Cointegration And Unit Roots Flashcards
How do you test whether a process is stationary or not
Dickey-Fuller (DF) or KPSS
What happens to the AR(1) process when ø1 = 1
It becomes a random walk model: yt = yt-1 + et
Integration process I(1) for example
Yt needs to be differenced 1 time to become stationary it is integrated to the order 1
How to detect non-stationarity
Augmented Dickey-Fuller test
Phillips-Perron
KPSS
The Dickey-Fuller Hypothesis
H0: The series contains a unit root
H1: The series is stationary
when is ADF valid
If et in the test regression is considered white noise
What is expected of the error terms in the PP test
Serially correlated
KPSS test hypothesis
H0: yt is stationary
H1: yt is non-stationary
Difference between unit root test and stationary test
Unit root test H0: has a unit root
Stationary test h0: stationarity