Hi,
I am running a Panel-ARDL model to test the relationship between financial development and renewable energy consumption by type. Having found evidence for cross-sectional dependence in my N=45 T=27 dataset (I also have the same problem for N=45 T=7 dataset), I am not sure how to test for cointegration in my model. I am aware of Westerlund's tests, but I have read these are only valid if T>N, is that true? I ran a Westerlund (2007) test using the code: xtwest lhydropc lFD lFD2, constant trend lags(0 2) leads(0 2) lrwindow(3) bootstrap(100), but I am not sure if it is valid given my N,T dimensions. Also not sure how I should interpret Ga as opposed to Gt, and likewise Pa as opposed to Pt?
Many thanks,
Sagar
