One ARECOMS decision moved cobalt 58 per cent and gave every buyer a new due-diligence file to keep.

By Kanisa Magazine · August 2025

The best-documented price event in the cobalt market this year was written by a regulator, not a trader. On 22 February 2025 the Democratic Republic of the Congo’s market authority suspended all cobalt exports, and the Banque Centrale du Congo’s digest of 27 August shows London cathode cobalt at US$33,466 a tonne, up from US$21,171 on 3 February. That is a rise of 58 per cent, and almost all of it arrived within eight weeks of the decision.

The pattern hiding in plain sight is that in a market where one government controls more than half of the reserves, the text of an administrative decision is a data series. Buyers, lenders and insurers who read the decisions as carefully as they read the price screen see supply changes earlier than those who only watch the screen. The rest of this piece explains why, and what the evidence now says.

A rule with dates on it

The suspension was written as a temporary measure, extended once, and reasoned in terms of stock.

The measure was issued by ARECOMS, the Authority for the Regulation and Control of Strategic Mineral Substances’ Markets, as Decision No. 001/ARECOMS/2025 with Communication No. 2025/003, according to the International Energy Agency’s policy record. It ran from 22 February to 21 June, covered industrial, semi-industrial and artisanal mining alike, and cited excessive market stockpiles as its reason.

On 21 June ARECOMS extended it by three months, to 21 September. Its chairman, Patrick Luabeya, signed the statement, which gave the reason as the continued high level of stock on the market, as reported by Eastleigh Voice. Note what the text does and does not say: it states a date, a scope and a diagnosis. It says nothing about what happens on 22 September.

What the price sheet recorded

The price jumped, then went flat, which means the market is trading the rule and not the metal.

The BCC’s digest reproduces London cathode quotations at specific dates. The series reads US$21,171 on 3 February, US$26,578 on 28 February, US$33,652 on 31 March, US$33,311 on 27 June and US$33,466 on 22 August. After the March step it has barely moved. A flat line at a high level is what a market looks like when it has priced the rule and is waiting for the next one.

The same digest gives a control. London copper, quoted on the same dates, went from US$9,365 a tonne on 3 February to US$9,780 on 22 August, a rise of 4.4 per cent. Cobalt rose 58 per cent over the same span. Whatever lifted cobalt was specific to cobalt, and the specific cause on the record is the export suspension.

The same digest carries the central bank’s own output series, which shows cobalt production of 140,121 tonnes in 2023 and 198,777 tonnes in 2024, a rise of 42 per cent. That is consistent with the glut that ARECOMS cited, and with prices at what Eastleigh Voice describes as a nine-year low before the ban. One detail deserves a reader’s care: the BCC’s table books all of the 2024 tonnage under partners of the Gécamines ventures and other operators, and none under Gécamines itself. Anyone using the series for due diligence should reconcile it with export records before relying on it.

Who the rule helps and who it binds

Producers disagree, traders have declared force majeure, and the pipeline delays the pain.

Producers are split. Eastleigh Voice reports that Glencore supports production quotas, while China’s CMOC Group, the largest cobalt producer, has argued for lifting the ban. The same report puts Congo’s holdings at about six million tonnes of cobalt reserves, over half of an estimated 11 million worldwide, a figure that explains why the rest of the market must absorb the decision.

Downstream, the commercial instruments strained first. S&P Global Commodity Insights headlined at the end of June that trader IXM had declared force majeure on cobalt contracts because of the ban. A suspension is the kind of event that sits in the fine print of supply contracts, and a buyer who never read the force majeure clause found out what it said.

The mechanism has a delay built in. Project Blue has noted that cobalt hydroxide takes roughly 90 days to reach China for refining, and that stocks outside Congo would be very low by 21 September if nothing changed. So the squeeze on refiners lands after the ban lifts, not while it runs, and the price evidence in the digest understates the pressure still to come.

For a Congolese miner the sequence runs the other way. Output continues, unsold metal accumulates at mine sites and in warehouses, and working capital stays tied up while the gate is shut. The digest carries no cobalt export volumes, so the carrying cost of that waiting is invisible in official statistics.

What a buyer or lender does differently

Keep a regulatory calendar beside the price file, and write the next decision into the contract.

Three habits follow from this. First, record every ARECOMS decision, with its number, date, scope and expiry, in the same file as the price history, because it is the earliest evidence of a change in supply. Second, read force majeure and government-action clauses in cobalt supply contracts now, against the scenario of a further extension and the scenario of a quota system, since Glencore’s reported preference suggests quotas are on the table. Third, reconcile the BCC production series, export permits and the counterparty’s own declarations before accepting any tonnage figure at face value.

Lenders and insurers carry the same exposure by another route. A borrower whose revenue depends on cobalt shipments from Congo has a covenant problem on any day an export decision lapses without a successor, and an offtake agreement that names no regulatory event as a trigger leaves the financier guessing. Writing the ARECOMS calendar into the credit file turns that guess into a monitored risk, with a named person responsible for checking the regulator’s channels weekly.

As of 31 August the authorities have given no public answer on what follows 21 September. The firms that will cope best are those who already know what each answer does to their contracts. In a market with one dominant regulator, the decision text is the cheapest piece of evidence available, and it is published first.