WisdomTree senior economist Jeremy Siegel has outlined a scenario in which the Nasdaq could fall between 20% and 30% if the enormous investment behind artificial intelligence fails to generate expected demand and financial returns. The warning is a conditional downside scenario rather than a prediction that a technology crash is imminent.
In his October 19 commentary, Siegel argued that the AI buildout has become one of the strongest forces influencing technology stocks, capital demand and long-term interest rates. His analysis of the AI buildout and market rotation said disappointment in AI spending or token demand could trigger a 20%-30% Nasdaq correction, while the broader S&P 500 might decline only 5%-10% as falling long-term yields support less growth-sensitive sectors. Siegel explicitly distinguished that scenario from the 2000 dot-com collapse.
That distinction rests partly on valuations. Siegel estimated the Nasdaq at roughly 22-24 times forward earnings, far below the approximately 100 times earnings reached near the peak of the internet bubble. His concern is therefore not that today’s technology companies lack real businesses, but that the returns required to justify extraordinary AI infrastructure spending could fall short of expectations.
AI Spending Creates a Growing Market Divide
Other analysts have identified related warning signs without attaching Siegel’s 30% figure to them. JPMorgan strategist Jason Hunter highlighted a divergence earlier this year between AI hardware companies and the hyperscalers making enormous infrastructure investments. Semiconductor and memory stocks had surged while several major capital spenders weakened, resembling a pattern that appeared before the dot-com downturn.
JPMorgan subsequently estimated hyperscaler AI capital expenditure at more than $750 billion in 2026 and noted that investors were becoming increasingly selective about companies generating acceptable returns from that spending. JPMorgan’s analysis of the widening AI trade The emerging question is whether rising AI revenue can grow quickly enough to justify continually expanding data-center, chip and infrastructure budgets.
Market Maven editor Gabe Marshank has raised a similar concern. An October 9 discussion centered on whether the roughly $7 trillion AI infrastructure debate could develop characteristics associated with the dot-com bubble. That argument focuses on overbuilding and return on capital rather than predicting a specific Nasdaq drawdown.
The pressure is already visible in financing markets. S&P Global warned this week that major hyperscalers may carry more AI-linked debt exposure than publicly recognized, with profitability from their infrastructure investments becoming increasingly relevant to credit quality.
Nasdaq Risk Does Not Translate Directly Into Bitcoin
A technology correction could spill into crypto through general risk appetite, leverage and financing conditions, but a Nasdaq decline would not mechanically imply an equivalent Bitcoin move. Recent data have shown Bitcoin’s correlation with the Nasdaq fluctuating substantially rather than remaining structurally fixed.
That relationship has recently weakened enough for Bitcoin’s 90-day Nasdaq correlation to approach zero, supporting the argument that Bitcoin’s rally has become less dependent on Federal Reserve and technology-market movements. NYDIG has similarly argued that equities explain only a limited portion of Bitcoin’s longer-term price movements. Temporary correlation during market stress should therefore be separated from a permanent economic relationship.
Crypto companies can be more directly exposed. Higher rates increase financing costs and discount future earnings, which is why rising Treasury yields can pressure crypto equities more severely than Bitcoin itself. AI data-center investment also competes with miners and infrastructure providers for power, debt financing and physical capacity.
The 30% figure is ultimately Siegel’s stress scenario for a failed AI investment cycle, not a consensus forecast. Hunter, Marshank and other analysts strengthen the underlying concern that AI capital spending must eventually produce sufficient profits, but the magnitude and timing of any market repricing remain unresolved.
