Volumes MomentumVolumeMean Reversion Available in Builder
Money Flow Index (MFI)
MFI
MFI reads like RSI on a 0–100 scale but weights every price move by the volume behind it. That makes it a way to ask whether a move has participation — with one large caveat in FX, where the volume MT5 supplies is tick count rather than traded size.
- Default period
- 14
- Common levels
- 80 / 20
- Indicator range
- 0–100
Illustrative — synthetic data, not a live quote.
What MFI tells you
MFI splits each bar's typical price move into positive or negative money flow, multiplies it by that bar's volume, and expresses the balance on a 0–100 scale. The intuition is that a move backed by heavy participation matters more than the same move on a quiet bar.
- Above 80: buying flow has dominated the window
- Below 20: selling flow has
- Around 50: participation is balanced between the two sides
- 1 Multiply typical price by volume for each bar
- 2 Split into positive and negative flow
- 3 Express the ratio on a 0–100 scale
Show the formula & calculation detail
MFI weights each bar’s price movement by the volume that accompanied it:
- Compute the typical price of each bar:
TP = (High + Low + Close) / 3. - Multiply by the bar’s volume to get raw money flow:
MF = TP × Volume. - Classify it. If today’s typical price is higher than yesterday’s, the flow is positive; if lower, negative.
- Sum positive and negative flow separately over the period, form their ratio, and convert to a bounded index:
MFI = 100 − (100 / (1 + positive flow / negative flow)).
That last step is identical in form to RSI’s — which is why the two look so similar and why MFI is often described as a volume-weighted RSI. The differences are upstream: MFI is built on typical price rather than the close, and each bar’s contribution is scaled by how much trading happened during it.
The scaling is where the practical caveat lives. On an exchange-traded instrument, volume is contracts changing hands. In retail FX there is no central exchange, so MT5 supplies tick volume — how many times the price changed during the bar. That tracks activity reasonably well, but it is a different quantity, and it varies between brokers because it depends on their price feed.
In MQL5 the call is iMFI(symbol, period, ma_period, applied_volume) and it returns a handle to a single buffer. The last argument is a volume type (VOLUME_TICK or VOLUME_REAL), not an applied price — there is no price selector, because typical price is part of the definition.
One consequence of the shared shape is worth stating: sharing a scale with RSI does not mean sharing its thresholds. MFI is conventionally read at 80 and 20 rather than 70 and 30 precisely because the volume weighting changes how far the readings travel. The Ultimate Oscillator is the same lesson taken further — it runs 0 to 100, is drawn with RSI’s own 70 and 30 lines, and measured across identical series spends 1.13% of its time above 70 where RSI(14) spends 4.94%. Whenever two oscillators share a range, the thresholds still have to be measured rather than assumed.
Signals
The distinct, backtestable signals MFI gives — and the regime each one suits.
Centreline Cross
Builder readyDivergence
Advanced logicFailure to Confirm
Advanced logicMT5 implementation
What MetaTrader 5 itself computes and draws — the reference every rule on this page is written against.
Buffers
| Index | Buffer | Drawn in MT5 as | What it holds |
|---|---|---|---|
| 0 | MAIN_LINE | Line | The MFI value, 0–100. One buffer only — there is no signal line. |
Platform notes
Works best / use carefully
No indicator is a universal edge. Here is where MFI helps — and where it misleads.
Works best
- Instruments with meaningful volume
- Ranging markets
- As a participation cross-check
- With a volatility filter
Use carefully
- FX, where volume is tick count
- Extremes as standalone entries
- Strong one-way trends
- Comparing MFI across brokers
Build a strategy with MFI
Wire the signal into entry and exit rules, then export a compilable MT5 EA — no code.
- MFI (14)
- Fast StdDev (10) below slow StdDev (40)
- MFI below 20 in a quiet market → Open Buy · SL 40 / TP 80
- MFI above 80 in a quiet market → Open Sell
Combine MFI with other indicators
One indicator rarely stands alone. These pairings cover MFI's blind spots.
MFI + Standard Deviation
- MFI below 20
- Fast StdDev below slow
- Buy
MFI + RSI
- RSI < 30
- MFI > 40
- Question the signal
No single-flow template pairs MFI with RSI. Start from MFI + StdDev and add an RSI node with its own Compare.
Open Builder →MFI + Bollinger Bands
- Price at lower band
- MFI below 20
- Buy
No single-flow template pairs MFI with Bollinger Bands. Start from MFI + StdDev and add band nodes with a Price node into a Compare.
Open Builder →Parameters
Starting values to validate on your own pair and timeframe — not guaranteed settings.
| Parameter | Default | Suggested test range | What it does |
|---|---|---|---|
| Period | 14 | 7–30 | How many bars of money flow are summed. 14 mirrors RSI's default; longer periods make extremes rarer and more meaningful. |
| Overbought / Oversold | 80 / 20 | 70–90 / 10–30 | Your levels, not indicator inputs — iMFI has no threshold argument. Note these are wider than RSI's conventional 70/30. |
| Volume type | Tick volume | Tick / Real | Which volume feeds the weighting. VOLUME_REAL exists in MQL5 but most FX brokers do not supply it, so VOLUME_TICK — the number of price changes — is what you get in practice. |
| Shift | 0 | 0–3 | Which bar back the value is read from. Shift 1 compares against a fully closed bar. |
Starting presets
Reactive 7 Frequent extremes
Standard 14 The Builder template's baseline
Smooth 28 Rare, swing-level readings
Market examples
Where MFI works, where it fails, and how a filter changes the outcome.
Flow extreme in a quiet market
MFI drops under 20 while dispersion is contracting, and the fade back toward the middle plays out — the configuration the template trades.
Extreme during an expansion
MFI hits 20 as the market accelerates away. The reading is correct — flow really is one-sided — and fading it means standing in front of the move.
Dispersion filter suppresses the trade
The same extreme appears while the short-window standard deviation is above the long one, the entry is skipped, and the move continues.
FAQ
- Is MFI just RSI with volume?
- Close, but two things differ. MFI weights each bar's flow by volume, and it is defined on typical price rather than the close. Where the two disagree, the gap is telling you that the price move and the participation behind it are not aligned — which is the whole reason to run MFI instead of RSI.
- Does MFI work in forex?
- With a caveat. FX has no central exchange and most brokers report tick volume — the number of price changes in a bar — rather than traded size. That is a reasonable activity proxy and MFI remains usable, but the readings are broker-specific and 'money flow' is a metaphor rather than a measurement. On instruments with real exchange volume the indicator does what its name says.
- Why are the levels 80 and 20 instead of 70 and 30?
- Convention, and it reflects that MFI moves further than RSI on the same data because the volume weighting amplifies active bars. They are your levels either way — iMFI takes no threshold argument — so test them rather than inheriting them.
- Can I use real volume instead of tick volume?
- MQL5 supports it: iMFI takes an applied-volume argument that accepts VOLUME_REAL. Whether you get anything back depends on the broker and the instrument. On most retail FX feeds real volume is absent or zero, which is why the Builder uses VOLUME_TICK.
- Can I build an MFI EA without coding?
- Yes for level rules and centreline crosses — one MFI node plus a Compare, with the period as an EA input. The MFI + Standard Deviation template also shows the dispersion filter that makes a fade survivable. Divergence and 'a breakout with no participation' are the exceptions and need custom MQL5.
- How should I validate settings before going live?
- Backtest on quality tick data first, then run the EA on a demo account before risking money. With MFI specifically, remember the tick-volume caveat — a strategy tuned on one broker's tick counts may behave differently on another's. When optimising, hold back data the search never saw.
Related articles
Glossary