How Does Automated Trading in the Treasury Market Respond to Economic News?

Project: Research

View graph of relations



Algorithmic trading, i.e., automated trades executed by computer program, has become more prevalent in financial markets in recent years. The effect of algorithmic trading, nevertheless, is still a relatively unexplored topic in the finance literature. In particular, there is little research yet in the fixed income market.We aim to fill this gap in the literature and study the impact of algorithmic trading in the U.S. Treasury market. The U.S. Treasury market is one of the largest financial markets in the world, with trading volume comparable to the U.S. equity market. Algorithmic trading has increased significantly in the Treasury market in recent years. According to Safarik (2005), over 50% of trading volume on BrokerTec, a major electronic trading platform of the U.S. Treasury securities, originates from automated trades. Given the importance of the Treasury market, we are interested in the role of algorithmic trading and its effect on market activities and the price discovery process of Treasury securities.The data used in our research is from BrokerTec and contains information of all transactions and limit order book over the period of 2003 to 2011. We identify an order as originated from algorithmic trading if the speed of reaction to changes of market conditions is beyond human capacity. The long sample period and vast dataset allows a detailed analysis on algorithmic trading in the US Treasury market and provides more statistical power than existing studies using data from equity market.This research project focuses on the following set of questions. The first question is how algorithmic trading is used around macroeconomic news announcements. Pre- and postannouncement periods represent different market environment. Traders have to assess the impact of announcement quickly to place new orders or cancel/modify existing orders. Computer programs are better adapted in handling such trades than human traders due to fast moving market conditions. The second question is whether algorithmic trading adds or reduces market liquidity. As algorithmic trades respond to information shocks quickly and potentially with large volume, do they supply or absorb market liquidity? The third question is whether algorithmic trading facilitates or hinders the incorporation of public information into bond prices. That is, does algorithmic trading help to speed up the price discovery process? Finally, we examine the impact of automated trading on price stability. It is generally believed that algorithmic trading is at least partially responsible for recent high market volatility and extreme returns.


Project number9041972
Grant typeGRF
Effective start/end date30/09/1227/09/16