Backtesting

Backtesting is a lightweight, composable Python toolkit that lets you model portfolios and rigorously evaluate rule-based trading strategies against historical market data.
It is designed for engineers, quantitative researchers, and data-driven investors who need transparent, reproducible backtests without the drag of heavyweight monolithic frameworks.

---

Key Features

Feature	What it gives you	Where to look
Object-oriented portfolio model	Portfolio, Asset, Currency classes with audit-ready transaction ledger	src/backtesting/portfolio.py, asset.py, currency.py, ledger.py
Strategy abstraction	Create a strategy by subclassing TradingStrategy; run multi-period evaluations with MultiPeriodBacktest	trading_strategy.py
Unit-aware analytics	pandas_units integrates pint units with pandas for type-safe calculations	pandas_units.py
Tutorial notebooks	End-to-end examples in Jupyter (see tutorial/)	tutorial/portfolio_introduction.ipynb
Python ≥ 3.11, zero heavy deps	Pure-Python core, only standard libraries plus pandas	requirements.txt

---

Installation

# Recommended: create a fresh virtual environment first
python -m venv .venv && source .venv/bin/activate

# From PyPI (when released)
pip install backtesting

# Or from source
git clone https://github.com/<your-org>/backtesting.git
cd backtesting
pip install -e .

---

Quick Start

import pandas as pd
from backtesting import Portfolio, TradingStrategy, MultiPeriodBacktest

class SMACrossover(TradingStrategy):
short = 20
long  = 50

def generate_signals(self, prices: pd.Series) -> pd.Series:
fast = prices.rolling(self.short).mean()
slow = prices.rolling(self.long).mean()
return (fast > slow).astype(int)  # 1 = long, 0 = flat

# Load your price series (index must be datetime) e.g. from CSV
prices = pd.read_csv("EURUSD_daily.csv", index_col=0, parse_dates=True)["close"]

portfolio = Portfolio(base_currency="EUR", cash=10_000)
strategy  = SMACrossover()
bt        = MultiPeriodBacktest(portfolio, strategy, prices)

tearsheet = bt.run()
tearsheet.plot()          # equity curve & metrics
print(tearsheet.summary)  # table of risk/return stats

See the tutorial notebooks for a deeper walk-through.

---

Project Layout

backtesting/
├── src/
│   └── backtesting/           # Core package
│       ├── portfolio.py
│       ├── trading_strategy.py
│       ├── ...                # other helpers
├── tutorial/                  # Jupyter notebooks
├── tests/                     # (coming soon)
├── requirements.txt
└── pyproject.toml

---

Running Tests

pip install -r requirements.txt
pytest -q

---

Contributing

Pull requests are welcome!
If you find a bug or want to suggest a feature, please open an issue first ensuring it hasn’t been reported.

1. Fork the repo & create your branch: git checkout -b feature/my-awesome-feature
2. Commit your changes with clear messages.
3. Run the test suite (pytest) and keep coverage ≥ 90 %.
4. Submit a PR and describe why and how you changed things.

Please follow Conventional Commits and run pre-commit install after cloning to keep the style consistent.

---

Roadmap

• Vectorised execution engine for intraday data
• Plug-in metric system (Sharpe, Sortino, max drawdown, etc.)
• Slippage & transaction-cost models
• Exchange adapters for live paper-trading
• Interactive Streamlit dashboard

---

Disclaimer

This project is for research and educational purposes only and does not constitute financial advice.
Trading involves substantial risk; use these tools responsibly and at your own risk.

---

License

Backtesting is distributed under the MIT License. see LICENSE.