Simulating the Synchronizing Behavior of High-Frequency Trading in Multiple Markets
Benjamin Myers and Austin Gerig
Abstract Nearly one-half of all trades in financial markets are executed by high-speed autonomous computer programs—a type of trading often called high- frequency trading (HFT). Although evidence suggests that HFT increases the efficiency of markets, it is unclear how or why it produces this outcome. Here we create a simple model to study the impact of HFT on investors who trade similar securities in different markets. We show that HFT can improve liquidity by allowing more transactions to take place without adversely affecting pricing or volatility. In the model, HFT synchronizes the prices of the securities, which allows buyers and sellers to find one another across markets and increases the likelihood of competitive orders being filled.
Financial markets have changed considerably over the last 20 years. During this time, most exchanges have switched from floor-based to fully electronic trading where orders can be sent to the market and executed with little or no human involvement (MacKenzie 2012). As a result, automated trading has flourished. One particular type of automated trading, known as high-frequency trading (hereafter HFT), has especially grown in size and importance. HFT exploits short-term price fluctuations and seeks a small profit per transaction many times throughout the day, without taking on significant overnight positions. Although difficult to determine
Department of Physics, University of Oxford, Oxford, UK e-mail: myers. benjamin. s@gmail. com
A. Gerig (H)
CABDyN Complexity Centre, Said Business School, University of Oxford, Oxford, UK e-mail: austin. gerig@sbs. ox. ac. uk
© Springer International Publishing Switzerland 2015
A. K. Bera et al. (eds.), Financial Econometrics and Empirical Market
Microstructure, DOI 10.1007/978-3-319-09946-0_____ 13
its true size, most studies estimate that about one-half of all transactions on major exchanges are due to HFT.
This study focuses on one particular effect linked to HFT—the synchronizing of price responses across multiple related securities (Gerig 2012; Gerig and Michayluk 2010). Figure 1 (taken from a recent article) shows this effect. Here, to analyze price synchronization in more detail, we simulate two markets where an identical security is traded and compare investor welfare when the prices in these markets are and are not aligned by the actions of HFT.
In our simulation, investors are modeled in a zero-intelligence framework (Gode and Sunder 1993; Farmer et al. 2005). This treatment strips out the idiosyncrasies of individuals’ behavior and assumes only local interactions are of significance— investors are only interested in meeting their own specific price expectations and they do not use complex strategies. To consider the effect of HFT, we simulate two zero-intelligence markets where an identical security is traded and allow HFT to
connect orders between the two markets when their prices cross. We show that HFT activity (as defined in the model) increases the probability that a typical investor entering the market will transact. Furthermore HFT activity reduces volatility so that prices are closer to their fundamental value.