This is an idea taken from a John Bollinger (of Bollinger Bands fame) talk. Instead of showing volume with a moving average overlay, we show volume relative to its moving average:
Now if we get a value of 1, that means the current volume is the same as its historical average. Under 0, less than average, and above zero, greater than its average.
If we get a value like 2, then current volume is twice its average. I hope the implication of this being displayed visually is becoming clearer.
I am still exploring volume as trading data point, but we see some ideas from this visual representation:
avgVolume = sma(volume, 10) // several configurable MAs allowed
relativeVolume = volume / avgVolume
Now if we get a value of 1, that means the current volume is the same as its historical average. Under 0, less than average, and above zero, greater than its average.
If we get a value like 2, then current volume is twice its average. I hope the implication of this being displayed visually is becoming clearer.
- We plot this relative volume as columns.
- We then plot horizontal levels, like 1, 2, 3 to see the magnitude of the current volume relative to its average.
- Consecutive rising or falling relative volume is shown in the same colour.
I am still exploring volume as trading data point, but we see some ideas from this visual representation:
- How do volume patterns change across timeframes? Do we get better signals or higher or lower time frames (e.g. big relative volume spikes)
- Can consecutive rises or falls indicate a big potential move, even though price is just fluctuating.
- What about a switch from rise to fall.
- If we get pinbars/spikes with a big relative volume spike, can we then infer more clearly whether buyers or sellers are in control.