Volatility Modeling and Forecasting Using Range-Based GARCH Models

Abstract

This paper investigates the forecast performance of symmetric and asymmetric GARCH models in comparison with symmetric and asymmetric range-based GARCH models. Specifically, we explore whether including the range and assuming asymmetry in the conditional variance equation significantly impacts the forecast performance of range-based GARCH models. The models examined in this study include GARCH(1,1), TARCH(1,1), RGARCH(1,1,1), and RTARCH(1,1,1). Our evaluation of these models utilizes different loss functions.Using daily, weekly and monthly opening, closing, highest and lowest all-share historical prices of the Nigeria Stock Exchange from 2014 to 2024, the results of data analysis reveal that incorporating the range and accounting for asymmetry in the conditional variance equation enhances the forecast performance of range-based GARCH models. Importantly, this finding holds for daily, weekly, and monthly forecast horizons.