Answers are grounded in this paper and cite their sources.
This analysis demonstrates improved volatility modeling in heteroskedastic time series, highlighting variance regimes identified by a Markov switching approach.
Variance breaks in time series data result in heteroskedasticity, influencing the distribution and modeling of the series. Volatility models, like as GARCH and ARCH, have exhibited constraints in accurately modeling this time series data. The constraints of current models have prompted the creation of regime-switching volatility models, which have proven effective in modeling such time series. Markov switching volatility models provide improved efficiency by incorporating regime dependencies in the examination of regime transitions. Nonetheless, the models rely on accurate identification and modeling of the fundamental Markov chain that dictates the variance regimes. This work presents the Large Deviation Markov Switching GARCH model. The huge deviation principle is employed to discern the fundamental variance regimes. The concealed states of the heteroskedastic time series are modeled and then employed to fit the Markov switching GARCH model to the heteroskedastic series. The model is subjected to a comparison study with the MSGARCH and standard GARCH models, which are implemented using the MSGARCH and rugarch packages in R, respectively. The in-sample prediction forecasting accuracy of the three models was evaluated using RMSE, MAE, and correlation coefficient metrics. The findings indicated that the proposed model displayed enhanced predictive accuracy relative to the conventional GARCH and MSGARCH models. The LDP’s identification of variance regimes within the series improved its capacity to estimate volatility, which enhances the model’s suitability for volatility modeling.
No takes yet. Share an insight, caveat, or question.
Benedict et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: