AI Stocks Plunge as Nasdaq Falls 5.5% in Post-Election Selloff and Nvidia, AMD and Chipmakers Lead the Slide

Introduction

Wall Street’s faith in the artificial intelligence boom suffered its sharpest blow in more than a year on Thursday, November 5, 2026, as a selling wave that began in AI infrastructure stocks spread across the whole market and erased the entire post-election rally in a single session. The technology-heavy Nasdaq Composite fell 5.52 percent to 24,315.00, the S&P 500 dropped 4.36 percent to 7,353.73 in its worst day since April, and the Dow Jones Industrial Average lost 1,833.06 points, or 3.44 percent, to close at 51,523.06. The Philadelphia Semiconductor Index plunged 12.25 percent. Nvidia, the chipmaker at the centre of the AI boom, fell 9.3 percent to 181.60 dollars, Advanced Micro Devices dropped 17.3 percent to 404.40 dollars, and memory-chip maker SanDisk sank 22.3 percent to 806.43 dollars. The selloff was the second straight day of heavy losses for AI-linked shares, and it arrived just 24 hours after the same market had surged on the Democrats’ sweeping victory in the midterm elections. What changed in a day was not the election result but what investors decided it meant, and the figures reported below show a market suddenly pricing in slower data-centre building, tougher regulation, and doubts about whether the hundreds of billions of dollars pouring into AI will ever pay for themselves. This is the story of the AI stocks plunge of November 2026, the Nasdaq selloff, and what Nvidia, AMD and the chipmakers now face.

1. What happened on Thursday

1.1 The indexes fall hard

The scale of Thursday’s decline was remarkable even by the standards of a volatile year. The S&P 500 closed at 7,353.73, down 334.89 points or 4.36 percent from its November 4 close of 7,688.62, according to figures reported by the Associated Press. It was the index’s worst daily fall since April 4, 2025, when markets were reeling from the tariff shock of that spring. The Nasdaq Composite fell 5.52 percent to 24,315.00, putting it more than 10 percent below its October 6 record close and into technical correction territory. The Nasdaq 100, which tracks the largest non-financial companies on the Nasdaq exchange, fell 6.36 percent to 26,629.81, finishing 14.7 percent below its October 6 record close of 31,224.47. The Dow, which had surged more than 1,500 points on Wednesday, gave back all of that gain and more, falling 1,833.06 points to 51,523.06. The selling was not confined to technology. By the close, the decline had spread across the market, and even the European Stoxx 600 index finished down 0.38 percent, with its technology sector sliding more than 3 percent. The speed of the reversal was what stunned traders most: on Wednesday the S&P 500 had risen 1.58 percent to 7,688.62 in a broad post-election rally, and barely a day later that entire advance had been wiped out and replaced by a rout.

1.2 The biggest losers

The heaviest losses were concentrated exactly where the AI boom had been hottest. Among the largest decliners reported at the close were SanDisk, down 22.3 percent, Lumentum, down 18.7 percent, Micron Technology, down 17.5 percent, Western Digital, down 17.4 percent, and Advanced Micro Devices, down 17.3 percent. Nvidia fell 9.3 percent to 181.60 dollars, a decline that alone erased tens of billions of dollars of market value from the world’s most valuable chipmaker. Broadcom fell about 5 percent, Qualcomm dropped about 10 percent after issuing a weaker-than-expected outlook, and Oracle declined close to 5 percent. The pain extended beyond chipmakers to the companies building and powering data centres. CoreWeave, which rents access to AI chips through the cloud, lost about 15 percent. GE Vernova, a supplier linked to data-centre power demand, fell 9.1 percent. Dell Technologies, one of the biggest sellers of AI servers, dropped about 9 percent on Thursday after falling sharply the day before, erasing about 20 billion dollars in market value, according to market reports. Data-centre cooling stock Modine fell 9 percent on Thursday after a 21 percent decline on Wednesday. Tesla slid more than 10 percent. Even the largest technology companies were not spared: Amazon declined 3.9 percent, Alphabet lost 2.7 percent, and Meta Platforms fell 3.4 percent. The Philadelphia Semiconductor Index, which tracks 30 leading chip companies, plunged 12.25 percent to 9,011.14, a drop of 1,257.52 points and its lowest level in more than six months.

2. Why investors are selling AI stocks

2.1 Doubts about returns on AI spending

The deepest reason for the selloff is a question that has been building for months: will the enormous sums being spent on artificial intelligence ever generate enough revenue and profit to justify them? JPMorgan estimates that global AI-related spending this year is approximately 900 billion dollars, potentially reaching 5.5 trillion dollars cumulatively by 2030. That is an extraordinary amount of money, and Wall Street has increasingly questioned whether the returns will match it. The doubts sharpened after recent results from the biggest spenders. According to a Reuters analysis, recent results and outlooks from Alphabet, Amazon and Intel disappointed investors: Alphabet posted its largest single-day drop since its initial public offering after missing expectations on advertising and search revenue, and Intel fell nearly 29 percent in a single day after issuing weak guidance. When the companies spending the most on AI report results that fall short, investors start to ask whether the spending itself is the problem. The 10-year United States Treasury yield, which remains near 4.9 percent, adds to the pressure, because high borrowing costs make it more expensive to finance data centres and reduce the present value of profits that may arrive years in the future. The Nasdaq 100 had already fallen more than 10 percent from its October 6 peak by November 3, marking its third technical correction of the year, which shows that the anxiety predates the election and was already building through late October.

2.2 Earnings disappointments pile up

Thursday’s selling had a fresh trigger in company results. Qualcomm reported fourth-quarter results that beat estimates, but its shares still dropped about 10 percent because its outlook for smartphone chip sales disappointed, with the company citing sluggish demand and declining revenue linked to Apple. Viavi Solutions crashed 18 percent in morning trading even though it reported that quarterly revenue had jumped 57 percent to 469.3 million dollars, raised its outlook, and approved a 500 million dollar stock buyback. When a company can beat expectations and still fall by nearly a fifth, the market is telling investors that it no longer believes the growth story at any price. The pattern repeated across the sector: good news was ignored and bad news was punished. On Wednesday, data-centre builders and power-equipment makers had already fallen as much as 20 percent or more, and Dell and optical-networking makers dropped 7 to 10 percent, even as the S&P 500 rose. By Thursday morning the selling had spread to the biggest names, with Nvidia’s drop nearly doubling to about 6 percent in early trading and Arm’s to about 10 percent, and this time the broader market fell too. The message from the market was consistent: investors are no longer willing to pay any price for AI growth, and they are demanding proof that the spending produces profits.

3. The midterm election factor

3.1 Democratic sweep raises regulatory fears

The proximate trigger for the two-day selloff was the result of the midterm elections held on Tuesday, November 3. Democrats won comfortable majorities in both chambers of Congress and victories in key governors’ races, and the market’s first reaction on Wednesday was euphoria, with the S&P 500 rising 1.58 percent and the Dow surging more than 1,500 points. But by Thursday, investors had begun to focus on what Democratic control might mean for the AI industry specifically. With Democrats now in charge of both chambers, investors are bracing for more oversight of an industry that many governments are already trying to regulate, and for possible limits on how quickly new data centres can be built. The fear is that a Democratic Congress will scrutinise data-centre permitting, electricity use and community impact far more closely than the previous Congress did. The day after winning, Democratic Congresswoman Yassamin Ansari publicly called for a pause on advanced AI research and stronger industry oversight upon taking office, including a ban on superintelligent AI. In September, Independent Senator Bernie Sanders and Democratic Representative Greg Casar jointly introduced the Ban on Artificial Superintelligence Act, with Ansari, Ro Khanna and other Democratic lawmakers co-sponsoring it. The political risk is real but unsettled: Congress does not change hands until January, and any state-level pause would still have to be written and passed. But markets rarely wait for laws to be written, and Thursday’s selling showed investors pricing in the risk before any bill is introduced.

3.2 Data-centre pauses and the UBS warning

The regulatory fear has a concrete number attached to it. UBS analysts estimated that if construction pauses spread, nearly a third of the new United States data-centre capacity expected to come online in 2028 could be delayed or lost. That is a striking figure for an industry whose entire investment case rests on building capacity as fast as possible. The estimate landed in a market already nervous about local opposition: data from Citi Research shared by The Information shows that United States data-centre bans and pauses remained near zero through mid-2025, reached roughly 20 by December, exceeded 300 in June 2026, and surpassed 500 by October. In other words, the local backlash against data centres has gone from almost nothing to more than 500 restrictions in less than a year. Investors who had assumed that data centres could be built anywhere, quickly, are now discovering that communities and local governments can slow or stop them. Asian semiconductor shares fell sharply overnight on concerns that a Democratic-controlled Congress could tighten data-centre development restrictions and weaken demand for AI infrastructure and chips, showing that the fear is global. The Nasdaq 100 was heading toward a 1.8 trillion dollar decline in market value, according to Reuters, a figure that captures how much wealth the selloff is destroying.

4. The data-centre backlash on the ground

4.1 Voters reject data centres

The political pressure on data centres is not only coming from Washington. On November 3, voters across the United States delivered a verdict of their own. According to Ballotpedia figures reported after the election, 41 local measures concerning data centres were on ballots in 10 states, and voters rejected data centres in 40 of them. In Harwood, North Dakota, the sole exception, voters rejected extra review by 325 votes to 105. In Ohio, 21 measures in 18 jurisdictions went against data centres. In Alhambra, 82 percent of voters backed a ban. In Joliet, 79 percent voted no. In Morgan County, 88 percent backed restrictions. In El Dorado, a 10-year ban passed by 71 percent to 29 percent. The results were not confined to one region or one party: they came from Texas, Georgia, Ohio and beyond, and they followed a pattern of growing local resistance to the noise, water use, electricity demand and construction disruption that large data centres bring. An Embold poll in August found that 75 percent of voters were opposed to a data centre near their home. Exit polls reported by CNBC and NBC found that 56 percent of voters called AI and data centres a crucial or major factor in their vote. The ballot results gave investors a second reason to worry: even if Congress does nothing, local voters are already saying no.

4.2 Bans and pauses multiply

The local backlash is already changing the economics of the AI buildout. OpenAI’s 8-gigawatt PORTS-Pike campus and other giant projects depend on the assumption that land, power and permits will be available on schedule. When more than 500 bans and pauses are in force, that assumption looks shaky. December votes are still to come in Clarksville, Arkansas and North Dakota, and each one is now watched by investors as a signal of how far the backlash will go. Some of the measures are advisory rather than binding, and the final results are still unofficial and subject to certification, but the direction of travel is clear. The industry’s response has been to argue that data centres bring jobs, tax revenue and investment, and that the demand for computing power will keep growing regardless of local politics. The market’s response on Thursday suggested that investors are no longer sure. When a company like CoreWeave, whose entire business is renting access to AI chips, falls 15 percent in a day, and when the neocloud companies CoreWeave and Nebius both fall sharply for a second day, the market is saying that the business model itself is under question, not just one company’s results.

5. The money at stake

5.1 Trillions in market value

The sums involved in the selloff are difficult to grasp. The Nasdaq 100 was heading toward a 1.8 trillion dollar decline in market value, according to Reuters, and CNBC reported that the Nasdaq 100 was set to lose more than 2 trillion dollars in market value in October alone as big technology earnings intensified questions about returns on AI spending. Individual companies lost staggering amounts in a single session: Dell’s 9 percent fall erased about 20 billion dollars in market value, and Arm’s 10 percent fall erased roughly another 20 billion dollars. Nvidia, which had been the most valuable company in the world, lost about 9.3 percent in one day. These are not paper losses in the abstract: they represent pension funds, retirement accounts and ordinary savers watching the value of their investments fall. Retail investors sold a net 358 million dollars of individual United States stocks as of 2:58 p.m. in New York on Thursday, on pace for the largest single-day net selling since March 17, 2020, when the figure was 502 million dollars. That is a sign that the selling was not only driven by large institutions but by ordinary investors heading for the exit at the same time. Bitcoin, which had been down only about 1 percent through the morning, slid nearly 6 percent by about 11 a.m. Eastern Time once the broader market started tumbling, showing that the risk aversion had spread beyond stocks.

5.2 The spending mountain

Behind the market value figures sits the spending mountain that the AI industry has built. JPMorgan’s estimate of about 900 billion dollars of global AI-related spending this year, rising to a cumulative 5.5 trillion dollars by 2030, is the largest corporate investment programme in history. The money is going into chips, data centres, power plants, cooling systems and the fibre-optic networks that connect them. The companies spending it argue that AI will transform every industry and that the returns will be enormous. The sceptics argue that the spending is being financed with debt, that the data centres may become obsolete as chip technology advances, and that the electricity and water demands are unsustainable. The truth is that nobody knows yet, because the largest AI investments are only a few years old and their returns are still mostly in the future. What Thursday’s selloff showed is that the market’s patience is not unlimited. When the 10-year Treasury yield sits near 4.9 percent, borrowing to build data centres is expensive, and investors want to see evidence that the money is earning a return. The October employment report, due on Friday, will be watched closely for further clues about the health of the economy, and any sign of weakness could add to the pressure on stocks.

6. The winners and the rotation

6.1 Money moves elsewhere

Not everything fell on Thursday. The selloff was, in part, a rotation: money leaving AI and technology stocks and moving into other parts of the market. On Wednesday, while AI stocks fell, the broader S&P 500 still rose, because investors were buying other sectors at the same time. Thursday was worse because the selling spread to the whole market, but even then some areas held up better than technology. The rotation reflects a judgment that the post-election world may favour different kinds of companies: those that benefit from lower energy prices, from the resolution of conflicts in the Middle East, and from a government that spends more on infrastructure and less on technology subsidies. Candriam’s head of multi-asset, Nadege Dufosse, pointed out that the capital rotation was based on assumptions of slowing AI capital expenditure, increased opportunities for resolving conflicts in the Middle East, and falling energy prices. The Dow’s early gain on Thursday, when it opened 322 points higher and briefly rose as much as 354 points before reversing sharply, showed that buyers were still willing to step in for non-technology stocks before the selling overwhelmed them. The question for the coming days is whether the rotation continues or whether the selling spreads to every sector, as it began to do on Thursday afternoon.

6.2 AppLovin surges

The clearest winner of the day was AppLovin, the AI-powered advertising company, which surged 32 percent after delivering better-than-expected results and issuing guidance that beat estimates. AppLovin’s rise showed that the market is not rejecting artificial intelligence itself but the price investors have been paying for it. A company that can show real profits from AI, rather than just promises of future profits, was rewarded handsomely. The contrast with Viavi Solutions, which crashed 18 percent despite strong results, is instructive: AppLovin’s business is advertising, where AI is already generating revenue, while Viavi’s business is selling equipment into a buildout that investors now doubt. The lesson of Thursday may be that the AI trade is splitting in two: companies that make money from AI today are being separated from companies that are spending money on AI for tomorrow. That split, if it continues, could reshape the technology sector for years, with the winners being those that can prove their AI investments pay off and the losers being those that cannot.

7. What analysts and investors are saying

7.1 The case for buying the dip

Not everyone believes the selloff will last. AMD’s chief executive, Lisa Su, said the market is a little short-sighted and that she sees no slowdown in demand for AI chips. Her argument is that the underlying demand for computing power is still growing, that the data-centre buildout is a multi-year programme, and that short-term political noise should not change long-term investment decisions. Some analysts agree, pointing out that the Nasdaq 100 would need another 6.2 percent decline to reach the conventional bear-market threshold of 20 percent below its record, and that previous corrections this year were followed by recoveries. The Motley Fool’s analysis of the selloff asked whether it was time to buy the dip, noting that Nvidia, Dell and the data-centre stocks had fallen hard but that their long-term prospects might be unchanged. The argument for buying is that the AI revolution is real, that the companies leading it are still growing, and that moments of panic are when the best bargains appear. History offers some support: technology stocks have fallen sharply before, in 2022 and again in early 2025, and each time they eventually recovered to new highs. The buyers of the dip are betting that this time will be no different.

7.2 The case for caution

The case for caution is equally strong. The Nasdaq has now entered correction territory, down more than 10 percent from its October 6 record, and the Nasdaq 100 and the semiconductor index have each crossed the 10 percent threshold. The S&P 500’s 4.36 percent fall was its worst since April 2025, and the Dow’s 1,833-point drop was one of the largest in its history. The political risk is unresolved: Congress does not change hands until January, and nobody knows what legislation a Democratic Congress will actually pass. The local backlash is still growing, with more than 500 bans and pauses already in force and more votes to come in December. The spending question is unresolved: JPMorgan’s 900 billion dollar estimate for this year is only the beginning, and the returns are still largely theoretical. And the broader economy is sending mixed signals, with the 10-year Treasury yield near 4.9 percent and the October employment report due on Friday. The cautious view is that the AI trade had become too crowded, too expensive and too dependent on the assumption that nothing would go wrong, and that Thursday was the day the market discovered how much could go wrong at once. The truth will probably lie somewhere between the two views, but for now the sellers are in control.

8. What remains unconfirmed and what to watch next

8.1 The closing numbers and the next sessions

Several things about Thursday’s selloff remain unconfirmed or subject to revision. The closing figures reported here come from market data providers and wire reports published after the close, and different providers showed small differences in intraday figures during the session, so the final numbers should be treated as reported rather than absolute. The Nasdaq 100’s 1.8 trillion dollar decline in market value is a Reuters estimate of the direction of travel rather than a final audited figure. The final certification of the November 3 ballot measures is still pending, and some of the local measures are advisory rather than binding, so their legal effect is not yet settled. The UBS estimate that nearly a third of new United States data-centre capacity expected in 2028 could be delayed or lost is a scenario, not a prediction, and it depends on construction pauses spreading further than they have so far. The next sessions will be watched closely: if the selling continues on Friday, the Nasdaq could move deeper into correction territory, and if buyers step in, Thursday could be remembered as the bottom. The October employment report, due on Friday, could move the market in either direction, and any comment from the Federal Reserve about interest rates will be scrutinised for clues about borrowing costs.

8.2 What to watch next

The weeks ahead will answer the questions that Thursday raised. First, the December votes in Clarksville, Arkansas and North Dakota will show whether the local backlash against data centres is still growing or has peaked. Second, the new Congress will take office in January, and its first moves on AI regulation, data-centre permitting and technology oversight will set the terms for the industry for the next two years. Third, the next round of earnings from Nvidia, AMD, Alphabet, Amazon and the other big AI spenders will show whether the spending is producing the profits that investors are demanding. Fourth, the Federal Reserve’s decisions on interest rates will determine how expensive it is to finance the data-centre buildout. And fifth, the companies themselves will have to decide whether to slow their spending, as some analysts expect, or to press ahead, as AMD’s Lisa Su suggests they will. For now, the market has spoken clearly: the AI boom is no longer a one-way bet, and the companies that built it will have to prove, quarter by quarter, that the hundreds of billions of dollars they are spending are worth it. The coming weeks will provide the answers, and the investors who sold on Thursday will be watching to see whether they were right to leave or wrong to panic.

Comments

comments