Chaikin Money Flow: volume with the nuance the OBV lacks

The CMF exists to fix one specific flaw in the OBV: that it adds the entire volume merely by looking at whether the close rose or fell. The CMF weights. It does the job better, and in exchange it has a blind spot of its own worth knowing before you trust it.

How it is calculated, in two steps

The first is the money flow multiplier, and it looks inside a single candle:

((Close − Low) − (High − Close)) ÷ (High − Low)

It is +1 if price closes right at the high, −1 if it closes at the low and 0 if it closes exactly mid-range. The diagram above shows it with four candles identical except for the close.

The second step multiplies that number by the candle's volume — that gives the money flow volume — then sums the last twenty and divides by the sum of those twenty volumes. The result is the CMF: a number between −1 and +1 answering what share of the last twenty candles' volume traded with price closing in the upper part of its range.

The blind spot: it does not look at the previous candle

The multiplier is computed inside each candle. It knows nothing about where the previous one closed.

The consequence is the diagram's second panel: a candle that falls hard and then recovers to close at its own high scores +1, maximum buying pressure, even though its close sits below yesterday's. To the CMF it was a buyers' candle. To whoever held the position it was a losing day.

That is not a bug, it is what it measures: pressure within the session, not direction between sessions. But it explains why the CMF can sit in positive territory through a stepped decline, and why it does not work as an entry signal on its own.

Do gaps matter in crypto?

It is only fair to the indicator to say this. The diagram's extreme case — an opening gap — is typical of equities, where the market closes overnight and price reopens wherever it likes.

In crypto the market never closes: it trades seven days a week, so between one daily candle's close and the next one's open there is no gap at all, except on very illiquid pairs or during an exchange outage. On Bitcoin or Ethereum a literal gap is extremely rare.

What does not disappear is the underlying problem, which is broader than gaps: the multiplier still never compares with the previous close. A candle that opens flush with yesterday's close, collapses eight per cent and recovers to its high still scores near +1 with its close far below yesterday's. That happens constantly in crypto.

The indecisive candle that counts for nothing

There is a second effect, less known and sometimes more annoying. If price closes exactly mid-range the multiplier is zero, and then that candle's volume vanishes from the numerator entirely.

A session with enormous volume and an indecisive close — a capitulation doji, say — does not move the CMF at all, even though it is probably the most informative candle of the week. The denominator does include that volume, so the real effect is that it dilutes the indicator: it pushes it toward zero.

The same goes for wicks. Since the range is defined by the high and the low, a long wick widens the multiplier's denominator and lowers the reading even when the candle's body is convincing.

How to read it

The theoretical scale runs from −1 to +1, but in practice the CMF rarely strays beyond the ±0.30 band. That is why the usual thresholds sit at 0.20 rather than 0.50: it is not an arbitrary convention, it is where the indicator actually lives.

CMF, OBV and MFI

All three try to read the same thing and differ in where they put the detail.

That the CMF is bounded is a practical advantage over the OBV: it can be compared across assets and across eras as it stands. The OBV's running total cannot, because it starts from an arbitrary point.

Common mistakes

Taking the zero crossing as a signal. In a range the CMF crosses zero again and again. If the BBWP says volatility is compressed, a zero crossing means nothing.

Shortening the period so it reacts sooner. At 5 or 7 candles the CMF becomes nearly as erratic as price itself, and then it adds nothing price had not already said.

Mistaking it for a measure of direction. The CMF can be positive while price falls, and that is not a failure: it measures where candles close within their ranges, not whether price rises.

Using it on bad volume data. Everything it measures comes from reported volume. On small platforms with inflated figures, the CMF inherits the whole mess.

On each asset

On Bitcoin the CMF works better than on almost any other crypto for a prosaic reason: its volume is the cleanest. On Ethereum read it beside dominance, because strong buying pressure in ETH with dominance rising usually means general market flow. On XRP treat its readings more sceptically: its volume episodes concentrate in a few news-driven sessions, and a twenty-candle average is easily contaminated.

Frequently asked questions