Fears About an AI Bubble Are Fading. That’s an Opportunity.

Dot-com era parallels are worth studying as AI powers markets higher. (Michael Nagle/Bloomberg)

Key Points

• Media coverage and internet searches related to an “AI bubble” peaked in August and have since decreased.
• Current market patterns, including high stock prices and equity valuations, show similarities to the dot-com era.
• Some analysts suggest that missing the final rally in the current market could be a significant error for investors.

“History doesn’t always repeat, but often rhymes,” is a useful paraphrasing of a Mark Twain quote that applies quite well to the current debate over the impact of artificial intelligence on global financial markets .

Proponents of the new technology claim it will be the biggest boost to economic activity since the internet. Skeptics warn that the trillions being invested—and the eye-watering stock prices its leaders command—will ultimately fall far short of the promised returns.

In the meantime, patterns are developing in both U.S. stocks and the broader economy that draw parallels to the dot-com era tech bubble, even as recent data suggests the public at large has grown more comfortable with AI’s seemingly inevitable advance.

Deutsche Bank analysts Adrian Cox and Stefan Abrudan, ironically using AI techniques, found that media coverage and internet searches linked to fear of an “AI bubble” peaked in August, and have slowed notably since.

The pair suggest that could indicate a more realistic assessment of AI’s capabilities, the time and investment required to deploy it at scale, and the challenge in implementing it across a host of sectors and businesses.

“That involves integrating it into well-governed enterprise systems that employees actually use,” Cox and Abrudan wrote in a report published Tuesday. “Evidence is still emerging of where the dollar and cents of value will come.”

That takes investors back to the debate over stock prices, which are near the highest on record, and equity valuations, which are some of the richest levels since the dot-com era.

Back then, the market provided a host of false alarms on the state of the tech boom, as the Nasdaq surged and fell by 10% or more on seven different occasions in the late 1990s.

Federal Reserve rate cuts, the Tax Relief Act signed by President Bill Clinton, and powerful market momentum ultimately carried the tech-focused benchmark to its March 2000 peak. It then collapsed more than 78% into its early October nadir in 2002.

And with global central banks cutting rates, the One Big Beautiful Bill Act providing a fiscal tailwind to an already-solid economy, and retail investors committed to the current rally, there are eerie similarities to the market today.

“It absolutely feels dangerous to buy tech stocks right now,” said Joe Tigay, portfolio manager at Catalyst Hedged Equity Fund. “Their valuations are high, and they could correct at any moment.”

“But my core thesis is that we are likely not at the bubble’s end yet…we are heading into the liquidity phase that preceded the final, astronomical surge of 1999—2000,” he added.

Vendor financing, a key component of the dot-com era rally that saw the largest tech companies lending money to highly valued but largely unprofitable customers, is also re-emerging in today’s market through a series of deals involving ChatGPT creator OpenAI.

“On the positive side, they are clearly seeing demand for their services inflect sharply and believe strongly they are on track to be a much bigger business in time,” said John Belton, portfolio manager and Gabelli Funds.

“On the negative side, one could argue that investing so aggressively ahead of comparable revenue generation adds systemic risk to the AI ecosystem as a whole,” he added.

And that, of course, raises the issue of both identifying the existence of an AI bubble and predicting when it will burst.

Cox and Abrudan at Deutsche Bank suggest the best method is actually not trying to do so at all.

In fact, they cite data showing that a $10,000 investment in stocks 1996 would be worth around $170,000 today, “but less than half as much if you’d missed the 10 best days and a quarter as much if you’d missed the 20 best days.”

“Timing the markets is notoriously hard,” the pair wrote. “Evidence suggests that staying invested over a long time horizon seems to be the best way to compensate investors for their risk.”

Tigay at Catalyst Funds puts it even more bluntly: “Missing the final rally would be a catastrophic mistake.”

Write to Martin Baccardax at martin.baccardax@barrons.com