Latency and liquidity provision in a limit order book

Abstract

We use a recent, high-quality data set from Nasdaq to perform an empirical analysis of order flow in a limit order book before and after the arrival of a market order. For each of the stocks that we study, we identify a sequence of distinct phases across which the net flow of orders differs considerably. We note that some of our results are consistent with the widely reported phenomenon of stimulated refill, but that others are not. We therefore propose alternative mechanical and strategic motivations for the behaviour that we observe. Based on our findings, we argue that strategic liquidity providers consider both adverse selection and expected waiting costs when deciding how to act.

Keywords:

• Limit order books
• Liquidity provision
• High-frequency trading
• Price impact

Acknowledgements

Julius Bonart thanks the Institute for Pure and Applied Mathematics at UCLA for hosting him as a visitor while conducting part of this research. We thank Jean-Philippe Bouchaud, Rama Cont, Jonathan Donier and Charles-Albert Lehalle for useful discussions. We thank Jonas Haase and Ruihong Huang for technical support. We also thank two anonymous reviewers for their helpful comments.

Notes

No potential conflict of interest was reported by the authors.

1 The term ‘zero intelligence’ is used to describe models in which aggregated order flows are assumed to be governed by specified stochastic processes. In this way, order flow can be regarded as a consequence of traders blindly following a set of rules without strategic considerations.

2 Many authors use the term ‘market maker’ to describe an institution that performs this role. However, in the context of an LOB, this term does not imply that a given institution is a designated ‘specialist’ with elevated status in the marketplace, as was the case for market makers in older, quote-driven markets.

3 To ensure that our results are robust to the choice of time period, we also repeated our calculations using data from 1 March 2013 to 1 September 2013. We found that our results for this period were qualitatively similar to those for 1 March 2015–1 September 2015.

4 We use data from the Fidessa Fragmentation Index to estimate the fragmentation of global equity markets. For more details, see http://fragmentation.fidessa.com/fragulator/.

5 Kirilenko and Lamacie (2015) argue that this type of latency consists of two components: market-feed latency, which is the time it takes for an automated trading platform to disseminate market data, and communication latency, which is the time it takes for a message to travel between a trader’s computer and an automated trading platform.

6 Similarly to equation (3(3) ), we use the indicator function to align buy-side and sell-side activity (see section 6.2).

7 We conjecture that this small difference in system-latency times is caused by the extra processing required for trade confirmation and clearing after a market order arrival.

8 Gareche et al. (2013b) provides a brief remark that this is indeed the case on Nasdaq.

9 This is indeed the case on many other trading platforms (Hautsch 2011).

Additional information

Funding

Julius Bonart gratefully acknowledges support from CFM and Martin D. Gould gratefully acknowledges support from the James S. McDonnell Foundation [grant number 220020411].