Nasdaq Tech Selloff: Stocks Fall 2.6% on AI Fears

Nasdaq tech selloff deepened on Thursday, October 22, as investors who had spent months cheering the artificial intelligence boom suddenly started asking a much harder question: who pays for all of this, and when does it start paying back?

The Nasdaq Composite fell 2.6%, losing 694 points to close at 26,008.86. It was the technology-heavy index’s worst single day since early June. The broader S&P 500 fell 1.2% to 7,663.43, its worst day since late July, while the Dow Jones Industrial Average ended essentially flat, up just 0.02% at 52,867.98.

That split is the story. This was not a day when everything fell. It was a day when technology fell, and fell hard, while much of the rest of the market held its ground. For a market that has been carried for months by a handful of giant technology companies, that distinction matters.

Key takeaways:

The Nasdaq fell 2.6% on October 22, its steepest daily drop since early June.
The S&P 500 fell 1.2%, while the Dow ended essentially flat.
Meta fell about 7% after raising its 2026 capital spending forecast to $135 billion to $145 billion, mostly for AI.
The Philadelphia Semiconductor Index fell 6.7%, its biggest daily drop since June 23.
Chipmakers led the losses: AMD and Intel fell about 8% each, Micron about 6%, and Nvidia 4.4%.
The sell-off spread to Asia the next morning, with Japan’s Nikkei down more than 2% and South Korea’s KOSPI down as much as 3.5% at one point.
Despite the fall, the Nasdaq remained about 5.8% below its all-time high reached on October 6, and still well up for the year.

What happened on October 22

The trigger was not one bad earnings report. It was a batch of high-profile results that, taken together, reminded investors just how expensive the AI race has become.

Meta was at the centre of it. The Facebook parent beat analyst expectations on revenue, yet its shares slid about 7%. The reason was on the spending side of its accounts: Meta raised its projected capital spending for this year to $135 billion to $145 billion, reported its first cash outflow in more than a decade because of AI-related spending, and announced plans to issue non-voting shares in a two-for-one split to help raise capital for the technology.

In plain terms, Meta told investors it needs to spend even more, and it may need to raise money to do it. For shareholders who had hoped AI investment would soon start to ease, that was unwelcome news.

Tesla fell 3.1%, even though it reported positive free cash flow and cut its 2026 capital spending plans. Alphabet also fell 3.1%, Microsoft declined 2.2%, and Amazon fell 4.6%. Among the so-called Magnificent Seven — the giant technology stocks that make up more than a third of the wider market — only Apple escaped with a gain, rising 1.2%.

Why the Nasdaq tech selloff hit chips hardest

If Meta supplied the spark, chipmakers supplied the fuel. The Philadelphia Semiconductor Index, known as the SOX, plunged 6.7% in a single session.

That matters because chips have been the purest way to bet on AI. When companies announce huge data-centre spending plans, chipmakers are supposed to benefit first. When investors start to doubt those plans, chipmakers are therefore hit first, and hardest.

AMD fell about 8%, Intel about 8%, Micron about 6%, and Nvidia, the largest US company by market value, fell 4.4%. Broadcom slid 4.3%. CoreWeave, a newer cloud provider whose debt-heavy business model has been closely watched, fell about 10% as worries about financing costs for AI infrastructure resurfaced.

With Thursday’s fall, the SOX was more than 26% below its record closing high reached on June 22, according to Reuters reporting. Yet it remained up 51.5% so far this year. That single fact captures the tension of the whole sell-off: a sector can be down sharply from its peak and still be up hugely for the year. A correction, in other words, is not the same as a collapse.

A rotation, not a rout

Look beyond technology, and October 22 looked very different. The Dow’s flat finish, while the Nasdaq fell 2.6%, suggests investors were not fleeing stocks altogether. They were moving money out of the sector that had run furthest, and into parts of the market that had lagged behind.

Market strategists described the same pattern in different words: technology, which had been the market’s golden child until about June, was the part suffering, while other sectors held up better. One strategist noted that investors were starting to see holes in the AI story, and because technology now makes up such a large share of the major indices, that reassessment dragged the indices down with it.

There were other pressures in the background. Brent crude rose about 4% on Thursday and continued climbing on Friday morning, taking the international oil benchmark above $114 a barrel, its highest since May, amid Middle East tensions. Rising oil prices raised fresh worries about inflation and about whether the US Federal Reserve would be forced to keep interest rates higher for longer. Treasury yields fell as investors sought safer assets, another sign of a cautious, defensive mood rather than outright panic.

Asia felt it next

The sell-off did not stop at the US close. Asian markets opened lower on Friday. Japan’s Nikkei 225 fell more than 2% before recovering some ground, and South Korea’s KOSPI dropped as much as 3.5%, according to reporting on the session.

Some analysts cautioned against reading too much into a single morning’s move in Asia, describing it as nervous sentiment rather than a sudden change in company fundamentals. That caution is worth keeping in mind. A global sell-off can reflect genuine concern, but it can also reflect traders in one region reacting to headlines from another before they have time to judge the underlying numbers.

Is the AI bubble finally cracking?

That is the question every investor, and every reader of this story, wants answered. The honest answer, based on October 22 alone, is: not proven, but no longer unaskable.

The case for a bubble is straightforward. Companies are committing hundreds of billions of dollars to data centres, chips and power, often funded with debt or new shares. Returns on that spending are uncertain and may take years to arrive. When even strong revenue results are punished because spending is rising faster, the market is signalling that it wants proof, not just promises.

The case against a bubble is equally straightforward. AI demand, as reported by the chipmakers and cloud providers themselves, remains strong. The Nasdaq, even after its worst day since June, was still only about 5.8% below its all-time high from October 6. The S&P 500 was still up 11.9% for the year at that point. Bubbles do not usually correct by 2.6% in a day and then stop; they tend to unwind over months, with repeated rallies that fail.

What October 22 most likely represents is a repricing. Investors are no longer willing to pay any price for AI exposure. They want to know how the spending will be financed, what return it will generate, and which companies have the balance sheets to survive a longer, more expensive build-out. That is a healthier question than blind enthusiasm, even if it makes for an uncomfortable day on the screens.

What happens next

Earnings from the other giants. Amazon, Alphabet and Microsoft had already reported in this round, but the next set of results from major technology and chip companies will be watched even more closely. Any sign that AI spending is slowing, or that returns are improving, could stabilise the market quickly.
The financing question. Meta’s plan to issue shares put financing in the spotlight. If other companies follow, investors will ask whether shareholders are being diluted to fund a race with no clear finish line. If companies can fund spending from cash flow instead, the mood may improve.
Oil and interest rates. Brent above $114 a barrel complicates the picture. If energy prices stay high, inflation worries could keep interest rates elevated, which makes expensive growth stocks harder to justify. If oil cools, some of that pressure eases.
Whether chips stabilise. The SOX has already fallen a long way from its June high. A period of stability there would suggest the chip sell-off was a valuation reset. Another sharp leg down would suggest investors are questioning demand itself, not just prices.
A test of nerves, not just numbers. The most important signal may not be a single earnings report, but whether buyers return on the next dip. In a healthy bull market, sharp falls attract buyers. In a market that has lost confidence, they attract more sellers.

For now, October 22 should be read as a warning, not a verdict. The AI boom has not ended in a day. But the market has made clear that the boom will no longer be funded on trust alone.

FAQs

Why did the Nasdaq fall 2.6% on October 22? Renewed fears about the scale and financing of AI spending, triggered by a batch of technology earnings, led investors to sell technology and chip stocks. Meta’s higher capital spending forecast was the main catalyst.

Why did Meta fall about 7% despite beating revenue expectations? Because investors focused on spending, not sales. Meta raised its 2026 capital spending forecast to $135 billion to $145 billion, reported its first cash outflow in more than a decade due to AI spending, and announced plans to issue shares to help fund it.

Which stocks fell hardest? Chipmakers led the losses. AMD and Intel fell about 8% each, Micron about 6%, Nvidia 4.4%, and Broadcom 4.3%. CoreWeave fell about 10% on financing worries.

Why did the Dow stay flat while the Nasdaq fell? Because the sell-off was concentrated in technology. Investors appeared to rotate money into other sectors rather than sell stocks across the board, so the Dow, which is less weighted to technology, held its ground.

Did the sell-off spread outside the US? Yes. Asian markets opened lower the next morning, with Japan’s Nikkei down more than 2% and South Korea’s KOSPI down as much as 3.5% at one point, before partially recovering.

Is this proof that AI is a bubble? No, not on one day’s evidence. It is proof that investors are now demanding clearer evidence on how AI spending will be financed and when it will pay off. Whether that becomes a longer downturn depends on upcoming earnings, financing plans, and chip demand.

Sources: Wall Street Journal sheet source https://www.wsj.com/finance/stocks/u-s-tech-stocks-drop-on-ai-spending-concerns-05400b55; Reuters via Al Jazeera; Invezz.

This article reports market movements as reported on October 22-23, 2026. It is not investment advice.