Skip to content

PLUTUS: Pair-Trading Toolkit

PLUTUS Flyer

Python PyPI - Version PyPI Downloads License: MIT GitHub

PLUTUS is a Python-based toolkit for performing pair-trading analysis. This project is designed for educational purposes and provides analysis tools for:

  • Fetching and processing financial data.
  • Conducting statistical tests (stationarity and cointegration).
  • Performing feature engineering.
  • Visualizing financial time-series data.

🌟 Key Features

➊ 📥 Data Acquisition

  • 📡 Fetch historical financial data using Yahoo Finance API.
  • 🗃️ Store and manage time-series data in a structured format.
  • 🔄 Combine and preprocess data for analysis.

➋ 📊 Statistical Tests

📈 Stationarity Tests

  • 🧪 Augmented Dickey-Fuller Test (ADF) tests whether a time series is stationary.
  • 📉 Phillips-Perron Test (PP) handles autocorrelations and heteroskedasticity.
  • 📊 KPSS test for trend stationarity.

🔗 Cointegration Tests

  • ⚖️ Engle-Granger identifies long-term equilibrium relationships.
  • 🔍 Phillips-Ouliaris handles residual-based cointegration testing.
  • 🔄 Johansen Test detects multiple cointegration vectors.

➌ 🛠️ Feature Engineering

  • 📅 Compute periodic returns (daily, weekly, monthly).
  • 🔢 Apply logarithmic and exponential transformations.
  • 🔗 Calculate correlation matrices and filter securities based on thresholds.
  • 🔬 Identify cointegrated pairs for pair trading.

➍ 📊 Data Visualization

  • 📈 Plot financial time-series data.
  • 📊 Generate dual-axis plots for comparing securities.
  • 🕵️‍♂️ Visualize correlation matrices.
  • 🔄 Plot autocorrelation and partial autocorrelation.

Explore the documentation to learn how to customize and make the most of PLUTUS Pair-Trading Toolkit for your project!