Goldman Sachs: Seven major "recovery" signals have appeared in the U.S. stock market, but investors must first endure until 2026

*Author: [Xu Chao](https://wallstreetcn.com/articles/3784976#from=ios?ivk=1), Wall Street Journal*

The U.S. stock market closed October with a round of intense fluctuations. Goldman Sachs derivatives strategist Brian Garrett outlined seven major market "repair" signals in the latest weekly report, believing that current positioning and sentiment have released a basis for contrarian buying—however, he also admitted that the real test is not over: "We just need to get through 2026 first."

The Nasdaq 100 index fell about 7% in October, marking the second worst month since the AI rally began in 2023, and it entered its third adjustment of the year. Chip stocks plunged 7% to 8% in a single day on Friday, marking the third largest single-day relative decline in over 15 years; the Philadelphia Semiconductor Index fell below the 200-day moving average on Friday after standing above it for 355 consecutive trading days and is currently testing the July low. The S&P 500 index fell 0.6% for the week, closing at 7580 points.


However, Garrett pointed out that despite spot prices remaining under pressure, "some contrarian indicators are emerging, suggesting that the current trend may have certain repair characteristics." Goldman Sachs Prime Brokerage data shows that the current net exposure level of hedge funds is historically usually accompanied by a drawdown of over 20%, while the S&P 500 is currently only down about 3% from its historical peak, and this divergence itself constitutes a contrarian signal.

### Chip Stocks End 355-Day Record, Valuations Return to AI Cycle Low

Goldman Sachs TMT expert Peter Callahan noted that the Philadelphia Semiconductor Index closed below the 200-day moving average on Friday, ending a record of 355 consecutive trading days above it, the third longest strong cycle in the past decade, and is currently retesting the July low. The forward price-to-earnings ratio of the semiconductor sector has fallen to about 16 times, reaching the valuation low of this AI cycle.

The trigger for this round of sell-off was a week of intensive earnings reports—Callahan believes that **the related data points "to some extent show characteristics of a growth inflection point," but he emphasized that these signals "come from outside the core AI complex,"** citing Amazon's guidance on AWS revenue visibility for the next two years as evidence.

The next day's reaction was even more severe: NOW fell 25% in a single day, INTC fell 30%, RDDT fell 30%, FFIV fell 20%, with many stocks posting their worst performance after earnings reports in history. **Callahan attributed the reasons to crowded positions, compressed earnings report calendars, and ongoing structural disputes in the industry.**

Garrett further pointed out that **the direction of market volatility has switched—stocks "are no longer surging upwards."** For example, Intel's average daily gain was 6.5% in June this year, while its current average daily loss has exceeded the average gain. Both the semiconductor ETF and QQQ exhibit the same pattern: the average amplitude on down days is again greater than on up days. **In Garrett's words, stocks are "operating like stocks again—downward volatility is more extreme than upward volatility."**


### Index Volatility Rebounds, but Gamma Remains Unmoved

The first contrarian signal listed by Garrett is: **the implied volatility of stocks at the macro level has finally begun to "rebound," catching up with the single-stock volatility and credit market that had fully priced in panic sentiment.**

Goldman Sachs' derivatives team's weekly report shows that the one-month implied volatility of the Nasdaq 100 surged nearly 2 volatility units in a single trading day, with clients first appearing to buy index downside protection and changes in the flow of semiconductor put options, while the S&P 500 briefly fell below the mid-term CTA (trend-following strategy) threshold during the day. The deviation of QQQ's actual volatility from implied volatility is one of the most significant single events in the past year.

The volatility of volatility also soared simultaneously: VVIX hit a high of 124 during the day, the highest level since the outbreak of specific geopolitical conflicts, before retreating. U.S. stock options traded nearly 100 million contracts in a single day on Friday, setting a record for the fourth highest in history.

The only thing that remained unchanged was Gamma. Garrett pointed out, "Despite the significant actual daily volatility of the index, the pricing of 0DTE straddles has not shown a clear response." For example, in the case of the S&P 500, the cost of straddles covering both Apple's earnings report and the midterm elections was only 150 basis points—lower than the actual daily drop of 180 basis points on Friday. In terms of year-end directional implied volatility, the S&P 500 is at ±5.1%, the Nasdaq 100 at ±7.8%, and the Russell 2000 at ±6.3%, with large-cap volatility again exceeding small-cap.

### Hedge Fund Net Exposure Reaches "Bear Market Level," but Index Has Not Fallen

The second contrarian signal comes from Goldman Sachs Prime Brokerage data: **Hedge funds net bought U.S. stocks for the first time in four weeks (+1.1 standard deviations), but 73% of the purchases came from macro products, mainly in the form of short covering—U.S.-listed ETF short positions decreased by 4.8%, the largest single-week short covering in over six months, led by technology, large-cap equity, and credit ETFs.**

Although technology was the most net bought sector that week, according to Goldman Sachs' Prime Brokerage team, this buying (as of Thursday) only accounted for about 40% of the selling volume over the past three weeks, with the "Magnificent Seven" (Mag 7) experiencing net selling on four of the past five trading days, **"indicating that risk appetite has shown almost no substantive recovery."**

Notably, Garrett pointed out that the Mag 7 remains "one of the few areas in TMT where positions continue to accumulate," with long positions still close to year-to-date highs—while the market sells technology, it is reluctant to part with the most concentrated holdings.

More crucial data shows that **the net leverage ratio of U.S. fundamentals has dropped to 47%, at the 7th percentile over the past year.** After conducting a regression analysis of this data against the S&P 500's drawdown, Garrett found that this net exposure level is historically usually associated with drawdowns of over 20%—while the current index is only down about 3% from its historical peak. In other words, hedge funds' positions are preparing for a "bear market that has not yet occurred," which precisely constitutes a contrarian bullish basis.

### 80% of Constituents Rise, but Index Falls: "Defensive Stocks Are Expensive"

The fourth and fifth signals come from the internal structure of the market. **On Thursday, 406 constituents of the S&P 500 rose, the highest proportion recorded since 2026, but the index still closed down that day. Garrett plotted all down trading days against the proportion of rising stocks on those days, and Thursday's data point was isolated—80% of constituents rose while the index fell, which is almost an anomaly in history.**

****

Behind this is the ongoing divide between AI and non-AI stocks: **Goldman Sachs' AI winners portfolio continues to be under pressure, while the S&P 500 excluding AI (S&P ex-AI) is expected to record its best single-month performance since 2022 this month. Meanwhile, U.S. Treasuries (TLT) and the AI winners portfolio show a two-week correlation of about -0.7, close to the low range of the past two years, presenting a classic risk-averse pattern of "AI down, bonds up."**

In terms of valuation, the S&P 500 has accumulated a gain of 10.7% this year, but the forward price-to-earnings ratio has fallen to its lowest level since 2026, about 19 times—indicating that while the index rises, valuations are "catching up" to earnings, forming a contrarian signal. However, "defensive stocks are already very expensive": the dividend yield factor is at the 90th percentile over the past decade, and the valuation factor is at the 80th percentile; while the momentum factor is only at the 30th percentile, and the growth factor at the 40th percentile. The forward valuation multiple of the low volatility index has expanded to growth stock levels.

Callahan wrote that the dialogue about "broadening the portfolio" is increasing, but "it's easier said than done"—trading days led by non-AI sectors often give a sense of "positions being dismantled," and it is difficult to find truly non-AI targets that are not disturbed by the narrative of AI disruption. He listed currently effective stocks including: AAPL, SNPS, ADI/TXN/NXPI, SPOT, V/MA, EBAY, and LYV.

### Apple Hits Historical Valuation High, Midterm Elections Priced at a Discount

This week, market catalysts are highly concentrated: Apple's (AAPL) earnings report after the market on Monday, AMD and ANET earnings on Tuesday, the midterm elections on Tuesday (Kalshi predicts a 65% probability of a complete Democratic victory), Wednesday's Treasury refinancing details (QRA), and Friday's non-farm payroll data (last month was +29,000). The implied volatility of the S&P 500 this week is 1.84%.

Callahan focuses on five stocks. Apple reached a historical absolute valuation high with a forward price-to-earnings ratio of 36 times, a premium of 17 times over the S&P 500, setting a record for relative valuation—this last remaining mega-cap stock still seen as a safe haven is also the most expensive in history, confirming the logic that "defensive is very expensive."

Regarding the midterm elections, Garrett expects that "it will not constitute an event" for the overall index, but believes it is a moment for thematic basket strategies to shine. He named the most noteworthy pair trade: going long on renewable energy and grid-related sectors while buying downside protection for mega-scale data centers and data center equipment—"who will pay for the electricity used by data centers," he believes is the "most undervalued election-sensitive theme" in U.S. stocks.

### The Last Two Signals: AI Financing Expectations Reverse, Retail Investors Start "Asking Before Buying"

Garrett's last two contrarian signals are more flexible. **The sixth: Credit market experts are beginning to suggest that AI financing demand may be below expectations. If mega-scale cloud providers' reliance on the bond market weakens, it will help alleviate market anxiety, but "it will take time to realize."**

**The seventh, which Garrett is particularly excited about: the behavior of his college classmates and father-in-law has begun to send warning signals—this group "has shifted from unconditional buying on dips to at least asking a question before buying." Garrett believes that retail investors are on the "eve of an intelligent transformation," citing an advertisement as an example: during the kids' trick-or-treating, the ad promises "set up your intelligent options strategy in 7 minutes."**

Overall, Garrett believes that the signals from positioning support a repair: net exposure corresponds to historical drawdowns of over 20%, panic-like market breadth anomalies have emerged, defensive stock valuations are severely overvalued, and retail sentiment has begun to waver. However, the flow of funds indicates that deleveraging is far from over: various trading desks are still selling technology stocks, long positions in the Mag 7 remain close to year-to-date highs, the CTA system is about to trigger a new round of short signals, and the options market's pricing for the combination of Apple's earnings report and the midterm elections is even lower than the actual volatility cost on Friday. As for 2027, the options market has priced the internal correlation of the index at about 20%, a historical low—at that time, the index may remain stable while the internals are tumultuous.