To The Moon

Nasdaq at a record. How many stocks are actually driving it? A breadth check through my sell-put lens

Written Wednesday morning SGT, using Tuesday's US close.

The headline

On Monday, the Nasdaq Composite gained 2.26% for a record close of 27,122.09, its first since June. It did it again on Tuesday: the tech-heavy index gained 0.45% to 27,244.28, while the S&P 500 closed relatively flat at 7,764.64. The Nasdaq-100 crossed the 30,000 level on September 21.

That's what the index did. The stocks inside it tell a different story.

The breadth data

On Monday, the S&P 500 had seven new 52-week highs and 24 new lows. On the Nasdaq, there were 64 new 52-week highs and 127 new lows. Put simply, the index jumped while about twice as many Nasdaq stocks hit fresh 52-week lows as highs.

It goes deeper. Nearly 23% of the S&P is at a one-month low and more than 12% is at a three-month low. Only a minority of stocks are even in a short-term uptrend. S&P 500 members above their 50-day moving average make up 30.4% of the index, which ranks in the 14th percentile, an unusually weak reading relative to long-term history. Two out of three stocks are below their 50-day average while the index sits near a record.

There's also a historical stat that has been going around. Analyst Jason Goepfert found only two days in history when the S&P rallied at least 1% to within 1% of a new high while more of its stocks fell to new lows than new highs. The prior two were July 23, 1929, and December 21, 1999. Those dates sound scary. But two data points are a warning light, not a forecast.

So who's actually lifting the index?

Mostly chips. On Monday, Intel and Arm Holdings surged over 12%, and the PHLX semiconductor index jumped 4.3%. AMD closed 10% higher to reach a market capitalization of $1 trillion for the first time, and Qualcomm increased more than 9%.

On Tuesday, the leadership was even more concentrated. Of the SOX's 30 constituents, 27 advanced and 3 declined. Memory led the gains: SanDisk rose 6.82%, Micron gained 5.00%, Seagate climbed 4.85%, and Western Digital rose 3.67%.

Big Tech was split. Amazon fell 1.34%, Google dropped 1.05%, Microsoft slipped 0.72%, and Meta declined 0.63%. Outside tech it was worse. Financial stocks took 286 points off the Dow.

So the honest answer to the question in the title is: a group of semiconductor and memory names, plus a few mega-caps. That is a narrow base for a record.

The other side of the argument

It would be biased to only show the scary numbers. There's a real bull case here too.

CNBC's Mike Santoli notes that the S&P 500 has gone nowhere since June 1 but hasn't slipped more than 3% from a record high over that span. He also points out that market breadth has eroded, but that also means internal oversold conditions are building. In other words, the laggards have already been sold, and money could rotate into them.

Positioning helps too. Goldman Sachs' John Flood argued that light positioning by professionals, elevated short interest in Nasdaq stocks and persistent earnings momentum skewed the risk-reward higher for the market into year-end. Monday's burst partly looked like shorts getting squeezed. TheStreet's Rev Shark put it down largely to poor positioning, with the reaction outsized because so many people were leaning the wrong way.

So both things can be true. A narrow rally isn't a crash signal. It's a fragility signal. The index can keep rising, but it rests on fewer stocks, and if those few crack, there's not much underneath to catch it.

The calendar doesn't help. The week after September options expiration is one of the weakest seasonal stretches of the year. Xi Jinping's US visit is on Thursday, and Micron reports on September 30.

How I use this in my sell-put strategy

I sell cash-secured puts on individual US stocks, not on the index. So a Nasdaq record by itself tells me almost nothing. What the breadth data tells me is how careful to be. Here's how each part of my process applies to a narrow tape like this.

1. Judge the stock, not the index.

My hard structural gate is that a stock must trade above its own 200-day moving average. When two-thirds of the S&P is below its 50-day, far fewer names pass my checklist, even with the index at a record. I don't loosen the rule because the headline looks bullish. If a name fails the gate, I don't sell a put on it.

2. A low VIX doesn't mean low single-stock risk.

The VIX closed Monday at 14.87. That's the options market pricing the index as calm. But in the same two days, Intel moved 12%, AMD 10%, Meta 11% and SanDisk nearly 7%. When stocks move in different directions, they cancel each other out in the index, and the VIX stays low while individual names swing hard. That's why I check the volatility gates on each ticker separately: IV Rank of at least 40%, IV Percentile of at least 45%, and implied volatility at least 1.2 times historical volatility. Some names will have rich premium this week. Many won't.

3. Count the AI names as one trade.

Most of my watchlist is some form of the AI trade: NVDA, AMD, CRWV, PLTR and others. In a broad market, those are separate positions. In a market where the index depends on one theme, they are effectively a single bet. If the chip leadership breaks, they will likely fall together. So I look at total theme exposure, not just position size per ticker. Five 2% positions in correlated names is closer to one 10% position than five separate ones.

4. Respect the earnings calendar.

My rule is no new position within 7 calendar days of earnings. Micron reports on September 30, so it's blocked from today onward. It doesn't matter that it led Tuesday's rally. The same applies to anything whose 30 to 45 day expiry would run into earnings.

5. Sizing, strikes and exits don't change.

Position size is capped at 5% of the account for my core names (NVDA, AMZN) and 2% for everything else.

The strike is the lower of two anchors: the one-standard-deviation expected move, or the nearest key moving average times 0.98. Delta must fall between 0.15 and 0.25, with expiries 30 to 45 days out.

Every position gets a 50% take-profit order the moment it's filled.

Stops are at 2 times the premium for core names and 2.5 times for the others, with one roll at most.

In a narrow tape, the leaders are often stretched far above their moving averages. As I showed in my Meta post, that pushes the moving-average anchor so far down that the strike doesn't fit my delta range. That's the chart telling me to wait, not to cut corners.

My verdict: selective, not aggressive

A narrow record high isn't a reason to sell more puts. It's a reason to sell fewer, better ones. This week, that means:

only stocks that pass every gate on their own chart,

no Micron before earnings,

not stacking correlated AI names,

keeping extra cash in the account in case the chip leadership stumbles.

If breadth improves, with more stocks back above their 50-day and new highs outnumbering new lows, I'll widen the net again. Until then, I'd rather miss a bit of premium than find out the hard way how much of this rally was carried by a few names.

How are you reading this? Is it healthy rotation, or a market running on a few stocks? Let me know in the comments.

This is my personal trading framework, shared for education. It is not investment advice. Options carry significant risk, including assignment and losses larger than the premium collected. Always verify live chain data before trading.

$QQQ $SOXX $NVDA $AMD $MU

Before you post, check two things:

Point 3 (counting AI names as one trade). I wrote the theme-exposure point in your voice, but it isn't one of your written v8.0 rules. Edit it if you'd rather not present it as your own practice.

The Goepfert stat. It comes from TheStreet's commentary. Cite it as his finding, not as your own backtest.