Stylized facts on the temporal and distributional properties of daily FT–SE returns

Abstract

This paper investigates the temporal and distributional properties of the London Stock Exchange FT–SE daily indices by examining the autocorrelations and distributions of a family of return transformations. Power transformations of absolute returns are more highly autocorrelated than actual returns, with the strongest autocorrelation occurring for powers around unity. Such transformed returns do not, however, display long-term memory. Absolute returns, after outlier reduction, are approximately exponentially distributed and the analysis suggests that they could be modelled by asymmetric GARCH processes