Investors probe how long Nvidia’s AI windfall can last amid cash flow and capex worries
Nvidia is still delivering blockbuster numbers, but markets are acting as if its AI story has hit an inflection point – a warning sign for investors relying on the AI trade to drive long-term returns.
Shares of Nvidia fell more than 5 percent on Thursday in their worst session since April, even though the company reported “stellar” quarterly results, according to CNBC.
CNN reported that profits nearly doubled in the fourth quarter and sales hit an all-time high, yet the stock still closed 5.5 percent lower.
CNN described how headlines flipped from “Nvidia eases bubble fears” to “Nvidia fails to impress,” noting that after three years of consistent beats, “very good” now looks ordinary for the “most valuable public company on the planet.”
For many investors, the core issue is no longer whether Nvidia can beat near-term estimates, but whether today’s AI spending boom can last.
Richard Clode, portfolio manager at Janus Henderson, told CNBC that “the debate has shifted away from near-term results and toward the sustainability of AI capex spending, amid concerns around its quantum, monetization and potential cashflow degradation.”
He pointed out that Nvidia guided to US$78bn in revenues, “well ahead of even the most bullish buyside expectations” and marking a fourth straight quarter of accelerating growth.
Nvidia’s data centre unit powered this surge, generating 91 percent of sales, with data centre revenue at US$62.3bn versus expectations for US$60.69bn.
Despite that, the market reaction was cautious.
Tom Graff, chief investment officer at Facet, told CNBC he expects a “bumpy ride” for at least the next couple of quarters.
He said investors already assumed a strong quarter after customers like Microsoft and Amazon projected higher data centre spend, but “what we didn’t get were details about the future guidance.”
Graff added that if “players like OpenAI might be slowing spending, that would show up in actual revenue one to two quarters from now,” so “the lack of specificity about the revenue outlook is generating some concerns.”
Uncertainty around a proposed US$100bn Nvidia–OpenAI arrangement is another pressure point.
According to CNBC, Nvidia’s 10-K says it is finalizing an investment and partnership agreement with OpenAI, but also makes clear there is no guarantee a deal will be signed or completed.
Gil Luria of D.A. Davidson told CNBC he remains optimistic that Nvidia will make a large OpenAI investment, despite what he called “confusing” language in the filing.
Some investors also question the broader tone from management.
After an analyst asked CEO Jensen Huang how confident he was that customers could continue “shoveling hundreds of billions of dollars” into Nvidia chips, Huang replied that he was “confident in their cash flow growing,” according to CNN.
Mike O’Rourke, chief market strategist at JonesTrading, wrote that “investors know there is trouble in paradise,” pointing out that Amazon, Meta, Microsoft and Google recently reported free cash flow “either collapsing or flattening.”
He argued that “if management is not going to be candid about information that is well known, investors become fearful of what they don’t know.”
Michael Burry, the “Big Short” investor, highlighted Nvidia’s rapidly growing commitments.
He noted that Nvidia’s “purchase obligations” jumped to US$95bn from US$16bn a year earlier.
He said TSMC has demanded more cash for complex custom chips, forcing Nvidia to place non-cancellable orders “well before demand is known,” and described the shift as “structural” rather than temporary.
Burry warned that the figure could be “catastrophic” if AI enthusiasm fades.
CNN said Nvidia did not immediately respond to a request for comment.
Competitive dynamics add another layer of risk.
As the industry pivots from training-heavy to inference-driven workloads, some investors expect more competition from alternative chipmakers.
Hardika Singh, economic strategist at Fundstrat, wrote that “where [Nvidia] did miss was easing investors’ concerns about its narrowing moat in the evolving world of compute,” according to CNBC.
She told CNBC the report shows “just how emotions, not logic, is driving the stock market right now,” and argued that more players in the field does not mean Nvidia cannot “emerge a winner.”
Singh added that the architecture of Nvidia’s Vera Rubin chip is designed to be very strong for inference.
Even so, formal analyst views remain broadly positive. Of 66 analysts covering Nvidia, 61 rate it a buy or strong buy, with the average price target implying about 37 percent upside from here, according to LSEG data.
Adam Phillips of EP Wealth Advisors told CNBC “the odds were stacked against them” given how far the stock has run, and said “it’s becoming harder and harder to impress the Street,” with many investors now asking, “Is all this investment going to pay off?”
