Re: doubts regarding cointegration. emergency
Posted: Tue May 26, 2015 3:14 am
Difficult to answer the questions without specifying what you are aiming to accomplish. If you are trying to asses whether X follows from Y with granger causality then those series need to be stationary. Generally prices are I(1) and returns are I(0), and your results are kind of indicative of that also. I would convert both the series to returns and run the Granger causality then.