
Nvidia’s earnings report on Wednesday evening was, on paper, a spectacular success: profits nearly doubled in the fourth quarter, and sales hit an all-time high.
In reality, it was a massive letdown.
Shares of the chipmaker, situated at the epicenter of AI trading, declined in late Wednesday trading and closed down 5.5% on Thursday. One could observe financial media headlines pivoting from the initial tone of “Nvidia lessens bubble fears” to a grimmer “Nvidia fails to impress.”
Part of this stems from Nvidia’s own curse: the company has so consistently outperformed earnings estimates quarter after quarter for the last three years that it has conditioned investors to anticipate blowout results. “Very good” is no longer noteworthy when it comes to the planet’s most valuable public company.
However, another element is a genuine shift in sentiment on Wall Street—globally, regarding AI and specifically about Nvidia.
“Investors know there are cracks in paradise,” penned Mike O’Rourke, Chief Market Strategist at JonesTrading, following Nvidia’s investor call on Wednesday.
There’s much more to any company than just the black-and-white facts and figures of its revenue, and this is what O’Rourke and other AI watchers focused on. During the call, an analyst questioned Nvidia CEO Jensen Huang about how certain he was of his customers’ capacity to continue committing hundreds of billions of dollars to acquiring Nvidia’s chips.
Huang responded that he was “confident in their free cash flow growth.” The issue, as O’Rourke points out, is that this isn’t actually happening. In fact, the leading “hyperscalers”—Amazon, Meta, Microsoft, and Google—reported earnings a month ago where free cash flow either plummeted or flatlined.
“If management isn’t candid about well-known information, investors start to fear what they don’t know,” O’Rourke wrote.
A Potentially “Catastrophic” Figure
Added pressure on Nvidia Thursday came from a blog post by Michael Burry, the “Big Short” investor who has become a vocal critic of the AI enthusiasm. In the post, Burry highlighted a figure from Nvidia’s financials that he suggested could prove “catastrophic” if interest in AI begins to wane.
To put it plainly, Nvidia’s “purchase commitments” (agreements to buy goods at a set future date) soared to $95 billion from $16 billion a year prior. The reason, he notes, is that TSMC—a crucial supplier—demanded more money upfront for all the complex, custom chips it fabricates for Nvidia.
This implies that Nvidia “was forced to place non-cancellable orders long before demand was certain,” and the trend appears “structural” rather than temporary, Berry wrote.
Nvidia did not immediately respond to a request for comment.
More broadly, traders are increasingly debating whether AI is simply overhyped or if it will prove so efficient at replacing human labor that it could upend the economy. In either scenario, the potential for seismic shifts is prompting some to consider shorting the market, Gregory Zuckerman of the Wall Street Journal wrote this week.
“One hedge fund manager is positioning for a crash in Nvidia chip sales. Another is unofficially betting against OpenAI, which is private. Some are shorting Oracle stock,” Zuckerman notes.
In other instances, as the Financial Times reported, some investors are “turning to complex options and hedging strategies” to navigate a market so tense that a single blog post or headline could wipe out tens of billions. One strategist told the FT they likened the hedging to a relentless game of “whack-a-mole market roulette.”
At least one tech titan, Apple, has been insulated from the jitters. The iPhone maker has often (and perhaps unfairly) faced criticism for lagging in AI prowess.
But by skipping the massive spending spree its rivals are undertaking, Apple has become a relatively safe harbor, Bloomberg recently reported. Apple shares rose 7% in the last month while the Nasdaq fell 2.7%.