Portfolio Risk Engine ⚙️

A Python-based portfolio analysis and optimization tool that uses Monte Carlo simulation to construct and evaluates a user-defined number of portfolios using historical market data (yfinance API), selecting the allocation that maximizes the Sharpe ratio (measures unit of return per unit of risk).

---

Overview

The Portfolio Risk Engine:

• Downloads real historical price data
• Converts prices into daily returns
• Simulates thousands of random portfolios (in this project - 5000)
• Evaluates risk and return metrics
• Selects the portfolio with the highest risk-adjusted performance (Sharpe Ratio)
• Visualizes results through equity curves and risk–return plots

The optimization is based on the Monte Carlo simulation method, meaning results vary slightly between runs as different random portfolios are sampled.

---

Features

• Data ingestion using real market prices
• Daily return calculation from historical prices
• Monte Carlo portfolio simulation (long-only, fully invested)
• Annualized Metrics
  • Expected return
  • Volatility
  • Sharpe ratio
• Risk Analysis
  • Maximum drawdown (MDD)
  • Value at Risk (VaR)
  • Conditional Value at Risk (CVaR)
• Visualization
  • Optimal portfolio equity curve
  • Risk–return scatter of all simulated portfolios

---

Project Structure

portfolio-risk-engine/
├── src/
│   ├── main.py            # Runs Monte Carlo simulation and analysis
│   └── risk_engine.py     # Core financial and risk calculations
│
├── data/
│   └── prices.csv         # Historical price data
│
├── make_prices_csv.py     # Downloads and prepares price data
├── requirements.txt
├── LICENSE
└── README.md

How It Works

1. Extracting Price Data

   • Historical daily prices are downloaded for a set of equities.
   • Prices are converted into simple daily returns.

2. Monte Carlo Simulation

   • Thousands of random weight vectors are generated.
   • Each portfolio is fully invested and long-only.
   • For each portfolio, return and risk metrics are computed.

3. Optimization

   • The portfolio with the maximum Sharpe ratio is selected.
   • This represents the best risk-adjusted allocation found in the sampled space.

4. Evaluation

   • The optimal portfolio’s equity curve is constructed.
   • Drawdown, VaR, and CVaR are calculated to assess downside risk. (how much the portfolio can lose during unfavorable market periods)

Libraries Used

• NumPy
• pandas
• yfinance
• Matplotlib