Tearing apart the DRC cobalt pricing disconnect

Cobalt.

The financial press has a simple story: EV demand is soft, Chinese processors are sitting on mountains of inventory, and cobalt is a busted commodity trade. The charts confirm it. Prices are down more than 60% from the 2022 peak. The short sellers are comfortable. The consensus has moved on.

Here is what they are not looking at. Battery cathode manufacturers — the people who actually buy refined cobalt and bolt it into chemistry — have been quietly signing multi-year offtake agreements at prices above current spot for the last four months. Not one. Several. Johnson Matthey, Umicore, a handful of Chinese cell makers. Nobody is announcing this. They're not required to. But the physical contracts are there if you know where to check the filings.

That spread — $12.40 spot, $16–18 in signed offtake — is the whole story. When cathode buyers are willing to pay a 30% to 45% premium above spot to lock in forward supply, they are telling you they do not believe the spot price reflects what's actually available. They are paying not for today's cobalt. They are paying for cobalt they are afraid won't be there in 2028.

I traded the 2018 cobalt squeeze wrong. I was watching the exchange price. I should have been watching the Katanga loading reports. Expensive lesson. Different mechanism this time, same category of mistake if you stay glued to the screen.

Story off. Numbers on.

The DRC produces approximately 70% of global mined cobalt. Not roughly. Not about. Seventy percent of the world's supply comes out of a single country whose eastern provinces have been in active armed conflict for the better part of two decades. That concentration is not news. What is news is what's happening to the structure of who mines it.

Artisanal mining is not a romantic sideshow. It represents roughly a quarter of DRC cobalt output. These are not companies with balance sheets and revolving credit facilities. These are individual miners and small cooperatives operating on pure cash-flow arithmetic. When spot price drops below their all-in cost — somewhere between $13 and $15 per pound depending on the operation — they stop. They don't hedge. They don't wait for recovery. They leave.

The large-scale industrial mines — Glencore's Mutanda, CMOC's Tenke Fungurume — are running on Chinese offtake contracts locked in at prices above current spot. They are not cutting production. They are not going anywhere. But they also cannot absorb the volume that artisanal operations produce, and they are not trying to. The cobalt that was coming from 400,000 artisanal miners in Lualaba province is not being replaced. It is simply gone.

The cathode buyers aren't dumb. They've run this calculation. The offtake premiums they're paying today are an insurance policy against a supply gap that the spot market is not pricing yet. The question is not whether they're right. The question is when the spot market figures out what they already know.

The consensus narrative says cobalt is oversupplied because Chinese refiners are sitting on 60,000 tonnes of refined inventory. That number is real. The narrative around it is not.

That inventory is not freely available. The majority of it is pledged against warehouse financing arrangements — essentially collateral in short-term lending structures used by Chinese metal trading houses. The metal is nominally "in stock." In practice, it cannot be sold without triggering a loan unwind. This is not speculation. This is the same structure that blew up in the nickel market in 2022 and in aluminum warehousing in 2013. The market learned nothing either time, which is sort of the market's signature move.

If freely available Chinese cobalt inventory is closer to 25,000–30,000 tonnes rather than the reported 60,000 — which is what a careful reading of the warehouse financing data suggests — then the market's entire bearish thesis is built on a number that includes metal that cannot move. The overhang is a fiction. Half of it is pledged. The bears are short a market they haven't fully mapped.

Meanwhile, the ASM miners who supplied the last structural buffer are leaving the market permanently. They don't come back when prices recover. It takes six to eighteen months for artisanal operations to restart — retooling, equipment, safety arrangements, cooperative agreements. By the time spot price rises enough to call them back, the cathode buyers will already be in shortage. This is the gap nobody is talking about. The market is staring at an inventory number. The inventory number is lying.

The sequence here is slow and then sudden. That's the cobalt market's rhythm — it has done this twice in the last fifteen years, and both times the move was violent because the positioning was one-sided when it broke.

The probable sequence: one more leg down on the flush of pledged inventory hitting the market. The bears read it as confirmation. Positioning gets even more one-sided. Then a cathode buyer — one who didn't sign enough forward offtake — goes to the spot market for a meaningful volume and finds the freely available metal is not there. First wave: spot spikes. Shorts cover. Second wave: the remaining cathode buyers who were waiting for the dip realize the dip is the shortage. Third wave: DRC government, watching the price, imposes an export levy or quota review. They have done this before, in 2018. It is in their regulatory playbook.

Where does capital go? Not into diversified battery materials ETFs — those are diluted with lithium and manganese exposure and will lag a cobalt-specific move badly. The edge, if there is one, is in the handful of cobalt-weighted pure-plays with production outside the ASM segment: companies with operational mines, unencumbered inventory, and revenue directly tied to spot. They're priced for the consensus narrative. They should be priced for what the cathode buyers already know.

The cobalt market is not a story about EVs being slow. It is a story about a supply base that is structurally exiting at current prices, an inventory overhang that is half fiction, and a group of cathode buyers quietly paying a 35% premium above spot because they've already done this math. The paper market will catch up. It always does. The only question is whether it happens at $16 or $22, and whether you're positioned before or after the loading docks in Lualaba go quiet enough to hear the difference.