markowitz-portfolio-optimization with CVXPY

What This Notebook Does

Implements the classic Markowitz Mean-Variance portfolio optimization model using CVXPY — the open-source Python library for convex optimization.

Covers

• Real stock data (AAPL, MSFT, GOOGL, JPM, GS, JNJ, PFE, XOM, CVX) via yfinance
• Minimum Variance Portfolio
• Maximum Sharpe Ratio Portfolio
• Full Efficient Frontier with Capital Market Line
• Real-world constraints (long-only, weight caps, sector caps)
• Portfolio comparison summary table

Why CVXPY?

Portfolio optimization is a Quadratic Program (QP). CVXPY lets you express it exactly as the math states:

minimize    w^T Σ w
subject to  1^T w = 1
            w >= 0

The same workflow is used at quantitative hedge funds and trading desks globally.

Tech Stack

• Python 3.10+
• CVXPY 1.8+
• yfinance, NumPy, pandas, matplotlib, seaborn

Run It

pip install cvxpy yfinance numpy pandas matplotlib seaborn
jupyter notebook markowitz_portfolio_optimization.ipynb

Author

Dhruv | Quantitative Finance + Open Source
GSoC 2026 Applicant — CVXPY Sub-Org (NumFOCUS)
GitHub: github.com/dhru189