Volume Spike vs. 20-Day Average: Predictive Power for Next-Day Returns

What You're Actually Asking

You've seen it: price surges on three-times-normal volume and traders call it confirmation. The specific version researchers test is straightforward—compare today's volume to the trailing 20-day average, flag a spike above some threshold, and measure whether the next-day close is higher or lower. The appeal is intuitive: unusual participation should signal conviction.

The harder question is whether that intuition survives rigorous out-of-sample testing with realistic transaction costs. Our database covers 660,005 backtests across 903 assets and 382 indicators, tested across 1-Hour, 4-Hour, Daily, and Weekly timeframes. Volume-based signals appear throughout that dataset, and the results are more sobering than most traders expect.

What the Data Actually Shows

The closest analog to a volume-spike signal in our test set is Money Flow Index, which combines volume and price direction in a way that tracks the same intuition. Its median win rate across assets is 72.2%. That sounds strong. But only 9% of assets where MFI was the best available indicator actually beat buy-and-hold after costs. Win rate and profitability are not the same thing.

Across the full dataset, only 26% of all indicator-and-timeframe combinations beat buy-and-hold. The median best Sharpe ratio across assets is 0.62. Volume-based signals are not uniquely penalized—most indicators underperform—but they are not spared either.

The Win-Rate Trap

A volume spike followed by a next-day gain is memorable. A volume spike that goes nowhere is forgettable. This asymmetric attention means traders tend to overweight the hits. When MFI records a 72.2% median win rate but converts that into a beat-buy-and-hold result on only 9% of assets, the gap is explained by the losses on the remaining trades being large enough to offset the gains.

Several other indicators in our data show the same pattern—high win rates masking poor overall performance. RSI Mean-Reversion posts a 71.7% median win rate but beats buy-and-hold on only 10% of assets. CCI hits 71.0% and beats buy-and-hold on 9%. The pattern is consistent: signals that fire frequently in trending markets inflate win rates without delivering net edge.

Where Volume Does Show Edge

Volume information is not useless across the board. Delta Volume Rising—a proxy for cumulative volume delta (CVD), which tracks buying versus selling pressure rather than raw volume magnitude—appears among the top five indicators for crypto assets in our data, with wins on 4 assets. That is a narrower, more direction-aware reading of volume than a simple spike-vs-average comparison.

For stocks, the top-performing indicators in our data are Fibonacci Pivots, Projection Bands, Intraday Momentum Index, and Camarilla Pivots—none of which are pure volume signals. For forex, the Fisher Transform dominates with wins on 17 assets. Volume contributes to some of these indirectly, but a standalone 20-day comparison does not rank as a decisive edge in the data we have.

Hypothetical Backtests, Not Advice

Everything on this page describes simulated historical performance. All 660,005 backtests are run out-of-sample with realistic transaction costs built in, but simulation cannot guarantee future results. Market conditions change, execution differs from backtests, and no indicator—however well it performed historically—constitutes a recommendation to trade. You are responsible for your own decisions. This is not financial advice.