Is Oracle’s stock spike the strongest indicator yet that we’re in an AI bubble?

Even if you only pay a little attention to the business world, it’s been hard to miss Oracle’s phenomenal week. The company’s shares jumped nearly 40% on Wednesday, and CEO Larry Ellison briefly overtook Elon Musk as the world’s richest person (he’s now essentially tied). Oracle also signed a staggering $300 billion deal with OpenAI for computing power over the next five years.

Even in the fast-moving world of AI, that’s a lot in a short time. Investors cheered, but some Wall Street bulls are wondering if this is further inflating the AI bubble—and making a potential collapse all the more alarming.

Oracle’s rocket ride

Oracle shares spiked after the company reported fiscal first-quarter earnings Tuesday afternoon. It missed analyst expectations on earnings per share and revenue, but investors looked past that shortfall, thanks to booming cloud demand. Oracle said it has $455 billion in remaining performance obligations (that is, contracted revenue that has not yet been recognized), up 359% from a year ago. Wall Street was expecting that number to be closer to $180 billion.

By Wednesday, The Wall Street Journal reported the OpenAI deal, one of the largest cloud contracts ever signed. (It will require about as much power as 4 million homes.) The one-two punch gave Ellison the largest single-day jump in net worth ever recorded by Bloomberg, which tracks billionaire holdings.

Oracle’s rocket ride is not unlike the one Nvidia found itself on beginning in 2023 (a ride that is still going strong). But the high level of the stock is giving some observers—and some insiders—pause.

“Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes,” said Open AI founder Sam Altman last month. “Is AI the most important thing to happen in a very long time? My opinion is also yes.”

Oracle shares were trading at nearly 50 times the company’s 12-month forward earnings on Wednesday. (The stock was down 5% in midday trading Thursday, to $311, but remains well above pre-earnings levels.) That’s the highest multiple since the dot-com crash (when the forward projected earnings hit 120).

Still, Oracle has some solid footing. Its expected cloud revenue and the OpenAI deal do give those stock escalations some footing. Investors are betting on informed company forecasts rather than blind hope.

Nvidia, meanwhile, has seen share prices jump 390% in the past two years and double since April. It has a market cap of $4.3 trillion, but is heavily reliant on two unknown customers who made up 39% of its Q2 revenue—a red flag for the bulls.

An AI bubble?

Talk of an AI bubble didn’t start this week, of course. Some analysts have been sounding alarms for months.

In July, Torsten Slok, chief economist at Apollo Global Management, warned that AI stocks are even more overvalued than dot-com stocks were in 1999, putting the market at serious risk.

“The difference between the IT bubble in the 1990s and the AI bubble today is that the top 10 companies in the S&P 500 today are more overvalued than they were in the 1990s,” he wrote in a note to investors.

Put another way: Investors are betting so heavily on AI that the stock price of companies like Nvidia, Microsoft, Apple, and others has become detached from their earnings.

Slok isn’t alone. Alibaba Group chair Joe Tsai has warned that U.S. AI stocks are in a bubble; longtime tech executive (and former C3.ai CEO) Tom Siebel did so, too.

Part of what has them so nervous is the fact that the top five companies in the S&P 500 now hold 30% of the index’s wealth. That’s a higher share than during the dot-com era and well above the “Nifty Fifty” that dominated markets in the 1970s.

That concentration doesn’t clearly identify a bubble, but it does underline how dependent the market is on just a few companies that are all part of the same industry. If something goes wrong in AI, the ripple effects could be disastrous for the market.