Modeling long memory in the EU stock market: Evidence from the STOXX 50 returns

Abstract

This paper examines the persistencebehaviour of STOXX 50 returns. To this end, weestimated the GARCH, IGARCH and FIGARCHmodels based on a data set comprising the daily returnsfrom January 5th, 1987 to December 27th, 2013. Theresults show that the long-memory in the volatilityreturns constitutes an intrinsic and empiricallysignificant characteristic of the data and are, therefore,in consonance with previous evidence on the subject.Moreover, our findings reveal that the FIGARCH is thebest model to capture linear dependence in theconditional variance of the STOXX 50 returns as givenby the information criteria.