Managed Money holds short positions against 70.0% of palladium's entire futures open interest, the largest bearish concentration in the precious-metals complex and the only outright net short among the four. The usual reading of a position that crowded is squeeze risk. Palladium's balance sheet argues otherwise — but not because metal is abundant. Primary supply is falling, with Russian output at a two-decade low, and the market still turns to surplus because demand is contracting more than three times as fast.
TL;DR
Positioning: Managed Money held 17,434 gross short contracts on 20 October against open interest of 24,916, equal to 70.0%. The net short of 11,994 represents 48.1% of open interest and has grown 2.8 times since 8 September.
Price: The LBMA Palladium Price AM was $1,059.10 on 23 October. GFEX PD2612 fell 5.20% in the week to 22 October to ¥249.75/g, reaching a new contract low of ¥247.25.
Balance: Johnson Matthey forecasts a 214koz surplus in 2026 against a 416koz deficit in 2025, ending a run of deficits stretching back to 2012.
The surplus is demand-led: Total supply falls 265koz while demand falls 895koz. Supply contracts 2.7%; demand contracts 8.8%.
Not mainly autocatalysts: Automotive demand falls 398koz, or 4.8%. Every other demand category falls 497koz combined, or 26.0%.
Russia: Primary supply falls from 2,950koz to 2,450koz, roughly 90% of the entire global primary decline. Nornickel guided in January to 2.415–2.465Moz against 2.725Moz produced in 2025.
Trade policy: A 132.83% dumping margin on Russian palladium was determined in April, but the ITC made a negative injury finding in May and no duty orders were issued.
Market Overview
The LBMA Palladium Price AM, set in auctions administered by ICE Benchmark Administration, was $1,059.10 per ounce on Friday 23 October 2026.
In China, the most-traded Guangzhou Futures Exchange contract, PD2612, closed at ¥249.75 per gram on 22 October against ¥263.45 on 15 October, a fall of 5.20%. The weekly range was considerably wider than that net figure suggests: the contract traded from a high of ¥273.45 to a low of ¥247.25, a span of approximately 9.6%, with SMM identifying the low as a new low since the contract's listing and the third established since the National Day reopening. Umicore's sales price for fabricated palladium, a euro-denominated product price rather than a market reference, fell 6.85% over the same 16 to 23 October window.
The relationship with platinum is the more telling observation. On the Guangzhou exchange, platinum's PT2612 contract closed at ¥389.45 per gram against palladium's ¥249.75 — platinum approximately 56% above palladium. For most of the decade to 2022 that relationship ran the other way, and its inversion is the clearest market expression of the divergence in the two metals' balances.
Positioning: Short Against Seventy Per Cent of Open Interest
The Commodity Futures Trading Commission's Disaggregated Commitments of Traders report for positions held on 20 October shows Managed Money holding 5,440 gross long contracts against 17,434 gross shorts, a ratio of more than three to one and a net short of 11,994 contracts.
Palladium, Managed Money
Contracts
Share of open interest
Gross longs
5,440
21.8%
Gross shorts
17,434
70.0%
Net position
−11,994
−48.1%
Source: CFTC Disaggregated Commitments of Traders, Futures Only, positions as at 20 October 2026. Total futures open interest 24,916 contracts.
The measure requires care. Every futures contract has a long side and a short side, so a gross short equal to 70.0% of open interest means this one reported category holds short positions against 70% of outstanding contracts — not that 70% of the market is bearish. The counterparties are other categories, including commercial hedgers, swap dealers and other reportable traders, whose positions the Managed Money figures do not describe.
Within that qualification, the concentration still distinguishes palladium from the rest of the complex. The net short alone represents 48.1% of open interest, against a net long of 25.9% for gold, 9.3% for platinum and 3.1% for silver, and palladium is the only one of the four in which Managed Money is net short at all.
The position has been built quickly. On 8 September the net short stood at 4,339 contracts; by 20 October it had reached 11,994, approximately 2.8 times its earlier size. The latest week added 1,431 shorts against 136 longs, so the book continued to expand on the bearish side even as open interest rose 1,743 contracts. Palladium is also the smallest of the four markets, at 24,916 contracts against gold's 402,279.
The Balance Turns, After Thirteen Years
Johnson Matthey's PGM Market Report of 14 May 2026 forecasts palladium moving into surplus for the first time since 2012, ending a run of deficits in every intervening year.
Johnson Matthey, '000 oz
2025
2026 forecast
Change
Primary supply
6,589
6,032
−557
Secondary supply
3,135
3,427
+292
Total supply
9,724
9,459
−265
Total demand
10,140
9,245
−895
Market balance
−416
+214
+630
Source: Johnson Matthey PGM Market Report, 14 May 2026. Supply and demand sum to the stated balance in both years.
The balance swings 630,000 ounces, and the composition of that swing is not what a surplus usually looks like.
Total supply does not rise. It falls by 265koz, or 2.7%, as a 557koz decline in mine production outweighs a 292koz recovery in recycling. The market moves into surplus because demand falls by 895koz, or 8.8% — more than three times the rate at which supply contracts.
A surplus arrived at this way has different properties from one created by new production. Supply-led surpluses persist until capacity exits, which is slow and capital-intensive. A demand-led surplus closes if demand stabilises, which requires no supply response at all. At 214koz the forecast surplus is also modest against total demand of 9,245koz, equivalent to roughly 2.3%.
The Surplus Is Demand-Led — and Not Mainly by Cars
The conventional explanation for palladium's decline is the displacement of gasoline engines by battery-electric vehicles, eroding autocatalyst demand. Palladium's own demand table does not support that as the dominant mechanism for 2026.
Demand component, '000 oz
2025
2026 forecast
Change
Share of decline
Automotive
8,227
7,829
−398
44.5%
All other demand
1,913
1,416
−497
55.5%
Total
10,140
9,245
−895
100%
Automotive demand falls 4.8%. Every other category combined falls 26.0%, and contributes more to the total reduction despite starting from less than a quarter of automotive's base. Johnson Matthey attributes part of this to exchange-traded fund investment turning negative during the first quarter, though the published balance does not break the 497koz down by component.
Automotive consumption therefore remains overwhelmingly the market's foundation, rising from 81.1% of total demand in 2025 to 84.7% in the 2026 forecast — its share grows precisely because the other categories fall faster. The autocatalyst story is real and the forecast direction is negative, but a 4.8% annual decline is an erosion rather than a collapse, and it is not what moves the balance this year.
Russia Falls, South Africa Holds
The supply side is where the conventional reading inverts most sharply.
Primary supply, '000 oz
2025
2026 forecast
Change
Russia
2,950
2,450
−500
South Africa
2,246
2,235
−11
Russian primary production falls 500koz, which Johnson Matthey describes as the lowest level in at least two decades. That single decline accounts for approximately 90% of the 557koz global reduction in mine supply. South African output is effectively unchanged, down 11koz, or half a per cent.
Company-level guidance is consistent in magnitude. Nornickel, the largest Russian producer, guided in January 2026 to palladium output of 2.415 to 2.465 million ounces for the year against 2.725 million ounces produced in 2025, a reduction of approximately 260 to 310koz, or 9.5% to 11.4%. That guidance is from January and is not an updated October figure, and company production guidance and national primary-supply estimates differ in coverage and reporting basis, so the two corroborate one another without being directly comparable.
The market is consequently tightening at the mine while loosening overall. Recycling supplies the offset, rising 292koz as higher prices improve collection economics, but even combined with South African stability it does not prevent total supply from falling.
Trade Policy: A Dumping Margin That Never Became a Duty
The United States Department of Commerce made an affirmative final determination in its antidumping investigation into unwrought palladium from the Russian Federation on 28 April 2026, establishing a Russia-wide dumping margin of 132.83%.
No duties resulted. On 29 May 2026 the United States International Trade Commission made a negative final injury determination, and no antidumping or countervailing duty orders were issued from those investigations.
The sequence is worth stating precisely, because the headline margin has continued to circulate as though it were an operative tariff. A margin calculated by Commerce establishes the extent of alleged below-market pricing; an order additionally requires the Commission to find that the domestic industry was materially injured. Without that finding the investigation closes and no duty attaches. Russian palladium faces no antidumping duty in the United States as a result of this proceeding, and any assessment of trade-driven rerouting should not assume one.
China: Arrivals Slow, Spot Stays Comfortable
Chinese customs data show September imports of unwrought palladium and palladium powder at approximately 2.90 tonnes, down 13.43% from August's 3.35 tonnes and 54.26% below the approximately 6.34 tonnes imported in September 2025.
The annual comparison is the steepest in the platinum-group complex, but it measures against an exceptionally strong base. Cumulative imports for January to September stand at roughly 36 tonnes, still more than 50% above the same period of 2025, though that growth rate has decelerated from 116% at the half-year. The trajectory is a slowing of an unusually rapid expansion rather than a contraction.
Physical conditions remained unstressed. SMM reported improved purchasing interest as futures fell, but also described recycled and imported availability as relatively ample, with mainstream spot quotations approximately at parity to the PD2612 contract. That parity is itself evidence: in platinum, by contrast, ingot suppliers quoted premiums of ¥1 to ¥2 per gram over the futures contract during the same period, with available ingot described as relatively tight. Palladium's spot market shows none of the prompt-availability pressure visible in platinum's.
Substitution: Forecasters Disagree on Direction and Timing
Substitution is frequently cited as palladium's route back to strength, on the reasoning that the price inversion should encourage automakers to re-engineer catalysts toward the cheaper metal. Research providers disagree on both the direction of the next move and its magnitude.
Johnson Matthey points the other way in the near term, noting that the Euro 7 standard may initially raise the platinum-to-palladium ratio in European catalysts, because platinum-rich systems were specified for some models during the period when palladium was the more expensive metal. It expects that substitution to reverse only as post-Euro 7 thrifting progresses.
The World Platinum Investment Council's January 2026 research addresses the quantity available to reverse. It estimates that the earlier substitution of platinum for palladium in automotive catalysts reached approximately 845koz a year, and forecasts that roughly two-thirds of those gains could reverse by 2030 as platinum's premium encourages greater palladium use — approximately 563koz, or around 17.5 tonnes, of annual reverse substitution on that assumption. Heraeus likewise anticipated some reverse substitution during 2026 in its annual precious-metals forecast.
The disagreement is principally about sequencing. Catalyst reformulation requires qualification, process redesign and homologation, so neither direction responds quickly to a change in relative prices, and a forecast horizon extending to 2030 addresses a different question from the 2026 balance.
Outlook
The immediate question is whether the demand forecast underpinning the surplus holds. The 2026 reduction is weighted towards non-automotive categories rather than autocatalysts, and Johnson Matthey attributes part of that to exchange-traded fund investment turning negative in the first quarter. The published balance does not disaggregate the 497koz decline, so the proportion attributable to investment flows rather than to other industrial uses is not established here — and with it, how much of the reduction could reverse quickly.
The second is whether Russian supply falls further. At 2,450koz the forecast already represents a two-decade low, and the projected surplus is 214koz. A further reduction of that order would be sufficient to return the market to balance on Johnson Matthey's own arithmetic, provided every other supply and demand component held at its forecast level — a strong assumption across a full year.
The third is positioning. A gross short against 70% of open interest in a market of fewer than 25,000 contracts is unusual by any standard, and the CFTC data describe the scale and direction of that exposure but not the price levels, risk limits or hedging arrangements behind it. The 2026 surplus forecast is driven by contracting demand rather than expanding supply. That makes the outlook sensitive to revisions in investment flows, automotive consumption and Russian production, but the CFTC data do not establish the assumptions behind individual speculative positions.
Key Risks
Forecast status. The 2026 balance, supply and demand figures are Johnson Matthey projections published on 14 May 2026 and are subject to revision. They describe a full year rather than current conditions, and five months have elapsed since publication.
Concentration measurement. A gross short equal to 70% of open interest describes one reported category's positions against outstanding contracts. Because every contract has both sides, it does not measure the proportion of the market positioned bearishly, and the offsetting positions sit in categories the figure does not describe.
Positioning is not prediction. Managed Money's net short describes reported exposure at the Tuesday reporting cut-off. It does not establish participants' price expectations, nor that the position will be unwound.
Demand composition. The 497koz non-automotive decline is not disaggregated in the published balance. Its reversibility cannot be inferred from the investment attribution alone.
Guidance vintage. Nornickel's production range is January 2026 company guidance, not an updated forecast, and is a different measure from a national primary-supply estimate.
Instrument comparability. The LBMA benchmark, GFEX futures, Umicore's fabricated sales price and SMM physical assessments are different instruments in different currencies under different conventions. No cross-market premium or arbitrage is inferred between them.
Intelligence Monitoring Points
Positioning: whether Managed Money gross shorts extend beyond 17,434 contracts and whether the net short continues building from 11,994 after growing 2.8 times since 8 September.
Russian supply: whether Nornickel revises its January guidance range of 2.415–2.465Moz, and whether national output estimates move below 2,450koz.
Investment flows: whether exchange-traded fund positions stabilise after turning negative in the first quarter, given their weight in the demand decline.
Chinese imports: whether monthly arrivals fall further from 2.90 tonnes and whether cumulative growth decelerates below 50%.
Spot conditions: whether Chinese palladium quotations move away from parity to the GFEX contract, as platinum's have.
Substitution evidence: whether catalyst reformulation data support reverse substitution during 2026, as Heraeus anticipated, or the later timing Johnson Matthey describes.
Balance revisions: whether subsequent PGM market reporting moves the 2026 surplus estimate materially from 214koz.
FAQ
How large is the speculative short position in palladium? Managed Money held 17,434 gross short contracts on 20 October against total futures open interest of 24,916, equal to 70.0%. The net short of 11,994 contracts represents 48.1% of open interest.
Does that mean 70% of the market is bearish? No. Every futures contract has a long and a short side, so the figure measures one reported category's short positions against outstanding contracts. The counterparties sit in other categories, including commercial hedgers and swap dealers.
Why is the market moving into surplus if supply is falling? Because demand is falling faster. Total supply declines 265koz in the 2026 forecast while demand declines 895koz, producing a 630koz swing from a 416koz deficit to a 214koz surplus.
Is the decline in demand caused by electric vehicles? Only in part. Automotive demand falls 398koz, or 4.8%. All other categories together fall 497koz, or 26.0%, and account for the larger share of the reduction, with Johnson Matthey attributing part of that to exchange-traded fund investment turning negative.
Do US duties apply to Russian palladium? No. Commerce determined a Russia-wide dumping margin of 132.83% on 28 April 2026, but the International Trade Commission made a negative injury determination on 29 May 2026 and no duty orders were issued.
Will automakers switch back to palladium now that it is cheaper than platinum? Forecasters differ. Johnson Matthey expects Euro 7 to raise the platinum-to-palladium ratio initially, reversing only as post-Euro 7 thrifting progresses, while the World Platinum Investment Council estimates roughly two-thirds of around 845koz of earlier substitution could reverse by 2030.
Data and source note: Positioning figures are from the CFTC Disaggregated Commitments of Traders, Futures Only, Managed Money category, for positions held on 20 October 2026 and released 23 October, with open-interest percentages calculated against total futures open interest. The LBMA Palladium Price AM for 23 October 2026 is the benchmark set in auctions administered by ICE Benchmark Administration; the underlying historical series is available through the LBMA data portal. GFEX contract prices, trading ranges and Chinese physical-market commentary are from SMM's platinum and palladium weekly review of 22 October 2026, and Chinese customs figures from its import and export analysis of 23 October 2026. Supply, demand and balance figures are Johnson Matthey forecasts published on 14 May 2026 and are annual projections rather than observations. Nornickel's production range is company guidance issued in January 2026. Substitution estimates are from the World Platinum Investment Council's January 2026 research and are forecasts to 2030. Umicore's quoted figure is a sales price for fabricated palladium in euros per kilogram and is not a market reference price. Instruments are quoted in their original currencies and units; no foreign-exchange conversion is applied and no cross-market premium is inferred.
SMM — Platinum and Palladium Weekly Review, 22 October 2026ews.metal.com/ru/newscontent/104173941-platinum-and-palladium-suffer-heavy-losses-again-at-weeks-end-spot-market-premiums-rise-with-active-trading-smm-platinum-and-palladium-weekly-review
