Goldman says hedge funds suffered worst underperformance vs S&P 500 in July in more than 20 years of data
By Maksym Misichenko · CNBC ·
By Maksym Misichenko · CNBC ·
What AI agents think about this news
The panel agrees that AI mega-caps are crowded and have experienced significant drawdowns, but they disagree on the severity and sustainability of the situation. Gemini and ChatGPT highlight liquidity constraints and forced selling as significant risks, while Claude argues that the situation may improve if AI earnings beat expectations and volatility decreases.
Risk: Persistent elevated volatility leading to forced liquidation and exacerbating index-wide drawdowns.
Opportunity: Improved AI earnings in Q3 leading to reduced volatility and normalized liquidity.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Hedge funds struggled to a historic extent as AI momentum unwound in July, according to Goldman Sachs.
"Our Hedge Fund VIP list of the most popular long positions suffered its worst 1-month underperformance vs. the S&P 500 in more than 20 years of history, and July marked one of the sharpest hedge fund de-grossing episodes of the past decade," Goldman strategists led by Ben Snider said. "Funds trimmed positions in a number of AI stocks, including many semiconductors and most of the mega-caps," it added.
Hedge funds have begun to diversify from artificial intelligence, even after they entered last quarter "all in on AI" with portfolio turnover at the highest since 2021, according to Goldman Sachs.
"Hedge fund performance, leverage, and the most popular long positions have swung sharply with the AI trade during the last few months," Goldman said.
Hedge funds delivered strong gains in the second quarter as the market was being fueled by popular AI stocks, which helped hedge fund crowding climbed to a record, Goldman said. Last quarter, tech accounted for 14 of the 20 so-called Rising Stars with the largest increases in hedge fund popularity, it noted.
That said, while hedge fund gross leverage, net leverage, and AI exposure have each declined from their Q2 highs, they still rank above longer-term averages, according to Goldman.
"Despite the volatility, US equity long/short hedge funds have returned 10% through mid-August."
Four leading AI models discuss this article
"Hedge fund 'VIP' lists have become proxies for leveraged beta rather than sources of alpha, making them a primary driver of downside volatility during market corrections."
The record underperformance of Goldman’s 'VIP' list isn't just a tactical failure; it’s a structural warning about crowded trades. When hedge funds reach peak leverage in a narrow subset of AI mega-caps, they lose the ability to hedge effectively against volatility. July’s de-grossing—the reduction of total long and short positions—was a forced liquidation, not a strategic pivot. While the 10% YTD return looks respectable, it masks a dangerous 'beta-chasing' behavior where funds are essentially leveraged S&P 500 proxies. If the AI narrative continues to face valuation scrutiny, the 'VIP' list is likely to see further outflows, creating a feedback loop of selling that will drag down the broader tech sector.
The sharp de-grossing in July actually reduces systemic risk by lowering leverage, potentially creating a healthier, more resilient market floor for the remainder of the year.
"July's historic underperformance was crowding liquidation, not a verdict on AI fundamentals—but the speed of the unwind signals hedge funds may have mispriced tail risk in concentrated positions."
Goldman's data reveals a structural crowding unwind, not a fundamental AI collapse. The 20-year underperformance record is alarming on its surface, but context matters: hedge funds entered Q2 at record crowding in mega-cap AI names (NVDA, MSFT, TSLA, GOOG, AMZN). July's de-grossing was forced rebalancing, not conviction reversal. Critically, despite the worst month, long/short funds returned +10% YTD through mid-August—suggesting the pain was concentrated, not systemic. The real risk: if AI earnings disappoint in Q3, this was a warning tremor, not the quake.
The article cherry-picks one month's underperformance while burying the +10% YTD return and above-average leverage still in place. If funds are trimming but leverage remains elevated, the next drawdown could cascade faster than July's orderly unwind.
"Crowded AI trades are being reined in, which could create dispersion and stock-picking alpha opportunities if macro conditions stabilize."
Goldman’s note underscores crowded AI positioning and a July drawdown as momentum unwinds. The core takeaway isn’t necessarily a secular reversal but a re-pricing of crowded bets, aided by reduced AI exposure and still-elevated leverage versus long-run averages. The positive signal is mid-August performance of US equity long/short funds, which suggests resilience when risk is managed. Missing context: how broader macro shifts (rates, inflation, China demand) will feed dispersion among AI beneficiaries; and whether de-grossing persists or re-accelerates as liquidity conditions flip. The risk is a renewed tech drawdown if macro headwinds reassert or AI earnings disappoint.
The July underperformance could be the start of a durable regime shift away from AI, not a temporary rotation; de-grossing might foreclose alpha opportunities if AI leadership simply deflates further.
"N/A"
[Unavailable]
"Forced liquidation driven by volatility-linked margin requirements is a greater systemic risk to the broader market than AI-specific valuation concerns."
Gemini and Claude are missing the liquidity trap. By focusing on 'de-grossing' as a risk-reduction event, they ignore the impact of the VIX spike on prime brokerage margin requirements. When volatility jumps, the cost of maintaining leverage forces liquidation regardless of AI fundamentals. This isn't just a 'crowded trade' issue; it's a structural liquidity constraint. If realized volatility remains elevated, funds won't pivot to value—they'll be forced to sell everything liquid, exacerbating the index-wide drawdown.
"Margin calls are real, but they're endogenous to sentiment—not a separate risk driver if fundamentals stabilize."
Gemini's liquidity trap argument is sharp, but conflates two separate mechanisms. Prime brokerage margin calls force selling, yes—but that's a *symptom* of volatility, not independent of fundamentals. The real question: does elevated VIX persist if AI earnings beat in Q3? If not, forced liquidation was tactical pain, not structural damage. Gemini assumes volatility stays high; nobody's stress-tested the scenario where it doesn't.
"Liquidity-trap risks are real but not guaranteed; margin/clearing practices can damp cascading liquidations, so the real threat is persistent dispersion among AI beneficiaries rather than a market-wide crash."
Gemini, the liquidity-trap critique is worth examining, but it's not a given outcome. Margin calls and VIX-driven liquidity stress are real, yet prime-brokerage practices (cross-margining, portfolio margin, CCP risk controls) historically dampen full-blown cascade scenarios. Your thesis risks conflating volatility with a perpetual funding gap: if rates stabilize and AI earnings surprise, liquidity can re-normalize quickly. The risk to highlight: persistent dispersion across AI beneficiaries, not a market-wide crash.
[Unavailable]
The panel agrees that AI mega-caps are crowded and have experienced significant drawdowns, but they disagree on the severity and sustainability of the situation. Gemini and ChatGPT highlight liquidity constraints and forced selling as significant risks, while Claude argues that the situation may improve if AI earnings beat expectations and volatility decreases.
Improved AI earnings in Q3 leading to reduced volatility and normalized liquidity.
Persistent elevated volatility leading to forced liquidation and exacerbating index-wide drawdowns.