Skip to content

PROFESSIONAL_README.md

Latest commit

 

History

History
163 lines (128 loc) · 4.58 KB

PROFESSIONAL_README.md

File metadata and controls

163 lines (128 loc) · 4.58 KB

Trading Bot

CI

A professional-grade Python framework for backtesting intraday trading strategies on US equities with comprehensive performance analysis and market microstructure simulation.

Key Features

Market Data & Infrastructure

  • Multi-format data support: CSV files, REST APIs (extensible)
  • Microstructure simulation: Realistic bid/ask order matching
  • Latency modeling: Configurable order-to-execution delay
  • Transaction costs: Variable fees and slippage simulation

Trading Strategies

  • Mean reversion: Z-score based entry/exit with lookback windows
  • Momentum: Trend-following based on price velocity
  • Market making: Two-sided limit order placement
  • Extensible architecture: Easy to add custom strategies via BaseStrategy

Backtesting Engine

  • Real-time simulation: Tick-by-tick order matching
  • Position management: Average cost accounting with realized/unrealized P&L
  • Risk controls: Maximum position limits per strategy
  • Performance tracking: Comprehensive equity curve and trade analytics

Analytics & Visualization

  • Professional metrics: Sharpe ratio, Calmar ratio, maximum drawdown, win rate
  • Publication-quality plots: Equity curves, drawdown analysis, P&L distributions
  • Parameter optimization: Grid search with performance ranking
  • Trade analysis: Per-fill P&L attribution and execution statistics

Installation

git clone https://github.com/your-username/trading-bot.git
cd trading-bot
pip install -r requirements.txt

Quick Start

Basic Usage

# Run with synthetic data
python main.py --synthetic --strategy mean_reversion

# Use your own data
python main.py --csv data/your_ticks.csv --strategy momentum --lookback 120

Advanced Configuration

python main.py \
  --csv data/AAPL_ticks.csv \
  --strategy mean_reversion \
  --lookback 60 \
  --threshold 0.8 \
  --max-position 200 \
  --latency-ms 100 \
  --fee-per-share 0.001 \
  --slippage-bps 3

Parameter Optimization

from data_loader import make_synthetic_orderbook
from tuning import grid_search_strategy

data = make_synthetic_orderbook(periods=5000)
results, best_metrics, best_bt = grid_search_strategy(
    data, 'mean_reversion',
    {
        'lookback': [30, 60, 120],
        'threshold': [0.5, 1.0, 1.5],
        'max_position': [50, 100, 200]
    }
)
print(f"Best Sharpe: {best_metrics['sharpe_ratio']:.2f}")

Data Format

Your CSV should contain these columns (case-insensitive):

  • timestamp: ISO format datetime (UTC preferred)
  • bid, ask: Best bid/offer prices
  • bid_size, ask_size: Quantities at best prices
  • symbol (optional): Security identifier

Example:

timestamp,bid,ask,bid_size,ask_size,symbol
2024-01-02T14:30:00Z,100.00,100.01,500,600,AAPL
2024-01-02T14:30:01Z,100.01,100.02,450,550,AAPL

Architecture

trading-bot/
├── main.py              # CLI entry point with argparse
├── data_loader.py       # Data I/O and synthetic generation
├── backtester.py        # Core simulation engine
├── strategies/          # Strategy implementations
│   ├── base.py         # Abstract base class
│   ├── mean_reversion.py
│   ├── momentum.py
│   └── market_making.py
├── metrics.py           # Performance calculation
├── plotting.py          # Visualization utilities
├── tuning.py           # Parameter optimization
└── tests/              # Unit tests

Performance Metrics

  • Sharpe Ratio: Risk-adjusted returns (annualized)
  • Calmar Ratio: Return/max drawdown ratio
  • Maximum Drawdown: Peak-to-trough equity decline
  • Volatility: Annualized return standard deviation
  • Win Rate: Percentage of profitable trades
  • Average Trade P&L: Mean per-trade profit/loss

Development

Testing

pytest -v

Code Quality

# Optional: Install dev dependencies
pip install black ruff mypy

# Format code
black .

# Lint
ruff check .

# Type checking
mypy --strict main.py

Contributing

  1. Fork the repository
  2. Create a feature branch
  3. Add tests for new functionality
  4. Ensure CI passes
  5. Submit a pull request

License

MIT License - see LICENSE file for details.

Disclaimer

This software is for educational and research purposes only. It is not intended for live trading or as investment advice. Past performance does not guarantee future results.

Developed for quantitative research and algorithmic trading education.