Volume is conventionally constructed as positive volume + negative volume (both typically within 20% of one another) and then color coded by the price change for a given interval. Is that really useful for characterizing how a stock moves through a base and reacts at a pivot point? I would have thought a trader would be interested in net volume (positive volume - negative volume) and allow negative values to plot below the zero line instead of color coding them above the zero line. What am I missing? Why has the current construction of volume been adopted almost universally on trading platforms? Note: most platforms must get volume as a single number per interval and simply accept that volume aggregates positive and negative volume. Higher end trading platforms like TradeStation disaggregate volume into positive and negative volume.
I think changes in volume is what matters the most, and subsequent changes in selling/buying volume in response to some past sharp changes of buy/sell volume, i.e. autocorrelation. This should be a vector or research.
Exactly what I do in TradeStation - percentage volume colorization. Though it is most effective on higher timeframes so not very useful on 1 and 5 minute timeframes I trade on.
Percentage of what? What is the basis or lookback period? Does that basis or lookback affect your determination of the usefulness on fast charts? Why is knowing "why" (ie changes in volume in response to...) beneficial? FWIW, I completely agree recognizing changes in volume PACE is an important piece of volume analysis. Just asking questions.
Comparing up/down volume (delta) seems pointless to me. Majority of trades will have a counter party right? I mean do we seriously think that a momentary delta is actually going to provide meaningful edge? Buy/sell volume will generally equalize. I personally just use volume qty. High volume = right time to trade, i.e more participants.
Total volume is relevant to VWAP, and immediate/near-term VWAP. You can combine volumes. Interestingly, my algo (the color coded price) sometimes agrees with volume buy sell ratios and sometimes not. Still though, the study is useful for my analysis. White is VWAP.
Vol percent works, even on indexes. Here's is the bank index $BKX which plainly shows divergence at the swing high earlier - up vol (blue) way less than prior down vol (red) from the open / chopped off a little on the left:-
Another look at volume (Bitcoin Daily):- where the breakout higher failed June 5th because the $volume was not there $25.22b vs $46.93b but OTOH breakout lower succeeded on June 11th because it had more $volume $34.50 vs $11.64