Nvidia's market value touched $5.78 trillion on Monday, an intraday record for the chipmaker. The milestone comes as one veteran tech analyst argues that consensus forecasts on Wall Street are still undershooting the company's actual demand.
Dan Ives, a partner and senior managing director at merchant bank Yorkville Ives & Co., told CNBC's Closing Bell that Street numbers for Nvidia are likely 25% to 30% too low. He pointed to supply chain checks in Asia showing chip demand running at 13 to 14 times available supply.
The scale of Nvidia's presence in the market is hard to overstate. It leads a group of 10 stocks that together account for roughly 39% of the S&P 500. That concentration has drawn comparisons to earlier eras of market dominance, and not all of them are comforting.
Record valuation meets skeptical money
Not everyone is buying the momentum. Michael Burry holds Nvidia put options through September 2027 and has said the AI bubble could burst early. His position is a direct bet that the current trajectory won't hold.
DBS Group in Singapore offered a different framing, telling Bloomberg that Nvidia trades at 17 times forward earnings. For context, Cisco traded at 100 times forward earnings before the dot-com crash. That gap suggests either a more reasonably priced leader or a warning that the comparison itself is imperfect.
The bear case has history on its side. Lehman Brothers' $639 billion bankruptcy in September 2008 stood as the largest one-day market loss at the time. Enron's 2001 collapse took down its auditor, Arthur Andersen. And by October 2002, the Nasdaq Composite had fallen to 1,139 from a March 2000 peak of 5,048. Each of those episodes started with a concentration of belief in a single narrative.
The financing layer underneath the trade
Nvidia isn't just selling chips. In August, the company announced preliminary agreements with six financial firms for about $500 billion in customer financing, according to CNBC. That arrangement effectively helps customers fund their purchases of Nvidia hardware — a structure that keeps demand flowing but also intertwines the company's fortunes with the balance sheets of its buyers.
The broader semiconductor supply chain is already feeling the pull. South Korea's chip-led exports lifted the country's nominal output 26.4% year on year in the second quarter. That's a concrete number from a major economy, not a projection.
What the bulls are actually saying
Ives isn't arguing that Nvidia is cheap. He's arguing that the demand signals are stronger than the models show. His 13-to-14-times-supply figure, if accurate, implies a shortage that won't resolve quickly. That kind of gap typically takes years of capacity expansion to close, and chip fabrication plants don't get built in a quarter.
But the concentration risk cuts both ways. When 10 stocks make up 39% of the S&P 500, a rotation out of those names doesn't just hit one sector — it moves the entire index. Passive investors who own an S&P 500 fund are effectively making a concentrated bet on Nvidia and its peers, whether they intend to or not.
The puts held by Burry expire in September 2027. That's a long runway, and options can expire worthless even when the underlying thesis eventually proves right. Timing is the hard part, and it always has been.
The next marker
Nvidia's forward earnings multiple of 17 stands in contrast to the dot-com-era extremes. The question now is whether the $500 billion in customer financing agreements announced in August will translate into reported revenue that justifies the record valuation, or whether the structure itself becomes a vulnerability if demand cools.
For now, the stock sits at a level no company has reached before. The next quarterly report will give the first real test of whether Ives' 25-to-30% upside call holds — or whether Burry's puts start looking prescient.
