Westinghouse’s IPO Has a $30 Billion Shadow Price — but the Services Base Must Justify It

Westinghouse confidentially submitted a draft U.S. IPO registration statement, with the filing announced on July 31, 2026. Share count, price range, proceeds, listing venue and timetable were not disclosed, so the filing starts price discovery without revealing a price.

The cleanest way to value Westinghouse is to separate recurring operating-plant work from project-driven new builds. According to Cameco, Westinghouse serves roughly 63% of about 417 operating reactors, fuels about 40% of the fleet and performs outage services across nearly two-thirds of it. That footprint anchors value in recurring plant work before assigning anything to the AP1000 pipeline.

A 2025 binding term sheet links Westinghouse to a U.S. government program intended to facilitate at least $80 billion of reactor investment. If the government participation interest vests by January 2029 and Westinghouse is expected to list at $30 billion or more, the government can require an IPO. At listing, that interest can convert into a five-year warrant tied to 20% of public value above $17.5 billion, so a higher valuation increases potential government participation.

The arithmetic above applies the term sheet’s stated formula; it does not capture exercise price, ownership structure, tax effects, offering primary-versus-secondary mix or later definitive amendments. Those details can materially change economic dilution. The eventual prospectus therefore needs to reconcile headline valuation with fully diluted value.

AI demand reaches Westinghouse through a multi-step chain: data-center growth raises the value of round-the-clock clean power; long PPAs improve generator revenue visibility; stronger utility economics support project equity; only then do reactor frameworks become component orders and OEM revenue. The first visible proof is in operating assets: Microsoft signed a 20-year agreement with Constellation Energy for approximately 835 MW from the planned Crane Clean Energy Center restart, expected online in 2028 subject to approvals.

The Department of Energy conditionally committed up to $17.5 billion for long-lead items supporting as many as ten 1.1-GW AP1000 units across five two-reactor projects. The financing targets components with the longest manufacturing lead times and fixed-price procurement. That structure moves bottleneck spending earlier than a conventional project timeline.

The non-obvious risk is that the financing program can improve supplier visibility while increasing Westinghouse’s own capital needs. If all five projects proceed under the disclosed structure, Westinghouse could be required to commit an aggregate $2.5 billion before accessing loan funds. The IPO may therefore finance equity obligations as well as growth, making use of proceeds a central prospectus question.

BWX Technologies is the clearest listed component linkage. Its 2024 memorandum with Westinghouse covers potential production of AP1000 and AP300 steam generators, reactor vessels, pressure vessels and heat exchangers. Because these are precisely the kinds of complex, long-lead items targeted by the DOE program, funded procurement could convert the memorandum into executable orders.

Centrus Energy is a second upstream read-through, but the linkage is structural rather than contractual: reactors require secure enrichment capacity, and no direct AP1000 award to Centrus Energy was established. Its $3.9 billion backlog extends to 2040, yet well over half of expected LEU deliveries through 2027 depend on TENEX. Nuclear growth therefore raises the value of domestic enrichment while geopolitical dependence keeps execution risk high.

Constellation Energy and Microsoft demonstrate why the downstream market matters to Westinghouse. Constellation Energy owns generation that can respond sooner than a greenfield reactor, while Microsoft supplies the long-duration demand contract. This pairing monetizes nuclear scarcity before new builds arrive.

This timing gap is central to the IPO. A buyer paying today for ten U.S. units is underwriting projects whose equity, permits, customers and final loan documents are incomplete. By contrast, service revenue from operating reactors and the Constellation Energy-Microsoft PPA are tied to identifiable assets and contracts. Public investors should therefore discount greenfield volume more heavily than installed-base cash flow.

Current nuclear-linked equities carry very different multiples because their cash-flow quality differs. Cameco trades at approximately 56.7 times TTM enterprise value to EBITDA, BWX Technologies at 29.3 times, Centrus Energy at 30.1 times and Constellation Energy at 14.1 times. These are not direct IPO comparables, but they show that investors already capitalize nuclear optionality aggressively.

The base-case thesis is deliberately narrower than “AI creates a nuclear renaissance.” Westinghouse already owns a large service and fuel franchise; AI-linked power demand and federal financing can add a second growth engine if utilities commit capital. The IPO succeeds as an investment only if disclosure prices recurring cash flow separately from contingent reactors.