Hi everyone,
I would like to know if I'm doing a time-series analysis using the vecm model, do I need to do Optimal number of lags twice?
Here's the VECM model steps that I have performed:
1.ADF test (my variables are stationary at first difference)
2.Optimal number of lags (varsoc command in stata)
3.Johansen test for cointegration (I have cointegrating equation)
4.VECM estimation results
Due to the result of the vecm model is differential (D_x, D_y, D_z) . Do I need to add another Optimal number of lags before step 4 to determine the optimal number of lags of the vecm model?
Thanks in advance!
I would like to know if I'm doing a time-series analysis using the vecm model, do I need to do Optimal number of lags twice?
Here's the VECM model steps that I have performed:
1.ADF test (my variables are stationary at first difference)
2.Optimal number of lags (varsoc command in stata)
Code:
varsoc x y z
4.VECM estimation results
Due to the result of the vecm model is differential (D_x, D_y, D_z) . Do I need to add another Optimal number of lags before step 4 to determine the optimal number of lags of the vecm model?
Code:
varsoc d.x d.y d.z
Comment