Tearing apart the DRC cobalt quota execution gap

The battery metals narrative this year is about EV demand growth, clean energy transition, critical minerals geopolitics. The financial press has been running variations of that story for three years. It's not wrong. It's also missing the actual price driver in cobalt right now by about nine thousand kilometres.

The Democratic Republic of Congo produces roughly 74% of the world's cobalt. In February 2025, its government imposed a full export ban to address a nine-year price low caused by chronic oversupply. By October, the ban was replaced with an annual quota system — 96,600 tonnes for 2026, roughly half of 2024 export volumes. Cobalt metal went from $21,502 per tonne at the start of 2025 to $56,000–62,000 per tonne by early 2026. Hydroxide CIF China hit $25.90 per pound in late April — up 350% from pre-ban levels. The price move is extraordinary. The press noticed the price. Nobody is explaining what's actually happening to the physical shipments.

The market knows the quota exists. What it has not fully priced is the gap between what the DRC government says it will permit and what the physical infrastructure can actually execute. That gap is the machine. And right now, the machine is broken in ways the spot price hasn't fully absorbed.

Bridge collapses on key trucking routes from mine to port. Mandatory government testing protocols — which require pre-payment of a 10% mining royalty within 48 hours and a liberatory receipt before customs clearance — that no one in the DRC's administrative apparatus had been built to process at scale. Discrepancies between private and government cobalt content testing. All of it compounding. The quota was issued. The trucks couldn't leave.

Numbers. No narrative.

The supply chain geometry is unusually simple. Cobalt hydroxide flows from DRC mines to the port at Lobito via truck and rail. Ships carry it to Chinese ports, primarily. Chinese refineries process it into metal, sulfate, and tetroxide for battery manufacturers. That chain has exactly one choke point: the DRC government's export administration apparatus. Right now that apparatus is processing approximately one-third of the material it has nominally cleared for export.

Indonesia is the secondary supply pole — cobalt-in-MHP output is forecast to climb 39% to 53,318 tonnes in 2026, up from 38,324 tonnes in 2025, as Chinese-backed HPAL plants expand. But Indonesian cobalt is a byproduct of nickel processing. Its output is capped by nickel ore quotas and — via the Strait of Hormuz disruption — by sulphur feedstock availability. You can't run HPAL without sulphur. The secondary supply source has its own chokepoint, and it's the same war that's driving everything else right now.

CMOC — the company that produced 117,549 tonnes of cobalt in 2025 and whose overproduction triggered the export ban in the first place — has been allocated 31,200 tonnes of export quota for 2026. Twenty-seven percent of what it produced. The company that flooded the market is now the biggest loser from the policy its flooding provoked. I've been watching commodity markets for long enough to appreciate the specific poetry of that.

The weak link is not the quota. The quota is the headline. The weak link is the assumption baked into current pricing: that the execution gap is a temporary administrative hiccup that normalises in Q2. That assumption may be wrong, and nobody in the financial layer is stress-testing it properly.

The DRC has twice extended the deadline for Q4 2025 quota utilisation — first into Q1 2026, then adding another 30 days after that — because shipments were running so far below allocation. The government itself is extending deadlines because its own system can't process the exports it authorised. That is not a bureaucratic refinement. That is a system that does not work. Building a new administrative apparatus — mandatory pre-payment systems, government testing labs, ARECOMS enforcement capacity — in a country that ranks 176th on the Human Development Index does not happen on a quarterly timeline.

The demand story is its own paradox. Battery makers have spent seven years trying to engineer cobalt out of their cathode chemistries. LFP adoption among Chinese EV manufacturers has been aggressive. Tesla pushed toward NMC low-cobalt formulations. The industry consensus circa 2023 was that cobalt demand growth would disappoint. That consensus ran directly into two reality checks: first, lithium cobalt oxide is experiencing a quiet renaissance in consumer electronics — smartphones, laptops, and wearables depend on LCO's volumetric energy density in ways no alternative chemistry currently matches. Second, defense and aerospace are adding cobalt demand that is entirely inelastic to EV battery trends.

I've seen this pattern in other byproduct metals. When the primary-use narrative goes bearish, the market misses the secondary applications that won't bend. Palladium was supposed to be dead once EV adoption killed catalytic converters. It took years before anyone properly modelled the industrial and electronics demand that was still there regardless. Cobalt is doing something similar in a different corner of the periodic table.

There is also a second-order demand risk that Fastmarkets flagged explicitly: if prices stay elevated long enough, EV manufacturers accelerate their shift to cobalt-free chemistries faster than projected. The DRC's quota policy is designed to raise prices and capture more value. If it raises prices too far for too long, it accelerates the structural demand destruction it was trying to prevent. The DRC president said the country had been a "victim of predatory strategies" for too long. That is true. It is also possible to overcorrect.

Two sequences are competing in real time. The market is oscillating between them and hasn't resolved which one wins.

The counter-sequence: the DRC administratively normalises by mid-year, Q2 quota utilisation reaches 70–80%, Chinese inventory rebuilds, prices ease back toward $40,000 to $45,000 per tonne. That is the base case embedded in most sell-side models right now. It requires the DRC government to have solved, in six months, administrative infrastructure problems that compounded across eight months of ban and quota implementation. Possible. Not certain.

The IRA FEOC rules are the structural overlay that makes this more than a short-term price story. Under the Inflation Reduction Act, EVs are ineligible for the full $7,500 tax credit if their batteries contain critical minerals from a Foreign Entity of Concern — a category that explicitly includes Chinese-refining operations. The threshold rises from 60% in 2025 to 80% by 2027. Every automaker selling in the US market has a hard deadline to source cobalt through non-Chinese refining channels. The Lobito Corridor — the rail and port project linking the DRC copper belt to Angola's Atlantic coast — is the structural solution. It is years from operating at scale. The deadline is not years away.

The cobalt market entered 2024 as a textbook oversupplied basket case. It entered 2026 in structural deficit, with the world's dominant producer administering a quota system its own infrastructure cannot execute, a secondary supply source constrained by a geopolitical sulphur crisis, and Western automakers facing regulatory deadlines to find supply that doesn't yet exist at the necessary scale. The interesting thing isn't the price. The interesting thing is how many of those problems were entirely predictable and entirely ignored until the trucks stopped moving in Katanga.