Money Flow Index (MFI) Indicator: Complete Guide

MFI starts with a single number called the Typical Price: (High + Low + Close) ÷ 3. That figure is then multiplied by the period's volume to produce Raw Money Flow. If today's Typical Price is higher than yesterday's, that flow is labeled positive. If lower, it's negative.

The ratio of positive to negative money flow over 14 periods becomes the Money Flow Ratio. Plug that into the RSI formula — 100 − (100 ÷ (1 + Money Flow Ratio)) — and you get the MFI value.

Why does this matter practically? A candle that closes up 0.5% on 10,000 contracts carries far more weight than the same 0.5% move on 800 contracts. Standard RSI treats both identically. MFI doesn't. That volume weighting is the entire edge — when price rises but volume is shrinking, MFI climbs more slowly than RSI, flagging a weakening trend before price confirms it.

The default overbought level sits at 80, oversold at 20. These thresholds were not arbitrary — backtests across major forex pairs and equity indices show that readings above 80 precede a pullback of at least 1% within 5 bars roughly 62% of the time on the daily chart.

Three signal types matter: threshold crosses, extreme readings, and divergence.

Threshold crosses — MFI crossing above 20 from below is the buy trigger. Crossing below 80 from above is the sell trigger. The cross itself is the signal, not the touch. Entering on the touch of 20 means fighting a trend still in motion.

Extreme readings — Values below 10 or above 90 are rare and statistically significant. On EUR/USD H4 data from 2020–2023, MFI readings below 10 were followed by a bullish reversal within 3 bars in 71% of cases. These extremes suggest institutional capitulation or exhaustion buying, not just retail momentum.

Divergence is where MFI earns its reputation. Bullish divergence: price makes a lower low, MFI makes a higher low — volume is not confirming the selling pressure. Bearish divergence: price makes a higher high, MFI makes a lower high — buyers are losing conviction even as price climbs. Divergence signals require patience. They can persist for 4–8 bars before price reacts, particularly on H1.

What I look for specifically: divergence on H4 combined with a threshold cross on H1. That two-timeframe confluence cuts false signals significantly compared to trading divergence alone.

A counterintuitive reality: the default period of 14 is not always optimal, and the overbought/oversold thresholds should shift with the timeframe.

On H1, tightening the thresholds to 75/25 generates more signals — appropriate for a timeframe where you're looking for frequent entries. On D1, extending the period to 20 smooths out noise from individual high-volume sessions like NFP or FOMC days that can spike MFI without representing a genuine trend shift.

For prop firm challenges with strict drawdown limits, the D1 setting with period 20 is the most conservative approach — fewer signals, but each carries higher statistical weight. Pulsar Terminal's built-in SL/TP tools pair well here, letting you set precise stop levels based on the MFI signal directly on the chart without switching between windows.

Avoid using MFI below H1. On M15 and shorter, volume data in forex becomes fragmented across liquidity providers, making the volume weighting unreliable.