S&P 500 Portfolio Optimization

Overview

This project performs portfolio optimization using historical data of S&P 500 companies. The goal is to create an optimized portfolio that maximizes the Sharpe ratio while minimizing volatility. The optimization process involves:

1. Fetching historical price data for S&P 500 companies.
2. Calculating sector-wise average closing prices.
3. Generating random portfolios and calculating their metrics.
4. Running Monte Carlo simulations to evaluate potential portfolios.
5. Applying Random Forest Regression for further portfolio optimization.

Prerequisites

Ensure you have the following libraries installed:

• openpyxl
• pandas
• yfinance
• matplotlib
• scipy
• scikit-learn
• PyPortfolioOpt
• cvxpy

You can install these libraries using pip:

bash pip install openpyxl pandas yfinance matplotlib scipy scikit-learn PyPortfolioOpt cvxpy

Instructions

1. Setup and Data Fetching:

   • Fetch the list of S&P 500 companies and their sectors from Wikipedia.
   • Replace ticker symbols for consistency and handle missing data.
   • Fetch historical adjusted closing prices for each sector.

2. Data Processing:

   • Calculate the average closing prices for each sector.
   • Compute the logarithmic returns and remove negative returns.
   • Generate random portfolio weights and normalize them.

3. Optimization Metrics:

   • Calculate expected returns, volatility, and Sharpe ratio for the portfolios.
   • Save the results to Excel files for further analysis.

4. Monte Carlo Simulation:

   • Run simulations to generate various portfolios and their metrics.
   • Analyze the results to identify the portfolio with the maximum Sharpe ratio and minimum volatility.

5. Machine Learning Enhancement:

   • Use Random Forest Regression to predict portfolio weights based on portfolio metrics.
   • Evaluate the model performance using Mean Squared Error.

The code is organized into sections as follows:

1. Library Installation - Install necessary libraries.
2. Data Fetching - Retrieve and preprocess data from the web.
3. Data Processing - Calculate averages and log returns.
4. Optimization Metrics - Compute portfolio metrics.
5. Monte Carlo Simulation - Simulate and analyze portfolios.
6. Machine Learning - Train and evaluate a Random Forest model for portfolio weight prediction.

Results

The final output includes:

• Portfolio weights with the highest Sharpe ratio.
• Portfolio weights with the lowest volatility.
• Predicted weights from the Random Forest model.