I used an inbuilt Matlab function to check for Granger Causality between two time series P and T that have a correlation coefficient of 0.6.

The function is : https://www.mathworks.com/matlabcentral/fileexchange/25467-granger-causality-test/content/granger_cause.m

On running it, I got the following answer:

>> [fs,cv] = granger_cause(P, T,0.05,2)

fs =


cv =


where FS is the F-Statistic and cv is the critical value from the F-Distribution.

Does this show causality? Actually... what does it show?

  • $\begingroup$ In the source file, it is reported that "Note that if F > c_v we reject the null hypothesis that y does not Granger Cause x". Please clarify your question if it is not a MATLAB issue but rather a statistical question. $\endgroup$
    – chl
    Oct 28 '13 at 8:10

FS > critical value, which means that the result rejects the null hypothesis(T didn't granger caused P). Or simply, T caused P.

I think the parameter for calculating the F statistics is weird and just wrote a post asking the author.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.