|
Part I
The Mechanism
|
Lithium.
The sell-side is calling it oversupply. Bloomberg ran a chart last week showing prices down roughly 55% from the May peak and filed it under "demand concerns." CNBC had someone explaining how EV growth is "moderating." The word "glut" keeps showing up in headlines written by people who haven't opened a warehouse manifest since 2024.
Spot lithium carbonate in China settled this week around 89,550 yuan per tonne. Down from above 200,000 in May. It looks like a market drowning in material. It is not.
Benchmark Mineral Intelligence presented the numbers at their Shanghai conference: a 171,000-tonne supply shortfall in 2026 and another 44,000-tonne deficit in 2027. Chile's Cochilco pegs 2026 demand at 2.08 million tonnes LCE against 2 million tonnes of supply. SQM — Chile's biggest producer — just told investors on its Q3 call that it expects demand to grow 30% this year.
Prices are falling into a deficit. I've been on this side of the trade before — most recently watching uranium equities bleed for months while the spot market tightened underneath everyone's feet. The screen says one thing. The warehouse says another. Eventually the warehouse wins.
Just a few years ago, they laughed at him…
And said that what Elon was planning to do was impossible.
In 2018, the New York Post even called Elon “a total fraud”...
And it proves the media was 100% wrong about Elon.
|
Part II
The Diagram
|
Strip the story. Numbers only.
Mysteel's weekly survey puts China's lithium carbonate stock at 93,200 tonnes as of November 18. Down 1,090 tonnes from the week before. Combined carbonate and hydroxide inventories are off 10% year-to-date. Monthly consumption in Q4 is running approximately 180,000 tonnes. At this drawdown pace, inventory days of cover are approaching 15. That's two weeks of buffer in a market consuming material at six thousand tonnes a day.
The supply side is contracting, not expanding.
CATL's Jianxiawo mine — China's largest lithium operation by capacity — has been largely idle since August 2025, aside from a brief informal restart in July 2026. Regulators revoked its Environmental Impact Assessment this year and sent it back to care and maintenance. Benchmark cut its 2026 output forecast by 49%, from 62,500 to 32,000 tonnes LCE. That's 30,500 tonnes erased from a market already running short.
Greenbushes — the world's biggest hard-rock lithium mine — slashed full-year spodumene guidance to 1,375–1,425 kilotonnes, down from 1,500–1,650kt. Ore grade is declining structurally. IGO shares dropped 14% on the announcement. This is not a one-quarter miss.
Albemarle shut its Kemerton lithium hydroxide plant in Western Australia entirely. Originally designed for 100,000 tonnes per annum at full build-out, it was running one train at 25,000 before they pulled the plug. Around 275 workers went home. Pilbara Minerals, MinRes, and Arcadium all mothballed or cut capacity through 2025 when the oversupply narrative was loudest.
Zimbabwe banned all raw lithium concentrate exports on February 25. Enforcement was immediate — cargoes already in transit were halted at the border. Fastmarkets had Zimbabwe penciled in for 124,000 tonnes LCE this year, roughly 7% of global supply. That metal is now stuck behind a resource-nationalism wall, and though Prospect Lithium Zimbabwe's sulphate plant began exporting in April, domestic processing capacity remains nowhere near enough to absorb it.
The supply base shrank during the glut. The glut ended. The supply base didn't grow back.
|
Part III
The Weak Link
|
Everyone models lithium demand as an EV story. It was, mostly, until about eighteen months ago. The variable nobody's sizing correctly is stationary energy storage.
Lithium demand from battery storage systems jumped 71% in 2025. UBS projects another 55% surge this year. This year, BESS will account for roughly 30% of global lithium demand — up from 15% two years ago. Grid-scale batteries, renewables integration, data center backup power. They all consume lithium carbonate and hydroxide, and none of them show up in the EV sales charts that analysts keep refreshing on their second monitor.
Here's where it gets uncomfortable. The 2022–2025 oversupply narrative — the one that tanked prices from 500,000 yuan to below 70,000 — triggered a wave of mine shutdowns and capacity deferrals across three continents. Producers did exactly what economics tells them to do: they shut in. Albemarle mothballed Kemerton. Pilbara pulled back. MinRes went idle. Arcadium paused. The market cheered because lower supply was going to "rebalance" things.
What nobody modeled was both pillars of demand — EVs and storage — accelerating at the same time the mines went dark. SQM just told investors demand is growing 30% this year and will reach roughly 2.5 million metric tonnes by 2027. The mines that were supposed to supply that metal are sitting cold. MinRes's Bald Hill restart shipped first product only in July 2026, with a six-month ramp to full capacity still underway. That's not an instant supply response. That's a slow climb back.
I watched this exact setup in cobalt in 2022. Oversupply narrative, production cuts, then a demand shock that caught everyone flat. The lithium version is bigger because the inventory cushion is thinner. Fifteen days of cover in a 180,000-tonne-a-month market is not a buffer. It's a countdown.
Buy Stocks… A Crash Is Coming
Think stocks are overpriced? Expect this bull market to end in a spectacular crash?
Famed economist Jim Rickards says you're 100% right.
But don't sell a single share until you've seen his 2027 prediction.
Jim believes that Congress is set to act by November 30th…
Triggering a historic buying panic that will push stocks higher than most people can imagine.
|
Part IV
The Chain Reaction
|
Walk through it mechanically.
Inventory days fall below 15. That's approximately where China sits today. Downstream cathode and cell makers who've been buying hand-to-mouth realize the cushion is gone. They hit the spot market simultaneously — the same converters who'd been sitting on the sidelines for eighteen months, waiting for lower prices that aren't coming. Spot spikes. Producers who shut in capacity try to restart, but Jianxiawo is gated behind an environmental review that may not conclude until late 2026 at the earliest, possibly stretching into 2027. Kemerton is cold indefinitely. Greenbushes is fighting declining ore grades with no quick fix.
The supply response doesn't arrive in time for the restocking wave. It never does. MinRes's Bald Hill restart — the one the market keeps pointing to as proof that supply elasticity still works — shipped first product only in July 2026, with a six-month ramp to nameplate still in progress. That's one mine slowly climbing back. It doesn't fill a 171,000-tonne hole.
Where does capital flow? Not into the diversified battery-metal ETFs diluted with cobalt plays, graphite companies, and miners hedged so far forward they'll sit out a spot rally entirely. The asymmetry — if it exists — sits with mid-tier producers running operational mines, unhedged output, and direct exposure to spot carbonate pricing. Companies whose revenue actually moves when the number on the Guangzhou Futures Exchange moves. SQM expects to produce 290,000 tonnes in 2026 from its Novandino operation and is still talking about demand exceeding 3 million metric tonnes by 2030. Pilbara has restarted and returned to full rate — but most shuttered capacity industry-wide remains cold.
The market priced lithium for a world where supply was abundant and demand was decelerating. Supply is not abundant — it's contracting. Demand is not decelerating — it's growing 30% this year. Those aren't vibes. That's Benchmark's deficit estimate, Cochilco's demand forecast, Mysteel's weekly inventory number, and SQM's earnings call, all pointing at the same hole. Fifteen days of cover. The clock doesn't care what the screen says.
Sources: Benchmark Mineral Intelligence, Mysteel, SMM, Cochilco, SQM, Fastmarkets, UBS, Reuters, IGO, CATL/Mining.com, Albemarle, MinRes, Pilbara Minerals, Zimbabwe Ministry of Mines


