Goldman Sachs: AI Trade Congestion Drops to One-Year Low, Deleveraging Nears End, Market Returns to Fundamentals

Following a sharp correction in global AI-related stocks, Goldman Sachs' latest research indicates that the most crowded AI trades have rapidly completed deleveraging, with technology stock trading congestion dropping to a near one-year low. This suggests the market's dominant logic is shifting from past 'position博弈' to 'fundamental validation,' with corporate earnings and AI investment returns becoming key factors influencing stock prices.

Lee Coppersmith, Goldman Sachs' Chief Trader, stated that the view of 'excessive position congestion' as the biggest risk in the AI sector is rapidly fading after this round of adjustment. According to Goldman Sachs Prime Brokerage data, last Friday and this Monday saw the largest two-day long liquidation in the global information technology sector since January 2021, and the second-largest concentrated deleveraging event in the past 10 years.

The report noted that the global memory chip sector simultaneously experienced the most severe sell-off on record, and capital flows into tech stocks have fallen into one of the weakest historical ranges, indicating the market is rapidly digesting previously accumulated leveraged positions.

Goldman Sachs stated this risk release ranks among the most intense deleveraging events in recent years. On the same day, the three major mainstream investment strategies—hedge fund long/short, macro, and equity long/short—all declined more than 1%, a situation last seen during the global market plunge triggered by the COVID-19 pandemic in March 2020.

Additionally, the net exposure and long/short ratio of the global 'Magnificent Seven' tech stocks are now at near one-year lows, around the historical 3rd percentile, with previously high mid-term momentum positions having rapidly declined.

Goldman Sachs pointed out that historical experience shows that after each crowded trade unwinds, momentum strategies typically experience a reverse overshoot. The current position cleanup is among the most severe in history, suggesting large-scale passive liquidations may be nearing an end. Currently, the volatility of the momentum factor is about nine times that of the broader market, a level only previously seen in Q4 2020 during the U.S. presidential election and the rollout of COVID vaccines, which triggered style rotation.

The South Korean market has become the epicenter of this AI deleveraging. Goldman Sachs noted that the KOSPI 200 index plunged 46% from its peak in just 27 trading days. Although SK Hynix's latest earnings failed to restore market confidence and instead triggered a new wave of selling, the index has recently found support near its 200-day moving average, and the 14-day RSI is approaching 30, signaling oversold technical conditions.

The market's sharp volatility has drawn attention from South Korean regulators. Market sources indicate South Korea is considering further tightening leverage trading restrictions. Rich Privorotsky, Head of Goldman Sachs One-Delta, believes that if such measures are implemented, they will help reduce market volatility and enhance market resilience in the long term.

However, Privorotsky maintains a cautious view on the memory chip sector. He stated that even after the sharp correction, any rebound is likely just a technical bounce in trading, and future trends will still depend on individual company fundamentals rather than sector-wide momentum.

Goldman Sachs also noted that this correction has been concentrated in AI-related assets and has not evolved into a systemic market risk. In contrast, the S&P 500 Equal Weight Index recently hit new highs, indicating that capital has not broadly exited the stock market but has instead focused on adjusting AI and semiconductor-related positions. This resembles a high-concentration trade position clearance rather than a systemic financial risk spread.

Coppersmith stated that the trading structure in the South Korean market may have fundamentally changed, and leveraged capital may not quickly return in the short term. However, compared to a week ago, the most positive change is that position pressure has significantly decreased, and market sentiment is gradually recovering.

As position factors gradually fade, Goldman Sachs believes the market will refocus on corporate fundamentals in the next phase. Privorotsky noted that after such a sharp correction, investors with long-term capital characteristics can reassess investment opportunities in momentum trading, memory chips, and hardware equipment sectors. The semiconductor capital expenditure supply chain remains the most favored due to its long-term competitive advantages.

Goldman Sachs stated the next key observation point will be the U.S. Federal Reserve's FOMC meeting. However, the S&P 500 index options implied volatility currently reflects only about a 70-point swing on the meeting day, suggesting investors generally believe this rate decision will not become a new market risk source.

The report stated that the true determinant of whether AI trading can restart will be the latest outlook on capital spending, profitability, and AI investment returns from upcoming tech giant earnings. After the market completes large-scale deleveraging, the next round of stock price discovery is gradually returning to corporate fundamentals.