| Abstract: | The residual-based cointegration test suggested by Engle and Granger (Co-integration and error correction: representation, estimation, and testing. Econometrica. 1987;55(2):251–276]; EG) have been used widely in the literature. This pioneering EG test is quite intuitive and easy to implement. However, despite popularity, the power property of the EG test has been be an issue. In particular, the EG test imposes implicitly the so-called common factor restriction, which can lead to significant loss of power when the signal-to-noise ratio increases. In this paper, we consider a simple modification procedure. The proposed modification is shown to improve drastically the power properties of the EG test. In contrast to the conventional EG test, the power in the modified test improves (rather than decreases) significantly as the signal-to-noise ratio increases. |
