Arbitrage-cpp

A modular C++ implementation of an energy futures statistical arbitrage pipeline.
The project replicates and extends a MATLAB workflow, providing efficient execution with C++ (Boost, NLopt, CMake).

Core features:

• Data loading, trimming, and outlier filtering
• Ornstein–Uhlenbeck (OU) model calibration and bootstrap
• Transaction cost estimation
• Optimal trading bands computation with stop-loss & leverage sweep
• Results exported to CSV for further analysis

---

Project Structure

The repository is organized into modular components:

include/utilities/

Header-only utilities used across the project:

• DataOrdering.hpp – Functions for data trimming and splitting
• Loaders.hpp – CSV loader and preprocessing
• StatisticalBootstrap.hpp – OU model bootstrap estimation
• OptimalBands.hpp – Optimal trading bands computation

---

src/

C++ source code implementations:

• main.cpp – Main pipeline entry point
• DataOrdering.cpp – Implementation of data ordering utilities
• Loaders.cpp – CSV loader implementation
• StatisticalBootstrap.cpp – OU model estimation & bootstrap logic
• OptimalBands.cpp – Optimal bands optimization (NLopt + Boost)

---

Root files

• CMakeLists.txt – Build configuration (CMake)
• HO-LGO.csv – Example dataset (not included in repo)

---

outputs/

Generated results and logs after running the pipeline.

---

Dependencies

This project requires:

• C++17 or later
• CMake ≥ 3.15
• Boost – Math & statistics utilities
• NLopt – Nonlinear optimization library

On macOS (Homebrew):

brew install boost nlopt cmake