Learn R Programming

⚠️There's a newer version (2.5.8) of this package.Take me there.

BGVAR: Bayesian Global Vector Autoregressions

Estimation of Bayesian Global Vector Autoregressions with different prior setups and the possibility to introduce stochastic volatility. Built-in priors include the SIMS, SSVS and NG prior. Post-processing functions allow for doing predictions, structurally identify the model with short-run or sign-restrictions and compute impulse response function, historical decompositions and forecast error variance decompositions. Plotting functions are also available.

Installation

BGVAR is available on CRAN. The latest development version can be installed from GitHub.

install.packages("BGVAR")
devtools::install_github("mboeck11/BGVAR")

Note that Mac OS needs gfortran binary packages to be installed. See also: https://gcc.gnu.org/wiki/GFortranBinaries.

Note that Windows OS needs the R package Rtools installed that you can compile code with Rcpp. There are some common issues which you find here: https://thecoatlessprofessor.com/programming/cpp/installing-rtools-for-compiled-code-via-rcpp/.

Usage

The core function of the package is bgvar() to estimate Bayesian Global Vector Autoregressions with different shrinkage prior setups. Calls can be heavily customized with respect to the specification details of the model, the MCMC chain, hyperparameter setup and various extra features. The output of the estimation can then be used for a variety of tools implemented for the BGVAR package.

Predictions are invoked with predict(), impulse responses are computed with irf(), forecast error variance decompositions can be called with fevd() and historical decompositions with hd(). Furthermore, counterfactual impulse responses are computed with irfcf() and conditional forecasts with cond.predict().

The package comes with standard methods to ease the analysis. The estimation output can be inspected with print(), summary(), fitted(), coef(), vcov() and residuals(). Default plot() is available for most outputs. All classes features print() methods. Various other helper functions to ease analysis are also available.

References

Crespo Cuaresma, J., Feldkircher, M. and F. Huber (2016) Forecasting with Global Vector Autoregressive Models: A Bayesian Approach. Journal of Applied Econometrics, Vol. 31(7), pp. 1371-1391.

Doan, T. R., Litterman, B. R. and C. A. Sims (1984) Forecasting and Conditional Projection Using Realistic Prior Distributions. Econometric Reviews, Vol. 3, pp. 1-100.

George, E.I., Sun, D. and S. Ni (2008) Bayesian stochastic search for var model restrictions. Journal of Econometrics, Vol. 142, pp. 553-580.

Huber, F. and M. Feldkircher (2016) Adaptive Shrinkage in Bayesian Vector Autoregressive Models. Journal of Business and Economic Statistics, Vol. 37(1), pp. 27-39.

Pesaran, M.H., Schuermann T. and S.M. Weiner (2004) Modeling Regional Interdependencies Using a Global Error-Correcting Macroeconometric Model. Journal of Business and Economic Statistics, Vol. 22, pp. 129-162.

Copy Link

Version

Install

install.packages('BGVAR')

Monthly Downloads

839

Version

2.3.1

License

GPL-3

Maintainer

Maximilian Boeck

Last Published

September 10th, 2021

Functions in BGVAR (2.3.1)

eerData

Example data set to replicate Feldkircher and Huber (2016)
bgvar

Estimation of Bayesian GVAR
fevd

Forecast Error Variance Decomposition
excel_to_list

Read Data from Excel
gfevd

Generalized Forecast Error Variance Decomposition
hd

Historical Decomposition
lps

Compute Log-Predictive Scores
plot

Graphical Summary of Output Created with bgvar
logLik

Extract Log-likelihood of Bayesian GVAR
pesaranData

pesaranData
residuals

Extract Residuals of Bayesian GVAR
rmse

Compute Root Mean Squared Errors
add_shockinfo

Adding shocks to 'shockinfo' argument
monthlyData

Monthly EU / G8 countries macroeconomic dataset
matrix_to_list

Convert Input Matrix to List
vcov

Extract Variance-covariance Matrix of Bayesian GVAR
summary

Summary of Bayesian GVAR
list_to_matrix

Convert Input List to Matrix
irf

Impulse Response Function
predict

Predictions
fitted

Extract Fitted Values of Bayesian GVAR
resid.corr.test

Residual Autocorrelation Test
get_shockinfo

Create shockinfo argument
conv.diag

MCMC Convergence Diagnostics
dic

Deviance Information Criterion
eerDataspf

eerData extended with expectations data
BGVAR-package

BGVAR: Bayesian Global Vector Autoregressions
coef

Extract Model Coefficients of Bayesian GVAR
avg.pair.cc

Average Pairwise Cross-Sectional Correlations
eerDatasmall

Example data set to show functionality of the package