Actually this seems reasonable to me but needs more information to to be sure. Could you give your variable descriptions, and examples of your proposed models?

Thanks very much, David Booth

David Eugene Booth thank you so much for your response. Actually I have little knowledge in econometrics but if I’m to answer your question, I have two dependent variables (gdp proxied for economics growth and MDI proxied for poverty) then my independent variable which is Microfinance bank activities has 6 variables, the ones for economic growth are Microfinance banks asset, loan, deposit then additional X which is agricultural production (control x variable) the independent variable for poverty are Microfinance banks liquidity ratio, loan to deposit and no of active borrower. The variables annual data(20 years) were Collected for five different countries. I’m think of running a panel ARDL because the order of integration is I(0) I(1) after doing the unit root. i want to know if there is a way I make comparison between the countries?

Ah let me suggest these: https://b-ok.cc/book/5825760/ed313f

and

https://b-ok.cc/book/3588849/f27729

You shouild find them a big help I think. Best, D. Booth

Seems like this is exploratory - you are running several models to see what differences are in the data, withih & between countries. This is fine as long as you are clear that your results are descriptive. Any p-values are meaningless. You can formulate hypotheses; but need an independent dataset to test them.

Hello Kehinde Hammed, This makes your second question clearer, the PMG is the solution. you will have two models, one with GDP as dependent model and the other with MDI as dependent variable . the output of the PMG will show the effects of all the independent variables on the dependent in all countries and then in each country bur don't expect any post estimation test except normality