Determinants of Inter-Trade Durations and Hazard Rates Using Proportional Hazard ARMA Model

This paper puts focus on the hazard function of inter-trade durations to characterize the intraday trading process. It sheds light on the time varying trade intensity and, thus, on the liquidity of an asset and the informations channels which propagate price signals among asymmetrically informed market participants. We show, based on an exogenous information process, that the way traders aggregate information has implications for the shape of the hazard function. We use semiparametric proportional hazard model which is augmented by an ARMA structure very similar to the wide spread ACD model to obtain cinsistent estimates of the baseline survivor function and to capture well known serial dependencies in the trade intensity process. From an inspection of conditional transaction probabilities based on Bund future transaction data of the DTB we find a decreasing hazard shape providing evidence for the use of non-trading intervals as an indication for the absence of price information among market participants. However, this information content seems to be diluted by a high liquidity bade level, particularly with respect to a large inflow of potential traders from the U.S. Furthermore, we provide evidence that past sequences of prices and volumes have significant impact on the trading intensity in accordance with theoretical models.