Can a time series be stationary in levels and non-stationary in first differences?
On a related note
My dependent variable is I(1), whereas my independent variables are I(0)
If I estimate a regression of the first differences in my dependent variable on all the I(0) variables
Is there a risk of a spurious regression?