Nearly Half of Nasdaq-100 Stocks Are in a Bear Market. The Index Is 3% From a Record

Research note. Computed from our maintained daily price database (TradeStation) with official Nasdaq-100 membership, 2011–present, as of July 14, 2026. Methods are in How we checked it below; our editorial and AI standards are in How we use AI.

Start with the two numbers side by side, because each one makes the other strange. The Nasdaq-100 closed July 14 within 3.5% of its all-time high. On the same close, 45 of its 101 members — the ones our database covers — were at least 20% below their own one-year highs. Not a correction: a bear market, name by name. Only 27 members were within 10% of their own highs.

There is no trick in the arithmetic. The index is capitalization-weighted, a few enormous companies are near their records, and the average member is not. Cap-weighting is a feature, not a bug — it’s why the index rides its winners. But it means “the Nasdaq is at highs” and “most Nasdaq stocks are in bear markets” can both be plainly true, and right now they are.

Is half the Nasdaq in a bear market bearish?

My instinct when I first plotted this was the same as everyone’s: it looks like a foundation giving way under a roof that hasn’t noticed. So I ran the test before writing a word of narrative. Take every session since 2011, sort them into quartiles by how many members were in bear markets, and measure what QQQ did next.

The days with the most members in bear markets were the best days to buy, not the worst — nearly a full quarter-percent-per-week edge over baseline at the 63-session horizon. This is the same U-shape we keep finding across our contrarian studies: TICK washouts, bears-strengthening tapes, the Hidden Bear regimes. Widespread damage marks stretched rubber bands more often than it marks tops.

If the essay ended there, the takeaway would be simple: scary chart, contrarian data, buy the fear. It doesn’t end there.

The catch: those precedents were washouts

Look at when the top-quartile days actually happened. August 2011. The 2015–16 earnings recession. December 2018. March 2020. The long 2022 grind, which peaked at 80.9% of members in bear markets on June 16, 2022. In every one of those, the bear share was high because everything was falling — the index included. Buying those days meant buying an index already 15, 25, 30% off its high. Mean reversion had something to revert.

Today the index is 3.5% from a record. The damage is entirely beneath the surface. So the honest question isn’t “what happens after high bear share” — it’s “what happens after high bear share while the index is at highs.” And there, the sample almost vanishes. Since 2011 there have been exactly 22 sessions with 40% or more of members in bear markets while QQQ sat within 5% of its one-year high. Two of them were in October 2011. 20 of them are from this year — February, March, April, and the run we are in right now, which started July 1.

The 2026 instances have resolved upward so far — the February and April readings were followed by the spring melt-up. I’m not going to build a forward-return table on that: a study whose sample is twenty days of the same regime it’s trying to judge, plus two days from 2011, is a pattern, not a statistic. The uncomfortable truth, and I’d rather say it than paper over it: history hasn’t run this experiment. 2026 is running it live.

Why the rally is this narrow

The mechanical answer is concentration: the index’s largest names are near their highs and carry enough weight to hold the roof up alone — tech’s share of the S&P 500 is at a record. It shows up everywhere we look, not just in the Nasdaq. Across our full ~5,500-stock database as of the latest close, the median US stock is 19.4% below its own one-year high, and 49% of all stocks are in bear markets — while the cap-weighted market sits within about 1% of its record. The S&P 500’s own bear share is 32.7%; at publication the Russell-style small-cap cohort stood at 43.6%.

Whether that repairs itself is the thing I watch daily, and there’s a tool for exactly that: our Market Repair Flow nets how many stocks are climbing out of damaged drawdown buckets versus falling into worse ones. As of the latest close it reads Stuck and falling — 14% of stocks moved to a healthier bucket over the past 20 sessions against 17% that got worse. That matters because the one divergence edge we have validated on that dataset points the wrong way: when the index is near highs while repair flow deteriorates, forward returns historically lagged. Not crashed — lagged. The narrowness itself has been contrarian-bullish; narrowness that keeps worsening at the highs has not.

What I’d actually do with this

I would not sell the index because 45 of 101 stocks are in bear markets — that exact impulse has been the losing side of this data for fifteen years. I also wouldn’t wave the chart away as “just cap-weighting doing its job,” because the configuration genuinely is novel and the one validated divergence signal we have is deteriorating rather than healing. What this reading does, practically, is set the burden of proof: the bull case now needs participation to broaden — the bear share falling as the index makes highs — and until that shows up, gains that arrive on five names deserve less trust than gains that arrive on fifty.

What would change my mind

• ConstructiveThe bear share falls below ~35% while QQQ holds or makes highs — broadening, the healthy resolution. Watch it daily on Bear Market Breadth, alongside Repair Flow flipping to Healing.
• BearishThe bear share takes out its 2026 high (57.4%, March 30) while the index stays within a few percent of records — the thinning accelerating at the top, with repair flow still deteriorating. That is the configuration our divergence study penalizes.
• ReconsiderIf 2026 keeps resolving upward and the divergence sample grows into something statistically real, the contrarian read earns the benefit of the doubt even in this shape — and I’ll update this piece with the numbers.

How we checked it

Every number above comes from our own daily computation — here’s what each one means, in plain terms:

• ·“In a bear market” means the stock closed at least 20% below its own highest close of the past 252 trading sessions — its own high, not the index’s. “In a correction” is the same idea at 10%.
• ·The percentile is where today’s share ranks among all daily readings since 2011 — the 90th percentile means it has been this high on roughly one day in ten.
• ·The quartile study sorts all ~3,900 sessions by bear share and measures average QQQ returns over the following 21 and 63 sessions. The windows overlap, so consecutive days aren’t independent observations — treat the edge as directional, not tradeable precision.
• ·One honest limit: we apply TODAY’s index membership across history, because point-in-time membership lists aren’t freely available. That flatters the past (yesterday’s failures aren’t in today’s list), which most likely means historical bear shares are understated — the divergence we describe would look rarer still, not less rare, with perfect data. Prices are as revised today, not real-time vintages.