A Report On Engle Granger Cointegration Test

2320 Words Aug 9th, 2014 10 Pages
4. Empirical Results
In this section, we discuss our findings of Engle–Granger cointegration test which we applied in order to identify whether there is cointegration relationship between dependent variable – the real non-oil GDP and independent variables - real credit to the private sector and non-oil sector real effective exchange rate. The steps of the EG approach have been undertaken in order to obtain the long-run model that explains the relationship between these variables.

4.1. Unit Root Test
First of all, variables should be given in log levels in order to alleviate the problem of serial correlation and the elasticity of the coefficients. The results of ADF unit root test in levels concludes that all three variables - seasonally adjusted real non-oil GDP (ln_rgdp_noil_sa), non-oil sector real effective exchange rate (ln_reer_noil), and real credit to private sector (ln_rcred_to_ps) are non-stationary as the ADF test fails to reject the null hypothesis of stationary series indicating that the series have unit root. If the variables in the model are non stationary estimated regression model can be spurious or it is non-sense model. We cannot accept non-sense model.
To identify the number of unit roots the variables are furthermore examined in the 1st difference form. That is, in the first difference, all the variables reject the null hypothesis even at 1% significance level. This indicates that all three variables are stationary in their first difference and should…

More about A Report On Engle Granger Cointegration Test

Open Document