Total crypto market capitalisation is dominated by its two largest constituents. When bitcoin and ether rise, the total rises, and a reader looking only at the headline concludes the market is up. Whether most assets participated is a separate question, and answering it is what breadth measures do.
The simplest is the count: what proportion of tracked assets closed higher. A day where the total gains while a minority of assets advance is a narrow market — capital concentrating rather than expanding. The opposite, a market where most assets rise while the index is flat, is a broadening one, and it usually accompanies a period in which risk appetite is moving down the capitalisation scale. Neither reading is bullish or bearish on its own; both describe where money is going, which the index cannot.
Second-order measures add resolution. Comparing an equal-weighted basket to a capitalisation-weighted one shows the same thing continuously: when the equal-weighted version lags badly, the average asset is being left behind. The proportion of assets above a long moving average distinguishes a broad uptrend from a narrow one. Bitcoin dominance is a crude form of the same measurement, with the denominator problems set out in our piece on it.
The reason breadth is worth tracking is that divergences between it and price tend to appear before changes in character rather than after. A market making new highs on progressively narrower participation is one where fewer and fewer assets are supporting the move, and it has historically been an uncomfortable configuration in every asset class where it has been measured. The converse — an index still falling while the proportion of assets making new lows shrinks — describes selling that is running out of participants.
Two cautions specific to this market. The universe is not fixed: tokens are listed and delisted constantly, and a breadth measure computed over whatever is tracked today is subject to survivorship effects that are much larger here than in equities. And the long tail is thin enough that a large share of the constituents can move on negligible volume, so an unweighted count treats an asset with almost no liquidity as equal to one with deep books. Restricting the sample to assets above a liquidity threshold makes the measure far more meaningful.
The practical use is as a second opinion on the headline. When the index and breadth agree, the reading is straightforward. When they disagree — the total rising on narrow participation, or falling while most assets have stopped making lows — the disagreement is the information, and it is not visible in the number most readers see first.