Goldman Sachs: Seven Major "Recovery" Signals Emerge in US Stocks, but Investors Must Endure Until 2026

U.S. stocks ended October with a round of sharp volatility. Goldman Sachs derivatives strategist Brian Garrett summarized seven major "recovery" signals in his latest weekly report, suggesting that current positioning and sentiment have released a contrarian buy signal—but he also admits the real test isn't over yet: "We just need to make it through 2026 first."

The Nasdaq 100 Index fell approximately 7% in October, making it the second-worst month since the AI rally began in 2023, and the index entered its third correction of the year. Chip stocks plunged 7% to 8% on Friday alone, marking the third-largest single-day relative decline in more than 15 years. The Philadelphia Semiconductor Index, after staying above the 200-day moving average for 355 consecutive sessions, broke below the average on Friday and is now testing July’s lows. The S&P 500 fell 0.6% for the week, closing at 7,580 points.

However, Garrett points out that despite ongoing downward pressure on spot prices, "some contrarian indicators are emerging, suggesting current trends may have some recovery characteristics." According to Goldman Sachs Prime Brokerage data, the current net exposure levels of hedge funds historically accompany more than a 20% pullback, while the S&P 500 is currently only about 3% below its all-time high. This divergence itself constitutes a contrarian signal.

Goldman Sachs TMT expert Peter Callahan points out that the Philadelphia Semiconductor Index closed below its 200-day moving average on Friday, ending a streak of 355 consecutive sessions—a record of continuous strength, the third longest in the past decade. The sector is now retesting its July lows. Forward P/E ratio for the semiconductor sector has fallen to about 16x, touching the cycle low valuation of this AI uptrend.

The trigger for this sell-off was a week packed with earnings reports—Callahan believes relevant data points "somewhat indicate an inflection in growth," but emphasizes these signals "all come from outside the AI core complex," citing Amazon's visibility into AWS revenue for the next two years as evidence.

The following day saw even more dramatic moves: NOW dropped 25% in a single day, INTC fell 30%, RDDT fell 30%, FFIV fell 20%, with many stocks recording their worst post-earnings performances in history. Callahan attributes this to crowded positioning, compressed earnings schedules, and ongoing structural controversy in the industry.

Garrett further notes that the direction of market volatility has shifted—stocks are "no longer surging explosively upward." Taking Intel as an example, its average daily gain in June reached 6.5%, whereas the current average daily loss has surpassed the average gain. Both the semiconductor ETF and QQQ are showing the same pattern: average swings on down days now exceed those on up days. In Garrett's words, stocks are "starting to behave like stocks again—downside volatility is now more extreme than upside volatility."

The first contrarian signal Garrett lists is: implied equity volatility at the macro level has finally started to "catch up," aligning with the previously priced-in panic in single-name volatility and the credit markets.

Goldman Sachs derivatives team's weekly report shows that one-month implied volatility on the Nasdaq 100 jumped nearly 2 volatility points in a single session, as clients for the first time were seen chasing downside index protection and buying semiconductor sector puts. The S&P 500 briefly broke below the mid-term CTA (trend-following strategies) threshold. The deviation between QQQ realized volatility and implied volatility was one of the most notable single events over the past year.

The volatility of volatility also soared: VVIX intraday peaked at 124, the highest since specific geopolitical conflicts broke out, before pulling back. U.S. equity options saw nearly 100 million contracts traded in a single day on Friday, the fourth-highest day in history.

The only thing unmoved is Gamma. Garrett notes, "Despite large single-day realized index moves, 0DTE straddle pricing hasn’t shown a notable reaction." For the S&P 500, the cost of a straddle covering both Apple’s earnings and midterm elections catalysts was only 150 basis points—below the actual single-day move of 180 basis points on Friday. For year-end directional implied volatility: the S&P 500 is ±5.1%, Nasdaq 100 is ±7.8%, Russell 2000 is ±6.3%, with large-cap volatility again above small-cap.

The second contrarian signal comes from Goldman Sachs Prime Brokerage data: for the first time in four weeks, hedge funds were net buyers of U.S. equities (+1.1 standard deviations), but 73% of the buying came from macro products, primarily driven by short covering—U.S.-listed ETF shorts fell 4.8%, the largest weekly short covering in more than six months, led by tech, large-cap equity, and credit ETFs.

Though tech was the sector with the most net buying during the week, according to Goldman Sachs prime brokerage, those purchases (through Thursday) were only about 40% of the amount sold in the prior three weeks. The "Magnificent 7" were net sold on four of the past five trading days, "showing risk appetite has seen almost no substantive repair".

Worth noting, Garrett points out that the Mag 7 is still "one of the few areas in TMT where positions continue to accumulate," with long positions still near year-to-date highs—while the market is selling tech overall, it seems reluctant to part with its most concentrated holdings.

More importantly, U.S. fundamental long-short net leverage has dropped to 47%, in the 7th percentile of the past year. Garrett regressed this data against S&P 500 drawdowns and found that such low net exposures historically correspond with over 20% pullbacks—yet the index is only roughly 3% off its record high. In other words, hedge fund positioning is prepping for a "bear market that hasn’t happened yet"—which is precisely a contrarian bullish factor.

The fourth and fifth signals relate to market internals. On Thursday, 406 S&P 500 members closed higher—the highest single-day ratio since 2026—yet the index still finished lower. Garrett plotted all losing trading days against the share of rising stocks, and Thursday’s datapoint stands alone: 80% of members gained while the index declined—an almost unique historical occurrence.

This reflects the ongoing split between AI and non-AI stocks: Goldman’s AI winners basket remains under pressure, while the S&P ex-AI looks set for its best monthly gain since 2022. At the same time, U.S. Treasuries (TLT) and the AI winners basket show a two-week correlation of about -0.7, near the bottom of the last two years, reflecting the classic "AI down, bonds up" risk-off regime.

On valuations, the S&P 500 is up 10.7% year-to-date, but forward P/E is back to the 2026 low of roughly 19x—showing the index rose while valuations "catch down" to earnings, a contrarian signal. However, "defense is already expensive": dividend yield factor is at the 90th percentile of the past 10 years, value factor at the 80th percentile; momentum factor only at the 30th percentile, growth factor at the 40th. Low-volatility indices’ forward valuations have expanded to growth stock levels.

Callahan writes that talk of "broadening portfolios" is increasing, but "easier said than done"—trading days led by non-AI stocks often feel like positions are being unraveled, and it’s hard to find truly AI-immune non-AI names. He lists current effective stocks as: AAPL, SNPS, ADI/TXN/NXPI, SPOT, V/MA, EBAY, and LYV.

This week’s market catalysts are highly concentrated: Apple (AAPL) earnings after the bell on Monday; AMD and ANET earnings Tuesday; midterm elections Tuesday (Kalshi estimates a 65% probability of a Democratic sweep); Treasury QRA details Wednesday; and nonfarm payrolls Friday (last month +29,000). The S&P 500’s implied weekly move is 1.84%.

Callahan focuses on five main stocks. Apple hit a record 36x forward P/E, a 17x premium to the S&P 500—by far its highest ever relative valuation. The ultimate safe haven megacap has become the most expensive stock ever, confirming that "defense is expensive".

As for the midterms, Garrett expects the event to "not be material for the overall index," but sees it as a moment for thematic basket strategies. He highlights the best pair trades: long renewable energy and power grid themes while simultaneously buying downside protection on hyperscale data centers and data center equipment—a reference to the "who pays for data center electricity" theme, which he calls the "most underestimated election-sensitive theme in U.S. equities."

Garrett’s last two contrarian signals are more nuanced. Sixth: credit market experts are beginning to hint that AI financing demand may be lower than expected. If hyperscale cloud operators rely less on the bond market, it may help ease market anxiety, though "it will take time to materialize."

Seventh, and Garrett's favorite: his college classmates and father-in-law are now flashing warning signs—their group has shifted "from always buying the dip unconditionally, to at least asking questions first." Garrett believes retail investors are on the "eve of an agentification transformation," and gives an example: during kids’ trick-or-treat breaks, an ad promises "set up your agent options strategy in 7 minutes."

Overall, Garrett sees positioning signals that support a recovery: net exposures reflect historical drawdowns over 20%, there's abnormal panic in market breadth, defensive stock valuations are severely overstretched, and retail sentiment is wobbly. Yet, fund flows suggest deleveraging is far from over: trading desks are still selling tech, Mag 7 long positions near year highs, CTA systems about to trigger another wave of shorts, and the options market pricing for Apple earnings plus the midterms is even below actual realized volatility cost on Friday. By 2027, options markets are pricing intra-index correlations at historic 20% lows—suggesting the index itself may seem stable while underlying internals remain in turmoil.