QQQ holds the 100 largest non-financial Nasdaq companies and the top ten are over half the fund. The index has come back to the $700–705 floor it has held five times since June, with the dealers' largest put strike at $700 and their book short gamma, so the next move from here is likely to be fast in whichever direction it goes. This page looks underneath the index: how many holdings are trending, how many are making new highs or lows, whether the average stock is keeping up with the giants, where the dealer gamma levels sit, and a scorecard of the signs that have marked this year's lows.
The state of QQQ in one paragraph. Updated Friday 30 October before the open, with Thursday's close and Thursday's preliminary open interest. The index is about to open below its floor for the first time since June. QQQ closed Thursday at $715.34 (−1.4%), 6.2% below the 6 October record of $762.86, with Alphabet's 12.6% fall doing most of the damage while 390 of the S&P 500 rose; the pre-market print is $697 (−2.6%) after Amazon (−7% on $75.9 billion of quarterly capex despite AWS +39%) and Intel (−26% on a light fourth-quarter guide), which is under the $700–705 shelf tested five times since June and just above the dealers' $690 put wall. The fund underneath is still in two halves. The average stock is fine: 65% of holdings are above their 200-day average (76% by weight), 53% above the 50-day and the 20-day, the equal-weight basket is ahead of the fund by 3.7% over one month and 1.8% over three, 34 holdings made 52-week highs in the last ten sessions against 14 new lows, and the fund took in $4.3 billion on Thursday while it fell. The problem is the top of the fund: Alphabet, Amazon and the semiconductors (a third of the weight) are the names being sold for their capex, and the fund cannot bounce without them. On Thursday's close the scorecard reads five of ten, forming: the shelf held, short-term breadth is healthy, the McClellan oscillator is just above zero (+2.6), Lam and Applied hold their lows and the group that led the way down (Nvidia, Broadcom, AMD, Micron) is +7% in five days; missing are a reversal candle (Thursday's range was 0.7 times average), participation (51% of holdings up on the last up day), fewer new lows, Nvidia back above its 50-day (Apple is), and the dealers, who are short $10.8 billion per 1% at the pre-market price with the flip at $721, the nearest call gamma at $720 and the call wall at $725: below the flip their hedging sells into weakness. Our read: the set-up is the one SMH printed on Monday, a break of a well-tested shelf on a specific piece of news, and the resolution is the same test. A close back above $705 today would be a failed breakdown on a hyperscaler's print, the strongest reversal signal this scorecard can produce, and the hedging flow reverses at the same time; a close below $700 has only the $690 strike and the 200-day at $679 beneath it, 3% down, in a book that amplifies. Apple (the second-largest holding) reports Monday after the close; AMD on 3 November; Nvidia on 17 November. For a call buyer the trigger is the close above $705 with the semiconductors leading, not the touch of $700; for a put seller, strikes below $670, December or later. Update, Friday afternoon New York: QQQ trades $689 (−3.6%), under the $700–705 shelf and at the $690 strike; re-priced there the book is −$11.0 billion per 1% with the flip at $721 and the put wall now $680 (tiles and gamma chart updated); the 200-day at $679 is the next support and the failed-breakdown case needs a close back above $705. Monday 2 November, pre-open: Friday closed $690.49 (−3.5%); scorecard 3 of 10, book −$7.8 billion per 1%, flip $727, put wall $680. Apple reports tonight, the midterms are Tuesday, the jobs report Friday.
| Signal | Now | What green needs | Status |
|---|
Blue candles closed up, red down (Yahoo daily bars). Orange dashed: the call wall (strike with the most call gamma above the price), put wall (most put gamma below) and gamma flip (below it the dealer book turns short gamma), from CBOE open interest across all QQQ expiries.
Dollars of QQQ dealers must buy or sell per 1% move, by strike (price on the vertical axis, like the candle chart). Blue: gamma from calls (dealers long, they lean against moves); red: gamma from puts (dealers short, they chase moves); black: the net of the two at each strike. Dashed black: today's QQQ price. Hover for the strike and gamma at the cursor.
Dealers are net short about of gamma per 1% move ( in the next 35 days), all of it from puts struck between $550 and $605. The gamma flip at is now 7% above the price, the call wall has moved down to (the strikes above it have lost their gamma as the index fell away), and the put wall at is the last large put strike before $500–520. In negative gamma dealers chase the move: a break of $550 is sold into and can extend quickly; a recovery through $560–575 is bought back just as fast. Friday's monthly expiry removes a large share of these puts, which is the mechanical reason this regime often ends at an expiry. Positioning uses Tuesday's open interest; Wednesday's arrives mid-morning New York time.
Gamma comes from end-of-day open interest and a sign convention (dealers long customers' sold calls, short customers' bought puts); it shifts after each expiry, most of all the monthly ones (16 October, 20 November).
The classic breadth gauge. Above 70% on the 200-day is a healthy bull market; the 50-day swings faster and often turns before the index (a falling line while QQQ rises is a negative divergence). The surge from 8% above the 50-day on 14 September to 100% on 6 October was a breadth thrust; it has unwound to 48% in five sessions, which is a washout, not a divergence: the index fell with it.
Solid: share of holdings. Dashed: share of QQQ's weight. When the weighted line sits well above the count, the big names are in better shape than the typical stock; that is still the case ( of weight above the 50-day against of names): this week the weighted line is below the count for once, because the semiconductors (a third of the weight) are the holdings under their averages while the typical stock is not.
Bars: holdings closing at a new 52-week high (up) or low (down) each day. Line: share within 5% of their 52-week high. Healthy advances show a steady stream of new highs; there have been almost none since the summer, while QQQ itself rallied, because most holdings are still well below their June–July peaks.
Running total of holdings up minus holdings down each day. If QQQ makes a new high while this line does not, fewer stocks are carrying the index.
The 19-day minus the 39-day exponential average of daily net advances (as a share of names trading). Above zero, short-term breadth momentum is rising; a cross below zero after a thrust marks momentum rolling over; readings below about −10, as in late July and mid-September, have marked short-term washouts. It is now.
Both rebased to 100 a year ago, rebuilt from the holdings' daily returns. A rising equal ÷ cap line means the average Nasdaq-100 stock is beating the giants (broad rally); a falling one means the giants are carrying the index. Over the year the average stock won; over the last three months the giants did (equal-weight versus cap-weight), and in the sell-off week the two fell together.
Share of holdings with the 50-day above the 200-day (the slow trend), and with a 14-day RSI above 70 (overbought) or below 30 (oversold). Washouts (many below 30) tend to mark lows; clusters above 70 mark extended rallies. Today no holding is above 70 and only SK hynix is below 30: the pullback has removed the froth without yet reaching an oversold extreme.
The top five holdings are of QQQ and the largest (Apple and Nvidia, level at the top) each; the "effective number" of stocks (one over the sum of squared weights) is , so QQQ behaves like a portfolio of about 30 equal positions, not 100. Weights for the top 25 are the fund's own; the remaining 75 are spread by market capitalisation, which is close to the index's method.
Year to date the fund is up 19 points and the gain has been broad in count but narrow in source: Micron (3.6 points), AMD (2.7), Apple (2.2), Intel (1.4) and Nvidia (1.1) supplied 58% of it, three of the five being the semiconductors that are now falling. Over three months the holdings are up 3%: Microsoft (1.7 points), Palantir, Nvidia, Tesla and CrowdStrike added more than the whole net gain, while Applied Materials, Broadcom, Lam, KLA and Arm took away 2 points, which is what the concentration card shows. Over one month the fund is down 5 points and the five largest detractors (Micron −1.0, Intel −0.9, AMD −0.8, Meta −0.6, Nvidia −0.5) explain 71% of the loss; outside them Microsoft, Apple, Walmart and Costco were up, and the equal-weighted line beat the fund by .
For a QQQ holder: the fund's risk this month is a third of its weight (the semiconductors) plus Meta and Alphabet; the other two-thirds are at or near highs. That is why the index can sit on a floor while its breadth looks healthy, and why the floor's fate is the semiconductors' fate: if they turn, the index has 4% of air above it to the flip; if they make another leg down, the healthy two-thirds will not hold $700 on their own.
| Holding | Weight | Trend | 1W | 1M | 3M | YTD | vs 50d | vs 200d | From 52w high | RSI | 3M contrib. |
|---|
Trend: Leader above the 50- and 200-day and within 5% of its 52-week high; Uptrend above both averages; Pullback above the 200-day but below the 50-day; Recovering above the 50-day but still below the 200-day; Downtrend below both. 3M contribution is the holding's weight three months ago times its return, in percentage points of QQQ's return. SK hynix is measured on its Seoul listing (KRW).
Holdings: the top 25 with exact weights from stockanalysis.com, the rest of the Nasdaq-100 from Nasdaq's constituent list with the remaining weight spread by market capitalisation (as of ), 99 holdings with price history. Daily prices: Yahoo Finance, two years, aligned to QQQ's trading calendar (closes carried over local holidays). Weights through time are rebuilt from today's weights and each holding's price path, so cap-weighted measures move with the market rather than freezing today's mix. Moving averages are simple; RSI is Wilder's 14-day; a new high is a close at or above the highest close of the prior 252 sessions. McClellan uses ratio-adjusted net advances ((advancers − decliners) ÷ (advancers + decliners) × 100). Equal- and cap-weighted indices are rebuilt from daily returns (cap-weighted tracks QQQ closely but ignores fees and rebalancing). Dealer gamma: CBOE delayed chain, recomputed with Black-Scholes, built .
Scorecard thresholds: above the 200-day ≥70% green, ≤40% red; above the 50-day ≥60% / ≤40%; above the 20-day ≥60% / ≤40%; within 5% of a high ≥40% / ≤15%; equal- minus cap-weighted over 3 months ≥0 / ≤−5%; top-five share of the 3-month return ≤60% / ≥100%; RSI above 70 ≤30% / ≥60%; gamma regime positive / negative. These are conventions for a 25-stock index, not optimised signals. The prose is written for the date shown; the charts and tiles update when the page is rebuilt (uv run python -m options_scanner.smh_breadth).