This guy has an interesting approach. Multiplying factors. For example, z score, x yield, time discount.
https://seekingalpha.com/article/4522005-the-high-high-low-c…
Multiplying factors. For example, z score, x yield, time discount.
That’s one of the basic MI techniques used around here for (literally) decades.
It’s a quick way to ensure that two or more criteria are all good, rather than one excellent and the other poor.
And it doesn’t require putting in a tuning parameter for each factor, which leads easily to overtuning.
As a random example, YLDEARNYEAR from 2003 multiplies earnings yield and dividend yield.
EG_PELA from 2001 multiplies earnings growth and projected earnings growth rates.
Jim
6 Likes