Only big tech (QQQ) broke out as the S&P 500 stayed range-bound and market breadth narrowed

What changed

Stocks rose this session, but only mega-cap tech broke out of its usual trading range, while the S&P 500 stayed boxed inside a tight one. The gain came from a handful of large tech names rather than broad buying, and fewer stocks took part even as the index climbed.

SPY, the S&P 500 ETF, is in what technicians call a Bollinger squeeze. Its Bollinger Bands, which track how widely price swings, have narrowed inside its Keltner Channel, a measure of the usual trading range around a moving average. That means the market has compressed without choosing a direction. ADX, a gauge of trend strength, reads 10.2, below the 15 level under which we treat the market as trendless.

QQQ tells a different story. It trades 0.5% above the upper edge of its Keltner Channel, outside its usual range. Its Bollinger bandwidth, the width of the band relative to price, is 8.9% against SPY’s 2.8%, so the two indexes have split into a compressed one and a stretched one.

Two breadth gauges show that fewer stocks joined in. RSP/SPY, which compares equal-weight stocks with the cap-weighted index, sits 1.6% below its 20-day average. IWM/SPY, small caps against large caps, sits 1.7% below its own. On a day the index rose, the stocks outside the top tier did not follow.

A rising index and a healthier market were two separate stories this session.

What hasn’t changed

A healthier market would need support from the data, and the inflation and jobs backdrop underneath did not move at all.

• Core PCE at 0.2% month on month – the August reading, released Sep 30; the September figure is due Oct 29.
• Fed target range at 3.75-4.00% – the next decision comes at the Oct 28 meeting.
• Payrolls at 29.0K – the September report, released Oct 2; the next jobs report is due Nov 6. Our labor turning-point scan, three checks for a job market that is breaking rather than slowing, still shows none triggered.

With those anchors fixed, reflation at 31% and stagflation at 30% stayed exactly where they were. Because inflation and jobs held still while the price structure narrowed, we read this session’s concentration as a change from inside the market rather than from the data. That gives the concentration more weight as a clue to the next move than it would usually carry.

Our read

The concentration inside the market is where our read starts. We put the odds that SPY’s squeeze resolves to the downside over the next 20 trading days at 58%. On the surface, big tech breaking out of its range looks like the start of an advance; underneath, fewer stocks hold up the index and a few names carry it.

That estimate comes with a caveat. It rests on only 8 past cases of a trendless squeeze. The sample is small, so the range around the estimate is wide, at 53-63%. The direction matters more than the exact figure: when the squeeze breaks, few stocks are in place to absorb the move.

The assumption that failed this session is that the rest of the market follows when big tech leads. No second line of leaders showed up. Chips dropped out of the lead, and small caps and equal-weight stocks both sat below their averages.

It is too early to call this a collapse, though. Neither breadth ratio made a new low, and our proxy for the McClellan oscillator, a gauge of breadth momentum, improved. This looks like narrowing that has stalled rather than one that is breaking down.

So the implication depends on a condition. If IWM/SPY falls below its 0.3600 breakdown line, no group is left to take over the lead, and a correction is more likely to spread across the market. In that case we think portfolios concentrated in single tech names without hedges would take a bigger hit than the index itself.

The more useful question today is whether anything in a portfolio would cushion a break. The 2-year Treasury yield slipped 0.2bp while the 10-year rose 3.0bp to 5.31%. The long end sold off without any new policy signal.

That long-end selloff pushed the 10-year yield above the S&P 500’s earnings yield of 5.26%. The equity risk premium, the extra return stocks offer over Treasuries, stands at -0.05 percentage points, so owning stocks currently pays nothing extra over owning government bonds.

The idea that bonds cushion stock declines is not working right now either. The 20-day correlation between SPY and long Treasuries (TLT) is +0.6. A positive reading means the two move in the same direction, so long bonds can fall alongside stocks instead of offsetting them.

The counter-evidence is real. CBOE Skew, the price of far-out downside protection on the S&P 500, is easing, and near-term implied volatility still sits below three-month volatility, a calm term structure. The high-yield spread, the extra yield on junk bonds over Treasuries, is 310bp, below the 350bp line where we would start to treat credit stress as moderate. Credit markets are not sounding an alarm.

Bond volatility has already met one of two conditions for a larger stagflation weighting. The MOVE index, the bond market’s implied-volatility gauge, is 113.6. The other condition, a 60-day stock-bond correlation above +0.50, is not met yet. If it is, our framework adds 10 percentage points to stagflation.

For portfolios built on long bonds as a shock absorber for stocks, that +0.50 line is the boundary to watch. If the short-term lockstep extends to the 60-day window, we would read losses on both legs as structural rather than a passing mismatch.

Put simply, the index screen looks like a bull market, but beneath it selling into the rally and unease in the bond market are building at the same time.

Options positioning

• Institutional flow: net premium was negative but stayed inside a neutral band, so it gives no directional read.
• Index downside protection: owned at a high price, while single-stock longs are left largely unhedged.
• Dealer positioning: in SPY, dealer hedging amplified moves in 6 of the last 10 sessions; in QQQ, dealers still dampen moves in most sessions, but that support is thin. This points to wider swings once the range breaks, not to a direction.

Checking our last call

Wider swings are a question for the sessions ahead; several of the lines our last edition was waiting on got an answer in this one.

• September CPI (Oct 14): not yet released.

None of the lines we were watching was crossed. Long-term yields rose but stopped short of the threshold, and the stock-bond lockstep drifted slightly away from its line rather than toward it.

The new September ISM services reading kept expanding but fell short of expectations. Its prices-paid index came in at 74.0, the highest since July 2022, so it did nothing to ease inflation pressure.

The 324bp high-yield spread we cited last time came from an earlier snapshot; the official series puts it lower, at 310bp today. The probabilities of the two lead scenarios are unchanged from our last edition.

Key dates

With the two lead scenarios unchanged, the data that can actually break the tie sits on the calendar.

• What to watch: whether IWM/SPY closes below its 0.3600 breakdown line, which it currently sits 1.6% above

A close below that line would mean no stocks are left to support the concentrated rally, and in our framework it would also take 10 percentage points off reflation.

What it means for investors

While small caps hold that line, we check whether any equity setup has earned real conviction with our six-axis alignment check: six frameworks that each either support owning broad equities, sit neutral, or argue against it.

○ means the axis supports owning equities, △ is neutral, ✗ points the other way. We look for five or more aligned axes before calling a setup a fat pitch, a rare case where most axes line up. Only three support today, and SPY’s own confluence check scores 1/6, so nothing qualifies as a high-conviction setup.

• Dollar-cost-averaging index investors – the signal is whether the Oct 14 CPI shows core prices rising 0.3% or more again. With the equity risk premium below zero and 3-month bills yielding about 4.22%, a new lump sum in the index is a bet on further upside with no extra yield for the risk. Regular contributions are a separate decision; a cautious reader would see the Oct 14 CPI and the Oct 29 core PCE before committing a lump sum.
• Portfolios heavy in big tech – the gauge to track is whether RSP/SPY gets back to its 20-day average. Many investors hold single-name tech longs with little protection, which puts these accounts on the losing side of the equity pain trade, the move that hurts the most investors given how they are positioned. If the squeeze breaks lower, concentrated tech books absorb the first blow. With QQQ’s RSI at 68.4, an extended reading, chasing the breakout has poor odds, and index protection does not extend to single names.
• Long-bond holders – the levels are whether TLT closes below its lower Bollinger band near $76.8, or reclaims about $79.7 alongside confirmed cooling in inflation. Long bonds have stopped acting as a cushion for stocks, and TLT’s RSI of 22.0 is deeply oversold without a macro reason for a rebound yet. If inflation cools and hike pricing collapses, a long-bond rally would become the pain trade on the bond side, so a cautious reader would treat duration as a two-sided risk and wait for the conditions to arrive rather than pick a side early.

As long as big tech’s breakout leaves the S&P 500 inside its range and fewer stocks join the rally, we would not read a rising index as safety for the whole portfolio.

Glossary

• A/D volume – a running tally of whether trading volume leaned toward buying or selling; a negative reading means selling volume dominated.
• Bollinger squeeze – when price bands narrow inside the usual trading range, a sign the market is coiled without a direction.
• Dealer gamma – whether options dealers’ hedging dampens or amplifies price moves; negative gamma amplifies them.
• MOVE index – the bond market’s implied-volatility gauge, the Treasury counterpart of VIX.
• RSI – a momentum gauge comparing recent gains with recent losses; high readings mark stretched rallies, low readings oversold markets.