Overnight US stock market crash: A reflexivity storm triggered by a $100 billion “early exit”

On Thursday, Goldman Sachs' phones were "exploding."

"The pain is extremely high."

This was the message that John Flood, Goldman Sachs’ top equities trader, sent to clients in his report that day. As he described it, his phones were “exploding.”

The Nasdaq 100 closed down 6.4% on Thursday, wiping out about $2.3 trillion in market value in a single day—the biggest one-day drop since March 2020. The S&P 500 fell 4.4%, and the Philadelphia Semiconductor Index crashed 12%. The "fear index" VIX surged 70%, Nvidia's gains for the year were wiped out, and Western Digital plunged 22% in just one day.

The immediate trigger for this crash was the Democrats' sweeping victory in both houses of Congress in the midterm elections. Goldman Sachs quickly warned that this “blue wave” could lead to a pause on about 45% of planned data center expansions in the US—the core narrative propping up valuations for AI and tech stocks was suddenly put into question by political reality overnight.

But the extent of the market carnage far exceeded what policy shocks alone could explain.

Nomura derivatives strategist Charlie McElligott offered another layer of explanation in an intraday report on Thursday—a significant part of this crash was self-inflicted by “the smartest people in the room.”

McElligott described a market paradox: midday on Thursday, after Morgan Stanley calculated a $52 billion mechanical sell order for the close and the whole of Wall Street found out, everyone tried to sell early to buy back at lower prices. The frontrunners themselves became the very force driving the decline. He warned that if the usual “2–3 pm low then rebound” failed to materialize, the frontrunners would become the force pushing markets down even further. That’s exactly what happened by close; combined with $47 billion in leveraged ETF pressure, $61 billion from volatility control funds, and negative gamma on the part of dealers, over $100 billion in mechanical selling was unleashed—and would continue into Friday.

Today, the plunges in the Korean and Japanese markets are providing answers to this scenario.

A $100 Billion Time Bomb, and All of Wall Street Knew

To understand McElligott’s logic, you first need to understand a particular mechanism.

There’s a class of products called leveraged ETFs, such as “2x Long Nasdaq.” Before every close, these ETFs must do one thing: rebalance profits and losses to return to their targeted leverage ratio. When the market crashes, this means they have to sell a large volume of stocks before the close—the steeper the fall, the more they sell. This is automatic, mechanical, and unavoidable.

There are also Volatility Control Funds (Vol Control). When market volatility spikes, they systematically reduce their equity positions—also mechanical, but usually executed with a one- to two-day lag.

On Thursday, the combined force of these two mechanisms reached shocking scale.

Morgan Stanley’s derivatives team calculated at noon local time on November 5: assuming prices stayed at 1 pm levels, forced selling by leveraged ETFs before the close would total about $52 billion. Of that, semiconductors alone accounted for $21 billion, and Nasdaq 100 another $15 billion.

McElligott at Nomura gave a slightly lower figure but one that was equally striking: about $47 billion from leveraged ETFs—Nomura’s second largest one-day rebalancing event ever, behind only the $51.3 billion on June 5 this year. He added that the sell orders were “especially painful” because 86% of the assets were concentrated in tech and AI-related stocks.

Add to that the roughly $61 billion in Vol Control Fund sales and the $7 billion in selling pressure for each 1% decline due to negative gamma among dealers, and the market faced well over $100 billion in mechanical selling.

The key characteristic of these numbers is: they can be calculated in advance.

After Morgan Stanley calculated that $52 billion number at noon, as McElligott noted, “On Wall Street, no information stays secret for more than 45 minutes”—the number was rapidly known everywhere.

The Smart Money’s Game Plan: Sell Early, Wait for the Dip, Buy Back

Knowing about these mechanical sell orders, what do traders do?

According to McElligott, discretionary traders tend to “front-run” these mechanical flows.

The logic is simple: Since there’s going to be a large passive selloff before the close, I’ll sell now and buy back cheaply after the market gets hit.

And this strategy has worked well recently. McElligott pointed out in his report that during recent “large single-day swings,” the market’s intraday low typically forms between 2 and 3 pm ET, after which frontrunners and hedgers reverse, take profit, or close positions, driving a rebound into the close. The closing price typically ends up higher than the lows that the mechanical selling was expected to create.

Just last Thursday, McElligott reminded the market “it’s happening again”—leveraged ETFs faced $35 billion in sell orders, and the market followed the classic “2-3 pm low then closing rebound” pattern.

The June meltdown in the Korean markets this year followed a similar script.

On the day that tariffs were suspended in April, the mechanism worked in reverse: McElligott calculated the largest ever +$34.3 billion pre-close rebalancing buy, and frontrunners bought in early, ultimately driving a surge higher.

So on Thursday, with Morgan Stanley’s $52 billion number circulating by noon, every trader aware of it was gaming out the same thing: sell now, wait for the 2–3 pm trough, then buy back in.

The Reflexive Trap: The Cavalry Is the Seller

Then, McElligott penned the chilling worry in his report: “But what if that ‘assumed 2–3 pm low, then rebound’ never actually comes?”

Obviously, the scary thing now is: these numbers are so big that ‘what if’ we don’t get that almost universally expected late-session ‘rebound after a 2–3 pm low’ (a standard bounce of one or two percent off the low), what if there’s no rebound in the ‘standard TWAP’ window and instead new lows are made…and after feinting at a rebound, the market actually closes at the session low.

This is the core of the reflexivity paradox.

When everyone knows the script and everyone acts according to it at once, the script itself fails.

The $100 billion in mechanical sell orders originally meant to hit just before the close was spread out over the day by discretionary traders “front-running”—these frontrunners themselves became the force driving down the market.

At the same time, Vol Control Fund sell orders lag by a day or two—the crash on Thursday would trigger more selling on Friday. McElligott wrote that the 3 pm drop would prompt about $61 billion in Vol Control Fund selling, which could swell to $70 billion if the plunge reached 10%.

What’s even more deadly: when all the frontrunners have already sold and are waiting for “the low”—who's left to buy?

In his report, McElligott used a meme to describe this scenario: everyone is sitting on their chairs waiting for the same rebound—but there aren’t enough chairs to go around.

ZeroHedge summarized it as follows:

“Sometimes the cavalry comes at 3 pm. Sometimes, the cavalry is the one selling.”

On Thursday, the cavalry did not come. The market closed at the daily low: Nasdaq 100 -6.4%, S&P 500 -4.4%

Stampede Scene: Best Trade of the Year Gets Wiped Out in 48 Hours

Before the reflexive mechanism was detonated, there was another powder keg in the market waiting for a spark.

McElligott called Wall Street’s “best trade of the year” the “semiconductor-energy barbell strategy”—long semiconductors (AI compute demand beneficiary), long energy (geopolitical risk premium + inflation hedge). A few weeks ago, the strategy was up 66% with a Sharpe ratio of 3.3.

Then, over the past 48 hours, both legs got wiped out at once:

• Energy: down 10.2% this week

• Semiconductors: down 13.1% this week

Weekly Sharpe ratios were -17.2 and -10.1 respectively.

• The collapse in the energy leg was triggered by the market repricing Iran risk—McElligott observed the White House signaling possible flexibility on Iran, war risk was quickly priced out, and Brent crude saw its biggest weekly drop since the Covid pandemic.
• The collapse in the semiconductor leg stemmed from market worries about data center regulation after the Democrats swept Congress.

Both legs collapsing at once means funds holding the barbell strategy not only faced losses, but were forced to unwind positions. The sentiment described by McElligott as “locking in this year’s gains just to survive to next year” was completely shattered in this moment.

Walleye Capital announced it would return client funds and fired its equities head, calling it their worst month since Covid. Goldman’s John Flood described long-only institutions as “deadlocked again”; the massive cloud-computing holdings bought in yesterday’s rebound were dumped just as quickly, and “no buyers have yet stepped in to take the other side.”

Abnormal ETF trading volumes confirmed the degree of panic: SPY volume was 160% above its 20-day average, QQQ up 100%, semiconductor ETF SMH up 150%.

According to Vanda Research senior strategist Jai Malhi, Thursday’s retail selling was “set to become the biggest single-day retail liquidation since late July last year.”

Bond Shorts Squeezed, the Other Leg Bleeds Too

There’s yet another leg quietly breaking at the same time.

CTA trend funds have built large short bond positions this year—they’re a part of the “short bonds / long inflation” trade and integral to the barbell strategy.

The stock market crash triggered a flight to safety into Treasuries, forcing rapid price jumps and short covering in bonds. McElligott’s estimates show CTAs face a “massive short squeeze” in bonds: another 2% rise in Treasury prices would trigger about $225 billion in forced global bond buying—$48.5 billion in US 2-year bonds, €69.5 billion in Eurozone 2-year, and ¥23.9 billion in Japanese 10-year bonds.

This means the bond short squeeze itself is now fueling the bond rally and putting further pressure on equity longs—yet another reflexive cycle.

The Crash Isn’t Over: Lagged Selling Pressure on Friday

The close on Thursday did not mean an end to this wave of mechanical selling.

Vol Control Funds act with a lag. McElligott wrote that Thursday’s decline would trigger more selling from these funds on Friday—“today’s drop is causing further orders to load up in the queue.”

UBS’s estimate is more direct: its risk-control rebalancing model shows that should the S&P 500 fall 2%, systematic funds would sell between $95 billion and $105 billion.

The Asian market is already reflecting this reality.

In Friday’s early session, Korea's Composite Index crashed 8.8%, broke below the 5,000 mark, and triggered circuit breakers—a level that Korean President Lee Jae-myung once considered a milestone of economic recovery, now quickly breached. The Nikkei 225 fell nearly 5.8%, dropped below 60,000 intraday, and is off more than 20% from its June high—making it an official technical bear market.

US equity futures saw a modest rebound in Asia’s morning session, with Nasdaq futures briefly turning green, signaling some investors were dipping back in on weakness. But UBS warned the pressure from systematic selling wasn’t fully released yet.

There’s one more variable on Friday: the US October nonfarm payrolls report is due out, with consensus expecting 59,000 new jobs and the unemployment rate holding at 4.2%. The direction of the data will decide whether this correction finds a window to breathe.