Markets are finally putting numbers where the rhetoric is. In April 2026, X-energy listed in the US and its shares jumped, lifting the market cap to about 11.5 billion dollars on day one, according to MIT Technology Review reporting on May 28, 2026. On May 13, 2026, Fervo Energy debuted on Nasdaq at 27 dollars, then rose roughly 33 percent intraday, pushing its valuation past 10 billion dollars, TechCrunch reported the same day. Two IPO pops in two months is not a thesis by itself. Paired with rising power demand from AI data centers and a favorable policy lane for firm clean power, it looks like a scarcity premium for 24-7 supply is getting priced in.

What the market is buying

Investors are not paying for vibes. They are paying for capacity that shows up at 3 a.m. and 3 p.m., which data centers value because uptime is revenue. Fervo’s pitch is enhanced geothermal, a firm resource that runs around the clock. TechCrunch on May 13, 2026 framed the pop as demand driven by AI data centers seeking dependable power. MIT Technology Review on May 28, 2026 added that the 2026 cohort features firms racing to meet higher electricity demand, with Fervo’s market cap cited at about 12.4 billion dollars at that time.

X-energy’s story is similar on the demand side, different on the technology. Next generation nuclear, high temperature gas cooled reactors, and a smaller unit size, are all part of the plan. In April 2026 the company went public in the US and shares surged on debut, per MIT Technology Review on May 28, 2026. The same outlet noted that the company is still years from commercial demonstration, which is a reminder that equity enthusiasm is not project execution.

A hot IPO is not a turbine spinning. Markets are paying for firm capacity options, not delivered megawatt-hours.

The AI load and the policy lane

The underlying mechanism is straightforward. Hyperscale AI facilities are power hungry, and they prize reliability. MIT Technology Review on May 28, 2026 tied this year’s listings to the rise in demand, partly from data centers, and noted strategic links with tech buyers. Google has backed Fervo and is a customer. Amazon is both an investor in and a client of X-energy. This is not philanthropy. It is power procurement meeting corporate growth plans.

Policy does not hurt either. The May 28, 2026 MIT Technology Review piece reported that while federal support has cooled for wind in this period, geothermal and especially nuclear have retained favorable treatment through tax credits and grants. That matters, because capital costs for firm resources are front loaded. Lower after tax costs and clearer regulatory paths improve project bankability, which lowers the discount rate investors apply to future cash flows. Translate that into plain English and you get higher equity valuations today.

Scarcity shows up first in price

If you want to see scarcity, look at what clears first in the market. Firm clean power is clearing at a premium. Fervo’s May 13, 2026 IPO was upsized multiple times, TechCrunch reported, and the stock pop moved its value past 10 billion dollars. MIT Technology Review’s May 28, 2026 snapshot put it at about 12.4 billion dollars later in the month. X-energy’s April 2026 debut also saw shares rise. These moves do not forecast flawless delivery, but they do reveal what investors are prioritizing.

Geothermal and nuclear are not identical on costs, timelines, or risk. Any comparison is project specific. On July 13, 2026, Oilprice.com, bylined ZeroHedge, published single-source drilling and cost updates for Fervo’s Cape Station, including a reported 70 percent reduction in drilling time on a new well and estimates that Phase II could track to about 5,500 dollars per kilowatt with longer term targets lower. Those figures, and comparisons to nuclear cost overruns, require independent confirmation. They do, however, signal what the market hopes to buy, repeatable learning curves that turn capex into scale.

Execution risk is not priced out

Investors love a learning curve, but steel still has to meet rock. MIT Technology Review on May 28, 2026 noted that X-energy remains years from a commercial project and that timelines matter for the broader sector. Fervo’s pipeline is ambitious, and TechCrunch on May 13, 2026 reported that proceeds would help scale projects like Cape Station in Utah. Ambition meets geology and supply chains in the field, not on an S-1 cover page.

Solar plus storage is part of the same demand story, but treat the specifics with care. MIT Technology Review on May 28, 2026 reported that Solv Energy went public in February 2026 and cited a 6 billion dollar figure. Those details remain a single outlet’s account in that piece, and corroboration has been flagged as an open question. The through line still holds. Markets are paying for capacity that can be deployed quickly and, if possible, run when the wind is not blowing and the sun is not shining.

What to watch next

First, watch offtake. Data center buyers can solidify the scarcity premium through long term contracts. Second, watch permitting and incentives, because policy clarity continues to lower capital costs for firm clean projects. Third, treat any single-source performance claim as provisional until independently verified. Oilprice.com’s July 13, 2026 drilling and financing notes for Fervo and Quaise fall in that bucket.

The upshot is simple. In 2026, the market is revising what counts as scarce. Intermittent megawatt-hours are abundant in spreadsheets. Firm clean capacity at hyperscale is not. April and May put a price on that gap. Execution, as ever, will decide whether today’s premiums age like assets or like anecdotes.