hey everybody....
i have a question concerning granger causality wald test when the dependent series is trend-stationary... i tried the following and was wondering if this is correct...
y(t) = a[y(t-1)] + b[x(t-1)] + cT + dT-squared ...
where T and T-squared are trends T=1,2,3,4 ... and T-squared=1,4,9,16...
and just looked the statistical significance of the x-variable to decide if there is causality or not [also ran "test x(t-1)=0", which confirmed the t-test statistic]...
is this a valid kind of approximation to the (one-way) granger causality wald test assuming only one lag?
thanks a lot for your help ... take care, matthias
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
