Dear Experts,

Good Day!!

I have 137 cross sections (countries) and a T (time) of 17, as well as more than 13 covariates, and I've had a lot of trouble using advanced techniques to extract long and short run associations between variables. First, I used some sample adequacy and variable selection procedures to ensure that all of my variables are appropriate, all variables are relevant according to result. Cross sectional dependency and slope heterogeneity are also present in the variables. After that, I performed the CIPS and CADF unit root tests and discovered that there was a mixture of level and first deference stationarity. Since of this, I decided to do my empirical work using cointegration tests (Westerlund and Pedroni), CCE, MG, AMG, DCCEMG, ARDL, CS-ARDL, or CS-DL, but no one running because these approaches only use limited number of variables (In Stata). Could you help advise me on the techniques I should use to find long and short term relationships in this situation?

Note: Please suggest some advanced techniques. Thanks!!!!