Wall Street's surge to all-time highs this year amid war in the Middle East and soaring financing costs was driven by a single investment thesis — the relentless rise in stocks driven by the boom in artificial intelligence.
After weeks of warning signals from credit markets, this dominant narrative imploded spectacularly on Thursday (5), in a $3.2 trillion debacle triggered by the political fallout from the US midterm elections.
According to investors, the wave of sales was fueled by the brutal dismantling of leveraged positions in technology accumulated by hedge funds, which seek returns through flexible investment strategies, and by the exit of individual investors from the market.
"It's like a chemical reaction," said one stock executive. "You have one, two, three elements, and then there's an explosion. The election was the explosive element."
The technology-heavy Nasdaq 100 fell more than 6% on Thursday in one of the worst days on record for chipmaker stocks, before rising 1.4% on Friday and ending the week down 4.9%.
The significant drop in the technology sector shocked investors. Stocks had initially risen on Wednesday following the Democratic "blue wave" in Congress, as investors unwound their pre-election protections in the face of an undisputed result and returned to buying shares. This optimism was fueled by a sharp drop in oil prices as hopes rose that the war with Iran would come to an end.
"While the risks surrounding both Iran and AI funding are likely to remain present, we believe investors have sufficiently reduced their risk exposure and that stocks can move higher," Citi said in a note on Wednesday, summarizing the market mood.
But the lull was broken on Thursday when oil prices rose again and the market focused on the possibility that newly elected Democrats would slow infrastructure development for the $1 trillion in AI investment that some expect next year.
"The midterm elections have put an end to the assumption that there would be a blank check for AI," said Ludovic Subran, chief investment officer at Germany's Allianz. “The market has just discovered that ‘the sky is the limit’ is not a valuation methodology.”
This year, a stock market heavily focused on technology had benefited from a dual effect of the AI boom on share prices: it boosted profits at companies that supply key technology-related inputs, such as chipmakers, and also supported economic growth prospects.
Now, investors face the risk that these two forces will go into reverse and that markets will have to price in an impact on U.S. economic growth, just as delays and regulations squeeze the profit prospects of companies directly involved in AI infrastructure.
“The AI investment thesis has been undervalued,” said Nikhil Srinivasan of Singapore-based Benzaiten Macro. Stocks "price the future, and the future beyond the next few years is now anyone's guess," he said. He predicts more moderate growth in the US next year, due to the slowdown in investment in AI.
The election of Democratic governors who are expected to restrict the construction of data centers in states like Texas, coupled with the possibility of a broader backlash against AI in a Democratic-controlled Congress, was enough to shake up a market already worried about the effect of higher financing costs on hyperscaler spending.
The third-quarter earnings season had already generated volatility: Tech giants detailed huge investment plans but also weak revenues in their core businesses. The collapse of debt securities from risky AI-linked borrowers such as neoclouds has fueled fears of a credit reckoning.
Now, managers fear that the expansion of data centers, already included in profit estimates, is in question. "What we're most concerned about is that executing all of this [capital investment] is going to be more complicated with... more delays and all these restrictions," said Stephen Jones of Aegon Asset Management.
Traders said one of the triggers for Thursday's selloff was individual investors, who in previous bouts of volatility had shown themselves to be reliable buyers on dips. According to Vanda Research, they had net outflows of US$388 million in individual shares in the first two hours, which put the trading session on track for the group's biggest day of sales since Covid-19, before they repurchased part of the shares at the end of the day.
As stocks fell, hedge funds had to "reduce gross exposure", that is, close leveraged positions, which intensified sales. "Fear [was] in charge. It's classic human psychology," said a senior trader at a major bank. “People need to raise cash because they lost money on the AI thesis.”
JPMorgan reported that Thursday was the biggest daily loss for hedge funds since the "Liberation Day" fees, with a 1.6% drop among its hedge fund clients.
As technology stocks rallied on Friday, some investors said the political risks were overblown. "This was a vote on data centers. It wasn't a condemnation of AI," said Cresset's Joe Zidle, who says the restrictions will be a medium to long-term problem, as "there's a lot of work going on and it's unlikely to be stopped."
Others, however, see the turbulence as a sign that the AI boom has reached its peak. “I think the market will go down from here,” Srinivasan said. "It could go up this week, next week... but it will still go down further."
