Money Flow Index strategy — on every asset
The Money Flow Index is a volume-weighted version of RSI: instead of counting price changes alone, it weights each day by its trading volume (using the typical price - the high, low and close averaged - times volume), then measures how stretched recent buying pressure is versus selling pressure on a 0 to 100 scale. Below 20 is called 'oversold', above 80 'overbought' - a gauge of stretch, not a prediction of a turn. Buy when it is oversold, betting on a bounce, and exit when it becomes overbought. It is the same mean-reversion bet as the RSI Reversion rule already in this library, with volume as the only extra information - so on constant, thin or volume-less data it behaves much like RSI (RSI-like, not identical: it scores each bar by its money-flow level, not the size of the move). Like any mean-reversion oscillator it fights the trend and whipsaws in choppy ranges, so a strong move can run it over (a falling knife), and volume itself can be noisy or patchy.
An illustration of where the rule triggers — not real prices, and not a signal to act. The dashed forks mean the outcome is undetermined; whether the rule made money is what a backtest measures.
When price is stretched from its mean — By construction it fades a stretched move — the same mean-reversion bet as RSI, with volume as the only extra input — entering oversold and exiting overbought.
In a sustained one-way move — A sustained one-directional move runs it over — like any reversion oscillator it fights the trend and can buy into a falling market, and volume itself can be noisy or patchy.
Whether that behaviour produced excess return on your asset over your window is exactly what your backtest measures — we publish no profit claims.
How it works — step by step
- • Entry — Go long when it reads oversold (below ~20), betting on a bounce.
- • Exit — Go flat when it reads overbought (above ~80).
- • Fills, scoring & data — Every backtest fills at the next real open and scores on the real price against a real buy-and-hold, never a smoothed or derived series; on constant, thin or volume-less data it behaves much like RSI, and volume itself can be noisy or patchy.
Most of these did not beat buy & hold — but a real few did, and finding out which is the whole point. This one strategy was tested on 91 different assets. With that many attempts, some will beat buy & hold by chance alone — which is why these are listed neutrally, not ranked by which ones came out ahead. To see whether any single result survives scrutiny, run the out-of-sample, walk-forward and multiple-testing checks on its backtest.
Of the 83 with enough data to judge, 36 beat buy & hold · 8 didn't have enough data to judge yet. Results are simulated backtests over the last ~5 years (less for younger assets), shown whole — losses included — not live or future performance. These are famous names that still trade today. Names that delisted, went bankrupt or were wound down aren't in the set (survivorship bias) — and that history isn't available from the free public data this tool runs on — so this rate isn't a representative base rate for every stock or ETF that has ever traded.
Listed in a neutral catalog order — not ranked by performance. Combinations we can't judge yet — too little data — are grouped separately below.
Here, “beat” means only a higher total return than buy & hold over this one past window — not risk-adjusted, not luck-checked. The out-of-sample, walk-forward and luck checks live on each individual backtest.
Past performance is not a reliable indicator of future results. This is an educational, analytical tool — not investment advice, and not a recommendation to buy or sell anything. You make all decisions and execute on your own broker.
HYPOTHETICAL OR SIMULATED PERFORMANCE RESULTS HAVE CERTAIN LIMITATIONS. UNLIKE AN ACTUAL PERFORMANCE RECORD, SIMULATED RESULTS DO NOT REPRESENT ACTUAL TRADING. ALSO, SINCE THE TRADES HAVE NOT BEEN EXECUTED, THE RESULTS MAY HAVE UNDER- OR OVER-COMPENSATED FOR THE IMPACT, IF ANY, OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY. SIMULATED TRADING PROGRAMS IN GENERAL ARE ALSO SUBJECT TO THE FACT THAT THEY ARE DESIGNED WITH THE BENEFIT OF HINDSIGHT. NO REPRESENTATION IS BEING MADE THAT ANY ACCOUNT WILL OR IS LIKELY TO ACHIEVE PROFITS OR LOSSES SIMILAR TO THOSE SHOWN.
On the 83 assets with enough data to judge, the same rule swung from +254 pts vs buy & hold on Ethereum to -1216 pts on Micron Technology. The typical (median) result was -22.2 pts — and even across these runs, the deepest drawdown you'd have sat through was -92% on Polkadot.
That range is the whole point: one rule is not universally good or bad — it depends entirely on the asset, and with 91 tried, some beat by chance. Shown to reveal the variation, not to pick one to trade.
Money Flow Index on stocks
Money Flow Index on etfs
Money Flow Index on commodities
Money Flow Index on crypto
Backtesting is one stage of preparing to trade, not the finish line. The full discipline runs five: Understand, Test, Stress-test, Forward-test, then Decide — eyes open. The rare idea that survives all five is the one worth taking into the market with your eyes open — and every stage before it is a place to stop, for free, as most ideas wash out in the first three. Finding that rare idea before you risk a cent is the whole point.