DP4456 Price Discovery in Tick Time

Author(s): Bart Frijns, Peter C Schotman
Publication Date: June 2004
Keyword(s): microstructure, nasdaq, price discovery, tick time models, ultra-high frequency data
JEL(s): C32, G15
Programme Areas: Financial Economics
Link to this Page: cepr.org/active/publications/discussion_papers/dp.php?dpno=4456

In this Paper we propose a tick time model for dealer quote interactions using ultra-high-frequency data. This model includes duration functions to measure the time dependence of volatility as well as information asymmetry. In order to assess price discovery we define several measures in tick time. These measures can be aggregated to calendar time and we define a comparable measure to Hasbrouck (1995) information shares. In our empirical part we examine the Island and Instinet Electronic Communication Networks, and three wholesale market makers for 20 actively traded stocks with varying liquidity at Nasdaq. Our results include that volatility does not increase with the duration between quote updates, and that longer quote durations lead to lower price discovery. In terms of price discovery we find that ECNs tend to dominate the liquid stocks, whereas market makers dominate the less liquid stocks.