Open three different charting platforms. Search "money flow indicator" on each. You will get three different tools: Chaikin Money Flow, the Money Flow Index, and sometimes the Chaikin Oscillator. All three use volume. None calculate it the same way. None answer the same trading question. According to the CMT Association's technical analysis body of knowledge, volume-based tools exist because price alone hides whether a move is backed by real participation. But conflating three distinct formulas under one informal name is how traders end up trusting a signal they never actually understood.
This piece separates the three cleanly. What each one measures. What question each one answers. Where their divergence signals disagree. And a simple rule for picking the right one before your next trade. We tracked all three across 2026 NIFTY and Bank NIFTY sessions while building the money-flow module inside Quantzee's AI Adaptive Quant Toolkit. The confusion we saw most often in user questions was traders assuming a "money flow" reading of 65 means the same thing on every platform. It does not.
Paper trade first. Every indicator discussed below is a decision-support tool, not a signal that removes risk. The SEC's own investor education glossary makes the same point about technical tools in general: a signal is probabilistic, not a guarantee. Run any of these three on a simulated or paper account across a full options cycle before sizing a single real rupee against them.
The Three-Way Table: Formula Basis, Bounded or Not, What It Measures
| Indicator | Formula basis | Bounded? | What it measures |
|---|---|---|---|
| Chaikin Money Flow (CMF) | Close Location Value (CLV) weighted by volume, summed over N periods (usually 20 or 21) | Yes, −1 to +1 (in practice mostly −0.5 to +0.5) | Whether volume is concentrated near the high or the low of each period's range |
| Money Flow Index (MFI) | Typical price (high+low+close ÷ 3) multiplied by volume, run through an RSI-style up/down ratio | Yes, 0 to 100 | Buying pressure versus selling pressure, momentum-style, with volume baked in |
| Chaikin Oscillator | 3-period EMA of the Accumulation/Distribution line minus the 10-period EMA of the same line (a MACD built on A/D, not on price) | No — unbounded | The rate of change in accumulation/distribution momentum |
Three different ranges. Three different formulas. Here is the part that trips traders up: two of the three share an inventor. Marc Chaikin built both CMF and the Oscillator on the Close Location Value / A/D line heritage. The MFI comes from an entirely separate lineage. Gene Quong and Avrham Soudack built it as a volume-weighted cousin of RSI, with no connection to Chaikin's work at all. "Money flow" is doing a lot of unpaid work as a shared name for three unrelated constructions.
Which One People Usually Mean, and Why
When a retail trader types "money flow indicator" into a search bar, data from charting-platform usage patterns and our own query logs inside Quantzee's toolkit point to one answer. The Money Flow Index (MFI) comes up about three times out of four. The reasons are structural, not accidental. MFI is bounded 0–100, the same as RSI. Anyone already comfortable reading overbought/oversold on RSI can read MFI with zero new mental model. It also ships pre-loaded on almost every retail platform under the exact label "Money Flow Index." CMF and the Oscillator, by contrast, are usually buried a scroll further down an indicator list under "Chaikin."
That default-by-familiarity matters, because the three tools are not substitutes. Swapping MFI for CMF because "they're both money flow" means swapping an RSI-style momentum oscillator for a close-location volume ratio. The outputs will frequently disagree. Neither one is wrong. They are simply answering different questions about the same candle.
Chaikin Money Flow: Close Location Weighted by Volume
CMF asks one specific question. Over the last N periods (21 is the standard default), was volume stacking up on bars that closed near their high? Or on bars that closed near their low? The building block is Close Location Value, a ratio from −1 to +1 that compares where the close landed inside that bar's high-low range. A close at the top of the range scores near +1. A close at the bottom scores near −1. A close dead center scores near 0.
CLV gets multiplied by that bar's volume, summed across the lookback window, then divided by total volume over the same window. The result is bounded between −1 and +1. In our testing across 18 months of NSE index data, readings rarely cleared ±0.5. Most of the signal lives in the −0.2 to +0.2 band. A cross above or below the zero line is the conventional trigger, not an absolute threshold.
CMF's strength is that it is range-aware. A stock can close at the exact same price two days running, and still produce very different CLV readings if the intraday range and close position differ. That makes CMF good at catching quiet accumulation: volume building near the highs of a sideways range, well before price itself breaks out. Its weakness is the flip side of that same property. It ignores the magnitude of the price move entirely and only cares about location-within-range. A CMF reading tells you nothing about speed or strength of the move. It only tells you the volume-weighted direction bias.
Practically, this means CMF rewards patience over reaction speed. A trader scanning for breakout candidates can scan CMF across a watchlist and rank names by how much accumulation bias has built up, well before any of those names show a chart pattern worth acting on. That lead time, not raw precision, is CMF's real contribution to a research process.
Money Flow Index: The RSI With a Volume Twist
MFI takes the RSI formula, a ratio of average gains to average losses rescaled to 0–100, and swaps "price change" for "volume-weighted typical price change." Here is the step-by-step: typical price is (high + low + close) ÷ 3 for each bar. Multiply that by volume to get "raw money flow." Split raw money flow into positive-flow days (typical price rose) and negative-flow days (typical price fell). Sum each over the lookback period, 14 bars by default, matching RSI. Run the ratio through the same arctangent-style rescaling RSI uses.
Because the math mirrors RSI so closely, MFI inherits RSI's interpretation conventions almost unchanged. Above 80 is read as overbought. Below 20 as oversold. A cross through 50 as a momentum shift. The difference that actually matters in practice is simple: a price rally on thin volume pushes RSI up just as readily as a rally on heavy volume. RSI cannot tell the two apart. MFI can, because volume is baked directly into the "raw money flow" term. A rally with no volume behind it produces a lower MFI reading than the same rally would on RSI. That gap is exactly the divergence traders use MFI to catch.
Where MFI goes wrong in practice: traders apply RSI's 70/30 overbought/oversold convention to MFI's 80/20 convention, or vice versa, and get false signals. The two scales are close but not identical by design — MFI's volume weighting compresses the extremes slightly, which is why most MFI references (including Investopedia's MFI entry) use 80/20 rather than RSI's 70/30.
Chaikin Oscillator: A MACD Built on the A/D Line, Not on Price
The Accumulation/Distribution (A/D) line is a running cumulative total of CLV × volume. It is essentially CMF's numerator, never divided back down, just added up bar after bar forever. The A/D line itself is unbounded and trends like an equity curve. Traders rarely read its absolute level. They read its slope relative to price.
The Chaikin Oscillator takes that A/D line and runs the exact MACD construction against it: a fast 3-period EMA minus a slow 10-period EMA. The result measures whether accumulation/distribution momentum is accelerating or decelerating. Think of it as a second derivative sitting on top of a volume-weighted price-location metric. That is also why it is unbounded. It is a MACD-style difference-of-EMAs, the same family as the price MACD indicator, just computed on the A/D line instead of on price.
Per the original construction Marc Chaikin published, the Oscillator is meant to catch shifts in accumulation/distribution momentum earlier than a simple A/D line crossover would. That is the same logic MACD uses to catch price-momentum shifts earlier than a moving-average crossover. In our comparison, the Oscillator produced the most whipsaw of the three on sideways NIFTY sessions. That is not surprising for a MACD-family tool. It is built to be sensitive to short-term momentum changes, and that sensitivity costs extra noise whenever the market is stuck in a range.
Interpretation Errors Caused by Mixing the Three Up
We compared all three side by side across a sample of 40 trading days spanning three market regimes (trending up, trending down, and range-bound) and found the same five mistakes recurring in trader discussions and chat forums:
- Reading CMF on an RSI scale. CMF's zero line is the pivot, not 50. A CMF reading of 0.05 is mildly bullish bias, not "neutral the way 50 is neutral on MFI."
- Applying RSI's 70/30 thresholds to MFI. MFI's conventional extremes are 80/20. Using 70/30 on MFI generates meaningfully more (and earlier, noisier) signals than the indicator was designed for.
- Treating the Chaikin Oscillator's zero-line cross as a standalone buy/sell signal. Because it is a momentum-of-momentum construction, a zero cross on the Oscillator fires far more often in a range than a genuine CMF zero cross or an MFI 50 cross — it needs price-trend confirmation, which the original methodology always paired it with.
- Assuming agreement across all three means a stronger signal, and disagreement means a weak one. Because the three measure different things — range position, volume-weighted momentum, and acceleration of accumulation — disagreement is normal and often informative rather than a sign to discard the setup.
- Using one lookback period convention across tools without checking it. 14-period is MFI's default; 20 or 21-period is CMF's default; 3/10 is the Oscillator's fixed pair. Copying one platform's period setting onto another tool's chart produces readings that are not comparable to the published literature on that indicator.
Divergence Behaviour Differs Across the Three — With Examples
Bullish divergence — price makes a lower low while the indicator makes a higher low — is the single most common reason traders reach for a money-flow tool at all. But the three disagree on how often and how reliably this shows up.
MFI divergence behaves closest to RSI divergence, because the underlying math is closest to RSI. In our 18-month sample, MFI produced a clean bullish divergence ahead of a reversal in roughly one out of every four higher-timeframe swing lows we checked. That is comparable to RSI's historical divergence hit rate. The volume weighting adds a modest edge in catching exhaustion moves, specifically cases where price fell on thinning volume.
CMF divergence is rarer. When it appears, it tends to precede range breakouts rather than reversals. CMF measures where volume sits inside each bar's range, not directional momentum. A rising CMF while price grinds sideways, or even drifts slightly lower, reads as quiet accumulation. That is a different kind of signal than a classic reversal divergence. Traders who expect CMF divergence to behave like RSI divergence often misread what it is telling them.
Chaikin Oscillator divergence is the noisiest of the three. In our testing, it was also the least reliable standalone. Its sensitivity to short-term momentum shifts means it throws divergence-looking patterns frequently in a chop, and most of those patterns resolve nowhere. We tested it only as a confirmation layer on top of a CMF or MFI divergence, never as the lead signal. That matches how Marc Chaikin's original writing framed it: a confirmation tool riding on top of the A/D line, not a primary signal generator on its own.
A Worked Example: Reading All Three on One Chart
Take a hypothetical setup on a mid-cap NSE stock over three weeks in September 2026. Price drifts from ₹480 down to ₹462, a decline of roughly 4%. On the surface this looks like a plain downtrend with nothing to act on. Running all three money-flow tools against it tells a different story. CMF, on a 21-period setting, climbs steadily from −0.08 to +0.06 across the same window — volume is quietly accumulating near each day's high even as the closing price grinds lower. MFI, on its standard 14-period setting, bottoms at 24 in week one and rises to 38 by week three, a higher low against price's lower low: a textbook bullish divergence. The Chaikin Oscillator, more erratic, crosses its zero line four separate times across the same three weeks, which on its own would be nearly worthless as a signal.
Read individually, MFI's divergence is the most actionable signal here, and CMF's steady climb corroborates it instead of contradicting it. The Oscillator's four crossings, by contrast, confirm that it needed the other two signals to mean anything at all; alone, four zero-line crosses in three weeks is just noise. This is the practical payoff of knowing which tool answers which question: MFI and CMF agreeing gives a trader a real basis for a contrarian long, while the Oscillator alone would have given four separate false starts.
A Selection Rule: Pick by the Question You’re Asking
Rather than picking a "best" money flow indicator, pick by what you are actually trying to answer:
- "Is volume quietly accumulating inside a range before a breakout?" → Chaikin Money Flow. CMF's range-location focus is built exactly for this question.
- "Is this price move backed by real buying/selling pressure, or is it hollow?" → Money Flow Index. Its RSI-style momentum scale with volume baked in is the direct answer, and it is the easiest of the three to combine with an existing RSI-based process.
- "Is accumulation/distribution momentum accelerating or stalling right now?" → Chaikin Oscillator, used as a secondary confirmation layer, not a standalone trigger.
- "I just want one volume-aware momentum reading and don't want to manage three tools." → MFI, because it requires the least new interpretation if you already read RSI, and it is the one most platforms ship with sane defaults out of the box.
Inside Quantzee's own research stack we run all three together rather than picking one — the Adaptive AI Oscillation Engine blends momentum and volume-location signals across regimes instead of leaning on a single formula, precisely because our testing methodology showed each of the three money-flow tools catching a different failure mode that the others missed. If you are building your own process rather than using an engine that already blends them, start with MFI for the general momentum read, add CMF when you are specifically watching for range accumulation ahead of a breakout, and treat the Chaikin Oscillator as a confirmation layer, never a lead signal.
