Systems & Timeframes
Breadth and the wider market
A market index is a weighted average, and an average can be dragged upward by a handful of enormous members while the majority of its constituents quietly fall. When that happens the headline says "market at record high" and the actual experience of most investors is a slow bleed. Breadth indicators exist to tell you which of those two stories is true, and they usually know first.
The classic framing: in a healthy advance the generals — the largest, most visible companies — lead and the soldiers follow, so almost everything goes up together. In a late-stage advance the generals keep marching while the soldiers desert: money crowds into fewer and fewer names because those are the ones still working. The index, being cap-weighted, reports only what the generals are doing. Nothing in the price chart of the index can show you the desertion, which is why you need a second instrument that counts the soldiers.
- ADVANCE/DECLINE LINE — a running total of advancing minus declining issues. When the index makes a new high and this line does not, participation is narrowing. It is the most-watched breadth divergence there is.
- NEW HIGHS minus NEW LOWS — how many individual names are making 52-week extremes. A rising index with shrinking new highs is the same warning from a different angle.
- PERCENT ABOVE A MOVING AVERAGE — the share of index members trading above, say, their 200-day average. It converts "how broad is this?" into one number between 0 and 100.
- All three answer the SAME question — how many participants are in this move — so use one, not three. Their value is entirely in disagreeing with the index, and they only do that near turning points.
An index rallies from 4,000 to 4,400 over three months and prints a new high. Over the same period the advance/decline line fails to exceed its own earlier peak, and the count of new 52-week highs falls from 180 a day to 60. Nothing in the index chart is broken — no lower high, no trendline break, no reversal pattern. But the readings say the move is now being carried by a small group. The correct response is not to short an index that is still making new highs; it is to reduce size, tighten stops on the weakest positions, and stop adding new longs. Breadth changes your EXPOSURE long before it changes your direction.
Sentiment measures — the put/call ratio, volatility indices, advisor bullishness surveys — are contrarian tools with a narrow window of usefulness. In ordinary conditions they wander in the middle of their range and say nothing. At genuine extremes, when almost everyone is positioned the same way, they matter for a mechanical reason rather than a psychological one: if nearly everybody has already bought, the pool of remaining buyers is empty, and a market with no marginal buyer falls on its own weight. Note the asymmetry, though — fear spikes are sharp and brief, while complacency can persist for a very long time. Extreme fear is a usable timing signal; extreme greed is only a warning.
Murphy's intermarket work maps the classical relationships: bonds and stocks often move together, commodities move inversely to bonds, and a strong dollar weighs on dollar-priced commodities. These are genuine tendencies with economic mechanisms behind them, and they are also regime-dependent — the stock/bond correlation has flipped sign for years at a stretch depending on whether inflation or growth is the market's dominant worry. Use the map to know WHERE to look for confirmation and what a surprise would look like, never as a mechanical trading rule. A relationship that requires a regime is not a law; it is a hypothesis you should check is still true.
Breadth divergences can run for many months. The advance/decline line has failed to confirm index highs for the better part of a year before anything happened, and traders who shorted the moment they spotted it were carried out long before they were vindicated. Treat breadth as a RISK dial, not a signal generator: it tells you how much exposure a market deserves, and it belongs in the same conversation as position size, not in the same conversation as entry timing.
Breadth requires many constituents, so it exists for stock indices and, imperfectly, for crypto via dominance measures. There is no advance/decline line for a single currency pair. If you trade FX or a single commodity, the analogous questions are answered elsewhere: by the correlated markets on the intermarket map, by positioning data such as the futures Commitments of Traders report, and by whether related pairs are confirming the same move. The principle survives — count the participants — even where this particular instrument does not.
The index tells you the score; breadth tells you the size of the team still playing. Because the index is weighted, it can keep setting records while the underlying market deteriorates — and that gap is precisely the information a price chart of the index cannot contain. This is the same lesson as volume in lesson 13, scaled up from one instrument to a whole market: price is the conclusion, participation is the evidence, and evidence moves first.
An index is at an all-time high, the advance/decline line peaked four months ago, and the number of new 52-week highs is a third of what it was. What position do you take, and what would actually trigger a directional trade? (You take no new longs, trim the weakest holdings and tighten stops — breadth adjusts exposure, not direction. A directional trade needs a price event: a lower high followed by a break of the prior trough on the index itself, which is module 2's definition and can be marked on the chart in advance.)
A cap-weighted index can rise while most of its members fall, so read breadth — the advance/decline line, new highs minus new lows, or percent above a long moving average — as a second opinion the price chart cannot give you. Use one of them, not three. Divergences can persist for months, so let breadth set your exposure while price sets your entries. Read sentiment only at genuine extremes, and treat intermarket relationships as regime-dependent tendencies to check, never as mechanical rules.
Add one breadth chart to your weekly routine — the advance/decline line of the index most related to what you trade is the standard choice — and put it beside the index itself. Every Friday, write one word: "confirming" or "diverging". Nothing else. After three months you will have a record showing how early the warning came and how long it lasted before anything happened, which is the calibration nobody gives you and everybody needs.
Systems & Timeframes