Inside the race to rebuild America’s fuel supply chain for a ‘second nuclear age’

The article argues that the U.S. nuclear buildout tied to AI demand is being constrained by a weak fuel supply chain, with about 98% of uranium used by U.S. reactors imported and current enrichment capacity meeting only about 7% of a hypothetical quadrupling of nuclear demand. Cameco says roughly 30% of its uranium mining capacity is shut in, while Urenco plans to expand its New Mexico enrichment facility by nearly 50% by 2036 and multiple startups and incumbents are pursuing new U.S. enrichment plants. The policy backdrop includes a 2028 ban on Russian enriched uranium imports, which could tighten supply and lift uranium and power prices if domestic and allied capacity does not ramp fast enough.

Analysis

The market is still underpricing the bottleneck that matters most: not reactor construction, but fuel logistics. That creates a multi-year, not quarterly, earnings bridge for the fuel-cycle names because capacity additions in mining, conversion, and enrichment are constrained by permitting, capex, and long lead times, while demand can re-rate immediately on headline reactor wins. The implication is that the scarcity premium should show up first in midstream fuel-cycle margins, then later in reactor economics if utilities cannot secure supply at fixed prices.

Cameco is the cleanest beneficiary, but the better risk-adjusted exposure may actually be the non-obvious toll collectors in conversion and enrichment rather than the miners. If western supply remains structurally short and Russian material is effectively excluded, contract pricing can reset for years, not months, because buyers will pay up for certainty before they pay up for volume. The second-order winner is any company with licensed capacity, brownfield expansion optionality, and a political narrative around energy security; the loser set includes reactor developers that have announced data-center PPAs without locking fuel, as their project IRRs can deteriorate if fuel procurement shifts from fixed-cost to spot-like economics.

The biggest contrarian risk is policy reversal or waiver creep: if Russian supply is partially normalized, the western supply deficit can disappear faster than new Western capacity can earn its cost of capital. That makes the trade path-dependent on enforcement credibility over the next 12-24 months, while the physical supply response remains a 3-7 year story. Another underappreciated risk is that government funding for enrichment can compress private returns by socializing the buildout just as the equity market starts to price scarcity.

For hyperscalers, the market is focused on the data-center power narrative but not the embedded commodity hedge problem. If they scale nuclear PPAs without upstream fuel equity, they inherit basis risk and timing risk that could leak into their long-duration power costs; that is a quiet margin headwind if AI capex remains as aggressive as expected. MSFT is a relative beneficiary of reliable power access, but the better trade is against the suppliers of that reliability, especially where project delays can gap against already-extended sentiment.

Key Decisions for Investors

• Long CCJ vs. short a basket of unlevered clean-power/utility names over 6-12 months: benefit from fuel-scarcity repricing with limited balance-sheet risk; downside is a policy-led Russian supply normalization or faster-than-expected government intervention.
• Initiate a tactical long LEU on pullbacks, 3-9 month horizon: asymmetry favors licensed enrichment capacity as the market eventually prices a domestic bottleneck; pair with a stop if enrichment policy funding stalls or if Russian waivers get extended.
• Buy MSFT and hedge with a smaller short in a uranium fuel-cycle basket over 12-18 months: Microsoft benefits from power security, but the hedge captures the risk that nuclear PPAs overpay for scarce fuel and raise long-run data-center electricity costs.
• Use options on CCJ or LEU rather than outright equity if entering now: the catalyst path is binary and policy-driven, so upside can be large while a surprise Ukraine/Russia détente could re-rate the entire theme lower quickly.
• Avoid chasing reactor developers until fuel contracting evidence appears: the better entry is after fuel supply agreements are signed, because current enthusiasm likely front-runs the harder part of the value chain.