United States the stock market fell 10% compared to its peak reached last month amid concerns about growing hurdles facing the high-flying artificial intelligence industry.
The fall in the S&P 500 is strong enough for Wall Street to give it a name: a “correction.” Such declines have occurred regularly for more than a century, and market professionals often view them as a potentially healthy erasure of excessive euphoria, which could send stock prices too high if nothing is done to control them.
But the corrections are disconcerting at the time. And the biggest fear that always accompanies a correction is that it could be a harbinger of an upcoming “bear market,” what Wall Street calls a decline of at least 20%.
Here’s a look at what history shows from past corrections and what market watchers expect going forward.
What is behind this fix?
Actions linked to the boom in artificial intelligence pushed markets to all-time highs in October. But a confluence of events has investors wondering whether this enthusiasm is overblown and whether the AI industry can live up to the high expectations.
On the one hand, investors fear that stock prices of AI companies will rise too high and that strong demand for memory, processors and other data center building blocks could falter if AI proves less profitable than promised.
Another concern is the large sums required to invest in technology and data centers. The industry is working to develop better models and infrastructure, requiring billions of dollars, but companies must increasingly raise money by selling bonds to investors. The Federal Reserve is now expected to raise its short-term interest rate next month. This could increase borrowing costs for AI companies, among others.
More recently, investors have become increasingly concerned about the regulatory environment put in place by the new Congress. Newly elected Democrats pledge to crack down on industry amid a backlash against data centers. And two big states offering data center incentives, Texas and Georgia, elected Democratic governors who pledged to stop new construction.
Then there’s Nvidia. The company’s chips power the AI industry and also provide the money that helped companies in the sector to develop. The sudden fall of Jensen Huangits former CEO and co-founder, sanctioned the company’s actions and played a significant role in the selloff.
“Jensen Huang’s departure may have been the catalyst for this selloff, but underlying concerns in the tech sector have been building for some time,” said Karan Ramchandani, managing partner of Post Oak Group. “Investors have poured huge amounts of capital into AI based on expectations of transformational growth, and we are now entering a period where those expectations must translate into measurable profits and returns.”
Not even the Successful market launch of the Anthropic AI laboratory this week could stem sales.
Was there a tech bubble? And did it burst?
Given their high valuations, Wall Street has been anticipating a correction in AI stocks for some time. Some who have called the rise in stocks a bubble believe it has not completely burst.
“The bubble is bursting and there is still a lot to do,” said David Trainer, director of investment research firm New Constructs. “We are in the early stages of coming to terms with the economic reality of blatant overinvestment in AI. Like all other booms, this one is collapsing.”
Others see the selloff as another example of the fear and anxiety that periodically grips markets. Dan Ives, a financial analyst at Yorkville Ives & Co., compared the current downturn to the pandemic-related market corrections of March 2020 and April 2025 following the Trump administration’s escalation of the global trade war. But he sees the selling as driven by panic, not a “moment of alarm indicating darker times ahead for technology and the market.”
“This is a painful technological liquidation that strongly dismisses the thesis of the 4th industrial revolution which will take place over the next few years,” Ives wrote. “And it’s a bad decision in our opinion.”
Ramchandani sees the sale as a reassessment of the AI sector rather than the bursting of a bubble.
“AI is not going away,” she said, “but the market could enter a phase where investors become much more selective about which companies can actually deliver on their promises.”
Mark Hackett, chief market strategist at Nationwide, says the decline caused by AI’s actions could ultimately prove positive for the overall market.
“If this evolves into a real surge, it could serve as a necessary reset to improve positioning, sentiment and the chances of a more sustainable rally by the end of the year,” he said in a research note.
How often do corrections take place?
Every two years, on average. Even during the historic nearly 11-year bull market for U.S. stocks, from March 2009 to February 2020, the S&P 500 stumbled through five corrections, according to research firm CFRA. Concerns about everything from interest rates to trade wars to the European debt crisis have driven these pullbacks.
The last correction in the US market was in March 2025. The last correction that turned into a bear market was in 2022. During the 2022 bear market, the S&P 500 fell 25.4% from January 3 to October 12.
What usually happens after a fall like this?
Looking at the 25 corrections since 1946 that managed to recover before turning into a bear market, the S&P 500 took an average of 130 days to bottom and lost an average of nearly 14.1% along the way, according to CFRA. The index took an average of 112 days to recover its losses.
For the 14 declines that turned into bear markets, the damage is much worse. Since 1946, the average bear market has caused a 32% loss for the S&P 500, according to CFRA.
___
Business writers Bernard Condon, Kaitlyn Huamani and Damian Troise contributed.
Gn bussni
