Investment
October 14, 2026 · 7 min read
The Two Camps of Nuclear Financing, and Where They Intersect
October 2026
Every large infrastructure buildout tackles the same questions, whether it’s highways, transmission lines, LNG terminals or new power plants: who carries which risk, for how long, and what has to be true before ordinary investors will carry it instead?
At the Atlantic Council’s Nuclear Energy Policy Summit on September 22, moderator Stephen Comello of the EFI Foundation put that question to four experts who finance energy. He asked each one to name the single risk that most keeps new reactors out of conventional capital markets.
The answers split down the middle. The split says a lot about where public money should go next.
Camp one: “Show me the customer”
Two panelists pointed to revenue risk, meaning the uncertainty over whether a reactor that takes years to build will earn enough over its 60-to-80-year life to repay the people who paid for it.
Robert Rudich of SGE, a developer building out small modular reactor fleets in Europe, put it in terms of policy durability. A developer can deliver a reactor on time and on budget, he argued, and still lose if a future government isn’t on board. His answer is an enhanced contract for difference that locks in a fair long-term price for the power and also protects against changes in policy and regulation.
David Livingston of Galvanize, which runs a growth equity fund of more than $1 billion focused on energy and industrial innovation, framed the same risk from the investor’s side. Venture and growth capital can back nuclear most responsibly, he said, where demand is already under contract. He pointed to the offtake agreements that hyperscalers and industrial buyers have signed with reactor developers. Some investors will take on more risk than that, but for most fiduciaries, a contracted pipeline is what makes the case.
Camp two: “Show me you can build it”
The other two panelists named a different risk: completion.
Daniel Hanna of Barclays described the core problem as unbounded completion risk. Uncertain costs and uncertain schedules, with no ceiling on exposure, are something private capital struggles to price. Long construction periods make it worse, because when one thing slips, the delay ripples through the supply chain and into policy risk.
Melinda Chow Alankar of the Anthropocene Institute went a step further. She tied completion risk directly to first-of-a-kind risk, the cost of learning while building. She also made the most pointed argument of the session. Revenue risk is real, she said, but it can be contracted around. Contracts for difference, power purchase agreements, tax credits and clean energy attributes are all predictable and within someone’s control. What no contract can fully anticipate is everything that goes wrong the first time a new design is built.
That’s why, in her view, government is often the party that has to absorb first-of-a-kind risk. How long that support should last is a more open question. Comello framed the discussion on the premise that nuclear eventually has to stand without government support. Livingston pushed back, arguing that nuclear is a public good that warrants continued public investment, a point he returned to at the close of the session.
Where the two camps agree
The camps are less far apart than the opening round suggested. As the discussion went on, three points of agreement emerged.
Think in fleets, not single reactors. Rudich argued that no one makes money building a single reactor, and that building one unit, demobilizing the supply chain and starting over has failed for 30 years. His company’s approach in Poland and the UK frames each program as a multi-unit package from day one. Each reactor gets its own go-ahead, but the terms for the whole fleet are set upfront, so suppliers know whether they’re staffing up for one unit or fourteen. Alankar pushed the idea into finance. The standard model of one special-purpose vehicle per project, funded with non-recourse debt, may need to give way to a single entity holding every project in a fleet. That spreads risk across units in a way investors can price.
Unpack the risks, then match the capital. Hanna argued that there is little truly new in nuclear finance, since every instrument has been tried. What’s new is how carefully the risks get separated and sequenced, so the right capital comes in at the right stage. He pointed to the UK’s regulated asset base (RAB) model for the Sizewell C plant, which brings in revenue during construction and defines what government backstops, as an example. Livingston put it simply: the goal isn’t to eliminate every risk upfront. It’s to bound each one and retire it at the right point in the project.
Separate technology bets from project bets. Rudich drew a line that public markets often blur. Investors are excited about nuclear technology companies, but the industry also needs investment in concrete, steel and reactors actually being built, and there is far less of that.
The two cautions worth hearing
Alankar raised two warnings that deserve more attention than they usually get.
The first is about ratepayers. Models like RAB lower the total cost of a plant by collecting money during construction. But Georgia’s Vogtle expansion showed what happens when customers pay for years before a plant produces power and costs keep climbing: a lasting loss of public trust. Serious community engagement, she argued, has to come first, not as a communications afterthought.
The second is about markets. A growing number of nuclear companies are going public before they earn revenue, which effectively makes the general public a new class of nuclear investor. That can raise a lot of capital at a good price if the timing is right. But IPOs take a year or more to prepare, and markets turn fast. Alankar also noted that nuclear’s public-market story is now closely tied to AI and data centers, so pushback against one becomes pushback against the other. Within days of the panel, Holtec formally withdrew its planned IPO. It had already postponed the offering the night before it was set to price, citing a sharp shift in market sentiment against the AI data-center economy.
What would tell us nuclear has become a normal asset class
Comello closed by asking each panelist what milestone would show that nuclear has moved from exceptional financing to conventional infrastructure. Their answers make a useful scorecard:
- Hanna:
- when each new operating plant is financed more cheaply than the one before it.
- Livingston:
- when several developers raise fleet-level capital before their first unit is finished, a sign that markets trust the delivery process, not just the reactor design.
- Rudich:
- interim structures that share risk and mobilize private capital will matter a great deal over the next decade, even before textbook project finance arrives.
- Alankar:
- when fleets themselves become the financing unit, held and funded as a portfolio rather than one project at a time.
None of those milestones depends on a technical breakthrough. They depend on building the same thing more than once and showing the second one went better than the first.
Livingston offered the closing image. Nobody building the interstate highway system asked how fast government could get out of highways once the first two were finished. The real question was what judicious public investment would let private capital and the wider economy build on top of the system. For clean, reliable power, the panel’s answer was that public money should carry the risk markets can’t yet price, and step back as each unit makes that risk smaller.
This piece draws on the panel “Novel Approaches to Nuclear Project Finance” at the Atlantic Council’s Nuclear Energy Policy Summit 2026, featuring Melinda Chow Alankar (Anthropocene Institute), Daniel Hanna (Barclays), David Livingston (Galvanize) and Robert Rudich (SGE), moderated by Stephen Comello (EFI Foundation / Nuclear Scaling Initiative). Watch the full conversation: https://www.youtube.com/watch?v=eN8rs6vbNUc
About the Author
Written by the Anthropocene Institute newsroom. The Institute funds and publishes research on nuclear energy, solid-state fusion, grid electrification and ocean carbon, and convenes the entrepreneurs, investors and policymakers building clean firm power.