TVGC: Stata module to perform Time-Varying Granger Causality tests

tvgc implements the VAR-based time-varying Granger causality tests proposed by Shi, Phillips and Hurn (J. Time Series Anal., 2018). These are sequences of Wald statistics based on forward recursive estimation, rolling estimation, and recursive evolving estimation. The command also supports estimation of these three sequences of Wald statistics in the context of a Lag-Augmented VAR (LA-VAR) model, as recommended to allow for the possibility of integrated variables; see Shi, Hurn, Phillips (J. Fin. Econometrics, 2020) and the references therein. tvgc computes 90th, 95th, and 99th percentile bootstrap critical values following the bootstrap scheme advocated by Shi et al. (2018, 2020).