All (Useful) Speed Bumps Are Asymmetric
On Friday, the SEC rejected the proposal by EdgA to apply an asymmetric speed bump. The 4-millisecond speed bump would have been applied to incoming orders, but not to cancelations/modifications of resting orders. In its rejection, the SEC stated that
“The proposed rule change is discriminatory in that the Exchange would delay incoming executable orders by 4 milliseconds, which would allow market participants with orders on the EDGA book that are not subject to the delay up to 4 milliseconds to cancel or modify their orders”.
The rationale behind the rejection is striking since, for any speed bump to be effective, it must create an asymmetric effect on some orders. Otherwise, faster traders would still “pick off” slower traders – it would just occur with a slight delay.
The IEX Speed Bump
For example, the existing IEX speed bump is itself asymmetric and has been its entire life. The primary benefit of the IEX speed bump is that it prevents faster traders from “picking off” pegged orders whose (pegged) prices have yet to be updated. While incoming orders must pass through the speed bump, the prices of pegged orders are updated immediately, without the 350-microsecond delay. In effect, the resting pegged orders are able to have their limit prices modified immediately, while incoming orders (and modifications) must pass through the speed bump, creating an asymmetry between pegged and unpegged orders.
A similar asymmetry exists in the recent IEX proposal to extend its novel Crumbling Quote Indicator (CQI) to lit orders, referred to as D-Limit orders. The limit prices of these D-Limit orders would be automatically adjusted whenever the CQI detected a crumbling quote. However, these CQI-driven modifications would not be subject to the IEX speed bump, giving D-Limit orders an edge over other orders – both marketable orders attempting to hit those orders as well as other (non D-Limit) resting orders being canceled/modified by the traders themselves.
Can speed bumps be effective and symmetric?
For speed bumps to be effective, the sequence of events has to be altered in some way, as in the applications of the IEX speed bump above. Some messages must be able to avoid the speed bump, otherwise the speed bump simply delays the inevitable. It’s like watching those cartoons where the coyote and his nemesis stop chasing each other for the day when the work whistle blows, only to pick up right where they left off the next day. In the case of EdgA and IEX, the speed bump changes the sequence by allowing the slower liquidity providers an opportunity to cancel/modify their prices (or to have the exchange modify their orders on their behalf). The end result is that some trades that would have otherwise occurred will no longer occur. If the speed bump were symmetrically applied, though, the same trades would have occurred – they just would have been delayed by the length of the speed bump.
Therefore, the debate surrounding “asymmetric speed bumps” is really a debate over speed bumps themselves, as the term “asymmetric speed bump” is effectively redundant. It’s like saying “tall giant”, “round ball”, or “cheating Astro”.
So what does this mean for other speed bump-related proposals?
The SEC’s rejection of EdgA’s proposal on the basis of being discriminatory would suggest the SEC will reject similar proposals that give liquidity providers an edge over liquidity “takers”, such as the IEX’s D-limit. In fact, one could argue that the asymmetry is more severe under the IEX D-Limit than under the EdgA proposal since the IEX asymmetry would only benefit resting D-Limit orders. Cancelations/modifications to resting non-D-Limit orders must pass through the speed bump. If the reason for rejecting EdgA was asymmetry (i.e., discrimination), then apparently the D-Limit order would be rejected as well.
With that said, the SEC allowed the original IEX speed bump to be implemented and applied asymmetrically on pegged orders. So, a case could be made either way. It really boils down to whether the SEC is going to hold firm on its definition of “discrimination” going forward, or whether it will be flexible if the discrimination (i.e., asymmetry) favors more than just high-speed traders.
The author is the Founder and President of The Bacidore Group, LLC. For more information on how the Bacidore Group can help improve trading performance as well as measure that performance, please feel free to contact us at firstname.lastname@example.org or via our webpage www.bacidore.com.
For an overview of our other services, please click HERE.
And please check out our other blog posts available HERE.
And if you'd like to receive notification of new posts, please join our mailing list below.
Copyright 2020, The Bacidore Group, LLC. All Rights Reserved.
 See page 11 of the SEC order.
 Of course, the indirect benefits of reducing adverse selection costs to liquidity providers would include potentially tighter spreads and greater depths.