README - Reproducing "Large datasets for the Euro Area and its member countries and the dynamic effects of the common monetary policy", Barigozzi M., Lissona C. and Tonni L. (2026)

This folder contains the MATLAB code that reproduces the empirical results reported in the paper. The code was developed and tested in MATLAB R2026a.


CONTENTS

  ccvar_monthly_iv.m         Estimates the CC-VAR identified with the high-frequency
                             instrument; produces Tables 3 and 4 and saves the results
                             used by the other two scripts.
  plot_figures_iv.m          Loads the saved results and produces Figures 1, 2 and 3.
  correlation_analysis_IRFs.m  Reproduces Table 6.

  Functions/                 Helper functions called by the scripts.
  data/                      Monthly country datasets and the high-frequency instrument.
  institutional_data/        Country characteristics used in the correlation analysis.
  outputs/                   Saved estimation results (Results.mat).
  Figures/                   Figures produced by plot_figures_iv.m.


1. ccvar_monthly_iv.m

   Reproduces:
   - Table 3: estimated number of factors
   - Table 4: share of variance explained by the common factors

   Tables 3 and 4 are printed in the command window, together with the first-stage
   F-statistic and the Olea effective F-statistic for the high-frequency instrument.

   The estimation results are saved to:
       outputs/ccvar_monthly_iv/Results.mat

   Confidence bands. The script has a "conf" flag at the top:
       conf = 0   point estimates only, no confidence bands (about 1 minute)
       conf = 1   bootstrap 68% confidence bands (about 40 minutes on a 2021 MacBook Pro)
   Set conf = 1 to obtain the confidence bands shown in Figures 1 and 3 and the
   significance encoding in Figure 2.


2. plot_figures_iv.m

   Loads outputs/ccvar_monthly_iv/Results.mat and produces the three main-text
   figures, saved in the Figures folder:

   - Figure 1  (Figure1_EA.pdf)        EA-wide IRFs
   - Figure 2  (Figure2_countries.pdf) Country-level IRFs
   - Figure 3  Country-minus-EA IRF differences, split for readability into one
               subfigure per variable:
                 Figure3a_IP.pdf, Figure3b_HICP.pdf, Figure3c_LTIRT.pdf,
                 Figure3d_SHIX.pdf, Figure3e_UNETOT.pdf

   If Results.mat was produced with conf = 0, the figures show point estimates only
   (no bands; in Figure 2 all responses appear dotted). Run ccvar_monthly_iv.m with
   conf = 1 first to reproduce the figures exactly as in the paper.


3. correlation_analysis_IRFs.m

   Reproduces Table 6, printed in the command window. It uses the estimation results
   in outputs/ccvar_monthly_iv/Results.mat.


ORDER OF EXECUTION

   1. ccvar_monthly_iv.m
   2. plot_figures_iv.m
   3. correlation_analysis_IRFs.m
