What a difference a day makes : on the common market microstructure of trading days

This paper analyzes the interday stability of the price process using transaction data. While the vast majority of empirical studies on the microstructure of financial markets rests on the tacit assumption that observed prices are generated by a time-invariant price process, we question this assumption by means of a minimum distance estimation framework. Starting from estimates specific for each day's price process, this procedure enables us to work out a common structure across trading days and allows us to disentangle the pecularities of trading days which are marked by certain news events. The determinants of transaction price changes for the BUND future trading at the LIFFE on the basis of 22 subsequent trading days are analyzed. Our empirical findings confirm that trading days do share a common structure to a large extent. However, single event dominated days are likely to show a differing price process. On the one hand, this fact renders pooled parameter estimates inconsistent. On the other hand, this procedure opens an avenue for an in depth analysis of information processing in financial markets.