NYSE Rulemaking:
Order Approving Proposed Rule Change
To Rescind Exchange Rule 390

SECURITIES AND EXCHANGE COMMISSION
(Release No. 34-42758; File No. SR-NYSE-99-48)

May 5, 2000

Self-Regulatory Organizations; New York Stock Exchange, Inc.; Order Approving Proposed Rule Change to Rescind Exchange Rule 390

On December 10, 1999, the New York Stock Exchange, Inc. ("NYSE" or "Exchange") filed with the Securities and Exchange Commission ("SEC" or "Commission"), pursuant to Section 19(b)(1) of the Securities Exchange Act of 1934 ("Exchange Act")¹ and Rule 19b-4 thereunder,² a proposed rule change to rescind Exchange Rule 390. The proposed rule change was published for comment in the Federal Register on February 28, 2000.³ The release publishing notice of the proposed rule change also included a Commission request for comment on issues relating to market fragmentation. The comment period relating to the rescission of Exchange Rule 390 expired on March 20, 2000. The Commission has received twelve comments letters explicitly addressing whether Rule 390 should be rescinded. These comments are summarized in section II below. The comment period on issues related to market fragmentation has been extended for two weeks and now expires on May 12, 2000.⁴

Off-board trading restrictions such as Rule 390 have long been questioned as attempts by exchanges with dominant market shares to prohibit competition from other market centers. On their face, such restrictions run contrary to the Exchange Act's objectives to assure fair competition among market centers and to eliminate unnecessary burdens on competition. The NYSE has defended Rule 390 on the basis that it was intended to address market fragmentation by promoting interaction of investor orders without the participation of a dealer, which also is a principal objective of the Exchange Act. Even granting the importance of this objective, however, Rule 390 is overbroad as a tool to address market fragmentation - it applies in many situations that do nothing to promote investor order interaction. In the after-hours context, for example, it creates an artificial incentive for trades to be routed to foreign markets. Rule 390 also effectively restricts the competitive opportunities of electronic communications networks ("ECNs"), which use innovative technology to operate agency markets that offer investors a high degree of order interaction. To avoid the anticompetitive effect of the Rule, some ECNs even have indicated that they would accept the very substantial regulatory responsibilities associated with registering as a national securities exchange, thereby foregoing the streamlined requirements available under Regulation ATS. Rescission of Rule 390 will eliminate these distortions of competition. The Commission will address legitimate concerns about assuring an opportunity for interaction of investor orders in the context of its ongoing review of fragmentation issues.

In an age when advancing technology and expanding trading volume are unleashing powerful forces for change and new competitive challenges for the U.S. securities markets, both at home and abroad, the continued existence of regulatory rules that attempt to prohibit competition can no longer be justified. Such rules typically succeed only in distorting competition and introducing unnecessary costs. The NYSE operates a market of very high quality. It recognizes that success in the future will depend on its ability to adapt and meet competitive challenges by continuing to provide a market that well-serves the interests of investors. The NYSE's proposed rule change to rescind Rule 390 is approved.

I. Description of Proposed Rule Change

The proposed rule change rescinds Rule 390, which generally prohibits NYSE members and their affiliates from effecting transactions in NYSE-listed securities away from a national securities exchange. Two Commission rules already limit the reach of Rule 390. Exchange Act Rule 19c-3⁵ limits the application of Rule 390 to stocks listed on the NYSE as of April 26, 1979. Exchange Act Rule 19c-1⁶ permits NYSE members to trade as agent in the over-the-counter market with another person, except when the member also is acting as agent for such other person. In addition, Rule 390 itself contains ten specific exceptions for unusual situations, such as a transaction that is part of a primary distribution by an issuer.⁷ Finally, an interpretation of the Rule permits members and their affiliates to trade as principal or agent on any organized foreign exchange at any time, and to trade as principal or agent in a foreign country's over-the-counter market after regular trading hours.⁸

The NYSE stated in its description of the proposed rule change that the intended purpose of Rule 390 was to maximize the opportunity for customer orders to interact with one another in agency auction markets and be executed without the participation of a dealer. The NYSE also discussed its concerns that broker-dealer internalization practices and market fragmentation would increase in the wake of Rule 390's rescission. It asserted that internalization - broker-dealers trading as principal against their customer order flow - results in the most objectionable of all forms of market fragmentation: the execution of captive customers' orders in a manner that isolates them from meaningful interaction with other buying and selling interest. The NYSE asserted that such practices not only decrease competitive interaction among market centers, but also isolate segments of the total public order flow and impede competition among orders, with no price benefit to the orders being internalized.

To address these concerns, the NYSE requested the Commission to adopt a new market-wide rule prohibiting broker-dealers from trading as principal against their customer orders unless they provide a price to the order that is better than the national best bid or offer against which the order might otherwise be executed. The NYSE asserted that this market-wide rule would assure that investors receive the fairest pricing of their internalized orders and would eliminate broker-dealer conflicts of interest in trading against their own customer order flow to capture the spread. The Commission's Concept Release sets forth the NYSE's proposal as one of the six potential options on which comment is requested.⁹

II. Summary of Comments

The Commission received twelve comment letters explicitly addressing whether Rule 390 should be rescinded.¹⁰ No commenter asserted that the Rule should be retained. Nearly all believed that the Rule imposed an unnecessary burden on competition. Four commenters, however, believed that the Commission should not approve the proposed rule change until it also addressed fragmentation concerns.

Many commenters supported rescinding Rule 390 on the ground that it is an unnecessary or inappropriate burden on competition.¹¹ The STA asserted that the rule is "an anachronism that limits liquidity and competition and therefore constrains investors from always obtaining the best possible price." ITG stated that the rule "imposes an unnecessary barrier to competition in listed securities between exchanges and other markets" and "imposes unnecessary costs on market participants." Instinet stated that "[a]mong the most significant factors that make such [off-board trading] rules obsolete is the development of electronic intermarket linkages that will ensure nationwide access to the best bids and offers available in any marketplace." Although supporting the rescission of the rule, AGS Specialist Partners stated that Rule 390 was "not intended as an anti-competitive initiative but as a protection for the public to ensure the proper exposure of their orders."

Several of these commenters also noted that rescission of the Rule would enhance the opportunity for competition between exchange markets and alternative trading systems.¹² The SIA stated that "technological advances and recent regulatory developments [have] led to the development of a host of alternative trading systems that provide a similar capability operating alongside the established markets in an intensely competitive environment," and that "[t]here is simply no justification for regulations such as Rule 390 that restrict off-board trading." Morgan Stanley noted that "the rule still may hinder the establishment and development of alternative OTC trading systems and markets in non-19c-3 listed stocks."

Other commenters believed that the Commission should take action to address possible collateral effects that could occur in the wake of rescinding Rule 390.¹³ Ashton Technology stated that it supported the rescission of the rule "if conditioned upon adoption of the NYSE Proposal as modified by an order exposure alternative, applying equally to upstairs market makers and exchange specialists, and calling for a new high powered routing mechanism with auto-execution capabilities to access and trade against `exposed' orders." The ICI supported the NYSE's recommendation that the Commission adopt "a market-wide requirement that broker-dealers not be permitted to trade as principal with their own customer order unless they provide for `price improvement,' i.e., a price to the order that is better than the national bid or offer against which the order might otherwise be executed." Nevertheless, ICI believed that the rescission of Rule 390 should not be delayed while the Commission considered whether to adopt a price improvement requirement.

Other commenters did not support the NYSE's proposal. The Knight/Trimark Group stated that the "NYSE's recommendation that the Commission adopt a new rule requiring broker-dealers to improve on the NBBO if they trade with customer orders as principal is an attempt to replace an Exchange rule that is explicitly anticompetitive with a Commission rule that is implicitly anticompetitive." The NASD criticized the NYSE proposal because it believed the proposal would "allow NYSE specialists to match the NBBO, while requiring market makers to attempt to improve [the NBBO] and also to bear the risk of the NBBO moving away in the interim." The NASD stated that best execution and order display obligations could achieve the same objectives as the NYSE's proposal.

Other commenters believed that the Commission should not approve the rescission of Rule 390 until it addressed market fragmentation issues.¹⁴ The AIMR noted that while it tentatively supports the rescission of the Rule, it "strongly believes that the present issue and those surrounding market fragmentation, which the Commission highlighted in its official request for public comment, are so closely related that the Commission cannot meaningfully consider each issue in isolation of the others." It requested that the Commission delay its decision regarding Rule 390 until the Commission had reviewed all public comments addressing possible market fragmentation and related issues. Finally, the BSE stated that "[a]t the very least, perhaps the Commission should deny the NYSE's requests to rescind Rule 390 until the Commission is satisfied that its rescission will not have a deleterious impact on the market, or until it has decided on the solution to any such anticipated deleterious impact."

In contrast, other commenters did not believe that the approval of Rule 390 should be delayed.¹⁵ The STA stated that "the question of internalization of customer orders touches upon a great number of important, compelling and interrelated issues regarding the roles of the exchanges, market makers, ECNs and investors," and that it was "inappropriate to link this complex and possibly contentious proposal with the proposal to rescind Rule 390." Morgan Stanley also believed that the Commission should not delay in its approval of the proposed rule change "pending its determination of what regulatory action should be taken to address the fragmentation issues."¹⁶

III. Discussion

The Commission finds that the proposed rule change is consistent with Section 6 of the Exchange Act¹⁷ and the rules and regulations thereunder applicable to a national securities exchange.¹⁸ In particular, the Commission finds the proposed rule change is consistent with Section 6(b)(5), which requires, among other things, that the rules of an exchange be designed to promote just and equitable principles of trade, to remove impediments to and perfect the mechanism of a free and open market and a national market system and, in general, to protect investors and the public interest, and Section 6(b)(8), which requires that the rules of an exchange not impose any burden on competition not necessary or appropriate in furtherance of the Exchange Act. The rescission of Rule 390 also is consistent with Section 11A of the Exchange Act,¹⁹ which sets forth the findings and objectives that are to guide the Commission in its oversight of the national market system. Rescinding Rule 390 will help further the national market system objective in Section 11A(a)(1)(C)(i) to assure the economically efficient execution of securities transactions and in Section 11A(a)(1)(C)(ii) to assure fair competition between exchange markets and markets other than exchange markets.

Rule 390 long has been questioned by the Commission and others because it directly restricts a certain type of market center competition - competition between exchange markets and markets other than exchange markets.²⁰ Given the explicit national market system objective to assure fair competition among market centers, as well as the requirement that the rules of a national securities exchange not impose any burden on competition not necessary or appropriate in furtherance of the Exchange Act, Rule 390 has been suspect on its face.

The NYSE has defended Rule 390 on the basis that its purpose was not to protect the NYSE's competitive position, but to protect customer interests by assuring a greater opportunity for interaction of investors' orders without the participation of a dealer. This type of order interaction is also a principal objective of the national market system set forth in Section 11A(a)(1)(C)(v) of the Exchange Act. Over the years, the Commission has sought to cut back on Rule 390 in ways that would reduce its anticompetitive nature without inappropriately reducing the opportunity for investor orders to interact. Exchange Act Rule 19c-1 allows NYSE members to execute trades in markets other than exchange markets as agents for their customers. Exchange Act Rule 19c-3 systematically has reduced the scope of Rule 390 over time as more and more companies have listed their stocks on the NYSE in the years since 1979. Nevertheless, the Rule still applies to securities that generate nearly one-half of total NYSE trading volume, including many of the most active NYSE issues.²¹

The Commission believes that whatever beneficial effect Rule 390 may have in enhancing the interaction of investor orders can no longer justify its anticompetitive nature. To the extent the Rule promotes the interaction of investors' orders, it does so in an undesirable way - by attempting a direct restriction on competition. Such attempts can never be wholly successful and typically succeed primarily in distorting, rather than eliminating, competition and introducing unnecessary costs. An egregious effect of Rule 390 is the artificial incentive it provides for NYSE members to route orders to foreign OTC markets for execution after regular trading hours. Such distortions can no longer be justified in an increasingly competitive international environment.²²

In addition, Rule 390 is much too broad even when considered solely as a tool to address market fragmentation and to promote the interaction of investor orders. As noted by several commenters, the Rule effectively restricts NYSE members from participating in markets operated by ECNs or ATSs.²³ These market centers offer their customers, among other things, agency limit order books that provide a high degree of investor order interaction. Using advanced technology for communicating and organizing information, ECNs can offer a number of advantages to investors, including low costs, fast display of limit orders, and fast executions against displayed trading interest.

These ECN limit order markets also can benefit the national market system as a whole by enhancing the process of public price discovery. Displayed limit orders are perhaps the most significant source of price competition in the securities markets. Limit order markets also allow for both investor and broker-dealer participation, but minimize principal-agent conflicts by adopting trading rules that establish a level playing field for the trading interest of both investors and broker-dealers - principally through price/time priority rules. Whatever limit order is first in line at the best price, whether submitted by investor or broker-dealer, such limit order has the right to trade first at that price. Price competition is invigorated and spreads are narrowed because those who improve the best bid or offer through limit orders know that they will be the first to trade. The price/time priority rules of limit order markets also can enhance depth and liquidity by providing an incentive for trading interest to stack up at prices that are at or around the best bid and offer. Because the second, third, and fourth orders in line at a price will be the second, third, and fourth to trade at that price (and so on), there is a strong incentive to submit limit orders even at prices that match or are outside the best bid or offer. The deeper a market, the less vulnerable it will be to excessive short-term price swings.²⁴

In recent years, the Commission has taken a number of steps that have paved the way for ECNs to compete with established market centers and be integrated into the national market system. In 1996, the Commission adopted the Order Handling Rules,²⁵ which required, among other things, the inclusion in the consolidated national best bid and offer ("NBBO") of limit order prices and sizes that improved the market for a security (by either improving the price of the NBBO or adding significant depth to the NBBO). These rules applied to both customer limit orders handled by specialists and market makers, as well as the limit orders of specialists and market makers themselves if they were displayed in an ECN. In 1998, the Commission adopted Regulation ATS,²⁶ which provides a streamlined regulatory regime for trading systems (including ECNs) that choose to be regulated as ATSs. In addition, ATSs with significant trading volume are required to display publicly their "top-of-book" trading interest in the consolidated national quote stream, even if such interest is not associated with a specialist or market maker. Most recently, the Commission approved a proposed rule change by the NASD that would enable ECNs to participate in the Intermarket Trading System that links market centers trading listed securities.²⁷ With the rescission of Rule 390, yet another regulatory barrier to competition will be eliminated.

The Commission emphasizes strongly, however, that its desire to clear away regulatory barriers to competition from ECNs in the listed market should not be interpreted as an indication of whether the ECNs will or should attract a significant amount of listed market share. That will be determined by competition. Similarly, the Commission's criticism of Rule 390 should not be interpreted as a criticism of the quality of the NYSE's market. To the contrary, studies repeatedly have demonstrated the merits of the NYSE's market, both in terms of its execution quality and its public price discovery function.²⁸

The NYSE offers a multi-faceted trading mechanism that can accommodate a wide variety of participants and trading strategies. Like the ECNs, it offers a limit order book with price/time priority among orders on the book. In addition, the NYSE, through its floor, offers a mechanism for investors with large trading interest to be represented in the market. Such investors typically will not display their full interest in a limit order because it likely would move the market against them, thereby increasing their transaction costs or even precluding any execution at all. The NYSE floor allows this large trading interest to interact with trading interest of all sizes on the other side of the market.²⁹ This enhanced opportunity for interaction can benefit both large and small investors. Indeed, the NYSE's very substantial price improvement rate for smaller orders is attributable to such interaction - more than 50% of market orders of less than 500 shares routed to the NYSE floor in stocks with a quoted spread of greater than 1/16^th are executed at a price better than the NBBO.³⁰

Finally, the NYSE has adopted a comprehensive set of trading rules that address the potential principal-agent conflicts that can arise when both broker-dealers and their customers trade in the same market center. These rules are intended to prevent NYSE members and professionals from obtaining unfair advantages in trading. In addition, the NYSE incorporates one market maker - the specialist - into its trading mechanism. Specialist trading is limited to help assure that it supplements, but does not supplant, public trading interest and thereby contributes to a fair and orderly market.³¹ The NYSE also monitors the actual performance of its specialists to assure that they comply with their affirmative and negative market-making responsibilities.

The outcome of the competition between the NYSE and other market centers will depend on which market centers are most able to serve investor interests by providing the highest quality trading services at the lowest possible costs. The Commission's regulatory task is removing unwarranted regulatory barriers to competition between the NYSE and other market centers. Its approval of the rescission of Rule 390 is intended solely to free the forces of competition and allow investor interests to control the success or failure of individual market centers.

Freeing the forces of competition to serve investor interests underlies the Commission's comprehensive review of issues related to market fragmentation. As discussed in the Concept Release, the Commission is concerned about certain broker-dealer practices that may substantially reduce the opportunity for investor orders to interact. Reduced order interaction may hamper price competition, interfere with the process of public price discovery, and detract from the depth and stability of the markets.

Currently, brokers that handle customer orders have a strong financial incentive either to internalize their orders by trading against them as principal or to route their orders to dealers that will trade against them as principal and share a portion of the profits with the broker. Internalization and payment for order flow arrangements provide dealers with a guaranteed source of order flow, eliminating the need to compete aggressively for orders on the basis of their displayed quotations. Instead, the dealers can merely match the prices that are publicly displayed by other market centers. These prices in many cases will represent limit orders that are displayed by agency market centers (such as the NYSE or an ECN). The limit orders may be denied an opportunity for an execution if dealers choose not to route orders to the market center displaying the limit orders and instead match the limit order prices.³²

Price-matching dealers thereby take advantage of the public price discovery provided by other market centers (which must make their best prices publicly available pursuant to Exchange Act price transparency requirements), but do not themselves necessarily contribute to the process of public price discovery. Moreover, if a substantial portion of the total order flow in a security is subject to dealer price-matching arrangements, it reduces the ability of other dealers to compete successfully for order flow on the basis of their displayed quotations. In both cases (unfilled limit orders and disregarded dealer quotations), those market participants who are willing to participate in public price discovery by displaying firm trading interest at the best prices are not rewarded for their efforts. This creates disincentives for vigorous price competition, which, in turn, could lead to wider bid-asked spreads, less depth, and higher transaction costs. These adverse effects would harm all orders, not just the ones that are subject to internalization and payment for order flow arrangements. Consequently, a loss of execution quality and market efficiency may not be detectable simply by comparing the execution prices of orders that are subject to such arrangements with those that are not.

Moreover, an agent-principal monitoring problem may tend to perpetuate rather than alleviate the isolation of investor orders that are subject to internalization and payment for order flow arrangements. It can be very difficult for retail customers to monitor the quality of execution provided by their brokers, particularly in fast-moving markets.³³ Given the difficulty of monitoring execution quality, the most rational strategy for any individual customer may be simply to opt for the lowest commission possible (which may be low in part because the broker is receiving payment for order flow, part of which is passed on to the customer). If many individual customers adopt this strategy, it could blunt the forces that otherwise would reward market centers that offer high quality executions.

Finally, the fragmentation concerns raised in the Concept Release are not limited to assuring that investors receive at least the best displayed prices, whatever they happen to be. Assuring that investors receive the best prices displayed anywhere in the national market system is crucial, but it is not sufficient to assure that the best prices displayed in the system are the most efficient prices reasonably possible. For example, the spread between the best displayed bid and the best displayed offer may be wider than it otherwise would be if a market structure fails to promote vigorous price competition.³⁴ Similarly, the depth of trading interest at the best displayed prices may be very thin, so that prices will be more volatile than they otherwise would be if a market structure does not reward traders for displaying multiple orders (and thereby adding depth) at the best prices. In addition, some market centers offer investors an opportunity for price improvement - an execution at a price better than the best displayed prices. To meet their best execution responsibilities, brokers must take these price improvement opportunities into consideration in deciding where to route customer orders.

Several commenters believed that the Commission should not approve the rescission of Rule 390 until it had addressed market fragmentation concerns. The Commission does not believe, however, that the potential fragmentation of the listed market due to an increase in internalization and payment for order flow arrangements warrants a delay in approving the proposed rule change. First, the Commission already has commenced its review of market fragmentation issues, and the comment period for the Concept Release ends on May 12, 2000. Several of the six potential options to address fragmentation set forth in the Concept Release would address internalization and payment for order flow arrangements.³⁵ The Concept Release also requests comment on any additional options, or modifications of any of the six options, that commenters believe would be useful in addressing fragmentation.³⁶ Second, the Commission intends to monitor any significant changes in the order-routing practices of NYSE members resulting from the rescission of Rule 390, particularly decisions to internalize their customer order flow. To comply with the duty of best execution owed their customers, brokers would need to assure that such changes further their customers' interests and not merely their own.

IV. Conclusion

IT IS THEREFORE ORDERED, pursuant to Section 19(b)(2) of the Act,³⁷ that the proposed rule change (SR-NYSE-99-48) is approved.

By the Commission.

Jonathan G. Katz
Secretary

Footnotes

¹ 15 U.S.C. 78s(b)(1).

² 17 CFR 240.19b-4.

³ Securities Exchange Act Release No. 42450 (February 23, 2000), 65 FR 10577 ("Concept Release").

⁴ Securities Exchange Act Release No. 42723 (April 26, 2000).

⁵ 17 CFR 240.19c-3.

⁶ 17 CFR 240.19c-1.

⁷ NYSE Rule 390(c)(i).

⁸ NYSE Rule 390, Supplementary Material .10, Interpretations of the Market Responsibility Rule.

⁹ Concept Release, note 3 above, section IV.C.2.b.

¹⁰ George Reichhelm, General Partner, and Andrew Schwarz, General Partner, AGS Specialist Partners, dated March 16, 2000 ("AGS Letter"); Deborah A. Lamb, Chair, Advocacy Committee, and Maria J. A. Clark, Office of General Counsel, Association for Investment Management and Research, dated March 15, 2000 ("AIMR Letter"); Fredric W. Rittereiser, Chairman and Chief Executive Officer, and William W. Uchimoto, Executive Vice President and General Counsel, Ashton Technology Group, Inc., dated March 20, 2000 ("Ashton Technology Letter"); George W. Mann, Jr., Senior Vice President and General Counsel, Boston Stock Exchange, dated March 17, 2000 ("BSE Letter"); Craig S. Tyle, General Counsel, Investment Company Institute, dated March 20, 2000 ("ICI Letter"); John Oddie, Chief Executive Officer, Global Equities, Instinet Corporation, dated March 20, 2000 ("Instinet Letter"); Timothy H. Hosking, ITG, Inc., dated March 17, 2000 ("ITG Letter"); Kenneth D. Pasternak, President and Chief Executive Officer, and Walter F. Raquet, Executive Vice President, Knight/Trimark Group, Inc., dated March 21, 2000 ("Knight/Trimark Letter"); Robin Roger, Managing Director and Counsel, Morgan Stanley Dean Witter, dated March 27, 2000 ("Morgan Stanley Letter"); Richard G. Ketchum, President, National Association of Securities Dealers, Inc., dated March 31, 2000 ("NASD Letter"); Marc E. Lackritz, President, Securities Industry Association, dated March 21, 2000 ("SIA Letter"); Robert C. King, Chairman, and Lee Korins, President and Chief Executive Officer, Security Traders Association, dated March 15, 2000 ("STA Letter").

In addition, the Commission has received other letters that address fragmentation issues, but do not address explicitly whether Rule 390 should be rescinded. Copies of all comment letters are available for inspection and copying in File No. SR-NYSE-99-48 in the Commission's Public Reference Room, 450 Fifth Street, N.W., Washington, D.C. Electronically-submitted comment letters are posted on the Commission's Internet web site (http://www.sec.gov).

¹¹ AIMR Letter; Ashton Technology Letter; ICI Letter; Instinet Letter; ITG Letter; Knight/Trimark Letter; Morgan Stanley Letter; NASD Letter; SIA Letter; STA Letter.

¹² ITG Letter; Morgan Stanley Letter; SIA Letter

¹³ AGS Letter; Ashton Technology Letter; ICI Letter.

¹⁴ AGS Letter; AIMR Letter; Ashton Technology Letter; BSE Letter.

¹⁵ ICI Letter; Morgan Stanley Letter; STA Letter.

¹⁶ Morgan Stanley also recommended that the NYSE file an additional proposal with the Commission to rescind Exchange Rule 393, asserting that it no longer serves "any valid regulatory purpose." Rule 393 requires members to obtain NYSE approval prior to participating in an off-board secondary distribution of an NYSE-listed security.

¹⁷ 15 U.S.C. 78f.

¹⁸ In approving this proposal, the Commission also has considered its impact on efficiency, competition, and capital formation. 15 U.S.C. 78c(f).

¹⁹ 15 U.S.C. 78k-1.

²⁰ See, e.g., Exchange Act Section 11A(c)(4), 15 U.S.C. 78k-1(c)(4) (provision added to the Exchange Act in 1975 directing the Commission to review exchange rules that impose off-board trading restrictions); Securities Exchange Act Release No. 11628 (Sept. 2, 1975), 40 FR 41808 (Commission commences proceedings under Exchange Act Section 19(c) to determine whether to amend or abrogate exchange rules that impose off-board trading restrictions).

²¹ Jeffrey Bacidore, Katharine Ross & George Sofianos, Quantifying Best Execution at the New York Stock Exchange: Market Orders, NYSE Working Paper 99-05 (December 1999) at 1 n.2 ("At the end of October 1999, 23 percent of NYSE issues accounting for 46 percent of the volume were subject to Rule 390.").

²² The trades executed in foreign markets also are not subject to NYSE surveillance or the Commission's regulatory oversight.

²³ For example, none of the various exceptions to, and limitations on, the scope of Rule 390 would generally allow an NYSE member to trade as principal in a U.S. market operated by an ECN.

²⁴ See Concept Release, note 3 above, at n.26 and accompanying text.

²⁵ Securities Exchange Act Release No. 37619A (Sept. 6, 1996), 61 FR 48290.

²⁶ Securities Exchange Act Release No. 40760 (Dec. 8, 1998), 63 FR 70844.

²⁷ Securities Exchange Act Release No. 42536 (Mar. 16, 2000), 65 FR 15401.

²⁸ See, e.g., Hendrik Bessembinder, Trade Execution Costs on NASDAQ and the NYSE: A Post-Reform Comparison, 34 J. Financial & Quantitative Analysis 387, 389 (1999) ("This study finds that trade execution costs remain larger on NASDAQ compared to the NYSE even after the new SEC order-handling rules are implemented, and that the difference in average trading costs is not attributable to variation in observable economic characteristics of the listed stocks."); Marshall E. Blume & Michael A. Goldstein, Quotes, Order Flow, and Price Discovery, 52 J. Finance 221, 232 (1997) ("The NYSE bid price equals on average the best bid price 97.1 percent of the time, and the NYSE ask price equals the best ask price 96.9 percent of the time."); Joel Hasbrouck, One Security, Many Markets: Determining the Contributions to Price Discovery, 50 J. Finance 1175, 1197 (1995) (an analysis of price discovery for equities traded on the NYSE and regional exchanges revealed that "price discovery appears to be concentrated at the NYSE: the median information share is 92.7 percent"); Justin Schack, Cost Containment, Institutional Investor, Nov. 1999, at 43 (worldwide survey of institutional investor trading costs found that "[f]or the first time ever NYSE-listed shares took top honors for the cheapest cost of execution anywhere in the world"); compare Louis K.C. Chan & Josef Lakonishok, Institutional Equity Trading Costs: NYSE versus Nasdaq, 52 J. Finance 713, (1997) (comparison of execution costs for institutional investors on Nasdaq and NYSE found that "costs are lower on Nasdaq for trades in comparatively smaller firms, while costs for trading larger stocks are lower on NYSE").

²⁹ Some ECNs offer an opportunity for large trading interest to interact by including a reserve size feature in their limit order book. See Concept Release, note 3 above, at text accompanying n.27.

³⁰ See Quantifying Best Execution, note 21 above, at Table 10 & Table 14. A market's price improvement rate is affected by the quality of the publicly displayed quotations that are "price-improved." The quality of the NYSE's public quotations is one of the issues addressed in the studies cited in note 28 above.

³¹ See Kenneth A. Kavajecz, A Specialist's Quoted Depth and the Limit Order Book, 54 J. Finance 747, 753 (1999) (comparison of spreads on NYSE limit order book with specialist's quoted spreads "suggests that the specialist plays an important role in narrowing the spread the market participants face when demanding liquidity, especially for smaller (less frequently traded) stocks.").

³² In February 2000, the agency markets operated by ECNs executed approximately 19% of the share volume in Nasdaq securities, a drop of 3% from September 1999. See NASD Economic Research Dept., http://www.marketdata.nasdaq.com (visited April 10, 2000) (In February 2000, ECNs that are ATSs collectively accounted for 19.2% of Nasdaq share volume, 25.1% of Nasdaq dollar volume, and 24.6% of Nasdaq trades.); NASD Economic Research Dept., <http://www.marketdata.nasdaq.com> (visited Dec. 11, 1999) (In September 1999, ECNs that are ATSs collectively accounted for 22.2% of Nasdaq share volume, 29.2% of Nasdaq dollar volume, and 28.0% of Nasdaq trades.). In calculating the market share of ATSs, the NASD adds orders executed internally on an ATS and the orders routed to an ATS for execution. Orders routed out to another market participant are not included.

³³ See, e.g., Lawrence Harris, Consolidation, Fragmentation, Segmentation, and Regulation, in Modernizing U.S. Securities Regulation: Economic and Legal Perspectives 269, 286 (Kenneth Lehn & Robert W. Kamphius, Jr., eds., 1992) ("[F]ew brokerage clients - and probably no small clients - can observe, monitor, and measure their brokers' efforts at low cost. Given the high volatility of securities prices, the general lack of real-time market information available to most brokerage clients, and the high cost of processing that information even when it is readily available, most clients cannot accurately determine whether their orders are well executed or not. Moreover, even if they could measure their brokers' performance, fairly evaluating that information is still more difficult. A fair evaluation would require that the clients compare the quality of service offered by at least a few different brokers.") (footnotes omitted). Retail investors have greater access to real-time market information today than in 1992. The other barriers to monitoring execution quality continue to exist.

³⁴ The spread between the best bid and offer is an indication of the premium that must be paid by investors seeking liquidity and therefore of the efficiency of the market. See Concept Release, note 3 above, at n.20 and accompanying text.

³⁵ See Concept Release, note 3 above, section IV.C.2.b.

³⁶ After the end of the comment period, the Commission intends to review expeditiously the comments submitted in response to the Concept Release and determine what, if any, further action is necessary.

³⁷ 15 U.S.C. 78s(b)(2).

http://www.sec.gov/rules/sro/ny9948o.htm