mvgc_adf (EXPERIMENTAL)

Augmented Dickey-Fuller unit root stationarity test

Syntax

   [tstat,cval] = mvgc_adf(X,alpha,q,pdeg)

Arguments

See also Common variable names and data structures.

input

   X          multi-trial time series data
   alpha      significance level
   q          number of lags; default: sqrt(number of observations)
   pdeg       time trend polynomial degree; may be
                 -1 : no deterministic part
                  0 : for constant term (default)
                  1 : for constant plus time-trend
                > 1 : for higher order polynomial (must be <= 5)

output

   tstat      matrix of ADF test statistics
   cval       ADF test critical value

Description

Calculates the Augmented Dickey-Fuller q-lag unit root t-test statistics tstat and critical value cval for a (possibly multi-trial) multivariate time series X at significance level alpha [1]. The time trend polynomial degree pdeg may take values between -1 and 5 (see output arguments above). The returned test statistics matrix tstat is N x n for N trials of an n-variate time series.

Note: Adapted from code written by James P. LeSage, modelled after a similar Gauss routine in a package called COINT.

References

[1] S. E. Said and D. A. Dickey, "Testing for Unit Roots in Autoregressive Moving Average Models of Unknown Order", Biometrika 71(3), 1984.

See also

mvgc_kpss | mvgc_demo_stats