What a Quota Can Do and What It Cannot

The Democratic Republic of the Congo produces roughly three quarters of the world's mined cobalt. In February last year its minerals regulator suspended all cobalt exports, industrial and artisanal alike, on the stated grounds that there was too much stock on the market. The suspension was extended twice and then replaced, from the sixteenth of October last year, by a quota system running to at least 2027.

The quotas are not generous. This year's total allocation is 96,600 tonnes, of which 87,000 is distributed to producers and 9,600 held at the regulator's discretion, which works out to about 7,250 tonnes a month and roughly half of what the country exported in 2024.

It worked, and then it stopped working. Cobalt bottomed at around 24,400 dollars a tonne in January last year, a nine year low. It reached 53,005 by the end of December, with cobalt hydroxide up 328 percent across the year. It trades near 39,100 dollars today, up about sixty percent from the pre ban trough and down about a quarter from the peak, and hydroxide fell 31 percent in the third quarter alone.

The Historical Echo

The cleanest precedent for a dominant producer restricting exports of a critical mineral, and for how such a regime ends, is China and the rare earths.

Beijing cut rare earth export quotas from roughly 50,000 tonnes in 2009 to about 30,000 in 2010, a reduction of some forty percent, and trimmed them again the following year. The price response was spectacular. Neodymium rose more than tenfold, to around 460 dollars a kilogram, between May 2010 and August 2011.

The restriction did not end because anyone was persuaded. The United States, Japan and the European Union filed at the World Trade Organisation in 2012, the ruling went against China in 2014, and the quotas were formally abolished with effect from the first of January 2015.

By then the prices had already collapsed, and they had collapsed for reasons that had nothing to do with the legal case. Engineers thrifted, using less of the metal per magnet. Designers substituted where they could. Mountain Pass and Lynas restarted outside China. The quota was still in force while the price went down, because the quota controlled tonnage and could do nothing about demand.

Where Patient Capital Is Positioning

The same three forces are now visible in cobalt, and the middle one is the most important.

Alternative supply is scaling. Indonesian mine output went from 35,000 tonnes in 2024 to 44,000 last year. Substitution is well past the theoretical stage. Lithium iron phosphate batteries, which contain no cobalt at all, passed half of global electric vehicle batteries in 2024 and exceeded 55 percent last year, and the International Energy Agency's assessment is blunt: the effect of cobalt price movements on overall battery costs is far more limited today than in the past. And the restricted metal eventually moves. Shipments of Congolese hydroxide into China exceeded ten thousand tonnes in June against about two thousand in May, which is what broke the price in the third quarter.

Where cobalt differs from the rare earth case, and differs in the regulator's favour, is that cobalt is a by product of copper. Restraint costs the producers relatively little. Glencore's first half cobalt output fell 46 percent, to 10,200 tonnes from 18,900, and the company simply prioritised copper and left the cobalt in solution to be processed later as the rules evolve. CMOC produced a record 117,549 tonnes last year and has kept its guidance while raising its copper target. Neither of them is in any distress. A producer that can wait can hold a quota regime up for years, and there is no trade case in view to force the question.

The observation for long horizon capital is about what an export restriction actually is. It is a tax on time. It raises the price of a thing that buyers must have now, and in doing so it funds every engineer who was already looking for a way not to need it. The Congo can control how much cobalt leaves the country. It cannot control how much cobalt the world decides a battery requires, and that second number has been falling for a decade and is still falling. Scarcity created by decision invites substitution. Scarcity created by geology does not, which is the difference between a policy and a moat, and it is why the assets worth owning across generations are the ones nobody can vote to expand.

The Capital Memo