My model is a dynamic model with three endogeneous variables (that is three equations). The unit root test is a mixture of i(1) and i(0). I cannot use PVECM because of the unit root result and i cannot use VAR beacuse of possible cointegration among the models. So, i need to perform BOUND TEST of cointegration to know the cointegration status of the models. But i don't know the stata code for executing BOUND TEST of cointegration.
You can install xtpmg if you assume there is no cross sectional dependence. However if there's cross sectional dependence you can install xtdcce2 in stats. You can aswell download the PDF documents that explains these routines..
Hamid Muili, I think this document could help you a little bit. All the best!
https://www.stata.com/features/overview/panel-data-cointegration-tests/
@Prince Fosu, thanks for the suggestion its indeed a great read. However, the questioner indicated that his model is a dynamic model. Thus, the Kao, westerlund and others are mostly appropriate for a static model. More so, I assumed he has correctly examined the unit root and stationarity property of his variables before detecting the model consist of mixed integration. Dimitrios et al (2020) also noted that panel ardl helps alleviate the problem of endogeneity by including lags...
Best Regards
- Badges
- Science topic