Volatility persistence in asset markets: long memory in high/low prices

Abstract

This study addresses the issue of volatility persistence in asset markets by analysing the behaviour of daily ratios of highest and lowest prices for a number of different assets both inter-war and post-war. These series include sterling exchange rates, S&P futures prices, the FT30 and the price of gold. It is found that each of these series can be characterized as having the property of long memory so that observations far apart in time are non-negligibly correlated. More importantly, that each series can be modelled as fractionally-integrated noise, so that its behaviour can be captured by a single parameter.