📈 Algorithmic Trading

Overview

This is a very simple algorithmic trading environment implemented in Python.

It utilizes historical crypto market data to optimize portfolio performance given a policy and a forecasting model.

The environment supports multiple trading strategy and risk management parameters.

🚀 Features

• Backtesting: Test your strategy on historical price data.
• Grid Search: Find your best parameters by parallelized search.
• Visualization Tools: View position opening, closing and returns.
• Strategy Customization: Implement and test different trading strategies.
• Risk Management: Stop-loss, take-profit, and exposure control.

🛠️ Installation

Requirements

Ensure you have Python installed (>= 3.8). Then, install dependencies:

pip install -r requirements.txt

📂 Project Structure

algo-trading/
│-- app/                        # Main trading application
│   │-- backtesting/            # Backtesting on historical data
│   │-- grid_search.py          # Grid search multiple parameters
│   │-- analyse_grid_search.py  # Analyse grid search (plot, stats)
│-- data/                       # Historical market data
│-- results/                    # Simulations results
│-- src/                        # Historical market data
│   │-- utils/                  # Utilities
│-- tests/                      # Unit tests
│-- config.py                   # Configuration settings
│-- main.py                     # Entry point for the system
│-- README.md                   # Documentation

📌 Usage

Running the System

The system is controlled through main.py, which centralizes all execution modes.

Backtesting a Strategy

Run a backtest with a given strategy:

python main.py --mode backtest --strategy moving_average \
    --fee 0.001 --init_cash 10000 --init_crypto 0.1 \
    --min_expected_returns 0.02 --stop_loss 0.05 --predict_len 10 \
    --verbose --plot_results

Grid Search for Parameter Optimization

Execute a parameter grid search:

python main.py --mode grid_search

Analyzing Grid Search Results

Run an analysis on the grid search results:

python main.py --mode analyse

🔗 Supported Data Sources

• Kaggle
• CryptoArchive
• Kraken

As long as you can format a CSV with column 'unix', 'open', 'high', 'low', 'close', 'volume', it'll work.

📈 Results & Performance Analysis

The system includes built-in visualization tools to plot:

• Cumulative returns
• Position opening / closing
• Grid search analysis

🔧 Configuration

Modify config.py to adjust parameters:

config = {
    "data": {
        "pickle_file": "/path/to/pickle",
        "start_position": -1 - 5000,        # Last n time steps
        "end_position": -1,                 # End of data
    },
}

🔮 Future Enhancements

• Advanced Machine Learning Strategies 🤖
• Multi-Asset Portfolio Optimization 📊
• Mke the fee time variable
• Separate the fixed fee in a maker and taker fee for selling and buying
• Implement take-profit and exposure control
• Switch from Pickle to all CSV

📜 License

This project is licensed under the MIT License.

🤝 Contributing

Contributions are welcome. Please submit a pull request or open an issue for discussion.

📧 Contact

For questions, reach out via email or open an issue in the repository.