Last week's report concluded that net positions were becoming less informative because the mechanics beneath them had diverged. This week the direction has become clearer. Investment funds became less bullish or more bearish in five of the six metals to 11 September, with nickel the sole exception. Copper saw substantial long liquidation while aluminium and zinc attracted new shorts. The complex is not rotating uniformly bearish so much as stepping back through different routes.
TL;DR
- Copper investment-fund net length stood at +40,773.38 lots on 11 September. The LME reports a weekly reduction of 5,233.83 lots, driven by 5,692.43 lots of long liquidation while funds also covered 458.60 shorts.
- That liquidation occurred across a week containing three-month copper's all-time high on 10 September and the same-session reversal that followed the Reuters report on US tariff hesitancy.
- Aluminium net length stood at +147,601.71 lots, back below the 150,000 threshold crossed a week earlier. The mechanism reversed: shorts rose 2,256.18 lots after falling 5,169.75 the previous week.
- Zinc's recent long accumulation paused. Longs fell 216.00 lots after rising 4,667.72 a week earlier, while shorts added a further 3,305.69, leaving net length at +59,123.26.
- Measured as a share of open interest, zinc rather than aluminium is where the fund footprint is largest: 31.45% of long open interest and 16.51% of short, the biggest presence on both sides in the complex.
- Aluminium's fund short share remains exceptionally small at 4.19% of open interest, below the short-side share in any other metal in the group.
- Lead is the only metal where funds are a larger share of shorts than longs, at 23.90% against 13.50%, and its net short widened to −25,779.92 lots as both sides expanded.
- Nickel was the only metal to strengthen, with the LME reporting an increase of 1,129.79 lots to +16,286.74 as longs grew 2,926.09 against 1,796.30 new shorts.
- All figures are LME Investment Funds, non-risk-reducing, reported in lots. Lot sizes differ by metal and positions cannot be aggregated across the complex.
Investment-Fund Positioning Board
| Metal | 11 Sep long | 11 Sep short | 11 Sep net | LME reported Δ net | Δ longs | Δ shorts |
|---|---|---|---|---|---|---|
| Aluminium | 186,017.82 | 38,416.11 | +147,601.71 | −2,481.69 | −225.51 | +2,256.18 |
| Zinc | 124,508.26 | 65,385.00 | +59,123.26 | −3,521.69 | −216.00 | +3,305.69 |
| Copper | 64,969.57 | 24,196.19 | +40,773.38 | −5,233.83 | −5,692.43 | −458.60 |
| Nickel | 49,582.53 | 33,295.79 | +16,286.74 | +1,129.79 | +2,926.09 | +1,796.30 |
| Tin | 3,119.00 | 671.00 | +2,448.00 | −169.00 | −87.00 | +82.00 |
| Lead | 33,441.26 | 59,221.18 | −25,779.92 | −1,615.09 | +751.24 | +2,366.33 |
Source: London Metal Exchange MiFID II Weekly COTR reports, positions as at 11 September 2026, published 15 September 2026. Weekly changes are those reported by the LME in the 11 September files. The implied 4 September positions differ slightly from the figures published in the 4 September COTR vintage, most notably in copper, where the difference is 190 lots. Bloodstone Research preserves each contemporaneous LME vintage rather than retrospectively replacing previously published figures.
Five of the six metals moved against the bullish case. Read across the complex, the weekly changes are best compared in tonnage rather than percentage terms, because lot sizes are 25 tonnes for aluminium, copper, zinc and lead, six tonnes for nickel and five for tin. On that basis the LME's reported weekly changes represent a reduction of 130,846 tonnes in copper, 88,042 in zinc, 62,042 in aluminium and 40,377 in lead's net short, against an increase of 6,779 tonnes in nickel and a reduction of 845 in tin.
That ordering matters because percentage moves flatter the smaller markets. Nickel's improvement and tin's decline both look substantial expressed as proportions of their own bases, but at six and five tonnes a lot respectively they involve quantities of metal an order of magnitude below the three large 25-tonne markets. Copper is the largest move on both measures, which is what gives this week's board its shape.
Market Share and Participation
| Metal | Fund long, % of open interest | Fund short, % of open interest | Position holders |
|---|---|---|---|
| Zinc | 31.45% | 16.51% | 300 |
| Aluminium | 20.29% | 4.19% | 309 |
| Copper | 15.22% | 5.67% | 238 |
| Nickel | 15.21% | 10.21% | 283 |
| Lead | 13.50% | 23.90% | 236 |
| Tin | 12.73% | 2.74% | 95 |
Source: London Metal Exchange MiFID II Weekly COTR reports, positions as at 11 September 2026. Open-interest shares are as published for the Investment Funds category, non-risk-reducing. Holder counts are the exchange's combined figure for all Investment Funds positions in each contract.
This is the comparator the positioning board cannot provide. Open-interest shares require neither a lot-size conversion nor a prior-period base, so they measure something the lot columns cannot: how large the reported Investment Funds positions are relative to each market's open interest.
On that measure the complex looks different from the way the net column presents it. Zinc carries the largest fund footprint, at just under a third of all long open interest and more than one lot in six on the short side. Aluminium's long share is substantial at 20.29%, but its short share of 4.19% is the smallest of any metal, which is the quantitative version of the point made here last week. Lead inverts the pattern entirely, with funds accounting for 23.90% of short open interest against 13.50% of long — the only metal in the complex where they are a larger presence on the bearish side.
Holder counts add a second dimension, though a narrower one. Aluminium and zinc draw the widest participation at 309 and 300 respectively, while tin's 95 holders confirm how small that market remains. These figures describe the breadth of participation within the category and should not be read as evidence about how positions are distributed within it.
Copper: The Long Book Contracts Through the High
Copper investment-fund net length stood at +40,773.38 lots on 11 September. The LME reports a weekly change of −5,233.83 lots, or 11.38%, against the prior-period base embedded in its latest COTR vintage. Compared with the +45,817.21 actually published in the 4 September report, the vintage-to-vintage decline is 5,043.83 lots, or 11.01%. Either measure makes copper the largest proportional and largest tonnage reduction in the complex, and the analytical conclusion does not depend on choosing between them.
The composition is unusually clean. Funds removed 5,692.43 gross longs while simultaneously covering 458.60 shorts, so total gross investment-fund positioning in copper contracted by 6,151.03 lots across the week. This was not bears arriving. It was bulls leaving, with a small number of bears leaving alongside them, and the long book finished the week at 64,969.57 lots against a short book of 24,196.19.
Copper carries 238 investment-fund position holders, fewer than aluminium's 309 or zinc's 300, while funds account for 15.22% of long open interest. That places the move in useful context: investment funds are a meaningful but considerably smaller share of copper's long side than they are in zinc. The holder count describes the breadth of participation in the category, but the COTR does not disclose how the week's 5,692.43-lot reduction was distributed among those holders.
The timing gives that reduction more weight still. The reporting week running to 11 September contains the sequence that last week's report could only anticipate. Three-month copper reached an all-time high on 10 September before reversing more than 3% within the session after Reuters reported that the White House had not decided whether to extend tariffs to refined copper — a question that rests on a Commerce Department report now more than two months past its deadline.
Last week's report established that investment funds entered that move less bullish than a week earlier. This week's data establish what happened across it: the long book shrank materially during the week in which copper set its record. The reports are point-in-time snapshots and do not reveal the sequence within the week, so it cannot be said that funds sold the high specifically. The divergence is nevertheless notable: record prices were accompanied by a materially smaller investment-fund long book rather than fresh long accumulation.
That sits comfortably alongside the structure of the rally itself. A price supported substantially by tariff-arbitrage flows and a drained LME live stock position carries a policy premium that investment funds appear to have been reducing rather than chasing. The gap between a record price and a contracting speculative long is the most informative single observation in this week's board.
Aluminium: The Mechanism Reverses
Aluminium net length stood at +147,601.71 lots, taking it back below the 150,000-lot threshold crossed the week before on a reported weekly change of 2,481.69 lots. The headline decline is modest, but the composition matters considerably more than the size.
Last week's increase was produced entirely by short covering, with shorts falling 5,169.75 lots while longs fell 4,035.55. This week that mechanism inverted. Shorts rose 2,256.18 lots while longs were almost unchanged at −225.51. The bears who had been leaving aluminium have started returning, and they account for practically the whole of the weekly deterioration.
The argument made here last week was that a net long increasingly dependent on disappearing shorts has a finite runway. With investment funds representing just 4.19% of aluminium short open interest, considerably below their short-side share in any other metal in the group, the scope for that mechanism was already becoming narrower. Its reversal this week is therefore more informative than the modest decline in net length alone suggests.
Aluminium remains by a wide margin the largest investment-fund net long in the complex at 3.69 million tonnes equivalent, with a gross book of 186,017.82 longs against 38,416.11 shorts, a ratio of 4.84 to one, and the widest participation at 309 holders. But gross long exposure has now declined in both reported weeks while the short base has begun rebuilding from an unusually low level, which leaves the largest position in the complex resting on a bullish book that is not growing.
Zinc: Accumulation Stops Where Funds Matter Most
Zinc was described last week as the strongest fresh-long story in the complex, though an increasingly contested one. This week fresh long accumulation stopped. Longs fell 216.00 lots after adding 4,667.72 the previous week and 6,311.37 the week before that, while shorts added a further 3,305.69 on top of the prior week's 3,067.46. Net length finished at +59,123.26 lots on a reported weekly change of 3,521.69, or 88,042 tonnes.
What the open-interest data add is that this is happening in the metal where the fund footprint is largest. Funds hold 31.45% of zinc long open interest, more than eleven percentage points above aluminium and more than double copper, and 16.51% of short open interest, second only to lead. Zinc is not a market in which fund positioning is one flow among many.
That changes how the flow reversal should be weighted. The progression across three reporting weeks moved zinc from near-uncontested accumulation, through a two-sided build in which longs still dominated, to a week in which only the short side grew. The gross short book now stands at 65,385.00 lots, the largest absolute short position among the five net-long metals. Zinc retains a substantial net long, and a single week of marginally negative long flow does not establish that the accumulation phase is over. But the buying has paused, the selling has not, and both are occurring in the contract where fund positions are largest relative to open interest.
Nickel: The Only Metal Moving the Other Way
Nickel was the sole exception, with net length at +16,286.74 lots on a reported weekly increase of 1,129.79, equivalent to 6,779 tonnes. The mechanics invert last week's precisely. Then, funds added 1,285.09 longs against 2,262.71 shorts and the net position deteriorated. This week they added 2,926.09 longs against 1,796.30 shorts and it improved.
Participation grew on both sides, with total gross positioning up 4,722.39 lots, so this is a market attracting engagement rather than one in which a single side is withdrawing. Nickel's book also remains among the most balanced in the complex, at 49,582.53 longs against 33,295.79 shorts, and the open-interest shares say the same thing more directly: funds hold 15.21% of long open interest and 10.21% of short, the narrowest spread between the two sides of any net-long metal. The improvement should therefore be read narrowly. Investment funds were incrementally more bullish nickel during a week when they were less bullish everything else, but the position remains genuinely two-sided and the tonnage involved is small relative to the 25-tonne markets.
Lead and Tin: Both Sides Build, and a Small Base Misleads
Lead moved from contraction to expansion. Last week both sides of the book shrank and the net short narrowed slightly. This week longs added 751.24 lots and shorts 2,366.33, more than three times as much, widening the net short to −25,779.92 on a reported weekly change of 1,615.09.
Lead is also the structural outlier in the complex. It is the only metal in which funds hold a larger share of short open interest than long, at 23.90% against 13.50%, so the net short is not an incidental feature of a small category but a position in which funds represent close to a quarter of all bearish open interest. Combined with this week's two-sided build, the pattern is one of short conviction rather than of attrition: participants are arriving on both sides, and the bears are arriving faster in the market where they already carry the most weight.
Tin fell 169.00 lots to +2,448.00, with longs down 87.00 and shorts up 82.00. At a five-tonne lot size that is 845 tonnes, against zinc's 88,042 for a broadly similar proportional move. Tin's book remains tiny and one-sided, at 3,119.00 longs against 671.00 shorts, with funds holding 12.73% of long open interest and 2.74% of short across just 95 position holders — the smallest participation in the complex by a wide margin. Weekly percentage moves in tin should not be placed beside those of the larger markets without that qualification.
Fragmentation Resolves Into Reduction
Last week's conclusion was that the net long had become less informative because superficially similar moves were being generated by opposite mechanics. This week the net direction has become consistent across almost the whole complex, but the routes to it have not.
Three distinct patterns produced the same directional outcome. Copper and tin reduced through long liquidation, with copper's 5,692.43-lot decline the dominant move in the complex. Aluminium, zinc and lead deteriorated through short building, with shorts adding 2,256.18, 3,305.69 and 2,366.33 lots respectively while their long books were flat or marginally positive. Nickel alone moved the other way, and did so with longs growing faster than shorts. A reader taking only the net column would see five metals moving together; the gross columns show that they are doing so for three different reasons.
The market-share data add a third layer. The two largest deteriorations by tonnage occurred in copper, where funds hold a moderate 15.22% of long open interest, and zinc, where they hold 31.45%. The same weekly flow therefore sits against very different backdrops in each. Meanwhile aluminium's short rebuild is starting from the smallest bear presence in the complex, which is why its reversal carries more weight than the modest lot change implies.
The important feature is that all of this is happening into price strength rather than weakness. Copper set a record during the reporting week. Aluminium's net long remains within 2,500 lots of its series high. Zinc's net length is still close to 60,000 lots. Investment funds are not capitulating on the complex; they are reducing exposure to it while headline prices remain elevated, which is a materially different signal from liquidation into a falling market.
What this does not establish is why. Positioning data record what funds did, not what they concluded, and the same pattern is consistent with profit-taking at elevated prices, with risk reduction ahead of policy events, or with a genuine reassessment of tariff-driven premia. Distinguishing between those requires the physical and spread evidence, not the COTR.
Outlook
Base case: Investment funds remain net long five of the six metals, but incremental flow continues to reduce exposure rather than add to it. Copper's long book stabilises after this week's liquidation without immediately rebuilding, aluminium's net long erodes gradually as shorts return from a low base, and zinc's short accumulation continues to outpace a paused long side.
Upside risk: Copper longs rebuild after the reporting date, indicating that the week's liquidation was a reaction to tariff uncertainty rather than to price. Aluminium shorts resume covering and zinc's long buying restarts, restoring the accumulation pattern of late August across the two largest net-long positions.
Downside risk: Copper's long liquidation continues while aluminium's short rebuild accelerates from its 4.19% open-interest base. Zinc's short book grows further in the market where the fund footprint is largest on both sides, and lead's two-sided build resolves further towards the short side. In that combination the complex moves from selective reduction to broad risk removal.
What would change the view: Simultaneous growth in gross longs and reduction in gross shorts across two or more of the 25-tonne markets would mark a return to genuine directional conviction. A resolution of the US refined-copper tariff question in either direction would also alter the positioning calculus materially, given how much of copper's recent price structure rests on that premium.
Key Risks
- Positions are point-in-time. The reports capture 11 September and reveal net weekly change, not the sequence within the week. Funds reduced copper length across a week containing a record high; they cannot be shown to have sold at the high.
- Percentages of position are not exposures. Lot sizes differ by metal, so nickel's improvement is 6,779 tonnes against copper's 130,846-tonne reduction on a similar-sounding proportional base.
- Weekly changes and vintage comparisons differ. The LME's reported change for copper is 5,233.83 lots; measured against the figure published on 4 September it is 5,043.83. Both are valid within their own vintage.
- Holder counts describe breadth, not concentration. The reports give the number of position holders in each category but not the distribution of positions or of weekly changes among them.
- Longs and shorts do not balance. The LME's delta-hedge methodology means the two sides need not reconcile, and the exchange reports the larger side as total open interest. No inference about total open interest should be drawn by summing the two sides.
- Basis matters within the category. These figures are the non-risk-reducing split. On the Investment Funds total basis, copper net length is +40,582.38 lots rather than +40,773.38, because risk-reducing shorts exceed risk-reducing longs.
- Investment Funds is one category. It should not be treated as capturing all speculative participation in LME markets.
- Copper's price structure rests partly on policy. A tariff premium contingent on an overdue Commerce Department report is capable of unwinding faster than positioning can adjust.
- Reduction into strength is ambiguous. Profit-taking, event risk management and genuine reassessment produce the same weekly numbers.
Intelligence Monitoring Points
- Copper gross longs, at 64,969.57 lots after 5,692.43 lots of liquidation, and whether they rebuild once the tariff question resolves.
- Aluminium fund short share, at 4.19% of open interest after adding 2,256.18 lots, as the most direct measure of how much covering capacity remains.
- Zinc fund market share, currently 31.45% of long open interest and 16.51% of short, and whether the paused long side resumes buying.
- Nickel's two-sided build, and whether the long side continues to outpace the short after inverting last week's pattern.
- Lead's short share, at 23.90% of open interest, and whether the two-sided build continues to skew bearish.
- The US refined-copper tariff decision and the outstanding Commerce Department report, as the principal policy variable in the complex.
- Position-holder counts, as an indicator of whether category participation is broadening or narrowing across the complex.
- Physical confirmation through LME stocks, cancelled warrants, live tonnage and spread structure alongside each positioning change.
FAQ
Which metal moved most this week? Copper, on both measures. Net length finished at +40,773.38 lots, with the LME reporting a weekly reduction of 5,233.83 lots, equivalent to approximately 130,846 tonnes.
Did funds sell copper's record high? The data do not show that. They show that gross long exposure fell by 5,692.43 lots across a week containing the record, but the reports are point-in-time snapshots and do not reveal when within the week the reduction occurred.
In which metal are fund positions largest relative to the market? Zinc. Funds hold 31.45% of long open interest and 16.51% of short, the largest presence on both sides in the complex, which is why the pause in zinc's long flow carries more weight than its lot total alone suggests.
Why did aluminium fall back below 150,000 lots? Because the short covering that carried it above the threshold reversed. Shorts rose 2,256.18 lots after falling 5,169.75 the previous week, while longs were nearly unchanged.
Is lead different from the other metals? Yes. It is the only one in which funds hold a larger share of short open interest than long, at 23.90% against 13.50%, and it is the only net-short position in the complex.
Why can't these lot figures be added together? Lot sizes differ: 25 tonnes for aluminium, copper, zinc and lead, six for nickel and five for tin. Aggregating lots across metals produces a meaningless total. Open-interest shares are the comparable measure.
Why does the weekly change not match last week's published figure? The LME's reported weekly changes imply prior-period positions that differ slightly from those published in the 4 September vintage, by 190 lots in copper and less elsewhere. Bloodstone preserves each contemporaneous vintage rather than restating earlier published figures.
Data and source note: London Metal Exchange MiFID II Weekly Commitments of Traders reports for aluminium, copper, zinc, nickel, lead and tin, positions as at 11 September 2026 and published 15 September 2026, with weekly changes as reported by the exchange in those files. Position figures, weekly changes and open-interest shares refer to the LME Investment Funds category, non-risk-reducing, and positions are reported in lots throughout; position-holder counts are the exchange's combined figure across the category and indicate breadth of participation only. Tonnage equivalents use contract sizes of 25 tonnes for aluminium, copper, zinc and lead, six tonnes for nickel and five tonnes for tin. The prior-period positions implied by the reported changes differ slightly from the figures published in the 4 September COTR vintage, most notably in copper at 190 lots; Bloodstone Research preserves each contemporaneous vintage rather than retrospectively replacing previously published figures, and both the LME-reported weekly change and the vintage-to-vintage comparison are given for copper. Open-interest shares are reported as published; total open interest is not disclosed in these files and has not been derived. Longs and shorts need not balance owing to the exchange's delta-hedge methodology, and the LME reports the larger side as total open interest. Copper price references are three-month LME throughout. COTR data are point-in-time positions and do not describe activity after the reporting date.
Sources
This document is published by Bloodstone Research for informational and institutional research purposes only. It does not constitute investment advice, an investment recommendation, an offer or solicitation to buy or sell any financial instrument, commodity or security, or a forecast of future performance. Market conditions and data may change without notice. Readers should conduct their own analysis and, where appropriate, seek independent professional advice before making investment decisions. For institutional enquiries contact research@bloodstonecapital.co.uk.
