Hi all, I’m having trouble determining if I can do this and appreciate thoughts of how I might approach.
My underlying goal is to investigate changing valuation levels of ranking systems over time. Examples might include: “Is the top quintile of Momentum relatively more or less expensive than it used to be?” Same for Quality or Growth or any other factor used. The end analysis might draw attention to factor cyclicality. Presently I’m planning to take measurements at every 6 months from 12-2005 forward.
What I’m having trouble understanding is how to restrict universes based on the results of a Ranking System, or how to create a new universe from the result of a screen that is based on the results of a ranking system.
I’d like to feed various top quintile universes at different points in time into the custom series tool so I can see median valuations at that point in time (like earnings yields, ebit/ev, etc) and compare that to develop a sense of what factors might be over or underpriced at different points in time compared to historicals.
While I can use something like Rating(“Growth factor ranking system”)>80 in a screen to restrict the results of the screen, I cannot use that same rule as an input to restrict the contents of a universe, and cannot use ranking rules to restrict results in the custom series tool - so not sure how to approach the problem. Is there another way I should think of it?
(edit: for now I’m going through and manually calculating median valuations from screen report results - I can get there this way - it’s just s l o w )
Any ideas on how to approach this efficiently? Thoughts appreciated.