Market Breadth Indicators: A Health Check on a Rally, and Its Limits

Market breadth measures how many stocks are taking part in a market move, rather than how far the index travelled. A rally where most listed stocks are rising is described as broad. A rally driven by a handful of large names while the majority fall is described as narrow. Breadth indicators are the arithmetic that turns that description into a number.

Breadth is genuinely useful, because a capitalisation-weighted index can rise on very few shoulders and give no hint of it. Breadth is also one of the most over-read families of market data in circulation, because the numbers are easy to compute, easy to chart, and very easy to narrate after the fact.

Why the index alone hides participation

India’s headline indices are weighted by free float market capitalisation, meaning each constituent’s influence is proportional to the value of its freely tradable shares. The construction is explained in how Indian indices are constructed and free float market cap explained.

A direct consequence is that the largest few constituents can move the index a long way on their own. If they rise while most other constituents fall, the index prints a gain that describes almost none of the market. Breadth measures exist to make that visible.

The main breadth indicators

Advance decline counts. The simplest measure: how many securities closed higher than the previous close, how many closed lower, and how many were unchanged. From these come the advance decline ratio and the cumulative advance decline line, both covered in detail in advance decline ratio explained.

New highs versus new lows. A count of securities making a fresh 52-week high against those making a fresh 52-week low. This is a slower measure than daily advances and declines, because reaching a one-year extreme requires a sustained move rather than a single session. Practitioners often read the net figure, new highs minus new lows, as a cumulative line.

Percentage above a moving average. The share of a defined universe trading above its own 50-day or 200-day average price. It converts “how many stocks are in an uptrend by this definition” into a single percentage between zero and one hundred. The choice of window is arbitrary, and the indicator changes character when you change it.

Up volume versus down volume. Instead of counting stocks, this sums the traded volume in advancing stocks against the traded volume in declining stocks. It weights participation by activity, which reduces the influence of tiny, barely-traded scrips that a simple count treats as equal to the largest company on the exchange.

Equal weight versus capitalisation weight. Comparing an equal-weighted version of an index to its standard weighted version is a clean, indirect breadth measure. If the equal-weighted line lags badly, the weighted index is being carried by its largest members. This connects to the size effects discussed in what is the size factor.

Share of constituents beating the index. For a defined index over a defined window, the percentage of members that outperformed the index itself. It answers the practical question of how hard it would have been to keep up by picking names at random from that list.

Where the inputs come from in India

No exchange publishes an official “breadth indicator”. What the exchanges publish is the raw material.

• Daily security-level price files. NSE and BSE each publish end-of-day files carrying open, high, low, close, previous close, and traded quantity and value for every security that traded that day. Every breadth measure above is computable from these.
• Daily market summaries. Both exchanges publish daily reports that include advance, decline, and unchanged counts for their traded universe, along with counts of securities hitting upper and lower price bands.
• 52-week high and low lists. Published as part of the daily reporting, and also derivable from the price history.
• Index constituent lists and weights. Published by the index providers associated with the exchanges, and revised at scheduled reviews. These define the universes that most professionally quoted breadth measures are computed on.

Because the indicator is computed rather than published, two sources can report different values for the same day. Almost always the difference is the universe, not an error.

How breadth is read

The habit that makes breadth useful is reading it against the index, not on its own.

When the index and breadth move together, breadth adds little. The interesting configuration is disagreement: the index rising while participation narrows, or the index falling while the number of declining stocks steadily shrinks. Practitioners call the first a narrowing advance and the second a contracting decline, and both are descriptions of composition rather than predictions.

The second habit is fixing the universe first and never changing it mid-analysis. Breadth computed on a large-cap index, on a broad 500-stock index, and on every scrip listed on an exchange will tell three different stories about the same day, and all three can be correct about their own universe.

The third is choosing an aggregation window. Daily breadth is extremely noisy. Ten-day and one-month averages of the same series behave far more like a description of a market phase.

What it does not tell you

Breadth data is over-read at least as consistently as flow data, and often in the same paragraph. The limits are structural.

• It is a description, not a forecast. Narrow breadth is a fact about the period that just ended. Narrow conditions have persisted for very long stretches and have also resolved quickly. Nothing in the arithmetic distinguishes those cases in advance.
• The universe decides the answer. Include the long tail of thinly traded and rarely traded scrips and the counts are dominated by securities almost nobody can transact in size. Restrict to an index and you have measured that index, not the market. Neither choice is wrong, but the choice is doing most of the work.
• Price bands and illiquidity distort counts. Securities in restricted trading categories, those subject to narrow price bands, and those that trade in a handful of shares still count as one advance or one decline, exactly like the largest company on the exchange.
• Equal counting ignores size. A simple advance decline count treats a company with a market value in the thousands of crores and a shell-sized microcap identically. Volume-weighted breadth reduces this but introduces its own distortion, because turnover concentrates in a small number of names.
• New listings and delistings change the base. The number of tradable securities is not constant. A cumulative line built over many years is accumulating across a changing universe, which is a milder version of the survivorship problem discussed in survivorship bias in backtests.
• It contains no information about businesses. Breadth is a count of price changes. It cannot tell you whether earnings are improving, whether cash conversion is deteriorating, or whether valuations are demanding.
• It is easy to narrate and hard to test. Because breadth throws off dozens of variants, the temptation to find the one that explained the last turn is enormous, and it is the same trap set out in what is overfitting in backtesting.

The honest summary is that breadth answers one question well: was this move shared or concentrated. That is a real and often overlooked fact about a market period, and an index level alone will never tell you. Treat it as a composition check on a move that has already happened, and it earns its place. Treat it as a timing device and it will supply an explanation for everything and a warning for nothing.

Related reading

• Portfolio metrics explained: the hub for the risk, return, and market analytics in this series.
• Advance decline ratio explained: the oldest breadth measure, computed and read step by step.
• FII and DII flows explained: the other market-level series that gets read as a signal and is really context.
• How Indian indices are constructed: why weighting decides how much the index can hide.
• Sector rotation strategy in India: where narrow and broad participation shows up as sector level dispersion.

This article is educational. Altys Labs is not a registered research analyst or investment adviser, and nothing here is investment advice or a recommendation to buy, sell, or hold any security.