MFI (Money Flow Index)
In Simple Terms: MFI is RSI with volume added as the tiebreaker. RSI tells you price momentum; MFI tells you whether that momentum is backed by actual money flow. A rally with rising RSI but flat MFI means price is climbing on thin volume — the move is fragile. A rally with rising RSI AND rising MFI means money is genuinely flowing in — the move has conviction. When MFI and RSI disagree (divergence), MFI usually wins because volume doesn't lie. The alpha: MFI divergence is a stronger reversal signal than RSI divergence — you need volume to confirm a trend change, and MFI captures that confirmation in a single oscillator.
The Money Flow Index, developed by Gene Quong and Avrum Soudack, is a volume-weighted momentum oscillator that operates on the same 0-100 scale as RSI but incorporates volume data into its calculation. Instead of comparing average gains to average losses (RSI's method), MFI compares "positive money flow" (price × volume on up days) to "negative money flow" (price × volume on down days). This volume integration makes MFI a more complete measure of market conviction — it answers not just "is price moving?" but "is money moving with price?"
In crypto markets where volume data is available and informative (unlike some traditional markets where volume reporting is fragmented), MFI provides insights that pure price-based oscillators miss. A price rise on declining volume (Rising RSI, falling MFI) is a different signal than a price rise on expanding volume (rising RSI, rising MFI). The former suggests distribution; the latter suggests genuine demand. This distinction is especially valuable in crypto, where low-volume rallies are frequently engineered by market makers and high-volume rallies represent genuine participation.
How It Works
MFI calculation steps:
1. Typical Price = (High + Low + Close) / 3
2. Raw Money Flow = Typical Price × Volume
3. Positive Money Flow = Sum of Raw Money Flow on days where Typical Price > Previous Typical Price
4. Negative Money Flow = Sum of Raw Money Flow on days where Typical Price < Previous Typical Price
5. Money Ratio = Positive Money Flow / Negative Money Flow
6. MFI = 100 - (100 / (1 + Money Ratio))
The typical 14-period MFI calculation accumulates positive and negative money flow over 14 periods, creating a volume-weighted ratio that expresses buying pressure vs selling pressure as a percentage. Readings above 80 suggest overbought (buying pressure may be exhausted); below 20 suggest oversold (selling pressure may be exhausted).
MFI vs RSI — when the difference matters. Consider this scenario: price rises 5% in a day on below-average volume. RSI spikes because the gains-to-losses ratio shifted. MFI barely moves because the actual dollar volume flowing into the asset was modest. Which indicator is right? Both are measuring what they measure correctly — but MFI's assessment of "this rally lacks money flow conviction" is the more actionable insight. The rally is technically overbought per RSI but not extended per MFI, suggesting the overbought reading may be a false signal caused by thin conditions rather than genuine excess.
The reverse scenario: price drops 3% on massive volume. RSI drops but MFI drops far more aggressively because the money flow out was substantial. This is a genuine oversold condition with volume confirmation — higher probability of a bounce than a comparable RSI oversold reading on low volume.
MFI divergence — the enhanced signal. Because MFI incorporates volume, its divergences carry more weight than RSI divergences:
- Bullish MFI divergence: Price makes a lower low, MFI makes a higher low. Selling pressure on the second decline was on LOWER volume — sellers are exhausting. This is a stronger buy signal than an equivalent RSI divergence because volume confirms the exhaustion.
- Bearish MFI divergence: Price makes a higher high, MFI makes a lower high. Buying pressure on the second rally was on LOWER volume — buyers are losing conviction. The distribution happening on the rally is visible only through the money flow lens.
The combined signal: when BOTH RSI and MFI diverge from price simultaneously, the reversal probability is significantly higher than either alone. When only one diverges, MFI divergence carries more weight — volume is the harder signal to fake.
MFI overbought/oversold in trend context. Like RSI, MFI can remain overbought (>80) for extended periods during powerful trends. In a strong crypto uptrend backed by genuine institutional buying, MFI may sustain readings of 70-90 for weeks. This is not a sell signal — it's confirmation that money flow is overwhelmingly positive. Shorting MFI overbought readings without trend context is as dangerous as shorting RSI overbought readings without context.
The nuanced approach: in an uptrend, look for MFI to pull back from overbought territory to the 40-50 zone (not below 20 — that's too weak for an uptrend) and then turn up. This "reset" in money flow indicates the buying has paused but not reversed — a continuation entry. In a downtrend, look for MFI to "reset" from oversold to 50-60 zone before turning down.
MFI failure swings. Identical to RSI failure swings but with volume confirmation: when MFI enters overbought territory (>80), pulls back, makes a lower high still in the overbought zone, then breaks below the intervening trough — that's a high-probability top signal. The volume weighting means the failure swing is confirming that even the retest lacked conviction (lower MFI high = less money flow on the retest).
The MFI 50-level as trend filter. Like RSI's 50-level, MFI's 50-level serves as a trend filter. When MFI consistently holds above 50 on pullbacks, the trend is bullish. When MFI consistently fails to reclaim 50 on rallies, the trend is bearish. The difference: MFI's 50-level is backed by money flow, not just price ratio. An MFI that can't sustain above 50 means buying pressure cannot overcome selling pressure in absolute dollar terms — a more fundamental weakness than RSI below 50.
Why It Matters for Traders
Volume-confirmed signals have higher reliability. Academic and empirical research consistently shows that volume-confirmed technical signals outperform price-only signals. MFI provides this confirmation in a single 0-100 oscillator — you get volume-weighted momentum without needing to analyze volume and momentum separately. For traders who struggle to integrate volume analysis with oscillator analysis, MFI does the integration for you.
MFI divergence catches distribution that price-based indicators miss. A common topping pattern in crypto: price grinds higher on declining volume (distribution). RSI makes a marginal new high (price pushed up, even if weakly). MFI makes a significantly lower high (money flow didn't confirm). The MFI divergence warns of the top before price breaks down, giving you time to reduce longs or prepare shorts. This pattern preceded many major crypto tops, including the November 2021 Bitcoin peak where daily MFI diverged sharply from price weeks before the decline.
Combine MFI with Kingfisher's TOF for layered volume analysis. MFI captures exchange-traded money flow direction. Kingfisher's Time of Flight (ToF) captures the nature of order flow execution (buying absorption vs selling absorption). When MFI is rising and ToF confirms persistent buying absorption, the uptrend has both money flow breadth (MFI) and execution quality (ToF). When MFI is rising but ToF shows selling absorption, the money flow is going in on weaker execution — the buying is hitting passive sell orders rather than lifting offers aggressively. This nuance in execution quality separates genuine demand from accumulation-at-a-discount.
Common Mistakes
- Trading MFI extremes without trend context. MFI > 80 is not a sell signal in a bull trend. MFI < 20 is not a buy signal in a bear trend. The trend determines whether extremes represent exhaustion (ranging market) or conviction (trending market). The same rules apply as RSI: filter extremes with the higher-timeframe trend, and wait for a reversal structure before acting.
- Ignoring the volume data quality in crypto. Not all crypto volume is created equal. Some exchanges wash-trade. Some volume is incentive-driven (zero-fee promotions, market maker programs). MFI on an exchange with inflated volume will produce inflated readings. Use MFI on data from reputable exchanges (Binance, Coinbase, Bybit, OKX) or aggregated volume from reliable sources. Garbage volume in = garbage MFI out.
- Expecting MFI divergence to time reversals precisely. MFI divergence can persist for weeks before price reverses. The divergence is an early warning, not a timing signal. Use it to reduce position sizes, tighten stops, and prepare for reversal — but don't enter reversal trades until price structure confirms (lower high in uptrend, higher low in downtrend). Divergence without structure confirmation is a thesis without evidence.
FAQ
Q: How is MFI different from Chaikin Money Flow (CMF)? A: MFI is a bounded oscillator (0-100) that compares positive and negative money flow as a ratio, producing overbought/oversold readings like RSI. CMF is an unbounded oscillator that accumulates money flow over time and applies a moving average, producing cross-above/below-zero signals. MFI is better for overbought/oversold analysis and divergence. CMF is better for trend confirmation (above/below zero) and sustained money flow direction. They're complementary — MFI for extremes, CMF for trend context.
Q: What MFI settings should I use for crypto? A: The standard 14-period works on daily charts. For crypto's faster cycles, some traders use 10-period for more responsive signals. The tradeoff is standard: shorter periods = more signals but more false ones. Stick with 14 unless you have specific evidence that a different period works better for your asset and timeframe. Parameter optimization without statistical validation is curve-fitting.
Q: Can MFI be used on lower timeframes? A: Yes, but with the same volume-quality caveat. On 15-minute charts, single large orders can spike volume and distort MFI, creating false signals. MFI is most reliable on 1-hour and above where the volume sample in each period is large enough to be representative. Below 1-hour, use MFI as a secondary reference, not a primary signal.
Deep Dive
Want to explore further? Check out:
- DeFi Yield Farming Guide: Earn Passive Income With Crypto 2026
- Understanding Crypto Market Structure: Order Flow, Liquidity and Price Discovery
- Leverage Trading Crypto: Complete Guide to Margin Trading 2026