I’ve been running some simple rank charts on single factors. When compared to similar screen performance (say using “quickrank”), the results in some cases have been polar opposites at worst, and inconsistent at best. I have checked and re-checked all settings and inputs as well.
Are there any known issues with the rank charts? When I recreate a new ranking system with the same factor, there seems to be a re-fresh and there is consistency.
However just this morning having more issues.
Any help would be much appreciated.
Thanks,
Ryan
Ryan -
This is a bit vague. Can you provide some specifics? Perhaps a link to the ranking systems and screens that are giving you trouble?
Thanks,
Yuval, that was vague, wasn’t it? 
Take the factor salesgrowth%ttm for example, rank higher as better.
I started with a ranking system where I was adding various factors for testing, call it a working rank system (Rank System #1). Currently it has roughly 20 factors, turning them off and on as required. See rank chart attached.
But when starting a new ranking system (#2), the performance is almost exactly inverse, see attached.
Using “quick rank” on a screen for highest ranked quintile (roughly 50 on a 250 stock universe), results most resemble rank system #2.
All settings the same.
The only thing I can think of is that Rank System #1 had too many nodes, however I have several rank systems w/ 20+ nodes.
Whatever help you can offer would be much appreciated.
Thanks,
Ryan



Ryan -
Could you please provide me with the URLs of the two ranking systems? I need to look at this behavior closely.
Thanks!
Yuval, just PM’d you on Seeking Alpha.
Thanks,
Ryan
Yuval, thanks for the feedback.
The rank arrow at the very top left of the rank system top line was somehow inverted to rank lower values, not higher (which is usually my default). I usually don’t change this, must have clicked it inadvertently.
Thanks for the help, I can sleep better now 
Cheers,
Ryan