A0819
Title: Dimension reduction in VAR-MGARCH models
Authors: S Yaser Samadi - Southern Illinois University Carbondale (United States) [presenting]
Toktam Valizadeh - Southern Illinois University Carbondale (United States)
Abstract: Economic and financial time series often exhibit heteroscedasticity, posing challenges for analysis. While multivariate generalized autoregressive conditional heteroskedasticity (MGARCH) models offer solutions, they often suffer from overparameterization, escalating parameter counts with dimensionality. A parsimonious approach is introduced to the vector autoregressive (VAR) model, incorporating a heteroscedastic structure within the error terms. The method offers notable efficiency improvements in model estimation by linking the conditional mean and conditional covariance functions of VAR-MGARCH models through minimal reducing subspaces. By parameterizing the method, redundant information within time series signals is eliminated, significantly reducing model complexity. The asymptotic properties of estimators are investigated, and the approach is evaluated through simulation studies and real-data analysis, offering enhanced modeling accuracy and computational efficiency for volatile financial time series.
