Greetings!
I'm working with the Panel dataset, where N = 4, T = 41, 9 variables (later, I exclude 2 variables for the cointegration test) & this is not a dynamic model. Here, cross-sectional dependency and slope heterogeneity exist. For that, the second generation unit root test (Pesaran-CIPS) was applied, which confirmed that 7 variables are I(1) and the remaining 2 variables are I(0). In this case, I used Westerlund (2007) for the cointegration test and Westerlund & Edgerton (2007) for the bootstrap cointegration test, and it reveals no cointegration exists.
My queries are as follows: 1. What is the stata command for the "Durbin-Hausman cointegration test, developed by Westerlund (2008)"? It is also a panel cointegration test that is used when there is present cross-section dependence and the series are in different orders of cointegration, I(0) and I(1).
2. Which one is the perfect model for capturing the short-run coefficient with cross-sectional dependence?
Thanks,
Rakib
I'm working with the Panel dataset, where N = 4, T = 41, 9 variables (later, I exclude 2 variables for the cointegration test) & this is not a dynamic model. Here, cross-sectional dependency and slope heterogeneity exist. For that, the second generation unit root test (Pesaran-CIPS) was applied, which confirmed that 7 variables are I(1) and the remaining 2 variables are I(0). In this case, I used Westerlund (2007) for the cointegration test and Westerlund & Edgerton (2007) for the bootstrap cointegration test, and it reveals no cointegration exists.
My queries are as follows: 1. What is the stata command for the "Durbin-Hausman cointegration test, developed by Westerlund (2008)"? It is also a panel cointegration test that is used when there is present cross-section dependence and the series are in different orders of cointegration, I(0) and I(1).
2. Which one is the perfect model for capturing the short-run coefficient with cross-sectional dependence?
Thanks,
Rakib
