The SEC Plan to Light Up Better Prices: Lots Still to Do
In this third part of our trilogy on the SEC’s proposed market structure reforms, John Ramsay looks again at relevant changes in the markets since Regulation NMS was put in place 18 years ago, what the SEC has already done and is proposing to do that affects how prices are displayed, and some of the questions that remain
Sometimes the most mundane and overlooked aspects of regulation can be among the most impactful. This is undoubtedly true for the rules around how exchanges express the best prices at which their members are willing to trade. A fundamental premise of regulation is that the markets depend on a high level of transparency about the prices people are willing to pay or receive for individual stocks. But the usefulness of displayed prices depends on the rules that determine how exchanges show them.
In this third part of ourtrilogy on the SEC’s proposed market structure reforms1, we look again at relevant changes in the markets since Regulation NMS was put in place 18 years ago, what the SEC has already done and is proposing to do that affects how prices are displayed, and some of the questions that remain. On the topic of how exchanges display prices, the SEC has already acted, as part of its Market Data Infrastructure Rule (MDIR), adopted in 2020.2 The question is therefore less about what change will come and much more about how and when it will be implemented.
What Rules Apply Today
Historically, stock markets have displayed their prices in “round lots.” In almost all cases, a round lot is equal to 100 shares. Why? Who knows. People like round numbers that are multiples of 10. Until very recently, the term “round lot”, notwithstanding its significance, was not defined in SEC rules but only in the rules of the listing exchanges. As of last April, only 11 listed stocks, with high per share stock prices, had a round lot size other than 100.3 Nine of these had a round lot of 10 shares, and two (including Berkshire Hathaway, the undisputed champion of high-priced stocks) had a round lot of a single share.
Of course, people also quote and trade (a lot) in “odd-lots”, a number of shares less than a round lot. The distinction between round and odd-lots matters for various reasons. First, only round lots count toward determining the national best bid and offer (NBBO), which is the basic equities benchmark all participants rely on to define the best quoted price and to measure the degree of “price improvement” an investor receives on a given trade. Second, only the best round lot bids and offers on each exchange are protected from being “traded-through” – meaning firms can’t execute a trade at a worse price without seeking to access the exchange quote. Third, the consolidated tape feeds published by the exclusive securities information processors (SIPs) don’t carry odd-lot quotes, so those quotes are visible only to people who pay for more expensive exchange proprietary data feeds. In contrast to quotes, odd-lot trades have been carried on the consolidated tape since 2013.4
One wrinkle is that exchanges often allow odd-lot orders from different traders to be added up to form a round lot. For instance, if two firms at virtually the same time each send displayed orders to an exchange to buy 50 shares at the same price, those orders will generally be aggregated to form one round lot to buy at that price. And if the price is the best displayed bid price for that stock on the exchange at the time the orders are received, the quote would be treated as protected.
What Has Changed?
It is interesting that the SEC’s release adopting Regulation NMS barely mentions and never interprets the term round lot. The same is true for “odd-lot” – the term is rarely used, and only to distinguish odd-lots from round lots.
It is clear that in today’s markets, the 100-share round lot is an anachronism, as average stock prices have increased and more quoting and trading has moved away from round-lots. It seems equally as clear that the absence of odd-lot orders from the consolidated tape means that people using those data sources are getting only a partial view of the prices people are willing to trade at – a “half-lit” view, you might say. Consider these points:
The average price of stocks in the Dow Jones Industrial Average index nearly quadrupled between 2004 and 2019, and it increased 18% from August 2020 to August 2022.5
Today odd-lot trades account for more than half of all trades in the markets, up from 43% at the beginning of 2020.6
Odd-lot orders are often at better prices than the NBBO, and the percent of the time this is true is substantially higher for higher-priced stocks.7
What Has the SEC Already Done?
The MDIR, adopted in 2020, created a substantial revision to the system for distributing consolidated market data. It greatly expanded the content of “core data” required to be contributed by all the exchanges and changed the distribution system to allow “competing consolidators” to replace the existing exclusive SIPs. In general, the MDIR, along with governance changes adopted around the same time, was meant to take consolidated data out of the control of the large exchange companies and increase its value so it could in some cases serve as a substitute for more expensive exchange proprietary data feeds.8
The MDIR made two changes relevant to lot size. First, it required the use of four new round lot sizes, based on stock price, as follows:
Second, the MDIR required that the consolidated data include all odd-lot quotes equal to or better than the NBBO, aggregated at each price level for each exchange. Thus, the consolidated data would show the total buy or sell odd-lot interest at each price level, covering all orders where the price is no worse than the NBBO. The odd-lot requirement will incorporate the new round lot sizes. So, in a stock with a 40-share round lot, an order for 50 shares will be displayed as a round lot, while one for 30 shares would be displayed as an odd-lot order, if its price was equal to or better than the NBBO. And, consistent with current practice, odd-lots will be aggregated, where possible, to form a round lot at the new sizes.
The SEC explained that the new round lot sizes were meant to “better ensure the display and accessibility of significant liquidity for higher-priced stocks” and to “reduce information asymmetries.”9 Translation: The new sizes will make the determination of the NBBO more rational across different types of stocks and put participants on a more level footing in reading and using market data. To the same effect, the new odd-lot order disclosure will increase the visibility of better prices, which can now be seen only by buying multiple individual exchange data feeds.
The MDIR called for an extended implementation schedule. The first major step involves the approval by the SEC (after proposal by the exchange SIP participants) of new fees to be charged to the competing consolidators. Full implementation will take at least two years after the approval of those fees. The round lot and odd-lot changes were scheduled to be implemented toward the end of the implementation timeline. The SEC explained it did not think it was necessary to require the existing processors to make these changes since they were expected to be “retired” with the addition of the new competing consolidators.
What Happened Next?
In two words: litigation and delay. The three large exchange companies, which had opposed the MDIR at each step of the way (IEX supported it at each step), sued in court to block it. They eventually lost that legal challenge last year. In the meantime, the existing tape plan participants, controlled by the same companies, proposed a fee schedule as required by the rule, but the SEC found that it was deficient in meeting the purpose of allowing competing consolidators to operate a commercially viable business (one might conclude it was designed for the opposite purpose). The participants have yet to offer a new fee proposal, and so the MDIR has been adopted and cleared by the court but is stuck in the mud, until the Commission determines how to get it unstuck.
To start the “unsticking” process, as part of its Regulation NMS changes, the Commission proposed to move up the implementation of the round lot and odd-lot changes by requiring that the SIPs make those changes on the existing consolidated data feeds, before competing consolidators come on board later. The Commission reasoned (reasonably enough) that its earlier logic – that the existing SIPs need not make these changes given their imminent retirement – no longer applies. The existing SIPs are looking at a much more leisurely path to retirement than had been foreseen, and the SEC believes the markets can benefit from moving up the timetable. IEX has previously written in support of the earlier implementation of these changes and continues to support doing so.
The Commission offered one additional requirement on top of what was adopted in the MDIR. It proposed that the consolidated data feeds would also include the designation of a new best odd-lot order, or “BOLO”. In essence, BOLO would reflect for odd-lot orders what the NBBO represents for round lots – the single best odd-lot price, with the associated size, that is available across all the exchanges. The BOLO would not itself affect the NBBO, but it would provide a new tool to help participants identify the best displayed price that is available for each NMS security.
Issues and Alternatives
Some in the industry have suggested changes to the tiering of the round lot sizes, but it is extremely unlikely the Commission will look to redo what it has already recently adopted. We expect the main questions will revolve around the timing of implementing the changes and whether and how to stage them. The SEC proposed to require the changes be completed 90 days after they adopt a final rule. Questions are likely to be raised about whether more time is needed.
One alternative the SEC identified would be to accelerate the round lot changes but not the implementation of the odd-lot information. One argument is that doing so will reduce short-term implementation costs for the “retiring” SIPs. Another is that the companies running the existing SIPs may themselves seek to register as competing consolidators, and adding the new odd-lot data to existing feeds could give them a “head start” over potential competitors down the road.
A counter to that is that the odd-lot data will provide short-term added value to consolidated data that many participants have been calling for, and companies running the SIPs are unlikely to gain any long-term competitive advantage from earlier implementation, if they decide to compete. The Commission also noted a connection to any reduced tick sizes that may be adopted, the topic of our first post. For stocks that have lower tick sizes, less liquidity will be shown at the NBBO, meaning there will be a greater need for odd-lot data, so deferring implementation would leave participants relying on SIP data with less transparency. The round lot changes would have a similar effect, because the NBBO for higher-priced stocks will show less liquidity in dollar terms. People who want to see the full picture will need the odd-lot component.
Each of the various aspects of the SEC’s proposed update to Regulation NMS is aimed directly at substantial, in some cases profound, market changes. These include enormous increases in the volume and speed of trading, a decline in displayed trading, changes in competition among market centers, the growing complexity of exchange fees and their different impact on classes of participants, and the barriers that have arisen separating, in some respects, investors from each other.
Each stakeholder group and constituency will debate the merits and details of the individual proposals.
Vigorous debate is necessary, but it should be clear that regulation cannot stay static for long, if the U.S. is to retain its leadership position in global markets and move closer to the goal of a national market system that works for all investors.