Bitget App
Trade smarter
Buy cryptoMarketsTradeFuturesStocksEarnInstitutionAI & More
US Stocks Last Week: Institutions Set Record for "Selling", Retail Investors Set Record for "Buying the Dip"

US Stocks Last Week: Institutions Set Record for "Selling", Retail Investors Set Record for "Buying the Dip"

华尔街见闻华尔街见闻2026/10/19 04:08
Show original
By:华尔街见闻

Last week, hedge funds recorded the largest net sell-off in the TMT sector in history, with net exposure dropping to the 0th percentile. However, at the same time, retail investors made record purchases through leveraged ETFs, with SOXL alone absorbing $7 billion. The issue is that the overall institutional exposure remains as high as 213.9%. In other words, fund managers opted for hedging rather than genuinely deleveraging and exiting the market.

The semiconductor sector just experienced its worst week of the year, yet retail investors are bottom-fishing with record-setting 3x leveraged buys, forming a rare standoff against hedge funds that hit a historic net selling high.

According to the latest data from Goldman Sachs Prime, net outflows from global tech stocks hit a record last week, bringing net exposure down to the 0th percentile; total gross exposure, however, remains high at 213.9%, in the 83rd percentile.

Meanwhile, retail investors are bottom-fishing through leveraged ETFs. Chris Lucas, Goldman Sachs ETF specialist, noted in a Sunday report that nearly $10 billion flowed into leveraged ETFs last week, the largest weekly inflow since data began in 2006. About $7 billion of this went into the 3x long semiconductor ETF SOXL, a record for the product, accounting for 74% of the week’s net inflows.

Goldman Sachs derivatives and flow strategist Brian Garrett believes positioning is already extremely low, market breadth is improving, and AI debt is not new information. The next move will “depend more on earnings confirmation than any other factor.” With institutions heavily underweight but not truly exited, and retail leverage highly concentrated, earnings season becomes the decisive battleground.

Semiconductors Face Heavy Sell-Off

Last week, the Nasdaq 100 closed at weekly lows, down about 5.5%, the steepest drop since “tariff parity day” on April 2. The Philadelphia Semiconductor Index (SOX) fell for eight consecutive sessions, sliding about 11.5% on the week.

US Stocks Last Week: Institutions Set Record for

Goldman Sachs TMT strategist Peter Callahan described in a Sunday internal report:

This was an extremely tough week... Institutional flows show clear selling pressure in tech/AI sectors as the market strives to lower growth expectations—driven by both rate and oil price pressures, and rising skepticism about linear AI capex growth.

Semiconductors have pulled back around 25% from their highs, and the RSI is approaching oversold territory for only the second time in four years. Callahan also pointed out that Goldman Research estimates AI capex now accounts for about half of S&P 500 EPS growth.

Retail’s “Triple Leverage Bottom-Fishing” Meets Institutional “Record Selling”

Facing an 11% weekly drop, retail investors doubled down with 3x leverage. In addition to the $7 billion into 3x long semiconductor ETF SOXL, the 3x long Nasdaq ETF TQQQ saw about $1.3 billion in inflows. This drove total leveraged ETF net inflows to roughly $9.5 billion (October 12–16), about twice the previous record high.

US Stocks Last Week: Institutions Set Record for

Goldman Sachs historical data shows leveraged ETF flows are countercyclical: these products see the strongest inflows on S&P 500 days with moves exceeding two standard deviations, averaging net inflows of 56 bps of prior AUM over the past five years; conversely, on best-performing days, outflows are about 17 bps. Lucas summarizes this as “buy panic, sell euphoria.”

But history also serves as a warning. Goldman’s Benny Quek notes that Korean leveraged ETF exposure has collapsed, with AUM shrinking from a $53 billion peak to $15 billion; the 3x long Korea ETF KORU has hit a 2026 low, down 80% from its June peak.

On the institutional side, Goldman Sachs Prime’s weekly net selling in global tech stocks hit a record; cumulative selling from October 8–15 matched late July levels, the largest since March 2022. In Asia alone, there was net selling of $7.8 billion, mainly in Taiwan. Quek believes institutional positioning is “about 6/10, significantly cleaner than in June and July.” The real worry lies in fundamentals—rates and credit, earnings, and ROIC.

“Reducing Longs” Has Happened, “Reducing Size” Has Not

Garrett revealed the nuance in institutional positioning: U.S. fundamental long/short net exposure fell 3.2 points last week, the largest drop since April 2, down to the 0th percentile, indicating “reducing longs” has occurred; however, gross exposure remains at 213.9%, the highest in three years (83rd percentile), so “reducing size” has not happened.

In other words, fund managers are choosing to hedge rather than truly de-leverage. They’re adding short hedges to longs, but not reducing overall holdings. This means those crowded long positions are still fully on the books.

On valuations, Nasdaq 100’s forward P/E has dropped to 20.5x, its lowest since the start of the AI cycle in 2023, narrowing its premium over the S&P 500 to about 1.8x, the tightest in nearly a decade; semiconductor forward P/E has dropped from about 29x to 17x. Meanwhile, nominal S&P 500 put option turnover surged to $3.3 trillion on Friday—a record—with Goldman’s “fear index” rising to 7.4.

Is This Bottom-Fishing Effective?

Lucas admits retail buy-the-dip is nothing new:

Flows into the leveraged ETF complex are historically strongest when markets experience violent selloffs (negative 2-sigma events) and weakest on market’s best days. It’s also worth noting that inflow magnitude is larger in absolute terms during sharp selloffs.

Over the past five years, this logic did hold—every time there’s a crash, the Fed has stepped in as backstop.

But Garrett’s APAC colleague Benny Quek provides a counterpoint: assets under management in Korean leveraged ETFs have plunged from a $53 billion peak to $15 billion.

Credit Market Warnings Louder Than VIX

Garrett notes that AI debt burden is “nothing new,” referencing an article titled “AI Borrowing Surges into the Bond Market” and points out, “This sounds like a weekend article, but it’s actually a Wall Street Journal piece from a year ago.”

He also acknowledges that CCC-rated spreads have broken out above 1,000 basis points, and that volatility in investment grade credit default swaps (CDX IG) is “much more pronounced than 1-month VIX futures”—IG spreads have widened to their highest since April, while VIX futures have only modestly risen.

Earnings Season Becomes the Decisive Moment

Lucas pointed to a “cocked spring”: last week’s buying spree drove cumulative YTD leveraged ETF flows (excluding inverse products) from about -$12 billion back to around -$2 billion. Leveraged ETFs are now managing $152 billion in assets with about $386 billion in gross exposure. If market conditions improve, the newly minted ETF units could become a multiplier for any rebound; if semiconductors keep falling, daily rebalancing will trigger mechanical sell pressure, with the unit base much larger than a week ago.

Garrett sets a timeline: 14% of S&P constituent companies report earnings by this weekend, and 44% report next week. Meta, Tesla, and Lam Research will release results on Wednesday (October 21), while Nvidia CEO Jensen Huang will deliver a keynote at GTC Berlin the same morning. Whether underweight institutions will be forced to cover on earnings beats, and if retail’s 3x leverage can hold out for a rebound, will become clear this week.

The next key point to watch is whether SOXL fund flows continue this week, and guidance from Meta, Tesla, and Lam on Wednesday.

0
0

Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.

Understand the market, then trade.
Bitget offers one-stop trading for cryptocurrencies, stocks, and gold.
Trade now!

You may also like

US Treasury "Liquidity Drain" Starts Again: Treasury Department Aggressively Issues Short-term Debt—Can Fragile AI Trading Withstand the Storm?

As artificial intelligence (AI) trading faces significant challenges, another liquidity “drain” triggered by U.S. Treasury issuance appears poised to impact the entire U.S. stock market.

智通财经•2026/11/09 07:11
US Treasury "Liquidity Drain" Starts Again: Treasury Department Aggressively Issues Short-term Debt—Can Fragile AI Trading Withstand the Storm?

Goldman Sachs: The U.S. stock market will have a "final surge" in 2027, but what follows may resemble the period just before the 2000 bubble burst

Goldman Sachs warns that the three core variables driving the strong performance of U.S. stocks over the past four years—earnings growth, capital flows, and the AI narrative—have all deteriorated at the margin. It is expected that the S&P 500 will complete its last round of gains, after which, in 2027, it will face the largest net equity supply shock since the bursting of the internet bubble in 2000.

华尔街见闻•2026/11/09 06:47

Nvidia (NVDA.US) stock plunges in after-hours trading amid Jensen Huang’s personal scandal and internal investigation

Due to personal scandals involving CEO Jensen Huang and the company's initiation of an independent investigation, Nvidia's stock price plummeted by over 4% during after-hours trading on November 8.

智通财经•2026/11/09 06:41
Nvidia (NVDA.US) stock plunges in after-hours trading amid Jensen Huang’s personal scandal and internal investigation

Tech stocks experience their worst week since the pandemic; Morgan Stanley: AI trading hits the "wall" moment in the marathon.

After more than two years of a strong bull run, the market once believed that the AI narrative was already more than halfway completed and victory was within reach. However, last week’s crash suggests that the real test may have only just begun.

华尔街见闻•2026/11/09 06:31
Tech stocks experience their worst week since the pandemic; Morgan Stanley: AI trading hits the "wall" moment in the marathon.