Autocorrelation of the trade process

Abstract

This study investigates how duration-based trading intensity modifies the first-order autocorrelation and the transitory variance of the trade process. Because prices are conditional expected values, a structural model in which the trade duration represents the rate at which prices incorporate new information is developed. This refined model is an extension of the one developed by Madhavan, Richardson, and Roomans (1997) and allows parameters characterizing the arrival rate of new information to be derived. Testing this model with data from the Helsinki Stock Exchange, I was able to determine that a model ignoring trading intensity effects on price changes would underestimate the transitory effects of the trade process. This finding suggests that trade duration captures neglected elements of implicit trading costs that increase with market microstructure effects