Skip to content

README.md

Latest commit

 

History

History
32 lines (20 loc) · 1.11 KB

README.md

File metadata and controls

32 lines (20 loc) · 1.11 KB
Raw

Cointegration Pair Finder

This Python script identifies cointegrated stock pairs among S&P 500 companies — a common statistical arbitrage strategy used in pairs trading.

It automatically downloads S&P 500 tickers from Wikipedia, fetches their historical price data from Yahoo Finance, and performs the Engle-Granger cointegration test on all possible stock pairs.

🚀 Features

  • Scrapes S&P 500 tickers directly from Wikipedia
  • Downloads historical price data using yfinance
  • Computes cointegration tests for all stock combinations
  • Outputs the top 10 pairs with the lowest p-values (most likely to be cointegrated)

🧠 How It Works

  1. Fetch S&P 500 tickers from Wikipedia.
  2. Download adjusted closing prices for a chosen subset of tickers.
  3. Run cointegration tests (statsmodels.tsa.stattools.coint) for every pair.
  4. Sort by p-value to identify the strongest candidates.

📦 Requirements

Install dependencies using:

pip install requests pandas yfinance statsmodels lxml