Volatility Forecasting Analysis

This project explores volatility forecasting techniques applied to a selection of exchange-traded funds (ETFs). We implement and compare various models, including basic forecasting methods and GARCH models, to predict future volatility based on historical price data.

Overview

The analysis covers ten major sector ETFs: SPY, XLB, XLE, XLF, XLI, XLK, XLP, XLU, XLV, and XLY. We examine daily returns and develop forecasting models to estimate future variance and volatility.

Data

The data/ directory contains:

• Individual ticker data for each ETF in CSV format
• Basic forecast results for each ETF
• GARCH model variance forecasts

All data is sourced from historical price information and processed for analysis.

Notebooks

The notebooks/ folder includes Jupyter notebooks detailing the analysis:

• 01_data_prep.ipynb: Data preparation and cleaning
• 02_estimators.ipynb: Implementation of various estimators
• 03_garch.ipynb: GARCH model development and forecasting
• report.ipynb: Comprehensive report generation

Report

For a detailed walkthrough of the methodology, results, and conclusions, view the interactive HTML report by visiting this link: https://rahulb0802.github.io/Volatility-Forecasting/

Requirements

To run the notebooks and reproduce the analysis:

• Python 3.8+
• Jupyter Notebook or JupyterLab
• Required packages: pandas, numpy, matplotlib, statsmodels, arch (for GARCH models)

Install dependencies with: pip install pandas numpy matplotlib statsmodels arch

Usage

1. Clone or download the repository
2. Navigate to the project directory
3. Open the notebooks in Jupyter to explore the code
4. View index.html for the complete report