Inference on Granger-Causality in the Frequency Domain

Implements unconditional and conditional Granger-causality spectra in the frequency domain, bootstrap inference for both spectra and their difference, and the Breitung-Candelon parametric tests. The bootstrap procedures follow Farnè and Montanari (2022) .

grangers

grangers implements estimation and inference for Granger-causality spectra in the frequency domain.

The methodological reference is:

Farnè, M. and Montanari, A. (2022). A Bootstrap Method to Test Granger-Causality in the Frequency Domain. Computational Economics, 59, 935–966. doi:10.1007/s10614-021-10112-x.

Exported functions

• Granger.unconditional() estimates the unconditional spectrum.
• Granger.conditional() estimates the conditional spectrum.
• Granger.inference.unconditional() performs bootstrap inference for the unconditional spectrum.
• Granger.inference.conditional() performs bootstrap inference for the conditional spectrum.
• Granger.inference.difference() performs bootstrap inference for the difference between unconditional and conditional spectra.
• bc_test_uncond() implements the unconditional Breitung–Candelon test.
• bc_test_cond() implements the conditional Breitung–Candelon test.

Data

The package contains euro_area_indicators, with quarterly Euro Area indicators used in the empirical analysis.

Installation

install.packages("grangers")