The Uranium Squeeze Washington Saw Coming Two Decades Ago

Uranium prices have surged past levels not seen in years, driven by a nuclear power revival, the electricity appetite of AI data centers, and a domestic enrichment bottleneck that mining alone cannot fix. Advanced reactors and small modular designs increasingly need a more highly enriched fuel that the United States has only limited capacity to produce, even as demand for it climbs. The headline framing treats this as a sudden shortage. The more accurate framing is that America gave up a large piece of its own enrichment capability decades ago, in a deal that made perfect sense at the time and left a predictable gap once the arrangement ended.

The Historical Echo

In 1993, the United States and a newly post Soviet Russia struck an unusual bargain known as Megatons to Megawatts. Russia agreed to convert 500 metric tons of weapons grade uranium, enough material for roughly 20,000 nuclear warheads, into low enriched fuel suitable for civilian power plants, and American utilities agreed to buy it. Over the following two decades, that converted Russian material supplied up to 10 percent of all electricity generated in the United States, quietly reducing the world's stockpile of weapons grade material while filling American reactors at commercial prices. It was, by most measures, a remarkable success, a rare case of former adversaries turning down bomb fuel into something as ordinary as a light switch.

But the program had a side effect that took years to become visible. Buying converted Russian uranium was cheaper than building and maintaining a full domestic enrichment industry, so American capacity in that specific capability atrophied over the two decades the deal ran. When Megatons to Megawatts concluded in December 2013 and geopolitical relations with Russia deteriorated in the years that followed, the United States found itself without the enrichment capacity it had allowed to wind down, precisely as a new generation of reactors began demanding more of it, not less. The lesson is not that the original deal was a mistake. It solved a real problem and did real good. The lesson is that outsourcing a strategic physical capability, even to a cooperative partner, can quietly leave a gap that only becomes obvious once demand returns and the capability is no longer there to meet it.

Where Patient Capital Is Positioning

Today's uranium market is living through the arrival of that gap. Enrichment, not mined uranium in the ground, is the actual bottleneck constraining new nuclear buildout, and rebuilding that capacity domestically takes years of investment and specialized infrastructure that cannot be conjured overnight. The pattern echoes what has already played out in rare earth minerals and is beginning to play out in copper, a strategic physical capability that was allowed to concentrate elsewhere for reasons that made sense in the moment, only to become a source of real vulnerability once demand shifted.

For long-horizon capital, uranium and the broader nuclear fuel cycle represent a version of the same thesis running through gold, silver, energy, and now these critical industrial metals: physical capacity, once lost, takes far longer to rebuild than it took to let go, and the cost of that rebuilding tends to show up in price long before the new mines or enrichment facilities are actually running. A family thinking in decades does not need to predict exactly how many small modular reactors get built. It only needs to recognize that the physical constraint underlying today's uranium price is a real one, built from a specific and traceable piece of history, not a speculative story invented for a headline.

The Megatons to Megawatts program is also a useful reminder that good policy in one decade can quietly create the conditions for scarcity in another, without anyone involved doing anything wrong. That is a more sobering, and more realistic, way to think about physical supply chains than assuming today's arrangement will simply persist. Domestic uranium royalties and mineral rights, like the energy royalties and mining interests already drawing long-horizon capital elsewhere, sit directly on top of that same recognition, that the physical capacity to produce a strategic resource domestically is worth something in its own right, independent of whatever geopolitical arrangement happens to be supplying it more cheaply this year.

The Capital Memo