William H. Donaldson
Chairman, U.S. Securities and Exchange Commission
Before the U.S. Senate Committee on Banking, Housing, and Urban Affairs
July 21, 2004
Chairman Shelby, Ranking Member Sarbanes, and Members of the Committee:
Thank you for inviting me to testify today on proposals to modernize the national market system for equity securities. I welcome your continuing interest in an issue of such vital importance to investors and the economy. The national market system encompasses the stocks of more than 5000 companies, which collectively represent more than $15 trillion in market capitalization. The Commission is committed to promoting the fairest and most efficient markets possible for these securities.
Since I appeared before you last October to discuss the state of the national market system, the Commission has made a great deal of progress. In February, we published for public comment Regulation NMS — a broad set of proposals designed to improve the regulatory structure of the U.S. equity markets. In April, we held public hearings on the proposals, then followed up in May by publishing a supplemental request for comment to reflect a number of important matters discussed at the public hearing. In addition, the comment period was extended until June 30 to give the public ample opportunity to prepare their views.
Consequently, this Committee’s hearings on the national market system are particularly timely. Your consultation and oversight will make an indispensable contribution to the Commission’s efforts as it moves forward in the rulemaking process. With your help, I am confident that we will succeed in our efforts to assure that the equity markets continue to meet the needs of investors and public companies.
I. National Market System Principles
The Regulation NMS proposals encompass four substantive areas — trade-throughs, market access, sub-penny quoting, and market data. Today, I intend to give an overview of the proposals, as well as offer a few thoughts on the Commission’s road forward. First, however, I would like to take a broader view of the market structure issues facing the Commission, as well as the policy objectives that the proposals are intended to achieve.
When assessing the current state of the national market system, the starting point is to recognize just how well it works overall. The system needs to be modernized, but it is far from broken. The U.S. equity markets have never been more fair and efficient for such a broad spectrum of investors than they are today. Since the national market system was created, investor trading costs have steadily declined. Not surprisingly, as trading costs have declined, the volume of trading has climbed inexorably upward. Indeed, our markets now routinely handle trading volumes that would have been nearly unimaginable just a decade ago. These are telling indicators of markets that are vibrant and healthy.
With all this success, inevitably, come problems. In the past few years, in particular, a remarkable confluence of forces has strained the existing components of the national market system. These forces have included technology advances, of course, but also the arrival of entirely new securities products and trading strategies. These include derivative products such as exchange-traded funds, which generate enormous trading volumes, as well as program trading in large baskets of stocks and statistical arbitrage trading. Moreover, the commencement of decimal trading in 2001 further transformed the equity markets. The number of quote updates exploded, and the quoted size at any particular price level dropped. Investors adopted new tactics to deal with the changed trading environment and found that they needed new trading tools to implement these tactics. In particular, investors have adopted automated order routing strategies that require exceptionally fast execution and response times from the markets. Finally, a variety of new, electronic markets have arisen that offer innovative trading mechanisms designed to meet the needs of those using the new securities products and trading strategies.
The proliferation of fast, electronic markets simultaneously trading the same stocks as slower, manual floor-based exchanges has complicated the task of making sure that an investor order receives best execution. The Commission’s challenge is to craft rules that reconcile different trading models without sacrificing the fundamental principle of assuring the best execution for investor orders. I believe this creates the conditions under which an investor can achieve the best available price.
In sum, the national market system needs to be modernized, not because it has failed investors, but because it has been so successful in promoting growth, efficiency, and innovation that many of its old rules now are outdated. Identifying and improving these outdated rules is the ultimate goal of the Regulation NMS proposals. To this end, the Commission has engaged in an exceptionally open and interactive process. It has actively sought out the views of a wide range of market participants. There are few areas of securities regulation in which the considered views of practitioners are more needed than market structure. When the Commission published the Regulation NMS proposals for public comment, it fully expected that the proposals would be revised and improved after hearing the views of commenters. Indeed, the public hearing on Regulation NMS in April produced such valuable suggestions for improvements that the Commission published a supplemental request for comment to incorporate these suggestions. This process is continuing. I fully expect that our review of the comment letters will promote additional improvements in the proposals as the Commission moves forward in the rulemaking process.
This process will be guided by those fundamental principles for the national market system that were established by Congress in 1975 and have guided the Commission over the years. Although the particular rules and facilities that implement these principles may be in need of updating, I believe that the principles themselves remain as valid as ever. In particular, the Commission has always sought to achieve the benefits of competition, while countering the negative effects of fragmentation from trading in multiple markets. The national market system has promoted the wide availability of market data so that investors can determine the best prices, ready access among markets to obtain those prices, protection of investor limit orders, and the duty of brokers to obtain best execution for their customer orders.
I particularly want to emphasize the importance of price protection and encouraging the display of investor limit orders. These orders typically represent the best displayed prices in a stock. They therefore are a critical source of public price discovery that is essential to the efficient operation of the markets. Competition among markets is a vital aspect of efficient markets, but we must also assure vigorous competition among the orders of buyers and sellers in a stock. If investor limit orders are neglected and trades occur at inferior prices without good reason, I believe that it harms both the particular investors involved and the integrity of the markets as a whole. Small investors justifiably may not understand why their order is by-passed by trading in other markets. But many of the largest institutional investors also have stressed to the Commission that they believe enhanced protection of investor limit orders is one of the weaknesses in the current national market system that needs to be addressed. Each of the Regulation NMS proposals is intended in large part to achieve this vital objective.
II. The Regulation NMS Proposals
The Commission is in the midst of extensive rulemaking process for its Regulation NMS proposals. This process has included a public hearing, a supplemental request for comment to reflect the topics raised at the hearing, and the extension of an already long comment period. Even prior to publishing the specific rule proposals in February, the Commission repeatedly sought the views of market participants and the public. It held multiple public hearings and roundtables, established an Advisory Committee on market data, and issued four concept releases on a variety of NMS issues. The Commission used the information and data gathered by these steps to form the basis of its NMS proposals. These proposals encompass four substantive areas — trade-throughs, market access, sub-penny quoting, and market data.
A. Trade-Through Proposal
The trade-through proposal has thus far garnered the most attention. It would, for the first time, establish a uniform trade-through rule for all NMS stocks. The rule would protect the best displayed quotes in a stock from trades at inferior prices. It is intended to encourage the placement of investor limit orders, which often provide the best displayed prices, as well as to promote the best execution of investor orders. As a uniform rule, it would cover both exchange-listed stocks, which are governed by an existing SRO trade-through rule, and Nasdaq stocks, which have never been subject to a trade-through rule.
I will briefly review the proposal’s application to each type of stock, but first want to emphasize a more general point. Some have debated the trade-through proposal as if it were a kind of referendum on the merits of exchange auction markets versus fully electronic markets. I do not approach the issue in these narrow terms. The Commission’s goal is neither to reward nor punish any particular type of market mechanism. Instead, the trade-through proposal is intended to address potential weaknesses in — and thereby improve — the markets for both exchange-listed and Nasdaq stocks.
For exchange-listed stocks, the proposal would address a serious weakness in the existing ITS trade-through rule, which was established by the exchanges and approved by the Commission. This weakness is caused by the disparate degree of access to quotes displayed by manual markets and those displayed by automated markets. Manual markets — those with traditional trading floors on which human beings effect trades — generally take from 10 to 30 seconds to respond to incoming orders. Automated markets respond much more quickly. Notice that I use the word “respond.” Some have confused speed of response with certainty of execution. Neither manual nor automated markets guarantee the execution of orders at their best displayed quotes. Such quotes, for example, may already have been executed against by previous incoming orders or have been withdrawn prior to order arrival. Indeed, according to Rule 11Ac1-5 execution quality reports from one active market center, even the fastest electronic markets generally have fill rates for marketable orders of approximately 60% to 75% in the most actively traded stocks.
Consequently, the problem that the trade-through proposal is intended to address for exchange-listed stocks is not differing certainty of execution, but differing speed of response and differing execution prices. With automated quotes, investors can know in less than a second whether their order has been executed and can adjust their trading strategy accordingly. With manual quotes, many traders — including large institutional investors seeking to trade in significant size — have emphasized that they may ultimately receive an inferior price if the manual quote turns out not to be available after waiting 10 to 30 seconds for a response. In such cases, some would prefer to send an order immediately to an automated market displaying an inferior price, rather than accept the risk of a slower response from a manual market and perhaps an execution at an even worse price.
As proposed, the trade-through rule would address the disparity of access between automated and manual quotes by providing an exception that would allow automated markets to trade-through manual markets up to a specified amount. Among the most interesting developments at our public hearing in April, however, were statements by representatives of exchanges with traditional trading floors that they were committed in the coming months to establishing auto-execution facilities for access to their quotes. In addition, other hearing participants noted that existing order routing technologies were capable of reacting, on a quote-by-quote basis, to indications from a market that its quote was, or was not, accessible through automatic execution.
The capacity to identify individual quotes as automatic or manual potentially would give exchanges with trading floors the needed flexibility to integrate effectively a trading floor with an auto-execution facility. Rather than being lumped into a single regulatory classification as “fast” or “slow,” markets would be allowed to offer choices to investors. In those particular contexts when a manual execution on a trading floor potentially could offer the most value — such as to generate additional liquidity for a large order or to offer price improvement on an order — the exchange could identify its quote as manual, thereby affording a brief period for human beings to participate in an auction. Such a manual quote would not, however, be entitled to trade-through protection. Investors therefore would have the freedom to send orders to markets with worse quotes if they believed they could obtain better executions in those markets. As I will note later, some believe that this freedom to by-pass all manual quotes could eliminate the need for the proposed “opt-out” exception to the trade-through rule.
The concept of an exception for manual quotes appears promising. One of the primary purposes of the Commission’s supplemental request for comment in May was to give the public a full opportunity to express their views on this concept. Their views will play a vital role in determining the course of any final rulemaking.
Switching to the market for Nasdaq stocks, the practical effect of the trade-through proposal would be quite different. Nearly all quotes in Nasdaq stocks currently are accessible through automatic execution. Nasdaq stocks have never been covered by the ITS trade-through rule, and therefore have not been given trade-through protection. Commenters note, however, that brokers must fulfill their best execution obligations when routing customer orders. One of the most significant issues currently before the Commission is whether application of a trade-through rule to Nasdaq stocks would enhance protection of investor limit orders and promote improved public price discovery. We currently are evaluating the information and data submitted on this issue in the comment letters. An important part of our policy analysis will be to consider the potential benefits – as well as any negative impact – of trade-through protection for the more than 3000 stocks of companies that are listed on Nasdaq, not just the relatively small number of stocks in the very top tier of trading volume.
The final issue regarding the trade-through proposal that I would like to discuss is the proposed opt-out exception. This exception would allow one market to trade-through a superior price displayed on another market if the customer submitting an order consents to disregarding the superior price. One objective of the proposed exception was to give investors the freedom of choice to access quotes with inferior prices if they were not satisfied with the level of automation or service of a market displaying the best-priced quote. The proposing release noted, however, that the exception may be inconsistent with the principle of price protection for limit orders and could undermine investor confidence that their orders will receive the best available price.
In sum, the opt-out exception as proposed presents a conflict between policy objectives. On the one hand, we want to promote competition among markets and freedom of choice for investors in choosing where to route orders. On the other hand, investors who post limit orders establishing the best prices contribute greatly to public price discovery. But these investors may not be rewarded for this contribution if their orders are by-passed by trades at inferior prices in other markets.
The comment letters have expressed strongly-held views both for and against the opt-out exception. A critically important issue will be to determine how best to reconcile the legitimate desire of investors to send their orders to the most accessible quotes with the policy objective of protecting limit orders. Panelists at the public hearing, for example, suggested that, if the only quotes that received trade-through protection were those that were truly accessible through automatic execution facilities, there would be no need to by-pass such quotes with an opt-out. Other panelists believed that an opt-out exception would remain necessary to discipline markets that fail to maintain truly automatic execution facilities.
Clearly, the Commission must work hard to evaluate the views of commenters and reach the best possible solution to this difficult issue. We are committed to the policy objective of strengthening public price discovery, without interfering with efficient operation of the markets.
B. Market Access Proposal
Discussion of the proposed opt-out exception highlights the importance of the market access proposal. A trade-through requirement that orders be routed to the best available bid or offer would be entirely unworkable if all markets did not provide fair and efficient access to their quotes. The market access proposal is designed to achieve this goal in two ways. First, to establish linkages between markets, the proposal would require quoting markets to allow non-discriminatory access to their quotes through members or subscribers. The proposal therefore does not mandate that the markets establish inflexible, “hard” linkage facilities, such as the current ITS linkage facility. Second, the proposal would limit the fees that a market could charge for access to its quotes. Currently, some markets are permitted to charge these access fees and some are not. There also are significant variations in the amount of fees charged by different markets. The proposal would create a more level playing field. It also would establish an outer limit on the amount that could be charged to market participants when they route orders to other markets. This limit could be particularly important when orders are routed to meet regulatory responsibilities, such as to comply with a trade-through rule or to obtain best execution for customer orders.
Clearly, the proposed fee limitation is the most controversial aspect of the access proposal. One commenter noted that fee issues have “vexed” the industry for years. On the one hand, all markets obviously must be permitted to charge for their services, particularly agency markets such as ECNs that do not trade as principal and therefore cannot earn trading profits. On the other hand, these ECNs typically do not retain the bulk of the access fees that they collect. Instead, these fees mostly are paid out as rebates to customers who post limit orders with the ECNs. Consequently, two limit orders offering to sell the same stock at $10 per share posted in two different markets may in fact not represent equally-priced quotes. One order may receive a rebate out of access fees charged by the posting market and therefore effectively is offering to sell, not at $10, but at $10 plus the rebated amount. The fee proposal would control the extent to which rebates detract from the comparability of orders with identical displayed prices. In this respect, it may be more accurate to view the proposal as a limitation, not on compensation to markets, but on the additional compensation paid to some traders beyond the price they place on their orders.
Some commenters have argued, however, that regulatory action is not needed to assure the comparability of public quotes because market forces alone will be sufficient to address the issue. The Commission will need to evaluate this view carefully, as well as the views of all commenters, to reach an appropriate resolution of what has been an intractable issue.
C. Sub-Penny Quoting and Market Data Proposals
The sub-penny quoting and market data proposals have not received as much attention as the other proposals, but are important parts of the proposed regulatory reform. Both are intended primarily to promote public price discovery. The sub-penny quoting proposal would prohibit markets from accepting or displaying quotes in price increments of less than a penny, except in stocks with prices of less than $1 per share. The proposal would help protect limit orders by addressing the practice of “stepping ahead” of displayed orders by economically insignificant amounts.
The market data proposal, among other things, would modify the current formulas for allocating revenues that are generated from fees for dissemination of the consolidated data stream. The revised formula would reward markets for the value of their quotes — those that reflect the best prices for the largest sizes and thereby contribute the most to public price discovery. In addition, the market data proposal would promote the public dissemination of market information beyond that which currently is provided through the consolidated data stream.
Many commenters on the market data proposal have suggested that the Commission should revisit the issue of the level of fees charged by the markets for the consolidated data stream. In particular, some believe that such fees are too high and that the Commission should adopt a cost-based approach for evaluating the reasonableness of fees. The Commission extensively addressed this issue in 1999 when it published a concept release on market data fees and information. The release specifically requested comment on the concept of a cost-based approach for evaluating fees. The responses to this concept reflected deep divisions in the securities industry. In an attempt to resolve these divisions and to obtain additional views, the Commission established an Advisory Committee on Market Regulation in 2000. The Committee included a diverse range of participants drawn from the securities and market data industries. It specifically considered whether the Commission should adopt a cost-based approach for evaluating fees, but rejected the idea as unworkable.
As evidenced by the comments on the Regulation NMS proposal, however, many continue to believe that some sort of cost-based limitation on market data fees has merit. Separately from the Regulation NMS proposal, the Commission currently is reviewing the governance and transparency standards that apply to SROs. The level of market data fees is closely related to these issues because such fees represent a very significant source of SRO funding. The Commission will need to select the most appropriate forum in which to address continuing concerns about market data fees.
III. The Road Forward
I will conclude by offering a few thoughts on the future of the Regulation NMS rulemaking process. The comment period ended only a few weeks ago, and we continue to review the large number of comment letters. It therefore would be premature to predict how the Commission ultimately will resolve the many difficult issues raised by the proposals. I do want, however, to express my appreciation to the public, to the members of Congress, and to market participants for the enormous effort and insight reflected in their comments. At the public hearing in April, I asked the participants temporarily to set aside their individual interests and to put on their public policy hats. I am very pleased that not just the hearing participants, but the public in general, have responded quite positively. The views expressed in their comment letters, though naturally reflecting differing perspectives and priorities, have almost uniformly focused their attention on the public welfare and on promoting the efficiency and fairness of the U.S. equity markets as a whole. The letters fully warrant close review, and their insights will be reflected in any final rulemaking.
Although I cannot predict the outcome of the Commission’s proposed rulemaking, I do believe it is extremely important that there be an outcome, and that the outcome be reached in a timely manner. Many of the issues raised by the Regulation NMS proposals have lingered for many years and caused serious discord among market participants. These issues have been studied and debated and evaluated from nearly every conceivable angle. Few would seriously oppose the notion that the current structure of the national market system is outdated in some respects and needs to be modernized. The Commission must move forward and make decisions with regard to final rules if the U.S. equity markets are to continue to meet the needs of investors and public companies.
I will conclude by emphasizing that the Commission recognizes the far-reaching nature of many of the proposals. If adopted, some would require significant industry efforts to modify systems and otherwise prepare for the new regulatory structure. We are sensitive to these concerns and will work closely with the industry on the process needed to implement any new rules efficiently. This process clearly would include appropriate time periods for the industry to prepare before the new rules become effective.
Thank you again for inviting me to speak on behalf of the Commission. I would be happy to answer any questions that you might have.