Volatility is a key measure of financial risk, and GARCH models are widely used to describe its dynamics. However, they do not account for the influence of news sentiment, which can significantly shape market volatility. Recently proposed News-Augmented GARCH model addresses it by incorporating sentiment signals in a nonlinear, asymmetric, and multiplicative form. This paper examines its theoretical properties and performs a simulation-based hyperparameter study. The analysis establishes the existence of a unique and causal solution, derives a stability condition, and evaluates model sensitivity and parameter recovery under controlled scenarios. Results demonstrate robust performance across various settings and provide guidance for informed hyperparameter selection and evaluation, enhancing model’s reliability for empirical applications.
This work is licensed under a Creative Commons Attribution 4.0 International License.