Comments on SR-NASD-99-53 (Foley, 09-12-2000)

September 12, 2000

U.S. Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, DC 20549

Attention: Mr. Jonathan Katz, Secretary

Re: SEC File No. SR-NASD-99-53

Ladies and Gentlemen:

Bloomberg Tradebook LLC appreciates the opportunity to comment on Amendment No. 7 to the SuperMontage proposal of the National Association of Securities Dealers, Inc. (the "NASD"), which the Commission published for comment in Securities Exchange Act Release No. 43138 (August 10, 2000) (the "August 10 Release"), and the letter Mr. Richard Ketchum of the NASD sent to the Commission's staff enclosing Amendment No. 7.¹

I. INTRODUCTION

Technological change and fair competition are in the best interests of investors and are essential to the growth and continued dominance of our securities markets. As technology has revolutionized the structures and mechanisms of the securities markets, ECNs have been a principal catalyst for that change. ECNs have made technological innovations accessible to investors and investors, in turn, have rewarded ECNs with a larger and larger share of the market. At the same time, both the Commission's staff and market participants, including Bloomberg Tradebook LLC, are aware that the technological innovations in the securities markets pose challenges as well as opportunities. In response to unprecedented and unpredictable technological change, our common goals ought to be protecting the interests of investors while encouraging innovation and ensuring fair and competitive securities markets. We do not believe those goals can be met by adopting a proposal that would, in effect, stifle competition, retard technological growth and innovation and unfairly privilege some markets and market participants at the expense of others.

ECNs have provided and, in the absence of governmental or self-regulatory organization interference, will continue to provide significant benefits to investors. Customer service is a significant part of the success of ECNs to date. ECNs provide an agency brokerage alternative to market makers. ECNs do not collect fees by charging principal spreads, but instead charge agency commissions on the orders they handle.

ECNs have substantially transformed the Nasdaq marketplace by offering technological and trading alternatives that Nasdaq, if not faced with their competition, would not have developed on its own. Indeed, the SuperMontage proposal copies many of the ECNs' prior inventions, such as the "reserve" (undisplayed trading interest attached to a disclosed order that can be accessed by a counterparty) and "auto-refresh" (after an execution exhausting the displayed quantity, refreshing the displayed quantity from untapped reserve quantity). While the NASD, several years after the ECNs introduced these innovations, would now attempt to catch up by mimicking the ECNs' trading features, the NASD's SuperMontage proposal is intended at the same time to eradicate the ECNs and to shield the new Nasdaq SuperMontage ECN from competition. If the NASD succeeds in this regulatory assault on the private ECNs, where will the creative energy to improve the market come from in the future?

As the Commission knows, there are serious competitive issues involved in the SuperMontage proposal. The NASD, however, has not ever made clear exactly how the order-routing algorithm would work or what design features have been built into the proposed SuperMontage ECN Nasdaq has been constructing for itself. Indeed, many significant features of the NASD's SuperMontage have only recently come to light.

As more fully explained below, we believe that a Commission order approving the SuperMontage on the basis of the record to date would be reversible as a matter of law. We are very concerned about having that eventuate because we are convinced it not only would delay the resolution of the issues but also would be bad for the markets. Nasdaq needs to improve its technology and SuperMontage would provide infrastructure improvements. If the NASD would remove the inappropriate burdens on competition and other deleterious aspects of its proposal, at least some of which probably are inadvertent, the SuperMontage proposal could be approved and implemented and the markets would benefit. If instead the Commission approves the SuperMontage on the basis of the record to date and litigation is brought to rescind the Commission order, the resulting delay and uncertainty would frustrate very real and useful progress toward improving the Nasdaq market. That result would not serve the public interest or the protection of investors.

The SuperMontage proposal should be measured against the need to preserve and strengthen fair and competitive market structures and to protect investors. Those principles are embedded in the Securities Exchange Act of 1934 (the "Exchange Act") and should govern the Commission's evaluation of the proposal.

There are two fundamental problems with the SuperMontage proposal as it now stands. Both are correctable in an Amendment No. 8 that the NASD should file. First, the NASD should delete aspects of the SuperMontage that would use NASD regulatory power (i) to give its proposed Nasdaq ECN order book unfair competitive advantages over ECNs, and (ii) to give an unfair priority over ECN participants, through the Nasdaq ECN order book, to the 450 market makers that apparently support the SuperMontage proposal. Ideally, that result would have been achieved by having separate technology for the NASD's collection and dissemination of quotation data as an exclusive securities information processor on the one hand, and its private ECN order-display functions on the other. That would imply, in turn, a need to segregate (i) the order- and execution-routing function to access Nasdaq quotations from (ii) the order- and execution-routing functions to access orders lodged in its private ECN. That segregation should have been reinforced by having separate design teams, insulated from one another by a functioning fire wall. Instead, however, the NASD elected to leverage its position as a regulator and to require its members to fulfill their quotation-display obligations under the Commission's Order Execution Rules² by providing orders to the new SuperMontage ECN. That use of regulatory power to enhance Nasdaq's monopoly market power offends the very notions of fair competition under the Exchange Act and, if it ever were implemented, would restrain trade and competition in this important line of interstate commerce.

In the absence of the strict technology-segregation approach we would have preferred, which would have avoided confrontation and controversy, there are ways of addressing the ill effects of the composite approach the NASD in fact has taken. These palliative measures, which are discussed below, would alleviate to some degree the burdens on competition of the SuperMontage and make it easier for the Commission to conclude that the remaining burdens are permissible under the Exchange Act.

The second, also curable, problem is the need for greater disclosure about the proposed workings of the SuperMontage. This rule filing is far from routine. It represents a far-reaching renovation of the Nasdaq market and requires searching public and Commission scrutiny. Even in the case of more routine filings, the Commission's review of self-regulatory organization rules should be done in a way that permits both the Commission and the public to understand key aspects of a self-regulatory organization's proposal. In the case of a conduct rule, the text of the rule, together with whatever interpretive or explanatory information the self-regulatory organization chooses to append, normally would suffice. In the case of a filing having the market-wide significance and impact of the SuperMontage, however, it is not sufficient for the NASD to publish just the text of what it chooses to print in its rulebook, together with a brief description. What is needed is a full explanation, either in the rule filing itself or on a website or other identified, publicly available source, of how the system would function, including the text of the specifications or, in the NASD's case, system "requirements" (the "SuperMontage Requirements"). Indeed, in this instance, we believe the SuperMontage Requirements, which instruct the programmers on what the programs are to achieve, should be regarded as an integral part of the proposed rules the Commission should review under the Exchange Act and, accordingly, should be published in an Amendment No. 8 to the NASD's filing for public inspection and comment.

It is our hope that, upon correcting the existing problems, the NASD can obtain Commission approval of its SuperMontage and move forward in a timely fashion to implement this important market initiative.

II. DISCUSSION

Turning first to the substantive aspects of the SuperMontage proposal, we note that the NASD has not addressed a number of the substantive issues that we presented in our memorandum of August 1, 2000, entitled "Issues Requiring Correction before Commission Approval of the Nasdaq SuperMontage Proposal," a copy of which is annexed to this letter as Attachment A, which we presented to the Commission in response to a request by Chairman Levitt. Before detailing the issues that the NASD did not address, however, we will focus on the issues discussed in Mr. Ketchum's August 7, 2000 letter, which accompanied the NASD's filing of Amendment No. 7. As noted above, we are convinced that the NASD should have separated entirely the design and programming of the aspects of the SuperMontage devoted to quotation collection and dissemination, which relate to the NASD's role as an exclusive securities information processor, from the aspects that are devoted to Nasdaq's proposed ECN, which is a private commercial venture that should not be supported by the NASD's regulatory power. The monopoly power the NASD enjoys as the exclusive market in Nasdaq stocks is problematic enough without adding to it the commercial incentive to use that power to crush the other ECNs that do not have regulatory subsidies.

In the absence of such an approach, which the NASD should have adopted from the outset, we believe there are lesser ways of addressing the problems that we currently can see with the SuperMontage (further problems may become evident when the SuperMontage Requirements are published). These lesser, palliative measures are discussed below.

Directing trading interest to specific market makers and ECNs. The monopoly power Nasdaq enjoys by virtue of the NASD's regulatory powers would severely alter competitive conditions in the market place and give Nasdaq's SuperMontage, as currently structured, an inappropriate, government-granted franchise. That franchise would consist of an integrated system of technology both for accessing and responding to quotations and for operating a private ECN. That combination would tend to restrain competition since the NASD's regulatory powers in maintaining the accessing and response technology would be used to nourish the private ECN. No other private party would enjoy that government-granted privilege.

Allowing the direction of trading interest to specific market makers and ECNs would go a substantial way towards achieving the separation of the NASD's technology for accessing and responding to quotations on the one hand and operating the SuperMontage ECN on the other. That separation has not been achieved under the proposed rule change as currently formulated because the NASD has not provided for the direction of trading interest to specific market makers. In response to criticisms made by Instinet and Bloomberg Tradebook, the NASD has now proposed to give market makers and ECNs the option to accept directed "liability" orders through the directed-order process of the proposed SuperMontage system. This is only a partially satisfactory solution to the problems we and other commenters earlier identified. Bloomberg Tradebook, and we expect other ECNs, would choose this welcome option, but market makers who might otherwise wish to do so would not likely make that choice because of the risk of double execution. That risk would be introduced unnecessarily by the SuperMontage's design.

Under the SuperMontage, market makers would be subject to execution messages as in the current Small Order Execution System ("SOES") but would not be offered the option of receiving directed execution messages, or in other words, directed SOES. If they were permitted to receive directed execution messages, they would not face a double execution risk since Nasdaq presumably would not send them duplicative execution messages. They thus would have the same ability to elect this option as the ECNs would have. Denying them this option would deny investors the ability to choose not to interact with the SuperMontage if they did not wish to do so. That would surely give the SuperMontage ECN an unfair advantage over all other market venues.

SuperMontage order priorities. With its SuperMontage proposal, the NASD plans to replace a system that does not have market-wide prioritization with a system of prioritization that is not based on time/price priority but on favoring market makers over the interests of ECN participants. Effectively, under the SuperMontage proposal, market makers are given a preference for their own proprietary quotation, under which they "jump the queue" and trade with orders they themselves enter for their customers, preempting other previously submitted contra orders at the same price.³ That preference, it would appear, would apply only to a market maker's order in the SuperMontage ECN, and would not apply to a proprietary order by the same market maker in a competing venue, such as another ECN.

To find a convenient pretext for such favoritism, Nasdaq has turned to access fees, which ECNs charge but market makers do not. The NASD may have done so in the belief that it should not assert more directly that agency-only brokers should get in line behind market makers trading for their own accounts. It is little wonder so many market makers support the SuperMontage proposal.

The NASD's argument in favor of that prioritization is a flimsy pretext, one which the Commission should examine and reject.

The NASD has argued that it is fair either to adjust ECN quotations by adding the amount of the access fee to the ECN's offer and subtracting it from the ECN's bid or to leave the ECN quotation unadjusted and to arrange the order-routing algorithm so that ECNs that charge an access fee go last in the queue at a given price. The problem with the first approach is that it would single out ECN access fees for special, discriminatory treatment. Commissions and access fees are routinely charged on transactions in the securities markets. In the case of Nasdaq securities, the transacting parties pay various fees, including fees for using SelectNet. None of those, in the NASD's estimation, should result in an adjustment to a quotation. In fact, the only fees that would be added or subtracted would be the ECN access fees. The NASD has yet to explain why this discrimination is consistent with the Exchange Act.⁴ The NASD has yet to clarify, moreover, whether it would charge fees for the Nasdaq SuperMontage ECN and, if so, whether those fees would be added or subtracted in the same manner.

The NASD's proposed alternative, putting the access-fee ECNs last in the queue, apparently would be based on the NASD's unsubstantiated and indeed untrue assertion that "[t]his prioritization is consistent with common industry practice today, where a market participant will route its orders first to market makers and ECNs that do not charge an access fee and then to ECNs that charge an access fee . . . ."⁵ Indeed, Bloomberg Tradebook routes a large proportion of its orders to market makers and ECNs and generally prefers to route to ECNs, including ECNs that charge access fees, instead of routing to market makers that do not. It does so because the ECNs, being automated with superior technology, respond more quickly than the market makers, many of which are not automated, and the ECNs offer additional undisplayed liquidity in the form of "reserve" quantities and in many cases offer price improvement over their published quotations. We understand that a number of other market participants adopt the same approach. We expect that the ECNs' technological and other innovative advantages will continue in the new environment.

Accordingly, the NASD's proposed discrimination against ECNs that charge access fees is unsupported by industry practice. In any event, the NASD did not present any data to substantiate its assertion that its discrimination is consistent with industry practice generally or is indeed appropriate in light of the statutory standards of the Exchange Act applicable to the NASD. Market makers and others know which ECNs charge access fees and could well make whatever discriminations they themselves wished, without having their commercial choices dictated or overridden by their regulator, the NASD. The NASD's stated reasons for putting access-fee ECNs last in the queue, or in any way discriminating against ECNs that charge access fees, do not have any sound basis in fact or policy.

NASD operation of order-routing system. We do not suggest that Nasdaq should not be able to provide order- and execution-routing services. We agree with Mr. Ketchum that such an idea would be "preposterous". Of course, Nasdaq should be free to provide technology for accessing the quotations it publishes as an exclusive securities information processor. What we stated instead, and continue to state, is that Nasdaq's proposed new ECN should have every opportunity to compete on the basis of quality of service and other commercial attributes of its proposed ECN but should not be able to leverage its ECN's power on the basis of the monopoly power the NASD itself has by virtue of being our regulator. To guard against that possibility, and indeed to control the risk that the NASD will have built into the SuperMontage algorithm any further features that would discriminate against the other ECNs and tilt the playing field in favor of the Nasdaq SuperMontage ECN, we have recommended and reiterate here our recommendation that the NASD's technology for responding to market maker or ECN quotations should be separated from the technology for submitting or responding to orders in the Nasdaq SuperMontage ECN.

Five seconds versus seven seconds. We think it is helpful that the NASD has now clarified that the time period for withdrawal of an order sent to an ECN and cancellation of an ECN's quotations is seven seconds rather than five seconds. The NASD originally proposed a maximum response time of five seconds for processing ECN orders, but it has since stated "that the five second maximum response time is in reality seven seconds." The NASD noted parenthetically that "two seconds is added to this time period for internal Nasdaq system processing."⁶

We do not object to the imposition of penalties on ECNs when they do not respond in a timely fashion to order messages sent by Nasdaq. The NASD's proposal, however, would continue its current practice of punishing ECNs when the failure is Nasdaq's. Indeed, it would worsen the current practice by making the punishment greater. A penalty is acceptable when the fault is the ECN's, but not when it is Nasdaq's. The penalty would be greater under the SuperMontage because it would consist of "zeroing out" the ECN's quotation on that side of the market, which would severely hamper the ECN's ability to do business in that stock. By tying the timing to Nasdaq's own clock, which starts running upon Nasdaq's transmission of an order rather than actual delivery of the order to an ECN, Nasdaq would leave ECNs with the risk that the SelectNet-style delivery system will fail, as it often has, to function properly. That in turn would relieve Nasdaq from an important incentive to improve its existing technology.

We note the NASD has not proposed to address the principal problem we identified, that its proposed technology is well below the industry standard in that it does not incorporate electronic "heartbeats" and acknowledgments. As a result, as we stated before, Nasdaq does not know, and cannot know, when it fails to receive a response to an order it believes it has transmitted, whether the intended recipient has yet received the order. A properly designed and implemented system of heartbeats and acknowledgments would protect market makers and ECNs from the current problem of miscued trades and unexpected market exposure. It is not evident why the NASD has been unwilling to address that problem.

It appears to be an improvement that Nasdaq would commit to transmit orders instantaneously. Nasdaq is supposed to do this today, however, and it repeatedly fails to do so. Would the new system be any better? How are we, or the Commission, or the investing public, to know?

We are encouraged by the NASD's suggestion that SuperMontage would protect ECNs against cancellations better than SelectNet does today. It is particularly helpful that an order entrant would not be able to cancel an order until the ECN had executed, partially executed, or declined the order. Nevertheless, while this new amendment would modify the current ten-second period before an order entrant could withdraw an order, it would put in its place a seven-second time period after which Nasdaq itself could withdraw the order from the ECN. As a result, from the point of view of an ECN, there would not be any net benefit and indeed the ECN's risk of market exposure from SuperMontage miscues would be increased.

Nasdaq's assertion that its approach would achieve the same end as the heartbeat-and-acknowledgment approach we recommended is untrue. For the reasons discussed above, Nasdaq's proposed approach would continue to leave ECNs exposed to the foibles of Nasdaq's technology. That, as the NASD states, Nasdaq lacks, and will continue to lack, the capacity to implement technology that other industry participants accept as standard goes far to demonstrate the substandard condition of Nasdaq's technology and capacity. That in turn reinforces our concern about having Nasdaq be the single point of failure for such a vital part of the securities markets' operation. Nasdaq should not be allowed to hide behind its sub-par technology if it is to run the SuperMontage system as a government-granted monopoly, one that could be relatively impervious to competitive challenge given the NASD's existing, government-granted monopoly power.

Now that the NASD has clarified its proposal to some degree, we are in a position to offer what we think might be a workable alternative. We believe the mechanism should rely on measurement of an ECN's performance and therefore should be geared to the ECN's own internal clock. The ECN should have an order taken away if, upon Nasdaq's receipt of the ECN's response, it is clear that the time elapsed between the ECN's actual receipt of an order and the ECN's transmission of a response exceeds a mandated maximum. Indeed, that maximum could comfortably be set lower than the five or seven seconds proposed by Nasdaq.

Appropriating order flow not directed to the SuperMontage ECN. The NASD makes an important point in noting that market participants would be able to avoid having their unexecuted orders or partially unexecuted orders flow automatically to the proposed SuperMontage ECN if the participants would only mark their orders as "immediate or cancel." What the NASD leaves unsaid, however, is that the default condition would be appropriation of the orders by the SuperMontage ECN. That would be not a neutral condition, but a playing field seriously tilted in favor of Nasdaq's private venture. Also, the NASD was not at all forthcoming about this aspect of its proposed technology. The ECNs had to divine it by imagining what anti-competitive features might be buried in the SuperMontage technology. That in turn raises the question of what was not disclosed, what other means exist by which the SuperMontage would appropriate order flow not directed to it.

These matters illustrate, moreover, the broader point we have made before, that the NASD's regulatory power to compel order display and order access should be separated by a reinforced fire wall from its commercial ECN venture. It is evident that the NASD and its senior management, who stand to profit personally in great amounts from a successful Nasdaq, should not be allowed to enlist the Commission's help in arranging for themselves a government-sponsored monopoly.

Non-disclosure of SuperMontage operation. What is even more disturbing than the matters discussed above is that these essential characteristics of the SuperMontage's operations are not spelled out in the text of the rules the NASD has filed but apparently are embedded in the system programming, which is not disclosed or fully discussed. That makes it impossible to do more than take the NASD's word for how the system works or to ask questions or make comments about the elements of the system that are not disclosed in the NASD's filing. The NASD has responded only partially to the questions previously raised by the ECNs and other commenters and it has done so largely by making piecemeal disclosure of previously undisclosed aspects of the SuperMontage. As noted above, publication of the SuperMontage Requirements as part of the rule filing would resolve this issue.⁷

Other issues not addressed by the NASD. In addition to the issues Mr. Ketchum discusses in his August 7, 2000 letter and to the points made above, there are a number of issues discussed in our August 1, 2000 Issues paper⁸ that the NASD does not address or respond to in any significant way and that continue to present legal impediments to a finding that the SuperMontage proposal is consistent with the provisions of the Exchange Act applicable to the NASD. These issues are listed below:

1. The parts of the SuperMontage that the NASD asserts would be voluntary are in fact coercive. As we noted in our August 1, 2000 Issues paper, the NASD has the power as an exclusive securities information processor to compel the submission of quotations to Nasdaq. In the SuperMontage proposal, it would use that power in a way that would in fact compel the submission of orders to the private SuperMontage ECN.⁹
2. Nasdaq's SuperMontage ECN would be allowed to display "three deep" quotations, that is, quotations on three price levels, the best bid and offer (the "BBO") and the two prices inferior to the BBO, whereas market makers and other ECNs would be allowed to display only the tops of their files.¹⁰ To provide for fair competition, the SuperMontage should make available to all market participants via its NQDS ("Level II") Nasdaq feed on a real-time basis three-deep information it currently collects from market makers and other ECNs.¹¹ That would eliminate the advantage Nasdaq currently proposes to make available only to its private-venture SuperMontage ECN.¹²
3. The proposed algorithm unfairly penalizes order-delivery ECNs that enter an order at one quantity and then do any of the following: (i) increase the size to add an additional quantity to reflect one or more customer limit orders, as contemplated by Rule 11Ac1-4; (ii) reduce the order size; or (iii) place or change an order on the other side of the market for the security. The algorithm would inappropriately treat the entire amended order as a new order and take away the ECN's or market maker's time priority. An order under similar circumstances in the SuperMontage ECN would not suffer a similar fate.¹³
4. The SuperMontage technology for responding to quotations would access reserve quantities in the SuperMontage only and would not and cannot provide the same service for reserve quantities attached to quotations of order-delivery ECNs. This is a further demonstration of the built-in and unfair advantages that would be accorded to the Nasdaq SuperMontage ECN over other market participants. SuperMontage should not prevent ECNs from having an accessible reserve if the ECNs elect to take orders instead of executions. This unjustifiable discrimination is not necessary or appropriate and is but one further instance of the NASD's effort to give its own private-venture ECN a unique competitive advantage. The remedy for this discrimination would be to abolish it, to require that the NASD not provide any service to its own ECN that it does not offer on identical terms to all other ECNs.
5. There remains a significant risk that Nasdaq would claim or imply that its system offers the only true path to best execution. The NASD's regulatory press on brokers and institutional investors to achieve best execution at peril of regulatory sanction would enhance the apparent attractiveness of using the Nasdaq SuperMontage ECN as a "safe harbor", if not more, with at least an implicit imprimatur from NASD Regulation, Inc.
6. We continue to be concerned that the SuperMontage proposal's reliance on mandatory, SOES-type messaging for market makers would cause substantial market-wide bottlenecks and make the Nasdaq technology a single point of failure. We believe several market makers have the technical proficiency to meet the same response standards as ECNs. We urge that these market makers be allowed to do so in exchange for being offered the same option as the ECNs to receive orders instead of SOES-type executions. This would go a long way to relieving the capacity concerns raised by the current system design of the SuperMontage.

The SEC or NASD should establish rules by which market makers and ECNs must respond to orders within a "few seconds" and, if they do not, they will receive executions. This would be similar to the standard the Commission has set for ECNs in the no-action letters previously issued to them.¹⁴ Establishing and enforcing response standards would be a better way than the SuperMontage to remedy a lack of response by market makers and it would avoid the SuperMontage's unjustifiable burdens on competition and its likely harm to the markets. The Commission can conduct independent audits and set these standards, much as it does today through OATS and other procedures.

7. As currently configured, the SuperMontage provisions for anonymity of orders would permit order-entry broker-dealers to enter unattributed orders that would be directed to ECNs that, in turn, would not know the identity of the order-entry firm and as result would not be able to know whether the order entrant was willing or unwilling to pay the ECN's access fees. That would constitute a direct assault on the ECNs that charge access fees because it would provide an easy way for order-entry firms not willing to pay access fees to nevertheless access the ECNs' liquidity.

In view of all these matters, the Commission cannot find that, as a substantive matter, the proposed SuperMontage is consistent with the provisions of the Exchange Act applicable to the NASD, particularly Sections 15A(b)(6) and 15A(b)(9).¹⁵ Indeed, the implications of approving the SuperMontage as currently constituted would go beyond the immediate proceeding. In evaluating whether the current filing is consistent with the Exchange Act as it applies to the NASD, we respectfully submit that it should also look to the similarly worded provisions of the Exchange Act applicable to national securities exchanges, particularly Sections 6(b)(5) and 6(b)(8). Almost 60 years ago, in the Multiple Trading Case, the Commission held that the public policy under the Exchange Act voided a rule of the New York Stock Exchange (the "NYSE") that purported to prohibit NYSE members from routing orders in NYSE-listed securities to the regional exchanges.¹⁶ Later, when the NYSE tried to impose a "Public Limit Order Protection Rule" (the "PLOPR") that would have required its members to clear limit orders on the NYSE specialists' books before taking orders to regional exchanges, the Commission once again objected and entered disapproval proceedings under the Exchange Act.¹⁷ In the latter case, the Commission was unpersuaded by the favorable market impacts the NYSE argued would arise from its proposed rule. In the order commencing proceedings to consider disapproval of the PLOPR, the Commission stated that the fact that an NYSE proposed rule change would limit the ability of member organizations to effect transactions on other market centers within the United States was, per se, a ground for disapproval of the proposed rule change.¹⁸ As in the Multiple Trading Case, the NYSE's arguments about the need to have the rule to promote market integrity and customer protection were irrelevant to that basic statutory issue.

If the Commission were to approve the SuperMontage proposal, which would be a market-wide equivalent of the now-discredited NYSE Rule 390, it might well have to allow other market centers, such as the NYSE, to have similar rules. That would permit the NYSE, for example, to establish an order-routing scheme similar to the SuperMontage scheme, which basically puts the ECNs and the regional exchanges at a substantial disadvantage. Putting roadblocks in the path of competing market centers restricts their ability to attract order flow and diminishes their ability to compete in ways quite similar to the market effects the Commission previously condemned in the Multiple Trading Case and its PLOPR order. We recommend, accordingly, that the Commission consider that precedential issue as it evaluates the NASD's SuperMontage, particularly in light of the private commercial interests the SuperMontage is designed to protect.¹⁹

Procedural matters. In addition to these substantive problems, the SuperMontage proceeding is materially insufficient as a procedural matter. These procedural deficiencies deprive the Commission of a lawfully sufficient basis for approving the SuperMontage proposal. The NASD's refusal, in the original filing and in the seven subsequent amendments, to provide the legally required discussion of the burdens on competition that would be proposed by its rule filing not only deprives the public of a basis to understand the important impacts the proposed SuperMontage would have on the markets and on investors, but it also deprives the Commission of a legally sufficient basis on which to determine whether the SuperMontage proposal is consistent with the provisions of the Exchange Act of 1934, particularly Sections 15A(b)(6) and 15A(b)(9), applicable to the NASD. The approach taken by the NASD in this proceeding is not, in our view, consistent with the Commission's own rulemaking approach during its long and distinguished history. In matters of this kind, particularly where valuable franchises would be awarded and the public interest would be vitally affected, sunlight is indeed the best disinfectant.²⁰

The NASD's rote incantation of the statutory standard in Section 15A(b)(9) in its filing, unsupported by any discussion or any demonstration at all, that "Nasdaq does not believe that the proposed rule change will result in any burden on competition that is not necessary or appropriate in furtherance of the purposes of the Act,"²¹ does not satisfy the requirements of the Commission's Form 19b-4 that such burdens be explained and justified in detail.²² As the Commission is aware, moreover, the courts have applied strict scrutiny to rule filings that do not meet statutory standards. The scope of this obligation is explained in Timpinaro v. SEC,²³ in which the U.S. Court of Appeals for the District of Columbia Circuit remanded an SEC-approved rule to the Commission for failure to substantiate its belief that the rule would achieve the desired objective.²⁴

In a context where the competitive impacts of an NASD rule proposal are certain to be trivial, an elaborate discussion is not warranted. That is not the case here, however. The NASD's statements do not elucidate the issue and, as a result, the NASD's filing does not provide a sufficient basis for public comment on, or Commission approval of, the proposed rule change.²⁵ Given the inadequacy of the record, the public is effectively deprived of a meaningful opportunity for comment on the proposed rule change. The Commission, in turn, is denied the benefit of the comments that could arise from the fully informed dialogue and genuine interchange of data, views and arguments the Congress envisioned in fashioning the rule-approval process embodied in Section 19(b) of the Exchange Act.²⁶ That is particularly the case where a valuable franchise or privilege such as the SuperMontage represents would be awarded, one that would give the newly privatized Nasdaq a government-granted monopoly over an important article of interstate commerce.²⁷

In addition to those problems, many important aspects of the actual operation of the SuperMontage algorithm remain unclear. It appears, as noted above, that significant aspects of the SuperMontage are not fairly disclosed in the rule filing even though they substantially affect the manner in which the SuperMontage would work, such as the priority routines built into the algorithm and the way in which orders would be captured by the SuperMontage ECN. For that to be the case after an original filing and seven filed amendments is a serious problem because it frustrates the ability of those who would be subject to the SuperMontage, including investors, market makers and ECNs, to understand its application and effects. That in turn is unfair to both the public and the Commission in their efforts to understand the meaning and implications of the proposed rule change. Those implications may well include a substantial likelihood that the SuperMontage, if indeed it were ever implemented in its current form, would have severely disruptive and dysfunctional effects, as has the NASD's current technology, effects that are not discussed in the NASD's filing. Accordingly, the NASD has plainly failed to comply in several significant respects with the requirements of Form 19b-4 and in that way has frustrated the purposes of the congressionally mandated rule-review process.

The NASD's rulebook, and its proposed rule changes, need not set forth detailed technical information of insubstantial importance. Nevertheless, the Commission's own Rule 19b-4 defines a proposed rule change as including any "stated policy, practice, or interpretation" of the NASD that is not "reasonably and fairly implied" by an existing" rule of the NASD.²⁸ The term "stated policy, practice, or interpretation" is in turn defined to include "any material aspect of the operation of the facilities of the [NASD]."²⁹ Viewed from the perspective of information the Commission and the public need if they are to understand rule changes proposed by the NASD, Rule 19b-4 should be interpreted to require disclosure in reasonable detail of any aspect of a system such as the SuperMontage that is likely to have a material bearing on whether the proposed rule change would or would not be consistent with the statutory standards applicable to the NASD. As the Commission knows, those standards include particularly the requirements in Sections 15A(b)(6) and 15A(b)(9) that the NASD's rules not unfairly discriminate between NASD members, that they remove impediments to and perfect the mechanism of a free and open market and a national market system, that they protect investors and the public interest and that they not impose any burden on competition that is not necessary or appropriate in furtherance of the purposes of the Exchange Act.³⁰

There is a practical limit in the case of technologically oriented rule filings to the utility of textual descriptions alone, no matter how conscientiously they have been prepared. In some cases, for the Commission and the public to have a legally sufficient basis for understanding a proposed rule change, it may indeed be necessary to include some fairly elaborate disclosure of technical matters, perhaps in appendices or on websites referenced in the filing. Depending on the circumstances, it may be important that disclosure be made of technical specifications, including, where appropriate, flow-charts and schematic diagrams, and specifications or system requirements, however denominated, that were used by the programmers in designing the system program code, to permit expert commenters to evaluate and critique the operation of a proposed system.³¹ At a minimum, moreover, the NASD should have to disclose, in terms its own members can understand and evaluate, a textual description of the operation of the system in enough detail to address the statutory standards. Generic descriptions that omit or disguise important aspects of a system's operation, such as in this case the ways in which the programs have been designed to benefit the Nasdaq SuperMontage ECN and to disadvantage other private ECNs, should of course not be permitted.

The SuperMontage Requirements, which as noted above are the directions given to the Nasdaq programmers as to the objectives they are to achieve, are indeed part of the proposed rule change the NASD should have filed and made part of the public record for the Commission's information and benefit and for that of public commenters. In the case of a proposed conduct rule, the text of the rule itself, together with whatever interpretive or explanatory material the NASD chooses to append, is the "proposed rule change" for purposes of the Commission's Exchange Act review. In the case of the SuperMontage, however, where a major order-processing system is involved, the "proposed rule change" is not whatever the NASD elects to put in its rule book as the SuperMontage rules (which, realistically, might well incorporate only a brief summary of many of the technical aspects of the SuperMontage), but is instead all the key elements of the SuperMontage that have a bearing on whether the SuperMontage is or is not consistent with the provisions of the Exchange Act that govern NASD rules, particularly in this case Sections 15A(b)(6) and 15A(b)(9). To hold otherwise would be to frustrate the very congressional purposes that underlie the self-regulatory organization rule review process envisioned in Sections 3(f) and 19(b) of the Exchange Act.

In the case of the SuperMontage proposal, the substantive and procedural defects in the NASD's filing go to the essence of the Commission's adjudicatory process with respect to this proposed rule change. They have effectively undermined the purpose of the notice-and-comment process. In and of themselves, the defects would make it lawfully impermissible for the Commission to approve the proposed rule change on the basis of the current record. Given those irregularities, the Commission would have been justified in returning the filing to the NASD without publishing it. In discussing the substantive requirements of Rule 19b-4 under the Exchange Act, the Commission has stated:

The Commission notes . . . that notices of proposed rule changes need to be clear in order to elicit meaningful public comment. Although a proposed rule change may be accepted as filed, the Commission believes that it should not be published until it has reached an adequate level of clarity regarding the issues raised by the filing. For complex filings, this can require more extensive review. Filings of rule changes also need to include information necessary to enable the Commission's staff to conduct a complete review. Any filings that fail to comply with the requirements of Form 19b-4 may be returned to the SRO and will be deemed not to have been filed with the Commission.³²

The NASD's approach in this proceeding, particularly its repeated failure to disclose exactly how the trading algorithm would work and the ways in which it has secreted built-in algorithmic advantages for Nasdaq's own proposed ECN, by no means comports with the statutory standards under the Exchange Act or the Commission's own procedural requirements for NASD rule filings. Indeed, the courts have underscored that accuracy in the notice given to the public is critical to the achievement of the objectives and policies underlying the policy justification standards of informal rulemaking, which as noted above self-regulatory organizations such as the NASD are supposed to meet:

The purpose of the comment period is to allow interested members of the public to communicate information, concerns, and criticisms to the agency during the rule making process. If the notice of proposed rule making fails to provide an accurate picture of the reasoning that has led the agency to the proposed rule, interested parties will not be able to comment meaningfully upon the agency's proposals . . . . In order to allow for useful criticism, it is especially important for the agency to identify and make available technical studies and data that it has employed in reaching the decisions to propose particular rules. To allow an agency to play hunt the peanut with technical information, hiding or disguising the information that it employs, is to condone a practice in which the agency treats what should be a genuine interchange as mere bureaucratic sport.³³

III. CONCLUSION

In view of these remaining, uncorrected problems in the NASD's filing, the Commission cannot lawfully approve the NASD's SuperMontage filing on the basis of the current record. A Commission order approving the filing would be reversible as a matter of law. Accordingly, the Commission should require the NASD to cure the substantive deficiencies noted above and in the other comments the Commission has received, the Commission should require the NASD to prepare and file a detailed disclosure, as required by Form 19b-4, of the ways in which the SuperMontage programming would actually handle orders, including the detailed programming requirements and specifications embodied in the SuperMontage Requirements and a detailed discussion of the burdens on competition that the SuperMontage would impose, and then the Commission should publish for public comment those items of information, which would finally give the Commission and the public an appropriate basis for evaluating the SuperMontage, and after that for Commission approval of the SuperMontage.

* * *

We appreciate the opportunity to make our views known to the Commission and the staff and we hope that our letter is helpful. If members of the Commission or of the staff believe we may be of further assistance in these matters, please let us know.

Very truly yours,

BLOOMBERG TRADEBOOK LLC

By: Kevin M. Foley by RDB

Kevin M. Foley

Attachments

cc(w/att.): The Hon. Arthur Levitt, Chairman
The Hon. Isaac C. Hunt, Jr., Commissioner
The Hon. Paul R. Carey, Commissioner
The Hon. Laura S. Unger, Commissioner
Annette L. Nazareth, Esq., Director,
Division of Market Regulation
Robert L. D. Colby, Esq., Deputy Director,
Division of Market Regulation
Belinda Blaine, Esq., Associate Director,
Division of Market Regulation
Mr. Stephen L. Williams, Senior Special Advisor
Division of Market Regulation
David M. Becker, Esq., General Counsel

ISSUES REQUIRING CORRECTION BEFORE COMMISSION APPROVAL OF THE NASDAQ SUPERMONTAGE PROPOSAL

At the request of the staff of the Securities and Exchange Commission (the "Commission"), Bloomberg Tradebook LLC has prepared the following list of unresolved issues relating to a proposed rule change creating a new Nasdaq Order Display Facility (commonly referred to as "SuperMontage") submitted by the National Association of Securities Dealers, Inc. (the "NASD").

As the staff is aware, Bloomberg Tradebook is very concerned with the NASD's efforts to use its regulatory advantages to create an anticompetitive "SuperECN" in the Nasdaq OTC marketplace. Accordingly, we oppose adoption of the SuperMontage proposal in its current form and suggest to the Commission that it is inconsistent with the Securities Exchange Act of 1934, particularly Sections 15A(b)(6) and 15A(b)(9) thereof. At the staff's request, however, we have prepared this list to assist the staff in identifying those operational issues that must be resolved before the SuperMontage proposal is approved. In light of the significant changes to the structure of the Nasdaq marketplace posed by the SuperMontage initiative, we seek the staff's assurance that any amendments to the SuperMontage proposal-including any amendments that attempt to address any issues raised herein-will receive thoughtful review and public consideration before Commission consideration of whether to approve or disapprove the proposed rule change.

I. SUPERMONTAGE PARTICIPATION IS MANDATORY AS A PRACTICAL MATTER; CHANGES ARE NEEDED TO REMEDY THIS ANTI-COMPETITIVE DEFECT

A. SuperMontage appears to be voluntary but would in fact be mandatory in some respects and coercive in others. Participation in the SuperMontage would be voluntary in some respects but sufficiently coercive that the SuperMontage would have unfair competitive advantages by virtue of the NASD's position as an SRO, as demonstrated below.

1. It would be voluntary for a market maker or ECN to put its order book into the SuperMontage ECN. A Nasdaq market maker or an ECN would not be directly compelled to do so. Participation in the SuperMontage would not be voluntary with respect to responding to orders in the SuperMontage ECN. Under the Order Execution Rules, a market maker or an ECN must publish its top of file in Nasdaq or another exchange which provides facilities for brokers to access the quotation. Nasdaq has to provide this access and does so today, but the SuperMontage would provide only one means of responding to market maker and ECN quotations and the SuperMontage ECN. Thus, users of Nasdaq's links to market maker and ECN quotations would be forced to interact with the SuperMontage ECN order book.

2. Because of its current position as the SRO for Nasdaq-listed securities, Nasdaq receives quotations in Nasdaq-listed stocks published by market makers and ECNs and provides electronic access to those quotations.

3. While putting orders in the SuperMontage would be de jure voluntary, it would be de facto mandatory because the same Nasdaq technology as would be used for accessing quotations in the quotation facility would automatically put the order into the SuperMontage ECN, feeding Nasdaq's SuperMontage ECN, and would make it available for other orders residing in that SuperMontage ECN order book. E.g., if Goldman Sachs published a quotation in the SuperMontage, an order from another NASD member seeking to access that quotation would be placed in the SuperMontage order-execution queue and would then be automatically available for interaction with other orders in the SuperMontage ECN.

4. There appear to be instances in which an order meant to access a market maker or ECN quotation would find itself not just engaged in the SuperMontage responding logic but in fact placed in the SuperMontage ECN order book. For example, please consider the situation in which an order entrant placed a marketable limit order with the intention of accessing a market maker's or ECN's quotation that was alone at the best bid or offer. If someone else had accessed the same quotation first, exhausting its trading interest, it would appear that the order entrant's marketable limit order would automatically be deposited "into the system" (i.e., the SuperMontage ECN order book) for delivery to the next quoting market participant who might come along, whenever that was. As another example, if Nasdaq determined that an ECN whose quotation was alone at the best bid or offer had failed to respond to such an order within five seconds, Nasdaq would deposit the order into the SuperMontage ECN order book. In both instances, SuperMontage in effect would preference itself as the default order depository. In addition, not only would the order be caught up in the responding logic, it would also be caught up in the SuperMontage ECN, in preference to all other market venues.

5. A different result would obtain if the SuperMontage ECN were separated from the NASD/Nasdaq quotation facility and if orders seeking to interact with a particular quotation could be directed solely and exclusively to a particular market maker's or ECN's quotation. Market makers and ECNs should have two separate pieces of technology: one for order display and to access quotations from market makers and ECNs and another to trade with orders in the SuperMontage ECN order book.

6. Because of the use of a single responding technology for both quotations and orders, Nasdaq would make mandatory the use of its SuperMontage ECN order book as a means of responding to both displayed quotations and the SuperMontage ECN order book and would do so to promote its own private, commercial advantage.

B. The part of SuperMontage that would be "voluntary" would in fact be coercive. The SuperMontage facility for responding to quotations and orders would build such advantages for the Nasdaq SuperMontage order book that ECN competitors might well feel compelled to enter their customer orders into the SuperMontage ECN order book rather than to succumb to this unfair competition.

1. Quotations from a market maker or ECN would receive price/time standing inferior to the standing of an order in the SuperMontage ECN order book. For example, please consider the situation of a market maker or ECN that is alone, on behalf of a customer, at the best bid with a bid for 1,000 shares. The market maker or ECN subsequently is joined at the best bid by orders in the SuperMontage ECN order book. If a second customer of the market maker or ECN then joined the first customer for another 1,000 shares at the best bid, the market maker or ECN quotation would be updated to 2,000 shares and would thereby be demoted to last place in the order-execution queue. Nasdaq has so designed its SuperMontage, however, that a similar order in the SuperMontage ECN order book would not suffer a similar fate.

2. Another inherent advantage Nasdaq would provide for the SuperMontage ECN would be that the technology for responding to quotations would access reserve quantities in the SuperMontage only and would not provide the same service for reserve quantities attached to quotations of market makers and ECNs.

3. There is a significant risk that Nasdaq would claim or imply that its system offers the only true path to best execution. The NASD's regulatory press on brokers and institutional investors to achieve best execution at peril of regulatory sanction would enhance the attractiveness of using SuperMontage as a "safe harbor", if not more, with at least an implicit NASD-R imprimatur.

C. Reliance on a single technology creates the anti-competitive evils that give the SuperMontage ECN an unearned and unfair advantage over its competitors. Accordingly, separating these functions into two separate pieces of technology, one for accessing all market maker and ECN quotations (including those of the SuperMontage ECN on an equal basis) and another just for accessing Nasdaq's SuperMontage ECN order book would alleviate many of the anti-competitive effects of SuperMontage.

D. The separation of quotation collection and accessing quotations could be largely achieved by restoring preferencing to market makers and ECNs. Bringing back preferencing would achieve the separation of technologies because market participants would be able to respond to quotations in the market place without placing their orders in the SuperMontage ECN order book. Accordingly, participation in the SuperMontage could then more appropriately be said to be voluntary.

1. The NASD should provide technology that allows preferencing and access to the tops of files (in other words, the published quotations) of all market makers and ECNs on an equal basis. This technology should not go below the tops of the files unless the NASD provided equal treatment below the tops of the files for all market makers and ECNS, including the SuperMontage ECN order book.

2. At our meeting last week, Richard Ketchum of the NASD stated that preferencing would be restored for ECNs accessing each other, but not for market makers. This would not go far enough to solve the problem. The Commission should require that this concession be extended to all NASD members accessing the quotations of market makers or ECNs. Only as so extended would preferencing restore the competitive balance and establish a level playing field.

3. The NASD should provide separate technology to access the SuperMontage ECN order book, but that technology should not access any other market maker or ECN. Without this change, the SuperMontage ECN order book would retain unfair advantages over its competitors that do not enjoy a government-sponsored advantage.

II. THE SUPERMONTAGE WOULD CREATE THE DOUBLE EXECUTION PROBLEM IT PURPORTS TO SOLVE

A. Purported solution to SOES issues. The harm faced by market makers under the current SOES system should be remedied, but not by making SOES-type executions universal across SuperMontage.

1. SOES executions are currently about 3% of the volume in Nasdaq stocks, an immaterial amount.

2. In addition, SOES is shut down when ECNs are at the best bid or offer.

3. ECNs are at the best bid or offer from 50% to 75% of the time, according to data provided by American Century.

4. Therefore, the risk of double execution exposure due to current SOES is minimal. "Solving" the current minor problem of double execution from SOES does not outweigh the harms inherent in the SuperMontage proposal. Moreover, if the Commission were to establish response standards, any double execution exposure would be solved and the SuperMontage harms would not occur.

5. At our meeting last week, Richard Ketchum of the NASD argued that ECNs and market makers can simply connect to each other. That argument is specious since, by restoring the primacy of SOES, this proposal would create the double execution problem; it would make it impossible for ECNs and market makers to interconnect outside of the SuperMontage ECN and avoid the problem of double or multiple execution.

B. Response standards. The SEC or NASD should establish rules by which market participants must respond to orders within a "few seconds" and, if they do not, they will receive executions. This would be similar to the standard the Commission has set for ECNs. Establishing and enforcing response standards would be a better way than the SuperMontage to remedy lack of response by market makers and it would avoid the SuperMontage's unjustifiable burdens on competition and its likely harm to the markets. The Commission has the ability to conduct independent audits and set these standards, much as it does today through OATS and other procedures.

III. SUPERMONTAGE WOULD EXACERBATE CURRENT "D.K." PROBLEMS UNLESS NASDAQ UPGRADES ITS TECHNOLOGY TO CURRENT INDUSTRY STANDARDS

A. Nasdaq should upgrade its technology or take responsibility for late messages. If the combination of quotation display and order collection were removed, many of the remaining issues would be far less significant. Nevertheless, the problems of Nasdaq technology would remain because they get in the way of efficient order handling and order execution. Nasdaq's technology lacks features that are standard elsewhere in the securities industry, particularly including "heartbeats" and acknowledgments. Consequently, Nasdaq's communications systems do not accord with the needs of the industry and Nasdaq simply ignores the problem or asserts that it does not exist. Under Nasdaq's technology, as a result, ECNs receive notices of cancellation with respect to orders they have already executed.

1. Examples of these technological problems follow:

Example 1. Order cancellation arrives too late to prevent execution of canceled order:

12:00:00 Noon: SelectNet order is placed.

12:00:01: Nasdaq transmits the order to an ECN as a SelectNet order.

12:00:06: Nasdaq, not having heard back from the ECN, takes the order away to give it to someone else.

12:00:07: The ECN receives the order from Nasdaq.

12:00:08: The ECN executes the order, believing it to be live, and sends an acceptance to Nasdaq.

12:00:09: The ECN receives the notice of cancellation from Nasdaq, leaving the ECN with market exposure.

Example 2: Order cancellation arrives before order and is discarded. Canceled order is then received and is falsely believed to be a live order:

12:00:00: Customer puts in an order.

12:00:01: Nasdaq transmits the order to an ECN.

12:00:06: Nasdaq transmits a notice of cancellation to the ECN.

12:00:07: The ECN receives the notice of cancellation.

12:00:08: The ECN discards the notice, believing the notice to be erroneous because the ECN does not have any order to cancel.

12:00:09: The ECN receives the original order and executes it.

The ECN has no knowledge of its exposure until it receives an indication, substantially later, that the "order" it had executed was no longer live.

2. As is demonstrated by both of these trading examples, which occur frequently in practice today when order-entry firms cancel SelectNet orders, the foibles of the Nasdaq technology today create miscommunications and impose costs on ECNs and other market participants that would not be caused if Nasdaq's technology functioned appropriately. This exposure would be exacerbated if Nasdaq were allowed to cancel and switch orders in a five-second window it has proposed would apply under SuperMontage. For instance, under example 1 above, after transmitting an order cancellation to ECN A (with the order itself to arrive after the cancellation) and waiting five seconds, Nasdaq then transmits the order and the order cancellation to ECN B. Assuming the same technology failures, Nasdaq could create an exposure for multiple market participants.

3. Nasdaq currently does not use the industry-standard protocol "FIX". All other market participants use FIX protocol when communicating with one another. Nasdaq's refusal to use the state-of-the-art software protocol creates errant multiple executions and exposure today. Nasdaq has no proposal to improve the current problems. Moreover, SuperMontage will make this exposure problem worse because, as noted above, use of the SuperMontage ECN would be de facto mandatory.

4. Nasdaq should have to develop a system of heartbeats (to confirm that the absence of an acknowledgment does not mean lack of a connection) and acknowledgments and should be able to start the clock for purposes of its five-second rule and for purposes of converting orders into executions, as we propose, only at the time the order is in fact received by the ECN or market maker. The existing problem would not be adequately addressed by increasing the interval before an ECN would be deemed to have rejected an order since the problem is not at the ECN's end of the communication; Nasdaq should be made to improve its communications if the problem is to be satisfactorily addressed. Any other approach would reward Nasdaq for its failures to upgrade its technology.

5. Richard Ketchum asserted at our meeting last week that it would be impossible for Nasdaq to upgrade its technology to insert heartbeats and acknowledgments. This may well be the case, but if it is, Nasdaq should not be allowed to introduce SuperMontage.

IV. INCLUDING ACCESS FEES IN PUBLIC QUOTATIONS WOULD HARM INVESTORS

A. There is not any good reason to include access fees in the quotation because: (i) the fees charged to access brokers by a particular ECN are not at a set amount but instead vary considerably from broker to broker; (ii) including the access fees would cause the optical spread to increase, which in turn would give market makers the ability to widen their own real spreads since, under current circumstances, an ECN is at the best bid or the best offer about 75% of the time and at both the best bid and best offer about 50% of the time, as the American Century data show; and (iii) in many cases, the access fees are absorbed by the access brokers and not passed along to their customers, with the result that the customers in fact receive the advantage of the published quotations without giving effect to the access fees.

B. Nasdaq itself proposes to charge fees to users of its SuperMontage ECN. Nasdaq, however, has not proposed to include these fees in its quotations. Furthermore, Nasdaq has not published the fees as part of its proposed rule change.

V. SUPERMONTAGE WOULD INVOLVE IMPROPER REGULATORY COERCION AND WOULD HARM INVESTORS

A. The proposed combination of quotation display and order collection is anti-competitive and not in the public interest.

1. There is no need to couple or integrate the quotation medium with an execution facility.

2. The SuperMontage would confiscate the order flow that today belongs to the ECNs and is a highly valuable asset. Instinet, for example, reports that it has been advised that its order flow adds between $500 million and $750 million to its enterprise value.

3. With Nasdaq becoming a privately owned, for-profit company, it should not have a regulatory subsidy, but should be required to compete on a level playing field with other liquidity pools, such as the existing ECNs.

4. Although the NASD has made a few minor changes to reduce the disincentives to private messaging away from the SuperMontage it previously had built into its proposal, a number of disincentives remain and these are designed to shore up Nasdaq's revenues. The NASD itself views private wire connections as a threat to the monopoly rents it currently extracts from SelectNet, as it told the prospective investors in Nasdaq.

SelectNet is Nasdaq's automated market service that enables securities firms to route orders, negotiate terms, and execute trades in Nasdaq securities. If pairs of market makers or ECNs determine that they do enough order routing traffic in a day so as to justify setting up an alternative proprietary network for their traffic, Nasdaq may be forced to reduce its fees or risk losing its share of the order routing business. A reduction in the order routing business could have an adverse effect on Nasdaq's business, financial condition and operating results.³⁴

The SuperMontage effectively forecloses investor choices by mandating the use of a central processing facility run by Nasdaq.
The cure for this problem would be for Nasdaq not to have the power to assign any execution messages, but only to route orders, and to separate the order collection facility, which can be run on a non-compulsory basis by Nasdaq, from the quotation display facility, which the NASD itself, and not Nasdaq, should run.

B. The Congress foresaw the problem of giving an exclusive securities information processor the kind of monopoly power the SuperMontage would give Nasdaq and it warned against doing so:

The Committee believes that if economics and sound regulation dictate the establishment of an exclusive central processor for the composite tape or any other element of the national market system, provision must be made to insure that this central processor is not under the control or domination of any particular market center. Any exclusive processor is, in effect, a public utility, and thus it must function in a manner which is absolutely neutral with respect to all market centers, all market makers, and all private firms. Although the existence of a monopolistic processing facility does not necessarily raise antitrust problems, serious antitrust questions would be posed if access to this facility and its services were not available on reasonable and nondiscriminatory terms to all in the trade or if its charges were not reasonable. Therefore, in order to foster efficient market development and operation and to provide a first line of defense against anti-competitive practices, Sections 11A(b) and (c)(1) would grant the SEC broad powers over any exclusive processor and impose on that agency a responsibility to assure the processor's neutrality and the reasonableness of its charges in practice as well as in concept [emphasis added].³⁵

VI. REPUBLICATION. The SEC should publish the changes in Amendments Nos. 5 and 6 and in the further amendment that will be necessary to respond to the problems identified at our recent meeting and in this outline before the Commission gives further consideration to the proposed rule change. The public comment process would be significantly frustrated if commenters were not given notice of the significant changes to the proposed rule change since Amendment No. 4.

August 1, 2000

In Figure 1, below, we present the illustration of the Nasdaq Order Display Facility proposed for SuperMontage that the NASD published in Securities Act Exchange Release No. 42166 (November 22, 1999), augmented by the addition of fields denominated "Hidden", in which we have hypothesized, on the basis of what the NASD information shows, the numerical information the SuperMontage display conceals.

In the NASD's published example, the total size displayed on the bid side is 9,000 shares at the best price, $20.00 in this example, and the four market sources of that liquidity are shown: MMA, SIZE, MMB@ and ECN1. The aggregate liquidity available at the second and third best prices, $19.95 and $19.90, respectively, show only the identities of the ECNs or other market makers for which those prices are, in each case, the displayed MMID's agency or principal top of the file. What would not be displayed on the SuperMontage would be the second and/or third best agency or principal price of an MMID that had a better first displayed price. We have adjusted that by showing in the "Hidden" fields what those second and third hypothetical prices would look like if the SuperMontage display were improved as we recommend. We have made the same adjustment on the ask side as well.

Figure 1

Displayed

Displayed

Best Bid

$20.00

9,000

Best Ask

$20.05

4,400

MMA

$20.00

1,000

ECN2

$20.05

1,000

SIZE

$20.00

5,700

SIZE

$20.05

3,400

MMB@

$20.00

1,500

4,400

ECN1

$20.00

800

9,000

Displayed

Displayed

2^nd Best Bid

$19.95

15,000

2^nd Best Ask

$20.10

5,000

MMB

$19.95

1,000

ECN1

$20.10

800

ECN2

$19.95

500

MMB@

$20.10

500

ECN3

$19.95

1,500

MMD@

$20.10

3,400

MMC@

$19.95

100

4,700

3,100

Hidden

Hidden

MMA₂

$19.95

900

ECN2₂

$20.10

300

ECN1₂

$19.95

11,000

Displayed

Displayed

3rd Best Bid

$19.90

25,000

3rd Best Ask

$20.15

15,000

ECN4

$19.90

1,000

MMA

$20.15

500

MMA@

$19.90

1,000

ECN4

$20.15

2,500

2,000

MMC@

$20.15

100

Hidden

MMB

$20.15

1,000

SIZE₂

$19.90

1,000

4,100

ECN1₃

$19.90

10,000

ECN2₂

$19.90

5,000

Hidden

ECN3₂

$19.90

7,000

ECN1₂

$20.15

9,000

23,000

SIZE₂

$20.15

400

ECN2₃

$20.15

1,500

10,900

In Figure 2, we present a composite display that would give the market a complete picture of the MMIDs that entered a bid or ask at any of the three best prices and also that MMID's second and third best price, on both an agency and proprietary basis, regardless of whether those individual second and third prices were themselves within the top three prices in the SuperMontage. For example, MMA, which had a $20.00 displayed bid and a $19.90 displayed agency bid in the Nasdaq proposed SuperMontage, would also have displayed, in the composite we recommend, its second best proprietary bid, at $19.95, its second best agency bid, at $19.85, its third best proprietary bid, at $19.85 and its third best agency bid, at $19.80.

_{Figure 2}

Second Best Bid/Ask

Bid

Ask

MMA₂

$19.95

900

ECN2₂

$20.10

300

ECN1₂

$19.95

11,000

ECN1₂

$20.15

9,000

SIZE₂

$19.90

1,000

SIZE₂

$20.15

400

ECN2₂

$19.90

5,000

ECN4₂

$20.20

1,200

ECN3₂

$19.90

7,000

MMC@₂

$20.20

700

ECN4₂

$19.85

2,000

MMB@₂

$20.20

1,000

MMA@₂

$19.85

400

MMD@₂

$20.25

400

MMC@₂

$19.80

1,000

MMB₂

$20.25

100

MMB₂

$19.75

1,000

ECN3₂

$20.25

7,000

MMB@₂

$19.75

1,000

_{Third Best Bid/Ask}

ECN1₃

$19.90

10,000

ECN2₃

$20.15

1,500

MMA₃

$19.85

500

ECN1₃

$20.20

4,500

ECN3₃

$19.85

800

ECN4₃

$20.25

2,000

ECN2₃

$19.85

1,600

MMC@₃

$20.25

800

SIZE₃

$19.85

1,000

MMB@₃

$20.25

1,000

ECN4₃

$19.80

2,000

MMD@₃

$20.30

600

MMA@₃

$19.80

400

MMA₃

$20.30

900

MMC@₃

$19.75

1,000

MMB₃

$20.30

1,200

MMB₃

$19.65

1,000

ECN3₃

$20.30

8,000

MMB@₃

$19.60

1,000

SIZE₃

$20.20

100

The advantage to the markets of these additional data would be that a much more complete and informative picture of the sources of liquidity would emerge. The NASD proposes only to show for any given MMID only its best price and then only if it is within the three best price levels. That outcome might serve the private commercial objectives of Nasdaq since the absence of MMID disclosure would homogenize under its own SIZE acronym much of the relevant data and would forestall the use of private connections away from Nasdaq to reach three-or-more-deep bids and offers. Not being able to connect outside of the Nasdaq system, in turn, would prevent market participants from avoiding paying SuperMontage fees to Nasdaq, which the NASD identified in its March 2000 private placement memorandum³⁶ as a major risk factor affecting the future commercial success of Nasdaq. The interests of the public, however, diverge markedly from those of Nasdaq in this regard. We believe the Commission should reject the NASD's approach and should insist that Nasdaq disclose the individual, three-deep bid and ask quotations submitted to the NASD in its SIP role by ECNs and market makers.

Footnotes

¹ Letter dated August 7, 2000 from Richard G. Ketchum, President, NASD, to Annette Nazareth, Director of the SEC's Division of Market Regulation, re SR-NASD-99-53-Amendment No. 7 and Response to Comments.

² Rules 11Ac1-1 and 11Ac1-4 under the Exchange Act.

³ See, paragraph (b)(1)(B) of proposed Rule 4710, text following enumeration of (i) through (v) priority list:

The following exceptions shall apply to the above execution parameters. First, if a Nasdaq Quoting Market Participant enters a Non-Directed Order into the system, before sending such Non-Directed Order to the next Quoting Market Participants in queue, the NNMS will first attempt to match off the order against the Nasdaq Quoting Market Participant's own Quote/Order if the participant is at the best bid/best offer in Nasdaq. . . .

⁴ The NASD's further proposed adjustment in Amendment No. 7, to restore the ECN quotation to its place in the queue if the ECN entering a particular Quote/Order indicated that the price improvement offered by the specific Quote/Order exceeded the separate quote-access fee the ECN charges, would not be workable since the access fees Bloomberg Tradebook and, we understand, other ECNs charge are not uniform, but vary considerably from order entrant to order entrant. Accordingly, an ECN would not know in advance whether, for example, a particular price improvement did or did not exceed the access fee that would be charged to the particular order entrant seeking to access its Quote/Order. The NASD proposal, therefore, would require the ECN to assume that the order-entry firm that might be charged only a very small access fee would be charged the highest access fee the ECN would charge to anyone.

⁵ Letter dated July 6, 2000 from Richard G. Ketchum to Belinda Blaine of the SEC's Division of Market Regulation, at page 2.

⁶ Letter dated July 18, 2000 from Richard G. Ketchum to Annette Nazareth, Director of the SEC's Division of Market Regulation, at page 3.

⁷ In saying that, we assume that the computer code actually built into the SuperMontage programming would accurately carry out the SuperMontage Requirements and not deviate from or supplement those requirements in any material way.

⁸ In his August 7, 2000 letter, Mr. Ketchum refers to our August 1, 2000 Issues paper as an August 2, 2000 submission by us.

⁹ Nasdaq reflects, in the very language it uses, this usurpation of its SIP role by its private ECN interests. The terms "quotes" and "quotations", denoting for nearly 30 years a fundamental principle of the Nasdaq market, have been replaced for the first time in the SuperMontage proposal by the term "Quotes/Orders". One searches in vain for a term resembling "Quotes/Orders" in the vocabulary of any other SIP or any other ECN. It seems to be a term that is useful or necessary only to an entity seeking to combine both functions.

¹⁰ In our initial comment letter on the SuperMontage proposal, we referred to this as the problem of the "unattributed attributables", by which we referred to the category of quotations where Nasdaq invites market participants to request MMID attribution but then declines to provide it because the quotations are in excess of the one attributable quotation Nasdaq would allow any market participant to have at any one time. Letter to the Commission from Bloomberg L.P. dated January 11, 2000 in SEC File No. SR-NASD-99-53.

¹¹ See diagram annexed as Attachment B, in which we set forth an illustration of the ways in which the SuperMontage display should be improved.

¹² In its initial SuperMontage filing, the NASD in fact advertised this unfair advantage its own ECN would have:

Nasdaq notes that if a market maker were to place an order into a qualifying ECN, that order would not be displayed in Nasdaq until it was at the top of the ECN's file. In the proposed Nasdaq system, however, the market maker's order in the Nasdaq Order Display Facility will be displayed when it is within the best three price levels on either side of the market. Thus, the Nasdaq Order Display Facility reduces fragmentation and increases transparency in that quotes/orders that might not be displayed to the market because they are in an ECN and not at the top of the ECN's book, may now be displayed in Nasdaq [emphasis added].

Form 19b-4, SR-NASD-99-53 (September 29, 1999) at n.17.

¹³ The NASD noted in its Amendment No. 7 that it had eliminated an overt discrimination against ECNs in its Amendment No. 4, in which the SuperMontage had determined execution priority on the basis of whether an ECN elects order delivery or execution delivery. See August 10 Release in text following n.12. The NASD knows that ECNs require order delivery by virtue of their agency-only business model. The NASD did not explain why it would retain the remaining SuperMontage features that would discriminate against order-delivery ECNs.

¹⁴ See, e.g., letter dated January 17, 1997 from Richard R. Lindsey to Roger D. Blanc re Bloomberg Tradebook, 1997 SEC No-Act. LEXIS 55.

¹⁵ Section 15A(b) sets forth both affirmative requirements for the NASD's rules and negative injunctions against NASD rules that would have improper purposes or effects. Not every NASD rule need promote each affirmative purpose so long as the NASD's rules as a group promote those purposes. At the same time, however, the NASD's proposed rule change would not be consistent with the Exchange Act if it violated even one of the negative injunctions (e.g., the provisions of Section 15A(b)(6) and Section 15A6(b)(9) referred to above). See Matter of Nat'l Ass'n of Securities Dealers, Inc., Order Approving Proposed Rule Change, Securities Exchange Act Release No. 17371 (December 12, 1980), in text following n.71. See also Matter of New York Stock Exch., Notice of Proceeding to Consider Disapproval of Proposed Rule Change, Securities Exchange Act Release No. 12249 (SR-NYSE-76-5) (March 23, 1976):

In the view of the Commission, a proposed rule change would not be consistent with the Act and the rules and regulations thereunder if, among other things, the Commission could not make the determinations required under Section 6(b) of the [Exchange] Act with respect to the rules of an exchange which included the proposed rule change prior to registration of an exchange.

Id. in text following n.8.

¹⁶ Matter of The Rules of the New York Stock Exch., 10 SEC 270 (October 4, 1941) (NYSE rule prohibiting dealings on other markets declared to be against public interest and illegal).

¹⁷ Matter of New York Stock Exch., Notice of Proceeding to Consider Disapproval of Proposed Rule Change, Securities Exchange Act Release No. 12249 (SR-NYSE-76-5) (March 23, 1976), 1976 SEC LEXIS 2116.

¹⁸ Id., in text following n.8.

¹⁹ Indeed, the argument that the SuperMontage would usefully resolve problems of "fragmentation" has a close resemblance to arguments in favor of former NYSE Rule 390. Fragmentation to some is competition to others. As Chairman Levitt observed:

I can't think of a more propitious time to reconsider other restrictions that distort competition and introduce artificial costs. One, in particular, comes to mind-NYSE Rule 390.

This rule has long prohibited NYSE members from dealing in listed securities off an exchange. For years, proponents have argued that Rule 390 prevents fragmentation. Others contend that the rule is an anticompetitive use of market power by a dominant market. As I see it, Rule 390 may very well be on its ninth life. Now is the time to ask ourselves: is there a valid justification for a rule that appears to be more a barrier than a benefit? And how, under any circumstances, could such an anticompetitive rule be sustained should the NYSE become a for-profit corporation? While rulemaking is certainly an option, one way or another, Rule 390 should not be part of our future.

Chairman Arthur Levitt, "Dynamic Markets, Timeless Principles," Speech at Columbia Law School (September 23, 1999).

²⁰ Cf. L. Brandeis, Other People's Money and How the Bankers Use It 62 (1914): "Publicity is justly commended as a remedy for social and industrial diseases. Sunlight is said to be the best of disinfectants; electric light the most efficient policeman."

²¹ Securities Exchange Act Release No. 42166 (November 22, 1999), Section II(B).

²² As the Commission is aware, the General Instructions to Form 19b-4, 5 Fed. Sec. L. Rep. (CCH) ¶ 32,356, are explicit on the point. They provide, with respect to "Information to be Included in the Completed Form," as follows:

4. Self-Regulatory Organization's Statement on Burden on Competition

State whether the proposed rule change will have an impact on competition and, if so, (i) state whether the proposed rule change will impose any burden on competition or whether it will relieve any burden on, or otherwise promote, competition and (ii) specify the particular categories of persons and kinds of businesses on which any burden will be imposed and the ways in which the proposed rule change will affect them. If the proposed rule change amends an existing rule, state whether that existing rule, as amended by the proposed rule change, will impose any burden on competition. If any impact on competition is not believed to be a significant burden on competition, explain why. Explain why any burden on competition is necessary or appropriate in furtherance of the purposes of the [Exchange] Act. In providing those explanations, set forth and respond in detail to written comments as to any significant impact or burden on competition perceived by any person who has made comments on the proposed rule change to the self-regulatory organization. The statement concerning burdens on competition should be sufficiently detailed and specific to support a Commission finding that the proposed rule change does not impose any unnecessary or inappropriate burden on competition [emphasis added].

Id. at p. 22,318.

²³ 2 F.3d 453 (D.C. Cir. 1993).

²⁴ See also Section 3(f) of the Exchange Act.

²⁵ To assist the Commission in its adjudicatory proceedings under the Exchange Act, the NASD must provide an adequate basis for comment on its rule proposals and, where significant competitive issues are involved, must provide an opportunity for the public to comment meaningfully on the issues involved. Perfunctory recitals do not provide that basis. See Connecticut Light and Power Co. v. NCR, 673 F.2d 525, 530-31 (DC Cir. 1982).

²⁶ The rigorous approach built into the Commission's Rule 19b-4 and Form 19b-4 responds to a direct, specific and unequivocal congressional mandate. See Securities Acts Amendments of 1975, Report of the Senate Comm. on Banking, Housing and Urban Affairs to Accompany S.249, S. Rep. No. 94-75, 94th Cong., 1st Sess. 29-30 (1975):

In order to facilitate expeditious Commission review and evaluation of [proposed rule changes] and to assure informed public comment on them, Section 19(b)(1) would require all self-regulatory organizations to file with the SEC in connection with any proposed rule change a "concise general statement of the basis and purpose" of the proposed rule change. It is the Committee's intention in adopting this standard to hold the self-regulatory organizations to the same standards of policy justification that the Administrative Procedure Act imposes on the SEC.

. . . [T]he Committee believes interested persons should have a meaningful opportunity to obtain accurate information about proposed changes in self-regulatory rules and to comment on the need or justification for these changes. Section 19(b)(1) would require the SEC to give notice and provide an opportunity for interested persons to participate in the process of reviewing a proposed change in a self-regulatory organization's rules. In addition, this section would require that all comment and all correspondence between the SEC and the self-regulatory agency concerning the proposal be available for public inspection. . . .

. . . The Committee believes the Commission has a responsibility to see that self-regulatory rules are fully responsive to regulatory needs. By explicitly providing that the Commission's oversight authority encompasses major self-regulatory policies, the bill would make this responsibility clear and substantially decrease the possibility of slippage between regulatory need and self-regulatory performance [emphasis added]. . . .

²⁷ Cf. Home Box Office, Inc. v. FCC, 567 F.2d 9 (1977). The anticompetitive aspects of the SuperMontage are of particular public importance in light of the market power Nasdaq has an exclusive processor of securities information. The Congress, in enacting the Securities Acts Amendments of 1975, warned particularly against possible abuses of that market power:

The Committee believes that if economics and sound regulation dictate the establishment of an exclusive central processor for the composite tape or any other element of the national market system, provision must be made to insure that this central processor is not under the control or domination of any particular market center. Any exclusive processor is, in effect, a public utility, and thus it must function in a manner which is absolutely neutral with respect to all market centers, all market makers, and all private firms. Although the existence of a monopolistic processing facility does not necessarily raise antitrust problems, serious antitrust questions would be posed if access to this facility and its services were not available on reasonable and nondiscriminatory terms to all in the trade or if its charges were not reasonable. Therefore, in order to foster efficient market development and operation and to provide a first line of defense against anti-competitive practices, Sections 11A(b) and (c)(1) would grant the SEC broad powers over any exclusive processor and impose on that agency a responsibility to assure the processor's neutrality and the reasonableness of its charges in practice as well as in concept [emphasis added].

Securities Acts Amendments of 1975, Report of the Senate Comm. on Banking, Housing and Urban Affairs to Accompany S.249, S. Rep. No. 94-75, 94th Cong., 1st Sess. 11-12 (1975).

²⁸ Rule 19b-4(c).

²⁹ Rule 19b-4(b)(1).

³⁰ The Commission often has had to take action to curb the anti-competitive tendencies of the NASD and other self-regulatory organizations. See, e.g., the Ætna proceeding, Securities Exchange Act Release No. 9632 (June 7, 1972) (Commission partially abrogated, as in excess of NASD authority and anti-competitive, part of former Art. III, Sec. 25 of NASD Rules of Fair Practice [now NASD Conduct Rule 2420]); Plaza Securities Corporation, Securities Exchange Act Release No. 10643 (February 14, 1974) (NASD disciplinary order under former section 25 set aside); Securities & Exch. Comm'n, Report Pursuant to Section 21(a) of the Securities Exchange Act of 1934 Regarding the NASD and the Nasdaq Market, 1996 SEC LEXIS 2123 (August 8, 1996); the "Multiple Trading Case", Matter of The Rules of the New York Stock Exch., 10 SEC 270 (October 4, 1941) (NYSE rule prohibiting dealings on other markets declared to be against public interest and illegal); Securities Exchange Act Release No. 12737 (August 25, 1976), 1976 SEC LEXIS 984 (order disapproving proposed NYSE rules 309 and 310); Securities Exchange Act Release No. 12249 (SR-NYSE-76-5) (March 23, 1976), 1976 SEC LEXIS 2116 (entry of order to disapprove NYSE Public Limit Order Protection Rule); and Matter of Applications of William J. Higgins and Michael D. Robbins, Securities Exchange Act Release No. 24429 (May 6, 1987), 1987 SEC LEXIS 1879 (NYSE action denying access on basis of policy not reasonably and fairly implied by existing rule set aside). See also Securities Exchange Act Release No. 12994 (November 18, 1976) in text accompanying nn. 14-23, and authorities cited therein. Cf. Silver v. New York Stock Exch., Inc., 373 U.S. 341 (1963).

³¹ While a self-regulatory organization may have some concerns that disclosure of the particulars about a new system may put at some competitive disadvantage a private, speculative commercial venture in which it has an interest, such as the SuperMontage ECN, those concerns must be subordinated to the needs of the Commission to understand the terms of substance of a proposed rule change if it is to determine whether the proposed rule change is consistent with the provisions of the Exchange Act applicable to the self-regulatory organization. Also, and equally important, the information must be made available to interested members of the public if the Commission is to provide the opportunity for submission of data, views and arguments the Congress envisioned under Section 19(b) of the Exchange Act. That is all the more important in a case such as this, where the NASD is attempting through the rule filing to use its quasi-governmental regulatory power and its monopoly market power to bestow on its private venture unique and inappropriate commercial advantages.

³² Securities Exchange Act Release No. 35123 (December 20, 1994), in text following n.35.

³³ Connecticut Light and Power Co. v. NRC, 673 F.2d 525, 530-31 (D.C. Cir. 1982).

³⁴ The Nasdaq Stock Market, Inc., Private Placement Memorandum (March 10, 2000) at A-20, attached to National Association of Securities Dealers, Inc., Notice of Special Meeting of Members to be Held on April 14, 2000.

³⁵ Securities Acts Amendments of 1975, Report of the Senate Comm. on Banking, Housing and Urban Affairs to Accompany S.249 , S. Rep. No. 94-75, 94th Cong., 1st Sess. 11-12 (1975).

³⁶ The Nasdaq Stock Market, Inc., Private Placement Memorandum (March 10, 2000) at A-20, attached to National Association of Securities Dealers, Inc., Notice of Special Meeting of Members to be Held on April 14, 2000.


Figure 1

Displayed			Displayed
Best Bid	$20.00	9,000	Best Ask	$20.05	4,400
MMA	$20.00	1,000	ECN2	$20.05	1,000
SIZE	$20.00	5,700	SIZE	$20.05	3,400
MMB@	$20.00	1,500			4,400
ECN1	$20.00	800
		9,000

Displayed			Displayed
2^nd Best Bid	$19.95	15,000	2^nd Best Ask	$20.10	5,000
MMB	$19.95	1,000	ECN1	$20.10	800
ECN2	$19.95	500	MMB@	$20.10	500
ECN3	$19.95	1,500	MMD@	$20.10	3,400
MMC@	$19.95	100			4,700
		3,100
Hidden			Hidden
MMA₂	$19.95	900	ECN2₂	$20.10	300
ECN1₂	$19.95	11,000

Displayed			Displayed
3rd Best Bid	$19.90	25,000	3rd Best Ask	$20.15	15,000
ECN4	$19.90	1,000	MMA	$20.15	500
MMA@	$19.90	1,000	ECN4	$20.15	2,500
		2,000	MMC@	$20.15	100
Hidden			MMB	$20.15	1,000
SIZE₂	$19.90	1,000			4,100
ECN1₃	$19.90	10,000
ECN2₂	$19.90	5,000	Hidden
ECN3₂	$19.90	7,000	ECN1₂	$20.15	9,000
		23,000	SIZE₂	$20.15	400
			ECN2₃	$20.15	1,500
					10,900

Bid			Ask
MMA₂	$19.95	900	ECN2₂	$20.10	300
ECN1₂	$19.95	11,000	ECN1₂	$20.15	9,000
SIZE₂	$19.90	1,000	SIZE₂	$20.15	400
ECN2₂	$19.90	5,000	ECN4₂	$20.20	1,200
ECN3₂	$19.90	7,000	MMC@₂	$20.20	700
ECN4₂	$19.85	2,000	MMB@₂	$20.20	1,000
MMA@₂	$19.85	400	MMD@₂	$20.25	400
MMC@₂	$19.80	1,000	MMB₂	$20.25	100
MMB₂	$19.75	1,000	ECN3₂	$20.25	7,000
MMB@₂	$19.75	1,000