More sobering reports about the small nuclear reactor industry, now from the Financial Times. I posted about the Wall Street Journal’s recent reality check here.
In January, I posted similar assessments that appeared in the conservative financial paper Barrons. (If you don’t know Barrons, it makes the Wall Street Journal look like Rolling Stone.)
I’m not anti-nuclear, I’m pro reality.
We have to accelerate solar, wind and batteries not just for the birds, the bees, and our grandbabies, but also because we need to keep the lights on, and these are the lowest hanging fruit and nearest in the pipeline, full stop.
Short sellers have reaped large profits from betting against small modular nuclear reactor companies, as the collapse of the “hype cycle” that had sent their share prices soaring wipes billions off their market value.
Funds made an estimated $2.1bn shorting three stocks — US-listed NuScale Power, Nano Nuclear and Sam Altman-backed Oklo — over the past year, according to data provider S3 Partners.
The three companies, which are lossmaking and have little or no revenue, surged last year as investors raced to capitalise on growing interest in nuclear energy among AI hyperscalers looking for new sources of power. Regulatory changes and funding announcements from the Trump administration also helped the sector.
But a total of $30.3bn has been wiped off their collective market value since their peak in October last year amid growing concerns over the lack of immediate revenue and the long build-out timelines for the technology.
Only two commercial SMRs are currently operable — in Russia and China, with more than 80 designs in various stages of development.
A key test of investor appetite towards the nuclear sector is expected in the coming weeks, when Holtec International and Westinghouse, two US-based companies with SMR divisions, are expected to list.
Meanwhile, X-energy has shed $5.8bn in market value since the surge that followed its initial public offering in April. Short sellers have earned an estimated $67mn from bets against the company since mid-May, according to S3 Partners.
The company, which is backed by Amazon and Ken Griffin and has yet to receive full regulatory approval to build its helium-cooled reactor, has 9 per cent of its shares out on loan.
“The sentiment has changed this year, people are far more critical,” said Putz. These companies “have almost zero revenue for the foreseeable future and, on top of that, there are very high capex [capital expenditure] requirements”.
The Trump administration vocally backed the nuclear sector last year, pledging to cut red tape and invest tens of billions of dollars to build new reactors and reopen old ones to generate the energy needed to “win” the global AI race.
In June, the Department of Energy announced $17.5bn of loans to help rebuild the US nuclear supply chain. However, timelines for delivery of the unproven reactors remain uncertain.
Analysts at BNP Paribas have expressed concerns over shortages of high-assay low-enriched uranium, a special type of nuclear fuel vital for SMRs.
The earliest some will come online is mid to late 2028, if manufacturers are able to speed up delivery while also satisfying regulators, although the majority will arrive during the 2030s, said Stein at the Breakthrough Institute.
“What we have seen in 2025 seems to me like an industry bubble that is already deflating,” said Siegfried Eggert, chief executive of activist short seller Grizzly Research, who has no short positions against the companies.
“I believe most knowledgeable investors understood for a while that the valuations seemed rather extended given the timeline of this industry,” he added.
The rate of new nuclear reactor construction would need to increase fivefold during the next decade if 2050 targets for the technology are to be met, according to the global industry group, which warns of a “gap” between ambition and action.
Roughly 60 to 65 gigawatts of new capacity would need to start construction every year from the late 2030s onward, up from about 12.6GW in 2025, if the collective international target of about 1,457GW of total capacity by 2050 is to be reached, according to the World Nuclear Association.
The US-Iran war has added to interest in nuclear technology as countries seek alternatives to fossil fuels. Several countries, including Brazil and China, have joined a goal — first announced at COP28 in Dubai in 2023 — to triple nuclear capacity to 1,200GW by 2050. Their combined aims would take total capacity to 1,457 gigawatts.
However, the technology has to contend with the high costs, long lead times and stop-start record that make it difficult to get projects off the ground. Meanwhile, competing technologies such as renewables with storage are rapidly gathering pace.
The roughly 3GW of new nuclear capacity that came online in 2025 compares with about 814GW of new solar and wind and 108GW of batteries, mostly of two-hour output, that were added to grids during the same year — a huge gap even after derating the renewables capacity to account for intermittency — largely because they are quicker and cheaper to build.
The increase in storage capacity is a particular threat to nuclear power because it eats into one of the sector’s biggest selling points — a low-carbon source of power regardless of the weather or time of day.
Cobb argued batteries would not necessarily be a disadvantage for nuclear because they meant it would potentially be able to run at a more constant rate rather than having to dial the fleet down when renewables’ output is high, eroding their revenues.