The Market Breadth Manual

Research note. A topic Manual — one subject, several gauges. Every figure computed at render from our daily breadth dataset (~4,750 eligible US common stocks in a typical session), the same files behind the live gauge pages. The single-gauge deep dives live in their own Manuals: the A/D line, the McClellan Oscillator and the Hindenburg Omen.

What is market breadth?

The index level is one number. Breadth is the question behind it: how many stocks came along? A market where 400 names rise is a different animal from one where the index is green because five mega-caps are, and the rest are quietly rolling over. Breadth is how you tell those two apart — built from the count of advancing versus declining stocks, new 52-week highs versus lows, and the share of stocks above their moving averages. The mistake most people make is treating breadth as one indicator. It isn't. It's three layers, and they answer different questions.

Is the market broad right now?

As of 2026-08-05: 56% of stocks are above their 50-day average and 55% above their 200-day — a majority participating. The last session's net advances printed -738. The Hindenburg Omen is not signalling (4.5% of issues at new highs vs 1.3% at new lows). Every number here updates after each close; the MA-breadth tracker and A/D line carry the full histories.

The three layers — read them in order

Layer 1 — Participation: is the move broad or narrow? Start with the base state. Moving-average breadth is the cleanest read on how many names are in uptrends. The advance-decline line tracks the daily tug-of-war and, cumulatively, whether participation keeps up with price (its Manual covers the divergence record and the cap-cohort decomposition). New highs vs new lows is the leadership read: fresh highs are breakouts, fresh lows are breakdowns.

Layer 2 — Momentum: is breadth accelerating or fading? Participation is the level; momentum is the direction of change. The McClellan Oscillator — two EMAs of net advances — turns up before the level does at real bottoms. The rare, powerful version is the thrust: breadth flipping from washed-out to broad buying in days. The Zweig Breadth Thrust and 90% up-volume days are the textbook triggers; a 90% up day right after a 90% down day is the classic capitulation-then-demand bottom.

Layer 3 — Divergence: is the move lying to you? Where breadth earns its keep: price and participation disagreeing. An index grinding to new highs on shrinking new highs and a flat A/D line is a narrowing market. The Hindenburg Omen formalizes one specific split — many new highs and new lows at once, inside an uptrend. Divergences don't time tops; they tell you the cushion under the index is thinner than it looks.

The breadth gauges compared

Each gauge answers a different question and fails in a different way. Read across the row — especially the last column — before acting on any single one.

Participation thresholds

Rough bands for the share of stocks above their 50-day average — orientation, not hard lines; the level and its direction both matter. Today's band is marked.

Every breadth thrust since 2010, with what followed

The strongest claim in breadth folklore is the thrust, so here is our record — every Zweig Breadth Thrust on our universe since 2010, with SPY's return over the following 3, 6 and 12 months, computed at render.

Across all 6 completed windows the 12-month return averaged +14.9%, positive 6 of 6 times. The pattern matches the signal's reputation — thrusts cluster near the start of durable advances — but a handful of observations is a small sample, and one bad outcome would move the average a lot. That is exactly why breadth is context, not a stand-alone trade.

The reading routine

• If breadth is diverging from price, check whether it's structural (the small-cap tail) or genuinely broad before calling a top — the A/D Manual's cohort decomposition exists for exactly this.
• If a thrust just fired, confirm it launched from a real washout (sub-30% participation, a recent 90% down day) — not a bounce inside a range.
• If the Hindenburg Omen prints, look for a cluster in an uptrend — the full cluster table shows single signals mostly fade.
• If new highs and new lows are both expanding, read it as a split, indecisive market — not an outright sell.
• Match the gauge to your horizon: McClellan and 90% days for swing turns; MA breadth, the A/D line and thrusts for weeks-to-months trend confirmation; the Coppock Curve and persistent divergences for regime shifts.

Where breadth will mislead you

• Divergences persist. Breadth can narrow for months while the index keeps rising. A divergence is a thinner cushion, not a sell date.
• Participation can stay elevated into a top. MA breadth held above 50% well into past distribution phases — direction matters more than level.
• Mega-cap masking cuts both ways. A cap-weighted index can look fine while the median stock is in a downtrend — that's the case for breadth, but it also means the index won't confirm the warning until late.
• Universes differ. Our counts run on ~4,750 US common stocks; NYSE-issue versions include bond funds and preferreds. Compare each gauge to its own history, never across sites.
• No price targets, no stops, no macro. Breadth tells you whether to trust the trend, not where it ends — pair it with the Bubble Tracker (where breadth is the confirmation modifier) and the volatility complex.

How we checked it

All breadth gauges are computed daily from our maintained price database across the US common-stock universe (~4,750 eligible names in a typical session; ETFs, preferreds and index pseudo-tickers excluded via the share-class snapshot). The thrust table lists every Zweig trigger since 2010 with SPY forward returns from each thrust date, graded only where the window is complete. Live figures on this page recompute at render from the same JSONs the gauge pages chart, so the Manual can never disagree with the tools. Full per-gauge methodology lives on each gauge's page and Manual.

Frequently asked questions