Nasdaq Loses $2.5 Trillion In Wipe Out As "AI Trade" Cracks

This time is different, all right, but not exactly in the way the Silicon Valley-AI hype machine meant. 

Like the blue tsunami that annihilated the Trump-MAGA coalition on Tuesday - and to which it is connected - the equity market wipe out had many different causes, none of which should have come as a surprise as they had been detailed for months: unrealistic financial projections, lack of discernible operating and economic improvements, slowing corporate uptake, dramatic increases in debt, refusal to acknowledge practical impediments to data center buildouts (aside from the rising political opposition) and the growing dependence of the economy and the markets on a shrinking core of AI companies. Markets are up slightly this morning, so this may either be a one-day freakout or a harbinger of more fundamental change, but either way, it is another warning that the prolonged AI trade is not likely to continue in the same way. JL 

David Uberti and colleagues report in the Wall Street Journal:

Thursday's stock-market wipeout left few corners of the tech sector unscathed. Asian chip makers, Texas power producers, New Jersey cloud-computing providers, Silicon Valley software makers, contractors and data-center developers all tumbled, in a selloff that appeared to feed on itself. Traders making highly leveraged bets got crushed, but so did those who doubted a recent pullback in AI stocks would get worse. It did, from the opening bell. cracks began to show in risky corners of the AI trade where companies amassed huge piles of debt to grow. Underwhelming earnings reports by Amazon and Alphabet sent tremors across trading desks. Then, the power shift in Washington, as well as the election of governors who promised data-center restrictions, exposed the stock market’s fragility. The yearslong bull market left the major U.S. benchmark indexes top-heavy and tech-dependent. The unwind of bullish bets using borrowed money compounded the pressure. 

decrease; down pointing triangleand Alphabet GOOGL -2.20%decrease; down pointing triangle sent tremors across trading desks.the power shift in Washington, as well as the election of several governors who have promised data-center restrictions, exposed the stock market’s fragility. The yearslong bull market left the major U.S. benchmark indexes top-heavy and tech-dependent.

Two days after the midterm elections swept a bloc of artificial-intelligence skeptics into power, the new reality slammed into Wall Street.

Thursday’s stock-market wipeout left few corners of the tech sector unscathed. Asian chip makers, Texas power producers, New Jersey cloud-computing providers, Silicon Valley software makers, contractors and data-center developers all tumbled, often in unison, in a selloff that appeared to feed on itself as the day wore on.

Traders making highly leveraged bets got crushed, but so did those who doubted a recent pullback in AI stocks would get worse. It did, from the opening bell.

“It seems like the rug is getting pulled out from under it,” said Jonathan Corpina, senior managing partner at Meridian Equity Partners. “The market continues to be extremely fragile with all this uncertainty in front of us.” 

By Thursday afternoon, some $1 trillion of the Magnificent Seven group of tech giants’ collective market value had vanished. The PHLX Semiconductor index posted its worst day since the depths of the pandemic. The Nasdaq composite stumbled into correction territory, shedding an estimated $2.5 trillion in total market value. That was the biggest drop in the benchmark’s history. And the poster child of the AI boom, Nvidia NVDA -9.26%decrease; down pointing triangle, closed the session in the red—for 2026. 

It was only a few weeks ago when AI stocks were still propelling the markets to record highs, with investors looking expectantly toward the planned initial public offerings of Anthropic and OpenAI.

Then, the pressure started to build. 

First, cracks began to show in risky corners of the AI trade where companies amassed huge piles of debt to grow. Underwhelming earnings reports by Amazon AMZN -2.19%decrease; down pointing triangle and Alphabet GOOGL -2.20%decrease; down pointing triangle sent tremors across trading desks. 

Now, the power shift in Washington, as well as the election of several governors who have promised data-center restrictions, exposed the stock market’s fragility. The yearslong bull market left the major U.S. benchmark indexes top-heavy and tech-dependent.

Thursday’s selloff broadened to drag every sector of the S&P 500 into the red, a departure from recent sessions when weakness in AI stocks spurred a large rotation into other areas of the market. The Nasdaq composite’s tumble marked its worst day since President Trump’s tariff announcements last year sent markets haywire. 

A data center under construction in Georgia. Dustin Chambers for WSJ

“When tech is leading the way, it’s very easy to get on board,” Corpina said. “When tech gets hit, it certainly makes investors a lot more uneasy.”

On Wednesday, the prospect of a slowdown in construction battered chip makers, contractors and other infrastructure firms. But gains in the shares of the biggest hyperscalers—whose massive spending plans could shrink with a pullback—helped push indexes higher, with the Dow Jones Industrial Average logging its biggest gain in almost 19 months.

That bifurcation between the biggest AI players and their suppliers continued on Thursday. While Microsoft, Amazon and Alphabet traded in the red, they sidestepped most of the pain coursing through the AI supply chain. 

There was one exception among the hyperscalers, though: SpaceX, which is teaming up with Tesla to build its massive Terafab chip factory in Texas to help support Elon Musk’s stratospheric ambitions. Shares in both firms fell by more than 9%, suggesting traders were particularly worried about Musk-owned companies given the tech titan’s close ties to the Trump administration. 

Investors had rested their hopes that the AI-fueled rally would extend into 2027 on the premise that firms with such capacity could continue printing blockbuster profits. The “picks and shovels” plays at the heart of the AI trade—companies like Micron and Nvidia—powered much of that eye-popping profit jump.

A change in the political backdrop “matters because AI infrastructure names, especially semiconductors and memory, are the key drivers of S&P 500 earnings growth,” said Mike O’Rourke, chief market strategist at JonesTrading. “Due to the expected policy shift, the earnings forecasts for next year can no longer be trusted.”

While few expect new data-center projects to grind to a halt, any future delays could have big consequences for a crucial part of the U.S. economy. 

“If you’re investing in a project and you think you’re six to 12 months away from generating revenue, but you’re really two to three years away, you’ve got a number of problems,” said Matt Rowe, a senior portfolio manager at hedge-fund giant Man Group. 

In Texas, which has more planned data-center projects than any other state, Wall Street expected that Republican Gov. Greg Abbott’s recently announced data-center moratorium was largely for show ahead of the election and would be reversed. 

Now, investors and analysts are rushing to assess what a longer pause on new projects could look like. Democratic Gov.-elect Gina Hinojosa has called for new approvals to remain on pause until the Texas Legislature can meet next year to pass “real protections.”

“I don’t think anybody in July or August thought that it would get this dicey,” said Mark Melchiorre, chief investment officer of credit-focused hedge fund Forza Investment Group. “Some projects will stall.”

Investors said Wall Street’s delayed reaction to the election stems in part from a steep decline in oil prices Wednesday after Trump and Vice President JD Vance acknowledged the war with Iran had dimmed Republicans’ political fortunes. That momentarily buoyed investors’ hopes that inflation eating into Americans’ paychecks would ease.

But a partial rebound in crude costs Thursday opened the door wider for a selloff. Late in the afternoon, individual investors were on pace for their biggest day of net stock sales since the Covid-19 pandemic, according to Vanda Analytics. Mechanical selling amplified the losses, said Ben Emons, founder and chief investment officer at FedWatch Advisors. 

Early in the session, Emons said, options traders scrambling to cover their positions helped spark indexes’ sharp declines. As losses mounted, systematic hedge funds piled in and juiced the downward momentum. 

The unwind of bullish bets using borrowed money compounded the pressure. Traders who piled into triple-leveraged exchange-traded funds or indexes linked to chips enjoyed magnified gains during the AI-driven rally earlier this year. But the dynamics reverse when prices fall. The Direxion Daily Semiconductor Bull 3X ETF, which seeks 300% of the daily performance of the NYSE Semiconductor Index, plummeted 39% Thursday. 

“When that momentum starts to go against investors, particularly when there’s leverage involved, it can lead to what might seem like extreme selloffs,” said Mike Bell, head of market strategy at RBC BlueBay Asset Management. “Sometimes markets go down because people sell, and sometimes people sell because markets are going down.”