Anyone who's looked at SPY daily price and SPY daily volume on the same chart will notice that there's a nice negative correlation between change in daily volume and daily return. Unfortunately, by the time I notice that daily volume has increased, I've already suffered today's loss! Causality bites! Of course, when a major move is happening, the volume trends typically persist for many days, so I can still catch much of the trend. It would be better if I could predict daily volume ahead of time. Obviously, it's impossible to predict an exogenous news event. That said, are people trying to make such predictions?
Great. Obviously, that indicator is a well-trodden path. I am thinking we are getting atypical volume behaviour as everybody is feeling out the start of the year in terms of direction and sizing. This is from the retail cheap seats, but the volume we bring is not insignificant. There is quite a bit of psychology along with change in portfolio compositions at the turn of the year.
The volume of SPY has been steadily ticking upward in each of the last several trading days, which doesn't portend well for the Santa rally, but maybe there are non-typical "new year" factors at work, as you suggest.
The ES futures in the Trading Technologies platform has the most accurate real time volume at price histogram that I have personally ever seen.
Isn’t high volume caused by high volatility? If so, wouldn’t you rather track volatility? Also, why are you correlating volume with trend? I am not sure I follow your premise.
Here's a scatterplot of SPY daily change in volume (x axis) versus daily % change in stock price. Pretty amazing correlation (maybe R^2 of 0.4?). It's the change in volume that's the key. Absolute volume actually isn't predictive at all. The problem is that by the time I measure daily change in volume, the daily change in price is already baked in. So to trade this, I'm relying on the "delta volume" trend to continue. This explains why I was asking for ways to predict daily volume before the day ends.