Speech by SEC Staff:
"Electric Money: Trading in the 21st Century"
Erik R. Sirri
Director, Division of Trading and Markets
U.S. Securities and Exchange Commission
Financial Times Conference
New York, N.Y.
November 14, 2007
Thank you for the invitation to talk today about U.S. equity market structure, particularly Regulation NMS, its implementation by the securities industry, and its effect thus far on trading. Before I begin my remarks, let me remind you that the views I express are my own and not necessarily the views of the U.S. Securities and Exchange Commission, the individual Commissioners, or my colleagues on the Commission staff.1 On October 8th, we reached the official end of phased-in compliance with Reg NMS. Now seems like a good time to step back for a few minutes, revisit the Commission's big picture policy goals for Reg NMS, and make at least a preliminary attempt to assess how things are going.
When the Commission voted to adopt Reg NMS in April 2005, it said that its overarching goal was to modernize and strengthen the national market system for trading U.S.-listed equities. The first question you might ask is — what is a national market system and why does the U.S. have one? A national market system is made up of multiple markets trading the same stocks and competing for order flow in those stocks. A national market "system" should be contrasted with a single dominant market. In the U.S., investors are fortunate to have a lot of choices on where to route their orders for execution. These include exchanges and ECNs that display quotes, as well as all sorts of different flavors of ATSs that offer a variety of trading services. In this respect, the U.S. stands in sharp contrast with other major equity trading countries, where a single exchange typically dominates public trading in its listed stocks.
In the U.S., Congress specifically directed the SEC to promote a national market system with competing markets. This type of competition among markets has some very important benefits. It creates incentives for markets to innovate in the types of trading services and technologies they offer their customers. It also imposes pressure on markets to keep trading fees low.
One potential problem with competition among markets is fragmentation — when trading interest in any particular stock is split up among many different trading venues, the quality of price discovery and efficiency of trading in those stocks can be impaired.
To address this problem, Congress also directed the SEC to assure that the multiple competing markets in the U.S. are effectively linked together in a unified national market system. The basic philosophy of the U.S. approach to market structure policy can be characterized as wanting to have your cake and eat it too. We want the benefits of competition, but without the adverse effects of fragmentation. In other words, promote competition — address fragmentation.
That is what the new NMS rules are designed to do. They are a big package of rules, but, from a trading standpoint, I want to highlight four very important elements — trade-through protection, private linkages, intermarket sweep orders, and the self-help remedy.
Clearly, the most significant part of Reg NMS is Rule 611 and its protection of displayed quotes against trade-throughs. Its key provision is that it only protects automated quotes that are immediately accessible on non-discriminatory terms. Reg NMS eliminated any competitive advantage manual markets might have had under the old ITS trade-through rule.
An important policy goal of trade-through protection is to provide an incentive for the public display of limit orders. Since the initiation of decimal trading in 2001, there have been a lot of reasons for market participants not to display their trading interest. One in particular is the ease with which other market participants can step ahead of displayed trading interest by an economically small amount. These disincentives for display have given rise to some of the dark pools and undisplayed liquidity that I will discuss a little later. When the Commission adopted Reg NMS, it emphasized the importance of displayed liquidity in promoting efficient equity markets. Displayed liquidity provides the price discovery that is the starting point for all types of trading, both on and off the public markets. Rule 611 was the Commission's attempt to do something positive to encourage displayed liquidity.
A common misunderstanding of Reg NMS is that it requires the display of liquidity and thereby drives trading off an exchange. This is not true. Attempting to force market participants to display their trading interest would be both fruitless and counter-productive. Instead, Reg NMS gives market participants a choice of whether and how much trading interest they want to display. Rule 611 simply tries to provide at least some positive incentive for display.
The second important element of Reg NMS is the private linkage approach to connecting markets, as opposed to the old ITS linkage system that was centralized and hard-wired. The new approach allows much more flexibility for market participants to fashion linkages in ways that make sense for their individual circumstances and business models.
A third important element of Reg NMS is the power that it gives order routers to control the handling and execution of their own orders in compliance with its rules. Rule 611 does this primarily through the intermarket sweep order, or ISO. An ISO frees the receiving market to execute the order immediately, without regard to better-priced quotes at other markets. The router of the ISO effectively assumes responsibility for trade-through compliance by routing additional ISOs to execute against any other better-priced quotes. ISOs have proven to be very popular with order routers. Early statistics indicate that as many as 50% of orders on some markets now are ISOs. It is important to recognize that while ISOs are an exception to Rule 611, they are most definitely not an exception to the Rule's policy of protecting limit orders. Every ISO identifier on a trade report means that quotes were protected and not traded through.
The fourth important element of Reg NMS is the self-help remedy it provides to deal with markets that experience systems problems. One of the most valid theoretical criticisms of a trade-through rule is that it could force market participants to route orders to a venue that cannot provide an immediate response. The self-help remedy frees market participants from depending on the problem market to do the right thing when it experiences systems problems and stop displaying automated quotes. Instead, if there is a reasonable basis to believe a market has a problem, market participants can adopt reasonable policies and procedures, stop routing ISOs to the problem market, and trade through its quotes.
Those are four important policy elements of the Reg NMS approach to equity market structure. With respect to options market structure, it is worth noting that the ISE and NYSE Arca submitted a proposal to the Commission last month to replace the existing options market linkage plan with a new plan that is based on Reg NMS principles, including trade-through protection, private linkages, ISOs, and a self-help remedy.
Now that Reg NMS implementation is complete, what can be said about its effects thus far? It is very interesting to think about how the equity markets compare today with where they were in April 2005 when the Commission voted to adopt Reg NMS. A great deal has happened and continues to happen. I'll briefly touch on three types of Reg NMS effects — the effect on automated trading, the effect on competition among markets, and the effect on market efficiency.
One goal of Reg NMS was to promote automated trading and particularly to eliminate any regulatory advantage for manual trading that existed under the old ITS rules. This goal certainly seems to have been achieved. Nearly every exchange has adopted a new, more automated trading system capable of displaying fast quotes that are entitled to trade-through protection. The private linkage approach has promoted much greater connectivity throughout the securities industry, both among exchanges and other trading venues, and among brokers-dealers and trading venues. In addition, the central processors for consolidated market data, SIAC and Nasdaq, have expanded their capacity. SIAC, in particular, has greatly reduced latency in collecting data from the SROs, consolidating the data, and distributing the data to the public. Finally, securities firms have revamped their order-routing and execution systems to comply with the new NMS requirements.
With all these rule changes and new systems, there was understandably a fair amount of concern in the securities industry about the potential for serious failures that might shut down the equity markets. So far, though, the new systems and procedures have held up remarkably well, even in the face of the sharp volume and volatility spikes that occurred this summer. This positive result is a credit to the expertise and hard work of those throughout the securities industry who were responsible for implementing Reg NMS.
The effects of Reg NMS thus far on competition have been nothing short of dramatic. Competition for order flow is extraordinarily fierce. There now is no single market with even a 50% share of matched order flow in its listed stocks. Instead, nearly all trading venues are competing to attract order flow in the whole range of NMS stocks, no matter the particular exchange where they are listed.
Probably most notably, NYSE Hybrid market share in NYSE-listed stocks is down from 78% in 2005 to 41% in October 2007. Even though NYSE Hybrid's market share percentage has nearly been halved, it actually is trading more shares in absolute terms because overall trading volume in NYSE-listed stocks has increased so much. Even prior to the volume surge this summer, trading volume in NYSE stocks had been increasing steadily over the last year, particularly when contrasted with volume in Nasdaq stocks.
Increased trading volume in NYSE-listed stocks was one of the most widely predicted effects of Reg NMS. I think to some extent NYSE stocks were like an undiscovered territory for electronic traders. Prior to the NYSE automating its quote, it was very difficult to trade NYSE stocks electronically. As a matter of necessity rather than choice, electronic traders were pretty much limited to Nasdaq stocks and a few ETFs. As of last fall when NYSE Hybrid fully automated its quotes for the first time, NYSE stocks became much more amenable to electronic trading. Moreover, many more NYSE stocks than Nasdaq stocks are large cap stocks. For example, NYSE stocks make up about 80% of the Russell 1000, which itself constitutes about 92% of the market cap of the U.S. equity market. The trading volume statistics seem to indicate that many electronic traders have shifted at least some of their attention from Nasdaq stocks to NYSE stocks.
The statistics on matched market shares across all U.S-listed stocks show that competition among exchanges is remarkably even. Indeed, non-exchange trading venues collectively have a greater market share than any single U.S. exchange. In October 2007, for example, non-exchange venues had a 30.0% share, Nasdaq had a 29.7% share, NYSE Hybrid had a 21.5% share, and NYSE Arca had a 15.9% share.
The single largest share of non-exchange volume is attributable to the BATS ECN — a relatively new competitor that entered the market after Reg NMS adoption. Another ECN that publicly displays quotes also has attracted a large share of non-exchange trading volume.
Of course, another substantial part of the non-exchange trading volume is attributable to the assortment of dark pools that have sprung up in recent years and that have received so much publicity. What is less recognized is the extent to which the public transparent markets that display quotes also offer undisplayed, or dark, liquidity. Most exchanges and ECNs offer order types to their customers that allow either all or part of the order to be undisplayed. Nasdaq, for example, recently said that approximately 15 to 18 percent of its executions are against undisplayed liquidity. Nasdaq also now offers a historical data product that shows how much undisplayed size was available for execution on its order book on a T+10 basis. Undisplayed size can be quite substantial and can even equal displayed liquidity at the inside quote, particularly in stocks outside the very top tier of trading volume where displayed liquidity may not be very large. As a result, it is likely that the transparent public exchanges with displayed quotes also offer the largest dark pools of liquidity in the U.S. equity markets.
The final type of NMS effect I will mention is market efficiency. SEC economists are examining this issue, likely with regard to the full range of statistical measures of efficiency. Up through the first and most important phase of NMS trading, their task was relatively clear. March of this year was the start of the Trading Phase of Reg NMS. At that time, it went into effect for the public quoting markets — exchanges and ECNs — that represent about 80% of trading volume. The next two phases, however, which applied to non-displayed OTC trading, did not occur until July and August — right in the thick of sub-prime mortgage problems and the return of fundamental volatility with a vengeance. As can be expected, the difficult trading conditions this summer resulted in major changes in market dynamics.
I will leave it to the Commission's Office of Economic Analysis to figure out how to disentangle the effects of the sub-prime crisis and the last two phases of Reg NMS compliance. Instead, I will focus on the first phase when NMS went into effect for 80% of the market and ask a basic question — were the markets better or worse off in the time since Reg NMS was adopted? In particular, I'll compare trading in April 2005, when the Commission voted to adopt Reg NMS, with trading two years later in April 2007, the first full month after the start of the NMS Trading Phase in March.
The place to begin in thinking about market efficiency is to recognize the enormous scale of trading in the U.S.-listed equity markets. In 2006, their dollar volume was almost 40 trillion dollars and likely will be significantly greater in 2007. With trading volume on this scale, even small per share improvements in efficiency can generate some very large annual dollar savings in transaction costs for investors. To illustrate, an improvement of a single basis point on 40 trillion dollars of trading adds up to 4 billion dollars in reduced transaction costs.
I'll mention two widely accessible measures of market quality — the publicly available Rule 605 execution quality statistics on execution speed and spreads, and statistics on the consolidated displayed depth at the inside quotes that are publicly available on the ArcaVision website. The statistics I'll mention should be considered rough and ready, back of the envelope, calculations. I offer them primarily to provoke thought about the commonly heard impression that fragmentation has harmed the markets since Reg NMS and led to reduced liquidity.
First, when the Rule 605 statistics in April 2007 are compared with those in April 2005, execution speed unsurprisingly had improved considerably, primarily for NYSE stocks. But quoted and effective spreads also had improved significantly across all stocks. Similarly, consolidated displayed depth at the inside quotes increased substantially from April 2005 to April 2007. For example, average consolidated depth at the inside quotes in Russell 1000 stocks had increased by more than 50%. And this displayed liquidity does not include the undisplayed liquidity at the inside that I mentioned earlier and that seems to be becoming a more frequently used trading tool by market participants.
To sum up, a tentative answer to the competition versus fragmentation policy question may be this. Although trading volume may now be more dispersed among different trading venues than it was in 2005 as a result of increased competition, the great majority of trading volume occurs at venues that publicly display quotes and whose liquidity, even when undisplayed, is immediately accessible to incoming orders. This transparency and accessibility allows liquidity to be consolidated on traders' screens rather than on any single trading floor or any single order book. If you can see liquidity, or reasonably know where it is likely to be, and access that liquidity immediately, then the market is not truly fragmented. It is a new world, though, and trading in such a virtual market requires a new set of skills and technologies than under the old, pre-NMS market structure.