Commodities23 October 2026 · 3,763 words · 17 min read
Nickel analysis — 2026-10-23
indonesian nickel production cutsnickel conversion marginsmhp nickel sulphatechinese refined nickel inventorylme nickel stocksnickel matte payable
Meta title: Nickel: Cuts Meet Heavy Refined Stocks | Bloodstone Research
Meta description: Indonesian matte output has fallen a fifth on sustained negative conversion margins, and SHFE nickel rose 3.05% on cut reports — but refined stocks barely moved.
Primary keyword: nickel market analysis October 2026
Secondary keywords: Indonesian nickel production cuts, nickel conversion margins, MHP nickel sulphate, Chinese refined nickel inventory, LME nickel stocks, nickel matte payable
Indonesian nickel-matte output has fallen more than 20% in a month as conversion margins remained negative through August, September and October, and the most-traded SHFE contract rose 3.05% on 21 October amid reports of deeper cuts. Those reports were unconfirmed, the rally appears to have been driven substantially by short covering, and Chinese refined inventories drew just 602 tonnes in the latest week. Indonesia's intermediate producers face demonstrable pressure; a corresponding shortage of refined metal does not yet appear in the physical data.
TL;DR
Indonesian matte output fell 21.3% from 21,100 tonnes of contained nickel in August to 16,600 tonnes in September, with October forecast at 15,800 tonnes.
The NPI-to-matte conversion margin has been negative for 48 consecutive trading days from 3 August, averaging negative $522 per tonne of contained nickel in August, negative $377 in September and negative $467 in the first half of October.
Operators have been switching RKEF lines back from matte to nickel pig iron as relative economics deteriorated.
Reports on 21 October of Indonesian cuts of 50% for MHP, 50% for matte and 20% for NPI were not confirmed by producers or officials in SMM's 22 October assessment.
The most-traded SHFE contract rose 3.05% on 21 October, with SMM attributing much of the rebound to short covering alongside a fall of approximately 14,000 lots in open interest during the morning session.
Upstream and downstream economics differ: sulphate production from MHP carried an estimated 5.8% margin on 21 October against 2.9% using matte.
Chinese refined nickel plate social inventory stood at 122,936 tonnes on 16 October, down 602 tonnes, while LME opening stock was 280,866 tonnes on 22 October.
Indonesian Output Has Fallen a Fifth in a Month
SMM reports Indonesian nickel-matte production of approximately 21,100 tonnes of contained nickel in August, falling to 16,600 tonnes in September — a reduction of 4,500 tonnes, or 21.3%, in a single month. The October forecast of 15,800 tonnes would take the cumulative reduction to 5,300 tonnes and leave output roughly 25% below August's level within two months.
Period
Matte output, contained Ni
Operating rate
August
21,100t
~30%
September
16,600t
~23%
October (forecast)
15,800t
~22%
The October figure is a projection rather than an observation; the August and September numbers are reported output.
Against reported monthly matte capacity of approximately 70,900 tonnes of contained nickel, utilisation has fallen from roughly 30% to 23% in a single month. The contraction is therefore occurring from a base already well below nameplate, which limits how far the reduction can tighten the market in absolute terms.
The mechanism behind the cuts is specific, and it is not a shutdown. SMM reports that operators have been switching RKEF lines back from matte production to nickel pig iron as the relative economics between the two deteriorated. Those lines are not idle; they are producing a different product. Substantial capacity could therefore be redirected towards matte if relative margins improved, though the speed of any switch back depends on operational constraints, feedstock arrangements and the commercial commitments made in the interim.
The Margins Behind the Switch
The contraction is economically explicable rather than operational, and the margin history establishes it as sustained rather than recent.
Period
NPI-to-matte conversion margin
August average
−$522/t contained Ni
September average
−$377/t contained Ni
First half of October
−$467/t contained Ni
SMM reports 48 consecutive trading days of negative conversion margins from 3 August, a sequence running through to mid-October. The broader matte production margin stood at approximately negative $857 per tonne of contained nickel in mid-October.
The duration is more informative than any single reading. A brief negative spread can be absorbed through contracted obligations or inventory management, but a run of that length forces an operational response, and the production figures are that response arriving. The margin improved during September, to negative $377, without halting the output decline — consistent with operators having committed to the switch rather than reassessing week by week.
Falling utilisation during a period of negative margins is a different signal from falling utilisation during a disruption. A disruption removes capability; a margin response removes willingness, and the distinction determines how quickly output can return. SMM's October assessment also describes weakening Indonesian ore prices and reduced NPI demand from stainless producers, so pressure is visible at several points in the Indonesian chain rather than at the conversion step alone.
MHP Economics and the Downstream Divide
Mixed hydroxide precipitate shows pressure through a separate route. Mid-October MHP spot prices sat below $13,000 per tonne of contained nickel against an estimated theoretical cash cost of approximately $14,500 after cobalt credits, an implied shortfall of roughly $1,500. That figure is a representative gap between a spot price and a modelled cost, not an audited loss for every producer: individual economics vary with ore grade, energy contracts, scale and cobalt realisation.
The important qualification lies downstream. SMM reports that producing nickel sulphate from MHP carried an estimated margin of 5.8% on 21 October, against 2.9% using matte as feedstock. Both conversion routes into sulphate were profitable on that date, and MHP was the more profitable of the two.
That changes the reading materially. It is not the case that the battery chain is uniformly uneconomic. Upstream intermediate producers face demonstrable pressure while some downstream processors earn positive conversion margins — which is what cheap intermediate feedstock produces. Low MHP prices that squeeze the producer simultaneously improve the economics of the converter buying it.
The pressure is therefore concentrated at a specific point in the chain rather than distributed across it, and the question for supply is whether intermediate producers can sustain losses long enough for prices to recover, not whether the chain as a whole is viable.
Reported Cuts Remain Unconfirmed
On 21 October, reports circulated of potential Indonesian production reductions of 50% for MHP, 50% for matte and 20% for NPI. No producer or official confirmation was identified in SMM's 22 October assessment, and they should be treated as reports rather than announcements.
SMM differentiates sharply between the three products. For MHP, negative production economics make reductions plausible, and a hypothetical 50% cut from current output could remove 15,000 to 18,000 tonnes of contained nickel monthly — a quantity large enough to matter to battery-chain feedstock supply. For matte, output has already contracted substantially, and halving the October forecast would remove approximately 7,900 tonnes monthly; SMM regards a sector-wide reduction of that scale as relatively unlikely given how much has already been cut. For NPI, cash margins remain positive, which makes large voluntary reductions less economically compelling.
The differentiation matters because the headline framing treats the three as a single coordinated programme when their economics point in different directions. The product with the strongest economic case for cuts is MHP, which is also where a 50% reduction would have the largest absolute effect.
The Rally Looks Like Short Covering
The most-traded SHFE nickel contract rose 3.05% on 21 October amid the reports. The coincidence of timing is clear, but the available evidence does not isolate the reports as the sole driver.
SMM's 21 October midday review attributed much of the rebound to short covering, reporting a fall of approximately 14,000 lots in SHFE open interest during the morning session. Falling open interest alongside a rising price is the signature of positions being closed rather than opened: new bullish conviction would ordinarily add contracts, not remove them.
That reading matters for how much the move should be taken as evidence. A rally driven by bearish positions being unwound tells you about the prior positioning rather than about expected supply, and it is consistent with the LME figures discussed below, which show funds carrying substantial gross shorts into the week. A market can rise several per cent on short covering without any participant forming a view that metal will become scarce.
Chinese Refined Inventories Showed Only Limited Destocking
The inventory evidence does not match the upstream contraction. Chinese refined nickel plate social inventory stood at 122,936 tonnes on 16 October, down 602 tonnes on the week — a reduction of approximately 0.5%. Shanghai bonded-zone nickel inventory was unchanged at 1,900 tonnes.
A draw of that size sits within the range of ordinary weekly variation. What the data establish is limited weekly destocking rather than a market drawing down material, and a single week's reading against a two-month production contraction establishes little in either direction.
Exchange stocks tell a similar story. LME nickel opening stock stood at 280,866 tonnes on 22 October, comprising 259,452 live warrants and 21,222 cancelled. Cancelled warrants represent metal earmarked for withdrawal, so the proportion — approximately 7.6% of the total — indicates how much is being drawn out of the system. Neither the headline stock nor the cancellation share suggests exchange-deliverable metal is being pulled away from the market.
One indicator moved differently. The Jinchuan refined nickel brand premium strengthened from ¥4,150 to ¥4,750 per tonne over the week, a rise of ¥600, against an SMM #1 refined nickel weekly average of ¥119,610. This is a movement in one producer's brand premium rather than a market-wide physical premium, and brand premiums can widen on preferences for particular deliverable material while aggregate availability is unchanged.
Product Segmentation and Transmission
The separation between intermediates and refined metal is structural, and reading the two datasets together requires it.
LME-approved refined nickel is the exchange-deliverable form held in LME and SHFE warehouses, though not every Class 1 product qualifies: delivery depends on approved brands, shapes and specifications. Matte and MHP are intermediates, converted principally into nickel sulphate for battery cathode manufacture, while NPI is a lower-grade product feeding stainless steel, which remains the largest single source of nickel demand.
These markets overlap. Matte is not categorically excluded from refining routes that ultimately produce Class 1 nickel, and sulphate and dissolved cathode compete for certain battery applications. But they do not clear together, and the conversion routes between them carry their own costs and lead times. Reduced matte and MHP output constrains battery-grade feedstock without reducing the availability of approved refined nickel already in warehouses or produced from other sources.
Whether the pressure transmits depends on three things the available data do not quantify: how much Indonesian intermediate output was destined for Class 1 conversion rather than sulphate, what alternative feedstock exists for the converters that relied on it, and how long the margin compression persists before prices adjust or capacity exits.
Prices and the Forward Curve
LME three-month nickel closed at $16,120 per tonne on 22 October, with official three-month prices quoted at $16,125 bid and $16,135 offered and cash at $15,925 bid and $15,935 offered. The SMM #1 refined nickel weekly average to 16 October was ¥119,610 per tonne. Indonesian high-grade matte was assessed at $14,145 per tonne of contained nickel FOB, 9.5% below its end-August level.
The matte decline requires care in interpretation. SMM reports a payable of roughly 93.5% against LME nickel since mid-May, so the fall in the outright matte price has largely tracked the benchmark rather than reflecting a collapse in matte's relative value. The outright decline is therefore not in itself an explanation for the deterioration in conversion margins, which depend on the relationship between matte revenue and NPI input costs rather than on the headline price alone.
The forward curve is the more telling observation. LME nickel trades in contango from cash through to the distant dates: cash bid at $15,925 against three-month bid at $16,125, with December 2027 at $16,805 bid and December 2029 at $18,380. A market pricing imminent physical scarcity would ordinarily produce backwardation, with nearby metal at a premium to forward. The curve currently indicates the opposite, which is difficult to reconcile with an expectation that Indonesian cuts will tighten deliverable supply in the near term.
Taken together, the exchange response to the 21 October reports was real but appears positioning-driven, the physical evidence behind it remains mixed, and the curve prices adequate nearby availability. That is consistent with the segmentation described above, and also with demand being the binding constraint: if battery-chain consumption has weakened enough to drive intermediate margins negative for ten weeks, removing intermediate output restores balance without creating scarcity in refined metal.
Positioning Has Gone Neutral
LME Investment Funds held 45,949.73 gross longs against 45,494.85 gross shorts in the non-risk-reducing book at 16 October, leaving net length of 454.88 lots — a 91.0% reduction on the week and approximately 0.99% of the gross long book.
The composition is as informative as the level. Gross longs fell 2,248.42 lots while gross shorts rose 2,356.39, so short accumulation accounted for 51.2% of the reduction and the decline in longs for the remaining 48.8%. That near-even split distinguishes nickel from the rest of the LME base metals complex in the same week, where one side generally dominated.
Funds have not established a net short. What the figures establish is that the category holds almost no net directional exposure, having carried a net long of 5,059.69 lots seven days earlier. The positions were recorded on 16 October, before the 21 October reports, so they do not capture any repositioning that followed — though a gross short book of 45,494.85 lots entering that week is consistent with SMM's account of the subsequent rally as short covering.
Outlook
The immediate test is whether October matte output meets the 15,800-tonne forecast, and whether any of the reported cuts are confirmed by producers. The MHP reduction is the one to watch: it has the strongest economic case and the largest potential effect, at 15,000 to 18,000 tonnes of contained nickel monthly.
The second question is whether conversion margins recover. The negative sequence since 3 August has produced a switch of RKEF capacity to NPI rather than closure, so that capacity could be redirected if the spread improved, subject to operational constraints.
The third is whether the physical evidence begins to move. Chinese refined inventories drawing faster than 602 tonnes a week, LME cancelled warrants rising from 21,222, a broader widening of brand premiums or a shift in the curve towards backwardation would each be evidence that intermediate contraction is reaching deliverable metal. None is currently present.
Key Risks
The reported production cuts are unconfirmed. No producer or official confirmation was identified in SMM's 22 October assessment, and SMM regards a sector-wide matte reduction of that scale as relatively unlikely given the contraction already recorded.
The October production figure is a forecast. SMM's 15,800-tonne estimate has not been realised, and the two-month contraction of approximately 25% depends on it.
Margin estimates are not audited accounts. The conversion-margin series and the MHP cost estimate are SMM assessments of representative economics, and individual operators' costs vary with ore grade, energy contracts, scale and cobalt realisation.
Upstream pressure does not imply downstream pressure. Sulphate production from MHP carried an estimated 5.8% margin on 21 October, so converters buying cheap intermediate feedstock may benefit from the conditions squeezing producers.
Price movement is not evidence of scarcity. The 3.05% SHFE rise coincided with the reports but appears substantially attributable to short covering, and the forward curve remains in contango.
Positioning describes placement, not conviction, and predates the news. The 16 October figures were recorded before the 21 October reports and the move that followed.
Cross-market comparison requires reconciliation. LME, SHFE and SMM prices cover different products under different conventions, and no physical premium or arbitrage is inferred between them.
Intelligence Monitoring Points
Reported cuts: whether producers or Indonesian officials confirm any of the 50% MHP, 50% matte or 20% NPI reductions.
Indonesian output: whether October matte production meets the 15,800-tonne forecast, and whether RKEF lines switch back from NPI.
Conversion margins: whether the NPI-to-matte spread emerges from its run of negative readings since 3 August.
MHP economics: whether spot prices recover towards the estimated $14,500 theoretical cash cost, and whether the sulphate conversion margin holds above matte's.
Chinese refined inventory: whether weekly draws exceed 602 tonnes, and whether Shanghai bonded stocks move from 1,900 tonnes.
LME stocks and curve: whether cancelled warrants rise from 21,222 tonnes and whether the contango flattens towards backwardation.
LME positioning: whether the 454.88-lot net long crosses into net short in the first report covering the period after 21 October.
FAQ
Has Indonesian nickel production actually fallen? Matte output fell from approximately 21,100 tonnes of contained nickel in August to 16,600 tonnes in September, a reduction of 21.3%. October is forecast at 15,800 tonnes, which has not yet been realised.
Have Indonesian producers announced 50% cuts? No. Reports of 50% reductions for MHP and matte and 20% for NPI circulated on 21 October, but no producer or official confirmation was identified in SMM's 22 October assessment.
Why did the nickel price rise if the cuts are unconfirmed? The most-traded SHFE contract rose 3.05% on 21 October, with SMM attributing much of the move to short covering alongside a fall of approximately 14,000 lots in open interest during the morning session. Falling open interest with a rising price indicates positions being closed rather than new bullish exposure.
Is the whole nickel chain losing money? No. Intermediate producers face negative margins, but sulphate production from MHP carried an estimated 5.8% margin on 21 October against 2.9% using matte. Cheap intermediate feedstock squeezes producers while improving converter economics.
Why have refined inventories not fallen alongside the cuts? Matte and MHP are intermediates feeding the battery chain, while LME-approved refined nickel serves different end uses and arrives by other routes. Chinese nickel plate inventory drew 602 tonnes to 122,936 in the week to 16 October, and LME opening stock stood at 280,866 tonnes on 22 October.
What does the forward curve indicate?LME nickel is in contango, with cash bid at $15,925 against three-month bid at $16,125 and December 2029 at $18,380. A market pricing imminent scarcity would ordinarily show backwardation, with nearby metal at a premium to forward dates.
Data and source note: Indonesian matte production figures, conversion-margin estimates and the reported production cuts are from SMM analysis dated 22 October 2026; the October output figure of 15,800 tonnes of contained nickel is a forecast rather than an observation, while the August and September figures are reported output. Operating rates are calculated against reported monthly matte capacity of approximately 70,900 tonnes of contained nickel. Conversion margins are representative estimates, with the 48-trading-day run of negative readings dated from 3 August 2026. MHP spot prices, theoretical cash costs, the nickel sulphate conversion margins of 5.8% and 2.9%, the short-covering assessment and the approximately 14,000-lot fall in SHFE open interest are SMM observations dated 21 October. The matte payable of roughly 93.5% against LME nickel refers to the period since mid-May 2026. Chinese refined nickel plate social inventory, Shanghai bonded-zone inventory, the SMM #1 refined nickel weekly average and the Jinchuan brand premium are SMM readings for 16 October 2026, compared with 9 October. LME figures are the three-month closing price of $16,120 per tonne, official three-month and cash bid and offer prices, forward closing and official prices, and opening stock, live warrants and cancelled warrants, all for 22 October 2026. Indonesian high-grade matte of $14,145 per tonne of contained nickel is an FOB assessment. The SHFE movement of 3.05% refers to the most-traded contract on 21 October. Positioning figures are from the LME MiFID II Weekly Commitments of Traders report, Investment Funds non-risk-reducing classification, positions dated 16 October 2026 and published 20 October, with the prior-week position reconstructed from the report's published change fields. LME, SHFE and SMM prices cover different products under different conventions; no foreign-exchange conversion or cross-market physical premium is inferred.
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.
