IJF_24

Replication code for the paper Labonne, P. (2024). “Asymmetric Uncertainty: Nowcasting using skewness in real-time data.” International Journal of Forecasting https://doi.org/10.1016/j.ijforecast.2024.05.003

The score-driven models are written C++ to make use of the automatic differentiation library CppAD. This code is then wrapped into a package using Rcpp. The package can be installed with:

devtools::install_github("paullabonne/IJF_24/RcppScoreDrivenDFM")

• R/intro.R loads/installs all the packages needed.
• R/sd_dfm.R contains functions for estimation, forecast evaluation, simulation etc.
• The quarto files in the notebooks folder replicate the paper, i.e., build the vintages, real-time recursive estimation, result analysis and simulations.