STUCTURAL BREAKS IN VOLATILITY OF STOCK MARKETS

Volatility can be defined as a value of risk of financial instruments over a specified time period. One of the most challenging practical problems is to understand and model a behavior of volatility dynamics through different time periods. In this paper, we deal with an impact of structural breaks on volatility persistence with a help of Iterated Cumulative Sum of Squares (ICSS) algorithm. The aim of this paper is to identify sudden breaks in volatility of financial time series which usually correspond to political or economic events, and measure an impact of these sudden breaks on volatility persistence. When incorporating those breaks into conditional volatility models, long memory property usually significantly disappears. Volatility persistence is estimated using (FI)GARCH family models in two regimes: without sudden breaks and with sudden breaks which are incorporated in volatility models in terms of dummy variables. Empirical analysis is provided on illustrative example using sample data from developed and emerging stock markets. Namely, we consider weekly data of S&P500, WIG20 and PX indexes in the period of 2004 - 2013 years which also includes the stage of recent global financial crisis of 2008-2009 years. Our findings mean that when ignoring an impact of sudden breaks on volatility it leads to overestimation of volatility persistence.