Portfolio Analyzer is a modular toolkit for advanced portfolio construction, optimization, and risk analytics. It leverages state-of-the-art quantitative finance techniques, including Black-Litterman blending, robust covariance estimation, and Monte Carlo simulation, all accessible through an interactive Jupyter notebook workflow.
- Automated Data Pipeline: Fetches and processes historical prices and market capitalizations for user-specified assets.
- Robust Statistical Estimation: Implements EWMA and Ledoit-Wolf shrinkage for stable mean and covariance estimates.
- Black-Litterman Integration: Blends equilibrium returns with forward-looking views, including DCF-based signals.
- Flexible Portfolio Optimization: Maximizes risk-adjusted return (Sharpe Ratio) with support for constraints and L2 regularization.
- Comprehensive Risk Analysis: Projects future portfolio outcomes using Monte Carlo simulation with configurable distributional assumptions.
- Interactive Workflow: Step-by-step analysis, optimization, and backtesting in notebooks/portfolio_analysis.ipynb.
- Advanced Visualization: Correlation heatmaps, network graphs, efficient frontier, and detailed backtest reporting.
- Historical Backtesting: Evaluates strategy performance with rolling rebalancing and benchmark comparison.
Portfolio Analyzer uses Poetry for dependency management.
-
Clone the repository:
git clone https://github.com/Vorckea/portfolio-analyzer.git cd portfolio-analyzer -
Install dependencies:
poetry install
The primary workflow is provided in notebooks/portfolio_analysis.ipynb:
-
Activate the virtual environment:
poetry shell
-
Launch Jupyter Lab:
jupyter lab
-
Open and execute the cells in notebooks/portfolio_analysis.ipynb to perform configuration, data preparation, optimization, simulation, and backtesting.
All key parameters (tickers, date ranges, model hyperparameters, and user-defined views) are managed centrally in src/portfolio_analyzer/config.py.
- For detailed module documentation, see inline docstrings and comments in the src/portfolio_analyzer directory.
- For troubleshooting or feature requests, please open an issue on GitHub.
This project is intended for research and educational purposes. Please review and validate all results before using in a production or investment context.