Why AI Stocks Fell Even Though Profits Were Strong
It sounds impossible: a company earns more money than ever, and its share price falls. But that is exactly what happened in the AI sell-off of November 2026. Intel grew 21%, AMD grew 45%, Nvidia's data-centre revenue grew 117% — and their shares fell anyway. How can good profits and falling prices exist at the same time? This is one of the most important lessons in the stock market, so let's understand it properly.
The market buys the future, not the past
Here is the key idea: a share price is not a reward for what a company already did. It is a bet on what the company will do next. When you buy a share, you are paying today for profits you hope the company will earn in the future.
That means results tell you about the past three months, but the price is decided by expectations about the next three years. If the results are good but the future looks a little less bright than people hoped, the price falls — even though the company just had a great quarter.
Guidance matters more than results
Every time a company reports results, it also gives "guidance" — its own forecast for the next quarter. In November 2026, guidance was the real story. Intel reported revenue of $16.51 billion, up 21%, ahead of the $16.44 billion analysts expected, with data-centre and AI revenue up 78% to $7.3 billion. But it forecast $15–16 billion for the fourth quarter, against $17.1 billion expected — and its shares fell 28.58% on 30 October, the worst single day since the 1970s.
AMD did the opposite: revenue of $13.4 billion, up 45%, data-centre revenue nearly doubling, a fourth-quarter forecast of $17 billion, and an expectation of more than 80% revenue growth in 2027. Even so, AMD shares fell 8.75% to $534.06 on 13 November, after rising 7.32% to $552.90 the day before. Good guidance helped, but it could not protect the shares from a market that was selling the whole sector.
Valuation: when the price already assumes perfection
After two years of huge gains, AI shares were priced as if everything would go perfectly forever. That is what "high valuation" means: the price already includes years of future growth. When a price assumes perfection, there is no room for even a small disappointment. A slightly weaker forecast, a slightly slower quarter, or a new doubt is enough to send the price down sharply.
Think of it like this: if everyone already expects a student to score 100%, scoring 95% is a great result — but it still disappoints everyone who expected 100%. That is what happened to AI stocks in November 2026.
Debt, data centres and doubt
Three new doubts arrived at the same time. First, investors asked whether the enormous spending on data centres would produce profits fast enough. Companies are spending billions building data centres and buying chips — if the returns come slowly, the spending looks less wise.
Second, politics entered the picture. After the US midterm results on 3 November, there were calls to restrict data-centre development. If data centres become harder to build, chip demand grows more slowly.
Third, financing. Some AI infrastructure companies carry very heavy debt — CoreWeave, for example, had about $35 billion of financial debt at 30 June plus $16.5 billion of lease liabilities, against only $5.5 billion of cash. When borrowing is expensive, heavy debt makes investors nervous. CoreWeave itself reported strong results — revenue of $3.66 billion, up 168%, and raised 2026 guidance to $13–13.3 billion — and its shares jumped 24.7% on the rebound day. But the debt question still hangs over the sector.
So is the AI story over?
Falling prices do not mean a technology has failed. The rebound on 12 November showed that clearly: the semiconductor index jumped 5.66% to 10,564.11 in a single day, with AMD up 7.32%, Intel up 7.52% and Micron up 11.9%. The market was re-pricing AI, not abandoning it. Anthropic is even reported to be planning an IPO — a sign that investors still see a long future for AI companies.
What changed is the mood. The market moved from "AI can do no wrong" to "prove it, quarter by quarter." That is actually a healthier place for beginners to learn from — prices that move on evidence, not on excitement.
What to watch next
The next big test is Nvidia's earnings on 17 November 2026, after the market closes. Nvidia is the biggest AI chip company — revenue of $96.2 billion in its last reported quarter, up 106%, with data-centre revenue of $89 billion, up 117% — and its results and forecast will set the tone for the whole sector. Watch three things in that report: the revenue number, the forecast for the next quarter, and what the company says about data-centre demand. Those three things will matter more than any single day's price move.
Final words
Strong profits and falling shares can exist together, because the market pays for the future, not the past. Results tell you where a company has been; guidance, valuation and confidence tell you where its price is going. If you remember that one idea, the November 2026 crash stops looking like a mystery — and starts looking like a lesson.
In the next post: Nvidia's CEO Jensen Huang has resigned — what happened, and what comes next for the world's biggest AI chip company.
Sources: Reuters, company earnings reports, November 2026.