📊 Stock Portfolio Risk Analysis

A probabilistic risk analytics dashboard for Indian equity portfolios.

Analyze historical risk, portfolio correlation and future uncertainty using Monte Carlo simulation + Geometric Brownian Motion (GBM).

✨ What this project does

• 📈 Pulls market prices from Yahoo Finance for NSE stocks
• 🧮 Calculates VaR, CVaR, volatility and loss probability
• 🎲 Runs configurable Monte Carlo portfolio simulations
• 🧩 Visualizes return correlations between holdings
• 🔮 Generates probabilistic future price forecasts with 95% uncertainty bands
• 📥 Exports risk results as CSV
• 🎨 Provides a polished Streamlit analytics dashboard

🛠️ Tech Stack

Layer	Technology
Language	Python
Dashboard	Streamlit
Data	yfinance, pandas, NumPy
Statistics	SciPy
Simulation	Monte Carlo, GBM
Visualization	Plotly
ML Utilities	scikit-learn, joblib

📁 Project Structure

Stock-Portfolio-Risk-Probability/
├── app.py                 # Streamlit dashboard
├── model_utils.py         # Data, risk and simulation utilities
├── train_model.py         # Reproducible training/artifact pipeline
├── requirements.txt       # Python dependencies
├── .gitignore             # Repository exclusions
├── data/                  # Generated market data / summaries
└── models/                # Generated model artifacts

data/ and models/ artifacts are intentionally ignored by Git because they are generated locally or during deployment.

🚀 Run locally

git clone https://github.com/KAVYA-29-ai/Stock-Portfolio-Risk-Probability.git
cd Stock-Portfolio-Risk-Probability
pip install -r requirements.txt
streamlit run app.py

The dashboard fetches market data directly, so a pre-generated model file is not required just to explore the application.

Optional: generate artifacts

python train_model.py
streamlit run app.py

📊 Risk Metrics

Value at Risk (VaR) estimates the loss threshold for a selected confidence level.

Conditional Value at Risk (CVaR) measures the average loss in the worst tail beyond VaR.

Monte Carlo simulation generates thousands of correlated portfolio outcomes from historical return and covariance estimates.

GBM forecasting produces probabilistic price paths rather than presenting a single deterministic prediction.

⚠️ Important Note

This project is an educational quantitative-finance application. Its simulations are based on historical statistical assumptions and are not financial advice or guaranteed predictions of market performance.

👤 Author

KAVYA-29-ai

Built as a hands-on project exploring Python, quantitative risk analysis, financial modeling and data visualization.

📄 License

MIT License