VS-LTGARCHX Research Project

Project Overview

The VS-LTGARCHX (Variable Selection Long-Term Generalized Autoregressive Conditional Heteroskedasticity with Exogenous variables) project represents a significant advancement in time series econometric modeling.

Objectives

• Develop enhanced GARCH modeling framework with improved variable selection
• Incorporate long-term memory components for better volatility forecasting
• Create robust implementation for practical financial applications

Methodology

The research combines:

• Advanced statistical theory for time series analysis
• Machine learning techniques for variable selection
• Extensive empirical validation using financial market data

Key Results

• Improved forecasting accuracy compared to traditional GARCH models
• Better handling of long-term volatility patterns
• Robust performance across different market conditions

Implementation

The methodology has been implemented in both R and Python, with comprehensive documentation and examples available for practitioners.

GitHub Repository: Log-TGARCHX-Subset-Selection

Publications

This research has resulted in peer-reviewed publications and conference presentations, contributing to the broader econometrics literature.

Preprint: HAL Science Archive

Published Paper: Journal of Time Series Econometrics - De Gruyter