If you're waiting for Michael Burry to tell you when the AI bubble will burst, don't hold your breath.
In Burry's new post on his Substack, the famed "Big Short" investor said that there "is no way to time or predict" the bubble pop, especially when the bubble may still have room to grow.
"Shorts are almost always short-term trades. Usually less than a year, maybe a couple years at most," wrote Burry. "Not 5 years, not 10 years."
"I believe today the stock market is in a phase that could become a blow off top of extreme magnitude on the upside, while at any time, maybe even today or tomorrow, making a generational top," Burry added.
In the lengthy blog post, Burry answered reader questions on his earlier posts. He argued that there is "supply-side gluttony," meaning massive data-center build-outs, GPU orders, and multibillion-dollar commitments without real end-user demand, which investors are mistaking for supply-chain activity. He attributed much of the hype to being driven by Nvidia CEO Jensen Huang's marketing.
"Even when it finally tops, it will not be for any specific reason," Burry said later in the post. "Even if the reason is an AI buildout bubble popping, that will likely not be apparent until a year or two later."
"Mostly, it is prudent neither to short stocks nor to buy puts on stocks. Stocks that are obviously overvalued tend to have the most upward momentum yet have puts that are very expensive," Burry added.
Nvidia did not immediately respond to a request for comment about Burry's latest digs.
In November, Burry launched a paywalled Substack called Cassandra Unchained. After posting charts on X that show circulatory investment deals between Nvidia and other tech giants, his first post took aim at Nvidia. He called the chipmaking giant "a Cisco" in the AI bubble debate, referring to the internet-networking giant whose stock plunged by over 75% during the dot-com crash.
Nvidia released a note to Wall Street analysts in late November, pushing back against some of Burry's claims.
Earlier in November, during Nvidia's Q3 earnings call, the CEO addressed concerns about the AI bubble.
"From our vantage point, we see something very different," Huang told investors, "We excel at every phase of AI, from pre-training and post-training to inference."
