'It's more likely they'll lose money': One strategist on why he's bearish on the AI trade

Business Insider

4 hours ago

Nearly three years after the launch of ChatGPT, the AI hype is still in full force as the biggest tech companies continue to shell out billions to build out their AI infrastructure.
Nvidia, meanwhile, hit a fresh all-time high this week, and some analysts see the gains piling up to push the premier AI chipmaker to a $6 trillion valuation.
But will investors see the billions in AI capex pay off? Peter Berezin, chief global strategist at BCA Research, thinks not.
"It's more likely they'll lose money," Berezin told Business Insider, referring to Big Tech companies spending big on the technology.
It's been hard to bet against the biggest trade in the stock market and its promise to turbocharge economic productivity. Investors quickly shook off the DeepSeek disruption earlier this year, and Big Tech companies plan to spend over $300 billion on AI investment. The rally in tech powered the Nasdaq 100 to record highs this week.
Still, Berezin thinks the market is missing the bigger picture. AI technology is tremendously expensive, and he sees monetization opportunities as slim. While AI certainly could boost productivity, that's no guarantee that higher profits will follow.
A race to the bottom
Venture capitalist Peter Thiel famously said "competition is for losers," and Berezin agrees with this sentiment.
"You don't make money in a competitive market. You make money as an investor in a monopolistic market. AI, so far, is very, very competitive, and that's a problem for investors in that area," Berezin said.
"If everybody can do it, then how do you charge money for it? OpenAI was bragging a few months ago about how they're actually losing more money than expected on their latest model, but to them that was a good thing because people were using it so much. The presumption here, of course, is that if they're using it, they're eventually going to pay for it," Berezin added.
"But why would they pay for it if Anthropic, if X, if DeepSeek, and many other companies now are offering similar products like this?"
It's not just the AI startups that are engaged in an AI arms race. Berezin points to the Magnificent Seven's capex spend as another example of fierce competition threatening profits.
He sees Big Tech's billion-dollar checks as a defensive investment to maintain market share, not an offensive strategy to expand it. At best, Berezin thinks these companies are investing in AI to maintain their current competitive standing, but there's a real possibility that some could lose their market dominance.
Undeniably, many Nvidia investors have already made a lot of money, but that's more of a testament to the chipmaker's near-monopoly on GPUs and less so a reflection of sustainable, broad-based AI monetization.
Unless there's more consolidation in other parts of the AI ecosystem, Berezin isn't optimistic that the AI trade will pay off.
What about the productivity gains?
Another key part of the AI bull thesis hinges on the possibility that AI could unlock massive efficiencies across every sector of the economy. While Berezin doesn't deny this potential, this outcome might not be a boon for shareholders, either.
Big-box retailers adopted IT technologies in the 1990s that led to productivity gains, but since these technologies were widely adopted, the competitive advantage was neutralized, resulting in lower prices for the consumer without a corresponding increase in profit margins, according to Berezin.
"You can increase productivity from using AI, but if everybody else is also increasing productivity from using AI and competing in the same market, then what you end up with is lower prices, but not higher margins," Berezin said.
In his opinion, the AI industry is likely on track to become something that looks like the airline industry: capital intensive, critical for the economy, but very low margin.
"I think that's the risk with AI, that the benefits of AI filter down more to the users of AI rather than the producers, which historically has been the case for most technological innovations," Berezin said.