The November Correction — Update

From the record: chips −33%, Nasdaq 100 −15%, the average S&P stock −3%

The selling is concentrated in the biggest tech and chip names. The average S&P 500 company (every stock counted equally) even made a record close on Nov 4, one day before today's drop.

QQQ fell straight through the Nov 3 zone and closed on its low

The shaded band is the $650–661 zone from our Nov 3 report. QQQ bounced after election day, then closed under the zone today. The Nasdaq 100 is now 4.8% under its 200-day average (the average close of roughly the last 10 months), a line it was still above on Nov 4.

Two sessions after election day, 2026 is weaker than every past midterm

From the election-day close (Nov 3), the S&P is down 2.8% and the Nasdaq 100 down 5.5% two sessions later, after rising 1.6% and 0.9% on Nov 4. The worst past starts were −1.4% for the S&P (16 midterms since 1962) and −2.3% for the Nasdaq 100 (10 since 1986).

Yet a year after every past midterm the S&P was higher, usually with a scare along the way

Every past case was also higher six months later. Typically the S&P gained 5.9% in three months and 13.2% in a year after a midterm, against 2.6% and 10.5% from any day. The Nasdaq 100 was higher a year later in all 10 of its cases, typically +28.5% (any day: +17.7%).

Every Nasdaq 100 drop of 10% or more since 1985, by where it finally ended

Each dot is one finished drop, 34 in all. All 15 that got as deep as today went past −15%, about half went past −20%, and only about 1 in 4 stopped within 2 points of here. Three (1993, 1995, 1996) stopped just short of today's line.

Their low came a typical 22 trading days (about a month) later. Today's low touched −14.8%; counted from the first close past −15%, the low came sooner, typically within 5 trading days, and about half got there inside that window.

Splitting them by speed or trend barely moved the typical end (−21% to −23%). Fast drops that started with the trend still up, like this one, ended deeper (around −25.5%), but only four did (1997, 1998, 2000, 2020), so treat that as a rough guide. The Nov 3 comfort (those fast drops usually stopped near −12%) no longer applies: the shallow ones ended before reaching here.

A year later, drops this deep were usually well higher

Typical Nasdaq 100 change, in %. From the first close at today's depth it was higher a year later in 13 of 15 cases, typically +28.4%. From the first close past −15% the picture is almost the same (+30.9% a year later).

S&P 500 dealer gamma by day: the deeper the red, the bigger the moves, either way

Dealer gamma is the hedging the option-selling banks must do every day, in billions of dollars. Above zero (green) it calms the market; below zero (red) it adds to moves both ways. Election day: −$25.1B. The Nov 4 bounce: −$6.8B, with the S&P near 7,686. Today: −$48.7B, the most negative since Oct 16 (−$85.3B).

Dragon's map stepped down about $35 on QQQ and $20 on SPY in two days

The King is the price with the most option weight on Dragon's map; the put wall and call wall carry the most put and call weight. QQQ now sits between them: put wall $645, King $650, call wall $653. The indexes moved the same way: SPX King 7,500 → 7,400 (put wall 7,300, call wall 7,900) and NDX King 28,000 → 27,000 (put wall 26,500, call wall 28,800).

Options now price bigger swings than on Nov 3, and protection costs more

The implied move is one standard swing, up or down, that option prices expect by that expiry; the white tick is where it was on Nov 3. For QQQ by Nov 20, ±6.4% is about ±$41 from today's close. Protection against a fall got pricier too: QQQ's skew (puts over calls, in volatility points) rose from 7.2 to 11.6 and its implied volatility from 30.4% to 39.0%. SPX and NDX moved the same way (Dec 18: SPX ±4.6% → ±5.7%, NDX ±7.2% → ±8.5%); puts per call barely changed.

The insurance moved down: QQQ's biggest put piles under the close are now $600 and $620

Puts are insurance that pays when price falls. On Nov 3 the biggest pile below price was at $680; QQQ has now closed under $650 (219K puts there in the midday read, since these counts update overnight). On the S&P, puts at 7,000 grew to 786K (437K of them expire Dec 18), and Dragon's live map counts 1.41 million contracts at that node.

The 30-day fear reading jumped over the 3-month one at today's close

The VIX is how much movement options traders expect over the next 30 days; VIX3M is the same over 3 months, and it normally sits on top. Today the VIX closed at 28.18 (16.61 the day before) against 25.55: "inverted", meaning traders pay more for protection now than later. It is also the VIX's first close above 25 in this drop. The Nasdaq's own fear meter (VXN) closed at 38.26.

In moments like this, the VIX went on to 30 about two times in three

After a first inversion in a Nasdaq pullback of 8–20% (only 13 cases since 2006), the VIX reached 35 within three months about half the time. The Nasdaq 100 was typically higher a month later (+2.8%) and three months later (+3.9%), but first fell a typical 8% further. After a first VIX close over 25, it reached 35 about 4 times in 10. The "days like today" overlap heavily, so they count for less than they look; three months on, the Nasdaq 100 was typically 10.3% higher after a further dip of about 4.5% first.

Across all drops the signs mattered; among drops this deep, they didn't

Across all 34 drops, more signs came with a deeper typical bottom. Among the 15 that got as deep as today it flipped: drops with two or three signs ended a bit shallower than those with zero or one. Few cases on either side.