Whale Rock's flagship fund falls 21.7% in July as AI stock reversal cuts 2026 gains

Whale Rock's flagship fund falls 21.7% in July as AI stock reversal cuts 2026 gains

Year-to-date returns dropped to 35.1% from 72.5% in June as SanDisk, Bloom Energy and CoreWeave sank in a memory-led AI selloff that also hit the firm's long-only portfolio.

Fact Check
Business Insider's report by Bradley Saacks confirms every specific figure in the claim: Whale Rock's flagship fund fell 21.7% in July 2026, cutting YTD returns to 35.1% from 72.5% through June, and the selloff hit SanDisk, Bloom Energy, and CoreWeave while also affecting the firm's long-only portfolio (which fell 18.8%). Bloomberg independently corroborates the 21.7% July drop and characterization as a major AI-selloff loser. Two credible independent primary sources align exactly with the claim.
    Reference12
Summary

Whale Rock Capital Management's flagship hedge fund fell 21.7% in July, cutting its 2026 gain to 35.1% from 72.5% a month earlier as artificial intelligence and semiconductor shares reversed sharply. The Boston-based $19 billion firm run by Alex Sacerdote was hit by losses in SanDisk, Bloom Energy and CoreWeave during a memory-led selloff in chip and AI infrastructure stocks, while its long-only fund dropped 18.8% in July but remained up 36.8% for the year. The retreat added to broader anxiety that markets have become overexposed to a single AI trade, a theme underscored by Leopold Aschenbrenner's Situational Awareness fund losing 67% in July after a forced unwind. Whale Rock's next test is the earnings season, with investor sentiment on AI infrastructure spending likely to determine whether the fund stabilizes or extends its losses.

Terms & Concepts
  • long-only fund: An investment portfolio that holds bought securities without taking short positions against stocks.
  • forced unwind: A rapid sale of holdings driven by margin pressure, funding stress or risk limits rather than routine portfolio changes.
  • AI infrastructure spending: Corporate spending on chips, servers, networking and data-center capacity needed to build and run artificial intelligence systems.