Goldman Sachs has pushed back against recent market concerns over an artificial intelligence (AI) bubble, calling them excessive. The bank argues that the AI investment boom is grounded not in mere expectations but in improving corporate earnings and sustained capital spending, making it likely that the related investment cycle will continue for some time.
Ben Snider, senior U.S. equity strategist at Goldman Sachs, said in a recent interview with Business Insider that "the AI investment cycle is still supported by solid fundamentals," adding that he continues to maintain his AI-related investment weighting.
Recently, the market has raised concerns that major cloud companies (hyperscalers) such as Microsoft, Amazon and Meta could scale back their data center and AI infrastructure investments if returns fall short of expectations relative to their AI spending. In that scenario, the entire AI supply chain—including semiconductors, servers, networking equipment and power infrastructure—could be affected.
But Snider does not see that as a high probability. He forecast that capital spending to build out AI infrastructure is still expanding, which will continue to support growth in related industries.
He also assessed concerns over technology stock valuations as an overreach. "It's interesting that investors treat high valuations as a warning sign while at the same time interpreting low valuations as a risk signal," he said. "The fact that this pessimism remains actually means there is still an equity risk premium in the market."
He added, "The real overheating signal is rather when every investor turns optimistic on AI and bubble concerns disappear entirely."
The key basis for Goldman Sachs' dismissal of the bubble argument is rising corporate profits. Snider explained that while the S&P 500 index has climbed more than 20% over the past year, its 12-month forward price-to-earnings ratio (PER) is actually lower than it was a year ago. This means the rise in stock prices is not simply the result of investors paying higher prices, but has been accompanied by improvements in corporate earnings.
He named AI infrastructure, power infrastructure and hyperscalers as investment areas to watch going forward. In particular, he assessed that AI infrastructure companies—including semiconductors, servers and AI networking equipment—remain at relatively low valuations despite steadily growing earnings.
"A significant portion of the semiconductor sector, including memory, has seen limited valuation multiple expansion even though profits have grown substantially," Snider said. "There is still room for further upside."
