NEW VERSION: This code computes the Wbar and Zbar test statistics of the Homogenous Non Causality (HNC) hypothesis. Under the null hypothesis of HNC, there is no causal relationship for all the cross-units. Under the alternative, there is a causal relationship from X to Y at least for one cross-unit. The panel must be balanced. The Wbar statistic corresponds to the cross sectional average of the N standard individual Wald statistics of Granger non causality tests. The Zbar tild statistic corresponds to the standardized statistic (for fixed T sample). This new version of the code defines the sample size T as the number of observations used to estimate the model (taking into account the initial conditions) and correct a bugg in the definition of Zhnc (equation (9)). No change for the definition of Zhnc_tild (equation (26)), which is the test statistic that we recommend for finite sample. A special thank to to Gareth Thomas (IHS Markit, Eviews).
Working Paper (2018)
Müller Gernot and Hettig Thomas