Granger causality tests

[Ulla Andersen]

Hello,

I'm trying to run a Granger causality test on panel data to check whether bank profits (dependent variable) Granger causes capital (independent variable). I have a strongly balanced panel so I was going to use the xtgcause command, however, I do not get any results, as seen in the capture below.
I cannot seem to fix it or find any information for why this does not work, so I was wondering if I could circumvent this issue with another stata command or approach? I would greatly appreciate your help and any input!



Best regards,
Ulla

[Will obv]

Hi Ulla,

Did you manage to find a solution to this problem? I am having the same issue for certain variables in my panel data set and I have found no luck in finding a solution for this so far.

Thanks

[Ulla Andersen]

Hey,
This was quite a while ago, but I ended up regressing each dependent and independent variable on 3 lags of itself and the other variable. I.e. regress dep variable on 3 lags of itself and 3 lags of indep variable and in another regression regress indep variable on 3 lags of itself and 3 lags of the dep variable. Finally, I tested the significance of the 3 lagged coefficients. Hope this helped!