Page 1 of 1

VECM WITH TWO COINTEGRATING EQUATIONS: HOW DO I INTERPRET?

Posted: Fri May 10, 2013 12:18 pm
by Osifo
Please i would love if you could help me in understanding my VECM model. Specifically, i have a multi-variable regression model. That is, GDP = f(DSXP, DRR , DSR, EXRT), I first converted all the variables to logs and then tested for unit root in the all the series. I found that each variable was nonstationary at levels but became stationary after first difference. Since all the variables are I(1), I proceeded to Johanse cointegration test. First I used one lag but the Trace Test and Max-Eigen test indicated no cointegration. Second i use two lags and both test (Trace and Max-Eigen) indicated two cointegrated equations. So i proceed to run the VECM model using two lags and two cointegrating equations. now my result is out and has two cointegration equations, both are negative and significant, but i dont know how to interpret the result. Also my independent variables now appear in log-difference or difference-log form with two lags. I have attached a copy of the result for your perusal.