Securities and Exchange Commission Chairman Jay Clayton and Brett Redfearn, the director of the SEC's Division of Trading and Markets, swung for the fences Friday as they highlighted potential changes the regulator is investigating regarding the content, performance, and governance of the Securities Information Processors.
It is about time the SEC and the industry review the positive and negative changes that Regulation NMS made to the equities market a decade ago. The continuous improvements in trading algorithms, compute power, and networking technology had brought changes to the market’s structure that few could have imagined before the SEC implemented Reg NMS.
Most of the SEC’s suggestions are perfectly reasonable. Adding depth-of-book data to the SIPs’ top-of-book data would acknowledge that the necessities for traders have changed over the years. Feature creep happens all the time. Who would buy a laptop that required a PCI wireless card or a car without a stereo or air-conditioning?
However, it is doubtful that the addition of the depth-of-book data would do more than raise the price the SIPs charge for their data and require firms to invest in fatter pipes to handle the higher volume of data.
According to a Greenwich Associates survey conducted for its 2018 Market Data Survey, the report’s authors wrote that only 2% of the respondents would trade with an electronic broker that exclusively relied on SIP data. The number grew to 36% if the clients had regulatory clarity that they would be fulfilling their best execution requirement in such a fashion. The vast majority would still rely on proprietary feeds from the various exchanges.
It is a matter of physics, and no matter how hard regulators try, they cannot regulate physics. A direct feed from an exchange always will be faster than aggregated feeds that need to travel further and undergo additional processing.
The only thing that might eventually change the equation is replacing today’s computer networking apparatus that would use quantum entanglement, which would make the issue of distance between servers moot.
The most overdue, and most straightforward, reform would be to allow more stakeholders to participate in the governance of the NMS Plan. The SEC did not expect to see a rash of demoralization just before implementing Reg NMS. If it had, it probably would not have entrusted oversight of the market infrastructure to a small cadre of for-profit companies.
The regulator should be cautious that it does not broaden representation too much. The larger an organization becomes, the less effective it becomes.
Imagine the state of the Consolidated Audit Trail if its operating committee was twice or three times its size with buy- and sell-side representation: It would languish in development hell far longer than it already has.