This repository is the culmination of my Semester 6 Project where I dove deep into the world of financial portfolio optimization. Coming from a Computer Science with AIML background, I wanted to challenge myself by applying mathematical modeling, Python, and data science to a real-world finance problem. This notebook is more than just code – it's a journey of learning, trial & error, and finally understanding the inner workings of modern portfolio theory.

Objective:
To build a system that optimizes a portfolio of stocks using Python such that the returns are maximized for a given level of risk, or risk is minimized for a given return — using concepts from Modern Portfolio Theory (MPT) by Harry Markowitz.

Initially, I was fascinated by how investors like Warren Buffet build portfolios. That curiosity led me to concepts like expected returns, volatility, Sharpe ratio, and efficient frontiers. I realized this would be a perfect intersection of Python, Data Science, and a practical problem.

1. Data Collection

   • Used yfinance to fetch historical stock prices.
   • Selected a basket of well-known stocks (like AAPL, MSFT, GOOG, etc.) for diversification.

2. Calculating Returns & Covariance

   • Computed daily returns.
   • Derived mean returns and covariance matrix of the portfolio.

3. Portfolio Simulation

   • Generated thousands of random portfolios with random weight allocations.
   • For each, computed expected return, volatility, and the Sharpe ratio.

4. Optimization

   • Used scipy.optimize to maximize the Sharpe ratio and minimize volatility.
   • Applied constraints like sum of weights = 1 and individual weights between 0 and 1.

5. Visualization

   • Plotted all simulated portfolios on a risk-return scatter plot.
   • Highlighted the efficient frontier, maximum Sharpe ratio portfolio, and minimum volatility portfolio.

• Real-world financial data handling with Python.
• Portfolio theory and its mathematical basis.
• Using scipy.optimize for constrained optimization.
• Visualizing data using matplotlib and seaborn.
• The balance between risk and reward in investments.
• Clean and structured Jupyter Notebook documentation.

• Python 3.10+
• pandas
• numpy
• matplotlib
• seaborn
• yfinance
• scipy

• Clone this repository

  git clone https://github.com/yourusername/PortfolioOptimization.git
  cd PortfolioOptimization
  

   

2. Run the notebook

Open PortfolioOptimizationFinal.ipynb in Jupyter Notebook or VS Code.

Add real-time stock data integration.

Support for cryptocurrencies and other asset classes.

Use machine learning to predict returns.

Add a simple Streamlit-based UI for non-technical users.

Concepts inspired by Modern Portfolio Theory by Harry Markowitz.

Tutorials and blogs by QuantInsti, Investopedia, and various YouTube channels on Quant Finance.

This project taught me how theoretical finance meets practical Python programming. It also gave me the confidence to work on multidisciplinary problems by blending coding, mathematics, and business logic.

I'm excited to expand on this in the future — maybe even build a full-fledged portfolio management tool someday!

Feel free to ⭐ this repo or suggest improvements via issues or pull requests.

Thanks for reading! — Shivam Chaubey