Pairs Trading Strategy: PHM & TOL

A statistical arbitrage trading strategy that identifies and trades correlated stock pairs. This project executes real-time pairs trading using the Alpaca API, focusing on PulteGroup (PHM) and Toll Brothers (TOL). The strategy profits from temporary mispricings between these two homebuilder stocks.

---

Features

• Cointegration & Correlation Testing: Identifies tradable stock pairs.
• Z-Score Based Trade Signals: Detects market inefficiencies.
• Backtesting Engine: Simulates historical performance.
• Live Trading via Alpaca API: Executes trades in real time.
• Risk Management: Implements position sizing and exit strategies.

---

Investment Thesis: PHM & TOL

PulteGroup (PHM) and Toll Brothers (TOL) are leading homebuilders in the U.S. housing market. Their stock prices are highly correlated and cointegrated, which makes them ideal for a pairs trade.

1. Macroeconomic Sensitivity: Both stocks react similarly to changes in mortgage rates, housing demand, and supply chain dynamics.
2. Market Segmentation: PHM targets the mass market while TOL focuses on the luxury segment. This difference can lead to temporary valuation divergences.
3. Earnings & Industry Trends: Their prices often move together following sector-specific news and earnings reports.

The strategy exploits temporary deviations in the spread (the difference between PHM and TOL prices), betting on a reversion to the mean.

---

How the Strategy Works

1. Identify a Stock Pair: Select PHM and TOL based on high correlation and cointegration.
2. Calculate the Spread & Z-Score: Standardize the difference in their prices to detect overbought or oversold conditions.
3. Generate Trading Signals:
   • Long PHM / Short TOL: When the Z-score is below -2.
   • Short PHM / Long TOL: When the Z-score is above 2.
   • Exit: When the Z-score returns between -0.5 and 0.5.
4. Execute Trades via Alpaca API: Automatically place orders based on the signals.
5. Backtest Performance: Use historical data to simulate strategy performance.

---

Project Structure

pairs-trading/
├── data/                  # Stored data files
│   ├── stock_data.csv
│   ├── cointegrated_pairs.csv
│   └── trading_signals.csv
├── docs/                  # Documentation files
│   ├── strategy.md
│   ├── backtesting.md
│   └── trading_signals.md
│   └── returns.md
├── src/                   # Core source code
│   ├── config.py          # API keys & strategy settings
│   ├── data_loader.py     # Fetches stock data
│   ├── pairs_selection.py # Identifies tradable stock pairs
│   ├── trading_strategy.py# Generates trade signals
│   ├── backtest.py        # Backtests strategy performance
│   └── live_trading.py    # Executes live trades
├── .gitignore
├── requirements.txt
└── README.md

---

Installation

Prerequisites

• Python 3.7 or later
• An Alpaca API account (for paper or live trading)
• Required Python packages (listed in requirements.txt)

Steps

1. Clone the Repository:

   git clone https://github.com/ronitbatra/pairs-trading
   cd pairs-trading

2. Install Dependencies:

   pip install -r requirements.txt

3. Configure API Keys: Update src/config.py with your Alpaca API credentials:

   ALPACA_API_KEY = "your_api_key"
   ALPACA_SECRET_KEY = "your_secret_key"
   ALPACA_BASE_URL = "https://paper-api.alpaca.markets"

4. Run the Trading Pipeline:
   • Backtesting:

     python src/backtest.py

   • Live Trading:

     python src/live_trading.py

---

Definitions of Key Terms

• Pairs Trading:
  A market-neutral strategy that profits from the relative price movements of two correlated stocks.

• Cointegration:
  A statistical property where two time series move together over the long term, allowing us to bet on their mean reversion.

• Correlation:
  A measure of how two stocks move relative to each other. High correlation (e.g., 0.8 or above) indicates similar price movements.

• Z-Score:
  A standardized metric indicating how many standard deviations the current spread is from its mean. It is used to signal trades.

• Spread:
  The difference between the prices of two stocks, calculated as:
  Spread = PHM - TOL
  A mean-reverting spread is key for pairs trading.

• Long/Short Positions:

  • Long PHM / Short TOL: Buy PHM and sell TOL when the spread is abnormally low.
  • Short PHM / Long TOL: Sell PHM and buy TOL when the spread is abnormally high.

• Alpaca API:
  An API that enables live trading by executing orders, fetching real-time market data, and managing positions.

---

License

This project is licensed under the MIT License. Feel free to modify and distribute.

---

Author

• Ronit Batra – GitHub Profile
• Email: xfb7hj@virginia.edu