Price Discovery in Tick Time

Abstract

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.

Keywords: Price discovery, tick time models, NASDAQ, ultra-high frequency data, microstructure

JEL Classification: C32, G15