Exchange Simulator for the Limit Order-Driven Market

This project is a modular and extensible exchange simulation platform designed for studying and evaluating trading strategies within a realistic limit order book (LOB) market environment. It bridges the gap between academic research and practical trading applications.

🔍 Project Overview

• Simulates a realistic limit order-driven stock exchange.
• Enables the integration of both manual and autonomous trading agents.
• Compatible with a wide variety of algorithmic and machine learning-based strategies.
• Includes manipulators to study market vulnerabilities.
• Provides tooling for analyzing simulations and visual comparison of trading strategy performance.

🧠 Research Objectives

• Analyze and simulate market microstructure dynamics.
• Evaluate the performance and stability of trading algorithms.
• Serve as a bridge between theoretical research and real-world application.
• Study market dynamics and the impact of strategic behaviors on liquidity, stability, and fairness.
• Use the platform for educational, regulatory, and experimental purposes.

💻 Key Features

• Modular simulation framework with realistic LOB mechanics.
• Autonomous agent support including ML-powered strategies.
• Fraudulent agent simulation for stress-testing market resilience.
• Web interface powered by Tornado + Bokeh for real-time monitoring.
• Analysis tools using Jupyter Notebooks for report generation and visualization.

🧪 Technologies Used

• Python 3.9 +
• Tornado (web server)
• Bokeh (web-based visualization)
• NumPy, Pandas (data handling)
• Jupyter (result analysis and reporting)

🚀 Getting Started

📦 Clone the Repository

git clone https://github.com/Jivl00/Exchange_simulator_for_the_limit_order-driven_market
cd Exchange_simulator_for_the_limit_order-driven_market

📥 Install Dependencies

pip install -r requirements.txt

🔧 Configuration

Edit config/server_config.json to adjust: server IP, ports, and API paths.

🟢 Start the Server

cd src
python server/server.py

Once the server is running, start the market simulation components:

python server/agents/market_maker.py
python server/agents/liquidity_generator.py

To resume from a previous state:

python server/server.py -l

• Simulations are saved to data/ for post-processing after server shutdown.

👥 Run Market Agents

Agents are defined in client/agents. Start one like this (from the src directory):

python client/agents/your_agent.py

🌐 Launch Web Interface

python viz/main_page.py

Access the GUI at http://<IP_ADDRESS>:<VIZ_PORT> as configured.

📊 Analyze Results

Open the reporting notebook:

viz/report/report.ipynb

Use it to:

• Compare strategy performance
• Analyze trade volume/frequency
• Summarize outcomes statistically

✅ Run Tests

cd tests
python -m unittest tests.py

📝 Documentation

• For more information on the web interface and algorithmic trading, refer to the trading maual.
• Official thesis document detailing the design, implementation, and evaluation of the exchange simulator: Thesis.
• Official thesis poster: Poster.
• Documentation of the visualization module: Visualization Documentation.