VaR Calculator for Indian Stocks

A command-line tool that calculates Value at Risk (VaR) for Indian stocks using three different methodologies by fetcheing stock data from Yahoo Finance API. This tool helps investors quantify potential portfolio losses under normal market conditions.

Features

• Three Calculation Methods:
  • 📊 Parametric (Variance-Covariance)
  • 📈 Historical Simulation
  • 🎲 Monte Carlo Simulation
• Exchange Support: Works with both NSE and BSE stocks
• Visual Reports: Generates comparative charts of results
• Configurable Parameters: Adjust confidence level, time horizon, etc.
• Automatic Data Management: Creates necessary directories at runtime
• Comprehensive Reporting: Saves portfolio data and generates comparison charts

Installation

1. Clone the repository:

git clone https://github.com/iammohith/VaR_Calculator.git
cd VaR_Calculator

2. Install dependencies:

pip install -r requirements.txt

Usage

Basic Command:

python main.py [TICKER] [EXCHANGE] [PORTFOLIO_VALUE] [OPTIONS]

Arguments:

Argument	Description	Example
TICKER	Stock ticker symbol	RELIANCE
EXCHANGE	NSE or BSE	NSE
PORTFOLIO_VALUE	Portfolio value in INR	1000000

Options:

Option	Description	Default
--confidence	Confidence level (0.90-0.99)	0.95
--horizon	Time horizon in days	1
--window	Historical data window (days)	252
--simulations	Monte Carlo simulations count	10000

Example:

python main.py RELIANCE NSE 1000000 --confidence 0.99 --horizon 5

Sample Output:

2025-07-16 10:19:28 - src.cli - INFO - Calculating VaR for RELIANCE on NSE
2025-07-16 10:19:28 - src.cli - INFO - Portfolio value: ₹1,000,000.00
2025-07-16 10:19:28 - src.cli - INFO - Confidence: 99.0%, Horizon: 5 days
2025-07-16 10:19:28 - src.cli - INFO - Data window: 252 days, Simulations: 10000

Value at Risk for RELIANCE (NSE) portfolio of value ₹1,000,000.00:
• Parametric VaR:   ₹67,788.39
• Historical VaR:   ₹74,384.95
• Monte Carlo VaR:  ₹69,182.25

Sample Report:

Project File Structure

VaR_Calculator/
├── config/                  # Configuration files
│   ├── logging.conf         # Logging setup
│   └── settings.py          # Application settings
├── data/                    # Auto-generated data
│   ├── logs/                # Application logs
│   ├── portfolios/          # Portfolio data
│   └── reports/             # Generated charts
├── src/                     # Source code
│   ├── cli.py               # Command-line interface
│   ├── data_fetcher.py      # Stock data retrieval
│   ├── exceptions.py        # Custom errors
│   ├── historical_var.py    # Historical VaR
│   ├── monte_carlo_var.py   # Monte Carlo VaR
│   ├── parametric_var.py    # Parametric VaR
│   ├── report_generator.py  # Report creation
│   └── utils.py             # Helper functions
├── main.py                  # Application entry
├── .gitignore               # Ignore files
├── LICENSE                  # MIT License
├── README.md                # This documentation
└── requirements.txt         # Dependencies

Dependencies

• Python 3.8+
• yfinance - Stock data fetching
• numpy - Mathematical operations
• pandas - Data handling
• scipy - Statistical functions
• matplotlib - Visualization

Key Features

1. Accurate Calculations:
   • Industry-standard VaR formulas
   • Proper handling of time horizon scaling
   • Robust input validation
2. Comprehensive Reporting:
   • Portfolio data saved as CSV
   • Comparative visualization of all three methods
   • Automatic timestamping of outputs
3. Exchange Support:
   • Clear identification of exchange in all outputs
   • Supports both NSE and BSE
4. User-Friendly:
   • Detailed logging
   • Clear console output
   • Easy-to-understand reports

Limitations

• Single-stock portfolios only
• Uses daily closing prices
• Monte Carlo assumes lognormal distribution
• Requires internet connection for data fetching

Contributing

Contributions welcome! Please fork the repository and submit pull requests.

License

MIT License - See LICENSE for details.