Modeling persistence and long memory under the impact of regime shifts in the PIGS stock markets

Abstract

This paper investigates the impact of regime shifts on the persistence and long memory characteristics of volatility from the vantage point of modeling volatility in the PIGS economies (Portugal, Italy, Greece, and Spain). We apply the AIT ICSS algorithm to identify the points of sudden changes in volatility and incorporate these sudden changes in the GARCH class of models (GARCH and FIGARCH models). Our findings indicate that when regime shifts in the unconditional variance are incorporated in the GARCH class of models, the estimated persistence and long memory fractional parameter in the volatility of returns exhibit a drastic decline. Furthermore, we find that sudden changes in the unconditional variance are largely associated with various macroeconomic events.