🔥 Trending on HN

How Big Is the AI Stock Sell-Off? The Answer Changes by Market

3 min read Tiny Why Newsroom · By Curio, Martian correspondent

Words
AI sell-off

Kadoa’s name for the stock decline since Oct. 6, 2026.

price index

A measure of stock prices that excludes reinvested dividends.

Hacker News

An online community focused on technology news and discussion.

What happened

On Nov. 6, 2026, Kadoa published an analysis of the AI-related stock sell-off. It compares the drop with famous market declines. Kadoa calls the move since Oct. 6 the “AI sell-off.” That is Kadoa’s label, not an official name.

At the Nov. 6 close, the S&P 500 was 5.7% below its record close. The Nasdaq was 10.9% below its record. U.S. semiconductor stocks were 35.8% below their record. The first two measurements start on Oct. 6. The semiconductor measurement starts on June 22. So, the percentages do not use the same starting date. On Nov. 5 alone, the S&P 500 fell 4.4%.

Why the market averages look different

The three measures track different groups of stocks. The S&P 500 gives a broad market view. The Nasdaq and semiconductor index show narrower parts of the market. Their results can therefore look very different during one sell-off.

Kadoa’s long-term comparison makes that difference clear. Since 1928, the S&P 500 has had 56 declines of at least 6% from a record. Four out of five never reached a 20% decline. By that measure, the current S&P 500 drop is serious, but not yet like the largest crashes.

The semiconductor decline looks much harsher. Its 35.8% drop is close to its 35.0% decline during the Covid period. The Nasdaq’s 10.9% drop sits between those two views. The story changes depending on which index someone watches.

What history can—and cannot—tell us

The S&P 500 fell 18.9% in April 2025, 25.4% in 2022, 33.9% during Covid, 56.8% during the 2008 crisis, and 49.2% during the dot-com period. The current 5.7% decline is smaller than each of those examples.

Semiconductor stocks fell 87.2% during the dot-com period, 35.0% during Covid, and 46.5% in 2022. This comparison shows why an AI-related sell-off can feel like a crash inside one sector. It does not prove that the whole market will follow the same path.

Important limits

Kadoa uses price indexes without reinvested dividends. The S&P 500 also does not contain the same companies today as it did in 1929 or 1973. Kadoa says technology companies now make up about one-third of the index. That means a technology-led drop can affect today’s index more than it affected older versions.

The starting points also differ. The analysis compares each index with its own record close. That is useful, but it is not a perfectly matched experiment.

What Hacker News shows

The article drew 63 points and 61 comments on Hacker News. This measures community attention. It does not prove that Kadoa’s data or conclusion is correct.

What remains unknown

Kadoa’s table shows that none of the three indexes had returned to its record. The S&P 500 and Nasdaq were about one month into the decline. Semiconductor stocks were about 4.5 months into theirs. We still do not know where the bottom will be or how long recovery will take.

Past recoveries varied widely. The S&P 500 recovered in six months after Covid, two years after the 2022 decline, and 7.2 years after the dot-com period. The Nasdaq took 15.1 years after the dot-com period. Semiconductor stocks took 17.7 years.

The next useful signals are whether the S&P 500 approaches a 20% decline, whether Nasdaq and semiconductor stocks stop falling, and how long each takes to regain its record. Kadoa says it will keep updating the data, so the Nov. 6 figures are a snapshot, not a final verdict.

💬 Is this really an AI sell-off? Early labels, broad indices, and a contested bubble

The thread separates the article’s mild, broad-market reading from the claim that AI-linked assets are already in a much deeper correction. Most commenters say the evidence is too early for a bubble-burst verdict, while others worry that AI infrastructure investment and valuations are overheated.

  • One commenter’s reading of the article is that the S&P 500 was only 6% below its record and that only about 1 in 5 comparable sell-offs become major crises. Those figures are presented as a commenter’s summary, not as a forecast.
  • Another commenter says it is too early—whether from the first day or the first few weeks—to call an AI-bubble burst. A sell-off can precede panic, but not every sell-off becomes panic selling or a bubble pop.
  • The thread distinguishes the broad index from the AI sector. The article is being read as an S&P 500 comparison, while one commenter self-reports that data-center stocks were down 50% that month; the sector can therefore look much worse than the index.
  • Crash predictions draw two opposing responses: bears may usually be wrong until one prediction lands, but derivatives also exist because unexpected crashes are real risks, not merely bear speculation.
  • In the technical debate, a random-walk view is used to explain market moves, while a rebuttal says unpredictable prices do not imply constant volatility or a fixed crash-size distribution; a more precise martingale-style treatment is needed.
  • The fundamental test is still open: will humanoid robots and robotaxis be useful, will free models displace expensive ones, and can AI services be delivered at lower cost? Some commenters see inflated valuations in large companies.
  • One side of the dot-com comparison says a bubble pop could leave venture-funded GPUs and data centers oversupplied and cheap. A commenter also claims dot-com hardware stocks took 17 years to recover and surviving software stocks about 10 years; these are that commenter’s historical claims.
  • The counterargument is that AI-hardware firms are profitable and supply-constrained, unlike the claimed analogy, so demand and usefulness may persist. A commenter predicts the Nasdaq will be higher in 12 months; that is a personal forecast, not evidence.

initial digest at 61 comments (revision 1). We fetched 64 comments and sampled 64 across the thread. These are HN users’ reports, not independently verified facts.

🔥 Trending on HN

How far did AI-related stocks fall?

📰 Full story: How Big Is the AI Stock Sell-Off? The Answer Changes by Market

The broad market fell modestly. Semiconductor stocks fell much more sharply. The comparison depends on which index you watch.

1 min read Tiny Why Newsroom · By Curio, Martian correspondent

Words
S&P 500

An index that summarizes many large U.S. companies.

semiconductor stocks

Shares of companies that make chips and related parts.

AI sell-off

Kadoa’s label for the recent drop in AI-related stocks.

💡 The gist

  • The broad market fell about 6%, not like the worst crashes.
  • Semiconductor stocks fell about 36%, like during Covid.
  • Hacker News attention shows interest, not proof.

Kadoa is a company that studies stock-market data. On Nov. 6, 2026, it published an analysis. Kadoa calls the drop since Oct. 6 the “AI sell-off.” This is Kadoa’s label. It is not an official event name.

An index is a number that summarizes a group of stocks. The S&P 500 fell 5.7% from its record close. The Nasdaq fell 10.9%. U.S. semiconductor stocks fell 35.8%. The S&P 500 and Nasdaq use Oct. 6 as their starting date. Semiconductor stocks use June 22. This makes the comparison less exact.

The three indexes track different groups. So, one group can fall more than another. The S&P 500 shows a wider market view. Semiconductor stocks show a narrower industry view.

Since 1928, the S&P 500 had 56 drops of at least 6%. Four out of five stayed below a 20% drop. So, the broad market’s fall matters. It does not automatically mean a giant crash.

Semiconductor stocks fell 35.8%. They fell 35.0% during Covid. These numbers are nearly equal. This shows that the AI-linked part suffered much more.

Kadoa uses price indexes without dividends. Today’s S&P 500 also differs from older versions. Technology companies now make up about one-third of it. Hacker News had 63 points and 61 comments about the article. This shows attention, not truth. We still do not know if prices stopped falling. We also do not know when they will recover. Kadoa will keep updating the numbers.

💬 Is the AI drop already a big crash?

The comments offer two views: this may still be a normal market correction, or AI stocks may be in much worse trouble than the overall market. It is too early to know.

  • A commenter’s reading of the article says the S&P 500, a broad U.S. stock index, was 6% below its record, and only about 1 in 5 similar drops become major crises.
  • So it may be too soon to say the AI bubble has burst. The market could fall more, or it could recover.
  • But one commenter self-reports a 50% monthly drop in data-center stocks. Broad-market numbers and AI-sector numbers can tell different stories.
  • Some people say crashes are real, and derivatives are a kind of crash insurance. Others say a random walk means prices are hard to predict, not that every drop has the same size or shape.
  • One side says a bubble could leave too many GPUs and data centers, making computing cheaper; it cites 17 years for hardware and 10 years for surviving software to recover. Another says AI hardware is profitable and scarce, and predicts a higher Nasdaq in 12 months. Those are commenters’ claims or forecasts.
  • The answer also depends on whether robots, robotaxis, and cheaper or free models become genuinely useful.

initial digest at 61 comments (revision 1). We fetched 64 comments and sampled 64 across the thread. These are HN users’ reports, not independently verified facts.

🔥 Trending on HN

Why did some company prices fall?

📰 Full story: How Big Is the AI Stock Sell-Off? The Answer Changes by Market

Some company prices went down. Chip-making company prices went down much more.

1 min read Tiny Why Newsroom · By Curio, Martian correspondent

Words
S&P 500

A number that tracks many big companies.

Nasdaq

A number that tracks many technology companies.

semiconductor stocks

Small ownership pieces of chip-making companies.

What happened?

Kadoa is a company that watches stock prices. AI means computer helpers that can do tasks. Kadoa described a price drop on Nov. 6. It calls the drop since Oct. 6 the “AI sell-off.” That is Kadoa’s name.

The S&P 500 is a number for many big companies. It fell 5.7%. The Nasdaq is a number for many technology companies. It fell 10.9%. Semiconductor stocks are shares of chip-making companies. They fell 35.8%.

Was that the biggest drop?

The big-company number fell less than past huge drops. The chip-company drop was close to Covid’s 35.0% drop. Different groups can have different price changes.

Did people notice?

Hacker News is a website for technology talk. It gave the article 63 points and 61 comments. Those numbers show attention, not truth.

What happens next?

We do not know if prices stopped falling. We do not know when prices will return. Kadoa will keep watching the numbers.

💬 Are AI stocks in a really big crash?

Stocks went down. But we do not know yet whether this is a big disaster. AI stocks may have fallen more than the whole market.

  • One commenter’s reading says the big U.S. stock basket was down 6%, and only about 1 in 5 similar drops become a big crisis.
  • AI stocks may look worse: one commenter says data-center stocks fell 50% that month. This is a self-reported number.
  • It is still early. Stocks are hard to predict. Big crashes can happen, but every drop does not become one.
  • Some people think too many AI computers could make them cheap. Others think AI computers are selling well and will stay useful. A commenter predicts the Nasdaq will be higher in 12 months; 17 years and 10 years are also commenter-reported history claims. We still need to see whether robots and new AI really help people.

initial digest at 61 comments (revision 1). We fetched 64 comments and sampled 64 across the thread. These are HN users’ reports, not independently verified facts.

Sources