Forecasting CSI 300 Volatility: The Role of Persistence, Asymmetry, and Distributional Assumption in Garch Models

This study investigates the daily volatility forecasting for China Securities Index-C300 series from 2002 to 2010 and identifies the source of performance improvements between volatility specification and distributional assumption. Empirical results suggest that CGARCH model achieves the most accurate volatility forecasts. Such evidence, along with the results of sign bias tests, demonstrates that modeling persistence components is more important than specifying asymmetric components for improving volatility forecasts of financial returns. Furthermore, the GARCH models with Gaussian distribution are preferable to those with more sophisticated error distributions.