Why Lithium’s Price Correction Is Less Bearish Than It Looks

Lithium carbonate has fallen roughly 12% from its May peak. It is also trading at more than double what it averaged a year earlier. Both statements are true at the same time, and that is where the story starts.

That dissonance is the whole analytical question. A commodity in correction that still sits 106% above its 2025 average is not behaving like a market in retreat, yet the surplus forecasts arriving in September 2026 read like a warning.

The timing is what makes this a genuine test. Mine restarts are now feeding concentrate back into the market, analysts are revising surplus projections sharply upward, and the 2025-2026 rally is finally being asked whether it was built on structural demand or on a speculative risk premium that could evaporate. What follows here separates the developments that are durable from the ones that are reversible, and explains what the wide gap between competing surplus forecasts actually means for anyone holding lithium-exposed assets or watching the space.

How far lithium prices have actually fallen, and from what starting point

Start with the numbers, because the interpretation only makes sense once the scale is fixed. Battery-grade lithium carbonate traded at roughly US$8.00-8.90/kg on a CIF Asia basis in May 2025. Twelve months later it had more than doubled.

Chinese spot carbonate peaked at approximately US$27,000/t on 13 May 2026. That was the top of an extraordinary recovery driven by demand re-acceleration, the end of supply chain destocking, and resilient electric vehicle and energy storage demand outrunning near-term supply.

The correction followed. Cochilco’s sector reporting assessed the August 2026 average CIF Asia spot price for lithium carbonate at US$19,525/t, a 12.2% decline from the May peak. Note that Cochilco’s exact spot figures could not be independently verified in publicly accessible 2026 reports; Benchmark Mineral Intelligence assessed battery-grade carbonate (CIF Asia, spot) at US$18,160/t on 10 August 2026 and US$19,750/t on 2 September 2026, tracking the same range.

Here is the number that reframes the rest. That August average sits 106% above the 2025 average. The correction looks large as a percentage of the peak, but it leaves the market at a price most producers would have called optimistic a year ago.

The pullback is also visible across the wider battery materials complex, not just carbonate. On 2 September 2026, Benchmark assessed lithium hydroxide (battery grade, spot, CIF Asia) at US$18,750/t and 6% spodumene (FOB Australia) at US$2,254/t.

Benchmark raised its 2026 average price forecast for Chinese carbonate to US$20,100/t, while flagging that the rally had moved ahead of the fundamentals.

“The mid-year rally had outrun market fundamentals,” Benchmark Mineral Intelligence noted in its September 2026 assessment.

Without this coordinate system, the correction reads as a bearish signal. With it, the read you should take is different: this is a normalisation event inside a structural recovery, not a reversal of it.

The supply-side events that broke the risk premium

If the risk premium collapsed, the reason is that supply came back from three directions at once. What matters is not any single restart but the fact they arrived together, from mechanisms that behave very differently.

Australia: price-elastic restarts return concentrate supply

Australian hard-rock supply moved fastest, because it responds directly to price. Mineral Resources placed the Western Australian Bald Hill mine into care and maintenance in November 2024 during the downturn, then reversed course on 19 May 2026, citing the sustained price recovery.

Crushing and mining commenced in June 2026, with first concentrate targeted for July 2026. Bald Hill carries an annual spodumene capacity of 140,000 metric tonnes, with first shipments from Esperance expected in Q1 FY27 and full capacity by Q2 FY27.

The Ngungaju processing plant resumed in July 2026. The speed of both is the point: Australian concentrate can switch back on within a quarter, which means it can also switch back off, making it the most reversible supply in this group.

Zimbabwe: policy mandates restructure export flows

Zimbabwe is a different animal. This is not a price-response restart; it is a government-mandated restructuring of export flows aimed at forcing domestic processing.

• 2022: Order prohibiting the export of unbeneficiated lithium-bearing ore.
• February 2026: Suspension of raw mineral and concentrate exports.
• April 2026: Export quotas introduced, with mandatory local processing plans and written timelines for lithium sulphate plants required before shipments resume.
• 1 January 2027: Full ban on unprocessed concentrate shipments takes effect.

On the levy, sources conflict. The original reporting cited a 16% concentrate levy; subsequent research could not confirm that figure and identified instead a 10% export tax on concentrates holding until the January 2027 ban. The confirmed figure is 10%, though the discrepancy is worth flagging. Either way, the mechanism here is structural, not cyclical, which means Zimbabwe’s export shift does not simply unwind if prices dip.

Argentina: DLE technology moves from pilot to proof

The most technically significant development came from Salta, Argentina. Eramet’s Centenario-Ratones project offered the first commercial-scale validation of direct lithium extraction (DLE), a process that pulls lithium directly from brine using absorption rather than the slow evaporation ponds the industry has relied on.

The plant was commissioned in July 2024 and delivered first production in December 2024. It produced 6,690 tonnes of lithium carbonate equivalent (LCE) in 2025 and 8,440 tonnes in the first half of 2026.

Centenario averaged 85% of nameplate capacity in Q2 2026 and reached 90% of its 24,000-tonne nominal capacity by June 2026, with 2026 guidance of 17,000-20,000 tonnes LCE. A Phase 2 feasibility study is underway targeting roughly 30,000 additional tonnes LCE per year, with a longer-term path toward approximately 150,000 tonnes annually.

That 90% figure is what turns DLE from a promise into a supply pathway. When you have price-elastic restarts in Australia, policy-driven restructuring in Zimbabwe, and technology-validated new production in Argentina all landing together, the read for you is that this supply increase is not easily reversed by a single price pullback. Only one of the three switches off cheaply.

What the surplus forecasts actually say, and why they disagree so sharply

The correction has prompted analysts to lift their surplus projections. What is striking is not that they agree the market is loosening, but how far apart they are on by how much.

Cochilco-sourced figures revised the 2027 LCE surplus upward from 9,300 to 18,900 metric tonnes, with the 2028 global surplus projected to widen to roughly 98,000 metric tonnes. Set that against Morgan Stanley, which forecasts surpluses above 650,000 tonnes in both 2027 and 2028, and the range is not a rounding error.

The gap is not about data. These models weight the same inputs differently: how fast demand grows, how quickly high-cost supply shuts in when prices fall, and how many announced projects actually execute on time. Change those assumptions and the output swings by hundreds of thousands of tonnes.

There is a second reason headline numbers mislead. A carbonate surplus can coexist with a hydroxide near-deficit, because the two products serve different battery chemistries. Benchmark’s separate assessments of carbonate and hydroxide reflect exactly that segmentation, and it explains how a market can post a surplus in one product while another category supports premium pricing.

Demand is the counterweight underneath all of it.

The International Energy Agency projects total lithium demand will rise from 165,000 tonnes in 2023 to over 531,000 tonnes by 2030, driven predominantly by EVs and battery storage.

For you, the divergence is the signal. When credible forecasters land between 18,900 tonnes and 650,000 tonnes, the market is telling you it sits at a genuine inflection point where the outcome depends on variables not yet resolved. That points toward scenario planning and position sizing, not conviction in a single price forecast.

The 2022-2024 bust as the correction’s interpretive frame

The current correction becomes far more legible once you set it against the last cycle. The market has done a version of this before, and the template it left behind is the most useful reference point available.

Prices exceeded US$80,000/t at the late-2022 peak. What followed was an 80-90% collapse, dragging carbonate down to roughly US$13,000/t.

The catalyst was a supply-demand mismatch. Global lithium supply grew by over 50% between 2022 and 2024, overwhelming a market that was simultaneously hit by slowing EV sales, expiring Chinese subsidies, and supply chain destocking.

The mechanism that ended that collapse is the same one now operating in reverse. Falling prices made high-cost production uneconomic, forcing shutdowns and care-and-maintenance decisions, which starved the market of the investment it needed and set up the next deficit.

That sequence tends to run in three phases:

1. Supply expands rapidly into optimistic demand forecasts, generating severe oversupply.
2. The oversupply collapse shuts high-cost mines and kills new investment.
3. The resulting underinvestment eventually tightens the market back toward deficit and a price recovery.

Recognising that template changes how you read September 2026. Wood Mackenzie has already warned that supply deficits could emerge as early as 2028 without significant new investment, though that projection carries a lower verification status and should be treated as a scenario rather than a settled call.

Anyone anchored to the 2022 peak as normal, or to the 2024 trough as the floor, is working without the right reference points. The cycle’s self-correcting logic is what makes today’s prices interpretable, and Benchmark’s US$20,100/t 2026 average forecast is the market’s current best estimate of where the cycle steadies.

Where the market stands as supply and demand forces reset

As of September 2026, the balance is genuinely two-sided. The correction is real, the surplus projections are material, and yet the structural demand trajectory and cost-curve dynamics mean the bear case requires a specific set of assumptions to hold true all at once.

The current coordinate set is Benchmark’s early-September assessments: carbonate near US$19,750/t, hydroxide at US$18,750/t, and spodumene at US$2,254/t. Beneath those prices sits the IEA’s demand trajectory toward 531,000 tonnes by 2030, which functions as a structural floor rather than a ceiling.

The question for you is not whether the correction continues. It is whether current prices still reflect a structurally improved market relative to 2025. On the evidence, the answer is yes, with conditions attached.

Three variables will decide whether the correction deepens or stabilises over the next 12-18 months:

• Demand pace: how fast EV and battery storage demand actually grows against the IEA’s trajectory, since even a 650,000-tonne surplus can be absorbed if demand compounds quickly.
• Cost-curve exits: how much high-cost production leaves the market as prices fall, because the faster mines shut in, the sooner the surplus narrows.
• DLE execution: whether projects like Centenario-Ratones deliver their Phase 2 expansion of roughly 30,000 additional tonnes LCE per year on schedule, with a longer path toward 150,000 tonnes annually.

For producers already operating, the 106% premium over 2025 averages means the current environment remains comfortably profitable. For those planning restarts, the same premium is the incentive that keeps new supply coming, and the reason the correction may prove self-limiting rather than self-reinforcing.

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, and financial projections are subject to market conditions and various risk factors. Forward-looking surplus and demand projections cited here are speculative and subject to change based on market developments.