Reforming the SIP with Odd Lots and Competing Consolidators
By Ivy Schmerken
With attention riveted on volatile stock and bond markets and the economic fallout from the COVID-19 pandemic, brokers and asset managers are focused on keeping up with the high volumes of market data and trades.
With so many unknowns, the situation could impact market data reforms to overhaul the securities information processor or SIP.
Last week, the U.S. Securities and Exchange Commission announced that it will extend the comment period for the January proposed order on modernizing the SIP’s governance due to challenges that market participants are facing with the COVID-19 pandemic. The 60-day comment period will expire in March. The SEC said it won’t act upon the final rule before April 24 in order to give commenters additional time if needed.
However, market data has been a lightning rod for debate as many market professionals have complained that the public consolidated data feed is inferior to the faster, proprietary data feeds sold by exchanges.
On Feb. 14, the SEC released a proposed rule to overhaul the content and speed of the consolidated stock data feeds, suggesting the creation of multiple competing consolidators to level the playing field. The regulator has not significantly updated the rules that govern National Market System (NMS) market data since their implementation in the 1970s.
“Both the content of NMS market data and the technologies used to collect, consolidate, and disseminate that data have lagged meaningfully behind proprietary data products and systems offered by the exchanges,” stated SEC Chair Jay Clayton in a summary outlining the proposed rule.
This came on the heels of the SEC’s Jan. 8 proposed order to increase transparency and address conflicts-of-interest in the governance of the National Market System (NMS) consolidated feed. In the proposed order, the SEC calls on the exchanges and the Financial Industry Regulatory Authority to unify the three equity data plans and asks them to submit one new plan called the New Consolidated Data Plan.
Currently, exchanges have multiple votes on the SIP Operating Committee, which is managed by exchanges (self-regulatory organizations) and FINRA through NMS plans. The new governance plan would reduce the number of votes for the exchanges and offer one-third of the voting structure to non-exchange members. It would also open up representation on the SIP Operating Committee to diverse participants: an institutional investor, a broker-dealer with primarily retail customers, a broker-dealer with primarily institutional customers, a securities markets vendor, a corporate issuer, and a retail investor. In addition, the administrator of the SIP would need to be independent and not offer a market data product.
Odd Lots & Auctions Data
The Feb. 14 proposed order seeks to update and expand the content of core data with odd-lot transactions for higher priced stocks [above $50 per share] and depth-of-book information, in addition to opening and closing auctions data, which are only available from proprietary data feeds sold by major stock exchanges.
“My personal opinion is they’re on the right track,” said Bernard Donefer, adjunct associate professor at the NYU Stern School of Business. ‘I think they’re doing a good thing and it’s about time,” said Donefer, referring to the SEC. “Having said that,” Donefer added that he thinks this is a first attempt, “a trial balloon,” that will be influenced by comments whereby market participants, investors and the public have an opportunity to weigh in on each proposal.
“You need to have richer, fuller content, depth-of-book, odd-lot information and order imbalance data. All of that data is critical to making it usable to the broker-dealer and institutional user community,” said Joe Wald, CEO of Clearpool Group, an institutional agency-only broker.
“Ultimately having a faster SIP and enriching the data, should result in a more cost-effective solution than the proprietary data feeds,” said Wald, in an interview.
Perhaps the most interesting aspect is the proposal to allow decentralized consolidators to enter the business of collecting, aggregating, and disseminating NMS stock data feeds.
“‘There is no question if they could pull off a distributed SIP which would focus on the speed and the focus on the content,” they are definitely moving in the right direction, said Wald, who participated in the SEC’s October 2018 market data roundtable, where many of these issues were discussed. Clearpool also submitted a comment letter supporting the SEC’s governance changes including expanding voting representation to non-SROs.
With stock [prices] at higher notional values, and stocks not splitting anymore, the proliferation of odd-lot information throughout the U.S. equity markets is important for determining best execution, said Wald.
“You’ve got meaningful liquidity that exists at odd-lot capacity and it makes the difference in knowing what the spread is and what it isn’t,” said Clearpool’s Wald. “There is definitely is a best execution rationale around having odd lots posted in the SIP because it will really show what the true market is,” he said.
According to the SEC’s analysis and comments made at the SEC market data roundtable in October 2018, “Odd-lot orders are frequently priced better than the quotation prices that are disseminated by the exclusive SIPs, yet these orders are not seen by investors or market participants that rely solely on SIP data.”
There are also significant decision-making points in the open and close auction data for passive and active managers and well as individual investors, he said. “Such an important concentration of liquidity takes place now at the beginning and end of every day just because of the way the markets have evolved such as how much passive interest is there through ETFs,” said Wald. “It’s important to understand order placement and how to achieve best execution for your client,” he said.
However, the big question is whether a faster, richer, and more robust SIP will be accepted as a viable alternative by market participants and broker dealers from a regulatory standpoint, said Wald. Some high-profile buy-side traders have publicly stated that they won’t send order flow to brokers unless they have direct data feeds with which to execute their orders.
“A lot of details still need to be fleshed out to determine whether it can be accepted as a viable alternative,” said Wald.
Another catalyst for beefing up the SIP has been cost pressures on brokers that must purchase the proprietary feeds to make decisions for their clients. A broker could easily spend tens of thousands of dollars per month on subscribing to proprietary market data feed products. According to the SEC’s February proposal, next to employee compensation, market data is a firm’s second highest expense.
“There was a spike in demand for exchange proprietary data products based on the feedback I received from third-party market data service providers after the 2018 SEC US Equity Market Structure Roundtables on Market Data and Market Access,” said Khody Azmoon, an electronic trading consultant specializing in electronic execution and platform solutions. “The crux of the matter is while larger brokerage firms could afford access, this could be challenging for smaller brokerage firms,” he said.
“By adding some of the proposed NMS market data changes, the SEC will help improve the competitiveness of those brokerage firms that rely on the SIP data products,” said Azmoon.
Supporters of decentralized consolidators point to this model as a way to solve the geographic latency that exists between the proprietary, exchange data feeds and the current single SIP consolidator. Although the SIP has been made faster by Nasdaq acting as the plan consolidator for Tape C securities and NYSE acting as the consolidator for all Tape A securities, the aggregation process adds latency.
While agreeing that the SEC is heading in the right direction, NYU Professor Bernard Donefer said there are some issues that the SEC’s proposed order doesn’t address.
“If the SEC decides to standardize on all of these SIP consolidators, then data has to leave at the same time to be consistent. However, that latency is a function of distance and infrastructure,” he said. For example, if a broker is in New Jersey and the SIP is in New Jersey, that transmission is going to be faster than if the broker is in San Diego. High frequency traders are always going to be faster because their servers sit next to the matching engine, he added.
Under the new plan, Wald said, there’d be a consolidator in NYSE’s data center in Mahwah, New Jersey, and there’d be one in Nasdaq’s in Cartaret, while CBOE’s data centers are located at NY4 run by Equinix in Secaucus. “And from those data centers that are disseminating the SIP, firms could have connectivity directly there, which will pick up the SIP at a more similar access [point] that [they]would get from a proprietary feed,” said Wald.
But, Donefer cautioned that having competing consolidators will introduce complexity into best execution. “If there are multiple SIPs, the question becomes what is the best price going to be? And if there are multiple SIPs, what is the best quote? “asked Donefer. “I don’t think anybody has thought out how competing SIPs will ensure how everybody gets the information on a timely basis,” said Donefer.
While he agreed that including depth-of-book data showing bids and offers at five levels-up and five levels-down the order book will help to fill an information gap between the SIP and the proprietary feeds, Donefer contends that HFT firms will still have an edge. “There will be high frequency traders with collocated servers that will be able to act on the order-book data before the SIP is able to aggregate the data and send it out again,” said Donefer. “This isn’t going to go away,” he said.
If the decentralized consolidator model makes the SIP more competitive, this could have a big impact on exchange revenues, Donefer pointed out. “If the decentralized, competing consolidator model is cheaper, there are firms buying the proprietary exchange feeds today, which will say that the SIP is good enough,” which is going to erode exchange revenues, said Donefer.
Donefer also questions what the addition of more data to the SIP feed mean in terms of infrastructure capacity to process all of that data. Expanding the content of the SIP with odd-lot data and depth-of-book could amount to transmitting four-or-five times the amount of data, increasing costs for the industry and retail brokers that send this data to active traders, he said.
Next Steps or Delays?
Market participants will have until May 26 to submit comments on the February proposed order.
Meanwhile, Nasdaq submitted a comment letter on Feb. 28, pointing out inconsistencies between the two proposals and requested clarification from the SEC. It also criticized the length of the second proposal at nearly 600 words, adding the SEC “is proposing Regulation NMS 2.0 under the guise of a proposal on market data.” In particular, Nasdaq notes that the proposed rule is changing the operation of rules pertaining to quotation display, locked and crossed markets, trade-throughs of displayed orders, and other matters.
It’s possible that a holistic review of Reg NMS may be necessary in order to tackle the market data reforms. With the potential inclusion of odd-lot quotes for high priced stocks (ranging from $50 to $1,000 per share) into the best-bid-and-offer (BBO), this could involve changes to displayed quotes and the Order Protection Rule which prevents trade-throughs for round lot orders (100 shares).
Others contend that it’s necessary to address the governance piece first in order to fix the problems with the SIP. However, Clearpool’s Wald said this is not the case, and that the SEC intentionally created two independent plans so that either could be implemented without the other.
Given the market uncertainty and global pandemic, it remains to be seen whether firms will be able to carve out time to comment on the market data reforms or whether they will request more extensions. On the other hand, economic repercussions from the public health crisis could intensify the need to reduce market data expenses for broker dealers.
“We’re at the beginning. It’s unknown how long it’s really going to take to see action on this proposal. But there are some timelines with respect to comments and then comments get vetted. And then the next thing happens.” Wald predicted there will be challenges in the process, but said he is optimistic that progress will be made.
“It’s a bold step forward for sure and it shouldn’t be diminished in any way. It really is that path toward meaningful change,” said Wald.
Data Service Integrations Available with FlexTrade’s Execution Management System Technology
- Kensho Actionable Intelligence
- Dataminr Global News Alert Feed
- Symphony Communication Services
- OTAS Trading Intelligence And Analytic Solutions
For further information, please contact us at firstname.lastname@example.org.
Past blog posts related to Data Issues
Consolidated Market Data Feeds Surpassing SIP in Trading
BIG Data: Getting Granular with ESG Factors
Data Science Platforms Help the Buy-Side Integrate Alternative Data
Algo Development 2.0 Looks to Open Source, Cloud & Big Data