Cointegration (?) of two time series

[kevinvanb]

Hi guys,

I'm struggeling with a problem regarding my thesis. I have two time-series, momentum and dividend yield. When I run a ADF test on momentum it says 'no unit root'
but on dividend yield, it do has a unit root.

When I take the first difference of log dividend yield, it no longer has a unit root.

Next to that, when I run a regression on momentum with dividend yield (first difference), and do an ADF test on the residuals, it rejects a unit root, which implies the
two time series are cointegrated (?) according to my Empirical Finance classes.

Can anyone confirm/reject this and tell me how to proceed?

Many thanks!