May 21, 1999

Mr. Jonathan G. Katz
Office of the Secretary
Mail Stop 6-9, Room 6507
Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, D.C. 20549-0609

Re: File No. SR-NASD-99-11; File No. SR-NASD-98-17

Dear Mr. Katz:

Knight Securities, Inc. ("Knight") submits this letter in response to the NASD's recent proposal (File No. SR-NASD-99-11) to modify the Small Order Execution System and SelectNet, in which the Commission determined to re-open the comment period on the NASD's previous proposal to establish a Central Limit Order Book. In response to prior comments from the securities industry on the impact of the proposed limit order facility, which were almost universally critical, the SEC has requested comments on whether the Central Limit Order Book should be adopted on a "pilot basis." For the reasons set forth below, Knight believes that the Central Limit Order Book proposal is anti-competitive and ill conceived and should not be approved, even as a pilot program.

Knight is the leading market maker in securities listed in The Nasdaq Stock Market and on the NASD OTC Bulletin Board Service, trading in approximately 7,000 issues. In April, Knight's Nasdaq volume was about 3.9 billion shares for an approximate 10.2% market share, while its OTCBB volume was about 450 million shares for an approximate 25% market share. Thus, we believe the Commission will find our comments useful in its consideration of the proposed Central Limit Order Book.

The Commission has re-opened comment on the Central Limit Order Book in light of another recent NASD proposal for an agency quote display structure. On March 11, 1999, the NASD, through Nasdaq, also proposed to permit the separate display of customer orders by market makers in Nasdaq through a market maker agency identification symbol. (See File No. SR-NASD-99-09) (the "Agency Quote" proposal). Under the Agency Quote proposal, a market maker in Nasdaq National Market Securities would be able to display in Nasdaq a second quotation, separate from its proprietary quotation, for the purpose of displaying customer interest. The Agency Quote proposal, in combination with the proposal to modify SOES and SelectNet, would provide market makers and other broker-dealers with additional resources to display and execute limit orders and to comply with their obligations under the Order Handling Rules.

Knight's Comments

The NASD has not demonstrated a need for a Nasdaq limit order facility nor has it analyzed or reviewed the present resources, or those being contemplated, that are devoted to limit order protection by various industry participants. In fact, the proposed Central Limit Order Book would place an unwarranted burden on competition that is not in the interest of investors. Of further significance, the NASD has neither provided information about the cost of implementing or operating a Central Limit Order Book nor determined who will bear those costs.

Further deliberations on a proposed Central Limit Order Book, including adoption of a Central Limit Order Book on a "pilot basis", are inappropriate given the overwhelmingly negative response that the proposal received from the securities industry and given the existence of a much less disruptive alternative in the form of the Agency Quote proposal. Because the principal problem with the proposed Central Limit Order Book is that it would result in a regulatory body having a monopoly over the limit order business, testing the book on a limited, short-term, "pilot" basis would not contribute to an evaluation of the potential anti-competitive effects of full-scale implementation.

The Agency Quote proposal does not give rise to the same conflicts of interest inherent in a self-regulatory organization, with oversight inspection and enforcement authority, operating a business in competition with its members. Accordingly, if the NASD desires to introduce a new limit order protection program on a pilot basis, the Agency Quote would be a far better candidate. Agency Quote is an important market structure change in and of itself.

A. No Need Exists for a Separate and Distinct Nasdaq Central Limit Order Book

The NASD has not demonstrated how the proposed Central Limit Order Book would enhance limit order protections already available to investors through market makers and ECNs. The Nasdaq Central Limit Order Book would duplicate services that are already available to customers in that Nasdaq already incorporates a substantial number of limit order books operated by ECNs and market makers. Nasdaq has not demonstrated that a need exists for a separate and distinct Nasdaq Central Limit Order Book, particularly one with serious anti-competitive potential.

With the adoption of the Order Handling Rules in 1996, which included amendments to Rule 11Ac1-1 ("Firm Quote Rule") and the adoption of Rule 1lAcl-4 ("Display Rule"), investors were assured limit order competition and display of the best prices to the entire market. The Display Rule requires market makers to display their customer limit orders that: (1) are priced better than a market maker's quote; or (2) add to the size of a market maker's quote when the market maker is at the best bid or best offer in Nasdaq. The amendments to the Firm Quote Rule require a market maker to make publicly available any superior prices that it privately quotes through an ECN by either: (1) changing its quote to reflect the superior price in the ECN; or (2) delivering better priced orders to an ECN that disseminates these priced orders to the public quotation system and provides broker-dealers equivalent access to these orders. Market makers and ECNs have already invested heavily in internal order handling technology to ensure compliance with the Order Handling Rules. The NASD's most recent proposal to modify features of SOES and SelectNet and improve automatic executions should also enhance limit order protection. Among other things, by facilitating automatic executions and reducing time delays between system executions, the proposal to modify SOES and SelectNet will allow Nasdaq participants to execute against customer limit orders more rapidly.

To provide market makers with more flexibility in determining how to handle customer limit orders, Nasdaq has filed the Agency Quote proposal which will permit OTC market makers to display their customer orders through a market maker agency identification symbol. Instead of having to display a customer limit order in their proprietary quote or in an ECN, market makers would be able to display the order in the Agency Quote. The Agency Quote would be used to display customer orders, but not the market maker's own proprietary interest or the proprietary interest of another market maker in the security at issue. In the release announcing the proposal, the NASD noted that its previous proposal for a limit order book, designed to address some of the same issues, had not obtained industry support.

One of the principal objectives of the NASD's proposal for permitting market makers to publish agency quotations in Nasdaq is to provide an additional resource for the display and automatic execution of customer limit orders. This is in recognition of the notion that market makers and other broker-dealers should be allowed to retain their limit order business and to allay the concerns of NASD members that Nasdaq should not operate a limit order book that competes with its members. The NASD affirmed in its filing that the Agency Quote proposal meets the objectives of Section 11A of the Securities Exchange Act of 1934 ("Exchange Act") and the Order Handling Rules, in particular the Display Rule.

In light of this recent proposal to permit the separate display of customer limit orders, it is vexing that thew SEC would re-open for comment an unnecessary and universally unpopular proposal for a limit order book. Without a compelling need for a separate limit order book, it would be exceedingly unwise and unnecessary to spend industry resources pursuing this alternative, particularly where the NASD would operate the limit order facility in competition with its own members, creating irreconcilable conflicts of interest that will ultimately harm public investors.

B. The Proposed Central Limit Order Book Would Place An Unwarranted Burden on Competition

The Nasdaq Central Limit Order Book, if implemented on a full-scale basis, would place the NASD in direct competition with its own members for the display of limit orders. Even worse, the Nasdaq Central Limit Order Book would benefit from a number of unfair advantages in this competition. Because of these unfair advantages, the Nasdaq limit order facility ultimately will have anti-competitive effects that hurt the investing public. Yet the NASD has not analyzed these anti-competitive effects in attempting to justify the limit order facility. Implementing the Central Limit Order Book on a pilot basis would not contribute to any meaningful assessment of the impact of the proposal.

First, the Central Limit Order Book would provide Nasdaq with an unfair advantage in competing with its members because the new facility would have design features that the NASD, in the exercise of its regulatory powers, has made unavailable to competing market makers. The Nasdaq system only allows market makers to display their best-priced orders or quotations, and requires them to update their prices after an execution. The Nasdaq limit order facility, on the other hand, would display the entire depth of limit orders at various price levels. This ability to display simultaneously a variety of orders priced away from the market would give the Nasdaq facility an obvious advantage over market makers in terms of visibility.

Second, because of the Nasdaq limit order facility's ability to hold and display multiple levels of interest at one time, limit orders displayed on the Central Limit Order Book may have certain execution advantages over those displayed by market makers, including a form of time priority. Moreover, while participation in the Central Limit Order Book is described as voluntary, the effect of the proposed rule will be to drive all limit orders to this facility, granting the NASD a monopoly in limit order handling. Because orders displayed in the Central Limit Order Book may have certain execution advantages, market makers may be forced to direct their limit orders to the Central Limit Order Book to fulfill their best execution obligations. In addition, market makers will have a strong incentive to send their limit orders to the Nasdaq facility because of the perception that the facility will provide a regulatory "safe harbor" for best execution and other regulatory obligations. It is also likely that markeet participants will expect preferential treatment for orders directed to the Central Limit Order Book.

In addition, the Central Limit Order Book would create a conflict of interest for the NASD, who would have the authority to examine the operations of, and impose disciplinary sanctions on, its own competitors, including its competitors' order routing obligations. As a regulator, the NASD also would have the opportunity to adopt rules and policies that favor the Nasdaq limit order facility to the detriment of its member competitors.

Purely in the interest of fairness, market makers should not be placed in the untenable position of competing with an affiliate of their own regulator. Market makers and ECNs already have spent substantial sums on creating systems for the display and execution of customer orders. Moreover, allowing Nasdaq to compete unfairly with its own membership for order flow is not in the interest of investors. Nasdaq's advantages in limit order handling will discourage NASD members from developing innovative systems for handling limit orders in the future. As a result, the entire market could become dependent on the capacity and security of a single system, which is a risky proposition in the best of circumstances.

Nasdaq's monopoly on limit order display and execution would have the additional, and perhaps disastrous, effect of discouraging market makers and dealers from participating in the Nasdaq market, ultimately resulting in decreased liquidity for Nasdaq stocks to the obvious detriment of investors. The absence of direct participating clients for market makers will cause market makers to commit less capital to market liquidity because there will be less profit to be had by it. The vast majority of Nasdaq issues cannot trade efficiently in an order-driven market and thus market maker capital is still crucial for maintaining the liquidity of these issues.

C. The NASD Has Not Sufficiently Analyzed the Cost of Implementing and Operating the Proposed Limit Order Facility

The Central Limit Order Book proposal is deficient as a matter of law because the NASD has undertaken virtually no analysis of the effects of the proposal on competition. Because of the nature of the probable longer-term effects of the proposal, introducing the Central Limit Order Book on a "pilot" basis would do nothing to cure this defect.

The NASD's proposal contains no analysis of the cost of the proposed limit order facility, which presumably will be financed by the dues and fees of its member-competitors. Nor has the NASD quantified the economic costs that the limit order facility would impose on market makers, even though the facility would result in far-reaching structural changes to the market. The proposal contains only the boilerplate language 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." Such boilerplate does not satisfy the requirements of Exchange Act Form 19b-4, which states that an SRO's Statement on Burden 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." Exchange Act Form 19b-4 also requires an SRO to "specify the particular categories of person and kinds of business on which any burden will be imposed and the ways in which the rule change will affect them."

Far from meeting these requirements, the NASD proposal does not even address the kind of fees it will charge for use of the limit order book. The overall competitive impact of the proposal cannot be assessed without an understanding of its fee structure, particularly since the facility would compete directly with systems of NASD member firms.

The NASD also has failed to address important systems and technology issues raised by its proposal, including potential systems and technology burdens on firms, such as extensive system modifications that may be required to protect limit orders sent to the Nasdaq book.

Instituting the Central Limit Order Book on a pilot basis would provide little assistance in evaluating the effects of the proposal. The most significant problem with the Central Limit Order Book is that it would create a Nasdaq monopoly over limit order handling, a problem whose effects could not be gauged if the facility were to handle only a sample of securities or if it were to operate for a short period of time. In addition, a pilot program probably would not reveal the burdens of system modifications in which members may be compelled to invest if Nasdaq were to operate the facility on a permanent basis.


By re-opening comment on the Central Limit Order Book, the SEC has resurrected a proposal that is unnecessary, anti-competitive, not in the interest of investors, without appreciable industry support, detrimental to market-making and therefore liquidity of Nasdaq stocks, and potentially unmanageable from a technological perspective. The effects of the proposal have not been thoroughly researched. In short, the Central Limit Order Book proposal is ill advised and should not be approved.

The need for further limit order protection by Nasdaq is questionable given the salutary effects of the Order Handling Rules. Of greater importance, in the current environment there is no lack of capacity or innovation on the part of NASD members in devising systems for handling customer limit orders and providing these orders with best execution. A quasi-governmental agency such as the NASD should not be permitted to usurp the commercial functions of its own members and capitalize unfairly on its regulatory power.

Please do not hesitate to contact us if you would like to discuss these issues in further detail.



Kenneth D. Pasternak   Walter F. Raquet
President   Chief Operating Officer

Cc: Lee A. Pickard, Esq.

Pickard and Djinis