IEX Responds to Exchanges’ Comment Letters

IEX is writing in response to comment letters sent by Nasdaq, NYSE, and BATS. Each of these incumbent exchanges has questioned the credibility and accuracy of IEX’s recent statements. IEX appreciates the opportunity to respond to our competitors to address these misunderstandings and to avoid further miscommunication.

Nasdaq’s Second Comment Letter Regarding its Router Approval from 20121:

First, Nasdaq’s comment letter suggests2that IEX claimed that IEX’s router functions in the same way as Nasdaq’s router. That suggestion is not accurate.

Second, Nasdaq seems to be missing the point that IEX set forth in our first and second response letters: the SEC granted Nasdaq’s router an “advantage” that no other Nasdaq member has to protect Nasdaq’s routing clients. This exclusive advantage: Nasdaq’s router receives a view of displayed and dark orders to maximize shares traded on Nasdaq without degrading its clients’ fill rate on away venues. We highlighted this fact to show that this is an advantage given only to Nasdaq’s router, and not to any other Nasdaq member. The premise of an exchange router having a unique advantage is similar to the IEX router not having to traverse the POP, which enables our clients to receive an execution from IEX without degrading their ability to receive fills on away venues. In the case of both the Nasdaq and IEX router, the commonality is not the method or type of advantage, but the fact that an advantage exists. More importantly, both Nasdaq and IEX have the same objective of improving fill rate when routing orders to away venues.

IEX has no issue with Nasdaq’s 2012 router change and applauds Nasdaq for seeking to protect its routing clients. We also commend the SEC for approving the change, at the time stating: “The proposed rule change meets these requirements in that it promotes efficiency in the market, and should, as represented by Nasdaq, increase the likelihood that a routable order will receive faster and better executions.”

IEX pursues the same objective, albeit via a different method; both Nasdaq and IEX have created what critics claim is an “advantage” for our routers versus our members. This comparison demonstrates that the SEC has already approved solutions that protect routable orders from being systematically disadvantaged by electronic front running.

In footnote 10 of its letter3, Nasdaq notes that its “routing broker” does not receive orders from members nor does it receive a “special market data feed.” We agree and recognize the separation between an exchange system (i.e. “routing logic”) and the exchange “routing broker” which is a simple courier to away markets. With this interpretation, we acknowledge that Nasdaq’s “routing logic” is entitled to have perfect visibility into the exchange order book (displayed and dark) as any exchange system should. Unfortunately, IEX’s “routing logic” is being mischaracterized by our critics as part of our “routing broker” when in fact, every other exchange’s “routing logic” is set forth in its rules as part of the exchange.

Why is IEX’s “router” deemed to include both the “routing logic” and the “routing broker” when every other incumbent exchange’s “router” is deemed to be just its “routing broker” (e.g. without the “routing logic”)?

Once again, this position unfairly creates an anti-competitive double standard against which IEX now must defend. So although we agree with Nasdaq’s point, since IEX has not been afforded this same interpretation, we felt Nasdaq’s “special market data feed” was an appropriate example of an advantage given to an exchange “router” that is not available to other exchange members. IEX’s “router” should be treated no differently from every other exchange’s “router,” namely to recognize the distinction between the “routing logic” component of the exchange and the “routing broker” facility of the exchange.

NYSE response to the IEX op-ed

In its second comment letter, NYSE responded to an IEX op-ed titled “The NYSE Speed Bump That You Weren’t Aware Of.” This op-ed highlighted the contradiction of NYSE attacking IEX’s speed bump, when NYSE had already created a delay variance of up to 400 microseconds within its own system. With so much potential debate being attributed to the IEX POP, it is fair to ask how NYSE can offer different means of access that vary by a comparable amount of time, especially without adequate public disclosure.

What is adequate access disclosure for an exchange? For an exchange with varying methods of access, we believe BATS sets a proper standard that clearly shows latency differences4for order entry and market data delivery. From our perspective, BATS disclosure seems to be adequate and NYSE’s does not (no public latency disclosure). We have not been able to identify rule filings by NYSE for implementation of the Binary Gateway which disclose the latency differences.5

NYSE argued that our op-ed implied that NYSE charges different prices for the different order entry protocols. In fact, we did not say or suggest this was the case. What we did say was that “Other exchanges offer products with different access speeds: connection ports, co-location, and market data – charging a premium to the fast which enables them to make money trading against the slow.” We unequivocally stand by the truth of that statement, and regret if it was misunderstood. However, we continue to believe that NYSE willfully offers different speeds of access in order to increase its profit margins and also to increase trading opportunities on its venue.

Exchanges have long relied on the tenet that it is the members’ responsibility to make choices regarding product and access offered, and that it is the exchanges’ obligation simply to disclose. We believe there is a fundamental gap between disclosure and transparency. The sheer volume of exchange communications makes it extremely difficult for members to gauge importance to their experience, especially when data and access products are described without critical details.

We think maintenance of these latency differences and a tiered access model promotes the value of speed and contributes directly and deliberately to outsized pricing power wielded by the exchanges over market access: order entry ports, market data, and colocation space. We do not think anyone, other than the exchanges, would disagree with the statement that the shift of exchange revenues from transactions to recurring, and ever-rising, fees for access and information is neither equitable nor congruent with its responsibility to its members, issuers, or investors.

This is another reason the US market urgently needs substantive competition among exchanges.

The BATS ALLB Router vs. Member Serial Router

In our third comment letter, we used the BATS ALLB routing option to illustrate that such a function, while perfectly reasonable for an exchange to offer, has an obvious advantage over a third-party broker accessing all four BATS markets using a serial router. We highlight this comparison to address the accusation that IEX’s router has an “insurmountable advantage” over a serial router asserted by some commenters. It’s irrational to determine that an exchange’s router has an “insurmountable advantage” by comparing it only to an illustrative serial method, which is only one design among many, and widely accepted to be far less efficient than the more common spray routing approaches. We never indicated that there was anything wrong with BATS ALLB; it was referenced as a comparison because BATS ALLB was very recently approved by the SEC, and is a design that has a fundamental advantage over a serial router.

Conclusion

IEX has not set out to misrepresent anything through intent, negligence or recklessness. We stand by our words and actions since our founding. To the extent there are misinterpretations of our message, we appreciate the opportunity to clarify them with this letter and will always seek to engage in healthy dialogue on aspects of market structure that impact investors and the trading community.