complicate the picture.

PMI transmission to commodity prices runs through two distinct channels: the production operating rate channel, which affects near-term raw material demand directly, and the sentiment channel, which shifts speculative positioning in futures markets before any physical change occurs, and the September processing PMI recovery activated the first channel without yet triggering the second.

New export orders sub-index, September 2026: 47.3%, still in contraction and the clearest crack in an otherwise positive headline.

Two structural imbalances sit beneath the recovery:

• Finished product inventory index at 56.8%, reflecting end-user hesitancy and slowing cargo pickup rather than healthy consumption
• New export orders at 47.3%, still contracting under external headwinds

The raw material inventory index at 58.2% and purchase price index at 50.8% show processors front-loading raw material purchases. That supports aluminium demand short-term, but it also means consumption-driven ordering has not yet arrived; firms are stocking inputs, not clearing output.

Put it together and the read sharpens. Production is running while finished goods pile up at 56.8% and export orders contract at 47.3%, which means output is outpacing what buyers are absorbing. Treat the PMI recovery as a constraint on the bearish case, a reason prices should not collapse, rather than a bullish catalyst that reverses direction.

Three variables to assess before trading aluminium through Q4 2026

The bearish forces here are identifiable and quantified. The bullish limits are structural rather than demand-driven. Which way the market resolves depends on how three variables sequence, and none of them clears before the holiday ends.

1. Post-National Day restocking pace in China. Bullish outcome: buyers return and the 638,000-tonne ingot trough deepens on genuine demand. Bearish outcome: the draw proves to have been pre-holiday timing, and inventories rebuild once mills reopen. Watch the first weekly SMM social inventory prints after the break.
2. The October Fed decision and the data preceding it. Bullish outcome: soft labour or inflation prints push the odds toward a hold, easing the dollar. Bearish outcome: the 50.4% hike materialises and the dollar-strength channel re-engages immediately. Watch US labour market and inflation releases feeding CME FedWatch.
3. Scrap compliance enforcement in Henan and Hunan. Bullish outcome: enforcement persists, keeping compliant invoiced scrap scarce and holding a structural cost floor under secondary alloy. Bearish outcome: enforcement eases and that floor softens. Watch regional scrap availability and A00-to-scrap spreads. This is the least-watched but potentially most durable tightener in the market.

Baseline read: SMM frames the market as a tug-of-war between macro and monetary headwinds and physical tightness, with a range-bound outcome carrying a downward adjustment bias. Physical tightness in primary ingot caps the downside; elevated billet stocks near 149,000 tonnes and macro uncertainty cap the upside.

None of the three variables resolves before National Day ends, which makes the week of 7-11 October 2026 the first genuine information window capable of updating a well-formed view. Positioning ahead of it is exposure to timing risk, not a bet on fundamentals. Knowing which releases to watch lets you act when the picture clarifies, rather than chasing price after the move.

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, and forward-looking statements are speculative and subject to change based on market developments.