Market Breadth & Internals

Today’s read posts after the close.

The index is an average, and averages lie by construction. Breadth is the set of questions you ask to find out how the average was produced, and in a concentrated market it is the difference between knowing what happened and knowing what the headline said.

The question breadth answers

An index went up one percent. That single fact is consistent with two completely different markets.

In the first, four hundred of the five hundred constituents rose modestly. Participation is broad, the move reflects something happening across the economy, and the index number is a fair summary of it.

In the second, four hundred and sixty constituents fell and five enormous ones rose enough to carry the whole thing. Participation is narrow, the index number is arithmetically correct and descriptively useless, and most portfolios that day lost money while the headline said the market was up.

Breadth is how you tell which one you are in. Nothing more mystical than that.

The measures worth knowing

Advance-decline is the simplest: how many names went up versus down. The cumulative version, added up over time, is watched because it sometimes diverges from the index for a stretch before the index catches up. It does not always, and the divergences that did not resolve are rarely mentioned by the people showing you the ones that did.

Percentage above a moving average tells you how many constituents are in some kind of uptrend on a chosen horizon. It is coarse but it is honest, and it is hard to fool.

New highs versus new lows counts how many names are making extremes. Expansion at the top of a range and expansion at the bottom mean different things, and a market making new index highs while the number of individual new highs shrinks is the classic narrowing pattern.

Equal-weight versus cap-weight is the one I find most immediately legible. Take the same index, weight every company the same instead of by size, and compare. When the cap-weighted version substantially outruns the equal-weighted one, the largest companies are doing the work. When they converge, participation is broad. It requires no interpretation and no special data.

Concentration changed what breadth means

This deserves saying directly, because a lot of breadth commentary was written for a market that no longer exists.

When a handful of companies make up an enormous share of an index's weight, the index and the average stock can diverge for long stretches without anything being broken. A narrow market is not automatically a sick market, and treating narrowness as an imminent warning has produced a great many premature calls over the past decade.

What narrowness reliably tells you is something about risk rather than about direction: fewer things are holding the thing up, so the outcome depends on fewer things. That is a genuine and useful statement. "Therefore it will fall soon" is not a genuine statement, and I have watched people lose money for years insisting that it is.

How I actually use it

As context, and mostly as a check on my own conclusions.

If I have a constructive read on the market and breadth is narrow, that does not cancel the read, but it does tell me the read depends on a small number of names continuing to work. That is a different bet than the one I thought I was making, and knowing which bet you are actually in is most of risk management.

If breadth is broad and improving, moves tend to have more follow-through, because more participants are involved in producing them. If it is narrow, single-name news can move the whole index, and index-level analysis gets noisier.

It is also the honest way to read an earnings season or a macro reaction. As covered on earnings season, "the index had a good quarter" and "companies had a good quarter" are separate claims, and breadth is the instrument that separates them.

Where breadth misleads

Divergences last far longer than anyone expects. This is the single biggest practical failure. A breadth divergence can persist for many months, and a trader who acts on the divergence itself is early in a way that is indistinguishable from wrong for long enough to be expensive.

It is a condition, not a trigger. Breadth describes the environment. It does not tell you when, and every attempt I have seen to turn it into a timing tool has quietly relied on something else for the timing.

Composition changes distort the history. Index membership changes, sector weights shift, and a breadth series measured across decades is not measuring the same object throughout.

And it can be trivially cherry-picked. There are enough breadth measures that on any given day, some of them are diverging. Choosing the ones that agree with a view already held is easy, common, and almost never labelled as such.

The one-sentence version

Breadth tells you how many things are carrying the market, which tells you how fragile the outcome is. It does not tell you when anything will happen, and any version of it that claims to is doing something other than measuring participation.

The mechanics of the individual measures, including how to read the internals day to day, are in market breadth and internals. How leadership rotates between groups while the index sits still is on sector rotation.

Educational market analysis only. Nothing here forecasts the market or recommends any position.

Questions traders ask

What is this Market Breadth & Internals page?

A dated, running read on market breadth & internals from a trading desk that scores its own calls publicly — hits and misses both.

How often is it updated?

After the market close on trading days.

Is this investment advice?

No — educational market analysis only. Nothing here is a recommendation.