Class ADFFiniteSampleDistribution

  • All Implemented Interfaces:
    ProbabilityDistribution

    public class ADFFiniteSampleDistribution
    extends EmpiricalDistribution
    This class computes the finite sample distribution of the Augmented Dickey-Fuller (ADF) test statistics. There are three main versions of the test and thus three possible asymptotic distributions:
    1. test for a unit root without drift or time trend (NO_CONSTANT);
    2. test for a unit root with drift (CONSTANT);
    3. test for a unit root with drift and deterministic time trend (CONSTANT_TIME).
    Note that our results are different from those in R. The p-values in R are interpolated using the values from Table 4.2, p. 103 of Banerjee et al. (1993). SuanShu computes the exact p-values (and hence the whole cdf). See http://numericalmethod.com/forum/index.php/topic,122.0.html
    See Also:
    • D. A. Dickey and W. A. Fuller, "Distribution of the Estimators for Autoregressive Time Series with a Unit Root," J. Amer. Stat. Assoc., vol. 74, pp. 427-431, 1979.
    • E. Said and D. A. Dickey, "Testing for Unit Roots in Autoregressive Moving Average Models of Unknown Order," Biometrika, vol. 71, 599-607, 1984.
    • A. Banerjee et al., "ch. 4, pp. 99-135," Cointegration, Error Correction, and the Econometric Analysis of Non-Stationary Data, Oxford, Oxford University Press, 1993.
    • Constructor Detail

      • ADFFiniteSampleDistribution

        public ADFFiniteSampleDistribution​(int sampleSize,
                                           TrendType trend,
                                           boolean lagAdjust,
                                           int lagOrder,
                                           int truncation,
                                           int nSims,
                                           long seed)
        Construct a finite sample distribution for the Augmented Dickey-Fuller test statistic.
        Parameters:
        sampleSize - the (finite) sample size
        trend - the type of Augmented Dickey-Fuller test
        lagAdjust - true if the distribution is adjusted for lags
        lagOrder - the lag order; lagOrder = 0 yields the original Dickey-Fuller distribution
        truncation - the number of truncated values
        nSims - the number of simulations
      • ADFFiniteSampleDistribution

        public ADFFiniteSampleDistribution​(int sampleSize,
                                           TrendType trend,
                                           boolean lagAdjust,
                                           int lagOrder)
        Construct a finite sample distribution for the Augmented Dickey-Fuller test statistic. The number of truncated values is 50.
        Parameters:
        sampleSize - the (finite) sample size
        trend - the type of Augmented Dickey-Fuller test
        lagAdjust - true if the distribution is adjusted for lags
        lagOrder - the lag order; lagOrder = 0 yields the original Dickey-Fuller distribution
      • ADFFiniteSampleDistribution

        public ADFFiniteSampleDistribution​(int sampleSize,
                                           TrendType trend)
        Construct a finite sample distribution for the original Dickey-Fuller test statistic. We do not adjust for the lag.
        Parameters:
        sampleSize - the (finite) sample size
        trend - the type of Augmented Dickey-Fuller test
      • ADFFiniteSampleDistribution

        public ADFFiniteSampleDistribution​(int sampleSize)
        Construct a finite sample distribution for the Augmented Dickey-Fuller test statistic. We test for a unit root with a drift and a deterministic time trend.
        Parameters:
        sampleSize - the (finite) sample size