Page 1 of 1


Posted: Sun Feb 14, 2021 4:08 am
by carlasophie

I have a question regarding the dickey fuller test in EViews. I'm estimating single and multi factor model (CAPM, Fama/French and Carhart). The times series returns of one portfolio is not stationary on level. After applying the augmented dickey fuller test with the first difference, it's stationary. Now I have to estimate the regression. My question is whether I have to use the first difference for all variables now or just for the time series that wasn't stationary at first.

Thank you for your help!
Best wishes from Germany