2. Fair competition among brokers and dealers, among exchange markets, and between exchange markets and markets other than exchange markets.

3. The availability to brokers, dealers, and investors of information with respect to quotations for and transactions in securities.

4. The practicability of brokers executing investors' orders in the best market.

5. An opportunity, consistent with the provisions of clauses (i) and (iv) of this subparagraph, for investors' orders to be executed without the participation of a dealer.

Translating these congressional imperatives into a set of conditions that define the national market system, the standards to be met are:

1. The system must execute orders quickly, cheaply and accurately.

2. The system must treat all market participants-brokers, market makers, investors and market center operators-equally.

3. The system must deliver the same information about quotations and executions to all market participants at the same time, and with the same information content.

4. The system must, by its design, allow all bids and offers in the same security to interact with each other to provide the lowest cost for the buyer and the highest sale proceeds for the seller.

5. The system must enable all investors' orders in each security to interact directly.

If these had been the standards the Commission had set forth in 1976, the Nation would have had the national market system the Congress ordered them to "facilitate" within a few years. Instead, the Commission has established a moving, ever-changing target of specifications that have attempted to meet the defects of the systems proposed by the institutions that had everything to lose and nothing to gain by meeting the congressional mandate.

Instead, these organizations-led by the New York Stock Exchange-used all their political and economic muscle to offer illusions of automation and forward progress. To demonstrate that fact, all you have to do is to read the NYSE's latest appeal for delay and confusion, issued under the title "Market Structure Report."

For example, the notion of "price improvement" is a fraud. When a trade is executed, either the seller hits the bid, or the buyer takes the offer. At the moment of a trade, the spread must be zero. The only important issue is: Who gets to know see the best bids and offers, and who can access them? Only if all market participants have an equal chance, can "best execution" be attained. The idea of "improving the NBBO (National Best Bid and Offer)" is smoke and mirrors. If the NBBO can be improved, by definition, it is not the national best bid and offer.

In the interests of brevity, here are my answers to the Commission's 65 questions in their Release:

1. To what extent is fragmentation of the buying and selling interest in individual securities among multiple market centers a problem in today's markets?

2. For example, has fragmentation isolated orders, hampering quote competition, reducing liquidity, or increasing short-term volatility?

3. Has fragmentation reduced the capacity of the markets to weather a major market break in a fair and orderly fashion?

4. Is fragmentation in the listed equity markets likely to increase with the elimination of off-board trading restrictions, such as NYSE Rule 390?

5. In the existing over-the-counter market, what are the incentives for investors and dealers to quote aggressively?

6. If fragmentation is a problem, are competitive forces, combined with the existing components of market structure that help address fragmentation (price transparency, intermarket linkages to displayed prices, and a broker's duty of best execution), adequate to address the problem?

7. Will the greater potential provided by advancing technology for the development of broker order-by-order routing systems, or for informed investors to route their own orders to specific market centers, address fragmentation problems without the need for Commission action?

b. Internalization and Payment for Order Flow

8. What proportion of order flow currently is subject to internalization and payment for order flow arrangements in the listed equity and Nasdaq equity markets?

9. Will the proportion increase in the listed equity markets as a result of the elimination of off-board trading restrictions?

10. Is it possible for a non-dominant market center to compete successfully for order flow by price competition, without using internalization and payment for order flow arrangements?

11. If not, is the inability to obtain access to order flow through price competition a substantial reason for the existence of internalization and payment for order flow arrangements?

12. To what extent can brokers compete as effectively for retail business based on execution quality (or implicit transaction costs), as opposed to commissions (or explicit transaction costs) and other services?

13. Do investor market orders that are routed pursuant to internalization and payment for order flow arrangements receive as favorable executions as orders not subject to such arrangements?

14. Even if these orders subject to internalization and payment for order flow arrangements receive comparable executions, does the existence of such arrangements reduce the efficiency of the market as a whole (by, for example, hampering price competition) so that all market orders receive less favorable executions than they otherwise would if there were no internalization or payment for order flow?

15. Even if internalization and payment for order flow arrangements increase the fragmentation of the markets, are any negative effects of increased fragmentation outweighed by benefits provided to investors, such as speed, certainty, and cost of execution?

c. Best Execution of Investor Limit Orders

16. Does increased fragmentation of trading interest reduce the opportunity for best execution of investor limit orders?

17. Are brokers able to make effective judgments concerning where to route limit orders so as to obtain the highest probability of an execution?

18. Does the opportunity for brokers to share in market maker profits through internalization or payment for order flow arrangements create an economic incentive to divide the flow of investor limit orders from investor market orders among different market centers?

19. If so, does this adversely affect the opportunity for investor limit orders to be executed fairly and efficiently?

20. Is it consistent with national market system objectives (such as efficiency, best execution of investor orders, and an opportunity for investor orders to meet without the participation of a dealer) for market makers to trade ahead of previously displayed investor limit orders held by another market center (that is, trade as principal at the same price as the limit order price)?

21. Does this practice significantly reduce the likelihood of an execution for limit orders by reducing their opportunity to interact with the flow of orders on the other side of the market?

22. Does the practice offer any benefits that outweigh whatever adverse effects it might have on limit order investors?

Possible Options for Addressing Fragmentation

If action to address fragmentation is determined to be necessary or appropriate to further the objectives of the Exchange Act, a variety of approaches could be considered. Six options are briefly described below, followed by requests for comment that relate specifically to each one. The options could apply either individually or in some combination with one another. If commenters believe fragmentation should be addressed, they also are encouraged to submit any additional options for addressing fragmentation that they consider feasible.

a. Require Greater Disclosure by Market Centers and Brokers Concerning Trade Executions and Order Routing

The Commission could require greater disclosure by market centers and brokers concerning their trade executions and order routing. Such disclosures could enable investors to make more informed judgments concerning the quality of executions provided by their brokers, as well as enable brokers and the general public to make more informed judgments concerning the quality of trade executions at all market centers. For example, all market centers could be required to provide uniform, publicly available disclosures to the Commission concerning all aspects of their trading and their arrangements for obtaining order flow. These disclosures could include the nature of their order flow (for example, the ratio of limit orders to market orders), their effective spreads for market orders for different types of securities (for example, securities that have different levels of trading), their percentage of market orders that receive price improvement, their speed in publicly displaying limit orders, their fill rates for different types of limit orders (for example, those with between-the-quotes and at-the-quotes limit prices), and their average time-to-fill for different types of limit orders. In addition, market centers could be required to make available comprehensive databases of raw market information that will allow independent analysis and interpretation by brokers, academics, the press, and other interested parties.

Brokers, in turn, could be required to provide disclosures to their customers (and to the Commission for public availability) concerning the proportion and types of orders that are routed to different market centers, their arrangements with market centers for routing customer orders, and the results they have obtained through these arrangements.

23. What would be the advisability and practicality of this option?

24. Would it effectively address the problems presented by market fragmentation?

25. Is there an effective and practical way to provide clear and useful disclosure to retail customers concerning execution quality?

26. If not, does the difficulty of providing such disclosure preclude brokers from competing effectively on the basis of execution quality?

b. Restrict Internalization and Payment for Order Flow

The Commission could restrict internalization and payment for order flow arrangements by reducing the extent to which market makers trade against customer order flow by matching other market center prices. Market makers would thereby be less assured of the profits that can be earned by trading against directed order flow and that are used to fund the economic inducements offered to brokers for their customers' order flow. For example, the NYSE has requested that the Commission take this type of action to address internalization. Under the NYSE proposal, broker-dealers would be limited in the extent to which they could trade as principal with their customers' market and marketable limit orders. A broker-dealer could buy from or sell to its customer only at a price that was better than the NBBO for the particular security. This type of prohibition could be extended to all market centers that receive orders pursuant to a payment for order flow arrangement, in addition to internalizing broker-dealers.

27. What would be the advisability and practicality of this option?

28. Would it effectively address the problems presented by market fragmentation?

29. Would restricting internalization and payment for order flow arrangements unduly interfere with competition among market centers to provide trading services based on factors other than price, such as speed, reliability, and cost of execution?

c. Require Exposure of Market Orders to Price Competition

As a means to enhance the interaction of trading interest, the Commission could require that all market centers expose their market and marketable limit orders in an acceptable way to price competition. As one example of acceptable exposure, an order could be exposed in a system that provided price improvement to a specified percentage of similar orders over a specified period of time. As another example of acceptable exposure, a market maker, before executing an order as principal in a security whose quoted spread is greater than one minimum variation, could publish for a specified length of time a bid or offer that is one minimum variation better than the NBBO. The Commission proposed such a rule for public comment in 1995 at the time it proposed the Order Handling Rules. Although it believed that an opportunity for price improvement could contribute to providing customer orders with enhanced executions, the Commission chose not to adopt the proposed rule at the time it adopted the Order Handling Rules.

Instead, it stated that it was deferring action to provide an opportunity to assess the effects that the Order Handling Rules would have on the markets.

30. What would be the advisability and practicality of this option?

31. Would it effectively address the problems presented by market fragmentation?

32. Are there effective means of representing undisclosed orders in markets in which trading interest is divided among many different market centers?

33. Would exposure of market orders through the quote mechanism provide a viable means of allowing the holders of undisclosed orders (particularly large orders) to interact with market orders?

34. What other means to facilitate the interaction of undisclosed and disclosed orders is feasible and practical?

35. Would requiring the exposure of market orders to price competition unwarrantedly delay the execution of those orders?

36. If so, should order exposure be offered as a choice to customers?

37. How would implementation of this option affect the opportunity for execution of displayed trading interest at the NBBO?

d. Adopt an Intermarket Prohibition Against Market Makers Trading Ahead of Previously Displayed and Accessible Investor Limit Orders

The Commission also could establish intermarket trading priorities as a means to address fragmentation. One option would be to adopt an intermarket prohibition against market makers (including exchange specialists) using their access to directed order flow to trade ahead of investor limit orders that were previously displayed by any market center and accessible through automatic execution by other market centers. Under this option, each market center would be responsible for providing notice to other market centers of the price, size, and time of its investor limit orders that were entitled to priority, as well as participate in a linkage system that allowed automatic execution against the displayed trading interest. To execute a trade as principal against customer order flow, market makers would be required to satisfy, or seek to satisfy, investor limit orders previously displayed and accessible at that price (or a better price) in all market centers.

To reward market makers willing to add liquidity to the markets through aggressive quote competition (as well as participate in public price discovery), a market maker could be allowed to trade with customers at its quote ahead of a subsequently displayed investor limit orders under certain circumstances. For example, a market maker could trade as principal against a customer order if, at the time it received a customer order, its quote was at the NBBO; its quote was widely displayed and accessible through automatic execution at a size at least equal to the customer order; and the market maker satisfied, or sought to satisfy, all investor limit orders that were displayed prior to the market maker's quote.

38. What would be the advisability and practicality of this option?

39. Would it effectively address the problems presented by market fragmentation?

40. Would prohibiting market makers from trading ahead of investor limit orders, regardless of where the order entered the national market system, facilitate a broker's ability to obtain best execution of its customers' limit orders?

41. Would an intermarket prohibition against market makers trading ahead of previously displayed and accessible limit orders encourage price competition and thereby enhance the efficiency of the market as a whole?

42. Would implementation of this option reduce the willingness or capacity of market makers to supply liquidity?

43. If so, would the problem be addressed by allowing market makers to trade at their quotations after satisfying previously displayed investor limit orders?

44. Would this option be feasible without the establishment of a single, intermarket limit order file?

e. Provide Intermarket Time Priority for Limit Orders or Quotations that Improve the NBBO As another option for encouraging price competition, the Commission could establish intermarket trading priorities that granted time priority to the first limit order or dealer quotation that improved the NBBO for a security (that is, the order or quotation that either raised the national best bid or lowered the national best offer). To qualify for such priority, the limit order or quotation would have to be widely displayed and accessible through automatic execution.

Only the first trading interest at the improved price ("Price Improver") would be entitled to priority. No market center could execute a trade at the improved or an inferior price unless it undertook to satisfy the Price Improver.

Subsequent orders or quotations that merely matched the improved price would not be entitled to any enhanced priority. If, prior to satisfaction of the Price Improver, another order or quotation was displayed and accessible at an even better price, the existing Price Improver would be superseded and permanently lose its priority. The subsequent trading interest at the better price would be the new Price Improver.

45. What is the advisability and practicality of this option?

46. Would it effectively address the problems presented by market fragmentation?

47. Would it discourage competition among market centers or reduce market makers' willingness to supply liquidity?

48. Would granting time priority only to the first trading interest to improve the NBBO provide an adequate incentive for aggressive price competition?

49. How difficult would it be to implement this limited type of intermarket time priority?

50. Would it require substantial modifications of currently existing linkage systems?

f. Establish Price/Time Priority for All Displayed Trading Interest

To assure a high level of interaction of trading interest, the Commission could order the establishment of a national market linkage system that provides price/time priority for all displayed trading interest. Under this option, the displayed orders and quotations of all market centers would be displayed in the national linkage system ("NLS"). All NLS orders and quotations would be fully transparent to all market participants, including the public. Orders and quotations displayed in the NLS would be accorded strict price/time priority. Market makers could execute transactions as principal only if they provided price improvement over the trading interest reflected in the NLS. Trading interest in the NLS could be executed automatically; however, the NLS would not be a market center itself: executions would continue to occur at the level of individual market centers. Public access to the NLS would be provided through self-regulatory organizations, alternative trading systems, and broker-dealers.

The NLS could be administered and operated by a governing board made up of representatives from the public and relevant parts of the securities industry.

51. What is the advisability and practicality of this option?

52. Would it effectively address the problems presented by market fragmentation?

53. Has advancing technology and increased trading volume created more favorable conditions for the establishment of a national market linkage system at the current time than at any time in the past?

54. What would be the respective benefits and costs of such a system?

55. Would a national market linkage system with strict price/time priority and automatic execution provide the most efficient trading mechanism?

56. If so, why have competitive forces failed to produce such a system without the necessity for Commission action?

"Most other self-regulatory organizations opposed creation of a Central File as described in the January Statement. These commentators argued that the kind of priority proposed to be afforded public limit orders entered into the Central File would have significant and deleterious effects on the exchange trading process. In essence, these commentators asserted that such a preference for public limit orders would provide a major trading advantage to those orders, thereby creating a disincentive to the commitment of market making capital by dealers, and would eventually lead to the elimination of exchange trading floors by inexorably forcing all trading into a fully automated trading system. In addition, several self-regulatory organizations suggested that, in lieu of the immediate implementation of a Central File, the Commission should permit the participants in the Intermarket Trading System ("ITS") sufficient time to attempt to provide limit order protection on an inter-market basis using the ITS. Specifically, the New York Stock Exchange, Inc. ("NYSE") and the MSE submitted proposals which envisioned the electronic dissemination and display of limit order information from each market center and use of the ITS to assure inter-market price protection of displayed limit orders in any market."

There was no justifiable reason for failing to produce such a system. It was just the political influence of those who wanted to continue doing it "the old fashioned way."

57. Are there any regulatory rules or industry practices blocking competitive forces that otherwise would produce such a system?

58. If so, what are they and how should they be addressed?

59. Would a mandated national market linkage system substantially reduce the opportunity for competition among market centers to provide trading services?

60. If so, would the costs of reduced market center competition outweigh the benefits of greater interaction of trading interest?

61. Would implementation of a comprehensive national market linkage system effectively require the creation of a single industry utility?

62. How should a national market linkage system be governed?

63. Should there be any exceptions from the requirement that all orders yield price/time priority to trading interest reflected in a national market linkage system?

64. For example, should there be an exception for block transactions or for intra-market agency crosses at the NBBO?

65. Should a national market linkage system incorporate a reserve size function to facilitate the submission of large orders?

There's a certain sense of "déjà vu all over again" in this Release. I thought it might be helpful if I attached a chapter I have written for the 2000 Edition of The World Guide to Equity, Derivative and Commodity Markets, "Decimalizing Humpty Dumpty: Finally a National Market System? Maybe!" It recounts a bit of history about the national market system, as well.

Needless to say, if there are hearings on this subject, I'm always ready!

Sincerely yours,

/s/ Junius W. Peake

Junius W. Peake

Comment letter on Fragmentation SR-NYSE-99-48-1
Attachment

Decimalizing Humpty Dumpty:
Finally a National Market System?
Maybe!
Junius W. Peake
Monfort Distinguished Professor
Kenneth W. Monfort College of Business
University of Northern Colorado

Preface

While this paper focuses on the decimalization of the U.S. stock markets, it is important that a proper foundation be laid that will place the issue in perspective. A quarter-century ago the U.S. Congress instructed the Securities and Exchange Commission ("SEC" or "Commission") to facilitate the development of a "national Market system for securities," using modern technology to do so.

25 years later, our country still does not have that national market system. In fact the systems for trading stocks are more fragmented and complicated than they were prior to the legislation. This article explains what happened and why. The good news, however, is that the prospect of decimalization will, for many reasons, finally force the Commission to scrap the Rube Goldbergesque¹ infrastructure that has been built, and replace it with a simple rule: "price-time priority," that will finally allow investors to have the modern, efficient, low-cost financial trading system the Congress wanted.

Executive Summary

In a course the author teaches, "Financial Institutions and Markets," students learn that one of the major risks a financial organization runs is "regulatory risk." This is the risk that regulators-congressional or congressional agency alike-can cost industry large sums of money for making ill-considered changes to laws and/or regulations.

The savings and loan débacle of the 1980s is a classic example of such regulatory error. In that case, the Congress allowed savings and loans and other thrift institutions to make investments in high-yield bonds and other financial instruments they were unqualified to judge. At the same time, Congress raised the deposit insurance tenfold, thus dramatically increasing the contingent liabilities to the taxpayers. The result is history: hundreds of billions of dollars lost.

What happened in the case of the national market system was a more subtle error, although its resultant consequences have placed the nation's securities markets into what could soon become an uncompetitive position internationally. The root cause of the problem was that the SEC, ordered by Congress to "facilitate a national market system for securities," decided to rely on attorneys to design the system through regulations, without understanding the power and potential of computers and telecommunications systems.

For the past 25 years, the Commission has spent (and wasted) its efforts in attempting to make the focus of the competition called for by Congress the location of where a trade took place, rather than the price at which the trade was made. Using Talmud-like logic, the Commission and its staff attempted to make the technology of the last quarter of the 20^th Century fit into the securities laws written in the early 1930s.

As time went on, the myriad of releases by the SEC (always claiming not to be micromanaging the design of the national market system) created an ever more Byzantine set of trading facilities and rules that drove systems designers almost to madness, and added more and more entities that allowed the same securities to be traded at the same time in different locations.

The Commission's solution for this situation was to require more and more information about the prices stocks were being quoted in multiple trading arenas, and demand that the financial industry figure out some way to make sure the investor received the "best" price for a trade, while adding costs at every step in the design.

Only early this year, after the third millennium dawned, the industry-led by some of its largest and most powerful firms-finally told the Commission what the investors want: a central market system using modern telecommunications and data processing systems to make sure that every bid has the opportunity to meet every offer in that security, first-come, first-served.

The inexorable push to reduce the minimum differentials of trading increments to cents from eighths and sixteenths will finally make alternative systems so complicated and expensive that the industry and the regulators must finally develop such a central market system to remain competitive globally.

Introduction and Background

A quarter-century ago, after several years of investigative and legislative hearings in both houses of the United States Congress, the Securities Reform Act of 1975 was passed and signed by the President. Among the many changes made in the securities laws was a new section inserted into the Securities Exchange Act of 1934: Section 11A.²

In that law, the Commission was ordered by the Congress "...to use its authority...to facilitate the establishment of a national market system for securities...in accordance with the findings and to carry out the objectives set forth [earlier in the Section]..."

The objectives of the Section were to assure:

i. economically efficient execution of securities transactions;

ii. fair competition among brokers and dealers, among exchange markets, and between exchange markets and markets other than exchange markets;

iii. the availability to brokers, dealers, and investors of information with respect to quotations for and transactions in securities;

iv. the practicability of brokers executing investors' orders in the best market; and

v. An opportunity...for investors' orders to be executed without the participation of a dealer.

If the language of the law were had stopped there, a true national market system would have been built within five years. However, another sentence was added, one that created a phantom loophole the New York Stock Exchange ("NYSE"), the regional exchanges, and the National Association of Securities Dealers ("NASD") used to obfuscate the congressional objectives for the next 25 years:

The linking of all markets for qualified securities through communication and data processing facilities will foster efficiency, enhance competition, increase the information available to brokers, dealers, and investors, facilitate the offsetting of investors' orders, and contribute to best execution of such orders. [Emphasis added.]

If any investor asks for the current "market" for a security, such as IBM, the response would certainly not be: "Boston Stock Exchange bid, offered at Island ECN!" Instead, it would be something like "105 ½ bid, offered at 105 ¾." After all, a "market" for a security (qualified or otherwise) is expressed as a price. It is not a marketplace! The Congress wanted bids and offers linked, not geographic locations. The SEC appeared to believe the same thing-at least for a couple of years.

In response to the legislation, the SEC appointed a 15-person "National Market Advisory Board ("NMAB")," which held periodic meetings and hearings on market structure for more than a year. The selection process attempted to make certain that every interest had representation. That diversity also guaranteed the Committee would never arrive at a majority opinion on almost anything, and the results were predictably obscure and vague.

On April 30, 1976, I, together with two colleagues, Professor Morris Mendelson of the Wharton School of the University of Pennsylvania, and fellow consultant R.T. Williams, Jr., presented our proposal for a fully-electronic auction market in which the total quantity of all bids and offers in each security, both from investors and broker-dealers, would be aggregated and displayed to all, and which could be executed electronically when a displayed bid or offer was hit or taken. We recommended the price increments between trades should be in decimal form, with the minimum possible spread becoming one cent. This proposal, known as the Peake-Mendelson-Williams ("PMW") National Book System³ became the paradigm for all electronic trading system proposals ("known generically as hard CLOB systems") discussed thereafter.⁴

In Professor Seligman's book, cited in the previous footnote, he noted:

"The Peake-Mendelson-Williams proposal discussed, among other topics, the advantages of designing the entire system as a coherent whole, rather than evolving a national market created in stages:

`If a National System should be implemented in stages, starting with a consolidated book with no execution capability, and evolving to an ultimate national system, the process will be time-consuming [between ten and fifteen years will probably be required between initial plan and ultimate realization] and complex. At each stage of implementation, a workable system has to be developed and implemented, only to be scrapped later. That will be costly.'"

The SEC did not follow this recommendation despite their longstanding commitment to a centralized auction market; a quarter-century has passed without consummation of congressional objectives.

For example, on December 19, 1975, the Commission issued a Release (34-11942), which stated unambiguously:

"Development of a central electronic repository for limited price orders would be of special significance to ensure integration of the markets and preservation of an opportunity for public orders to meet without the participation of a dealer. Such a step will certainly enhance competitive opportunities in market makings. For all these reasons, the Commission will utilize its new powers under the Act promptly to ensure implementation of a national mechanism for multi-market protection of limit orders. Nevertheless, it must be emphasized that it would be inappropriate to withhold from the markets the benefits to be derived from increased market maker competition indefinitely. Development of a national limit order mechanism is a further step in creating a national market system and must be expedited."

And again, in 1976 (SEC Release 34-12159):

"The Commission believes that there is a need for further modernization and improvement of our securities markets, not only for the purpose of utilizing new data processing and communication techniques, but also to insure economically efficient execution of securities transactions and fair competition among brokers and dealers and among various securities markets which either directly compete with each other or have the potential for such competition. Existing exchange mechanisms for the storage and execution of limited price orders appear to be in need of modification to meet the requirements of member firms and investors for expeditious handling of order flow in the context of a national market system, as well as to cope with an increasing volume of securities transactions (such as that experienced in recent weeks). Further, existing limit order mechanisms are unable to provide nationwide limit order protection and thus cannot always provide the degree of protection for limit orders which hopefully could be furnished by a composite book. Finally, a composite book appears to be well suited to assuring an opportunity for public orders to meet without the participation of a dealer."

Read these quotations from Securities and Exchange Commission Release No. 34-14416, dated January 26, 1978:

"The concept of a national market system was first articulated in the Commission's letter of transmittal accompanying its Institutional Investor Study, submitted to Congress on March 10, 1971. There the Commission stated that:

[a] major goal and ideal of the securities market and the securities industry has been the creation of a strong central market system for securities of national importance, in which all buying and selling interest in these securities could participate and be represented under a competitive regime."

Again, quoting from the 1978 Release:

"In addition to elaborating on the principles set forth in the Future Structure Statement, the Commission's Policy Statement articulated two new proposals to govern trading within a national market system: an auction trading rule, which would provide price priority protection for all public orders entered in a proposed central electronic repository, and a public preference rule, which would accord preferential treatment to public orders entered in the central electronic repository by preventing securities professionals acting as principal from competing for execution with such orders unless such professionals bettered public bids or offers entered in that system."

In reporting the legislative history of the 1975 Amendments, the Release stated:

"...the Senate Committee on Banking, Housing and Urban Affairs (the "Senate Committee") stated that:

[t]he rapid attainment of a national market system . . . is important . . . to assure that the country maintains a strong, effective and efficient capital raising and capital allocating system in the years ahead."

And again:

"The Senate Committee noted, however, that auction trading principles could not be perfected under existing circumstances because of fragmentation of the markets, particularly "the lack of a mechanism by which all buying and selling interest in a given security can be centralized and thus assure public investors best execution." Thus, the concept of implementing a nationwide system according price and time priority to all limit orders of public investors over all professional orders, regardless of where such limit orders originate or in what market center professional orders may be executed, received considerable support from the draftsmen of the 1975 Amendments."

Discussing "Progress to Date," the Release continued:

"The major problems to which the idea of a national market system is addressed are those arising from "market fragmentation," or the existence of multiple, geographically separated forums in which trading in the same security occurs, and from the institutionalization of the markets."

The Release then put forth:

"THE COMMISSION'S FUTURE PLANS FOR FACILITATING ESTABLISHMENT OF A NATIONAL MARKET SYSTEM."

In those plans, these prophetic words appeared:

"The adverse consequences of failing to achieve more rapid progress toward a national market system have become particularly apparent in the context of the Commission's pending proceeding concerning removal of exchange off-board trading restrictions. During the course of that proceeding, many elements of the securities industry, members of Congress and representatives of American business have urged the Commission to assume a leadership role in developing a national market system in order to overcome the impediments to development of that system inherent in the diversity of the securities industry, so that the benefits to the markets, the professional trading community and the public which the Congress and the Commission have long believed would inure from that system might finally be secured. Commentators in that proceeding, for example, were virtually unanimous in the view that the risks which many believe would attend removal of the remaining off-board trading restrictions could be minimized by assuring more effective integration of the markets for securities presently covered by those restrictions by means of national market system mechanisms."

And:

"Nationwide Limit Order Protection. The Commission continues to believe that one of the basic principles upon which a national market system must be based is the assurance that all agency orders in qualified securities, regardless of location, receive the benefits of auction-type trading protections. To this end, the Commission believes the several self-regulatory organizations should take joint action promptly to develop and implement a central limit order file (the "Central File") for public agency orders to buy and sell qualified securities in specified amounts at specified prices ("public limit orders")."

Finally:

"The Commission urges the self-regulatory organizations to prepare and submit to the Commission, preferably jointly, a plan or plans no later than September 30, 1978, contemplating the design, construction and operation of a Central File. However, should voluntary cooperation among such organizations to that end prove difficult, or involve undue delay, the Commission intends to commence rulemaking to consider the manner and timing of compulsory development of a Central File (including the question of whether that task should be assigned principally to a single self-regulatory organization.)."

But in 1979, thanks to a fear by the Commission that changing to a centralized, electronic market for securities trading might be too risky, coupled with intensive lobbying efforts by the NYSE and other exchanges and market makers to preserve the status quo, they suddenly reversed course, and permitted a trio of systems to be, as the Commission put it, the "cornerstones" of the national market system. That reversal sealed the unconscionable delay of a national market system, at least for the rest of the 20^th Century. In their April, 1979 Release (34-15770), the SEC stated:

"Most other self-regulatory organizations opposed creation of a Central File as described in the January Statement. These commentators argued that the kind of priority proposed to be afforded public limit orders entered into the Central File would have significant and deleterious effects on the exchange trading process. In essence, these commentators asserted that such a preference for public limit orders would provide a major trading advantage to those orders, thereby creating a disincentive to the commitment of market making capital by dealers, and would eventually lead to the elimination of exchange trading floors by inexorably forcing all trading into a fully automated trading system. In addition, several self-regulatory organizations suggested that, in lieu of the immediate implementation of a Central File, the Commission should permit the participants in the Intermarket Trading System ("ITS") sufficient time to attempt to provide limit order protection on an inter-market basis using the ITS. Specifically, the New York Stock Exchange, Inc. ("NYSE") and the MSE submitted proposals which envisioned the electronic dissemination and display of limit order information from each market center and use of the ITS to assure inter-market price protection of displayed limit orders in any market."

The systems that were substituted for an electronic central bid and offer file-the Consolidated Quotations System ("CQS"), the Intermarket Trading System ("ITS"), and the Consolidated Tape System ("CTS")-created the illusion, rather than the reality of a national market system. Rather than centralizing all orders in a single security, they left the markets fragmented, allowing each of the separate stock exchanges to accumulate and execute orders on their own exchange.

The ITS was supposed to allow bids and offers sent to one exchange to be executed on another with a better price. In reality, ITS was little-used, and was described that same year (in sworn testimony before two congressional subcommittees) by the then president of Merrill Lynch, William Schreyer, as follows:

"....The Intermarket Trading System, or ITS, which links the New York with some regional exchanges, is a communications device, and nothing more. It is as far from the concept of an automated, efficient marketplace as a tom-tom from a communications satellite." ⁵

CQS was supposed to provide a montage of quotes (best bids and offers on each exchange in ITS), so orders could be directed to the marketplace with the best price. However, not all broker-dealers belonged to all exchanges, and to keep their systems costs reasonable, and to receive reciprocal order business, programmed their computers to send all orders in a security to the same marketplace.

The CTS displayed the transactions after execution, although often late, or not in sequence (or both).

To analogize with another financial transaction system, the automated teller machine ("ATM"), the exchange-designed national market system required one system to obtain prices, another system to receive and execute transactions, and a third to display the results of the trades. As we start into the third millennium, this practice continues. Can you imagine the public's reaction if the ATM system required a customer of a bank to walk to the first ATM machine to find his or her account balance, to a second ATM machine to enter the transaction, and to a third ATM machine to find the remaining account balance and to receive a receipt (and/or funds)?

On this shaky foundation, the Commission based their national market system. It was to fracture further.

Over time, the markets changed. As expected, trading volumes mushroomed, and unfixed commissions led to the almost instant development of discount brokerage firms, that specialized in executing unsolicited orders from investors at low cost.

In January, 1994, the Commission's Division of Market Regulation's staff released a telephone book-sized report, "Market 2000,"⁶ intended to set forth areas of market structure they believed needed attention. The new Chairman and the rest of the Commission, however, distanced themselves from their staff report, noting prominently on the title page:

"This is a report of the Division of Market Regulation. The Commission has expressed no view regarding the analysis, findings, or conclusions herein."

The Report did address two important issues: a central limit order book and decimalization. With respect to the former, the Report noted:

"The Commission proposed a market-wide, consolidated limit order book ("CLOB") in the 1970s. The exchanges vigorously opposed the creation of a CLOB and continue to oppose it in their comment letters on Market 2000. Floor members likely fear that a CLOB would be the first step toward the complete automation of the exchanges. Likewise, broker-dealers with large trading desks would oppose automated executions of large trades and adherence to market-wide time priority."⁷

Clearly, the Commission had acquiesced to the politics of the establishment, rather than analyzing the merits of the arguments. The Commission capitulated to the views and economic interests of the exchanges and over-the-counter market makers.

Here's what the Report had to say about decimalization:

"...the Division believes it is appropriate to revise the current pricing system...The Division believes that decimal pricing is preferable, and may be inevitable...In particular, the SROs [Self-Regulatory Organizations] should consider whether adoption of decimal pricing would benefit investors and strengthen the competitive posture of the U.S. equity markets as they position themselves in a global market."⁸

Despite that wake up call to the SROs and the industry, no systems modifications were made that would have prepared any exchanges or the NASD to prepare to convert to the inevitability of decimals.

It was only three years later, when certain members of the House of Representatives introduced a Bill requiring decimalization, and scheduled hearings on decimalization, that the Commission, the exchanges and the National Association of Securities Dealers ("NASD") suddenly "got religion," and espoused decimalization. At the hearings, the NYSE argued against decimalization, and only one member of the Commission, Steven Wallman, a longtime supporter of decimalization, testified in its favor.⁹

The industry promised to make the switch to decimals promptly, and set September 1998 as the conversion date. However, not unexpectedly, that date was pushed off by the Commission, which made the argument (supported by the establishment) that the so-called "Y2K Bug fix" took priority over decimalization, and thus reducing the size of spreads would wait until mid-2000.

Finally, this year the Commission proposed a date certain for decimalization. In its January, 2000 release, the Commission stated:

"The Commission believes that decimal pricing could benefit investors by enhancing investor comprehension, facilitating globalization of our markets, and potentially reducing transactions costs, depending on the minimum price variant used. These benefits in turn will further the national market system objectives of economically efficient execution of securities transactions and fair competition."¹⁰

However, as usual, the Commission waffled a bit:

" While the Commission is not mandating the details of a Decimals Implementation Plan, the plan must provide that decimal pricing of at least some equities (and options on those equities) trading on the Participants' markets will begin no later than July 3, 2000, and decimal pricing of all equities and options on the Participants' markets will be completed within six months of that date. If the Participants adopt a phase-in plan for implementing decimal pricing, rather than pricing all equities and options on the Participants' markets in decimals on July 3, 2000, the plan must provide specific dates by which each phase will be completed and identify which securities will be priced in decimals during each phase. The Decimals Implementation Plan may fix the minimum increment during the phase-in period, provided that the minimum increment is no greater than five cents for any equity priced in decimals. The Commission believes that it is appropriate for the Participants to establish a minimum increment during the phase-in period to allow the industry to make a smooth transition to decimal pricing and to determine the impact of decimal pricing on trading rules and inter-market systems capacity."¹¹

Only a few weeks later, the NASD told the Commission they would not be ready by July 3, 2000, and asked for a delay until February 2001. They argued that their share volume had increased from 1998 to 2000 by 115%, while their quotation message traffic had increased by 278% for the same period.¹²

However, the Commission quickly responded in its usual equivocal manner:

"At this point, the Commission has not extended the July 3, 2000 date for implementing decimal prices...The Commission continues to discuss with the self-regulatory organizations a viable decimal pricing implementation date, and, more generally, how to implement decimals in a way and in a timeframe that does not threaten the well-earned confidence that U.S. investors have placed in our markets. The Commission remains committed to implementing decimal pricing in a timely and orderly manner."¹³

A letter sent to the Commission and to the NASD by five powerful members of the House of Representatives, ordered Frank Zarb, President of the NASD, and Chairman Arthur Levitt of the SEC "...to report by March 24^th, or sooner, on what steps the NASD, the New York Stock Exchange and computer-based electronic communications networks, or ECNs, must take to convert to decimals."¹⁴

On March 16, 2000, the last date before the completion of this paper, Chairman Levitt, in an address at the Northwestern University School of Law, proposed still another forum for discussing the fragmentation problem:

"The ability of all investors to see the depth of supply and demand in any stock would be a giant step towards a true National Market System. And it is a step that our markets should take now, before the uncertainty of decimalization is upon us. Toward this end, I have asked the Division of Market Regulation to take the lead in facilitating a dialogue on this issue by hosting a public roundtable with representatives of all markets, dealers, market data vendors and other interested parties...We simply must not allow decimalization to obscure liquidity in our markets."¹⁵

The Solution to Market Fragmentation (and to the Maximization of Liquidity) is Decimalization

There is always a limit as to the maximum number of data points that can be sent over a computer system. No system can handle an infinite amount of traffic. The NASD's plea for more time until decimalization is understandable. With each market maker (and there can easily be 50 or more for a single stock) being able to enter prices at one-cent increments, the specter of seeing a montage of quotations with those possibilities is frightening to any systems analyst or programmer. Add ECNs and exchanges to the broth, and you have a witches' brew of data almost incomprehensible.

But decimalization is coming. It is inevitable, despite the decades-long attempts of the establishment to prevent its arrival. The rest of the world trades in decimals. If for no other reason than global competitiveness-and that's the ultimate competition-the United States must decimalize, and should do so as quickly as possible.

With an electronic centralized bid-offer facility, most of the problems vanish instantly. While thousands of broker-dealers and others may be entering bids and offers into such a central system at almost the same time, the information displays need only to show the total quantity bid for and offered at each increment for a security. As any systems engineer will tell you, processor speeds are far faster than input-output devices.

In addition, the SEC has long had a rule demanding "best execution." Best execution is simple; the Commission's approach attempts to make it sound complicated. For a buyer, "best execution" means paying the least amount of money for the purchased security on each transaction. For the seller, it means receiving the greatest amount of money for a security sold on each transaction. All other conditions laid out by the Commission are mere rhetoric.¹⁶

When spreads can be reduced to one-cent increments, "best execution" requires that a price-time priority system be in operation. Otherwise there is no practical way for investors-or broker-dealers, for that matter-to be able to demonstrate they received the best price on each order execution. Thus, a central bid-offer book is the only solution that meets the congressional intent, and will bring together all buying and selling interest at one focal point. It also maximizes competition; it does not create the SEC's dreaded "monolith."

Economics 101

All things being equal, when the cost of something is lowered, more will be traded. That principle applies to securities trading, as well as to everything else. There are three components of cost in a securities trade: explicitly-charged commissions; bid-offer spread; and market impact. In May 1975, as part of the Securities Reform Act, fixed commissions were abolished; volume immediately soared. In 1997, bid-asked spreads were reduced 50%-from 12 ½ cents per share (1/8) to 6 ¼ cents per share (1/16). Volume soared again. The larger the volume in a security, the more can be traded without affecting price (in the absence of information that would cause a price change).

Broker-dealers make money on spreads, just as do used car dealers. Dealers that register as market makers (or specialists on exchanges) risk their capital by buying unwanted positions from investors. The service they perform is known as immediacy, since the dealer stands ready to buy the unwanted position at a price from which the dealer expects to make a profit. Immediacy may also be called "microliquidity," since only the investor selling the unwanted position has received the funds (or securities in the event of a short sale cover). The market as a whole has not received liquidity, despite all the assertions otherwise made by market makers and specialists, since the unwanted position is still for sale-but by the market maker, not the investor.

Market makers, like all other dealers, attempt to minimize risk, wherever possible. It is for this reason that some dealers guarantee to acquire unwanted positions at the bid (or offer) side of the stated market, the National Best Bid and Offer ("NBBO"). But these guarantees have limits that reduce the dealer's risk. For example, the size of the position taken is minimized or limited. And the type of investor to whom the guarantee is made is also limited (unsophisticated individuals; market orders only, for example).

In 1975, market makers were very aware of these realities. As a result, they recognized that any structural revamping of the markets that would lessen their involvement or reduce spreads, represented a threat to their income stream. Not surprisingly, they mobilized their political and economic forces to prevent (or at least delay) changes that would benefit investors and disintermediate their roles. In combination with the established exchanges, they have been eminently successful, although that success is finally nearing its end.

Centralization vs Fragmentation

On February 23, 2000, the Commission issued a "Concept" Release¹⁷, asking numerous questions about market structure. The Release proposed six possible structural options:

1. Require Greater Disclosure by Market Centers and Brokers Concerning Trade Executions and Order Routing.¹⁸

2. Restrict Internalization and Payment for Order Flow.

3. Require Exposure of Market Orders to Price Competition.

4. Adopt an Intermarket Prohibition Against Market Makers Trading Ahead of Previously Displayed and Accessible Investor Limit Orders.

5. Provide Intermarket Time Priority for Limit Orders or Quotations that Improve the NBBO¹⁹.

6. Establish Price/Time Priority for All Displayed Trading Interest.

All but the last one perpetuate fragmentation; only the last attempts to centralize order flow, albeit imperfectly.

There is no way a market for financial instruments can be constructed that permits the same security to be traded at the same time in more than one market center to be other than fragmented. Even the New York Stock Exchange understands that fact.

In 1968, the NYSE issued an economic analysis report in which they stated:

"The splintering of the central auction market would disadvantage small investors in particular. Their orders would no longer be brought to one place where most bids and offers are concentrated. Instead, orders would often be executed in the offices of securities firms to which they happened to be entrusted.

"In these circumstances, there would be very little likelihood that the best executions would or indeed could be obtained. The individual broker would find it impractical to check each fragmented market for each customer's order. As a result, neither individuals nor securities firms would know for certain whether each order received the best available price at the time it was executed."²⁰

No matter how people try to rationalize the possibility of having "competing" market centers that trade the same securities in the same time frame somehow create a "central" market, the difficulty and costs of constructing, operating and regulating such an enterprise makes it uneconomic and counterproductive.

By the way, the charges levied against the idea of a central bid-offer book are false. First, there is no need to prevent dealer bids and offers from being executed ahead of investors' at the same price unless the dealers know about the investors' orders before entering their own bids or offers. A "Chinese Wall" approach, as used to separate commercial banking from trust banking will do the trick, if properly policed.

The other canard is that institutional orders must not be included in the electronic central bid-offer book. So long as the orders ahead of the institution's order are guaranteed execution at the prices that were entered, the system will accommodate any sized order.

Prediction: Within the year, the Commission and the Congress will finally recognize that reality, and we shall finally be on our way toward the national market system the visionaries in Congress wanted a quarter-century ago. The beneficiaries will be investors, and the intermediaries will also benefit by having an efficient, low-cost, modern trading system. Even the Commission will benefit greatly; such a system will be far simpler to regulate, and the rules for its operation will be far fewer and simpler.

While "regulatory risk" has been severe, the broken fragments of the Humpty Dumpty that is our present market structure can be repaired in time to keep the United States at the forefront of the world's capital markets .◙

Appendix "A"

The relevant text of Section 11A of the Securities Exchange Act (as amended):

Securities Exchange Act of 1934

Section 11A -- National Market System for Securities

1. The Congress finds that--

A. The securities markets are an important national asset which must be preserved and strengthened.

B. New data processing and communications techniques create the opportunity for more efficient and effective market operations.

C. It is in the public interest and appropriate for the protection of investors and the maintenance of fair and orderly markets to assure-

i. economically efficient execution of securities transactions;

ii. fair competition among brokers and dealers, among exchange markets, and between exchange markets and markets other than exchange markets;

iii. the availability to brokers, dealers, and investors of information with respect to quotations for and transactions in securities;

iv. the practicability of brokers executing investors' orders in the best market; and

v. an opportunity, consistent with the provisions of clauses HREF="http://www.law.uc.edu/CCL/34Act/#a.1.C.i" and HREF="http://www.law.uc.edu/CCL/34Act/#a.1.C.iv" (iv) of this subparagraph, for investors' orders to be executed without the participation of a dealer.

D. The linking of all markets for qualified securities through communication and data processing facilities will foster efficiency, enhance competition, increase the information available to brokers, dealers, and investors, facilitate the offsetting of investors' orders, and contribute to best execution of such orders.

2. The Commission is directed, therefore, having due regard for the public interest, the protection of investors, and the maintenance of fair and orderly markets, to use its authority under this title to facilitate the establishment of a national market system for securities (which may include subsystems for particular types of securities with unique trading characteristics) in accordance with the findings and to carry out the objectives set forth in HREF="http://www.law.uc.edu/CCL/34Act/sec11.html#a.1" The Commission, by rule, shall designate the securities or classes of securities qualified for trading in the national market system from among securities other than exempted securities. (Securities or classes of securities so designated hereinafter in this section referred to as "qualified securities".)

3. The Commission is authorized in furtherance of the directive in HREF="http://www.law.uc.edu/CCL/34Act/#a.2" of this subsection--

A. to create one or more advisory committees pursuant to the Federal Advisory Committee Act (which shall be in addition to the National Market Advisory Board established pursuant to HREF="http://www.law.uc.edu/CCL/34Act/sec11.html#d" and to employ one or more outside experts;

B. by rule or order, to authorize or require self-regulatory organizations to act jointly with respect to matters as to which they share authority under this title in planning, developing, operating, or regulating a national market system (or a subsystem thereof) or one or more facilities thereof; and

to conduct studies and make recommendations to the Congress from time to time as to the possible need for modifications of the scheme of self-regulation provided for in this title so as to adapt it to a national market system.

Appendix "B"

The Peake-Mendelson-Williams "National Book System," submitted April 30, 1976 to the National Market Advisory Board of the SEC.

NOTE: Only Appendix "E" is attached, dealing with the basic unit of pricing (decimalization).

Letterhead of Junius W. Peake,

Morris Mendelson and R.T. Williams, Jr.

April 30, 1976

Mr. George A. Fitzsimmons Reference: File No. 57-619
Secretary of the Commission
Securities and Exchange Commission
Room 892
500 North Capitol Street
Washington, D. C. 20549

Dear Mr. Fitzsimmons:

In response to your request for comment on the issues related to the development of a composite limit order repository, we are enclosing our plan for a National Book System which, we believe, provides a trading, communications and execution system meeting the legislative intent of the Securities Acts Amendments of 1975 as well as the Commission's policy statements.

The attached plan, as well as our specific responses to the hypothetical system characteristics enumerated in the Commission's Release, describe what we believe to be the ultimate auction market, permitting investors of all classes to participate under equal rules and with fair trading information.

While it may appear that some of the elements of our proposed National Book System differ substantially from the present mechanism, the fact remains that our system will cost less to design, build, operate and regulate than any interim system. It will also be simpler to construct and will restore a centralized trading facility.

Any attempt made to obtain a system such as we present, in stages, must result in a sequence of fully developed systems, each operating only long enough to permit the next stage to be constructed before being discarded.

The National Book System is not a "Composite Book." We believe the term "Composite Book" implies that several "books" are to be integrated. Our conviction is that there should be a single book, national in scope, permitting full competition by all potential buyers and sellers.

The National Book System is primarily a communications system, using the power of modern technology to facilitate the meeting in a common electronic arena of the parties to any security transaction-the buyer and seller or their agents.

Trading strategies and decisions will, of course, continue to be made by investors and their brokers. The computer will attend to the clerical tasks of matching like priced orders; delivering reports of executions to the brokers involved, to clearing organizations and reporting facilities: It will validate, edit and queue new orders.

The need for such a new system is becoming increasingly apparent, as the old system unravels. The causes of the unraveling are well known, although viewed differently from individual perspectives. These causes include:

• Dramatically increased participation by institutions, thus undermining the historical fixed commission system;

• Improved communications technology, which has permitted access to regional exchanges and over the counter dealers with relative ease, and eliminated the need for physical trading arenas;

• A better-developed recognition of the fiduciary duties of financial intermediaries, including banks, investment advisors and pension fund trustees;

• A reluctance on the part of established trading organizations to recognize the pressures for change, and the resultant erosion of their economic competitiveness.

With the advent of unfixed rates last year, the pace at which the old system is unraveling is accelerating. We agree with the Commission's recognition of the need to implement a new system that will meet the needs and challenges of our Nation's third century.

Our design of the National Book System is premised on the belief that all present participants in the market structure should be permitted to remain so long as the serve an economic function, and no participants are entitled to special privilege. When economic regulation works, it is far more efficient than legislative regulation. Our proposed system, by providing only the vehicle for trading in an orderly and equitable environment, requires a much simpler regulatory apparatus and many fewer rules.

Since the information contained in any "book" of unexecuted orders has substantial economic value, we see no merit in attempting to devise a scheme whereby favored participants are permitted to have access to this asset. Because the technology exists to make this information available to all, we have concluded that the National Book, with all bids and offers summarized at each price, should be available for inspection on a real time basis.

Assigning duties and rewards for franchised market makers is very difficult, as well as unnecessary. One has only to review the Report of the Committee to Study the Stock Allocation System from the New York Stock Exchange dated January 27, 1976, to determine the complexity of this task. The National Book System, by providing free entry and exit by market makers, eliminates this burden. The economic need to provide a continuous market in a security in order to attract the order and inquiry flow from institutions, combined with the economic advantage in making markets accruing to those who attract those flows, will assure a reasonable continuity of market making. The ability of brokers to access the National Book System directly will give even thinly traded securities the advantage of an auction market.

In the National Book System, we have limited direct access to "qualified" broker/dealers. We specify that they be clearing members of the national clearing and settlement system. We have selected this standard to assure the operational and financial security of the system. Other broker/dealers may, of course, access the system with their own data entry devices, but would be "correspondents" of "qualified" brokers of their choosing. The financial commitment for a transaction would be with the "qualified" broker/dealer, thus permitting a more secure base of direct participants.

Finally, we are confident that the development and implementation of the National Book System is in the best interest of the entire broker/dealer community as well as the investing public. By providing a single "book", all bids or offers at each price have the opportunity to displace orders, thus assuring participants that every execution Is In fact the "best execution" available at that time. The complex routing systems being developed by brokers and contemplated to connect market centers may be abandoned in favor of a single path to the best market.

In summary, we believe the National Book System that we hereby propose is the best, fairest, and most economic system for trading, not only of securities, but also of any other asset, including futures contracts, options and all debt instruments.

We shall be pleased to meet with the Commission, its staff or its Advisory Committee to discuss our proposal at greater depth.

Very truly yours,

/s/ Junius W. Peake

/s/ Morris Mendelson

/s/ R.T. Williams, Jr.

Junius W. Peake
Senior Vice President
R. Shriver Associates
120 Littleton Road
Parsippany, New Jersey 07054

Morris Mendelson
Professor of Finance
Wharton School
University of Pennsylvania
Philadelphia, Pennsylvania 19174

R.T. Williams, Jr.
Senior Consultant
R. Shriver Associates
120 Littleton Road
Parsippany, New Jersey 07054

THE NATIONAL BOOK SYSTEM:

An Electronically Assisted

Auction Market

Junius W. Peake
Senior Vice President
R. Shriver Associates
120 Littleton Road
Parsippany, New Jersey 07054

Morris Mendelson
Professor of Finance
Wharton School
University of Pennsylvania
Philadelphia, Pennsylvania 19174

R.T. Williams, Jr.
Senior Consultant
R. Shriver Associates
120 Littleton Road
Parsippany, New Jersey 07054

April 30, 1976

Table of Contents

SECTION I

A DESCRIPTION OF THE NATIONAL BOOK SYSTEM

I. SECTION I

A. Introduction and Summary

The National Book System we propose is an electronic trading arena in which the principals to a trade, or their agents, may gain information about the market, and can execute an order. The system is substantially simpler, more powerful and less costly than other public proposals. It establishes a trading mechanism without commitment to any existing manual procedures.

Present trading systems were developed with the communications technology available at that time. These limitations of communications capability have led to many shortcomings in the present market system, including fragmented market places (regional and third markets), trading in fractional units of currency (1/8ths) rather than the decimal system, as well as isolating in a "second class" system, so-called "odd lots". All of these complicating elements will be eliminated with the development and implementation of the National Book System.

The advent of electronic communication, and computers that reliably process and switch information, have rendered some existing roles in the securities industry not only redundant, but even counterproductive. Therefore, we have designed and propose a system based upon the trading requirements of the securities markets, and using current technology. The use of computers for such applications as we are recommending parallels processes commonly used in elements of the broker/dealer and securities trading systems of today.

Order entry and routing systems are used by all larger brokers as well as many smaller ones. Matching and reporting executions by computer is routine. Creating output records for clearing and settlement is commonplace. Automatic execution is in place to a limited degree on several exchanges. An automated quotations system has been working efficiently and reliably for more than five years. Nothing we propose is technologically difficult or dangerous; it is only a reordering of many existing systems into a unified whole. In short, our proposal is simple because we have accommodated the requirements for trading, rather than duplicated the status quo in computer code .

The system we suggest is a communications facility that permits any investor, via a broker or market maker, to gain instant access to historical volume and price information and to see the entire book in summarized form. The investor may step into the market at any point and execute at any available price he finds attractive, or, if spreads become too large, he may enter the arena and act as a de facto market maker. In short, the public and any qualified broker/dealer may act as market maker if regular market makers permit a profit potential to materialize. True competition will exist in the securities market.

This system, then, becomes the ultimate auction market. It allows all buyers and sellers to compete or meet via an electronic intermediary.

The National Book System requires a single processor, though in fact that processor will need multiple systems to provide sufficient backup. Competition among prospective vendors should be encouraged at each step. The characteristics of the National Book System should be circulated to receive competitive bids for the design of the system. These bids will permit different ideas to be developed.

The winning design (or designs) should again be put out for construction and operator bids. Various parts of the system may have different computers and processors, although all must conform to unified standards. The contracts awarded should be for fixed periods of time, with provisions for new competitive bids as they approach expiration. These policies, supervised by the governing board, will assure that the system will take advantage of new technological developments.

Ironically, a single processor is required if the stated priorities are to be met. If so called "competing marketplaces" are permitted to continue, it is technically difficult, if not impossible, for the price/time priorities to be honored .

If participants are allowed to enter the national trading system through multiple marketplaces, the system must be unnecessarily complex to assure that all participants are treated fairly in all possible markets. We are proposing a better system, with all securities trading in a single arena. We believe that it is not the marketplaces that should be competing, but rather the investors, agents and market makers.

We propose that the National Book System provide communications, execution and record keeping capability essential to efficient trading. All other functions presently provided by exchanges and others, would be prohibited from being performed by the National Book System. These other services, if economically viable, may be offered by anyone who may wish to provide them. If the services are not independently viable, they should not be subsidized by the marketplace.

The balance of our submission examines the National Book System in greater detail and responds to the Commission's Release.

B. Criteria

1. The system should have the ability to become, in a central arena, a single national market for all eligible securities.

2. The system should be accessible to all broker/dealers, either directly, if qualified by objective criteria, or indirectly, through any qualified broker/dealer.

3. The system should present equal market trading information to all users, including broker/dealers, market makers and the general public.

"Eligible security" is a publicly traded security that meets capitalization, financial and trading criteria.

A "qualified" broker/dealer is one that is a member of the national clearing, depository and settlement system and has met objective criteria of operational, financial and professional fitness.

4. The system should be able to process all sizes of orders, including so-called "odd lots" and "blocks",

5. The system should provide for equitable sequencing, or "queuing," based upon price and time of entry, as well as public priority.

6. The system should provide for the execution of public orders without the necessity of dealer intervention.

7. The system should insure that all executions of orders are, by their processing, "best execution."

8. The system should have the fewest rules possible imposed upon its operation, consistent with the public interest, as well as technical and policy considerations.

9. The system should, as a by-product of its operation, drive the national online transaction reporting system and provide compared ("locked in") data for the national clearing, settlement and depository system. (This statement is not intended to mean that a broker/dealer may sell a security by hitting all bids without consideration of statistical and historical information, nor relieve him of his fiduciary duties and responsibilities.)

10. The system should display in summarized form, all bids and offers at each price. This will narrow spreads and make the market tighter.

11. The system should take advantage of proven technology to reduce expense to the public and provide prompt and accurate data to the investing community.

C. Organization and Duties of the Governing Board

We believe that a system meeting the stated criteria should be controlled and regulated by a national board selected from among experienced executives from three groups: (a) the broker/dealer community; (b) representatives of corporations, the securities of which are traded through the system and (c) members of the public who are knowledgeable in the securities field. These latter might include academicians, attorneys versed in securities law, accountants, economists, as well as experienced investors .

The National Securities Trading Board, as we shall call it, should have a governing board of seventeen. Twelve would be selected from broker/dealers nationally, giving due regard for geographic diversity as well as the various disciplines within the securities industry, including marketing, trading, computer applications, securities research and international arbitrage. Three members of the board would be selected from senior officers of corporations, the securities of which are traded in the system. Two should be selected to represent the public.

The governing board should make rules, subject to the oversight of the Securities and Exchange Commission, regarding the conduct and operation of the system. The board should have the right to discipline the broker/dealer participants for good cause, including the right to cease to act on their behalf.

The corporation should have regional offices in order to provide administrative and operational support to its participants. Each regional office should cover a specific geographic area to service participants located there. An advisory board should be elected by participants in each region that should be in composition, proportional to the national board. These advisory boards should meet at least annually with their participants and should, through their chairperson, advise the national board of regional needs and problems.

D. Participants

Participants should be of two classes:

1. Broker/dealers qualified by experience and operational and financial expertise to become "qualified" .

2. Corporations that shall, by objective criteria established by the Securities and Exchange Commission, have securities eligible for trading through the corporation.

Initially, operational and financial requirements for broker/dealer participants might be established by using clearing corporation membership standards presently in force by the exchanges and the NASD. Broker/dealers that do not meet clearing corporation standards may use the system through qualified broker/dealers of their choice.

Overview of System Design.

The National Book System should have at least three central computers, redundant and separated from each other. These will control the order flow and maintain queues. In addition, there should be regional computer centers, also redundant, providing additional processing capabilities and connected by communications lines to the broker/dealers in their region. Independent sources of power and voltage regulation should be provided at all central and regional sites.

The system should be so designed that risk of processing failure is minimal. Security techniques of the latest design should be installed, both physically and electronically. Passwords should be varied frequently, and security audits performed continually.

The participant broker/dealers have a choice of access to the system either through terminals or via their own data processing systems. The governing board of the system will make appropriate rules for such interconnections.

E. Operation of the National Book System

1. Agency Orders

A broker in a branch office of Smith & Co., a qualified broker/dealer in the system, would interrogate an inquiry device near his desk for a particular security, just as he does today. However, instead of only a bid/asked and last sale, he would be able to view the market in the security by seeing the "book" of bids and offers summarized at each price. (Bids would be presented from highest to lowest, while offers would be presented vice-versa.) In addition, statistical and historical information about the security would also be available, including volume and average spread statistics, as well as such information as earnings and margin eligibility. As the market changed, updated information would appear.

Let us assume that the market is 42.15 bid for 305 shares with 273 offered at 42.27. The broker receives an order for 175 shares at "the market". Depending on the firm's own internal policy he may enter the order directly into the system, or deliver it to a central order room within his office. In either event, the broker, or his firm's computer, would probably test the market by interposing his customer's bid between the bid and offer. He chooses to bid 42.23 for 175 shares, and the market is instantly updated, becoming 42.23 bid for 175 shares, 273 offered at 42.27.

Another broker, seeing the new bid, and holding an order to sell 100 shares, determines that the spread is reasonable, and enters his order to sell 100 shares at 42.23. This order triggers an execution. A report of the trade is sent by the system to buyer and seller, and a "locked-in" trade report is delivered to a computer file for retransmittal to the national clearing and settlement system.

Simultaneously, an execution message is sent to the national trade reporting system, which will print the transaction on a tape as well as update volume and last sale statistics for any interrogation purposes.

Concurrently, the market as displayed will change to 75 shares bid for at 42.23 and 273 shares offered at 42.27. The buying broker, having received a better price for 100 shares, may elect to raise his bid to 42.27 for his remaining 75 share order. The resultant execution will trigger the same type of process as before, and the displayed market will now be 42.15 bid for 305 shares, 198 offered at 42.27.

Note that the system provides an iterative, or "face to face" type of bargaining. Extraneous information, present in today's system, is eliminated. This includes the identity of the broker in order to effect a trade. The system operator and regulator, of course, may have access to these data, if required. The data are contained in the system, but displayed only to those authorized, and when needed.

Block orders may be negotiated by a market maker or block trader, and then executed through the system, clearing the book in the process.

2. Market Maker Orders

A registered market maker may enter his markets through a terminal device or may employ a computer to assist him in implementing his trading strategy, with the computer generating orders, depending on criteria (which may be continuously variable) established by the market maker. By advertising that he is a market maker in a security, he will attract flows of orders from institutions and others. The order and inquiry flow will insure the economic viability of his function. His firm's anonymity in the National Book with its reporting devices will assure him, moreover, that, unlike a quotation system, "follow the leader" and other strategies may not be employed by unscrupulous dealers.

The system, by its simple design and operating characteristics, will have the capacity to handle volumes not possible in today's era of limited floor space and telephone dialing speed.

Only Good Until Cancelled (GTC) limit orders should be permitted in the system. The broker, however can accept any type of order from his customer, and change limits or cancel in response to customer requirements. This will guarantee that the National Book will always conform to the broker's own open order file. There will be no possibility of "getting out of synch". This provision simplifies the design of the National Book, while permitting brokers to render any suitable services to their clients consistent with sound economic logic.

SECTION II

RESPONSES TO THE SECURITIES AND EXCHANGE

COMMISSION'S RELEASE NO. 12159/MARCH 2, 1976

SECTION II

A. Characteristic 1

We have proposed a National Book System that will become the central repository of all bids and offers in each listed security, rather than disparate "books" maintained by exchanges and others. The repository should use a computer system designed to facilitate competition among the purchasers or sellers of a security rather than among market centers. The system should maintain a single limit order book that is capable of queuing, in time sequence within price, and distinguishing between public and other orders.

With multiple books a variety of unnecessary operational problems, such as the determination of the opening price, suspension of trading, etc., arise. In addition, a system in which the limited price orders are stored in more than one repository rather than a single location is inefficient and the needless source of substantial cost and implementation delays. Finally, multiple "books" render time priority rules futile and may affect the price priority rules as well.

There is no economic merit in detouring agency orders through specialists or market makers, and there are considerable technological problems in doing so. If orders are detoured, there are two possibilities: either the specialist receives an order and keys it into the book or the exchange's computer relays the order directly. In the first case the possibility of unnecessary error is not only considerable, but the detour may be counterproductive. At the opening, a traffic jam may result as a consequence of attempts to enter a large volume of orders in time. Unnecessary delays may occur in the transmission of orders or cancellations.

The second alternative is more efficient, but in that case both the specialist and the relaying mechanism are redundant. If the design of the National Book System limits order entry to specialists and a limited number of other market makers, the policy questions that must be addressed are as follows:

1. Who may be a specialist?

2. How is he selected?

3. How (and by whom) is he compensated?

4. What are his duties?

5. How is his performance measured?

6. How many specialists may there be?

7. How is he removed?

8. Who is a market maker?

9. What are his duties?

10. May he also be a broker?

11. How does he qualify?

12. How does he cease to be a market maker?

Operational questions, as well, are introduced as follows:

1. How is communication between a broker and a particular specialist or market maker established?

2. What is the methodology to be used and routing techniques employed to enter and cancel orders?

3. How do you program a path from a broker's order room to the correct specialist who has the first order at a given price?

4. How do you display the consolidated book if a given specialist (or market maker) has the first, fourth, seventh and fifteenth bid (in time sequence) at a given price?

We believe the best solution, providing the fairest and simplest system is to permit all qualified broker/dealers to enter orders in eligible securities directly into the National Book System either as agent for another broker/dealer or a public customer .

B. Characteristic 2

Even if all qualified broker/dealers are permitted to enter orders in an eligible security, the universe will be finite. The system will merely determine the number of currently eligible "members" and registered market makers. Any member with the necessary financial and operational fitness may initiate and cease market making in any eligible security at any time by notification through the system. Since we believe that exposure to a substantial flow of order and inquiries is an economic requirement for a firm to make a competitive market such flows will only be forthcoming to a firm if the firm makes a continuous market. Competition will thus force market makers to maintain the desired continuity and provide liquidity, and thus there is no need for any commitment rules. Any attempt to require market makers to maintain bona fide two sided markets for a longer period of time than that firm considers economically justified will only result in the widening of spreads. Such wider spreads will be necessary to compensate the firm for the additional risk and capital commitment.

C. Characteristic 3

We believe that a National Book System should be capable of differentiating between public and dealer orders. It is relatively simple to design such a system. For regulatory and statistical purposes, such an identification is useful. Since a broker /dealer is required to know the identity of his account before entering any order, the distinction can be entered at the time of the order.

D. Characteristic 4

We have already addressed the question of time and price priority. We agree that there should be no precedence based upon size. So-called "odd lot" orders should also be included in the National Book System. The present inferior treatment of these smaller orders is unnecessary, and creates multiple systems for trading the same security.

If queuing is performed without size preference, the total quantity bid for (or offered} at each price in a security may be displayed in summarized form , thus providing complete market trading information to all investors .

E. Characteristic 5

We also concur with the concept of an "open book." If all bids and offers in the book are displayed in summarized form, the information available will be accessible to anyone, whether a member of the public, a broker or a market maker.

Full disclosure of bids and offers should narrow the spreads for a security, since temporary disparities will be instantly visible to all. Orders can be entered by both the public and professionals to "close up" the market. As long as the identity of the bidder or offeror is not displayed, the market can be summarized and will become simpler to design. Further, since the system allows execution, identity of the contra party is unnecessary.

The contents of the book are valuable: there is no reason to suppress them, thereby creating a monopolistic value for those who have access to the book's content. In addition, enforcing regulations relating to unauthorized access to a closed book are difficult. Problems relating to the use of inside trading information are eliminated if the book is open.

F. Characteristic 6

We have already stated our belief that any qualified participant should be able to execute against the book directly, and that we believe that the National Book System is the most economic device for achieving the priorities endorsed above. Furthermore, we question how these priorities can be achieved without a single book.

G. Characteristic 7

In the National Book System, all execution reports will be delivered to the involved participants immediately. The National Book System will permit the "locked-in" trade to become a reality, and the execution will furnish information of each trade to both the national clearing and settlement system and the national reporting system. If an order arrives too late to be executed at the price specified or is at a price that fails to trigger an execution, the system will treat that order as any other limit order and notify the broker involved of that fact.

H. Characteristic 8

We agree that the system should provide on-line update. Indeed, we do not see how the system could operate otherwise.

I. Characteristic 9

We do not believe it will be necessary to require explicitly that all orders be entered in the National Book System. However, no participant, having agency or fiduciary obligations will be able successfully to defend an execution in terms of price and time priority, unless that execution occurs in the National System.

Orders to buy (or sell) blocks, however, should be required to "clear" the book at the execution price. A block may be "crossed" by the broker/dealer by entering the bid and offer at the same price. Existing bids or offers at the cross price or better should be executed at the cross price and have precedence over the cross by virtue of time priority.

J. Characteristic 10

Sell orders should be identified as "short" or "long." Furthermore, the system will be designed to suppress entry of improperly priced short sell orders if this is necessary under the rules in force.

APPENDIX E: The Basic Unit of Pricing

It is always well to remember the circumstances under which a particular system has developed. The present practice of trading securities in fractional units arose because two brokers kept trying to "split the difference." When spreads of whole dollars become too large, "at a half" became the cry. This process continued until eighths were reached.

While some securities, notably low priced ones, trade at sixteenths, only U.S. Governments and similar securities trade at smaller fractions (U.S. Governments trade at fractions as small as 1/128). Over the decades, attempts have been made by many to encourage decimal unit trading. Brokers resisted, because of the possible confusion of prices in a face-to-face crowd environment. Quotations in tenths, rather than in eighths, could cause mistakes and losses through errors. Thus, the custom of trading in fractions has continued to the present.

Under the National Book System, such restrictions disappear. The computer is an ideal tool to aid the decision makers by serving as a "bookkeeper" to keep inventories, average costs, as well as markets displayed in any form. Narrower spreads are immediately possible if trading is conducted in minimum differentials

In addition, the National Book System has the advantage of making possible a central market with minimum numbers of rules. The basic price and time priority rules will themselves achieve the objective of "394" type rules. If pricing changes to cents, no broker can claim to have achieved "best execution" unless he has attempted to execute his trade between the bid and offer as long as the spread is greater than minimum unit of pricing. In addition, even when the spread is one unit, the broker cannot claim to have honored the time priority rule unless the order is executed through the system.

Decimalizing humpty dumpty.doc

3-21-00

Footnotes

¹ Rube Goldberg was an American cartoonist who depicted complicated, but totally useless machines.

² The relevant text of this Section is attached as Appendix "A."

³ The Peake-Mendelson-Williams proposal is appended as Appendix "B."

⁴ For example, in Joel Seligman's The Transformation of Wall Street, (Revised Edition), Northeastern University Press, Boston, 1995, p. 526, "...the most widely discussed hard CLOB proposal was that of securities industry consultants Peake and Williams and University of Pennsylvania Economics professor Mendelson. (Hereinafter "Seligman")

⁵ Joint Hearings before the Subcommittee on Oversight and Investigations and the Subcommittee on Consumer Protection and Finance of the Committee on Interstate and Foreign Commerce, House of Representatives, September 21, 24 and 25, 1979, Serial 96-89, U.S. Government Printing Office, Washington, 1979.

⁶ "Market 2000: An Examination of Current Equity Market Development," Division of Market Regulation, U.S. Securities and Exchange Commission, Washington, January 1994.

⁷ Ibid., p. III-6.

⁸ Ibid., pp. IV-9&10.

⁹ Testimony of Steven M.H. Wallman before the Subcommittee on Finance and Hazardous Materials, Committee on Commerce, U.S. House of Representatives on April 10, 1997

¹⁰ Securities and Exchange Commission Release No. 34-42360, January 28, 2000.

¹¹ Ibid.

¹² "NASD Calls On The SEC To Delay Decimalization," March 7, 2000, Nasdaq web page: http://www.nasdaq.com/reference/sn_nasd_calls_on_SEC_to_delay_decimalization_000308.stm

¹³ SEC Press Release 2000-28, March 10, 2000.

¹⁴ Judith Burns, "Lawmakers Raise Concerns About Decimal Delay," Dow Jones Newswires, March 14, 2000.

¹⁵ It is interesting to note that investors and issuers, for whose benefit the markets exist, were not named specifically.

¹⁶ For example, see speech by Commissioner Laura Unger, "Technology Bytes the Securities Industry: The New Millennium Brings New Investors and New Markets," March 14, 2000, Securities Industry Association: "The broker must take into account opportunities for price improvement along with a number of factors, including: (a) size of the trade; (b) speed of execution; (c) the likelihood of execution; (d) the type of security; (e) information about and access to competing markets; and (f) cost of access."

¹⁷ Release No. 34-42450; File No. SR-NYSE-99-48

¹⁸ Commissioner Unger, in the speech cited at footnote 15, addresses fragmentation as follows: "The six options provide a sort of sliding scale of alternatives to resolving market fragmentation, starting with disclosure and ending with time/price priority. I am somewhat partial to Option # 1 since it incorporates one of my report's recommendations-disclosure-to the Commission on best execution."

¹⁹ The "NBBO" stands for the "National Best Bid and Offer" displayed in the CQS.

²⁰ Seligman, p.407