A1865
Title: Stochastic coefficient GARCH-X models: Evidence from the US and ASEAN-5 equity markets
Authors: Manabu Asai - Soka University (Japan)
Tsubasa Furuichi - Soka University (Japan) [presenting]
Abstract: The Stochastic Coefficient GARCH-X (SC-GARCH-X) model extends the standard GARCH-X framework by allowing conditional variance parameters to vary over time as a Stochastic process driven by exogenous variables. Stationarity conditions and asymptotic properties of the Quasi-Maximum Likelihood (QML) estimator are discussed. Monte Carlo simulations examine the finite sample properties of the QML estimator to ensure its reliability. An empirical application to US and ASEAN-5 equity markets, incorporating range-based volatility as the exogenous variable, reveals that the SC-GARCH-X model provides superior in-sample fit and out-of-sample forecasts for developed markets by capturing dynamic volatility structures more effectively. However, certain emerging markets are better characterized by a simple asymmetric GARCH model, suggesting that the complexity of volatility dynamics varies across different stages of market development.
