I have used factor momentum as a core foundation to my strategy for over 20 years. I can tell you that it works very well.
I have about 35 factors I use, I settled on these after three years of testing. They have stood the test of time since 1997. Prior to P123 I did everything in Excel.
To give you a start, I have a ranking system with three composite nodes. The first calculates the percent change of each factor for one month, the second calcs the three month change. This is simply the percent change in value for each factor.
The factors inside these two composite nodes are given a weight of zero, or equal weight. The one month and three month parent composite nodes are given 1% weight.
The third composite node is (NodeRank(“1mo_Change”)-NodeRank(“3mo_Change”))/NodeRank(“3mo_Change”).
The third node is given 98% weight. This works for me but there may be other ways to go about this.
The trick is then what? Now you know what stocks have the highest factor momentum. But simply that a stock’s Inst%Owned went from 60 to 80 means nothing.
In my case I have developed a process which tells me which factors the market is currently rewarding. It was extremely complicated to develop on P123 (I crashed the servers a couple of times during development :). I spent years writing screens to do this, and wasn’t able to do it at all until variables were introduced. Not to mention my system uses about 1300 custom formulas to make it all come together. Once I finally put it all together it now takes about 30 seconds or so to run the screen on a 4500 stock universe.
The ranking system is just the starting point.
Bottom line, factor momentum works. Its been in use since the late 1980’s by various money managers.