Key Takeaways
- • Goldman Sachs Prime data show Asia-focused long/short funds fell 7% in the month, with year-to-date returns dropping from a June high of 40% to 16%, erasing more than half of their gains in four months.
- • TMT long/short funds fell 5% over the month, with year-to-date returns narrowing from 24% in June to 12%; global fundamental long/short funds fell 2% for the month, with year-to-date returns down from 17% to 10%.
- • Global AI stocks have seen net selling for six consecutive trading days, with the six-day cumulative net flow Z-score falling below -3, the most extreme position reduction since ChatGPT's launch according to Goldman Sachs.
- • Goldman Sachs analyst Mark Wilson argues that any AI stock rebound will be a position-adjustment rally rather than a fundamental re-pricing until the sell-off in junk-rated AI assets, led by CoreWeave, stabilizes.
- • CoreWeave shares fell 15% in a single day and its five-year CDS reached an all-time high, shifting market focus from AI demand to credit risk in AI capital structures.
Summary
Position-level pressure has been pronounced. Goldman Sachs Prime book data show global AI stocks (GSXGGLAI) have faced net selling for six consecutive days, with the six-day cumulative net flow Z-score falling below -3. This surpassed the second-fastest sell-off pace of the year set the previous week and coincided with the most severe single-day drawdown for quantitative strategies since August. Goldman Sachs analyst Mark Wilson highlighted a three-and-a-half-year long/short strategy going long "AI winners" and short "AI risk assets," which at one point returned more than 90% this year. Since late June, those gains have nearly halved, and after a failed rebound attempt in September the strategy is retesting recent lows.
Wilson attributed this week's volatility to two triggers. CoreWeave co-founder McBee said in an interview that life insurance companies are providing AI with "unlimited" capital, a remark that prompted investors to reassess private credit pricing and exposed concerns about the weakest link in the AI financing chain. Separately, OpenAI Chief Financial Officer Sarah Friar publicly warned companies to exercise caution when adopting low-cost open-source models, intensifying debate over the economics of frontier AI models. Goldman Sachs credit strategists have labeled bonds of emerging cloud providers and frontier-lab-related AI companies the "epicenter of the financing-cost debate." CoreWeave's stock fell 15% in a single day, and its five-year CDS hit an all-time high, with its July low now a key technical level.
Wilson remains only partly bearish. He expects the second derivative of capital spending and earnings growth to decelerate, but argues prices already reflect that. He cited a Goldman Sachs CFO remark that computing power purchased a year ago can now fetch three to five times its original price. Wilson expects market leadership to shift from AI-spending beneficiaries to the largest AI investors, which he says can fund capex internally. On the bond side, the 10-year U.S. Treasury yield failed to climb further after CPI and PPI releases and reversed sharply, which Wilson said signals weakness not seen before. He cautioned that December WTI futures were approaching $100 per barrel. Goldman Sachs' Ben Snider estimates net supply of U.S. publicly traded equities will post its largest increase since 2000 next year, though Wilson views this as manageable if the market believes equity financing funds productive investment.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.