Accumulation Distribution vs Chaikin Money Flow

Michael Whitman · Trading since 2017, full-time since 2020 · Published 2026-09-08 · Updated 2026-10-19 · Reviewed 2026-10-04

Two indicators built by the same man from the same weighting. One adds it up indefinitely; the other sums it over a window and divides by volume, which turns a total into a reading with a scale. On SPY’s daily bars the arithmetic link is exact, and the useful question is what the division by volume does.

What each one is

The accumulation distribution line accumulates weighted volume. Each bar’s volume is multiplied by where its close sat in its range, from +1 at the high to -1 at the low, and added to a running total. Accumulation distribution covers it. StockCharts notes Marc Chaikin first called it the Cumulative Money Flow Line.

Chaikin money flow sums the same weighted volume over 20 sessions and divides by the total volume of those 20 sessions, giving a reading between -1 and +1. Chaikin money flow covers it.

So the money flow is the line’s 20-session change, rescaled. The sum of the last 20 weighted volumes is exactly how far the line moved over those 20 sessions, and volume is never negative, so the two always share a sign.

Where they differ

Whether the level means anything. A running total starts from an arbitrary point, so only its direction carries information. The bounded version has a scale you can set a threshold on.

How much history is carried. Everything, against 20 sessions. The total remembers a year ago; the money flow has forgotten it entirely.

What the division by volume does. A 20-session rise of a few million shares in a quiet year and one of a billion shares in a crisis are very different line moves. Divided by their own volume, the quiet one can read higher.

What each is read for. Slope and divergence on the total; threshold crossings and divergence on the bounded one.

Which one to use: Accumulation Distribution or Chaikin Money Flow?

Use the money flow when you want a threshold or a comparison across years. A rule that says above or below a level needs a scale, and only one of these has one.

Use the running total when the long sweep is the question. Whether weighted volume has been accumulating for a year is something a 20-session window cannot report.

Use a buffer if you trade the money flow’s zero line. StockCharts suggests +0.05 and -0.05, and on SPY that cut the number of signals by more than half.

And run one, not both. Their 20-session agreement is arithmetic, not confirmation.

The same sign, every session

On SPY’s daily bars the identity held without exception. For every session from 31 January 1994, the first with 20 sessions of the line behind it, to 2 October 2026, the money flow was positive exactly when the line had risen over the previous 20 sessions and negative exactly when it had fallen. The two calculations differed by no more than rounding in the fourteenth decimal place.

So a chart showing both “confirming” each other over a month is showing one number twice. The line can still disagree with the money flow about longer stretches: a line that rose for a year and a money flow below zero describe one situation, a long build with a weak last month.

What dividing by volume changes

SPY’s highest money flow reading was 0.61, on 8 February 1996. Over the 20 sessions behind it, the line rose by only 6.0 million shares, because only 9.8 million shares traded in all of them.

SPY’s biggest 20-session rise in the line came on 19 December 2008: 1,732.1 million shares, in a period when 8,012.1 million traded. The money flow that day was 0.22, the highest of 2008 but only the 1,626th highest reading of the whole period.

Neither reading is wrong. The line measures shares; the money flow measures the share of the window’s volume that closed toward the highs. Over a period in which SPY’s median daily volume went from 0.73 million in 1996 to 256.7 million in 2008 and 68.1 million in 2025, only the money flow can be compared across years at all.

What the shared blindness to gaps costs

A gapped bar contributes almost nothing. The weighting looks at the close’s position inside the bar’s own range, so a session that opened far away and closed in the middle is treated as neutral.

StockCharts gives the example itself: a stock that gaps down and closes well lower can still lift the line, if the close sits above the middle of that day’s range.

On-balance volume does see gaps. It compares each close with the previous close, so it registers exactly the event these two miss, which is why it is the more useful second opinion.

The original data

The sample. SPY’s daily high, low, close and volume as Yahoo Finance reports them, 3 January 1994 to 2 October 2026; the running total starts from zero before the first of those sessions. Multiplier = ((close - low) - (high - close)) / (high - low), zero when a bar has no range; money flow volume = multiplier times volume; the line is the running total; the money flow is the 20-session sum of money flow volume divided by the 20-session sum of volume, all as StockCharts defines them.

Readings. The money flow was above zero on 5,672 sessions (68.98%). It crossed zero 747 times over 32.67 years, 22.87 a year. Counting only moves from above +0.05 to below -0.05 or back, the buffer StockCharts describes, there were 299 crossings, 9.15 a year, and the reading sat inside the buffer on 1,833 sessions (22.29%). Lowest reading: -0.39 on 20 December 2000. Biggest 20-session fall in the line: 2,657.9 million shares on 9 March 2009, money flow -0.32.

Every session with its money flow volume, line, 20-session change and money flow is in the SPY money flow file, and each year’s peaks in the yearly peaks file.

When it fails

The failure is treating agreement between them as a signal, and it is available constantly. A chart with the line in one pane and the money flow in another will show them rising and falling together over every 20-session stretch, because they must. A trader who counts that as two indicators agreeing has doubled the confidence and added no information, and sizes the position as if two independent readings had lined up. The disagreements that do appear are about time, a long total against a short window, and reading those as conflicting signals produces a fresh one whenever the last month differs from the last year.

The second failure is quoting the total’s level, a number set by whichever day the count began.

A third is comparing the line’s moves across decades. Volume changed by hundreds of times; only the money flow is scaled for that.

A fourth is using either on an instrument that gaps often. Both ignore the gap itself.

A fifth is using either on a partial volume feed. The input is a fraction of the market.

And a sixth is hunting for divergence after the fact. Some stretch of chart will always supply one.

Related

Accumulation distribution covers the running total and how it is built. Chaikin money flow covers the bounded version and its lookback. And volume analysis covers what the volume input can and cannot tell you.

This page is educational, not financial advice. Test every idea on your own charts before risking money.