A Deterministic Limit Order Book Simulator with Hawkes-Driven Order Flow

Battle Count: 90

Highly relevant to quantitative trading. Provides a production-ready simulator for LOB dynamics with realistic order flow clustering. Directly applicable to market making, execution algorithms, and risk management. The nearly-unstable regime finding (branching ratio ~0.8) is practically important for realistic simulation. The C++ engine with Python bindings is immediately usable for backtesting. Time-rescaling diagnostics provide rigorous model validation tools for trading systems.

Moderate-to-high complexity. Requires C++17 knowledge for the LOB engine, Python for analysis, and understanding of point process theory for the Hawkes generator. The architecture is well-modularized (stochastic layer, deterministic layer, analysis layer) with PyBind11 bridging. However, the mathematical foundations (spectral radius conditions, contraction mappings, compensator calculus) require graduate-level probability theory. The simulator itself is production-ready with Streamlit UI, but extending it requires significant effort.

Exceptional reproducibility: full source code (C++/Python), configuration files, processed datasets, deterministic build scripts, fixed random seeds, and all figures/tables are publicly available on GitHub. Includes Streamlit interactive interface. All experiments use publicly accessible data (Binance REST API, LOBSTER). The paper explicitly states 'deterministic scripts to regenerate all figures and tables' and provides seed control throughout.