Rule for the increase in Ranking over time

How do I create a rule that looks for the increase in rank of a stock within my RS over time?
Say I wanted to give a bonus if a stock has increased its rank by:

  1. percentage, or
  2. x number of positions
  3. Over the last: week, month, etc…


Check out RankPrev()

RankPrev doesn’t exist within a ranking system. It would be great if we have a function to get for example previous rank value (with a lookback period) of a specific node within a ranking system.

Right. Anyone else know a trick to do this? Seems like it would be a fairly helpful RS factor.


I use this rank formula to get the percent change in a factor over X time, in this case one month:

(Inst%Own- Fhist(“Inst%Own”,4))/Fhist(“Inst%Own”,4)

Change out the factor “Inst%Own” to whatever…

I don’t know how you could do it with a rank. But this formula with create a rank based on the change in the factor. You could create several composite nodes for whatever time frame you are measuring. For example I compare the one month change in a factor to the three month change.

Then rank based on greatest change for the past month over three month change:


*I use long hand percentile change. Keeps me on point.

1 Like

Thanks, this is good info. I am interrested in your factor momemtum investing. I recently build my own ranking system (42 factors) in a custom universe in an attempt to create a screen of screen (or merging two similar investment strategy - but with different sectors and size). It worked pretty well, learned a lot on the use of veriables. - still learning…

I am reading the reference paper that you provided in earlier posts.