How to Use the Money Flow Index
To use the money flow index, read it as a volume-weighted version of the relative strength index. It is bounded between zero and one hundred, so extremes mean the recent range has been one-sided in both price and volume, which is a statement about the past rather than the future.
The money flow index runs the relative strength index calculation on price multiplied by volume rather than on price alone. The output is bounded between zero and one hundred, and every property of a bounded oscillator applies to it unchanged.
Before you start
A volume feed you trust, because volume is half the calculation. In a fragmented market the reported figure is partial, which makes the weighting partial too.
A decision about whether the market is ranging or trending, made before any reading. The same value means exhaustion in one and strength in the other.
An understanding that it is the relative strength index with volume weighting. If volume were flat, the two would produce the same line.
The steps
1. Read it as a bounded position measure
A high reading means recent movement has been one-sided in price and that the one-sided days carried volume. It is a description of what happened, not a forecast.
2. Classify the market first
In a trend the reading sits above eighty or below twenty for long stretches. On this site’s shared series direction runs average 2.01 bars and the longest ran 11, so persistence is ordinary.
3. Set the lookback from your horizon
Fourteen bars is a fortnight of daily data. If you hold for months, the reading is answering a question about a shorter period than your trade.
4. Compare it against the unweighted version
They agree most of the time. The disagreement is the only thing this indicator adds, and it is usually caused by one or two unusual-volume sessions.
5. Define divergence before looking for it
How many bars, how large a gap. Any two series disagree somewhere over any window, so without a rule the pattern is always available in hindsight.
6. Require structure before acting
An extreme reading at a level price has already respected is a different event from one in open space, and the indicator cannot tell them apart.
7. Take the stop from the chart
Where the idea is wrong is structural. That distance sets the position size, and the reading plays no part in either decision.
How to tell it worked
The market type was decided before any reading was interpreted.
0 trades came from a reading alone, without a structural setup.
Divergence was defined in bars before the chart was examined.
And the chart carries 1 oscillator of this family, not this one alongside the unweighted version.
What the volume weighting buys
A small amount of extra information and a meaningful amount of extra noise. Volume is more erratic than price, so weighting by it makes the line jumpier without making it more decisive.
In an illiquid instrument the weighting actively hurts. A handful of trades can dominate a session’s volume, which then dominates the reading, so the indicator amplifies exactly the data you would want to discount.
When it is worth having over the plain version
On instruments with a complete, reliable volume figure. Centrally traded futures and single-listing equities qualify; fragmented equities and spot currency largely do not.
When you are specifically interested in whether volume supported the move. That is a narrow question and it is the only one the weighting answers.
Not as a second confirmation on a chart that already has the unweighted version. They agree the overwhelming majority of the time, so having both means a signal is almost always confirmed — which sounds like rigour and is duplication.
What the calculation actually does
It starts from the typical price of each bar — the average of high, low and close — rather than from the close alone. That alone makes it slightly different from the unweighted version before volume enters at all.
Then it multiplies that by the bar’s volume to get a money-flow figure for the bar, and sorts those figures into positive and negative depending on whether the typical price rose or fell.
Then it runs the same ratio the relative strength index runs. Positive flow over total flow across the lookback, scaled to a hundred.
Which means two different substitutions have happened: typical price for close, and volume-weighted for unweighted. When the two indicators disagree it is worth knowing that either change could be responsible, and the display gives you no way to tell which.
The original data
Of the 24,971 unique videos in research/search-study-corpus.jsonl, 44 mention this indicator in the
title, at a median of 3,160 views across 36 channels, and 59% of those titles are instruction-shaped.
The relative strength index appears in 154 instruction-shaped titles at 4,398 and on balance volume in
26 at 15,517. The counts come from site/corpus_count.py and site/rank_howto.py.
44 videos at a 3,160 median — a smaller audience per video than either of its neighbours. A hybrid indicator generally attracts less attention than either parent, which is what the numbers show here.
The answer to the question on that chart is that the gap between them is a volume fact. One or two sessions carried unusual volume on the up side — which is worth knowing and is not a reason to prefer one reading over the other, because they are answering slightly different questions.
When it fails
The failure is a trend read with range settings, and it is the same failure every bounded oscillator produces. The reading pins above eighty, every bar looks more extended than the last, and price continues. The volume weighting changes nothing about this — it can make the pinning more emphatic, because a strong trend usually carries volume, so the indicator most likely to look extreme is the one in the market least likely to reverse.
The second failure is trusting it on a poor volume feed. Half the input is wrong.
A third is running it with the unweighted version. They mostly duplicate.
A fourth is finding divergence by looking. It is always available.
A fifth is using it on illiquid instruments. The weighting amplifies noise.
And a sixth is stopping from the reading. It contains no prices.
Related
Money flow index covers the calculation. Relative strength index is the unweighted version. And volume analysis sets the ceiling on what the weighting can add.
Running this next to the relative strength index taught me the most about it: they agree almost all the time, and where they disagree is usually a day with unusual volume rather than a change in the market. That is a narrow amount of extra information for a second oscillator on the chart.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.