back to top
Sunday, November 22, 2026

Top 5 This Week

The AI Sell-Off: Nasdaq’s Worst Week Since 2020 as the AI Boom Faces Its Biggest Test

The artificial intelligence boom, the most powerful force in global stock markets for the past three years, faced its biggest test in the week ending Friday, 20 November 2026 — and it failed it. The Nasdaq Composite, the index most closely associated with technology and AI, fell by about 10 per cent in a single week, its worst weekly performance since March 2020, when the pandemic sent markets into freefall. By Friday’s close, the Nasdaq stood at 22,617.73, roughly 18 per cent below its record high of 6 October, and the gains it had built up across 2026 had been wiped out. The sell-off was not confined to one index or one stock. It spread across the market, across the AI supply chain, and even into cryptocurrency, raising a question that investors had been avoiding for months: what if the trillions of dollars poured into artificial intelligence do not pay off as quickly, or as surely, as everyone assumed?

The scale of the week’s losses was striking. The S&P 500, the broadest measure of the American stock market, fell by about 8.6 per cent for the week, its seventh consecutive daily loss — a losing streak that speaks to the persistence of the selling rather than a single day of panic. The Dow Jones Industrial Average fell by about 8 per cent. These are not the numbers of a normal market wobble. They are the numbers of a market reassessing a central belief, and doing so all at once.

Nvidia: the week’s defining fall

At the centre of the sell-off stood Nvidia, the chipmaker whose processors power the AI revolution and whose stock had become the single most important barometer of confidence in the entire AI story. On Friday alone, Nvidia fell 14 per cent to $125.53. For the week as a whole, it was down about 23 per cent. And measured against its October record of $239.24, the stock has now fallen by roughly 48 per cent — nearly half its value gone in a matter of weeks.

The immediate trigger for Friday’s plunge was a Reuters report that staff at the US Securities and Exchange Commission are reviewing Nvidia’s accounting. It is important to be precise about what that means: the report described a preliminary review by SEC staff, and it implied no finding of wrongdoing. A regulatory review of this kind is a routine part of the SEC’s work, and many such reviews conclude without any action being taken. But in a market already nervous about AI valuations, the report was enough to accelerate selling in the stock that mattered most. Investors did not wait to find out whether the review would lead anywhere; they sold first and asked questions later.

The deeper story, though, is not about a single news report. Nvidia’s valuation had come to rest on an assumption: that demand for its AI chips would keep growing at an extraordinary rate, that the data centres being built around the world would keep buying, and that the profits would justify a share price that had risen to levels few companies in history have ever reached. When that assumption began to wobble — when investors started to question the pace of data-centre construction and the returns on AI investment — the stock that had risen furthest had the furthest to fall.

Anthropic’s debut and the IPO paradox

The week’s other defining AI story was Anthropic, the AI company behind the Claude chatbot, which completed what was described as the largest initial public offering on record, raising $81 billion. The debut itself was a success: the shares opened at around $111.50, against an offer price of $83, and closed the first day up about 17 per cent. For a few hours, it looked like the AI story still had the power to excite the market.

Then came Friday. Anthropic’s shares fell by about 8.5 to 9 per cent, swept down in the same selling that hit the rest of the sector. The contrast between the debut and the fall captures the paradox of the moment. Investors are still willing to pay extraordinary sums for a stake in the AI future — $81 billion is not a small bet — but they are no longer willing to hold AI stocks at any price, through any doubt, without asking hard questions about when the money will come back. An IPO can be a triumph and a warning at the same time, and Anthropic’s week was both.

Bitcoin joins the slide

The sell-off did not stop at the stock market. Bitcoin, the world’s largest cryptocurrency, ended the week near $58,895, below the $60,000 level that traders watch closely, down about 9.8 per cent for the week and more than 50 per cent below its 2025 peak. Over the weekend it slid further, towards $56,000. Bitcoin’s fall matters because it has become, for many investors, a proxy for risk appetite — the willingness to hold speculative assets when confidence is high. When Bitcoin falls alongside the Nasdaq, it is a sign that the mood has shifted from caution about one sector to caution about risk itself.

The link between AI stocks and Bitcoin is not a direct one. Bitcoin’s price is driven by its own dynamics — regulation, institutional adoption, the halving cycle and the ebb and flow of speculative interest. But the two markets share an investor base, and they share a psychology. The same investors who bought Nvidia on the promise of an AI future also bought Bitcoin on the promise of a decentralised one. When confidence in one promise falters, confidence in the other tends to follow.

What spooked the market

Behind the week’s selling lies a single, uncomfortable question: will the trillions of dollars being poured into artificial intelligence actually pay off? The numbers involved are staggering. Technology companies, data-centre operators and their lenders have committed sums that would have been unimaginable a decade ago — building vast computing facilities, buying chips by the tens of thousands, and borrowing heavily to do it. The bet is that AI will transform productivity, create new industries and generate returns that justify the spending.

But in recent weeks, investors have begun to focus on the other side of that bet: the borrowing costs. Data-centre construction is capital-intensive, and it is financed with debt. As borrowing costs have risen and as the pace of data-centre buildouts has shown signs of slowing, the arithmetic has started to look less comfortable. If the facilities take longer to build, cost more to finance, and generate returns more slowly than expected, then the companies that borrowed to build them — and the chipmakers that supplied them — face a difficult period. That is the worry that has driven the selling, and it is a worry about the fundamentals of the AI investment cycle, not about a single company’s results.

It is worth putting the week in perspective. The AI boom has not been a bubble in the simple sense of a market with no underlying reality. Artificial intelligence is a real technology, and it is already changing how people work — including how AI is already changing everyday work in Pakistani businesses. The question is not whether AI matters, but whether the market had priced in too much, too fast. The history of technology booms — from railways to the internet — is full of moments when a genuinely transformative technology met a market that had run ahead of itself, and the correction that followed was painful but not fatal to the technology itself.

What happens next

Nobody can say with confidence where the market goes from here, and this article will not pretend otherwise. A sell-off of this scale can be the start of a longer correction, or it can be the sharp reset that clears the way for a healthier recovery. What matters is what happens to the underlying drivers: whether data-centre spending stabilises, whether AI companies can show the revenue growth that justifies their valuations, whether borrowing costs ease, and whether the SEC review of Nvidia’s accounting concludes without findings. Each of those is a genuine uncertainty, and the market will be watching each one closely.

For ordinary investors — and for the pension funds and savers whose money is invested in these markets — the week’s lesson is an old one: diversification matters, and concentration in a single theme, however exciting, carries real risk. The AI boom made many investors wealthy, and the sell-off has reminded them how quickly those gains can reverse. For the technology industry itself, the test is different. The companies building AI must now show that the technology can deliver the productivity and the profits that the market was promised, at a pace that justifies the extraordinary sums being spent. That is the test the AI boom now faces — and the week ending 20 November was the moment the market decided to start asking the question in earnest. As reported by TradingView, Nvidia’s near-48 per cent fall from its October peak has become the symbol of a week in which the AI story changed.

Shayan Hussain
Shayan Hussain
Shayan Hussain is the founder and owner of Pakistan Morning Post. He is also a YouTuber and Freelancer. The sole purpose of Pakistan Morning Post is to share updates independently about Pakistan and rest of the world.

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Most Read Stories