AI selloff drives quant funds' worst performance since August

By Nell Mackenzie and Amanda Cooper

LONDON, July 9 (Reuters) - Some hedge fund managers have in recent weeks posted their worst trading results in almost a year, as many got caught in crowded trades amid highly volatile markets, according to Goldman Sachs.

Goldman said in a note dated Wednesday that systematic managers — sometimes called quant funds — that use algorithms to trade market trends, have given back a quarter of their year-to-date returns. Returns for this group of traders are now up 10.8% for the year, down from a return of 14.4% on June 22. 

Losses came from bets against some of the biggest and most crowded parts of the market right now — U.S. equities, Asian developed-market stocks and, to a lesser extent, Europe, the note said.

Huge volatility in shares of chipmakers had made for a tricky trading environment anyway in late June and into early July. But hefty levels of leverage among retail investors in Korean markets in particular amplified a lot of the share-price moves.  

Quant funds made up roughly 10% of the largest hedge funds in 2025, according to data from S&P Global. 

Regulators, including those at the Bank of England, the Bank of Japan and the Bank for International Settlements, have been warning for some time about lofty valuations, particularly in the tech sector where shares in companies like Micron Technology, Intel or Marvell Technology have risen by around 200% in 2026 alone.

And they've voiced concern over how the growing role hedge funds play in financial markets adds to volatility and risk.

Goldman said fundamental managers, or stockpicking funds, were down 2.2% in the same period, having been caught up in crowded tech sector trades. But this group, it said, was still up 15.5% this year.

These stockpickers have "aggressively" fled trades related to AI, most of which had previously driven winning trading positions, Goldman said. 

This mass exit has brought hedge fund leverage to its lowest levels of the last year, a sign of the scale of their trading activity.

(Reporting by Amanda Cooper; Editing by Elisa Martinuzzi and Joe Bavier)