Citi Still Bullish on Lithium After Stop-Loss: AI Energy Storage to Drive H2 Restocking, 4% Supply Deficit Expected in 2026
Miles Bennett
Citi's lithium futures longs lost 17%–22% in a month, forcing a stop-out — yet the bank reaffirms its bull call, projecting a 4% global lithium supply deficit in 2026 as AI data-center power demand supercharges battery storage and reshapes lithium into a dual-engine commodity.
Why is Citi still bullish after losing money?
Two lithium longs opened May 12 were stopped out within a month: the July hydroxide contract lost roughly $4,270/t, the August carbonate contract roughly $4,450/t — drawdowns of 17%–22%.
Citi frames the stop-out as a risk-management action, not a directional reversal. This means → the bank attributes the selloff to three temporary pressures — GFEX inventory buildup, faster supply response, and near-term demand uncertainty — not to a fundamental shift.
Price targets stay unchanged: 0–3 month China lithium carbonate at $40,000/t, 6–12 month at $20,000/t, with a reiterated call that prices are highly likely to test the RMB 250,000/t level.
How certain is the 2026 supply deficit?
Citi's supply-demand balance traces a clear arc: 2024 surplus 133kt LCE (~9%) → 2025 narrows to 97kt (~6%) → 2026 flips to a deficit of 75kt LCE, roughly 4%. Demand of 2.08Mt LCE against supply of 2.005Mt.
In plain terms = lithium has been "too much to sell" for two years; in 2026 it becomes "not enough to go around."
Three supply bottlenecks persist simultaneously: a major Jiangxi mine restart stalled by environmental approvals, Australian mines constrained by high diesel costs and slow ramp-up, and South American brine projects scaling sluggishly. This reflects a reality where capacity cannot return as fast as demand grows, even when the price signal is loud.
How did AI become lithium's second demand engine?
The logic chain: AI data centers consume surging power → grid-level power gaps → rigid demand for battery energy storage systems (BESS — large lithium batteries that act as grid-scale "power banks" for peak shaving) → lithium consumption rises in lockstep.
The numbers confirm it: BESS demand grew 51% in 2025, far outpacing EV battery demand growth of 26%. Global BESS deployment is forecast to jump another 57% in 2026. This means → lithium's demand structure is shifting from "EV single-engine" to "EV + BESS dual-engine."
In the first four months of this year, China's total battery output hit 874 GWh, up 41% year-on-year; storage batteries alone surged 99% — nearly doubling.
What triggers the second-half restocking?
CATL's 2026 battery output target is 1.2 TWh, implying second-half production of 680 GWh — up 32% half-on-half. New capacity is expected online by August–September, adding roughly 5,000–6,000 tonnes of incremental lithium demand.
A second catalyst: China's 2027 export-tax-rebate cut is approaching, likely triggering a wave of front-loaded battery-cell exports in H2 2026. In plain terms = exporters rush shipments before the policy hits, propping up upstream raw-material prices.
China's visible lithium carbonate inventory remains below 20 days of consumption; as of the first week of June, total stock stood at 98,786 tonnes, down 630 tonnes week-on-week — physical inventory is low, leaving room for restocking.
What do the lithium majors' earnings tell us?
SQM posted Q1 2026 revenue of $1.76 billion, up 69.8% year-on-year; lithium-segment revenue hit $1.19 billion, up 135.7%. Realized lithium prices rose roughly 95% year-on-year; volumes climbed 25% to about 69kt LCE. SQM raised its full-year lithium sales growth guidance from 10% to 15%.
Albemarle (ALB) Q1 net income surged 547% year-on-year to $319.1 million; adjusted EPS of $2.95 crushed the Wall Street consensus of $1.09. Energy-storage adjusted EBITDA margin jumped from 24.8% to 46.5%.
This reflects a lithium price recovery already showing up in top-line profits at the majors — and the traditional peak season has not yet begun.
What is the biggest risk to this bull thesis?
Citi itself flags a longer-term cooling signal: 2027–2028 could see a return to slight surplus (roughly 25kt and 24kt LCE, respectively). This means → even if the 2026 deficit materializes, it may not herald a multi-year upcycle — more likely a "spike deficit" sandwiched between surplus years.
Two variables will decide the outcome: ① whether the 2026 supply gap actually arrives as modeled; ② whether BESS demand can keep absorbing incremental supply. If either falls short, the durability of lithium's price rebound weakens.
Citi's stop-out is itself a reminder: directional conviction and trade timing are two different things — short-term volatility can make the right call lose money.
Content is for reference only, not financial advice.