Back in the good old days, a broker would take a client's order to the specialist post of the NYSE or to the appropriate crowd on the floor of the CME or CBOT and his first question would be: "What's the market?" The broker would receive information about the state of the market and then he (or she) could use his experience to execute the order.
This process allowed hidden liquidity to be revealed, and custom liquidity created specifically in response to the question to be spawned.
In today's computerized market it is easy to assess some of the available liquidity from exchange-published market data feeds, but the significant amount of liquidity that is hidden or which would only become present under certain conditions, is overlooked. If implemented the right way, a brief pause of a liquidity-seeking order would allow hidden and custom-spawned liquidity to also compete to fill the incoming order, thereby providing potential price and size improvement.
The duration of the pause can be variable based on the preference of the liquidity seeker, but a pause of as little as 20 or even 5 milliseconds is enough to give market-makers the chance to respond with what would otherwise be hidden or non-existent liquidity. The order information can be shielded behind a firewall so that all unfilled order information is fully protected. Institutional traders could choose longer pause durations and an "all-or-none" option to create competition to for block execution with NO risk of order leakage or front-running.
The reintroduction of this component to market structure brings many benefits. It allows for aggregation of liquidity and order confidentiality as discussed, but it also prevents pennying, spoofing, and other forms of gaming because none of the orders on either side are publicly displayed. It reduces quote traffic because market-makers have no need to layer the book, or even continuously work any orders at all. And it also eliminates the need for imposing minimum quote duration, or even a minimum tick increment, again because quotes aren't displayed. A further benefit is the opportunity to implement a brief auction for the best price and size during the pause. (Related: Tracing Trades Through the Electronic Maze )
The pause allows both sides of the trade to "opt into" the trade. The taker of liquidity opts into the trade the same way he always has, by sending the order. But because he is willing to wait for the brief pause to occur, market-makers are able to respond to his order and opt into the other side of the trade if they desire. Because unfilled order information is not disclosed to the market-makers, and because the pause forces multiple market-makers to compete for the order, at least a portion of the efficiency created by this improved risk profile for the market-maker will be passed back to the liquidity consumer in the form of price and/or size improvement.
Although allowing clients to pause orders for a short time seems somewhat simplistic, giving them this option will go a long way toward improving market structure for all participants. Increased speed has improved markets substantially, but the current speed arms race has become a dead weight loss for the industry. The pause process allows the last 20 milliseconds (or whatever) of the current winner-take-all speed incentive to be diverted into an incentive for price and size improvement.
D. Keith Ross is the Chief Executive Officer of PDQ Enterprises, owner of PDQ ATS, an equity trading venue that emulates the interaction and dynamism of traditional floor brokers in a high-speed, electronic trading environment, where algorithms compete to fill orders. Prior to PDQ, Keith was the Chief Executive Officer at GETCO and began his career as an options analyst in 1976.
image: © Ewen Roberts