Statarb

Statistical Arbitrage Trading System

A production-grade statistical arbitrage (stat-arb) trading system that identifies market mispricings through quantitative factor analysis, portfolio optimization, and systematic execution. The system processes historical market data, generates alpha signals from multiple strategies, optimizes portfolio positions considering transaction costs and risk, and backtests trading strategies through multiple simulation engines.

Python 3 Migration Complete ✅

**Migration Status**: COMPLETE (v2.0.0-python3) **Python Version**: 3.8+ (recommended 3.9-3.12) **Migration Date**: February 2026 **Test Coverage**: 99% pass rate (101/102 tests)

This codebase has been successfully migrated from Python 2.7 to Python 3. All core functionality validated and production-ready. See [PYTHON3_MIGRATION_COMPLETE.md](PYTHON3_MIGRATION_COMPLETE.md) for migration details and [RELEASE_NOTES_v2.0.0.md](RELEASE_NOTES_v2.0.0.md) for changes.

**Key Changes:**

• Python 3.8+ required (tested with 3.12.3)
• Modern dependencies: numpy>=1.19, pandas>=1.3, scipy>=1.5
• scipy.optimize replaces OpenOpt (Python 3 compatible)
• 100% backward compatible function signatures
• Zero breaking changes to user-facing APIs

Overview

This system implements a complete workflow for statistical arbitrage trading:

1. Data Loading & Preprocessing: Loads and processes market data from multiple sources
2. Alpha Generation: Calculates predictive signals from 20+ trading strategies
3. Factor Analysis: Decomposes returns using PCA and Barra risk models
4. Portfolio Optimization: Maximizes risk-adjusted returns with realistic constraints
5. Backtesting: Simulates execution across multiple engines with transaction cost modeling

The system is designed for daily rebalancing across ~1,400 US equities with sophisticated risk management and execution cost modeling.

Table of Contents

• Features
• Architecture
• [Installation](#insta