Live · updated 8 minutes ago · 135 stories in 24h

← Stream
Hedge Funds Unverified First seen

AI-focused hedge fund VARA falls 33% in October tech selloff

Bloomberg reported, per Hedgeweek, that Value Aligned Research Advisors (VARA), a hedge fund heavily exposed to AI stocks, lost 33% in October amid a technology selloff. The Princeton, New Jersey-based manager told Class C shareholders the fund remained up 13% year-to-date despite the drawdown, following a 44% loss in July. The post also states the Nasdaq 100 fell 6.7% in October and the Philadelphia Semiconductor Index dropped 12% on a Thursday.

Why “Unverified”

  • Reported by Hedgeweek alone, with no second source or primary confirmation yet

1 independent source group across 1 post · evidence score 48/100. How labels work

Market around the story

  • NDX Still being measured. The price reaction is recorded once the market has traded for six hours after the first report.

What the sources claim

  1. VARA, an AI-focused hedge fund, suffered a 33% loss in October as tech shares sold off.

    “suffered a 33% loss in October as a sharp selloff in technology shares”
    Hedgeweek
  2. VARA told Class C investors its fund was still up 13% year-to-date despite the October drawdown.

    “its fund remained up 13% year-to-date despite the October drawdown”
    Hedgeweek
  3. The October decline followed a 44% loss for the fund in July.

    “The latest decline follows a 44% loss in July”
    Hedgeweek
  4. The Nasdaq 100 fell 6.7% in October, its steepest monthly decline in more than 18 months. NDX

    “The Nasdaq 100 fell 6.7% in October, its steepest monthly decline in more than 18 months”
    Hedgeweek
  5. The Philadelphia Semiconductor Index dropped 12% on Thursday as chipmaker selling intensified.

    “the Philadelphia Semiconductor Index dropped 12% on Thursday as selling pressure intensified across chipmakers”
    Hedgeweek

Sources

More on NDX

All NDX stories →

Summary and claims extracted by AI with prompt analyze-story@5 on 2026-11-06 10:20 UTC. Labels are computed from the sources, not by the model.