Cross-Asset Portfolio Optimization

Overview

This project focuses on building a diversified cross-asset portfolio using historical financial data. The goal is to maximize risk-adjusted returns while minimizing portfolio risk and drawdowns.

It applies modern portfolio theory concepts such as log returns, covariance analysis, and efficient frontier simulation.

---

Features

• Multi-asset portfolio construction (stocks, crypto, commodities, indices)

• Data preprocessing and cleaning

• Log return calculation

• Covariance and correlation analysis

• Monte Carlo simulation (5000 portfolios)

• Efficient Frontier visualization

• Optimal portfolio selection:

  • Maximum Sharpe Ratio
  • Minimum Volatility

• Portfolio backtesting

• Drawdown analysis

• Rolling Sharpe Ratio

• Stress testing (high volatility periods)

• Portfolio performance metrics

---

Project Structure

portfolio-optimization/
│
├── data/                  # CSV files (asset price data)
├── notebook.ipynb        # Main analysis notebook
├── requirements.txt      # Required libraries
└── README.md             # Project documentation

---

Technologies Used

• Python
• Pandas
• NumPy
• Matplotlib
• Seaborn

---

Installation

1. Clone the repository:

git clone https://github.com/your-username/portfolio-optimization.git
cd portfolio-optimization

2. Install dependencies:

pip install -r requirements.txt

---

How to Run

Option 1: Jupyter Notebook

jupyter notebook notebook.ipynb

Option 2: Python Script

python main.py

---

Methodology

1. Data Preprocessing

• Load multiple CSV files
• Convert Date column to datetime
• Set Date as index
• Keep only Price column
• Handle missing values
• Convert prices to float

2. Feature Engineering

• Compute log returns:

returns = np.log(prices / prices.shift(1))

3. Statistical Analysis

• Mean returns
• Covariance matrix
• Correlation heatmap

4. Portfolio Optimization

• Simulate 5000 random portfolios

• Compute:

  • Expected return
  • Volatility
  • Sharpe ratio

5. Optimal Portfolios

• Max Sharpe Portfolio
• Min Volatility Portfolio

6. Efficient Frontier

• Plot risk vs return
• Highlight optimal portfolios

7. Backtesting

• Compute portfolio value over time
• Compare performance

8. Risk Analysis

• Drawdown calculation
• Rolling Sharpe Ratio
• Stress testing during high volatility

---

Key Results

Metric	Max Sharpe	Min Volatility
Annual Return	20.2%	17.7%
Volatility	24.34%	23.57%
Sharpe Ratio	0.83	0.75
Max Drawdown	-44.39%	-40.99%
Sortino Ratio	1.04	-

---

Visualizations

• Asset price trends
• Efficient Frontier
• Portfolio equity curves
• Drawdown plots
• Rolling Sharpe Ratio
• Correlation heatmap

---

Assumptions

• Risk-free rate = 0%
• No transaction costs
• Daily rebalancing
• Missing data dropped (no interpolation)
• Log returns used for calculations

---

Conclusion

• The Max Sharpe Portfolio provides higher returns but comes with higher risk.
• The Min Volatility Portfolio offers more stability with lower returns.
• Diversification plays a crucial role in reducing portfolio risk.
• Correlated assets increase overall portfolio volatility.

---

Future Improvements

• Include transaction costs
• Add dynamic portfolio rebalancing
• Use machine learning for asset allocation
• Implement regime-switching models
• Stress test with extreme market scenarios

---

Author

Dharshini D

---