The sixteen exchanges in the US equity market compete fiercely to be the first to receive marketable order flow. Orders at one exchange can easily be “queue-jumped” by others posted on exchanges listed higher in brokers’ routing tables. Those who jump the queue are eating your lunch—leaving your passive orders with a lower fill rate and forcing your execution algorithms to cross the spread. And orders at the back of the queue are likely to experience higher adverse selection when they do get filled.
Selecting a venue for a passive order is a complex problem because it depends on the specific exchanges competing at the NBBO at the time of order placement. The speed of a fill at NASDAQ depends not only on NASDAQ’s market share, but whether orders at the same price are also posted at IEX or EDGX, for example, or some other combination of three, eight, or fifteen other exchanges. And there are 65,535 possible combinations of exchanges that could be competing simultaneously at the NBBO.
In this groundbreaking research, we provide specific examples of “queue-jumping” and describe quantitative methods that can be used to identify the exchanges with the highest likelihood of filling limit orders in real time. We apply a ranking system called ELO that is used to rank chess players and online video gamers. We assign ELO ratings to exchanges and show how it can be used to determine which exchange is most likely to receive the next marketable order for any of the 65,535 possible combinations in competition. In this paper, we rank the exchanges under varying conditions…and the results may surprise you.
Most importantly, we share how ELO and other methods we introduce can be used to turn the problem of queue-jumping on its head. At BestEx Research, we’re using them to optimize our SOR’s venue selection for both hidden and displayed orders based on all sixteen limit order books in real time.
This research is now published in the December 2021 special issue of the Journal of Investing on Trading. © 2021 PMR. All rights reserved.