Repository for the Master's thesis: "Existe-t-il des asymétries dans les ajustements des prix à la pompe ?"
Author: Gaël Pefourque
Academic Year: 2023–2024
Institution: Université Paris-Est Créteil (Master 1 Économie appliquée, parcours MASERATI)
This project investigates whether there are asymmetries in the adjustments of retail fuel prices in France in response to changes in the price of crude oil (Brent). The study focuses on the year 2022, leveraging high-frequency data at the gas station level.
It uses econometric modeling, specifically Error Correction Models (ECM) and cointegration analysis, to detect and quantify such asymmetries.
- Evidence of asymmetric price adjustments: fuel prices increase faster than they decrease following changes in crude oil prices.
- Competitive intensity (proxied by number of nearby stations) significantly affects price behavior.
- Margins are higher and price adjustments more frequent on highways.
- Fuel prices: Daily prices for diesel, SP95, SP95-E10, and SP98 from French gas stations (source: prix-carburants.gouv.fr).
- Brent price: Daily Brent spot price from U.S. Energy Information Administration.
- Exchange rate: USD/EUR from Boursorama.
The dataset includes:
- Over 9 million price observations.
- Station-level details: location, fuel type, station type (highway vs road), etc.
- Conversion of Brent prices into EUR/litre.
- Calculation of HT (tax-excluded) fuel prices.
- Computation of competition index (stations within 5km).
- Time series plots of fuel prices vs Brent price.
- Correlation analysis between different fuel types and Brent.
- Box plots of price spreads between pump and crude oil prices.
- Distribution of price durations and margins by station type.
- Regional maps of price averages and station densities.
- Interactive map showcasing station clusters. see here:
- Step 1: Estimate long-term relationship (cointegration) between Brent and fuel prices.
- Step 2: Apply ECM with asymmetry using Engle-Granger two-step procedure.
- Step 3: Use Wald test to detect significance of asymmetry.
- Python 3.9+
- Pandas, Numpy
- Statsmodels
- Matplotlib / Seaborn
- Scikit-learn
├── preprocessing/ # Jupyter notebooks with preprocessing
├── eda/ # Jupyter notebooks with exploratory analysis
├── model/ # Source code for modeling and processing
├── figures/ # Graphs and plots for the thesis
├── thesis.pdf # Full PDF of the final thesis (French)
└── README.md
Understanding how fuel prices respond to oil price shocks is essential for:
- Public policy (e.g., windfall taxes, price controls)
- Consumer welfare
- Debates around market power and transparency
If you use this work in your research or publication, please cite it as:
Pefourque, G. (2024). Existe-t-il des asymétries dans les ajustements des prix à la pompe ? Université Paris-Est Créteil.
Thanks to Ferhat Mihoubi for supervising this research.