The econometrics of ultra-high-frequency data

Ultra-high-frequency data is defined to be a full record of transactions an d their associated characteristics. The transaction arrival times and accom panying measures can be analyzed as marked point processes. The ACD point p rocess developed by Engle and Russell (1998) is applied to IBM transactions arrival times to develop semiparametric hazard estimates and conditional i ntensities. Combining these intensities with a GARCH model of prices produc es ultra-high-frequency measures of volatility. Both returns and variances are found to be negatively influenced by long durations as suggested by asy mmetric information models of market micro-structure.