I am studying the effects of FDI on the receiving nation and my previous method of VECM is not working at all, all my results are non significant! I am studying the effects on the real exchange rate and my variables are LNGDP LNFDI and many controls. I did unit root tests and found that the variables are all stationary in I(1) or I(0) and are co integrated by johansen. Therefore I chose VECM but my p values are massive !
How would you approach this problem from scratch? Would you even use a VECM?
Any guidance would be really appreciated as this is for my thesis,
Regards,
TBG