The Uranium Contracting Cliff
Reactors get announced at press conferences. Pounds get secured in contracts, and the contracts are not being signed.
There are two uranium markets, and almost all public commentary watches the wrong one. The spot market, which sat at 86.40 dollars a pound at the end of July, is thin, opportunistic and largely irrelevant to how reactors are actually fuelled. The market that matters is the long-term contract market, where utilities buy the fuel they will load five and ten years from now. That price has reached 94 dollars, its highest level in eighteen years. The spread between the two is the story.
A utility does not buy uranium the way a fund does. It buys coverage, in multi-year strips, because a reactor that runs out of fuel is not a trading loss but a national incident. The right question is therefore never what uranium costs today. It is what percentage of a utility's requirement for 2031 is already under contract, and at what price it will have to buy the rest.
Fourteen years of borrowing from the future
The industry needs roughly 185 million pounds a year to run the existing reactor fleet. Simply holding coverage flat, replacing the pounds burned each year, requires contracting on the order of 150 million pounds annually. As of the second week of August, about 37 million pounds had been contracted globally in 2026. The market is on course for a fourteenth consecutive year of below-replacement contracting.
Fourteen years is long enough that it no longer registers as unusual. It has become the baseline expectation of an entire generation of fuel buyers, most of whom entered the industry after Fukushima and have spent their careers in a market where uranium was abundant, secondary supply filled every gap and patience was always rewarded. That was a genuinely correct read of the market for about a decade. The inventories it drew down are now gone.
| Year | United States | European Union |
|---|---|---|
| 2026 | 98 percent | Fully covered |
| 2030 | 60 percent | 100 percent |
| 2031 | — | 81 percent |
| 2032 | — | 78 percent |
| 2033 | 9 percent | — |
| 2034 | — | 36 percent |
Read the American column again. Ninety-eight percent of next year's requirement is secured. Nine percent of 2033 is. That is not a forecast or a bearish analyst's model. It is the sum of contracts that already exist, and the only way the number improves is if utilities sign new ones, in volume, at whatever price clears. Europe reaches the same cliff about two years later.
Supply is not coming to the rescue
The conventional answer is that high prices cure high prices. In most commodities that is true. In uranium it has repeatedly not been, and 2026 is showing why.
Kazatomprom, the largest producer in the world, cut its 2026 nominal production plan by 3,000 tonnes of uranium, from 32,777 to 29,697, and declined to return to full capacity. The reason it gave was not geological. It was that the company did not consider market conditions sufficient to justify producing more. A dominant, state-controlled, low-cost producer choosing discipline over volume is not a temporary condition to be waited out. It is the structure of the industry.
Beneath that sit constraints that no price fixes quickly. Kazakh production depends on sulfuric acid supply that has been tight for years. Canadian operations have lost output to flooding at Key Lake and McArthur River. Namibian production has been affected by weather. Conversion and enrichment, the stages between the mine and the fuel assembly, are their own bottlenecks and were substantially reordered when Russian supply was removed from Western chains. None of these are solved by a higher spot price in a given quarter.
Meanwhile demand has been revised in one direction only. The United States has announced a programme of up to 80 billion dollars to build new reactors, the first such commitment in decades, and uranium has been added to the US critical minerals list. Data centre power procurement has brought a new class of buyer, one with no institutional memory of cheap uranium and considerably less price sensitivity than a regulated utility.
What I think happens
The contracting cycle restarts, because it must, and it restarts into a supply base that has spent a decade learning that volume discipline pays better than growth. Utilities will return to the market not gradually but in a cluster, when coverage gaps become uncomfortable enough to force procurement committees to act, and they will be competing with each other for term commitments from a producer group with every incentive to hold the line on price.
The distinction worth holding onto is the one between announcements and pounds. A reactor programme is a political commitment that can be restated indefinitely. A fuel contract is a legal obligation to deliver specific material on a specific date, and the gap between the two is currently measured in the tens of millions of pounds a year. Somebody has to close it. The price at which it closes is the only real question, and the contract market has been answering it quietly for eighteen months.
Prices and contracting volumes cited are as reported at 31 July and 10 August 2026. Disclosure: the author is Chairman of Saddleback Uranium, a uranium exploration venture with claims in Wyoming, and is therefore not a disinterested observer of this market. Nothing here is investment advice.