This study examines the impact of sudden changes on volatility persistence, or the long memory property, in six Asian stock markets: Hong Kong, Korea, Indonesia, Malaysia, Thailand, and Singapore. We examine sudden changes associated with global financial and political events, specifically, the 1997 Asian currency crisis, the 1998 Russia crisis, the IT dot com bubbles of 2000, the 9/11 terror attack of 2001, and the recent financial crisis of 2007-2010(sub-prime mortgage crisis and Lehman Brothers bankruptcy). When these sudden changes are incorporated into GARCH and FIGARCH models, the evidence of persistence or the long memory property vanishes from volatility. This result suggests that ignoring the effect of sudden changes overestimates volatility persistence. In addition, out-of-sample analysis confirms that volatility models, which incorporate sudden changes,
provide more accurate one-step-ahead volatility forecasts than their counterparts without sudden changes. Thus, incorporating information on sudden changes in conditional variance may improve the accuracy of estimating volatility dynamics and forecasting future volatility for researchers and investors.
Keywords : Volatility Forecasting, Sudden Changes, Long Memory, ICSS Algorithm