Why Did the Nasdaq Fall Nearly 6%? AI Rally Under Pressure

On October 6, 2026, the Nasdaq Composite closed at a record high. By October 23, it had fallen almost 6%. Yet over the same period, the Dow Jones Industrial Average rose, a divergence that suggests investors were reassessing technology stocks more sharply than the broader US market.

The pressure came from several directions. Higher oil prices and inflation complicated the interest-rate outlook, concerns about funding AI infrastructure reached the bond market, and potential competition challenged chipmakers’ profit expectations.

Together, these developments sharpened a question behind the AI rally: strong demand may support growth, but how much will companies need to spend and borrow before that growth delivers lasting profits?

Nasdaq trend from October 6 to the latest available close

The Nasdaq Composite declined 5.96% from its October 6 close to October 23. The broader S&P 500 fell about 2.50%, while the Dow gained 1.35%.

Source: Nasdaq daily closing data via FRED. 

The S&P 500 and Dow comparison above uses the same October 6-23 closing window. These benchmarks contain different businesses and use different construction methods. Their divergence shows why the Nasdaq decline should be examined separately from the broader US market.

There is no October 26 regular-session Composite reading yet at this update. Premarket Nasdaq-100 futures are a separate instrument and are not added to this closing-index series.

The timing matters, too. By October 16, the Nasdaq was already 5.29% below its October 6 record. It then recovered 2.15% through October 21, before falling 2.60% on October 22 and another 0.21% on October 23.

What changed after the October 6 record?

Oil and inflation made future growth more expensive to finance

The record close came during a temporary easing of pressure from oil and Treasury yields. On October 7, renewed increases in yields and oil concerns weighed on stocks again.

The September inflation report, released on October 14, reinforced the problem. Consumer prices rose 3.7% year on year, up from 3.4% in August. Gasoline accounted for more than half of September's 0.6% monthly increase.

By October 23, Brent crude futures had settled at $116.72 a barrel, as renewed attacks in Saudi Arabia raised concerns about energy supplies.

For technology investors, the connection runs through both valuation and financing. Higher returns on government bonds make distant, uncertain stock-market profits less attractive. Expensive borrowing also raises the cost of building the infrastructure intended to produce those profits.

That second effect became particularly important for companies relying on lenders to fund their expansion.

The AI financing concern reached the bond market

CoreWeave, which trades under CRWV, became a visible example.

The company rents computing capacity to AI customers. Expanding that capacity requires large investments before the resulting customer payments necessarily recover the cost.

On October 15, the yield on its dollar bond maturing in October 2031 briefly exceeded 16%. Its shares also suffered heavy selling during that week.

A rising bond yield means investors demand a higher return to own that debt at its current market price. It does not automatically reset the interest payments on existing fixed-rate bonds. It can nevertheless signal that future borrowing or refinancing may be harder or more expensive.

That is a more consequential concern than ordinary stock-price volatility. A business can have growing demand and still face a difficult gap between paying for capacity and collecting enough cash from it.

The potential effect extends to suppliers. If cloud providers cannot fund planned projects on acceptable terms, chip purchases could arrive later than expected. This is a financing risk to the spending cycle, rather than evidence that all those orders have already been cancelled.

Chipmakers faced a separate challenge to their profit expectations

Reports of progress by Chinese memory-chip producer CXMT added another concern around October 13: future competition.

Part of the attraction of memory suppliers rests on the possibility that tight supply supports strong prices. More competition could weaken that pricing power, even if the overall market for AI computing continues growing.

This helps explain why “AI demand is strong” became an incomplete answer to falling semiconductor shares. Shareholders also need to know what suppliers will earn from that demand and how long unusually favourable conditions can last.

The reported developments did not establish that Chinese competitors had already displaced established suppliers. They introduced uncertainty into expectations that had helped support the rally.

Why didn't falling bond yields stop the Nasdaq selloff?

This is the detail that a simple “stocks fell because interest rates rose” explanation misses.

On October 22, the 10-year Treasury yield fell approximately 0.19 percentage point to 4.953%, as investors sought safety in government bonds. The Nasdaq still fell 2.60%.

Earlier in October, elevated yields had made the backdrop difficult. But during this session, lower yields reflected a move away from risk rather than confidence in a better growth outlook.

That distinction matters for what happens next. If doubts centre on financing and the returns from AI spending, a decline in government-bond yields may offer only partial relief. Investors would still need more confidence in the businesses themselves.

Author's take: Is this a dip or a deeper warning?

My reading is that this was a reassessment of the financial assumptions behind the AI rally. The evidence supports taking the funding concern seriously, while distinguishing between the companies exposed to it.

The Dow's gain over the same period shows that investors could find reasons to own other parts of the market. Meanwhile, the pressure on AI-linked stocks suggests that earlier expectations for spending, pricing power and eventual returns were facing greater scrutiny.

The important distinction is between a company whose growth becomes less profitable and one whose growth becomes difficult to finance.

For the first, the debate is about the price investors should pay for future earnings. For the second, it can extend to whether expansion must slow, borrowing becomes more expensive or additional shares must be issued. A similar percentage fall in their stock prices would not make them equivalent opportunities.

That is why I would not treat the Nasdaq's 6% decline as a sufficient measure of value. A company's shares can fall while its financial outlook weakens by even more. Conversely, a lower price can become more attractive when the underlying earnings outlook remains intact. The evidence needs to be assessed at the level of the holding.

The next useful test is therefore the quality of the growth being reported. For large technology companies, that means cash generation after infrastructure spending and credible evidence of returns on that investment. For debt-dependent providers, it means financing terms, project funding and the timing of customer cash receipts. For chipmakers, it means whether orders and pricing remain consistent with the profits investors expect.

A recovery supported by those improvements would carry more weight than a short rebound driven mainly by renewed spending promises.

The October selloff has raised the standard of evidence needed to justify the AI rally. Demand remains an important part of that evidence. The ability to fund that demand and retain an attractive share of the resulting profits is equally important.