UBS Trading Desk: Last Week’s US Stock Market Performance Resembled the Time of the ‘AI Stock Guru Liquidation’

The US stock market did not experience a broad sell-off last week, but rather a targeted position reset. UBS trading desk trader Torsten Sippel pointed out in his latest weekly report that last week's market movement was highly similar to the deleveraging event triggered months ago by Situational Awareness (a well-known AI quantitative fund)—AI and momentum sectors continued to be under pressure, funds rotated into defensive and low-crowded sectors, but overall risk exposure was not systematically abandoned.

AI and momentum sectors became the epicenter of this adjustment. Semiconductors, storage, and AI infrastructure faced ongoing selling pressure with hardly any capital stepping in to bottom-fish; the storage sector saw a 16% weekly decline, speculative growth stocks fell 12%, and the momentum long-short strategy dropped 8%. Meanwhile, consumer staples rose 5%, the value long-short strategy increased 6%, and the market trend has clearly shifted from growth and crowded momentum towards quality and value.

Hedge funds’ net exposure has dropped to the 30th percentile of the last five years, but total exposure is still above the 90th percentile, a structure that means if P&L pressure continues, the market still faces further deleveraging risk. Sippel explicitly stated in the report that the recent market pattern is "fragile rather than outright bearish," with risk being redistributed rather than abandoned.

Deleveraging, Not De-Risking: The Nature of This Adjustment

Sippel characterizes last week's market action as “De-Grossing rather than De-Risking.”

From a flow perspective, US high-touch trading volume increased significantly, long-only funds had a buy/sell ratio of 36/64, and hedge funds’ buy/sell/short ratio was 42/50/8. Long positions continued to be reduced, but the addition of new single-stock short positions was relatively limited, which does not fit the pattern of complete risk-off behavior.

This pattern is identical to the previous deleveraging event led by “AI Stock God,” former OpenAI researcher Leopold Aschenbrenner's hedge fund Situational Awareness. At that time, crowded positions were compelled to unwind, not systemic selling driven by fundamentals. In this round, software, defensive, and low-crowded sectors absorbed some of the rotating funds mid-week, while consumer, housing, healthcare, and some industrial sectors also saw inflows. Small caps and equal-weight indices outperformed the broader market.

AI Financing Logic Under Scrutiny: CoreWeave as a Key Indicator

CoreWeave has become an important indicator for observing AI trades. Sippel pointed out that as credit concerns around the company continue to grow, market debate about AI has shifted from demand-side issues to more fundamental questions: Can the economics and financing structures of AI infrastructure development support the next phase of capital expenditure expansion?

It is worth noting that both ASML and TSMC released strong fundamental data, but this has not alleviated the above concerns. This divergence is significant—meaning that even with robust hardware demand data, the market has developed a distinct pricing logic for the sustainability of AI capital expenditure financing.

Against this backdrop, upcoming earnings reports from the hyperscalers will face a higher bar for validation—investors want to see that cloud revenue, backlog, and cash generation can truly support the scale and financing needs of AI capital expenditure.

Macro Factors Amplified but Not the Trigger for Deleveraging

Sippel emphasized that macro factors played the role of amplifier rather than trigger in this round of adjustment.

Inflation and interest rate shocks accelerated systematic selling, while Middle East tensions have increasingly transmitted into the volatility market. The VIX spiked on Friday, and the 1M/3M VIX term structure inverted for the first time in six months, indicating that hedging demand has spread from AI-specific risks to broader market risks.

Friday afternoon’s market weakness is especially worth noting—the sell-off in the tech sector was indiscriminate. Sippel believes this aligns more with ETF redemptions, options expiry, and hedge rolling rather than incremental sales based on individual stock fundamentals.

Earnings Season Has Yet to Become a Dominant Factor

The overall performance of the current earnings season is acceptable, but its impact on market trends remains limited. 10.4% of S&P 500 market cap has reported, Q3 expected revenue growth is 12.1%, EPS growth is 26.5%, actual profit beats by 4.1%, and 75% of companies have exceeded expectations.

However, strong results from the banking sector and positive AI-related data have not efficiently absorbed the ongoing selling pressure, so earnings remain subordinate to positioning logic for now.

In the next five trading days, 80 companies will report earnings, accounting for a combined 13.8% of S&P 500 market cap, including Netflix, General Electric, Meta Platforms, Tesla, Procter & Gamble, and Western Digital. Sippel specifically noted that Meta is reporting a week earlier than usual and rhetorically asked:

“Are they getting ahead of the (issuance) window closing?”

In summary, Sippel’s judgment on the recent market is "fragile rather than outright bearish."

Risk is being redistributed, not abandoned. However, with genuine bottom-fishing capital lacking in the AI sector, total exposure still high, and the volatility curve under fresh pressure, there is still room for further deleveraging. In addition, the recent rise in volatility has sparked a new round of questions about the sustainability of the capital market issuance window.