Nasdaq 100's AI Boom Meets Its Auditor: Funding Costs, Not Demand, Now Set the Tone

A 2.2% slide on Friday capped a bruising week for the Nasdaq 100, which has shed 5.5% over seven sessions as the market's AI conversation shifts from how fast the technology can grow to who can afford to build it. The trigger was not a collapse in orders but a repricing of the balance sheets behind the buildout.

Charu Chanana, chief strategist at Saxo, frames the change plainly: investors are demanding profitability and capital discipline rather than rewarding expansion at any price. The distinction matters because the operating picture and the financing picture are pointing in different directions.

Chipmakers Deliver, Borrowers Stumble

Chanana points to strong results from TSMC, Samsung and ASML as evidence that demand for computing power remains intact. Bloomberg reported Friday that there are still few signs of weaker semiconductor orders, which means the selloff cannot simply be read as a demand collapse that has already happened.

The pressure sits elsewhere. CoreWeave, Nebius and Oracle carry heavy debt loads, and CoreWeave in particular has become a warning sign for AI infrastructure financing. Its borrowing costs have climbed sharply, and some of the bonds it issued this year have touched record lows on the secondary market. Reuters reported Friday that investors are focused on a debt burden of roughly $50 billion at the company.

That is a company-specific strain, not a verdict on every index member. Still, it illustrates the link between the capital intensity of the AI buildout and the cost of credit — a connection that has moved to the center of the debate.

Inflation Surprise Sharpens the Rate Calculus

The macro backdrop made the financing question more urgent. Consumer prices rose 3.7% year over year in September, according to AP, above the 3.6% economists had expected and up from 3.4% in August. The hotter-than-forecast reading deepened uncertainty about the path of monetary policy, pushing Treasury yields higher and weighing especially on technology and AI shares.

Futures markets responded. Reuters reported that the probability of a rate hike at the Fed's October 27–28 meeting climbed to about 37%, up from roughly 22% before the inflation data landed. Reuters had already flagged rising rate anxiety as a source of selling pressure on Thursday.

The damage spread well beyond any single infrastructure financier. AP listed Micron Technology, Advanced Micro Devices, Broadcom and Intel among Friday's decliners, while Meta and Alphabet also fell.

Gartner Puts a Number on the Hurdle

How steep the financing math could become is the subject of a Gartner forecast. To cover investment and financing costs plus a 12% return on capital, AI data centers would need to generate a cumulative $8.5 trillion in revenue by 2030, the firm estimates.

Gartner expects at most $3.6 trillion in addressable AI revenue from businesses and consumers by that date. On its assessment, more than $3.5 trillion of infrastructure investment could be at risk of impairment.

Those are projections and risk scenarios, not booked losses. They do, however, explain why investors are looking past growth rates toward funding costs and the eventual utilization of the infrastructure being built.

Earnings Season Becomes the Proving Ground

The next concrete test arrives with quarterly reports. Netflix reports after the close on October 20, with Tesla and Meta following after the U.S. close on October 21. Meta's capital spending outlook is seen as a key read-through for AI stocks.

For the Nasdaq 100, two separate tests now run in parallel. Companies must confirm the demand and earnings expectations baked into their valuations, and the rate path will help determine the terms on which the next phase of the AI buildout can be financed. Strong results and cheap funding are different requirements — and earnings season can only settle the operating side of the ledger.