Energy Trading – Mean-Reversion Strategy on Heating Oil & Gasoil Futures

This repository presents a comprehensive analysis of mean-reverting statistical arbitrage strategies applied to a futures pair:
Heating Oil (HO) and Low-Sulfur Gasoil (LGO), using high-frequency intraday data.

The study includes:

• Data loading, filtering, and preprocessing
• Ornstein-Uhlenbeck (OU) process calibration using bootstrap methods
• Optimal trading band selection under transaction costs and stop-loss constraints
• In-sample (IS) and out-of-sample (OS) performance assessment

Two parallel implementations are provided:

• Energy-Trading-Python/: Complete Python pipeline
• Energy-Trading-MATLAB/: MATLAB modules and scripts

---

📁 Repository Structure

Root Files

• README.md: This file
• HO-LGO.xlsm: Raw futures data in Excel with macros
• requirements.txt: Python dependencies
• Final_Project_Group_8B_Report.pdf: Final report (PDF-style)
• .gitignore: Git tracking exclusions
• venv/: Python virtual environment (not tracked)

---

🔧 Python Directory — Energy-Trading-Python/

Python implementation of the project with utilities and crypto extensions:

File/Folder	Description
Final_Project_Group_*.ipynb	📓 Jupyter notebooks with full pipeline
Final_Project_Crypto_*.ipynb	Crypto adaptation of the strategy
crypto_dataset/	Additional datasets for crypto testing
utilities/	Modular utility functions (data, plots, stats, etc.)
path/	File and directory path utilities
PDF/	Generated PDF outputs

---

🧮 MATLAB Directory — Energy-Trading-MATLAB/

Modular MATLAB codebase structured into functional blocks:

Folder	Description
Cost & Analysis Utilities/	Transaction cost handling and performance metrics
Data/	Data files (converted/cleaned)
Data Processing/	Trimming and filtering scripts
OU model Estimation .../	OU model calibration and bootstrap
Trading Strategy & Bands/	Trading rules and band computation
Prints and Plots/	Visual results
Final_Project_Group_*.mlx	Main live script notebook (MATLAB)

---

⚙️ How to Run (Python)

1. Install dependencies (preferably in a virtual environment):

pip install -r requirements.txt

2. Run the main Jupyter Notebooks:

jupyter notebook Energy-Trading-Python/Final_Project_Group_*.ipynb

---

📚 Academic Context

This project was developed as part of a university coursework on quantitative trading and statistical arbitrage.
The methodology is based on the paper by Baviera (2019), available in the PDF/ folder.

---

✅ Evaluation

This project received the highest grade in the course, validating both the correctness of the implementation and the depth of the analysis.
All results and methods were thoroughly reviewed by the academic supervisors.

---

📄 License

This project is licensed under the MIT License — see the LICENSE file for details.

---

🔒 Note on Repository Privacy

Please note that the original repository, which contains the complete commit history and development process, is private.
This is due to the presence of academic material and internal coursework content that cannot be publicly shared.

This public version has been curated for portfolio and educational purposes, and faithfully reflects the structure, results, and methodology of the original work.