gc_tests

Purpose

Computes the tests for Granger causality of specified variables.

Format

GC_out = GC_tests(data, vnames, pmax, ic, test, Nboot)

Parameters:

• data (Txk matrix) – Data to be tested with k individual variables each in a separate column.

• vnames (String array) – Variable names.

• pmax (Scalar) – Maximum number of lags.

• ic (Scalar) –

  The information criterion used for choosing lags.

  1	Akaike.
  2	Schwarz.
  3	t-stat significance.

• test –

  Test option for Granger causality

  0	Granger causality.
  1	Toda & Yamamoto
  2	Single Fourier-frequency Granger causality.
  4	Single Fourier frequency Toda & Yamamoto.
  5	Cumulative Fourier-frequency Toda & Yomamoto

• Nboot (Scalar) – Number of bootstrap replications.

Returns:

GC_out (Kx5 Matrix) – Results matrix containing Wald stat~P-values~Bootstrap P-values~Lags~Frequency

Examples

library tspdlib;

// Number of bootstrap replications
Nboot= 1000;

// Number of observations
T  = 188;

// Number of lags in VAR model
pmax = 12;

// Information criterion
ic   = 1;

// Set to perform Granger Causality
test = 0;

// Load data matrix
GCdata = loadd(getGAUSSHome() $+ "pkgs/tspdlib/examples/TScaus.dat");
data   = ln(GCdata);

// Variable names
vnames = "y1"$|"y2"$|"y3"$|"y4";

// Toda & Yamamoto test
test = 1;

// Run Granger tests
GC_out = GC_tests(data, vnames, pmax, ic, test, Nboot);

Source

gctests.src

See also

Functions panel_fisher(), panel_zhnc(), panel_surwald()