Volume Oscillator: Formula, Settings, the Weekend Trap on Bitcoin, and What It Added to Breakouts
The Volume Oscillator shows how far a short average of volume runs above or below a longer one; here is the percentage formula, the 5 and 10 or 12 and 26 settings, why it dives every weekend on Bitcoin, and what a positive reading added to S&P 500 ETF breakouts since 1993.
On this page
The Volume Oscillator compares two moving averages of volume, a fast one and a slow one, and plots their difference, most often as a percentage of the slow average. Positive readings mean recent volume runs above its longer-term pace; negative readings, below it. It says nothing about price direction: a selloff and a rally on heavy volume both push it up. LuxAlgo notes that no single author owns the construction, which charting platforms adopted as standard equipment; the Interactive Brokers glossary describes it in the same terms, as a relationship between two moving averages of volume.
Volume Oscillator (5, 10) on daily Bitcoin
- Price, daily candles (hollow up, filled down)
- Volume Oscillator, % gap of the EMA(5) of volume over the EMA(10) (solid line)
How the Volume Oscillator is calculated
Take an exponential moving average of volume over a short period and another over a longer period, subtract the long one from the short one, and divide by the long one. The percentage form makes readings comparable across assets whose volume is counted in very different units. StockCharts’ Percentage Volume Oscillator (PVO) is the same construction with 12 and 26 days and a 9-day EMA of the result as a signal line, the settings of MACD. LuxAlgo writes that the difference is most often, not always, expressed as a percentage of the slow average.
How to read it, and which settings
StockCharts uses the PVO to validate support and resistance breaks: a break on expanding volume, with the PVO moving into positive territory, has more credibility than one on low volume. It also warns that the PVO can be choppy because volume does not trend, and that divergences are not well suited to it. LuxAlgo gives a fast average near 5 against a slow one of 10 to 20 as common. Shorter settings react faster and cross zero more often. The oscillator also spends more time below zero than above: on the S&P 500 ETF since 1993, the 5 and 10 version was positive on 44.3% of the days, a pattern that fits volume made of many ordinary days and a few heavy ones.
The 5 and 10 setting crosses zero about twice as often as 12 and 26 on both markets:
Zero crossings a year of the Volume Oscillator by setting
- S&P 500 ETF
- Bitcoin
| Setting (EMA) | ETF: days positive | ETF: zero crosses a year | Bitcoin: zero crosses a year |
|---|---|---|---|
| 5 and 10 | 44.3% | 54.9 | 78.4 |
| 12 and 26 | 44.7% | 27.1 | 36.0 |
| 14 and 28 | 44.8% | 23.8 | 32.1 |
The counts run from January 29, 1993 on the ETF and August 17, 2017 on Bitcoin, both to October 28, 2026, each from the first day both averages exist.
The weekend trap on Bitcoin
The chart shows the 5 and 10 oscillator on daily Bitcoin from July 1 to October 28, 2026. Bitcoin trades every day, but its weekends are quiet: over the year to October 28, 2026, the median weekend day traded 0.48 times the volume of the median weekday. A 5-day average reacts to that within the week. In the window, the oscillator was negative on 31 of the 34 Saturdays and Sundays, against 42 of the 86 weekdays, and its eight lowest readings all fell on weekends, the lowest being −18.0 on Sunday, August 30.
Days with a negative 5 and 10 oscillator on Bitcoin, July 1 to October 28, 2026
Its highest reading, +32.7, came on August 21, the third day of a rally that took Bitcoin from 64,725 on August 18 to 78,338, a 21.0% gain on rising daily volume. The oscillator had marked the expansion of volume, but not what came next: on September 2, twelve days later, Bitcoin closed at 77,340, 1.3% below the August 21 close. It went on to its highest close of the window, 86,620, on September 21.
Does it work?
Our bench tests chart patterns detected by rule against random entries, not indicators (see our methodology). We measured whether a positive oscillator improved simple breakouts, the first close above the highest close of the previous 20 days, on the S&P 500 ETF from January 29, 1993 to October 28, 2026, on dividend-adjusted closes, without costs. A breakout is counted only if the previous day was not one, and its 10-day change only if ten later closes exist.
| ETF breakouts | Count | Mean change over the next 10 days | Share of rises |
|---|---|---|---|
| 14 and 28 above zero | 232 | +0.54% | 66.8% |
| 14 and 28 at or below zero | 606 | +0.19% | 60.1% |
| 5 and 10 above zero | 272 | +0.37% | 62.5% |
| 5 and 10 at or below zero | 568 | +0.25% | 61.4% |
Mean change over the next 10 days after an ETF breakout, 1993 to 2026
Most ETF breakouts came with the oscillator at or below zero. Those with expanding volume did a little better over 10 days, more clearly with the slower setting. On Bitcoin, from 2017 to 2026, the 5 and 10 version gave +3.48% over 10 days for the 128 breakouts above zero against +2.14% for the 104 at or below, but the order reversed over 20 days, +4.99% against +6.38%. These are descriptive averages, not a trading test.
The oscillator works on raw trading volume, with the averages explained on the moving average page. The Klinger Volume Oscillator is a different tool despite the name: it signs volume by price direction first. See all volume indicators.
Frequently asked questions
What is the Volume Oscillator?
The gap between a fast and a slow moving average of volume, usually as a percentage of the slow one. Above zero, recent volume runs above its longer pace.
What are the best settings for the Volume Oscillator?
LuxAlgo gives 5 against 10 to 20 as common; the PVO uses 12 and 26. Shorter settings cross zero about twice as often on the S&P 500 ETF.
What is the difference between the Volume Oscillator and the PVO?
The PVO is the percentage Volume Oscillator with 12 and 26-day EMAs and a 9-day signal line, as StockCharts defines it.
Is the Volume Oscillator the same as the Klinger oscillator?
No. The Klinger oscillator signs volume by the direction of price before averaging it; the Volume Oscillator uses unsigned volume.
