Hi seniors and friends,
I am working with Johansen co-integration, by definition this test requires non stationary data to model.
My situation is:
In my dataset , 2 variables are stationary at level, trend and intercept, these two variables are non stationary at level and intercept.
My question is,
Can I use this mix of different order of integration in johansen co-integration analysis?
Yes, use this mix of a different order of integration in Johansen co-integration analysis?
Yasodhagayathri A The Engle-Granger test statistic for cointegration is equivalent to an ADF unit root test on the cointegration regression residuals: There is no cointegration if the residuals include a unit root. The ADF test's null hypothesis is that the residuals have a unit root.
- Badges
- Science topic