OPINION – What Future for SMRs if the AI Boom is a Bubble?

What Future for SMRs if the AI Boom is a Bubble?

Readers who follow the relationship between demand for artificial intelligence (AI) data centers and the prospects of SMR and microcomputer developers who want to build their power plants to run them, are probably aware that the market frenzy over publicly-traded AI stocks has produced over-the-top stock valuations that are a growing concern.

Why this is important: The financial health of the AI industry has a direct bearing on the number of data centers expected to be built which in turn potentially drives demand for SMRs and microreactors to power them. It is a daisy chain of connections that appears to contain a fair degree of risk.

A news analysis published in the New York Times on 10/14/25 by former members of the Council of Economic Advisers, Jared Bernstein and Ryan Cummings, warns that the current surge in Artificial Intelligence (AI) investment exhibits classic signs of a speculative bubble, similar to the dot-com era of 2000.

If the two economists are correct, then there is a lurking risk that if the AI bubble pops, that paper commitments, e.g., nonbinding MOUs, for data centers, will be tossed in the bit bucket. SMR and mircoreactor developers will wind up without the customers they expected to drive the growth of their businesses and provide returns for their stockholders.

So what is a “bubble?”

What Are the Stages of a Bubble?

Irrational exuberance is a concept that entered the American idiom based on the writings of the economists Alan Greenspan and Robert J. Shiller. They define it as “escalation of asset price driven by human psychology” and is not related to the actual value of the business for the stocks in question. Irrational exuberance has become synonymous with the creation of inflated asset prices associated with bubbles, which ultimately pop and can lead to market panic.  As a phenomenon it is one of the reasons behind over valuation of stock prices in a specific industry or sector of that industry.

According to the online investor newsletter MarketPsych Newsletter for May 15, 2025, in the following image of a speculative bubble (the Nasdaq 1996-2002 bubble), we see fundamentals improving (grey line), psychology turning bullish (moving from optimism to greed), and prices rising (blue line). Media coverage and the introduction of amateurs into the market was evident in the launch of dedicated financial news TV channels and magazines, and enthusiastic TV commercials for stock brokerages.

What the Economists Said About an AI Bubble

The Core Argument: The authors of the NYT analysis of the AI industrye argue that investment levels in the AI sector, notably over $200 billion in venture capital this year and a tripling of data-center investment since 2022, have become persistently detached from plausible near-term profits. For instance, OpenAI is seeking $1 trillion for data centers, yet its expected 2025 revenue is only around $13 billion. This detachment from fundamentals is the hallmark of a bubble.

Key Evidence and Market Impact: There is a valuation disconnect. The price-to-earnings (P/E) ratios for AI-heavy stocks, like chipmaker Nvidia (trading at approximately 55 times earnings), are at levels last seen during the dot-com peak.

Concentrated Growth: Just seven firms heavily invested in AI are responsible for driving more than half of the S&P 500’s nearly two-thirds swell since late 2022.

Economic Exposure: The share of the economy devoted to A.I. investment is estimated to be nearly one-third greater than the share dedicated to internet-related investments during the dot-com bubble.

The Fallout: Should AI adoption prove slower or less profitable than expected, a subsequent market correction is likely. This would trigger a wealth effect collapse, where stock losses curb consumer spending, potentially slowing real GDP growth (currently estimated to be boosted by 0.4 percentage points by the AI wealth effect).

Silver Lining: The authors conclude that while a crash could lead to a recession, the systemic damage would likely be less severe than the 2008 housing crisis. This is because the debt financing AI data centers appears less distributed and systemically embedded in the global financial system than mortgage-backed securities were.

What Happens to SMR Developers if the AI / Data Center  Bubble Bursts?

The falling dominoes that would follow from an AI bubble bursting are that many but not all of the nonbinding MOUs for nuclear energy reactors to power AI data centers would evaporate taking them the over valuation of some of the stocks of nuclear mirco and small modular reactor startups.

In short, the collapse of the AI bubble would negatively affect firms in the data center industry and, following that, impact the nascent nuclear energy industry especially developers of SMRs and microreactors who have built up investor confidence in their future, based on part, on stacks of nonbinding MOUs for AI data centers not yet built.

Given that AI platforms are a significant element of growth of the US economy, the question is whether there is “excessive exuberance” driving the AI industry is germane.

A key risk for investors is that none of the SMR and micro reactor developer have yet to deploy a commercial reactor for a customer. Licensing challenges with the Nuclear Regulatory Commission remain a significant hurdle. These companies currently have no revenue, meaning the investment case rests squarely on future contracts, partnerships, and government support. And with a rally this steep, any delays or regulatory snags could pressure the lofty valuations of their stocks.

What Experts in Nuclear Energy Finance are Saying

A knowledgeable senior level investor emailed Neutron Bytes with the following observation. The person’s name and title are withheld due to their position.

“The pseudo or exaggerated order books or projections will soon be called out. We will also see investors dry up, moving away from the wishful reactor developer class, as decisions get made and real customers choose from two or three, maybe four or so more credible designs.”

This isn’t the only warning. According to a report by the Bloomberg wire service on 10/07/25 “some on Wall Street are starting to get a little dubious about how frothy the market for nuclear stocks has gotten.”

Rama Variankaval, global head of corporate advisory for JPMorgan Chase & Co., acknowledged in an interview with Bloomberg that there is “unrealistic optimism” around the sector.

Bloomberg reports that Chris Gadomski, head of nuclear research at BloombergNEF, “ominously compared the enthusiasm around SMRs and data centers to the internet boom and bust of the early 2000s” in an interview with the Financial Times.

Gadomski also suggested that Wall Street might be underestimating the enormous costs it will take to get SMRs up and running.

“There’s a lot of cheerleading happening, but the amount of capital that you need to cross the finish line is huge,” he said.

If investor confidence in the future of the AI data center customer segment is negatively impacted, the daisy chain of effects of investors’ response could be one of being less   willing to support financing the run from prototypes to products for SMRs and microreactors.

Options for Actions by SMR and Micro Reactor Developers

SMR and micro reactor developers can do several things to prevent being caught up the consequences of a burst AI bubble.

First, they can get off the hype conveyor belt of popping off press releases every time the firm has an “agreement in principle” with an AI driven data center developer. Nobody knows, in aggregate, how many of these will be built or when.

Second, they can recognize that delays in getting the first-of-a-kind units built are inevitable pushing some expected gains in market share into the 2030s. Meanwhile, between now and then there are competitive factors that also come into play.

Gas Will Rule Data Center Power Deals through the End of this Decade

Even for the data centers that are being built now, and at least for the next four-to-five years, some of them will be powered by natural gas plants to generate electricity. For instance, FERMI America, which has plans to built four Westinghouse AP1000 1,1150 MW PWRs, plus some SMRs at the same site, says for now it will provide power for data centers by burning natural gas. Being located in Amarillo, TX, there is plenty of it. FERMI says it won’t start down the nuclear path until electricity demand makes the business case for it.

The outlook for nuclear energy to power data centers is now facing competition from oil companies who in addition to providing the gas are now moving up the value added chain by offering to build power plants specifically for AI data centers to burn the gas these firms produce.

Exxon and Chevron, two of the planet’s biggest producers of natural gas, announced earlier this year they intend to get in to the business of building private wire gas-fired power plants for data centers. This market strategy could lock up data center customers with gas for years as even rapid depreciation won’t make the case to swap one out for multiple nuclear plants of similar size, e.g. less than 300 MW/unit.

Another example already underway of using gas to provide power to date centers is that Meta, aka Facebook, announced a new $10 billion data center in Richland Parish, Louisiana. The data center will be powered by three combined-cycle combustion turbines with a capacity of 2.26GW built and operated by Entergy Louisiana. That’s the equivalent of seven 300 MW SMRs.

Beyond that some data centers will take a page out of Microsoft’s playbook with Constellation and ink a binding PPA with an existing nuclear power plant in Pennsylvania. A developing story is that NextEra, the owner of the Duane Arnold nuclear plant, is on a path to restart the reactor to meet anticipated electricity demand from multiple Google data centers being built in Iowa.

What SMR and Micro Developers Can Do to Secure Their Future?

Taken together, competition from natural gas, and existing nuclear power plants, raises questions about market strategies for SMR and microreactor developers. In the meantime, a plausible course of action for SMR and microreactor developers is to focus on getting NRC licenses and booking binding PPAs for power or contracts to build, and if not with AI data centers, then with any customer that wants 24X7 365 reliable power that doesn’t emit CO2.

With timelines in the 2030s, for many of them, it is a long road to revenue. The sooner deals with real contracts for performance are done with paying customers that are not AI data centers, the more secure their financial futures are likely to be.

Examples include DOD’s plans for microreactors on military bases, with the spillover effect to civilian sectors, is one of those pathways. Another is floating SMRs for export to developing nations. The cargo shipping industry is intensely interested in establishing a business case for SMRs for their firm’s future vessels. There is a future for SMRs and mircoreactors as long as the industry’s developers hang their hats on more than one hook.

Caveat: Nothing in this article is intended to be financial advice for investors. The comments here are solely the opinion of Neutron Bytes.

# # #