ρ < 1; it has a non-standard asymptotic
distribution when ρ = 1, being skewed to
the left.
Therefore we cannot apply the standard t-test to test the unit root null. This test will reject the null too often; the actual test size will be greater than the nominal test size.
White (1956) showed that under H0:ρ=1, τ does have a stable limiting distribution.
Dickey (1976) and Dickey and Fuller (1979) tabulated the percentiles of the this distribution. (How?) The percentiles of the Dickey-Fuller distribution are available in many time series textbooks (including the Enders and Hamilton texts).
See Table
Notes –
the unit root test is a one-sided test. Reject H0 if τ is “too negative”, i.e., if ρ-hat is too much less than one.
the median of the DF distribution is about
-0.5.
the 0.025 percentile of the standard normal is -1.96. The 0.025 percentile of the DF distribution is -2.23. The 0.05 percentile of the standard normal is -1.65. The 0.05 of the DF is -1.95
The asymptotic distribution seems to be appropriate even if T is as small as 25.
an equivalent test: Regress Δyt on yt and use as the test statistic.
Under H0 the test-statistic
also converges in distribution. The
percentiles of this distribution were also
tabulated by Dickey and Fuller. Unit
root tests based on this test statistic
seem to be less powerful and, so, less
widely used.
Suppose we modify the model and null hypothesis as follows:
yt = α + ρyt-1 + εt , εt ~ iid (0,σ2), -1 < ρ < 1
H0: ρ = 1 and α = 0
HA: ρ < 1
The difference between this and the previous case? Under the I(0) alternative, yt can have a non-zero mean. But, in neither case does yt have a deterministic trend component. (Once we allow the ε’s to be serially correlated, this case could be appropriate for testing for a unit root in the unemployment or inflation rates.)
Under the null hypothesis, the t-statistic
from the regression of yt on 1, yt-1, or, equivalently, the t-statistic
Share with your friends: |