Replication package for "Renewable Energy Penetration and Wholesale Electricity Price Volatility in Europe"
Authors/Creators
Description
This replication package accompanies the article "Renewable Energy Penetration and Wholesale Electricity Price Volatility in Europe" by Oksana Liashenko, Ihor Turskyy and Olga Demianiuk, submitted to Energies (MDPI).
It contains the data and Python code needed to reproduce every table and figure in the article. The data are a monthly panel of 29 European countries from January 2015 to October 2024 (3,224 country-months). The panel combines day-ahead wholesale electricity prices with generation shares by technology (wind, solar, gas, coal, nuclear, hydro), both compiled from Ember datasets released under CC BY 4.0.
The analysis covers:
(i) two-way (country and month) fixed-effects regressions of wholesale price volatility on wind-and-solar penetration, with Driscoll–Kraay standard errors;
(ii) robustness checks: four alternative volatility measures, a non-overlapping annual specification, lagged regressors, Oster bounds for unobserved selection, and instrumental-variable (2SLS) estimation;
(iii) heterogeneity analyses by dispatchable-thermal share, hydro share and the 2021–2023 energy crisis, plus a price-level (merit-order) check;
(iv) country-level wavelet coherence with Monte Carlo red-noise significance testing (fixed random seed).
Contents:
- data/energy_panel_final.csv — the analytical panel;
- data/variables.csv — variable definitions, units and sources;
- code/ — Python scripts: 01_panel_estimates.py (Tables 1–4, A1–A6 and diagnostics), 02_wavelet.py (Table B1), 03_figures.py (Figures 1–4 and B1), run_all.py (runs the full pipeline) and common.py (shared functions);
- output/ — all tables (CSV) and figures (PNG) as reported in the article;
- requirements.txt — exact Python package versions.
To reproduce, install the packages in requirements.txt and run python code/run_all.py from the package root. The panel estimates take seconds; the wavelet significance tests take about 30–40 minutes.
Source data: Ember, European Wholesale Electricity Price Data and Monthly Electricity Data (https://ember-energy.org), CC BY 4.0.