Three apparently unrelated price events in September 2026, lithium hydroxide's seaborne-domestic basis widening, cobalt's 17% monthly correction from $56,000 highs, and a sulphuric acid cost shock that has pushed processing costs above energy as the single largest C1 input, are all expressions of the same structural problem: China's export control architecture has become the dominant pricing variable across every battery raw material simultaneously, and the market has no clean way to model correlated cost risk at this scale.
Introduction
Three price stories are running simultaneously in the battery raw materials complex this week, and traders are treating them as separate trades. They are not. Lithium hydroxide (battery grade, CIF Asia) assessed at $18,750 per tonne on 2 September, up 2.7% week-on-week, while seaborne carbonate slipped 0.8% to $19,750 per tonne. Cobalt hit $46,430 per tonne on 7 September, down 4.27% in a single session and 17.52% over the trailing month. And sulphuric acid spot prices in Chile and Indonesia have climbed to $440 and $380 per tonne respectively, with acid now accounting for 11% of hard-rock lithium chemical production costs, up from roughly 3% before the Strait of Hormuz closed on 28 February 2026.
The connecting thread is not a demand shock. It is not a geology problem. It is a single correlated input, sulphur and its downstream derivative sulphuric acid, interacting with a Chinese export control architecture that has tripled the number of affected mineral tariff codes since 2023. That combination has created a cost floor that is simultaneously compressing margins at lithium converters in Australia and Chile, widening the seaborne-domestic basis in lithium, and masking the true H2 supply picture for cobalt by making any stockpile release through DRC trucking routes more expensive to execute.
Building on my analysis of S&P Global's structural cost benchmarks in August, the price architecture problem for Western critical minerals is no longer primarily about reference pricing. It is about a correlated cost variable, acid, that is now embedded in the forward curve of every battery metal at once, and that no single market participant currently prices with sufficient precision.
Lithium's Basis Trade: What the Hydroxide-Carbonate Split Is Actually Telling You
The directional split between lithium hydroxide and carbonate on 2 September is the most analytically interesting price signal of the week, and it is being underread. Hydroxide up 2.7%, carbonate down 0.8% seaborne while Chinese domestic carbonate (EXW) rises 0.3% to RMB 150,300 to 160,500 per tonne: that is not noise. That is the market pricing divergent chemistry demand in real time, with NCM sentiment softening on CATL curtailment news while LFP demand holds domestically.
The CATL Jianxiawo story is where the supply side gets genuinely complicated. The mine, carrying nameplate capacity approaching 150,000 tonnes LCE annually, had its environmental impact assessment revoked by the Yichun Ecology and Environment Bureau after more than 70 formal complaints. Benchmark has cut its 2026 mined output forecast for Jianxiawo to 32,000 tonnes LCE from 62,500, removing 30,500 tonnes from the supply side in a year when Chinese battery-grade inventories are already at their lowest since 2025. The EIA review process could stretch into Q1 2027, and Benchmark's 2027 forecast of approximately 99,000 tonnes looks increasingly contingent on a regulatory clearance timeline that has already slipped once.
The seaborne-domestic basis is the actionable read here for traders heading into Q4 contract season. Chinese domestic prices are being supported by genuine inventory depletion and the government's RMB 62.5 billion EV subsidy program, which is pulling through demand at the cell level. Seaborne prices are being pressured by late-August buyer lull, rising freight costs, and uncertainty about whether CATL's NCM curtailment signals a broader chemistry pivot or a temporary production adjustment. Off-take negotiators pricing long-dated hydroxide contracts against the CIF Asia benchmark need to be explicit about whether the relevant reference price is the seaborne spot or the DDP landed equivalent in Europe and North America, which retains a meaningful premium.
The Benchmark Q4 2025 lithium forecast had pencilled an $18.80 per kilogram CIF Asia annual average for 2026 under its upside demand scenario. Spot hydroxide at $18.75 per kilogram on 2 September puts the market almost exactly on that trajectory, but the Jianxiawo downgrade and the acid cost inflation discussed below both represent unmodelled upside risk to that figure heading into Q4.
Cobalt's Correction Is a Sentiment Trade, Not a Fundamental Reversal
Cobalt's 17.52% monthly decline to $46,430 per tonne on 7 September looks alarming in isolation. It is less alarming when you observe that the metal is still 39.28% above year-ago levels and that the correction has been driven almost entirely by sentiment around CMOC stockpile release timing rather than any actual change in the DRC quota framework or demand trajectory. This is a market that doubled off February 2025 lows, ran to $56,000 per tonne by year-end, and is now digesting that move. The structure of the trade has not changed.
The DRC quota framework remains the dominant supply constraint. The 96,600 metric tonne annual cap for 2026 and 2027 (87,000 commercial plus 9,600 strategic reserve) nearly halved peak 2024 export volumes from a country controlling over 70% of world supply. CMOC, which produced 117,549 tonnes in 2025, is now capped at a 31,200 tonne export allocation for 2026: 27% of its prior year DRC output. The company that triggered the DRC export response through aggressive overproduction is now the single most constrained actor under the regime its own behavior created.
The CMOC in-DRC stockpile, estimated at approximately 48,600 metric tonnes as of March 2025, is the market's central unknown. That figure exceeds Glencore's entire 2024 cobalt output. The pace of its release under the quota mechanism will determine whether 2026 closes in deficit or achieves partial rebalancing, and the answer is not currently knowable from public data. What traders are doing in the interim is correct: pricing the uncertainty as a short-term correction against a structurally tighter long-term backdrop. The IEA's Global Critical Minerals Outlook 2026 projects the cobalt supply gap widening from just over 15% to over 25%, reflecting the production impact of the quota regime. That is not a number consistent with a sustained bear position.
LFP's dominance in global EV deployments (over half of all battery deployments) has been used to argue that cobalt demand is structurally compromised. That argument is weakening. NCM still controls 80 to 90% of European and North American EV deployments, and the energy density and cold-weather performance limitations of LFP are reasserting cobalt-bearing chemistry as the premium-market standard. Recycled cobalt supply growing 43% in 2026 per Wood Mackenzie is a demand offset, but at 17% of total feedstock supply it does not change the structural tightness math at current price levels.
The Acid Shock: A Correlated Cost Variable Across Every Battery Metal Simultaneously
Sulphuric acid is the story that connects everything else, and it is not yet priced into forward curves with the precision the market will eventually require. When the Strait of Hormuz closed to commercial dry bulk traffic on 28 February 2026, the immediate focus was crude oil. The secondary casualty was sulphur, and the downstream casualty has been every critical mineral whose processing depends on acid leaching, which is to say: copper, lithium, cobalt, nickel, rare earths, and high-purity manganese sulphate for battery cathodes.
The IEA's Global Critical Minerals Outlook 2026 puts the exposure in stark geographic terms. The Middle East supplies approximately one-quarter of global sulphur, with roughly half of seaborne sulphur trade transiting the strait. Gulf sulphur loadings fell to nearly 400,000 tonnes in March alone, a 74.5% year-on-year decline. By April, more than 600,000 tonnes of sulphur were stranded on vessels queuing in the Gulf with no exit. China responded in May 2026 by implementing a full sulphuric acid export ban through August, replacing a 700,000 tonne annual quota with a complete cessation. The ban covers acid produced as a by-product of Chinese copper and zinc smelting, and analysts expect it may extend through year-end.
Benchmark Mineral Intelligence's cost analysis translates this into numbers that should be in every project model right now. Sulphur prices are up more than 50% since the start of the Middle East conflict. Sulphuric acid spot prices in Indonesia and Chile have climbed to $380 and $440 per tonne respectively. For hard-rock lithium processing, acid previously represented approximately 3% of C1 production cost. It now accounts for 11%, overtaking energy as the single largest individual cost component. The S&P Global Platts CFR Mejillones benchmark, the standard reference for Chilean copper miners, doubled in under seven weeks, with a 26.7% single-week spike following the China export ban announcement.
For cobalt specifically, the acid shock interacts directly with the DRC stockpile release dynamic described above. Moving 48,600 tonnes of stockpiled CMOC material through Katanga trucking routes, already disrupted by bridge collapses, against a backdrop of elevated processing acid costs at downstream refiners, compresses the economics of accelerated quota execution. The market is not fully pricing the possibility that the practical pace of stockpile release is constrained not by quota mechanics but by the landed cost of processing that material once it arrives at conversion facilities. That is an asymmetric risk the current spot price does not reflect.
As Benchmark's Will Talbot noted, the outstanding risk is that more critical minerals players cut production or shut down operations entirely. With acid holding in the $1.26 to $1.42 per kilogram range through July and into early August, still roughly eight to ten times pre-crisis levels, that risk is not abstract. It is a forward guidance revision waiting to happen at several mid-tier processors.
Institutional Flows and the Pricing Infrastructure Gap
Global critical mineral investment declined 9% in 2025, with battery metals down more than 20% and lithium capex cut by approximately 40%. The institutional capital pullback preceded the acid shock; the acid shock is now compounding the cost uncertainty that made institutional investors reluctant to commit in the first place. The problem is not risk appetite. It is the absence of a credible forward-pricing mechanism that allows investors to model acid exposure across a multi-year project timeline.
In my August coverage of S&P Global's Treasury-backed structural cost benchmarks, I noted that the $75 to $80 per kilogram NdPr oxide reference price represented the first credible non-Chinese cost anchor for rare earth off-take negotiation. The acid shock creates an analogous gap for every battery metal processed via sulphuric acid leaching. There is currently no IOSCO-compliant forward curve for sulphuric acid at any major exchange. Benchmark Mineral Intelligence's lithium forward curve (covered in August) is a model the acid market needs to replicate, because the acid price is now a primary input into the economics of every commodity that curve is trying to price.
The IEA's own recommendation is instructive here: for 11 high-risk materials, the annual cost of strategic stockpiling by non-dominant suppliers is estimated at less than $900 million, modest against the potential economic disruption of supply shocks. That figure should be benchmarked against the cost of the current acid disruption cascading through lithium, cobalt, nickel, and copper simultaneously. The market externality from having no buffer mechanism for a single correlated input is now measurable in margin compression across four major battery metal value chains at once. The IEA's largest-ever oil stock release in March 2026 provided a template. Critical mineral markets need an analogous institutional mechanism for acid, or at minimum for sulphur feedstock.
Trader positioning in cobalt reflects the uncertainty rather than a conviction view. The 4.27% single-session decline on 7 September has the character of stop-loss liquidation on long positions built during the DRC quota euphoria, not a fundamental reassessment. The bid-offer in physical cobalt metal has widened, reflecting dealer reluctance to make tight markets against an unclear CMOC stockpile release schedule. Lithium seaborne spot is similarly illiquid in the late-August buyer lull, with off-take negotiators for Q4 contracts using the widened bid-offer as negotiating cover to delay pricing commitments. The acid uncertainty is, in part, a negotiating tool. Both buyers and sellers are waiting for the cost picture to clarify before locking in multi-quarter exposure.
Key Levels to Watch
On lithium, the key number is 32,000 tonnes LCE: Benchmark's revised 2026 Jianxiawo output forecast. Any further downward revision as the EIA review extends into Q4 is a bullish catalyst for both seaborne hydroxide and domestic Chinese carbonate, and should be read as upside risk to the $18.80 per kilogram CIF Asia annual average. Watch the seaborne-domestic carbonate basis; the current gap between flat-to-falling seaborne and rising EXW China prices is the clearest signal of where demand is actually concentrating. A sustained seaborne discount to domestic would indicate that Chinese conversion economics, supported by domestic demand and the government subsidy program, are absorbing available feedstock faster than offshore buyers can compete.
On cobalt, $46,430 per tonne on 7 September is the near-term floor to monitor, with the caveat that it is a sentiment floor rather than a cost floor. The cost of production floor for high-grade cobalt hydroxide from DRC operations is substantially lower than current spot, meaning the technical support is the quota mechanism and CMOC stockpile uncertainty, not underlying production economics. A formal announcement from CMOC on the pace of in-DRC stockpile release under the 31,200 tonne 2026 allocation is the single most important forthcoming data point for cobalt price direction. Absent that clarity, the market will continue to trade sentiment rather than fundamentals, and the volatility profile will remain elevated on both sides. The IEA's projected supply gap widening from 15% to over 25% is the medium-term anchor for a long bias, but position sizing needs to account for the CMOC wildcard.
On acid, the actionable numbers are $380 per tonne (Indonesia spot) and $440 per tonne (Chile spot). Any movement below $300 per tonne in either market would signal meaningful supply rebalancing and should be treated as a margin relief catalyst for lithium converters and nickel HPAL operators. Conversely, any extension or expansion of China's acid export ban beyond August 2026, which analysts consider likely, would push costs further into unmodelled territory for off-take contracts currently being renegotiated for 2027. The $1.26 to $1.42 per kilogram acid price range sustained through July and early August is already eight to ten times pre-crisis levels; that is a permanent cost-curve reset for any processing operation that cannot source acid domestically.
The investment case across all three markets resolves to one underlying question: at what price level does sulphuric acid cost begin to impair project-level economics sufficiently to constrain supply in ways that create genuine long-term scarcity? We are not there yet for the largest, lowest-cost operators. We may already be there for the marginal tonne that Western governments are trying to bring into production to reduce China dependence. That is the structural tension that connects all three price stories in one trade.