Capturing Block Liquidity to Improve Algorithm Performance

Abstract

The authors discuss an advancement in algorithmic trading called block indication, also known as contingent or conditional orders. This concept implies the sending of non-binding indications of liquidity to a system that alerts a trader to a potential trade counterparty. The technique allows for algorithms to aggregate fragmented liquidity and helps create more natural crossing opportunities. In 2009, ITG integrated block crossing with its algorithms, allowing them to cross blocks with traders and other algorithms in its POSIT Alert^® block crossing system. The benefit of this integration is then quantified, revealing that overall execution cost (versus arrival price) decreases approximately 10 bps if half the order is executed in blocks (holding order difficulty level and market conditions constant.) The methodology can be used by other broker-dealers as well.