Granger casuality: which kind of variables can you compared?

#1
Good morning and thanks in advice for your time.
Can I use Granger casuality for studying if a financial variable (stock price of a company) can be caused by a non financial variable (like the number of hours the employers work)? If not, do I have to consider my non financial variables as exogenous variables referred to the financial variable?
Thanks
 

noetsi

Fortran must die
#2
You can use a granger causality test in theory to test if one time series influences another time series, it does not have to be applied to financial or non-financial data. Granger himself apparently questioned whether the test made sense for some non-economic methods- but I don't know if that actually makes sense or not (or when it would and would not make sense).

I believe Granger considers a variable exogenous if it influences Y without being influenced by Y. I am new to these tests, but at least some sources suggesting regressing X on Y and Y on x - ideally X would cause Y without being influenced by Y [although what exactly you do if this is not the case is something I am unclear on].