The GVAR model and hence the codes are set up to estimate VECMX models which are then aggregated to yield the global GVAR model. In my implementation, it turns out the cointegrating relationships are quite tricky and hence I am looking to estimate the model using stationary variables. One option I have in mind is to feed in I(1) variables (say in log levels) and then manually set the cointegrating ranks to zeros. Will this strategy work? Any hint will be greatly appreciated.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
This is effectively a strategy that ignores cointegration. As long as that is what you want to do then it is feasible within the GVAR program and it will work.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The GVAR model and hence the codes are set up to estimate VECMX models which are then aggregated to yield the global GVAR model. In my implementation, it turns out the cointegrating relationships are quite tricky and hence I am looking to estimate the model using stationary variables. One option I have in mind is to feed in I(1) variables (say in log levels) and then manually set the cointegrating ranks to zeros. Will this strategy work? Any hint will be greatly appreciated.
This is effectively a strategy that ignores cointegration. As long as that is what you want to do then it is feasible within the GVAR program and it will work.