Forward Variables

Research
1

Hey,
I'm trying to figure out if variables such as earnings increasing in the future are strongly associated with stocks going up, and I wanted to break it down by industry, sector, capitalization, share price etc.

Its a long project and part of what I would intend to do is find the results of tickers before and after an earnings increase.

To find this out though, it would make things much easier if I can screen for stocks that will have Netincome, sales, analyst ratings, costs, etc changing in the future, and I'm not sure how to code this in the screener tool. This would allow me to immediately find out how important net income increases are for certain subsections of the market.

Does anyone have any thoughts?

As an example;
We can do NetIncCurQ for current net income, but I havent figured out how to run a scan for NetIncNextQ, and on all the other variables ive mentioned

Any advice would be realy nice.
Thanks,
James

2

The best way to do this is to adjust the as-of date at the top of the screen and then use historical values for your inputs and present values for your "future" values. I do this quite often: I'll run the screen a dozen or so times with different as-of dates, then download the results and analyze them. So, for example, to test the correlation between current profit margin and future income growth, you might use an as-of date a year ago, use ShowVar to get values for NPMgn%PY and IncBTaxGr%A, and see if there's any correlation.

3

Thank you, Yuval. Ill do that!

4

There's one drawback to this approach, though, and that is survivorship bias. In your data you won't get any companies that were in your universe a year prior to the as-of-date and are no longer in your universe now.