Speculative themes are being thrown out! Goldman Sachs trading desk: Thursday's momentum trading in US stocks saw the largest decline since the DeepSeek impact

Concerns about the massive financing needs of the AI cycle are brewing, leading to a significant sell-off in U.S. tech stocks, a sharp decline in momentum trading strategies, and fierce selling in speculative sectors such as AI-related thematic baskets and Bitcoin-sensitive stocks.

The Nasdaq 100 index fell more than 2% on Thursday, closing at the day's low, with five out of the last six trading days showing declines. This sell-off has heavily impacted momentum trading strategies and AI-related assets. Although the Nasdaq 100 index is only about 5% away from its all-time high and is attempting to hold the critical support level of the 50-day moving average, market sentiment has clearly shifted towards defense.

Goldman Sachs data shows that its high-beta momentum pair trading (GSPRHIMO) plummeted 7% on Thursday, marking the second worst performance of the year and the largest single-day drop since the DeepSeek incident. AI-related thematic baskets, Bitcoin-sensitive stocks, and speculative sectors like quantum computing have all faced fierce selling.

Peter Callahan, a Goldman Sachs expert in technology, media, and telecommunications (TMT), stated that the trading desk has received numerous inquiries about "why the drop" and "what the catalysts are." Goldman Sachs attributes this volatility to a resonance of multiple factors: investors taking profits ahead of Nvidia's earnings report, a rise in hawkish comments from the Federal Reserve, and increasing doubts about the return on investment in AI infrastructure.

With only 30 full trading days left in the year, investors are waiting for market narratives and price trends to stabilize. Goldman Sachs points out that momentum exposure in the prime brokerage book remains high (76th percentile for one year, 88th percentile for five years), while seasonal pressures from year-end position reductions and tax-loss harvesting are becoming apparent. Currently, investors are awaiting clearer signals—whether Nvidia's earnings report will exceed expectations or whether the Federal Reserve's monetary policy path will clarify—to determine when this round of selling will stabilize.

Five Major Pressure Factors Triggering the Sell-off

Goldman Sachs' trading team has identified five major triggers for this decline.

First, investors are choosing to take profits and reduce holdings in AI leading stocks ahead of Nvidia's earnings report next week. Second, a report by the Financial Times regarding "inflated" power demand in data centers has raised concerns about an investment bubble in AI infrastructure.

Hawkish statements from Federal Reserve officials further dampened market sentiment. Comments from Boston Fed President Collins and Minneapolis Fed President Kashkari weakened expectations for a rate cut in December In addition, the news of layoffs and corporate cost-cutting announced by Verizon, along with the potential volatility from intensive economic data ahead of the Federal Reserve's decisions in the next three weeks, has increased market uncertainty.

Goldman Sachs TMT sector expert Callahan stated that the macro backdrop has continued to deteriorate over the past few months—from the performance of internet companies during the third quarter earnings season, to the non-linear characteristics of AI development momentum, to the signs of fatigue in the price trends of market-leading sectors (AI, large tech stocks), multiple factors have left investors in a state of "confusion."

Momentum Trading Strategies Highly Linked to AI Narrative

Goldman Sachs warned last Friday that the sell-off of high beta (GSP1BETA) and residual volatility (GSP1RSVL) could evolve into a momentum strategy unwinding wave before the end of the year.

Thursday's sharp decline confirmed this prediction. Data shows that the momentum factor is currently significantly correlated with high short ratios, high residual volatility, and high beta, while its correlation with high-quality factors is far below normal levels.

More notably, the long side of high beta momentum pairs trading, which outperformed the short side and the market, is precisely the disaster area. This is directly related to the rising correlation between momentum strategies and the AI narrative.

Goldman Sachs data shows that AI beneficiary stocks relative to the S&P 500, excluding the "Magnificent Seven" tech giants, have fallen by 9%, whereas similar pullbacks in the past typically reached about 20%.

Goldman's unconstrained momentum strategy (GSPRHIMO) is currently clearly biased towards high beta, high residual volatility, and cyclical sectors, going long on information technology and industrial sectors while shorting healthcare. This structure makes the strategy particularly vulnerable in this sell-off.

AI Skepticism Intensifies Speculative Theme Decline

AI-related sectors and the most speculative areas of the market experienced a widespread decline on Thursday. Bitcoin-sensitive stocks and themes highly correlated with momentum factors, such as quantum computing, plummeted significantly.

Goldman Sachs pointed out that skepticism towards AI is rising—factors such as Oracle's credit default swaps (CDS) widening and increased leverage, as well as SoftBank selling Nvidia shares, are impacting the bank's AI thematic basket.

The performance of momentum strategies was previously driven mainly by the long side, while the short side's performance was in line with the market. However, as the correlation between the AI narrative and momentum factors approaches historical highs, this structural exposure has now become a source of risk.

Goldman Sachs' trading team stated that Nvidia's earnings report next week will be an important catalyst. Prior to this, hedging momentum exposure is relatively cautious. The bank's momentum excluding AI basket (GSPUMOXX) can provide investors with a factor hedging tool while avoiding a direct short bias against AI