June 4, 1999
Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, DC 20549
Attention: Mr. Jonathan G. Katz, Secretary
Re: SEC File Nos. SR-NASD-99-11 and SR-NASD-98-17
Ladies and Gentlemen:
Bloomberg L.P. appreciates the opportunity to comment, in response to the Commission's request in Securities Exchange Act Release No. 41296 (April 15, 1999) (the "Release"), on the proposed rule change submitted by the National Association of Securities Dealers, Inc. (the "NASD"), through its wholly owned subsidiary, The Nasdaq Stock Market, Inc. ("Nasdaq"). The proposed rule change (the "NNMS proposal") would modify Nasdaq's Small Order Execution System ("SOES") and SelectNet Service ("SelectNet"). In that connection, the Commission reopened the comment period on Nasdaq's Integrated Order Delivery and Execution System proposal set forth in Securities Exchange Act Release No. 39718 (March 4, 1998)
(SR-NASD-98-17) (the Integrated Order Delivery and Execution System and Central Limit Order Book proposal hereinafter called the "CLOB proposal," and, together with the NNMS proposal, the "Proposed Rule Changes") and we are commenting on that proposal as well. In addition to the comments below, we refer the Commission to our letter to the Commission dated May 12, 1998 relating to the "Next Nasdaq System" (the "May 1998 Letter"), a copy of which is attached, as well as to certain views summarized below, with which we concur, that were expressed in the letter of Charles Schwab & Co., Inc. and its affiliate, Mayer Schweitzer, Inc. (together, "Schwab), to the Commission dated May 26, 1998 (File No. SR-NASD-98-17) (the "Schwab Letter").
Implementation of certain aspects of the Proposed Rule Changes could substantially affect the business of Bloomberg's wholly owned subsidiary, Bloomberg Tradebook LLC ("Tradebook"). As the Commission is aware, Tradebook operates a proprietary electronic communications network ("ECN") pursuant to a no-action letter from the staff of the Commission's Division of Market Regulation. Through Tradebook, Bloomberg offers its institutional and broker-dealer customers, and other broker-dealers that access the Tradebook system via Nasdaq's SelectNet, the opportunity to buy and sell equity securities via the BLOOMBERG service.
Nasdaq put forward the CLOB proposal last year. That proposed rule change met with significant opposition and concern among market participants, many of whom urged the Commission to conduct a full fact-finding investigation of the proposal's effects on competition, the marketplace, the securities industry and investors. The Commission has not yet approved the CLOB proposal. In the meantime, Nasdaq has recently proposed another rule change to introduce the NNMS proposal. In light of the NNMS proposal, the Commission has reopened the public comment period for the CLOB proposal.
The two NASD proposals attempt to serve many of the same improper objectives. Neither would benefit the marketplace or serve the public interest or the protection of investors. In fact, the proposals would have substantially deleterious effects: stifling competition, hampering innovation, reducing liquidity and diluting efficiency of the U.S. capital markets. In addition, the proposals are not supported by sufficient factual or legal bases. They therefore should not be approved. We call upon the Commission to decline to approve the proposals and to prevent Nasdaq from effectively using its quasi-governmental power to enhance its own competitive posture at the expense of NASD members.
Nasdaq National Market Execution System ("NNMS")
Overview. The Commission should not approve the NASD's NNMS proposal, which would combine SelectNet and SOES and create the Nasdaq National Market Execution System on an interim basis pending review and approval of the proposed Integrated Order Delivery System. The NNMS proposal and operations are unattractive to ECNs, and it is highly unlikely that ECNs will participate in the "full" NNMS, since the NNMS would place ECNs at a competitive disadvantage. The NASD has not disputed this view; in fact, in recent public meetings with market participants the NASD has stated that it believes ECNs will choose order-entry participation. It is accordingly clear that the NNMS proposal would impose substantial burdens on competition. The NASD has not demonstrated, and we believe cannot demonstrate, that those burdens are necessary or appropriate in furtherance of the purposes of the Securities Exchange Act of 1934 (the "Exchange Act"). As a result, the Proposed Rule Changes would violate Section 15A(b)(9) of the Exchange Act.
The NNMS would create a central market execution system in two tiers. Participation would be mandatory for market makers and "voluntary" for ECNs. Nasdaq states that two types of participation would be offered to ECNs: "full" and "order entry". "Full" participation would require ECNs for the first time to be subject to automatic executions, which would put ECNs at a severe competitive disadvantage. Order-entry participation would be a continuation of current SelectNet linkage and functions, but would marginalize the contribution ECNs could make to the marketplace. In turn, NNMS has inherent design flaws because it would shut down when an ECN was alone at the inside quotation and therefore would become increasingly ineffective as ECNs continued to grow.
Full Participation in NNMS. Full participation in NNMS would allow ECNs to connect with NNMS and allow the NNMS to "sweep" the ECNs' top-of-the-file orders into the NNMS, a function that would hit or take all market-maker, ECN or NNMS quotations at the best bid or ask price. Access to the sweep in NNMS, however, would involve a trade-off because of the following extremely disadvantageous effects of full participation NNMS on ECNs: (1) an ECN's reserve quotations would not participate in the sweep, and this would limit the value of the sweep since it would not reach the reserve of a market maker or an ECN; (2) ECNs would be subject to potential double executions; and (3) ECNs would be forced to provide access through NNMS to brokers who refuse to pay ECN access fees.
Market makers and institutional customers of ECNs often prefer to trade on an ECN because of the additional services and features offered on ECNs. Through intense competition, ECNs have developed a variety of innovative functions and capabilities to allow traders to tailor their trading process to their individual needs and strategies. One such feature is a reserve quotation. Nasdaq has stated that in NNMS any ECN reserves would be bypassed, but the reserve feature of NNMS itself on NNMS would function. As a result, Nasdaq would in effect preference its own additional functions in NNMS, at the expense of its "full participants".
Nasdaq states in the NNMS proposal that the NNMS would eliminate potential "dual liability" of market makers. We emphasize, however, that the NNMS modifications to Nasdaq's negotiation and automatic execution systems would not do so for ECNs, nor would they limit dual liability for any market maker that automatically executed customer orders immediately, as an ECN does. As discussed in our May 1998 Letter with respect to the CLOB proposal, since NNMS would subject full participant ECNs to executions instead of orders, ECNs could receive double executions resulting from the time lag between execution of an order (or an update of a quotation) on the ECN's own system and receipt of an execution for the same order from Nasdaq some time later. Market makers have reported that, in some cases, SOES messages can be delayed up to 15 seconds. Since Nasdaq does not propose to update its technology in connection with NNMS, the response time of NNMS would presumably be just as slow as Nasdaq's current technology and would not remove this liability problem for ECNs.
A third disadvantage to an ECN's being a "full" NNMS participant would be that ECNs would have to provide access to brokers that refuse to pay ECN access fees, which would reduce the incentive of any broker to pay access fees. This disadvantage would also apply to a market maker's agency quotations. Since a full participant ECN would receive executions from NNMS instead of orders, an ECN or market maker would have no recourse if a broker did not pay the appropriate access fee. In effect then, an ECN would have to provide its services and liquidity for free. There would result a significant adverse effect on competition, one that is not even mentioned in either the NNMS proposal or the NASD's access fee proposal. For the reasons outlined above, we believe that few ECNs would elect "full" participation in NNMS.
Order-Entry Participation in NNMS. If an ECN wished to avoid the pitfalls of full participation in NNMS, it would be penalized for that choice. Order-entry participation would omit from the NNMS "sweep" all orders in the ECN. An ECN that chose this level of participation could be marginalized by not having its full range of available liquidity accessible via the sweep.
On the other hand, it may be the NNMS that would be marginalized. If, as the NASD expects, most ECNs were to elect to be "order-entry" rather than full-entry participants in NNMS, the NNMS "sweep" would shut down during the significant amount of time that an ECN is alone at the best bid or offer, just as the SOES system does today. As ECNs grow, the NNMS tool would become less and less effective. We therefore believe that the NNMS is flawed in its design and is ill-conceived.
NNMS is Anti-Competitive. As we and others have stated in previous comment letters to the Commission and as we reiterate below, if Nasdaq and the NASD, as the applicable self-regulatory organization, mandate a monopolistic central execution system, such as NNMS, with which all NASD-regulated broker-dealers must comply, ECN innovation could be driven out of the market, and the marketplace and investors would suffer as a result. Under the NASD's own statutory mandate in Section 15A(b) of the Exchange Act, and the legislative intent underlying those requirements, the NASD is not free to use quasi-governmental power to damage what it seems to view as its private sector competitors, which are its own members.
In this respect, the same issues and concerns apply to the NNMS as we describe below in our comments to the CLOB proposal.
A Better Solution than NNMS. Instead of approving a system that would create disincentives for ECNs and other market participants to innovate, to offer investors cutting-edge technology and capabilities and to provide additional liquidity and transparency, the Commission should encourage a different type of market structure. The combination of private and public linkages that has formed the market network since the advent of the Order Handling Rules is the best model for growth and development of the U.S. capital markets into the next century. The Commission set the stage for this network with the Order Handling Rules and it should not allow the NASD to reverse the progress that the markets have made since then towards more open and efficient operation.
If the NASD believes it is important for the sweep to work, the only appropriate solution would be to send SelectNet order messages to those entities, like ECNs, that respond immediately and automatically to orders entered, and are audited by the Commission and the NASD for their responsiveness to such orders, and to send SOES messages to market makers that do not respond immediately and automatically. Under this approach, if an ECN or market maker did not respond immediately to all orders, it would open itself up to receiving executions. This decentralized market model would allow innovation and competition to flourish, resulting in a more efficient marketplace.
The Commission should encourage Nasdaq to upgrade its technology and to replace SelectNet with modern order-routing technology. The NASD designed SelectNet for negotiation and communications between market participants several years ago, not for its current function of order routing in a very quick market. From time to time the NASD has modified SelectNet on an ad hoc basis to add additional capabilities, but has not revamped the software system to correspond to SelectNet's current role. If Nasdaq were to replace SelectNet in its entirety with new technology, Nasdaq would be able to eliminate anachronisms such as the ten-second rule that exists solely because SelectNet cannot properly handle order flow in today's markets.
CLOB Proposal (Integrated Order Delivery and Execution System) (SR-NASD-98-17)
Bloomberg commented on the NASD's proposed Integrated Order Delivery and Execution System, the CLOB proposal, in our May 1998 Letter. We hereby confirm the concerns summarized in that letter and raise additional issues for the Commission's consideration in connection with the CLOB proposal.
The Commission should not approve or allow the NASD to adopt the CLOB proposal. In addition, we believe that the pilot program proposed with the re-opening of the comment period for the CLOB proposal would be an unnecessary use of time and resources. The NASD has not justified the CLOB proposal. Until it provides such information as we and several other commenters requested during the prior comment period and as we again request in this letter, the Commission should not allow a pilot program.
Interaction and Effect with Other NASD Proposals. The CLOB proposal is now over one year old and the Commission recently re-opened the comment period. Since its filing, however, the NASD has proposed several other rule changes, some of which (for example, OptiMark) have been approved and put into effect. The CLOB proposal does not consider the interaction with, or effect of, these other adopted and pending rule changes on the CLOB proposal. The NASD has not updated the CLOB proposal to reflect these developments and has not offered guidance on the effect of the CLOB proposal on its subsequently approved and pending rule changes. In this respect, Nasdaq has not complied with Form 19b-4 under the Exchange Act, and its filing is, as a result, materially deficient as a matter of law.
For example, in connection with the market-maker dual acronym and access fee proposal (SR-NASD-99-16 and SR-NASD-99-09), it is not clear whether market makers and ECNs that submit orders to the CLOB would be allowed to charge access fees on these agency orders. If not, members that wish to continue to charge access fees would effectively be denied the CLOB. The CLOB would thus not be a utility for NASD members, but rather a direct competitor with NASD members, both market makers and ECNs, as discussed in the May 1998 Letter and in the Schwab Letter. In addition, as stated in our comment letter relating to the access fee proposal, the effect of NNMS on market makers' ability to charge access fees is ambiguous at best. As explained in that letter, market makers may not be able to charge access fees if NNMS were to be implemented due to the interaction of the two proposals.
In the absence of a reasoned presentation and discussion of these considerations and effects, market participants and the public are unable to make fully informed judgments about the CLOB proposal or other pending NASD proposals. Without that full consideration, the CLOB and NNMS proposals are procedurally and substantively inadequate. The Commission should require additional information and research of the NASD for its own review and also for the public's benefit and review before further analyzing the proposals and certainly before approving any such changes.
CLOB Should Not Be Approved.
A centralized execution hub, such as the proposed Nasdaq CLOB, would be harmful to the marketplace. Reliance upon a single source for everything from order flow to execution to communications technology would cause the U.S. capital market structure to be subjugated to the NASD's market design and be limited, in turn, by the NASD's chosen technology, much of which is substantially behind that of other service providers in the securities markets and the securities industry.
The details of the Central Order Limit Book proposal determine whether it would assert a centralized, and therefore we believe negative, effect on the market. The Schwab Letter clearly sets forth examples of these details and articulates that it is the combined effect of these details that would create a centralized influence on the market. Even if such effect is inadvertent, it is nevertheless real. The details of the CLOB need to be examined closely to evaluate the worthiness of the proposal.
1. Insufficient Justification for Change. The CLOB proposal would create a fundamental structural realignment of Nasdaq and the NASD's role in the marketplace. The vice inherent in the proposal is that it would present imposing and unjustified competitive obstacles to the market makers and to the ECNs. The NASD has not, and we believe cannot, square those burdens with the statutory standards of Section 15A(b)(6) of the Exchange Act, that the NASD's rules be designed to remove impediments to and perfect the mechanism of a free and open market, and Section 15A(b)(9) of the Exchange Act, that the NASD's rules not impose any burden on competition that is not necessary or appropriate in furtherance of the purposes of the Exchange Act. As stated in the Schwab Letter and our May 1998 Letter, the NASD has not provided a sufficient legal basis to permit the Commission to approve the propose rule changes.
2. Proposed Change is Unnecessary and Negative. As the Schwab Letter notes, the CLOB is unnecessary because ECNs and market makers have invested significant funds and resources to create their own limit order books and provide efficient order handling capabilities. Spurred by the Order Handling Rules, these tools provide ready access to the market for customers and investors through limit order display and ECN display. The CLOB would not provide additional capability to the market and the NASD does not provide any factual basis for a belief that the current system of linked networks is not adequate. Indeed, the CLOB would cause serious market disruptions and unintended negative consequences for market liquidity, transparency, competition and transaction costs, ultimately hurting the investing public. The CLOB's centralization would be harmful to the market and in stark contrast to the diverse, private network linkage system that has blossomed in the wake of the Order Handling Rules. Interaction of order flow from multiple sources is a sound model, one that should be further developed instead of abandoned.
3. Anti-Competitive Effects. We share Schwab's concerns about the NASD's ability to establish the CLOB with the financial support of its members, which include market participants and ECNs that the CLOB would place at a severe competitive disadvantage. The Commission and the Congress have repeatedly recognized the value of competition instead of a centralized, single service provider to ensure innovation, service, capacity, value and efficiency. The CLOB proposal would take the market in the opposite direction.
Nasdaq plans to provide advantages for the CLOB that it will not permit for market makers and ECNs. The CLOB would display the full book rather than the top of the file or best bid or offer, providing users more visibility and greater trade-through protection. In its Form 19b-4 filings, the NASD does not justify, and we believe cannot justify, giving its own system this advantage over NASD members' own competitive quotations and over the existing private sector ECNs. For example, OptiMark would protect the entire Nasdaq book, but would protect only the top-of-file for market makers and ECNs, a fact that has met with considerable protest and criticism from attendees at the NASD's education seminars in New York last month. The CLOB would also give a time advantage to orders in its own book that are below the top-of-file. This inequity could give time priority to inferior-priced orders in fast markets because of the time it takes market makers and ECNs to update their quotations.
Unlike the Order Handling Rules, which improved liquidity and tightened the markets, the CLOB would draw order flow away from market makers and ECNs and substantially lessen liquidity. The CLOB would ultimately discourage market makers from committing capital to thinly-traded issues. In addition, ECNs provide a significant liquidity function and the CLOB could not adequately replace or replicate that capacity. ECNs anonymously and electronically access all liquidity in the Nasdaq market and immediately display customer limit orders or execute marketable limit orders. Using innovative trading technologies, ECNs achieve superior order handling and executions.
As suggested above, a fundamental harm that would arise from implementation of the NASD's CLOB proposal would be that the entire market for Nasdaq securities would become dependent on the capacity, integrity and security of a single, largely antiquated system. We believe that investors and the marketplace benefit from a range of alternative systems for routing, displaying and executing orders. Today, we have a decentralized order-routing and -execution process, through multiple direct links among market makers and ECNs. If Nasdaq were made the only choice, reliability and redundancy and the cleansing and strengthening forces of competition would be severely limited and compromised.
A generation ago, in the Future Structure Statement, the Commission endorsed in the exchange context the concept of a consolidated limit order book as a means of exposing all supply to all demand for a given security and providing for system-wide protection of customers' limit orders. Today, intervening technology has made it no longer necessary to envision a single, monopolistic limit order book to achieve those objectives. The rapidly declining costs of telecommunications bandwidth have made it possible to build and maintain redundant, competitive systems to handle orders and provide market mechanisms without relying on a single service provider. As a result, the public-utility model that informed the market structure debate twenty-five years ago is no longer necessary or even useful.
To be sure, building and maintaining duplicative public utility systems for the retail distribution of electric power, gas or steam may still not be economically sensible or environmentally acceptable. The securities markets are different, however. With the rapid deployment of new technology, electronically interconnected redundant systems that handle and execute orders serve a purpose. They provide a more competitive, innovative and stronger market structure than would ever be possible if the Commission were to allow the NASD to perfect a governmentally mandated monopoly through its unitary CLOB.
4. Inherent Conflicts of Interest. We take issue with the NASD's use of its members' funds to subsidize, develop and operate a CLOB in competition with its member firms. The NASD by definition would be conflicted in its loyalties to monitor and discipline its members with respect to order-routing obligations and to operate and profit from the CLOB. While we have every confidence in the integrity of the individuals at the helm of the NASD and its various subsidiaries, the conflict between the regulatory function and the promotional, business-oriented function, presents continual risks.
5. Filings are Deficient and Non-Compliant. As we described in the May 1998 Letter and as stated in the Schwab Letter, the NASD did not comment on the burdens on competition with respect to either the CLOB or the NNMS. As the Commission knows, self-regulatory organizations must comment pursuant to Rule 19b-4 on the specific and detailed effects and burdens on competition of any proposed rule change. The Congress has emphasized the importance of these findings to the Commission's obligations in approving rule changes and overseeing the marketplace. In the CLOB proposal and the NNMS proposal, the NASD has merely restated the language of Section 15A(b)(9) of the Exchange Act with the cursory conclusion that these proposals will have no effect or burden on competition. In this respect the proposals are materially deficient and cannot serve as the basis for adequate public comment or for Commission review and approval. These perfunctory statements do not provide the Commission with a legally sufficient basis to approve the proposed rule changes. As a result, a Commission order approving the proposed rule changes on the basis of the current record would be reversible as a matter of law.
The Commission should reject the NNMS proposal and the CLOB proposal because the proposals offer no additional benefits, but hamper competition to the detriment of investors and the marketplace. We believe that the current varied linked system of market makers and ECNs offers the best solution in a naturally competitive environment: innovative trading tools that anonymously and electronically work orders quickly and discretely across multiple sources of liquidity, with minimal market risk and impact.
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,
By: __/s/ Kevin M. Foley____
Kevin M. Foley
cc(w/att).: The Hon. Arthur Levitt, Chairman
The Hon. Isaac C. Hunt, Jr., Commissioner
The Hon. Norman S. Johnson, 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
Harvey J. Goldschmid, Esq., General Counsel
Mr. Richard G. Ketchum
The National Association of Securities Dealers, Inc.