financial-backtesting 0.2.2

A backtesting framework for machine learning and financial trading strategies.

Financial Backtesting

A Python library for backtesting financial trading strategies using machine learning models and custom signals. This library provides a robust framework for simulating trades, analyzing performance, and visualizing results.

---

Features

• Dynamic Position Sizing: Supports risk-based position sizing for better capital management.
• Customizable Parameters: Configure stop-loss, take-profit, leverage, transaction fees, and more.
• Comprehensive Metrics: Provides detailed performance metrics such as win rate, profit factor, drawdown, and total return.
• Visualization: Generates plots for equity curves, drawdowns, and trade signals.
• Trade Logging: Tracks and summarizes closed positions for detailed analysis.

---

Installation

Install the package using pip:

pip install financial-backtesting

---

Usage

1. Import the Library

from backtesting import Backtest

2. Prepare Your Data

Ensure your data is a pandas.DataFrame with at least the following columns:

• Price Column: The column containing price data (e.g., close).
• Date Column: The column containing date/time data (e.g., date).
• Signals: A list or array of buy/sell signals.

3. Run a Backtest

# Example DataFrame
import pandas as pd
df = pd.read_csv("your_data.csv")

# Example signals (0 = Buy, 2 = Sell)
signals = [0, 2, 0, 2, ...]

# Initialize the backtest
backtest = Backtest(
    data=df,
    signals=signals,
    price_column="close",
    date_column="date",
    BUY_SIGNAL=0,
    SELL_SIGNAL=2
)

# Run the backtest
results = backtest.run(
    stop_loss=0.01,  # 1% stop loss
    take_profit=0.03,  # 3% take profit
    initial_capital=10000,
    leverage=2,
    quantity=1,
    transaction_fee_rate=0.001,
    dynamic_position_size=True,
    risk_per_trade=0.02,
    max_slippage=0.001,
    plot=True
)

# Print results
print(results)

---

API Reference

Backtest Class

Initialization

Backtest(data, signals, price_column="close", date_column="date", BUY_SIGNAL=0, SELL_SIGNAL=2)

• data: pd.DataFrame - The input data containing price and date columns.
• signals: List - A list of buy/sell signals.
• price_column: str - The column name for price data.
• date_column: str - The column name for date/time data.
• BUY_SIGNAL: int - The value representing a buy signal.
• SELL_SIGNAL: int - The value representing a sell signal.

Methods

1. run() Runs the backtest with the specified parameters.

   run(
       stop_loss: float,
       take_profit: float,
       initial_capital: float = 10000,
       leverage: float = 1,
       quantity: int = 1,
       transaction_fee_rate: float = 0.002,
       dynamic_position_size: bool = False,
       risk_per_trade: float = 0.01,
       max_slippage: float = 0.00,
       start_date: str = None,
       end_date: str = None,
       plot: bool = True
   )
   

   • stop_loss: Stop-loss percentage (e.g., 0.01 for 1%).
   • take_profit: Take-profit percentage (e.g., 0.03 for 3%).
   • initial_capital: Starting capital for the backtest.
   • leverage: Leverage multiplier.
   • quantity: Fixed quantity per trade (ignored if dynamic_position_size=True).
   • transaction_fee_rate: Transaction fee rate (e.g., 0.001 for 0.1%).
   • dynamic_position_size: Whether to use risk-based position sizing.
   • risk_per_trade: Risk percentage per trade (used if dynamic_position_size=True).
   • max_slippage: Maximum slippage percentage.
   • start_date: Start date for filtering data.
   • end_date: End date for filtering data.
   • plot: Whether to generate plots.

2. get_positions() Returns the current open positions.

3. get_closed_positions() Returns the closed positions.

4. print_closed_positions(n: int = None) Prints a summary of closed positions.

---

Outputs

Backtest Results

The run() method returns a dictionary with the following metrics:

• Total Trades: Total number of trades executed.
• Win Rate (%): Percentage of profitable trades.
• Profit Factor: Ratio of gross profit to gross loss.
• Max Drawdown (%): Maximum drawdown during the backtest.
• Profit/Loss: Total profit or loss in dollars.
• Initial Capital: Starting capital.
• Final Capital: Ending capital.
• Return (%): Total return percentage.

Plots

• Price Action with Signals: Shows buy/sell signals on the price chart.
• Equity Curve: Portfolio value over time.
• Available Capital: Remaining capital over time.
• Drawdown: Drawdown percentage over time.
• Trade Performance Distribution: Pie chart of profitable vs. losing trades.

---

Example Output

Backtest Results

+-------------------+----------------+
| Metric            | Value          |
+-------------------+----------------+
| Total Trades      | 50             |
| Win Rate (%)      | 60.00          |
| Profit Factor     | 1.50           |
| Max Drawdown (%)  | -10.00         |
| Profit/Loss       | $1,500.00      |
| Initial Capital   | $10,000.00     |
| Final Capital     | $11,500.00     |
| Return (%)        | 15.00%         |
+-------------------+----------------+

Plots

• Price Action with Signals: Shows buy/sell signals on the price chart.
• Equity Curve: Portfolio value over time.
• Drawdown: Drawdown percentage over time.

---

Contributing

Contributions are welcome! If you'd like to contribute, please fork the repository and submit a pull request.

---

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Contact

For questions or feedback, contact Shashwat Chaturvedi at [e-mail].