Market Structure Issues
by William H. Donaldson
Chairman, U.S. Securities & Exchange Commission
Before the Subcommittee on Securities and Investment
of the U.S. Senate Committee on Banking, Housing, and Urban Affairs
October 15, 2003
Good morning Chairman Enzi, Ranking Member Dodd, and Members of the Subcommittee. I am delighted to be here to discuss some of the significant market structure issues that we are facing in the U.S. equities market today.
In the coming months, the Commission will be focusing with increased intensity on the structure of the U.S. equities markets, with particular regard to their fairness and efficiency. As you know, Congress formally directed the Commission to address market structure when it enacted the Securities Acts Amendments of 1975. That legislation instructed the SEC to facilitate the creation of a national market system for securities that would maintain fair and orderly markets, and tie together all buying and selling interest so that investors would have the opportunity for the best possible execution of their orders, regardless of where in the system they originate.
Congress specified five key objectives of the national market system: (1) economically efficient executions of securities transactions; (2) fair competition among markets and securities firms; (3) the availability of market information to investors; (4) execution of orders in the best market; and (5) direct interaction among investor orders. To achieve these objectives, Congress recognized that communication systems, particularly those designed to disseminate market data, would form the heart of the national market system. Rather than attempt to dictate the specific elements of U.S. market structure, however, Congress chose to rely on an approach designed to provide maximum flexibility to the Commission and the securities industry in its development.
The 1975 Amendments to the Exchange Act created a framework for fostering transparency, interconnectivity, and competition in our securities markets. As a result, today, equity market centers compete with one another in an environment where quotes and transaction prices are widely available to all market participants. Direct and indirect linkages among competing market centers help ensure that brokers can access the best quotes available in the market for their customers. Market centers (including exchange markets, over-the-counter market makers, and alternative trading systems) have an incentive to offer improvements in execution quality and to reduce trading costs in order to attract order flow away from other market centers. This competition among market centers encourages ongoing innovation and the use of new technology. Within all existing registered exchanges and a number of other markets, investor orders have the possibility of interacting directly without the intervention of intermediaries. This furthers Congress's fifth objective - direct interaction of customer orders - allowing investors to obtain executions at better prices than otherwise would be available.
The national market system has worked remarkably well for the past quarter century. And in recent years it has become increasingly efficient. At the same time, this very efficiency, arising from technological and other market developments, has put strains on existing national market structures. One significant change has been the proliferation of new electronic markets, such as ECNs, that offer fast executions and have spurred competition among market centers, but at the same time exacerbated concerns about market fragmentation, the feasibility of integrating different market models into the national market system, and maintaining a level regulatory playing field among functionally-equivalent market participants. The implementation of decimal pricing in 2001, and the concurrent move to a minimum tick of one penny in the equity markets, has narrowed spreads and enhanced the efficiency of the price discovery process, but at the same time reduced the liquidity available at each price point, made it easier to step ahead of limit orders, and placed economic strains on the dealer business. Decimal pricing has also put a premium on swift access to displayed prices so investors can quickly reach these smaller quotes before they change. The trend toward demutualization of exchanges, and their conversion to for-profit enterprises, has heightened concerns about the inherent tensions in the self-regulatory model, in particular the concern that the funding and vigor of the regulatory function might be sacrificed in favor of delivering returns to shareholders.
The issues surrounding intermarket access provide a good example of some of the strains impacting U.S. market structure in recent years. In a system with many competing market centers and pools of liquidity, participants clearly need to know what the best prices are and where they are available. But this information is of little use in the absence of effective access to the market centers with the best prices. Implementing market access, however, has raised a number of difficult issues. Offering access to one's market to competitors can conflict with the core business strategy and commercial self-interest of a market. Over the years, markets have sought to maintain strict control over access and often have erected barriers to achieve this objective. These barriers historically have taken the form of direct bans, restrictive membership requirements, discriminatory execution priorities, fees, and information restrictions. Finally, even setting aside intentional barriers to access, significant practical difficulties must be overcome to ensure the availability of access in an environment where scores of separate market centers - floor-based and electronic, both fast and slow - may be actively quoting and trading a security. The existing compulsory market-to-market linkage in stocks - the Intermarket Trading System (ITS) - applies only to NYSE and Amex stocks and, in the view of many, has been less than successful in overcoming these obstacles to providing effective intermarket access.
Over the last several years, the Commission has taken a number of steps to address concerns facing our national market system. In the Order Handling Rules and Regulation ATS, for example, the Commission broadened the class of market centers required to make their quotations and orders publicly accessible. In doing so, it sought to redefine the idea of an exchange to include not just traditional exchanges, but also trading systems where orders interact according to specified trading rules. The Commission also adopted rules to improve the disclosure by market centers of execution quality data, and the disclosure by broker-dealers of their order routing practices, in order to enable investors to "comparison shop" among the myriad market centers, and to stimulate competition on the basis of execution quality.
In addition, the Commission has developed ideas and solicited public comment on some of the more difficult market structure issues, such as the regulation of market data fees and revenues, the fragmentation of the U.S. securities markets, and the regulation of exchanges. A federal advisory committee also was convened to address market data concerns, and last year the Commission held public hearings on the full range of market structure issues.
There is no doubt that there are issues regarding our national market system that call for our attention. In my view, several aspects of equity market structure raise pressing questions. These include: (1) the implications of differences among markets in the means by which their quotes may be accessed by non-members and of access of transaction fees that are not included in displayed quotations ; (2) the role of trade-through rules in intermarket trading for very different types of markets and systems; (3) the manner in which market data is consolidated and distributed, and the resulting revenues allocated among the markets; (4) whether a mixed dealer and auction market such as Nasdaq should be allowed to register as a for-profit exchange; (5) whether the fragmentation of markets that results from competition is reducing the effectiveness of regulatory processes; and (6) the effectiveness of the current self-regulatory system for the securities markets.
That said, I firmly believe our system of multiple, competing markets - on balance - has worked remarkably well. We have the world's most competitive and efficient markets. Competition among markets has fostered innovation and led to the creation of a variety of trading platforms designed to meet the needs of different types of investors. New entrants, particularly those with fully electronic platforms, keep the pressure on established markets to innovate. However, new entrants also challenge our existing infrastructure, much of which was created in the 1970s before the dramatic advancements in technology.
As has always been the case in our competing markets model, our challenge as regulators is to ensure fair and efficient markets through a balance of competition and regulation. Fair and efficient markets, of course, are the key goals of securities market regulation. But fairness and efficiency are at least superficially different concepts, creating tensions in our regulatory mandate. Fairness suggests the use of regulation to ensure against unfair results. Efficiency, on the other hand, suggests reliance on free markets and competitive forces to achieve an efficient result, which may not necessarily be a "fair" one. Regulation and competition do not necessarily conflict, as regulation often seeks to remove barriers to competition or promote efficiency. In other cases, there will be a tension between regulation and competition. Striking the appropriate balance is the responsibility of the Commission.
The optimal equity market structure, in my view, is based on several fundamental principles. First, I believe we should seek to achieve the benefits of competition while countering the negative effects of fragmentation from trading in multiple markets, through widely available market data, ready access among markets, price protection principles, and best execution standards.
Second, to the greatest extent possible, I believe we should let market forces determine outcomes by seeking to have the marketplace, rather than the government or its regulators, choose the "winners" and "losers." We must seek to provide a level playing field in which all markets can compete fairly and aggressively. That said, regulation is necessary in certain situations, such as when an exchange exercises market power, or when externalities such as principal/agent conflicts obstruct otherwise competitive outcomes. Regulation is also appropriate when its benefits to the marketplace exceed its costs and reduce market frictions, such as when settlement date standards or quoting conventions are established.
Finally, I believe that market transparency, fairness, and integrity are key to the strength of our marketplace. These fundamental concepts underpin the Commission's approach to regulation, and contribute substantially to investor confidence in our markets.
With these general principles in mind, I would like to focus the remainder of my testimony on four key areas of the Commission's market structure initiative: (1) access to markets; (2) market data; (3) the self-regulatory model; and (4) the nature of a securities exchange.
I. Access to Markets
A. Fair Access
In our modern-day marketplace for securities, the New York Stock Exchange, Nasdaq, the American Stock Exchange, the regional exchanges, and numerous electronic communications networks, all compete with each other to offer the deepest pools of liquidity to investors at the very best prices. I believe that the Commission must resist the temptation to force these diverse securities markets to mimic each other, but rather to encourage them to compete over their differences within a single, robust, national system. In the end, there is little doubt in my mind that investors benefit from markets that compete, so long as the competition is truly fair.
With that in mind, a significant market structure issue on the Commission's agenda is making sure that access between markets is as fair and as efficient as it can be.
If best execution is to be achieved in an environment characterized by multiple competing markets, broker-dealers must be able to identify the location of the best available prices and obtain access to those prices routinely and efficiently. In contrast, a market center that is inaccessible does little to promote efficiency and fairness in the marketplace.
Most brokers send orders directly to the market that they expect will provide their orders best execution most of the time, and most of these orders are executed in the market that receives them. At times, however, the best price at that moment may be in another market. And traders in one market may need to access prices in another market to keep prices in line. For these reasons, markets need easy access to each other, either directly or indirectly through brokers.
The Commission's approval last year of the NASD's Alternative Display Facility pilot program has highlighted this issue. Rather than obtaining access through "hard" linkages directly between markets, in the way that competing markets can access the New York Stock Exchange, in the Alternative Display Facility competing market centers obtain access to each other directly through privately negotiated access agreements and indirectly through subscribers. The Commission will be evaluating this decentralized access approach to determine whether, as a practical matter, it would be an appropriate model for the national market system, and thus could be applied to other market centers.
B. Access Fees
Access fees charged to reach a quote create another difficult market structure problem. Some markets charge varied per-share transaction fees for access to their quotes. Therefore, a displayed price may represent the true price that a customer will pay or it may represent only a base price to which an undisclosed access fee will later be added.
These pricing disparities can impede access between competing markets, raise trading costs, and create confusion about the true quoted prices. The absence of a uniform quoting convention across all markets also raises the incidence of locked and crossed quotations. To ensure real access to public quotes between competing markets, it is important that these quotes be accessible to other market participants on clear and fair terms.
I should also mention that, because access fees have gradually shrunk to less than one cent per share in most markets, the imposition of the fees results in de facto subpenny pricing. Indeed, many market participants have suggested that these access fees have precipitated trading in subpennies, thus magnifying the strains caused by the move to decimal pricing. The Commission intends to work closely with the industry and investors to find appropriate solutions to the challenges raised by access fees and subpenny pricing. Whatever solution the Commission decides to adopt, we must assure that access fees will not function as a tollbooth that snarls traffic along the national market system.
C. Price Protection
As part of our examination of inter-market linkages, we also intend to re-assess the question of intermarket trade-throughs, which occur when orders are executed in one market at prices inferior to the prices disseminated on another market. Today's highly competitive securities markets include fully electronic markets that provide swift automatic execution of customer orders, as well as traditional floor-based markets that execute orders through human interaction. Although a market participant that desires an opportunity for price improvement may prefer that its order be routed to a floor exchange for execution, an investor who values speed and certainty of order execution over a marginally higher price may find such a delay intolerable. Accordingly, the challenge before the Commission is to devise standards that allow faster markets and slower markets to thrive within a single system of interconnected markets, while at the same time providing order executions to customers that display prices and for those customers who desire the best price on their orders.
II. Market Data
Another significant market structure challenge facing the Commission involves the collection and reporting of trading information and influence of the resulting revenues on market structure. Our existing market data system has strengthened the U.S. equity markets and has assured that investors have real-time access to accurate, reliable, and affordable information from all significant U.S. market centers. And yet the increasing number and diversity of U.S. market centers, has fueled demands for modernizing the current market data structure. Despite the sweeping changes that have taken place in the markets over the past 30 years, the structure for market data, including the collection and dissemination of a market's best bid and offer, the national best bid and offer, trading volume statistics, and last-trade prices, has changed very little.
The Commission recognizes that market data revenue is very important to our markets. Indeed, in recent years, self-regulatory organizations have drawn as much as 45% of their total revenues from market data revenue. In 2001, the Commission convened a panel of experts, chaired by Dean Joel Seligman, that looked into the structure of our market data system, as well as the compensation that markets have been receiving for their market data. The Seligman Committee noted that under the current system, securities information processors distribute market data revenues to self-regulatory organizations based primarily on each self-regulatory organization's reported trade volume. This compensation scheme has created a financial incentive for self-regulatory organizations to report as many trades as possible. As a result, markets are vying for ECNs and market makers to report their trades through them, as this allows markets to tap more deeply into the pool of available market data revenue and to rebate substantial portions of the additional revenue to the entity reporting the trade.
Significantly, in 2002 the Commission determined that programs for rebating market-data fee proceeds to market participants were creating incentives for traders to engage in transactions with no economic purpose other than to increase the amount of the market data revenues that they received. In this regard, the Commission abrogated several more extreme proposals to extend rebates of market data revenues to market participants, to allow more time for consideration of these issues.
It is my belief that market centers should be rewarded for providing better services. The recent developments call into question whether the current method of distributing market data revenue creates appropriate economic incentives, and whether it furthers the goal of rewarding markets that make valuable contributions to the market data being disseminated.
III. The Self-Regulatory Model
Another matter of great importance is the effectiveness of the self-regulatory system of securities markets. Recently, a number of concerns have been raised about the current state of self-regulation, including SRO conflicts of interest, SRO governance, and inefficiencies in self-regulation.
Congress and the Commission have long recognized that self-regulation has both benefits and weaknesses. The principle of self-regulation is based on the idea that regulation can best be done as close as possible to the regulated activity. However, an SRO that operates a market has an inherent conflict of interest between its roles as a market and as a regulator. I believe that the Commission must continue its work in ensuring that SROs vigorously fulfill their obligation to enforce their rules and the federal securities laws and rules. The advent of for-profit, shareholder-owned exchanges creates additional issues, including ensuring that self-regulatory obligations do not take a back seat to the interests of shareholders. The challenge for the Commission and the SROs is to ensure that as the securities markets grow more competitive, the SROs continue to dedicate their energies and resources to surveillance and enforcement. We also must prevent fragmentation of trading from creating gaps in SRO oversight of the markets.
As part of our review of the self-regulatory structure, I believe the Commission must thoroughly review the SROs' governance practices. Recent events at the New York Stock Exchange point to the need for this review. SROs play a critical role as standard setters for sound governance practices. Just as SROs have demanded that their listed companies strengthen their governance practices, we must demand that, at a minimum, SROs match the standards they set for listed companies. There are several topics that merit our consideration, including board composition and independence of directors; the independence and function of key Board committees; the transparency of the SRO's decision-making process; and the diligence and competence required of Board and committee members and ensuring their focus on the adequacy of regulation.
These are critical issues facing the SROs and the Commission. I am committed to ensuring that our system of self-regulation continues to serve as an effective and efficient means of overseeing our securities markets.
IV. Exchange Criteria
The last topic that I would like to touch upon is what it means to be registered as a national securities exchange. All currently registered exchanges have a limit order book in which better-priced orders take precedence. But a mandatory order book system is not easily reconciled with a dealer model, such as the Nasdaq stock market, in which there is no central limit order book.
I spoke earlier about the merits of price protection across markets. Nasdaq's application to register as an exchange places squarely before the Commission the issue of whether price protection within a market is a requirement of exchange registration. One issue is customer expectations. I suspect that customers generally expect their better priced orders to be protected within an exchange.
We do not expect all exchanges to be identical, much less to replicate any market's faults. Yet until now all exchanges have given their limit orders priority throughout their marketplace. If the Commission were to approve Nasdaq's application, other exchanges would likely seek to eliminate intra-market price priority from their rules. As a result, the protection of limit orders within markets would decrease. For this reason, Nasdaq's exchange application raises market structure issues that transcend the particular question of whether Nasdaq, or any other particular market, should be registered as an exchange.
In conclusion, I would like to reiterate that the market structure challenges that I have discussed today may shape the national market system for years to come. I look forward to continued input from this subcommittee on these important matters.
Thank you again for inviting me to speak on behalf of the Commission. I would be happy to answer any questions that you may have.