AI Stocks Lose $5 Trillion in 3 Weeks as Hedge Funds Go Underweight AI for the First Time Since 2023
Hedge funds just staged their biggest retreat from stocks since at least 2017, and Goldman Sachs says the unwind looks closer to its start than its end. What happens when NVIDIA reports earnings could determine whether a $5 trillion wipeout…
NVIDIA (NASDAQ:NVDA | NVDA Price Prediction) closed November 2 at $202.06. That was down 1.26% for the day and 15.54% below its October 6 close of $239.24, which was its highest close of the past year. Invesco QQQ Trust (NASDAQ:QQQ) fell 1.36% to $681.09, which leaves it 10.34% under its October 6 high of $759.66.
Yahoo Finance counted 47 AI and chip stocks that have lost about $5 trillion in market value since October 6. The biggest losers in Yahoo Finance’s count were NVIDIA ($835 billion), Alphabet (NASDAQ:GOOGL) ($835 billion), SpaceX (NASDAQ:SPCX) ($505 billion) and TSMC (NYSE:TSM) ($395 billion), while the 625 other large U.S. stocks in the same count added about $1.3 trillion.
Morgan Stanley runs a prime brokerage desk, which finances and clears trades for hedge funds. Morgan Stanley’s data reportedly show funds underweight the broader AI trade for the first time since early 2023, after they cut total stock exposure by more than 8% since June, the biggest retreat since at least 2017, and much of that positioning shift is likely still ahead.
Funds Sold Their Holdings Far Faster Than They Shorted
Goldman Sachs data show hedge funds sold AI stocks for five straight days in October, reversing every AI purchase made since the start of 2025. On the biggest day, per Goldman Sachs, they sold shares they owned about 3.5 times faster than they opened new short positions.
A short position profits when a stock falls. Selling existing holdings at that pace looks like de-risking to protect gains more than a bearish call on AI demand.
Underweight Overall but Crowded at the Top
Underweight means funds hold less of the broad AI trade than their benchmarks do. That usually helps a sell-off end, because fewer holders are left to sell.
Goldman Sachs also said the unwind looks closer to its start than its end. Goldman Sachs data show funds’ AI holdings still rank in the top 17% of the past three years, and the Magnificent Seven were net sold in four of the last five trading sessions. A Nomura note put hedge funds’ sensitivity to those seven stocks at the 97th percentile as of October 26.
NVIDIA was QQQ’s largest holding in the fund’s June 30 portfolio filing. That is why the fund fell less than NVIDIA but still fell sharply.
Because the rest of the market rose, money moved around within stocks. After hours on November 2, NVIDIA traded at $203.80 and QQQ at $682.58, slightly above their closes.
Household Wealth Ties the Sell-Off to Spending
Federal Reserve data going back to 1945 show stocks made up a record 48% of household and nonprofit financial assets in the second quarter, up from 39% at the dot-com peak. The richest 10% of households own 88% of stocks.
Those households do much of the country’s spending. That is why Bank of America economists called a sustained correction in the AI trade “one of the larger risks to the US and global economy.”
What NVIDIA’s Next Earnings Report Could Mean for the Sell-Off
NVIDIA’s fundamentals look stronger than its recent price action. Second-quarter revenue of $96.22 billion beat the $92.07 billion consensus, and the company guided third-quarter revenue to $108 billion, plus or minus 2%.
NVIDIA is still worth $4.88 trillion, but at roughly 13 times forward earnings, the stock already prices in much of the positioning risk. Crowding in the mega caps is still high, so a weak report on November 17, a date NVIDIA has not yet confirmed, could set off another round of selling.
Even a beat may not end the selling, because if NVIDIA beats estimates and still trades below its November 2 close of $202.06, bearish sentiment and broader market forces can drive the stock lower before a recovery.
Contact [email protected] for any questions or corrections.
Omor Ibne Ehsan is a writer at 24/7 Wall St. He is a self-taught investor with a focus on growth, cyclical, and dividend equities that have strong fundamentals, value, and long-term potential. He also has an interest in high-risk, high-reward investments such as penny stocks.
