Market Data and Trading Costs
The proliferation of data has greatly improved institutional trading desks’ ability to analyze their own transactions along with broader market activity for clues on how to trade better.
In U.S. equities, the issue is that the Securities Information Processor industry utility is as dated as avocado-colored kitchen appliances, and there are systemic issues that go beyond dollar figures, according to Ray Ross, CTO and a Founder at Clearpool Group.
As exchange data is a fixed cost for all market participants, the highest-volume banks and market makers end up trading for low cost or no cost, while smaller firms subsidize the largest firms and may get squeezed out and forced to consolidate. That’s according to Tyler Gellasch, Executive Director at Healthy Markets.
The U.S. Securities and Exchange Commission is looking at the issue. In a speech earlier this year, Chairman Jay Clayton said the SEC has heard industry concerns regarding the SIP, and the data utility may need to be upgraded across content, governance and infrastructure.
“There needs to be a wholesale look at this in a variety of areas,” Ross said. “For one, we need to look at what’s included in the SIP as opposed to the proprietary data feeds. Is there, for example, imbalance information, odd lots, and at least some level of depth of book? And then some thought and discussion needs to be around who’s getting that data and how it’s used.”
Shortcomings of SIP-generated core data effectively force broker-dealers to pay exchanges for proprietary data in order to provide competitive execution services and meet the best-execution requirements of regulators, Clearpool said. That could be remedied if SIP is upgraded to include depth of book and odd lots, and if the speed of SIP data dissemination is upgraded to be in-line with private data feeds.
Regarding governance, a sticking point is that today the exchanges govern the SIP, and thus “have a disincentive to either invest in the SIPs or to make SIPs competitive products,” Clearpool said. The firm suggested adding meaningful broker representation to the SIP and recognizing exchange operators as a single entity for purposes of voting, rather than having one vote per exchange.
Exchange operators maintain their data prices as fair, but at the same time they acknowledge the data-procurement framework needs an update. Intercontinental Exchange’s New York Stock Exchange has proposed enhancing the SIP by offering depth of book, odd-lot quotes and primary auction imbalance information in a three-tiered product suite.
The premium data product “would address concerns that the SIP’s current content may be insufficient for institutional investors and active traders,” NYSE wrote in an August blog post. “Reconstituting the SIP as a three-tiered product suite will better support industry needs, reduce the administrative burden on data subscribers and address concerns about the changing nature of core data in modern markets.”
While details of any way forward need to take shape, NYSE’s plan is at least a step in the right direction. “Even the exchanges now are starting to realize that a change is needed, and they are making reasonable proposals,” Ross said. “So maybe the needle has moved enough that we can have some consensus on moving forward and finally get some change.”
SIP provides essentially a government function, but for-profit companies are in charge along with Finra, noted Gellasch of Healthy Markets. “Exchanges have an obligation to maximize profit for their shareholders, which means maximizing the prices that they charge,” he said. “I don’t believe there’s a way to cure that conflict of interest if exchanges are still in charge of the SIP.”
“The best path for regulators is to go back to the future,” Gellasch added. “The Exchange Act and its amendments say exchange rules have to be non-discriminatory and can’t be undue burdens on competition, and fees have to be reasonable and equitably allocated. The Commission seems eager to do this.”
With the right changes, Clearpool said, “reliance on exchanges’ proprietary data feeds can be reduced…and market participants can rely on the SIP as an accurate indicator of the marketplace.”
Between 80-90% of enterprise data is unstructured so most firms cannot unlock its value.
The entire Jefferies family mourns the loss of Peg Broadbent.
First Canadian bank becomes member of RepoClear.
Covid-19 is accelerating the integration of sustainability concerns in research.
DeepFX uses deep learning technology to predict the short-term price movements of six currency pairs.