The US equity markets have become increasingly complex in recent years. In part due to regulatory changes such as decimalization and Reg NMS, and in part due to technology improvements such as fast data communication, the trader has a wide variety of choice of where to send each piece of a big order (see [Hasbrouck, 2007, Appendix] for a review). Furthermore, the human trader is increasingly replaced by an algorithmic trading engine, which makes routing decisions based on continuous real-time updating of its model of market liquidity. Doing this well is “smart order routing.”
The problem is made vastly more complicated by the presence of “hidden liquidity:” limit orders resting at a market center whose existence is not disclosed to outside traders. …
If all liquidity were displayed, then smart order routing would be easy: the difficulty comes from trying to discern the presence of hidden liquidity and to use that information in making routing decisions. …
Our goal is more ambitious: we want to maintain a dynamic estimate of the hidden liquidity present on each of several diff erent exchanges, and use this information to make routing decisions now and in the future.
No comments:
Post a Comment