Volume momentum
Flags a run of minutes where volume keeps running well above its recent average, with no view on price direction.
Each one-minute bar's volume is compared with the average of the last 20 bars, which includes the current bar. A bar counts as a surge when its volume is more than 1.3 times that average. The indicator keeps the surge flags of the last 5 bars and fires when more than 60% of them are surges, which at defaults means 4 of the last 5. It never reads price, so it shows that volume stayed heavy, not which way price moved.
How to read it
- Markers sit on the candles where it fires. A marker means heavy volume has held for most of the recent minutes; it makes no call on direction.
- At defaults it fires on about 5% of bars, around 13 separate runs a day in the month sample.
- Consecutive markers are one burst, not separate signals. The marker does not say whether the surge was buying or selling; read it beside the candle's direction.
What it outputs
- Volume surge run
- Fires on a bar when more than 60% of the last few bars were volume surges. Its value is that share: 0.8 or 1.0 at defaults.
When it fires
When more than 60% of the last momentum_window bars (4 of the last 5 by default) each traded above 1.3 times their recent average volume. It means heavy volume has held across several minutes, not one spike. Price direction plays no part in the test.
Settings
- volume_avg_window
- Bars in the volume baseline each bar is measured against. Longer makes the baseline steadier, so surges must stand out from a longer history. Shorter lets a burst lift its own baseline, so later bars pass less often. At 1 it never fires.
- momentum_window
- Recent bars that vote on the surge share. Longer needs a larger, denser run of surge bars before it fires. Shorter accepts a shorter burst: 2 needs both of the last two bars, 3 needs two of three. At 1, any single surge bar fires.
Pitfalls
- It ignores price. The description says volume momentum often precedes explosive moves, but the code only counts volume, so that claim is untested here.
- Surges look identical in a quiet drift and at a breakout, since no price or range filter is applied. Pair it with one.
- The 1.3 surge threshold and the 60% vote are hard-coded, so only the two windows can be tuned.