Replication package for "Macroeconomic Shocks, Institutional Quality, and Banking-Sector Credit Risk in Emerging Europe"
Description
This dataset contains the full replication package for the study "Macroeconomic Shocks, Institutional Quality, and Banking-Sector Credit Risk in Emerging Europe". The package documents the construction of a country-year panel for 14 European banking systems over 2007-2024 and the estimation workflow used in the paper. It includes the starting source panel, the FX-corrected panel, intermediate analysis datasets, reproducible code for all analytical steps, diagnostic files, tables, figures, logs, and a machine-readable manifest. The analysis covers dynamic panel estimation of banking-sector credit risk, threshold modeling, robustness checks, referee diagnostics, and scenario simulation. The package is designed as a single canonical release so that all reported tables and figures are internally consistent and traceable to one reproducible run. The archive includes Python scripts, one R script for the GMM diagnostic appendix, processed CSV and Excel data files, bootstrap draws, and publication-ready outputs. The package is intended for replication, verification, and reuse in research on banking-sector credit risk, macro-financial transmission, institutional quality, and emerging European financial systems.
Files
Steps to reproduce
1. Download and unzip the replication package. 2. Create a clean working directory and preserve the internal folder structure. 3. Install Python dependencies listed in requirements.txt: - python3 -m venv .venv - source .venv/bin/activate - pip install -r requirements.txt 4. Install the required R package for the dynamic-panel GMM diagnostic: - R -q -e 'install.packages("plm", repos="https://cloud.r-project.org ")' 5. Place the starting source file in the data/ directory: - panel_final_v4_liquidity_adjusted.xlsx 6. Run the full pipeline from the project root: - bash run_all.sh 7. The workflow will: - correct FX breaks around euro adoption, - construct the analysis panel, - estimate the baseline dynamic panel model, - estimate the threshold model, - run robustness and referee diagnostics, - simulate scenario paths, - generate tables, figures, logs, and a manifest. 8. Final outputs are written to: - outputs/tables/ - outputs/figures/ - logs/ - data/exports/ 9. Validate the release by running: - python3 code/validate_release.py
Institutions
- University of NisCentral Serbia, Niš
- College of Slavonski BrodBrod-Posavina, Slavonski Brod
Categories
Funders
- Ministry of Education, Science and Technological DevelopmentCentral Serbia, BelgradeGrant ID: 451-03-33/2026-03/200371