Reputation: 1
I am struggling with an econometric theory problem. I have a time series of yield daily data on government bond:
\{ y \}_{t=0}^{T}
which has clearly a unit root (DF test confirmed). Provided that the first-difference \Delta y_t = y_t - y_{t-1}
is instead stationary, I would like to run the following regression analysis:
\Delta y_t = \gamma (\Delta z_t) + \beta x_{t-1} + \epsilon_t
where x_{t-1}
and \epsilon_t
are both i.i.d. and z_t
is itself a time series with unit root.
Can I run this regression consistently and interpret the coefficient \beta
as the correlation between x_{t-1}
and \Delta y_t
, or is there anything I am missing here?
I have searched for references on the following textbooks, without finding anything which clarified my doubt:
Upvotes: 0
Views: 21