Is he right?

From the point of view of technical stock valuations, probably yes. From the point of view of fundamentals, maybe not.

A classic bubble exists when the assets being valued are fundamentally not worth their price (and never will be) or when the underlying value is close to zero. 

So, in the great Dutch “tulip mania” of 1637, it is clear in hindsight that the price of a tulip bulb should never be equal to 10 times an annual salary.

And in the Great Financial Crisis of 2008, it became clear in hindsight that many mortgages had been given to people who simply didn’t have the ability to afford them, and thus those mortgages were worth far less than banks’ balance sheets said they did.

So the question becomes whether AI is a bubble or not right now. From the fundamental point of view, the answer is no. OpenAI isn’t literally worth nothing. It’s not a tulip bulb or a tract home in the middle of nowhere. There is a real business there.

JPMorgan’s Brenda Duverce told clients in a recent note that “OpenAI’s ARR has reached ~$13bn (up 30% from Jun-25), and the company has reported it is on track to reach 700mn weekly active users (up 40% from Mar-25), while surpassing 5mn paying business users (up 66% from Jun-25).” She also noted that a “secondary market transaction that could push the company’s valuation to $500bn, up from the previously cited $300bn post-money figure from Mar-25, which would make OpenAI the most valuable private company in the world.” 

To put that bluntly, a company with a chatbot that often gets things wrong is somehow about to become the largest unicorn earth has ever seen. That does feel frothy.

But OpenAI isn’t worth nothing: $13 billion in revenues is a real thing. Maybe the value of its equity will decline in the short term but the company isn’t teetering the way Lehman Brothers was in 2007.

But how about the technical point of view? 

There is a lot of chatter on Wall Street right now about whether tech stocks are overvalued in a way that looks like a bubble. They have some scary charts!

Here is a real head-scratcher: The contribution to U.S. GDP growth from data center spending is now the same as that from consumer spending, according to Apollo Management. The obvious problem with that: This state of affairs exists because consumers have reduced their spending habits as data spending has increased. Unless data centers suddenly start buying cars or shopping at Home Depot, this isn’t good for the long run.

It’s especially not good because the run-up in value of tech stocks is now overpowering the rest of the S&P 500. John Authers of Bloomberg wrote this morning, “It’s unheard of for 2% of the index’s companies to account for virtually 40% of its value.” 

And here is a chart from Bespoke Investment Group. It shows the performance since 2015 of the Magnificent Seven companies vs. the rest of the market. “Bloomberg’s Mag 7 index vs. its 500 Ex Mag 7 index is pretty unbelievable. You can barely see the ‘Ex Mag 7’s’ 129% gain because of how much the 2,800% gain for the Mag 7 overshadows it,” the company says:

Goldman Sachs’ David Kostin has reportedly said that the Mag 7 stocks grew their earnings per share in Q2 by 26% year on year. So there is real money fueling a real business there.

The partial conclusion must be: This isn’t a bubble of fundamentals. No one thinks AI is made of tulips. But it does look a lot like some stocks are technically overvalued, and it should not surprise anyone if this “bubble” bursts.

Here’s a snapshot of the action prior to the opening bell in New York:

• S&P 500 futures were flat this morning, premarket, after the index closed flat yesterday near its record high. 
• STOXX Europe 600 was up 0.54% in early trading. 
• The U.K.’s FTSE 100 was up 0.31% in early trading.
• Japan’s Nikkei 225 was down 0.38%.
• China’s CSI 300 was down 0.38%. 
• The South Korea KOSPI was down 0.81%. 
• India’s Nifty 50 was up 0.42% before the end of the session.
• Bitcoin fell to $114.9K.