Granger Causality test

[tyzellar]

Dear Tom, Could you please tell me what this test does?
I am applying the panel var model and I want to check the Granger Causality between the variables but I am not sure whether this code the correct one is. In addition, I would like to know how to calculate the DFC?
Thank you in advance!

***Granger causality tests*****
dofor x=hp cpi
linreg(dfc=85,robust,print) x
# hp{1 to optlag} cpi{1 to optlag} gdp{1 to optlag} $
ir{1 to optlag}
summarize
# ir{1 to optlag}
exclude
# ir{1 to optlag}
end dofor x

[TomDoan]

I'm not sure the point (in this context) of the SUMMARIZE instruction---that computes the sum of the IR lag coefficients and the corresponding standard error, but I don't know what one would make of the value. The EXCLUDE does a "Granger causality" test for IR to whatever the X is on a given pass. However, this is looking at just one variable in a four variable system, so the economic significance is limited---see causality testing with 3 or more variables.

[tyzellar]

Dear Tom, Thank you a thousand times!!!

[tyzellar]

Dear Tom, Could you please tell me if it is possible to apply the block exogeneity in cases of panel data models.
Thank you in advance!