Sunday, February 9, 2025
Business & EconomicsTech

The Precarity of the Frothy AI Market

News broke over the weekend that a Chinese-developed, open-source large language model AI product is delivering comparable results to ChatGPT’s flagship o1 model. How could this happen, tech bros wondered everywhere, predicting certain doom for their stock portfolios on Monday morning. What was interesting to me was not that Nvidia cratered as much as 18% in intraday trading— that stock is already grossly overvalued, and I’ve written a bit about the hype– but rather that more stable, mature companies that aren’t manufacturing fancy chips also completely ate shit. I’m thinking about two in particular: Vistra and Constellation Energy.

At least since the peak era of Cold War nuclear enthusiasm, utilities haven’t been considered the most exciting business. The rates of profit are often fixed and regulated under statute. Other elements of the business are heavily regulated under federal law. Utilities have become less un-exciting in the past decade or so, amid a nationwide push for renewable energy in its many distributed generation formats, but also because of novel or novel-enough technologies like heat pumps and electric vehicles. Sure, these aren’t new, but technologies have improved and people are suddenly excited about what they could do for utilities from things that are more obvious (utilities could get paid– on a fixed rate of profit scheme- to install heat pumps for customers) and perhaps some that are less obvious (utility strategy nerds who get excited about the possibility of switching from summer peaking to winter peaking).

So, when AI came along, suddenly everybody was excited about power generation, once again. I wrote about the possibilities– and a few very specific reservations I had- associated with restarting nuclear power plants like Three Mile Island or Palisades in Michigan. But there were other things driving up Constellation’s stock price, like news of a new proposed acquisition. This is not necessarily what’s driving AI valuation in a market that we b-school bros define as “frothy”– where valuations become detached from reality, and where there is a lot of speculative capital being thrown around to try and get at growth opportunities.

The other thing that’s vital here is the standpoint of fundamental analysis: The charts here show that VST and CEG are trading at price-to-earnings multiples of 27-32, which is substantially cheaper than the highly inflated tech company prices like that of Nvidia (47x), Tesla (109x), or, say, Trump Media (-11x). So, while that’s still (in my mind) high, it shows that maybe the sell-off is perhaps less than entirely justified.

What the whole episode tells me, though, principally, is that if your AI ecosystem is so fragile that some preliminary reports about a Chinese open-source model can break your entire stock market, it suggests that maybe your product isn’t so great and resilient to begin with. These companies have spent billions on these products. Some of them work. I use ChatGPT basically every day and I confess to this. But will it change the world at the rate these tech types are claiming? Sure, I guess! Capitalism is, after all, an ongoing saga of Line Go Up. But it speaks to the fragility of these products and the ecosystems in which they’re being developed that this sort of thing can just happen.

This isn’t a stumble, it’s a selloff, at least in these couple of intertwined sectors.

It is perhaps time to re-evaluate the stratospheric valuations of these AI companies and, indeed, perhaps the entire stock market.

Nvidia's stock cratering today seems indicative that the vaunted AI boom is perhaps much more fragile than any of the tech bros would have cared to admit-- before today, at least.

Nat M. Zorach

Nat M. Zorach, AICP, MBA, is a city planner and energy professional based in Detroit, where he writes about infrastructure, sustainability, tech, and more. A native of Lancaster, Pennsylvania, he attended Grinnell College in Iowa, the Kogod School of Business at American University, the POCACITO transatlantic program, the SISE program at the University of Illinois Chicago, and he is also a StartingBloc Social Innovation Fellow. He enjoys long walks through historic, disinvested Rust Belt neighborhoods at sunset. (Nat's views and opinions are his own and do not represent those of his employer).

Leave a Reply