Busy busy! I’m hoping to start getting more organized with my new life schedule to make room for research. Outlook looks better as time goes on and I want to capitalize on it.
Here’s a baseline (time fixed) momentum stat I quickly did. There’s a trend in my work that I start with very basic ideas and questions and gradually add more and more elements to it, we’ll see where this one goes. The idea behind doing it this way (not to mention that it is the only way I can do it) is that it removes or confirms intuition about the theory you had about the work before you did any hard math on it giving you a solid foundation.
First I made a rolling weekly (1hrx120) pivot point. I marked every bar as being above or below the PP. Then I simply tracked the accuracy of the past positions to predict where the next point will be.
I didn’t expect to learn too much from this, but it actually produced some interesting take away points for being such a simple statistic. First, it’s that for most periods, you can follow the logic that if the current point is below the pivot, the next one will be as well (and vice versa for longs). The second, more interesting aspect, is that for durations of 120 and above, there seems to be evidence that the strength isn’t as high. More specifically, that as time goes on, there is an increased probability that the “trend” (if defined by the position relative the pivot) will end.