Hi,
Could someone please explain how the dickey-fuller test for cointegration works, and what is meant by the term "null hypothesis"? I should add that I've read the standard definitions Wikipedia, but I was hoping for something more intuitive.
I understand that the term "co-integrated" means that if you have two non-stationary time-series, a linear combination of the two is stationary. By stationary, I assume they mean the time series is mean-reverting.
Thanks!
Could someone please explain how the dickey-fuller test for cointegration works, and what is meant by the term "null hypothesis"? I should add that I've read the standard definitions Wikipedia, but I was hoping for something more intuitive.
I understand that the term "co-integrated" means that if you have two non-stationary time-series, a linear combination of the two is stationary. By stationary, I assume they mean the time series is mean-reverting.
Thanks!
