Last Thursday, US equities suffered a historic shock that caught the options market off guard. In a weekend report, Goldman Sachs derivatives strategist Brian Garrett noted that by market cap destruction, trading volume, volatility spike, and investor positioning, the day's market data echoed the COVID-19 shock, the Global Financial Crisis, and the tariff panic of "Reciprocal Tariff Day"—but its greatest distinction was that the decline's "surprise factor" relative to prior options market pricing was the largest in 19 years.
Last Thursday, the Nasdaq 100 plunged 6.4% in a single day, its worst one-day drop since March 2020; the S&P 500 fell 4.4%, its worst performance since so-called "Reciprocal Tariff Day." That day, Cboe recorded the largest single-day put option volume in history, with roughly $2.5 trillion in Nasdaq 100 market cap evaporating in one session—the largest single-day dollar loss in the dataset's history. Garrett characterized last week's market action as "impossible to describe in any words other than 'ugly.'"
Garrett's core judgment is that the shock "resembles a violent positioning reset rather than a systemic event," but he warned that "more mechanical selling pressure remains to be absorbed," and whether it ultimately proves to be a "positioning reset" or "something more consequential" remains an open question. The Goldman Sachs Panic Index spiked to 9.57 last Thursday, a recent historical high. Garrett noted that readings above 9 have historically been good buying opportunities.
"The Largest Underestimation in More Than Two Decades": MSR Hits 19-Year Record
Garrett constructed a specific metric to measure the "surprise factor" of the decline—the Move Surprise Ratio (MSR), defined as an asset's single-day percentage move divided by its prior-day 1-month implied volatility: relative to what the options market was pricing, how large was the actual move? The conclusion is striking: the S&P 500's MSR was the largest since at least 2007, exceeding any reading during the Global Financial Crisis, Volmageddon, and the COVID-19 pandemic. The Nasdaq 100 (measured via QQQ) had the second-largest MSR in history, behind only so-called "Reciprocal Tariff Day"; the Russell 2000's reading ranked in the top five historically. Garrett's assessment: "To say the volatility market underestimated this move is the largest underestimation in more than two decades." Nomura's Charlie McElligott also admitted on Friday that previously flat skew and shrinking VIX ETN size led him to misjudge the market's lack of "accelerant fuel"—"Boy, was I wrong."
Market Cap Destruction and Volume: A Historic-Scale "Hole"
Garrett began with the scale of losses. The Nasdaq 100 shed roughly $2.5 trillion in market cap in a single day—the largest single-day dollar loss in Goldman Sachs data going back to 2001, far exceeding any prior episode. The percentage decline looks less extreme only because the index's absolute size has never been larger. Volume also set records. Garrett noted that last Thursday's dollar notional turnover was the largest "regular trading day" volume in history (excluding option expiration and index rebalancing days), reaching approximately $1.75 trillion—far above levels during the April 2025 tariff panic. ETFs played a key role. Last Thursday was the third-largest ETF dollar notional trading day in history, with ETFs accounting for 33% of consolidated volume, above the 28% historical average. Garrett's explanation was blunt: "When VVIX rises more than 40 points in a day, investors' first instinct is to grab the blunt instrument that is ETFs."
Hedges Unwound Early: Why the Volatility Market Was Caught Off Guard
A key reason the crash was so violent is that buy-side investors had just proactively removed hedges. Garrett wrote: "Clients had aggressively removed protective positions during Wednesday's post-midterm election rally, leaving investors under-hedged heading into Thursday's volatility, and the explosive rise in convexity had few precedents." The volatility of volatility (VVIX) jumped about 44 points in a single day—the second-largest one-day gain since 2006, behind only the February 2018 "Volmageddon" event. On a delta-adjusted basis, the 1-month VIX's excess performance also ranked among the largest in the past five years, on par with the August 5, 2024 yen carry trade deleveraging and so-called "Reciprocal Tariff Day." Correlation also spiked sharply. Garrett noted that the market had long been in a state of extremely low single-stock correlation; last Thursday, the S&P 500's 3-month implied correlation jumped about 9 points in a day to slightly above 20—one of the largest single-day gains since 2023—reaching in one day the level where 12-month implied correlation had been.
Panic Index Jumps from "Near Zero" to 9.57
Six weeks ago, Garrett was lamenting that US equities "refused to price any panic"—the Goldman Sachs Panic Index closed below 1, while bond volatility and credit spreads were already signaling extreme fear. Last Thursday, that situation reversed completely. The Goldman Sachs Panic Index (composed of equal-weighted percentile rankings of VIX volatility, S&P 500 skew, S&P 500 term structure, and S&P 500 1-month at-the-money implied volatility) hit 9.57 last Thursday and closed the week at 9.2. Garrett noted that readings above 9 have historically been good buying opportunities, but cautioned that readings above 9 around so-called "Reciprocal Tariff Day" persisted for weeks before bottoming. Meanwhile, the way the market expresses bullishness is quietly shifting. Garrett wrote: "The 'volatility goes lower' trade has become the new 'spot goes higher' trade." On Friday, investors chased not S&P 500 or Nasdaq 100 calls, but VIX puts, with the VIX put/call volume ratio rising to about 1.5—the highest level in his two-year chart.
Positioning Data: Net Exposure Falls to 0th Percentile
Positioning data is equally striking. Goldman Sachs Prime Brokerage data shows last week's net selling was driven primarily by long liquidation. US fundamental long/short funds' gross leverage fell 8.7 percentage points to 207.5% (26th percentile of 1-year range), while net exposure fell 3.7 percentage points to 43.4%—a 1st percentile reading. On a beta-adjusted basis it was even more severe: gross leverage at 213.2% (-15.7 percentage points) and net exposure at 57.5% (-6.4 percentage points), both at the 0th percentile. At the single-stock level, single stocks saw net selling for an 8th consecutive week, with this week's supply the largest in over four months, and defensive rotation within: selling tech and industrials, buying utilities, real estate, and financials. Global tech stock selling matched the largest in Goldman Sachs data history, with four of the top five largest tech net selling weeks occurring in the past five months. As of last Thursday's close, TMT long/short funds were roughly flat year-to-date, having given back all of June's +24% peak. Goldman Sachs' high-beta momentum portfolio is now in a 50% drawdown—deeper than the 2021 deleveraging and the deepest since the series began in 2011. The six-month return spread of the "AI vs non-AI" trade has fallen to -37%, the lowest in the decade-long chart, compared with nearly +70% just over a year ago. Goldman Sachs' high-touch cash desk net sold $6 billion for the week, concentrated in tech, and has sold $30 billion in tech stocks (TMT plus communication services) since the current decline began—exceeding July's $21 billion supply. The one positive signal: at Friday's close, the semiconductor sector saw its first buying skew in more than six trading days.
Systematic Selling Pressure Not Yet Over
For investors hoping Friday's bounce marked a bottom, Garrett offered a warning: systematic traders are still selling stocks and buying bonds. Goldman Sachs data shows CTA signals have turned negative in about 60% of global equity markets, and the firm estimates roughly $28 billion in S&P 500 futures to be sold over the next week and about $45 billion over the next month. On bonds, CTAs have bought $33 million in DV01, with an estimated additional $28 million over the next week and $78 million over the next month—which Garrett called "quite substantial." This means trend followers are about to significantly lengthen duration, while nonfarm payrolls came in strong at 165,000 and the market has begun pricing a December Fed rate hike. Garrett also warned about defensive sector valuations. Consumer staples currently trade at about 22.5x 24-month forward P/E—roughly double the semiconductor index's ~10.5x. The Goldman Sachs Delta One desk favors shorting its diversified staples basket (GSXUSTAP), citing overbought valuation premiums and dividend yields relative to the S&P 500, and noting that "at current rate levels, shorting via swaps generates positive carry."
Reset or "Something More Consequential"?
Given the above data, Garrett's operational bias is cautiously long: "There may be more technical pressure to absorb in the short term, but I would buy December call spreads on the indices and the hardest-hit sectors (long delta, volatility neutral, fading call skew)—including DRAM, SOXX, etc." He left an open question: "The question now is whether this represents yet another painful positioning reset or the beginning of something more consequential." The head of Goldman Sachs Delta One leans toward the 1998 script rather than 2000; Tony Pasquariello sees one last rally before a tougher 2027; while BofA's Michael Hartnett believes this Democratic "blue wave" marks the "peak of AI." Garrett concluded that given ongoing systematic selling pressure, the possibility of Fed rate hikes during deleveraging, and unanswered questions about AI financing, the possibility of "something more consequential" remains.
Comments