The microstructure of a U.S. Treasury ECN: the BrokerTec platform

This paper assesses the microstructure of the U.S. Treasury securities market, using newly available tick data from the BrokerTec electronic trading platform. Examining trading activity, bid-ask spreads, and depth for on-the-run two-,three-, five-, ten-, and thirty-year Treasury securities, we find that market liquidity is greater than that found in earlier studies that use data only from voice-assisted brokers. We find that the price effect of trades on BrokerTec is quite small and is even smaller once order-book information is considered. Moreover, order-book information itself is shown to affect prices. We also explore a novel feature of BrokerTec--the ability to enter hidden ("iceberg") orders--and find that, as predicted by theory, such orders are more common when price volatility is higher.