1 Securities and Exchange Commission 2 3 Roundtable: Market Structure Hearing 4 Date: October 29, 2002 5 Pages: 1 through 214 6 Location: Securities & Exchange Commission 7 450 Fifth Street, N.W. 8 Washington, D.C. 20549 9 ATTENDEES: 10 SECURITIES & EXCHANGE COMMISSION 11 Commission, Chairman Harvey Pitt; 12 Commissioner Harvey Goldschmid, 13 Commissioner Paul Atkins 14 Commissioner Roel Campos 15 Commissioner Cynthia Glassman 16 Annette Nazareth 17 Bob Colby 18 Alden Adkins 19 Larry Harris, Chief Economist 20 21 22 23 24 Diversified Reporting Services, Inc. 25 (202) 467-9200 1 ATTENDEES Continued 2 SECURITIES & EXCHANGE COMMISSION: 3 Giovanni Prezioso, General Counsel 4 Meridith Mitchell 5 Steve Jung 6 Mary Head 7 Brian Stern 8 9 ADDITIONAL ATTENDEES 10 Professor Maureen O'Hara, Cornell 11 Wayne Wagner, Plexus Group 12 Sal Sodano, Amex 13 Gus Sauter, Vanguard 14 Larry Leibowitz, Schwab 15 Rick Ketchum, NASDAQ 16 Cathy Kinney, The New York Stock Exchange 17 Bernie Madoff, Madoff 18 David Whitcomb, Automated Trading 19 John Markese, American Association of Individual Investors 20 Ed Nicoll, Instinet Corporation 21 Gary Gastineau, ETF Advisors; 22 Prof. Joel Hasbrouck, NYU 23 24 25 1 P R O C E E D I N G S 2 COMMISSIONER GLASSMAN: Good morning. For those 3 who don't know me, I'm Commissioner Glassman. And on behalf 4 of the Commission, I want to welcome all of you today to this 5 market structure hearing, and to thank you all for your 6 participation. I know I speak for all of my colleagues when 7 I say that we've been looking forward to this interactive 8 session today, and our next session in New York next month, 9 with great anticipation. 10 The purpose of today's session is to educate the 11 Commission on issues relating to the structure of the U.S. 12 equity markets, with emphasis on how the public interest and 13 the protection of investors can best be served. A key sub- 14 theme is whether our well-intended rules to protect investors 15 and to protect the public are having unintended negative 16 consequences. We want the benefit of your views on the 17 topics we have chosen for discussion today. Even more useful 18 to us will be the opportunity to listen as you, the 19 panelists, discuss and debate this array of issues. 20 In my nine months as Commissioner, I've had 21 meetings with representatives of many different markets and 22 market participants, and I've learned a lot from each of the 23 meetings. But what I really wanted to do was to get all of 24 the differing points of view on a particular topic 25 represented in one room and let people advocating different 1 positions have the opportunity to counter each other with no 2 holds barred. And that's what I hope will happen today. 3 As far as our expectations, it's obvious that two 4 days of hearings, no matter how productive they are, are not 5 going to answer all of our questions. But today's discussion 6 will enhance our understanding of current market structure 7 policy concerns, as well as the possible tradeoffs that may 8 have to be made as we attempt to resolve the important issues 9 before us. 10 Before we get started, I'd like to thank the key 11 members of our market structure working group: Annette 12 Nazareth, Bob Colby, and Alden Adkins and their staff in 13 Market Reg; Larry Harris, our Chief Economist, and his 14 colleagues; our general counsel, Giovanni Prezioso, and his 15 principal associate, Meridith Mitchell; Steve Jung on the 16 chairman's staff; and last, but not least, Mary Head and 17 Brian Stern on my staff. Now let's get started. I'll turn 18 the session over to Annette. 19 MS. NAZARETH: Thank you, Commissioner Glassman. 20 I, too, would like to welcome the Commission, our 21 distinguished panelists, and members of the public, to this 22 first market structure hearing. Today the continue -- the 23 Commission continues in its more than 30-year tradition of 24 critically examining issues related to the structure of our 25 equity markets. Prior Commissions have faced many complex 1 and novel issues in this particularly challenging area of 2 securities regulation, both before and after the Securities 3 Act Amendments of 1975 that gave the Commission a mandate to 4 build a better National Market System. 5 To its credit, the Commission took actions that 6 have resulted in a vibrant, healthy, and efficient 7 marketplace -- A marketplace, indeed, that is the envy of 8 the world. As technology continues to transform our lives 9 and the way we do business, once again we stand at a 10 crossroads with respect to various market structure issues. 11 And the solutions we arrive at will have far-reaching 12 consequences, especially in this highly competitive 13 marketplace. 14 We hope that these hearings will achieve several 15 important objectives. First, as Commissioner Glassman 16 mentioned, we hope that these hearings will provide a 17 comprehensive picture of the structure of the U.S. equities 18 markets. Our marketplace consists of a sweeping variety of 19 market centers and trading methods that cater to different 20 types of securities and different categories of investors. 21 Some markets are centralized; others are decentralized. Some 22 markets are driven by the interaction of customer orders; 23 others are propelled by the willingness of designated market 24 intermediaries to commit capital. Some markets are grounded 25 in a physical trading floor where bids and offers meet face- 1 to-face; others float in the anonymous realm of cyberspace. 2 Another goal of these hearings is to discuss the 3 complex tradeoffs that we have to make in regulating market 4 structure. How should we reconcile the investor's need to 5 obtain current information about market activity with each 6 market center's desire to exploit the commercial value of the 7 data it generates? How should we reconcile the investor's 8 desire to have access to the best prices with the right of 9 individual market participants to employ the business model 10 of their choice? And when should a market be required to act 11 in the public interest, rather than in the interest of its 12 members or shareholders? These questions do not have simple 13 answers, and there are many difficult decisions to be made. 14 We are nevertheless committed to resolving these competing 15 interests in the fairest and most balanced manner. 16 A third goal of these hearings is to use what we 17 learn to maintain consistent policy decisions on the key 18 issues facing the Commission. In recent years, the 19 Commission has undertaken a number of market structure 20 initiatives. Just to name a few, the Commission adopted 21 order handling rules for market makers and ECNs in 1996; 22 Regulation ATS governing alternative trading systems in 1998; 23 and order execution and order routing quality disclosure 24 obligations for market centers and retail brokers in 2000. 25 While some may differ, most would say that these rules have 1 significantly improved the transparency and accessibility of 2 our markets. 3 The Commission has also solicited comment on a 4 number of specific areas of market regulation. Among others, 5 the Commission has issued concept releases on the regulation 6 of market data fees and revenues, the fragmentation of the 7 U.S. securities markets, and the regulation of exchanges. 8 The Commission also convened an advisory committee in 2000, 9 chaired by Dean Joel Seligman of the Washington University 10 School of Law in St. Louis, to consider issues related to the 11 collection and dissemination of market information. These 12 efforts to develop sound policies for market regulation must 13 continue if we are to make our markets fairer, more 14 efficient, and more transparent. To this end, the Commission 15 seeks to ensure that the decisions it makes in the coming 16 months will benefit from the wise counsel of the industry, 17 the academic community, and the public. 18 Before I introduce our participants today, I'll 19 briefly mention just a few housekeeping matters. This 20 morning there will be two sessions. In this opening session 21 we will touch on the overarching principles of market 22 structure, and discuss whether there is consensus on those 23 principles. This session will last only an hour, and will be 24 followed by a 15-minute break between 10:30 and 10:45. We'll 25 then follow with our first major session on market data, to 1 be moderated by Larry Harris, our Chief Economist. At 12:15, 2 we'll break for a one-hour lunch, and we'll pick up with our 3 afternoon program at 1:15. 4 As to general ground rules, in order to keep the 5 dialogue moving and to afford everyone an opportunity to 6 speak, I would ask that each of you attempt to limit your 7 responses to questions to about two to three minutes. Given 8 the limitations on time, we also ask that you try not to 9 repeat points already made. And, now, the next ground rules 10 get even tougher. Please attempt to leave all egos at the 11 door, and try to think about the questions and topics from a 12 public policy perspective, and not simply from the 13 perspective of your own business model. And, finally, keep 14 in mind that the Commission's goal ultimately is the 15 protection of investors. 16 I would also like to note that we will have 17 Commission staff occasionally coming around into the audience 18 with note cards, and the audience is invited to write 19 questions for our participants. We will try to address some 20 of those questions raised by the audience as time will allow. 21 So, with that out of the way, I'd like to introduce 22 each of the participants, and then circle back for any 23 opening remarks that each of you may wish to make. First, 24 I'd like to introduce our Commission, Chairman Harvey Pitt; 25 Commissioners Goldschmid, Atkins, and Campos; and, of course, 1 Commissioner Glassman. I would especially like to thank 2 Chairman Pitt for his leadership in calling for these 3 hearings, and Commissioner Glassman for her substantial 4 efforts in putting together this program today. We also have 5 some SEC staff at the table besides myself. We have Bob 6 Colby, who is the Deputy Director of Market Regulation; Larry 7 Harris, our Chief Economist; and Giovanni Prezioso, our 8 general counsel. 9 And I'll just briefly go around and introduce the 10 rest of the panelists, and then we'll circle back and give 11 anyone the opportunity to -- to make some introductory 12 remarks. Skipping our -- at my own risk, skipping our 13 Commissioners, we have Professor Maureen O'Hara from Cornell; 14 we have Wayne Wagner from the Plexus Group; Sal Sodano from 15 Amex; Gus Sauter from Vanguard; Larry Leibowitz from Schwab; 16 Rick Ketchum from NASDAQ. And then we circle all the way 17 back here to Cathy Kinney from the New York Stock Exchange; 18 Bernie Madoff from the eponymous firm Madoff; David Whitcomb. 19 We have -- I don't know -- have I... 20 MR. WHITCOMB: Automated Trading. 21 MS. NAZARETH: Oh, I'm sorry, David. I should have 22 written that down. 23 MR. WHITCOMB: That's okay. 24 MS. NAZARETH: Automated Trading. John Markese, 25 who is from the American Association of Individual Investors; 1 Ed Nicoll from Island (sic); Gary Gastineau, who is from ETF 2 Advisors; Joel Hasbrouck from NYU. 3 And with that, I will circle back, I guess, 4 starting with Maureen. If -- if there's anybody -- you're 5 certainly welcome to skip if you don't have any introductory 6 remarks. But if you would like to take your allocated two or 7 three minutes for some introductory remarks, you're welcome 8 to do so. Otherwise, you may pass. 9 MS. O'HARA: I'll pass. 10 MS. NAZARETH: Okay. 11 MR. WAGNER: My name is Wayne Wagner. I am 12 Chairman of the Plexus Group. We study institutional 13 trading. And if I represent any interest here, it would be 14 the interest of the institutional investors. And, you know, 15 the institutional trading dominates the -- the volume, and 16 not necessarily the thinking of the -- of the Commission 17 here. 18 When we look at institutional trading, we find that 19 it's really quite different from the -- from the retail 20 trading. We see that institutional orders average over 21 250,000 shares a -- a -- as the portfolio manager is making 22 the decision. And this gets reduced down through a process 23 of going from the trader to the broker to the exchange, to 24 average orders which are somewhere around 1000 shares here. 25 And this reduction process is a -- is a difficult one that -- 1 that leads -- what we find is that over half of the 2 institutional dollars traded exceed 20% of daily volume as 3 the portfolio manager is thinking of the trade that he wants 4 to make, or she now, while only 1/6th of the trading occurs 5 in a volume of more than 20% of daily volume. So there's 6 obviously a three-to-one dispersion of the orders that 7 portfolio managers want to fill in order to manage funds, and 8 what actually gets traded. 9 And, you know, that means that if those trades are 10 not traded on the day that the portfolio manager wants it, 11 they're carried over to further days. And that induces 12 delays here. One thing Plexus is most famous about is 13 describing an iceberg of transaction costs, where the bulk of 14 the costs lie unseen in the delay costs and opportunity costs 15 of the inability of institutional traders to trade fast 16 enough. Those costs represent approximately 75 to 80% of the 17 total cost of executing trades here. And they're sort of off 18 the -- out of the range of normal thinking about market 19 structure and -- and the costs of -- of executing here. 20 But to my mind, these institutional investors are 21 managing funds for individuals here, and the interest of -- 22 of breaking these orders down and -- to the degree that they 23 have to be broken down in order to get through the exchange 24 leads to these delay costs, and it is a friction and an 25 inefficiency in taking the research that these institutional 1 investors do, and allowing that to have its full impact on 2 the performance of the portfolios. Thank you. 3 MS. NAZARETH: Thank you. 4 Sal? 5 MR. SODANO: Thanks, Annette. I'm Sal Sodano, 6 Chairman and Chief Executive Officer of the American Stock 7 Exchange. And first I'd like to thank the Commission for 8 holding these very important meetings. The American Stock 9 Exchange looks forward to working with the Commission and 10 other market participants to address many very challenging 11 issues in the weeks and months ahead. And please note that I 12 use the time frame of weeks and months ahead, in that these 13 issues really are very important and require, as we know, 14 very timely focus. 15 The American Stock Exchange has a very long history 16 of innovation and is unique among U.S. security markets in 17 that. We are the only U.S. market that actively lists and 18 trades across three diverse product lines: equities, 19 options, exchange-traded funds, commonly referred to as ETFs. 20 Quite frankly, we invent investment concepts and products. 21 Therefore, we'd like to clarify that the term 22 "entrepreneurial entrent" as used in the outline for these 23 hearings could equally apply to exchanges, especially the 24 American Stock Exchange. 25 Indeed, in this context I'd like to make some 1 general recommendations. And I will basically only make two, 2 taking into account the limitation on time. First, as we 3 move forward, it is important that the Commission ensure, as 4 you said, Annette, in your early comments, that the 5 principles enshrined in the congressionally mandated National 6 Market System be restored: best execution, transparency, 7 equal regulatory oversight, and fair competition. To achieve 8 this, we believe the last year-and-a-half, especially the 9 recent weeks with Island going dark, demonstrates that 10 Regulation ATS did not and is not serving its intended 11 purpose. Therefore, we believe the first step should be to 12 abolish Regulation ATS and require all markets that meet the 13 definition of an exchange to basically register as such. 14 Further, we believe the Commission should take very 15 concrete steps to modernize the exchange regulation. Again, 16 Annette, as you pointed out, many of these regulations date 17 back to 1975. The disparity between markets is simply too 18 vast. These need to be addressed. 19 Finally, the Commission should bring an immediate 20 end to an obvious conflict of interest that has led to 21 nothing but regulatory abuse, reduced liquidity, and 22 therefore investor harm. Market-sponsored payment for order 23 flow must and should be eliminated immediately. Thank you. 24 MS. NAZARETH: Thank you, Sal. 25 Gus? 1 MR. SAUTER: Thank you, Annette. I'm Gus Sauter. 2 I oversee the portfolio management and trading of Vanguard's 3 internally managed equity funds. We believe that there is a 4 very significant role that the financial intermediary 5 community should fulfill, and we think of the -- this 6 community as including exchanges, specialists, market makers, 7 broker-dealers, institutional money managers, and so on. 8 We think that that role is to provide for the 9 efficient flow of capital from investors to businesses. Its 10 secondary role is to effect efficient exchange of ownership, 11 whether it be stocks or bonds, from one investor to another. 12 We believe that the marketplace should provide for best 13 execution. And we define "best execution" in several 14 characteristics, including consideration for immediacy and 15 certainty of execution, as well as execution at the most 16 favorable price. 17 To satisfy best execution, we believe that the 18 markets need to be liquid, and that the structure of the 19 market should be designed to provide for this greatest 20 liquidity. And I fully expect we will be discussing this 21 further, but in short, we believe that price-time priority is 22 paramount for a most liquid market. We also believe, to cut 23 to the chase, that a central limit order book satisfies the 24 best efficiency of the marketplace. Thank you. 25 MS. NAZARETH: Thank you. 1 Larry? 2 MR. LEIBOWITZ: Okay. I'm Larry Leibowitz. I run 3 the Equities Division of Schwab Capital Markets. We are one 4 of the largest retail firms in terms of customer accounts, 5 over nine million accounts, both traditional retail and 6 direct access traders. We're the 14th largest mutual fund 7 complex, a growing institutional business, regional 8 specialists, and one of the largest OTC market makers, and a 9 founding member of one of the most successful ECNs. Bottom 10 line is we're affected by almost any of the different 11 viewpoints that -- that come across this table, and we view 12 it with many different hats. The bottom line for us is that 13 markets that are open, transparent, with equal access for all 14 is what's going to be good for individual investors, 15 institutional investors, and, in the long run, good for the 16 market. 17 We believe in the critical importance of 18 competition to encourage maximum innovation. We're strong 19 supporters of a National Market System of multiple competing 20 markets. We're always wary of SROs abusing their regulatory 21 authority to advance their competitive interests. Allowing 22 markets to regulate their competitors creates inappropriate 23 conflicts of interest. We believe the SEC must unbundle the 24 functions of markets and the regulator. The concept of 25 competition among interacting orders is code for the CLOB. 1 While it may be a desirable outcome to some, order 2 interaction should never be promoted at the expense of 3 competition among the markets. 4 MS. NAZARETH: Thank you. 5 Rick? 6 MR. KETCHUM: Thank you, Annette. I want to first 7 agree with your characterization that, speaking from a NASDAQ 8 perspective, I think in the last ten years there has been, in 9 large part as a result of Commission oversight, a dramatic 10 change for the good in markets overall, and in the NASDAQ 11 market, in particular, encouraging competition, innovation, 12 while balancing that with greater rights for investors to be 13 represented immediately in the marketplace. 14 I think at the same time, though, that historic 15 high watermark for both individual and institutional 16 investors' participation in the NASDAQ market also sets up, 17 by those very same actions, risks that indeed, remembering 18 your stricture at the beginning, raise particular risks for 19 the protection of investors, unless the SEC steps in actively 20 to mold an efficient and competitively fair environment that 21 encourages both investor protection and risk-taking. 22 In short, we believe that market linkages are 23 needed now, perhaps more than ever, but not cynically 24 inefficient ones that degrade these existing trading 25 environment. We believe that best execution standards must 1 take into account the accessibility in markets in designing 2 what those standards are. We believe that competition must 3 be fair among ECNs and market makers, among ATSs and markets. 4 That requires, at a minimum, consistent rules with respect to 5 access fees and -- and greater flexibility with respect to 6 exchanges in markets such as NASDAQ, and being able to move 7 and respond to competitives threats. 8 And finally, we believe that the Commission must 9 aggressively step in to ensure that legitimate competition 10 among markets do not spawn -- does not spawn regulatory 11 fragmentation and arbitrage that seriously harms investor 12 protection. Thank you. 13 MS. NAZARETH: Thank you. 14 Cathy? 15 MS. KINNEY: I'm Cathy Kinney, the President and 16 Co-Chief Operating Officer of the New York Stock Exchange. 17 Mr. Chairman, members of the Commission, and senior staff, 18 the NYSE is pleased to participate in today's forum, and 19 would like to thank and commend the Commission for opening 20 this dialogue. 21 Nothing is more important to the future of our 22 capital markets than reestablishing the trust and confidence 23 investors place in our system. The trust and confidence has 24 been severely tested in recent times by accounting scandals, 25 by the breach of senior corporate officials and professionals 1 in discharging their obligations, by compromised analysts 2 betraying investors, and by questionable IPO practices. We 3 feel that we are one storm short of the perfect storm. Dark 4 clouds have already gathered over the secondary markets. For 5 notwithstanding the best of intentions, competition and 6 innovation, we've allowed a number of anti-investor practices 7 to gain a foothold in our market practices at the core that 8 fail to put the investor first. 9 Today in our market system, internalization, 10 preferencing, payment for order flow, and other forms of pay- 11 to-play compromise the agency obligations of brokers, and 12 interpose dealers between customer orders. Markets engaged 13 in regulatory arbitrage disseminate non-competitive quotes 14 and purchase tape prints that inflate their volume and convey 15 phantom liquidity. These practices are bad for the health of 16 the markets, and erode public trust and confidence. So far, 17 our secondary markets have escaped the assignment of 18 responsibility for the $9 trillion in missing investor 19 savings. If we do not move quickly, we will have -- all have 20 very much to answer for when the current storm comes, only to 21 find ourselves in the eye of the storm. 22 Mr. Chairman, members of the Commission, and staff, 23 we now have the opportunity to shore up the foundation of 24 transparency, and the resolution of conflicts in our markets 25 for current and future investors. Set the standard for 1 markets and for broker-dealers to do better, much, much 2 better to restore the investor to the premier position by 3 fostering order competition within markets, and at the same 4 time, foster competition among markets. Optimize public 5 price discovery, and delineate and demand best execution, 6 which are two sides of the same protection for investors. 7 Make size, price, and depth, breadth of services, compression 8 of execution costs, and quality of information the basis for 9 intermarket competition. Attack not just the symptoms, but 10 the root causes of anti-investor practices. Require that 11 customers go first. Preclude dealers from trading ahead of 12 public customers on exchanges, and outlaw practices like 13 internalization, preferencing, and phantom auctions. Ban 14 play to -- pay-to-play schemes of all varieties. Preclude 15 any market from hiding its quotes, blocking access to its 16 facilities, or ignoring better priced orders resident in 17 other markets. Restore tape integrity by making the place of 18 execution the place of tape reporting. Allow markets to 19 withdraw from the data consortia, end market data rebates, 20 and painting the tape by ending the subsidies that fund them. 21 Principles of the free market process should be 22 strengthened. Remove advocates of the free-for-all market 23 process from our system, and neutralize the systemic risk 24 they impose. Now is the time to bring investors back to the 25 safe harbor of the finest, most admired and tightly regulated 1 market system the world has known. To do anything less would 2 jeopardize America's tradition of leadership, and betray 3 America's 85 million investors. Thank you. 4 MS. NAZARETH: Thank you, Cathy. 5 Bernie Madoff? 6 MR. MADOFF: Good morning. I'm Bernie Madoff from 7 Madoff Securities. These are very difficult issues, and I 8 have been participating in -- in this process for certainly 9 close to 30 years of the 40 that I've been in the industry. 10 And I would like to thank the Commission for establishing 11 this type of forum. 12 I would encourage the Commission to not give up on 13 the industry's role in helping to participate in solving the 14 problems that we are constantly facing. That being said, I 15 think the Commission plays a pivotal role. And, as Rick 16 Ketchum said, the only way that these issues will ever hope 17 to be resolved is with the active participation of the SEC in 18 this process. I would also like to say that I think the 19 industry and investors would be very well served if all of us 20 that are participants in this marketplace focused in on what 21 we think is the right thing for the investor and for the 22 market, overall. And I've never been sure that those are two 23 related issues. And to lay aside the individual marketing 24 and the individual self-interest that the various market 25 participants all have in this whole issue. 1 So this is a difficult process. I'm sure it's -- 2 it'll be ongoing. And I think this is a great opportunity 3 for all of these issues to be aired. So thank you for having 4 this function. 5 MS. NAZARETH: Thank you. 6 David Whitcomb? 7 MR. WHITCOMB: Hi. I'm Dave Whitcomb of Automated 8 Trading Desk. I want to thank the Commission also for having 9 the hearings, and for inviting me. 10 Let me give you first my background and biases. 11 I'm a retired finance professor from Rutgers, and I'm also 12 founder and chairman of Automated Trading Desk. ATD develops 13 and operates expert systems for fully automated limit order 14 trading. We have two broker-dealer subsidiaries. ATD 15 Brokerage Services does proprietary trading, and trades for 16 institutions. Chicago Securities Group is a specialist on 17 the Chicago Stock Exchange. The two subsidiaries trade about 18 4% of the NASDAQ volume, 2% of listed volume, and about 10% 19 of total ECN volume. So you won't be surprised that I will 20 focus my comments on how to make the world safe for limit 21 order trading. You will be surprised that I still favor 22 decimalization, despite the fact that it cut ATD's profits in 23 half on the day it began. You'll also be surprised that I 24 favor radical reform of listed trading, despite the effect it 25 might have on our Chicago specialist unit. 1 The order handling rules and decimalization 2 revolutionized the NASDAQ market. They freed limit order 3 trading via ECNs to compete with dealers' quotes in what I 4 call E-market making. The result is much tighter spreads and 5 new pools of liquidity, although it's what I call distributed 6 liquidity. Market makers and E-market makers like us make 7 less money in this environment, but the market is growing and 8 modernizing, and I think it's good for us in the long run. 9 I want to complete the revolution by extending the 10 order handling rules to the listed market, ending a variety 11 of restraints on competition between limit orders and 12 specialists. Among the restraints are delays in displaying 13 limit orders, delays in executing market orders, stepping 14 ahead by specialists. Nothing wrong with stepping ahead, but 15 only if limit orders can compete. Stepping ahead for the 16 benefit of investors. 17 I also want to demutualize the exchanges and 18 trading systems to put rule making and enforcement where it 19 should be, at the SEC. SROs were a bad idea in the New Deal, 20 and they're a bad idea today. Thank you. 21 MS. NAZARETH: Thank you. 22 John? 23 MR. MARKESE: Good morning everyone. I'm John 24 Markese. I am President of the American Association of 25 Individual Investors. We're an educational group, and I have 1 to say that the one thing we can't seem to educate our 2 investors on are market structure issues. But if they listen 3 to the comments that we've heard already around the table, 4 they'd be even more depressed than they currently are about 5 the stock market. 6 (Laughter.) 7 MR. MARKESE: While they -- they know we're not 8 trading under the Buttonwood Tree any longer, markets have 9 gotten away from them. They don't quite know how their 10 orders are executed. I wouldn't give them a test on any of 11 this, but I have to say this, that they still demand and they 12 still want -- and there's no one individual investor, but 13 they still demand liquidity; they still demand regulation; 14 they still demand timely, pertinent information; they want 15 operational efficiency; and they don't want to pay anything 16 for it. That has not changed. 17 (Laughter.) 18 MR. MARKESE: I fear, looking at this from a view 19 -- I'm also an ex-academic and a few other things along the 20 way, that what I see in this marketplace today, we have 21 fragmentation of regulation; we have fragmentation of 22 execution; in fact, we have simply fragmentation. We need to 23 coordinate this. We need a national -- a true National 24 Market System. The technology, the products have run faster 25 than our abilities to make it all work well together. 1 So I'm here for best execution, and I don't think 2 there's any one definition that I would subscribe to of best 3 execution. At times it's -- it's certainty; at times it's 4 time priority; at times it's price; at times it's the 5 package, meaning the brokerage costs and everything else. So 6 it's a complex issue. Individual investors have an inherent 7 interest in this, but unfortunately they're not as well 8 informed as they should be. I think we have some serious 9 issues that we need to discuss, and I look forward to today. 10 Thank you. 11 MS. NAZARETH: Thank you. 12 Ed? 13 MR. NICOLL: Well, good morning, Mr. Chairman, 14 members of the Commission. Thank you for this opportunity to 15 participate in today's round table. Annette, I am Ed Nicoll, 16 CEO of Instinet. 17 MS. NAZARETH: I know. I -- I apologize. 18 (Laughter.) 19 MS. NAZARETH: I -- I've got to get with the 20 program. 21 (Laughter.) 22 MR. NICOLL: And Instinet is the world's largest 23 electronic agency broker serving its customers for over 30 24 years. Today Instinet handles almost a third of all the OTC 25 shares traded in NASDAQ listed securities. There could not 1 be a better time to have these meetings. We are probably 2 facing today a greater number and more diverse set of 3 challenging public policy issues than we've seen in years. 4 While some may be concerned about how to manage these issues, 5 I actually see them as an opportunity for all of us, market 6 participants and regulators, to shape a stronger, healthier, 7 and more robust marketplace. 8 Perhaps the greatest common theme cutting across 9 these challenges is how to integrate electronic markets into 10 the National Market System, while at the same time ensuring 11 that both traditional and electronic markets can offer their 12 unique products and services. Let us today not get 13 distracted into debating which is better; but, instead, let's 14 see if we can begin to lay out the regulatory framework that 15 would allow both models to compete fully and fairly, and 16 deliver the greatest value to all investors. Thank you. 17 MS. NAZARETH: Thank you. 18 Gary? 19 MR. GASTINEAU: I'm Gary Gastineau. In my career 20 I've been on the buy side, I've been on the sell side, and 21 I've even worked for an exchange. Effective Friday morning, 22 I will be an issuer. So I'm probably as close to John 23 Markese, in terms of what I'm interested in, as anyone else 24 here. 25 Over the years I've sat at the feet of gurus like 1 Wayne Wagner and Larry Harris talking about market 2 microstructure and transaction cost issues. And as I read 3 over the sort of talking points or discussion points for 4 today, I think if I'd been a Commissioner I would be either 5 extremely pessimistic or extremely optimistic. Pessimistic, 6 in that I doubt very much that it's possible to solve all 7 those problems in any kind of coherent way. Optimistic in 8 the sense that, unless I do something really bad, I can't 9 make it much worse. 10 (Laughter.) 11 MS. NAZARETH: Joel, can you one-up that? 12 MR. HASBROUCK: This is a bad position. I teach at 13 New York University, and one of the courses I teach is a 14 course in trading and markets. And I can certainly attest to 15 the complexity of the securities markets, at least my 16 students can. Because by the time I find I finish a 17 semester, we can just cover really only essentially the 18 basics. 19 I would also note that many of the things we're 20 talking about today are about tradeoffs. One of the basic 21 tradeoffs is going to be between what the SEC, in its 390 22 release, called "price competition," that is, all -- bringing 23 all of the trading activity together in the search for the 24 best price, with market center competition, which is the 25 innovation and competition of diverse trading mechanisms. 1 The SEC, in its Rule 390 concept release, outlined these two 2 goals. One or the other will not be chosen. The best we can 3 do is strike a balance. Thank you. 4 MS. NAZARETH: Thank you. Well, I think you can 5 all tell from these introductory remarks that we're in for a 6 very interesting dialogue in the course of the day. I think 7 that most of the major issues that we'll be grappling with 8 were, in one way or another, touched upon in -- in your 9 opening statements. 10 I thought before we move on, you know, in the 11 program to the -- to the more specific programs, we should 12 have a very general discussion about what the core precepts 13 are of the National Market System, to see whether, going 14 forward, we agree that those are the principles that -- that 15 we agree should be those that we're operating under, you 16 know, whether there is consensus on those. So I'll go 17 through them. 18 The core principles of the National Market System 19 are, first, transparency of information. That is, the 20 availability to brokers, dealers, and investors of 21 information with respect to quotes and trades. The second is 22 economically efficient execution of securities transactions. 23 The third is ease of obtaining best execution. The next is 24 fair competition among markets and intermediaries within 25 markets. And finally, the opportunity, consistent with the 1 aforementioned goals, for investor orders to be executed 2 without the participation of a dealer. 3 So again I think it would be helpful to have a 4 brief discussion to establish whether there is sort of 5 general consensus on these core principles before we move on 6 to a more detailed discuss of market structure. And if these 7 aren't the correct goals---and obviously these are sort of 8 our template, you know, this is what the legislation requires 9 that we have as our goals in -- in regulation---if they're 10 not the correct goals, what should the goals be. And 11 obviously, any changes would require legislation. 12 Since we're starting out, though, in this -- in 13 this brief segment with a somewhat theoretical topic, I 14 thought I'd turn to one of our master theoreticians here, 15 Maureen O'Hara, and ask you -- start with you, and ask you... 16 Sure, you have a... 17 MR. COLBY: Well, I just want to say if -- if you 18 don't believe there should be any National Market System or 19 any Commission involvement at all, this is probably the time 20 to say it. 21 MS. NAZARETH: Right. Well, that's true. If you'd 22 like us to lay off everything. I'd start with Maureen, but 23 we don't necessarily have to go, you know, around the room 24 with everyone opining. What -- whoever would like to 25 participate. 1 MS. O'HARA: Well, thank you, Annette. 2 Let me start off and say that I think that the 3 National Market System is a wonderful dream, and it's totally 4 impractical. We'll just start with there, in the sense that 5 much of the difficulties that people around the table have 6 alluded to is that we all agree on the notions of fairness 7 and competition. But, by definition, what becomes 8 competitive for one market oftentimes becomes anti- 9 competitive for another. So I think the challenge that this 10 Commission faces---and let me stress that I do think you play 11 an extremely important role---while I believe firmly that 12 where the -- that the competition that we've seen the last 13 decade amongst market centers has dramatically improved the 14 quality of the U.S. markets, the difficulty is that as we 15 move forward into an era of increasingly privately owned 16 trading systems, we need a central, shall we say, market cop 17 to keep things even. And so I would encourage the Commission 18 to recognize that their role is probably greater now than 19 ever. 20 Having stated that, I do believe that the National 21 Market System, as it's currently envisioned, is too 22 simplistic. The challenge that the Commission faces is 23 recognizing that for individual customers the ability to get 24 a fair price may not come at the same expense that 25 institutions face, as Wayne points out. Moreover, the 1 ability to compete by smaller markets is only really 2 feasible, given our current National Market System, if they 3 can somehow manage to compete in a way that may not 4 necessarily result in a better price for the consumer. 5 These are not all necessarily bad tradeoffs, but I 6 think they require us to think more seriously about the 7 difficulties of trying to balance features such as 8 transparency and efficiency. When I look at efficiency, I'm 9 more concerned about the ability of corporations to raise 10 funds efficiently, and investors to be able to trade at 11 prices that approximate the fair values of securities. 12 As we fragment, the only way fragmentation gets us 13 there is if market centers that have better ways of 14 accomplishing either of those goals are able to compete. But 15 our current system, I think, leads us far astray. Our 16 national best bid and offer to me has become problematic 17 because it isn't the national best bid or offer. You can 18 trade at better prices in a number of markets, including our 19 largest ones. 20 So I -- I urge the Commission to think seriously 21 about whether price competition is currently being met, 22 whether prices currently are efficient, whether things truly 23 are fair. I -- I have serious doubts that practices that 24 allow dealers to undercut best bids and offers necessarily 25 lead us to where we want to go. 1 MS. NAZARETH: Ed, I bet you have a view. 2 MR. NICOLL: Well, I don't know if I disagree. I 3 mean, we're -- we're on such a theoretical level here, I'm 4 not sure which way some of those comments cut in terms of, 5 you know, our -- where we are today. 6 I certainly am in favor of the Commission's role, 7 first of all, to answer Bob's question. And I believe that 8 it's absolutely critical today, more than ever, to have a 9 strong SEC play a role here. And, you know, I think the 10 issues have been laid out and -- and people have already sort 11 of staked out, you know, their positions in their opening 12 statements. 13 I obviously -- and we obviously are concerned 14 about, you know, our ability to compete with markets that are 15 protected by regulations that we don't think serve our 16 customers. And, you know, we find ourselves in a situation 17 in which, you know, we have the demand for a very efficient 18 marketplace where we have investors who have chosen our 19 marketplace over competing marketplaces, and where we've 20 become the de facto primary marketplace in some instances. 21 And yet, because of what we believe are -- are, you know, 22 antiquated regulatory structures that -- you know, that favor 23 one market structure over another, do not allow us to -- to 24 compete as efficiently as we could. We think that -- that 25 that competition benefits all investors, not just the 1 investors who trade over our networks. We think that the -- 2 when even the most sophisticated investors come together and 3 trade at the most efficient price, that that -- that that's 4 critical that we promote that. Because, you know, the 5 average retail investor---excuse me---who wades into that 6 market at any -- any time will benefit greatly from the 7 liquidity and the efficiency that's being created by even the 8 most sophisticated investors. 9 So, you know, what we're interested here in -- in 10 exploring is finding a way to promote the competition between 11 these new electronic marketplaces, of which we're certainly 12 one of the preeminent players, while at the same time 13 preserving the role of the -- of the traditional marketplace 14 when they make more sense, when investors choose them. And 15 -- and, of course, finding a regulatory framework that's 16 flexible enough for both of these markets to exist at the 17 same time. 18 MS. NAZARETH: Thank you. 19 Bernie? 20 MR. MADOFF: I think that it's important to 21 everybody to keep their focus. You know, we -- we operated 22 our firm saying that no good deed goes unpunished. And I 23 think that's what you have to look at when you look at this 24 National Market System. Number one, I do believe we do have 25 a National Market System. I think it's -- I think it's 1 important to have one. I -- I think that what this industry 2 has accomplished is nothing short of amazing. If you -- if 3 you look at the cost of executions, if you look at the 4 efficiency of executions, if you look at the transparency of 5 the marketplace that exists, it is nothing to be ashamed of. 6 Yes, there are problems that always arise when 7 you're trying to do what we're trying to do. In my mind, the 8 big risk is to try and undo rather than just try and fix and 9 to tinker. I think the SEC, all the years that I've been 10 involved in this process, has always taken the position 11 "let's go forward, let's run the risk of making mistakes in 12 allowing innovation and allowing competition to go forward, 13 and we can always circle back and correct the problems that 14 occur." There is no way of accomplishing what we're trying 15 to accomplish without having some fallout, and that clearly 16 will be negative. 17 But I don't think the markets are in disarray. I 18 don't think they're -- they're a disaster by any stretch of 19 the imagination. I think what you're seeing is dislocation 20 of certain -- certain markets, of certain participants in the 21 marketplace. Quite frankly, that's -- that's what is 22 supposed to happen, and I think that's -- that's beneficial. 23 I think that it's a mistake for the SEC to try and overreact 24 to some of these problems. Fortunately, in my experience, 25 they never have. And they have shown a great deal of 1 restraint in not taking the advice of those people whose oxes 2 have been gored in this competitive process. I remember 3 everybody saying ban payment for order flow. They're still 4 saying ban payment for order flow. We don't even pay for 5 order flow any longer, although we're considering going back 6 to paying for order flow. It is -- all of these things are 7 constantly resurface (sic). And I think that, you know, I -- 8 first of all, I enjoy this process, and I think the markets 9 have enjoyed the process, and I think that it's important for 10 everybody to constantly take a deep breath, including the -- 11 the regulators, and just try and fix some of the dislocations 12 that are occurred. It won't be easy, but it is a heck of a 13 lot better than going back to where we started before you had 14 the '75 Securities Act. 15 MS. NAZARETH: Rick? 16 MR. KETCHUM: Perhaps because I've lived with them 17 for 28 years, I may be prejudiced, but I think the core 18 principles are exactly the right principles for today, just 19 as they were in 1975. Although I do have to remark that I 20 think only someone with a boundless optimism of Bernie can 21 speak of some of these things as enjoyable. 22 (Laughter.) 23 MR. KETCHUM: I think they're the right core 24 principles, not because they drive you to a single 25 conclusion. In fact, they're the right core principles 1 because they don't. Because they are, by necessity, 2 overlapping; often inherently contradictory. Back when I was 3 at the Commission we liked that because it gave us a lot of 4 discretion. Now I find it a little bit more terrifying. But 5 nevertheless... 6 (Laughter.) 7 MR. KETCHUM: ...nevertheless, good, because there 8 is a tendency, with the magic words of "National Market 9 System," to think of something that could be tied up in a 10 bow, put under a tree, and be finished. And in the end, the 11 National Market System is messy, just as the core principles 12 are. It reflects a marketplace that continually changes as a 13 result of technology, as a result of investor demand, as a 14 result of innovation by entities such as Ed represents. And 15 if -- if the principles are not broad enough for the 16 Commission to step back, reapply them, and recognize that any 17 decision they make has to inherently be a balance, then 18 they're very dangerous. These allow you to do that, and 19 always require you to recognize that you're making a balance 20 and you're accepting tradeoffs, and to accept them knowingly. 21 If there's any piece that -- where I had the chance 22 to rewrite back in 1975 I would do, it is to make more 23 explicit economic efficiency of transactions. Not because 24 it's wrong, but because I think there is a risk of the 25 Commission not focusing on that means as much efficiency, and 1 a great part of efficiency is liquidity. And then one must 2 balance the needs, as were indicated before, of individual 3 investors with institutional investors throughout. 4 And the last piece I'd say is the standards 5 themselves are great precisely because they must be balanced. 6 I think the one thing we all have to watch, and -- and 7 certainly the Commission, as well, and all self regulatory 8 organizations, is to get too attached to what worked 20 years 9 ago, and recognize that, while the principles remain correct, 10 the solutions of 20 years ago aren't always correct. 11 MS. NAZARETH: Thank you. 12 Larry, did you want to go next? 13 MR. LEIBOWITZ: Yeah. I mean, I think there's a 14 tendency, in listening to a lot of what's going on here, to 15 say, "Oh, my God, what's happening to our markets?" The 16 reality is we're in a bear market, and that's what's happened 17 to our markets. 18 Over the last few years---and I've been in this 19 business for 20 years, so, Bernie, I'm junior to you by quite 20 a bit---I mean, I've seen -- we have seen so many changes 21 over the last few years, almost all of which have been to the 22 investor good, whether it's order handling rules, 23 decimalization, increased linkages, the 1-5 rule disclosure, 24 all in the name of increased transparency, increased 25 fairness. And they've had a major effect if you look at 1 profitability of dealers and -- and spread sizes and all of 2 these things. It shows that our markets are much more 3 efficient than they were. That's all been accomplished 4 through improved regulation and improved competition of 5 alternative trading systems. And I think it's -- it's a 6 mistake to sort of think, 'Oh, well, we can come up with one 7 big solution that's going to fix everything that's wrong.' 8 The reality is our job is -- is to remove any unfairnesses 9 and inequities that we see between the markets, some of which 10 are a result of regulation, and some of which are a result of 11 different constituents having more market power than others. 12 And what we have to do is balance these things as we figure 13 out how to move forward. 14 MS. NAZARETH: Thank you. 15 Sal? 16 MR. SODANO: I'd like to say that we do agree with 17 the definition of the National Market System. I guess the 18 point that we'd like to make has to do with the aspect of 19 fair competition. When you look at the players in the 20 marketplace, and I'll use ourself as an example, when -- and 21 obviously we look to and foster innovation and -- and new 22 ways to do things and ways to better the investor, none 23 better than at the Amex. We do that. 24 But when you have issues such as timing that's 25 required to effect changes in rule makings because of one 1 standard versus another and one market compared to another 2 and the regulatory costs overall to operate a marketplace 3 versus a competitor, the redundancy requirements to put up as 4 a primary market versus another market in the marketplace 5 where those requirements are just not there, and quite 6 frankly can shut down if they'd like to. 7 The investor protection rules that are evident and 8 required on one side versus another. Quite frankly, from 9 someone like ourselves, that would pose the -- the situation 10 of why bother to -- how do we protect investment that we put 11 into our marketplace and try to do things better. There's 12 just an incredibly unequal playing field, and the -- the one 13 section, fair competition, just needs a lot of work. 14 One last comment with regard to payment for order 15 flow. I guess I make that more timely with regard to the 16 options industry in relationship to the Philadelphia Stock 17 Exchange announcing that they want an organized payment for 18 order flow program. I can tell you, without any hesitation, 19 that many market makers will stop becoming market -- or 20 continue to be market makers in the market as a result of 21 those additional costs which basically are passed along to 22 order flow providers and not the investor, which has the 23 direct effect of reducing liquidity in the marketplace. So 24 we have an additional cost levied on the market which is not 25 advantageous to anyone. So, I mean, it goes on into the 1 equities markets, as well. And I don't want to get into that 2 long discussion. I think that'll happen this afternoon. But 3 in particular, with the options industry, it is absolutely a 4 very bad thing that needs to be stopped. 5 MS. NAZARETH: Okay. 6 Gus? 7 MR. SAUTER: Thank you. I'd like to agree with 8 many of the other comments, that the principles of the 9 National Market System we think are well-founded. It would 10 be hard to argue against them. They sound like motherhood 11 and apple pie. 12 I would, though, say that while we do believe the 13 markets are much better today than they were ten years ago or 14 even five years ago, we think there have been many great 15 innovations, the order handling rules, the Reg ATS, that have 16 greatly improved the situation for investors. Nevertheless, 17 we think there are additional improvements that can be made. 18 We believe there are many situations where the 19 rules really are designed for the financial intermediary 20 community. That instead of benefitting the investor, they 21 benefit the -- the intermediary community. We think it 22 should be the other way around. Ultimately, everything 23 should be designed around the benefit to the investor. 24 So we think there's -- there's a ways to go. We 25 think there are some very important changes that could be 1 made. Market linkages. We think that it's problematic to 2 have internalization. But yes, we do believe in these 3 principles. We do believe the markets are better, but 4 they're not perfect by any means. 5 MS. NAZARETH: Okay. 6 Cathy? 7 MS. KINNEY: I think the good news and bad news is 8 that there is nobody but the Commission who can help us work 9 through a lot of the issues and principles that you've laid 10 out today. As Gus said, it would be hard to argue with these 11 principles. These are the right principles. But I would 12 agree completely with his -- his view that unless these 13 principles are centered around the investor, we can all find 14 ourselves in a very difficult position. 15 I think some of the issues that we've heard around 16 the table today and the issues that you're going to grapple 17 with in the very subject topics are where there have been 18 misalignments or misapplication of the principles in certain 19 rule making that has occurred. And I think this is an 20 opportunity for us to step back. Because, to be sure, Rick's 21 correct. Technology, innovation, many of the things that we 22 all know about the success of the markets today really have 23 come and -- and created a certain amount of change. But I 24 think if we go back to the basic principles, we can all agree 25 they're the right principles if we keep the investor in front 1 of all of our deliberations about how to apply these 2 principles, I think we'll stay on track. And I -- I can't 3 imagine that we can go too far afield if we keep those 4 principles and the investor in mind. 5 MS. NAZARETH: Thank you. 6 Yes, Gary? 7 MR. GASTINEAU: Well, I -- I agree with everyone 8 else that the principles are -- are terrific. The problem 9 is, as Gus says, they're like motherhood and apple pie. I 10 think that what we need to do is we need to step back a 11 little bit and think about the possibility that the best 12 solution for some of these issues is deregulation. 13 Now, I realize that in the current political 14 environment it's going to be pretty hard to move very far in 15 that direction, certainly as a major agenda. But at the same 16 time, as the Commission looks at the market out here, we have 17 the best markets in the world. Now, there are two reasons 18 for that. One of those is the Commission, and one of those 19 is that markets work, and markets work to make themselves 20 work and to operate efficiently, unless they are frustrated 21 at every turn. 22 The Commission needs to think about how much of 23 this is due to the results of prior Commissions, and how much 24 of this is due to the fact that the market would do pretty 25 well in a lot of respects without the kind of intense 1 scrutiny and the intense regulation that it has today. 2 MS. NAZARETH: Okay. 3 John? 4 MR. MARKESE: This is a rather naive statement, and 5 I'm -- I'm admitting to it from the beginning. But when I 6 look at these principles, I see competition merely as 7 trumping efficiency and transparency. I think we have 8 probably stood in the way of competition many times with 9 regulation. I think if we had truly competitive markets, 10 we'd have more efficient and more transparent markets. But 11 what I'm aghast at is the fairness issue. I think that's 12 been lost somewhere along the line. I see payment for order 13 flow, and I can't think of a good thing that that does for 14 investors. I know some will argue it's liquidity, it creates 15 pools of liquidity. I see rebates, I see regulatory 16 arbitrage. And frankly I don't see the Commission having in 17 the past addressed those issues sufficiently from the 18 individual investor viewpoint. And, again, I'm taking that 19 simple viewpoint. 20 But if we look at a truly competitive market, we're 21 going to get low cost providers, we're going to get efficient 22 execution for all participants. And as far as transparency 23 was, we'll sell transparency, but it will be there, and 24 competition will push that price very low. So I see these 25 principles as being reasonable, but it starts with 1 competition, and I think we've hindered it. I'll agree with 2 Gary. And the fairness issue, I think we've lost it along 3 the way. I've -- I -- from the viewpoint of the individual 4 investor, I'm really saddened by that. 5 MS. NAZARETH: On that sad note... 6 MR. MARKESE: That note. I'm sorry. 7 (Laughter.) 8 MS. NAZARETH: Sorry. I think this has been a 9 tremendous opportunity to sort of tee up the issues that 10 we're going to be discussing for the rest of the day. I 11 think we'll take a 15 minute break and come back and talk 12 much more in depth about market data. 13 (Brief recess.) 14 MR. HARRIS: Welcome to the market data section. 15 Thank you for joining us. Before delivering my remarks, I'll 16 now note that now batting for the New York Stock Exchange... 17 (Laughter.) 18 MR. HARRIS: ...Richard Bernard. 19 MR. BERNARD: If you like my answers, I am -- I'm 20 Rich; if I'm not, I -- 21 MR. HARRIS: Oh, Michael Ryan for Amex. Thank you. 22 U.S. investors today have access to a consolidated 23 real-time stream of market information, data on trades and 24 quotes for each of the thousands of stocks traded in our 25 markets. Price transparency is a cornerstone of the U.S. 1 National Market System. Many people believe that it 2 facilitates the best execution of customers' orders, promotes 3 investor protection, and mitigates the fragmentation of 4 buying and selling interest among different market centers. 5 The best quotation and last sale are the most basic 6 bits of market information. The best quotation is the 7 highest bid and the lowest offer price currently publicly 8 available for a security. It is the national best bid or 9 offer, or NBBO. The last sale identifies the price at which 10 the most recent trade occurred, the size of that trade, and 11 the market in which the trade took place. 12 The National Market System legislation adopted by 13 Congress in 1975 as part of the Securities Acts Amendments 14 governs price transparency in the United States. The 15 amendments gave the SEC the authority to adopt rules to 16 implement the essential mechanisms of an integrated market, 17 such as the collection and consolidation of market 18 information. At the same time, Congress recognized the 19 importance of relying on competitive forces to the greatest 20 extent possible to shape the National Market System. 21 Under the market information rules adopted by the 22 Commission, stock exchanges, and the NASD, under -- I'm 23 sorry. Under the market information rules adopted by the 24 Commission, stock exchanges and the NASD make available 25 quotes, trade prices, and volumes for all exchange-listed and 1 NASDAQ stocks within seconds of the quote or trade. This 2 information is consolidated across markets and distributed to 3 the public. Presently, the current NBBO quotation, 4 individual market center, OTC market maker, and ECN quotes 5 and reports of all trades are disseminated for each stock on 6 a real-time basis. 7 The SROs have established several National Market 8 System plans to jointly disseminate consolidated market 9 information. These plans govern all aspects of the 10 arrangements for collecting and distributing market 11 information. They require the individual SROs to transmit 12 market information to a central processor, which then 13 consolidates the information to a single stream for 14 dissemination to vendors and some large -- larger end users. 15 In turn, market data vendors disseminate the information to 16 the public. 17 The plans also govern the fees that users pay for 18 market data, the distribution of revenues derived from those 19 fees, various reporting obligations, hours of operation, and 20 regulatory halt procedures. The resulting net revenues, 21 which currently total about $400 million a year, are divided 22 among the exchanges, NASD, and NASDAQ. They represent a 23 significant percentage of their revenues. 24 The implementation of decimal pricing in penny 25 increments in 2001 increased the demand for deeper levels of 1 market data. For example, in January 2002, the New York 2 Stock Exchange began to provide a real-time, dynamically 3 updated view of the limit order books for all NYSE traded 4 issues. NASDAQ's SuperMontage, which began to operate in -- 5 on October 14th, 2002, combines NASDAQ's stand-alone 6 quotation and order routing systems with the means for 7 aggregating orders and quotes at multiple price levels to 8 create an integrated trading system. Certain ECNs, such as 9 Island and Archipelago, make their limit orders book -- limit 10 order books available on their websites at no charge. In 11 addition, several Internet portals offer free access to ECN 12 limit order books and other market data. 13 Nevertheless, the impact of decimalization has led 14 some market participants to believe that additional 15 regulation is required to standardize the collection of such 16 information. Several market centers now rebate or seek to 17 rebate a significant percentage of their market data revenues 18 to their members or users based on the business that they 19 generate. These rebates have affected order routing 20 decisions, and have created incentives for tape shredding and 21 wash trading. The very existence of the rebates has led some 22 to suggest that market data fees are too high. Moreover, 23 many traders resent paying for data that they helped create. 24 They believe that the national market data belong to the 25 public, rather than to the exchanges and NASDAQ. The 1 exchanges and NASDAQ, of course, disagree. 2 In light of the wide range of views on the issue, 3 the SEC formed an advisory committee in August of 2000 to 4 assist it in evaluating issues related to the public 5 availability of market information in the equity markets. In 6 its final report dated September 14th, 2001, the committee 7 affirmed the importance of price transparency and 8 consolidated information to the health and vitality of our 9 securities markets. The committee also recommended, with 10 varying degrees of consensus, a number of measures to 11 increase the flexibility of market centers and data vendors 12 to make data available to the public. The committee also 13 suggested that technological and competitive developments may 14 have lessened the need for requiring market centers to act 15 jointly to consolidate data under existing National Market 16 System plans. 17 Our first session today will cover the way that 18 market data---in particular real-time market data---are 19 collected, processed, and made available to investment firms 20 and individual investors. The questions that we hope to 21 address are: What are the fundamental pieces of information 22 that professionals and investors use to determine whether and 23 when to trade a security, and at what price? What 24 information should be available to professionals and 25 investors, and on what terms? To what extent is regulation, 1 rather than market forces, required to ensure availability of 2 this information on a consistent basis? Who should benefit 3 from controlling market data? Who should determine how 4 information is made available to the public? How should 5 revenues from the sale of market data be allocated? Should 6 market data rebates be prohibited, or do they represent the 7 formation of a competitive market that should be promoted? 8 Finally, do regulators need to intervene to ensure that 9 market information is fairly priced? 10 Our purpose now is to examine this ambitious 11 agenda. Unfortunately, we have less than an hour-and-a-half 12 available to us to hear viewpoints from 13 highly informed 13 panelists. Accordingly, I'll be ruthless in allocating time. 14 I'll ask panelists to offer comments only when they can 15 introduce -- when they can introduce new ideas and 16 perspectives. We will measure success today not by our 17 ability to resolve these questions, but by whether we have 18 identified every important issue. 19 So let's start by asking what information is 20 necessary for traders to complete their trades? Wayne, you 21 had mentioned that you represent institutional investors. In 22 a post-decimal environment, is the best bid or offer that's 23 presently disseminated enough? 24 MR. WAGNER: The problem is the size associated 25 with that price, rather than the price itself, here. I mean, 1 the prices tend to be very quickly changing, which means that 2 there is a little bit of a problem of between the time you -- 3 you open your eyes and you -- and your first blink, the price 4 may have changed several times. But that's a fairly minor 5 problem. 6 The real problem is that it -- to my mind, that it 7 is -- there's -- what an institution will use the best bid 8 and offer for is -- is a sort of a voluntary means of 9 advertising interest in the -- I want to trade. It's a way 10 of placing an advertisement that says, "Come to me. I'm here 11 willing to trade on -- on this." 12 It's gotten very difficult for an institution to 13 make that kind of a statement, given the -- the ability to 14 move codes quickly at very low price ahead of that here. So 15 the quote I think means less than it did before, and the -- 16 and the question remains: Where is the liquidity? How do I 17 get to it? How do I get the -- find the other side, and how 18 do I message in order to be able to do that? 19 MR. HARRIS: Gus, are your traders getting all the 20 -- all the information that you believe that they need at 21 Vanguard? 22 MR. SAUTER: No. We believe that we need to see 23 the total book. And there are movements in certain venues to 24 provide that. We think it is important, not to see the BBO, 25 we need to see all the way up and down the book, both price 1 and size. We think we do need to see historical executions, 2 which obviously we do have access to at this point in time. 3 And we -- we also use historical best bid and offer, as well. 4 So the biggest piece of information we need at this 5 point is full transparency of the book. And that would be 6 consolidated across all markets. 7 MR. HARRIS: Are you willing to display orders into 8 a fully transparent book? 9 MR. SAUTER: With -- with appropriate protection on 10 those orders, absolutely. Limit orders, we would -- we 11 historically have been large users of limit orders. Under 12 the current conditions, in the penny environment, we have 13 pulled back a little bit from that. We do -- we do observe 14 limit orders being much smaller than they were just five 15 years ago, or even two years ago. And -- and part of that is 16 us pulling back a little bit away from limit orders. It's 17 now more advantageous, if you're doing floor trading, to use 18 a floor broker and -- and to hide. But we would be 19 absolutely willing to use more limit orders again if those 20 limit orders had protection. 21 MR. HARRIS: Let me turn now to retail investors. 22 And -- and, John, I'm going to ask you. In an environment 23 where the -- what's published as the national best bid or 24 offer, but where people often trade at better prices, are 25 retail traders getting all the information that they need to 1 make decisions? 2 MR. MARKESE: No, our average member holds 12 3 stocks and trades six times a year. I know there are day 4 traders out there. Declining species, probably. And there's 5 some people who thought they were day traders. But for the 6 average individual investor who's trading not particularly 7 often, the end national best bid or offer is a reasonable 8 benchmark, with the assumption that they can get, you know, 9 potential for price improvement. I would say the last 10 transaction is also important; obviously the last trade. 11 Are there -- are there investors willing to pay up 12 for better data or more data, more elaborate data? Of 13 course. But I think that is more than sufficient for their 14 trading, given the average investor. 15 MR. HARRIS: Dave, running an electronic trading 16 operation, you depend critically on market data. Are you 17 getting everything that you need? 18 MR. WHITCOMB: It depends on the market. In the -- 19 in the current NASDAQ market system, including the ECNs, we 20 get everything -- almost everything we need. We need more, 21 of course. But I think the important point is that the 22 structure of the market really determines a lot of these 23 issues. It determines whether the NBBO is meaningful, which 24 it is in a market where you have electronic trading 25 dominating, as -- as in the case in the NASDAQ market these 1 days. It's not in the listed market because of so-called 2 price improvement, which makes it unmeaningful. 3 Whether things like the deep book will be freely 4 given or given at marginal cost to investors again depends on 5 market structure. The ECNs typically give their entire deep 6 book free, both via websites and also by direct -- by a 7 direct transmission, again, because they don't have any 8 profit incentive for suppressing that information. In the 9 listed market, there's a strong profit incentive for 10 suppressing the information. The deep book that we now have 11 from the NYSE is only updated every ten seconds, which sounds 12 fast. But, in fact, for an electronic trader, it's a 13 lifetime. If that were really a real-time deep book, it 14 would convey more information. But that information would 15 hurt the interests of the specialists who have a pretty 16 important role in making the rules in that market. 17 So there's a tremendous linkage between whether -- 18 whether the markets are run on a pure profit-making basis or 19 whether they're a traditional SRO market, and the provision 20 of information. 21 MR. HARRIS: Now, Dave, would you have the New York 22 Stock Exchange provide that information? And if so, at what 23 price? 24 MR. WHITCOMB: If they were a private competitor in 25 the listed -- in the market for exchange services, I'll bet 1 they'd provide it free. They do provide the deep book that 2 they provide very, very cheap, but they don't update it as 3 often as I would like because it's not in the self-interest 4 of the specialists to let that information be out. 5 And, by the way, if people would look at that deep 6 book, they would see quotes on that book that are much better 7 than the transactions that are actually occurring. When you 8 send orders via Direct+, they execute at the NYSE quote, even 9 though there are limit orders on the deep book at the same 10 moment that are at better prices. That's again a problem of 11 market design. When you have a -- a free enterprise 12 electronic trading system, things like that don't happen 13 because they're not in -- in the profit interest of the 14 people who run the system. 15 MR. HARRIS: Rich, would you like to take an 16 opportunity to respond? 17 (Laughter.) 18 MR. BERNARD: Delighted. A couple of things. 19 First, on -- on the NYSE open book, which is our display 20 book, that ten seconds has nothing to do with specialists, 21 and everything to do with just a decision that was made about 22 the amount of data and the amount of data changes that we 23 were spewing out. And we would go to zero seconds if that 24 were proved to be the demand. You have to remember that that 25 is not an executable book. You want to execute against the 1 New York Stock Exchange, your general choice is to come 2 through Super DOT, NYSE Direct is a automated execution of 3 the best quote at the time it gets there. It's not a full 4 depth of market opportunity. And then just to speak to where 5 we're going, I -- we maybe filed, I can't remember, but we 6 are about to file, if we haven't, for the depth quote or the 7 dual quote or the liquidity quote, as -- as we've called it 8 in different incarnations. Remember that the New York Stock 9 Exchange is not just the book, it's also the trading 10 interests in the crowd, and it's the interest of the 11 specialists. Those interests will be rolled up into a single 12 depth quote at a -- a level that the specialist judges to be 13 a good place to reestablish what the institutional market is 14 looking for and -- and executable in -- in the regular way. 15 And for the point of information, because it has -- has 16 implications for capacity, just think a second that we were 17 quoting the inside quote, not five years ago, was at an 18 eighth. We now have inside quotes at pennies. That, as -- 19 as the institutions will tell you, that retail quote is not a 20 particularly useful piece of information if you're trying to 21 trade in size. 22 What we're going to be doing is going back and, in 23 effect, quoting at the eighth or whatever the right number is 24 for the depth quote, and then we will auto quoting (sic) the 25 real quote -- the realtime, the retail quote off of that 1 depth quote. But that does have profound capacity 2 implications, because it means that our retail quote will 3 change more often, it means that the markets that are machine 4 generating their retail quotes off of our retail quotes will 5 change more often, and God help the auctions exchanges who 6 are doing all the series that they do off that quote. So we 7 do have some capacity issues coming down the pike. 8 MR. HARRIS: Joel? 9 MR. HASBROUCK: Thank you, Larry. 10 I'd like to make two points. First, there's often 11 a presumption that the principle of transparency, more is 12 better, always renders visible that which was formerly 13 invisible. In fact, it changes the incentives. If you know 14 that one of your -- part of your trading interest is going to 15 be visible, it changes your incentives for putting it out 16 there in the first place. This ties into a point that Gus 17 Sauter made; namely, that the limit order books we have now 18 are much thinner. There seems to be less liquidity on them 19 than, well, of course, people would like, but also less 20 liquidity than many people suspect is really out there. 21 Now, the question is: Is this a consequence of the 22 visibility of these books, or a consequence of the inadequate 23 protection for the orders that reside on it? I know one 24 industry commentator has gone so far as to call it, "the 25 death of the limit order." Now, he's only exaggerating 1 slightly. What he means is that the old sense of the limit 2 order as being a provider of liquidity, much like a dealer, 3 has been replaced by the concept of the limit order as a 4 quick in-and-out sort of thing, something that can be 5 canceled within one second or two, and becomes almost more a 6 medium for negotiation, or an advertisement, than a true 7 provision of liquidity. 8 MR. HARRIS: Let's turn a little bit more towards 9 some regulatory issues, but not too far from where we've 10 been. The -- in the current regulatory environment, we 11 require that all markets deliver a certain amount of 12 information that goes into the formation of the best bidder 13 offer. The New York Stock Exchange and NASDAQ and others are 14 providing more information on a presumably value-added basis. 15 And so I'd like to ask: Where should the line be between 16 what is -- must be delivered to everybody, and what the 17 exchanges or the ECNs or dealers can retain and sell on a 18 proprietary basis, maybe. 19 Rick? 20 MR. KETCHUM: I think it's a very good question, 21 Larry. You know, there -- the great thing about truths are 22 that -- that you can say enough of them and -- and forget 23 that they -- they are relative vis-a-vis various different 24 points. The consolidated best bid and offer is not nearly as 25 valuable for active traders as it used to be. Just as John 1 indicated, it remains extremely valid -- valuable with 2 respect to monitoring the quality of the execution you 3 receive, and generally understanding and being able to 4 develop strategies to where the market is and where it's 5 moving. 6 And it is -- it has been the focus of SEC oversight 7 in this area, and a requirement from the standpoint of 8 consolidation of that information, it should be -- it should 9 continue to -- to be for a couple of reasons. One is that 10 it's still valuable and important for those reasons. 11 Secondly, unlike the montage of information that we proved, 12 either of quotes or now orders through SuperMontage or the 13 New York Stock Exchange provides, where we have enormous 14 incentives to be able to reach out and get as many people to 15 follow it as possible, both for them to feel good about our 16 market, to trade in our market, and also, frankly, to compete 17 from other people who collect orders. The consolidated best 18 bid and offer, because it's a reference price, essentially 19 everybody has to have, and it makes sense from that 20 standpoint to have a substantial amount of regulatory 21 oversight. 22 When you move away from that, I -- I'd say the last 23 couple of years are pretty clear indications, from how the 24 ECNs have made the information available to how NASDAQ makes 25 it available, we kind of -- we think it's a good idea to have 1 a depth quote, but we kind of think the best way to figure 2 that out is to have all the orders there and let you execute 3 against them. And -- and I think that there will be 4 continuing competition between markets, ECNs, and others who 5 collect orders, to make that information available as much as 6 possible. And the price will be impacted from that 7 standpoint. If I turn out to be wrong, the Commission will 8 always have authority to deal with market information with 9 respect to price, and be able to evaluate that. Although I'd 10 strongly urge that it start looking at it and recognizing 11 that we compete across categories, exchanges, ECNs, and 12 others, and that, therefore, that regulation should be very, 13 very light outside of the consolidated best bid and offer. 14 MR. RYAN: Just to add to that, I think Rick's 15 right about the Commission stepping in too -- too far and 16 hard into the area of market data in terms of getting away 17 from the national best bid and offer. I think that a 18 consolidated NBBO is going to continue to be important. I 19 think the Commission should look at the -- the depth of book 20 issue and providing that, the way that that's provided, and 21 look at certain -- at certain areas; for example, any anti- 22 competitive practices along those lines. 23 But beyond that, I think the Commission should 24 really focus on the NBBO and how that's handled, and how 25 market data revenue is distributed, and the -- and the impact 1 that that has on all investors in getting information out in 2 a -- in a timely manner, making sure that the information and 3 the systems that support that information, whether it's a -- 4 whether they're regulatory systems or technology systems or 5 competitive practices where markets try to differentiate 6 themselves from one another, and how that revenue is -- is 7 used in supporting that -- that infrastructure. But let 8 competitive forces deal with depth of book type of issues. 9 MR. HARRIS: Gus, please. 10 MR. SAUTER: Yeah, Larry, you mentioned whether or 11 not information should be restricted to only certain market 12 participants, and -- and not made available to all. I -- I 13 believe that would be inappropriate. I believe that, again 14 with -- using the principle that ultimately the market should 15 be designed in the interest of investors, then all 16 information should be available to investors. Making 17 information available only to certain financial 18 intermediaries necessarily gives them an advantage over the 19 investor. It may make them participate in certain situations 20 where they would not, but I do believe it's -- it's 21 definitely to the disadvantage of investors by giving certain 22 intermediaries information not available to all. 23 MR. HARRIS: Gus, would you like the SEC to require 24 all markets to deliver up enough limit order information that 25 they could create an analog to the NBBO, a consolidated book? 1 2 MR. SAUTER: We would love to see a consolidated 3 book. If we have a large execution, we would love to know 4 how far down that book we need to go to execute it all 5 immediately. We may then decide to sweep the, you know, 6 mythical consolidated book, or we may decide to execute in 7 smaller pieces, depending on how far down the book we would 8 need to go. But being able to determine the true liquidity 9 actually out there would great enhance our ability to trade. 10 MR. HARRIS: Now, can you get that without our 11 intervention? 12 MR. SAUTER: Data vendors do compile a lot of this 13 information and consolidate it for us. So we -- as -- as 14 various venues have opened up their books and made them 15 available for full viewing, we are getting more and more 16 transparency. We need to continue to do so, and make sure 17 that, one way or another, investors or institutions---the 18 "buy" side, if you will---has access to all of the data 19 available, whether it's aggregated by their vendors or 20 provided some other means. But it's important that the 21 complete book of every venue is -- is distributed. 22 MR. HARRIS: Larry? 23 MR. LEIBOWITZ: Yeah. I would say from -- from our 24 standpoint, and -- and with deference to John, while the NBBO 25 is -- is an interesting indication, I can tell you from the 1 number of call center calls that we get every day from 2 confused investors about flickering quotes that they can't 3 see, and about not being able to see what's really going on 4 in the market, that it does have a major impact on -- on 5 confidence in what's going on in the market. 6 I also think that we've got to get out of this 7 notion that this data belongs to the exchanges or belongs to 8 a market center. As you said in your opening remarks, in the 9 old days we used to give our data to the exchanges and then 10 have to buy it right back. Well, now, in some markets, we 11 actually have to pay to provide our data to the -- to the 12 market, and then have to buy it back. But the reality is, 13 while the vendors can compete, if the initial providers of 14 the data are allowed to have a monopoly pricing or monopoly 15 control of what they do with that data, it makes -- makes for 16 an inefficient market and an unfair market. 17 MR. HARRIS: Well, many people have spoken in favor 18 of competition. Why won't competition work here? Couple of 19 folks. 20 Ed? Want to answer the former question first? 21 MR. NICOLL: Right. Yeah. The first question -- 22 the first answer is you have an inherent conflict in the 23 Vendor Display Rule in which you require this consolidated 24 information, and as soon as you require it, that confers 25 monopoly power on the people who distribute it. If -- if one 1 particular market center, who may or may not -- who may or 2 may -- where demand in a free market otherwise might not 3 exist, is required -- you are required to purchase that 4 information from them, you're conferring certain powers on 5 them to sell that information. 6 And then you're essentially creating, I believe, as 7 an obligation on your part, to regulate, you know, how much 8 that is charged. Obviously, if you didn't regulate that, 9 there wasn't some sense of what a reasonable charge was, then 10 you would be conferring almost unlimited economic power on 11 that particular entity. So the Vendor Display Rule, as of 12 itself, is a -- is a real tension in that. 13 I want to, if I just could, just sort of tease out 14 some of the -- the conflicts that sort of Gus intentionally 15 or unintentionally indicated in -- in his viewpoint. I mean, 16 what we all should understand is that what everybody wants 17 is, everybody wants everybody else to display their 18 intentions to trade, without themselves being required to 19 display their own intentions to trade. 20 So you've got -- a particular example here, where 21 Gus is talking about, well, what I want is I want everybody's 22 -- I want a display of everybody's intentions to trade, and I 23 will sweep the book. Okay? Well, why isn't he out there 24 displaying the fact that he wants to purchase or sell 100,000 25 or 500,000 shares of a particular entity? Why does he get to 1 decide, and step back and hide his intention to trade, while 2 he views everybody else's displayed intentions to trade, and 3 choose to effect the option, the free option, in effect, that 4 he's being given by hiding his intentions. 5 Now, that intention to trade, we cannot regulate 6 that. We can only regulate that down so far. We can't go to 7 every individual and every institution in the United States 8 today and say, "Anytime you're thinking about trading, you're 9 required to raise your hand and say, 'I'm -- I have an 10 intention to trade here.'" And there is this inherent 11 tension between when we require people to display their 12 intentions to trade and when we don't. And in designing the 13 -- the market structure system, we have to be very cognizant 14 of the fact that if we overregulate and put a -- you know, 15 and take a -- what I would describe as a naive view of -- of 16 limit orders, then we may be setting up a situation in which 17 you have less transparency and less liquidity than you 18 otherwise would have. 19 MR. HARRIS: Gary had a hand up as well. 20 MR. GASTINEAU: Couple of comments. One relates to 21 something that -- that Joel had said in terms of the 22 protection of the orders. Gus's comment, as well. If you 23 provide for limit orders that are hidden, perhaps not on the 24 primary market, but perhaps on a competitive market, that's 25 an interesting way in which markets could compete. You have 1 a limit order book, but you also know that in some markets, 2 behind the posted market, there might be another 1000 or 3 million shares waiting to be done at that price. And if Gus 4 decided to hit the book at a certain thing, he may be 5 surprised to find that he's filled at the national best bid 6 and offer on a million share trade. I wouldn't want to bet 7 on that, but, you know, things like that could happen. 8 But the point is you need to have the ability to 9 innovate, and that's one of the places where the transparency 10 argument falls down, because it's clearly in the best 11 interest of all investors to permit some kind of hidden 12 orders, particularly for a large investor who at this point 13 is fairly significantly disadvantaged in terms of the way the 14 market structure works. The NBBO is great for John's people, 15 but doesn't work at all for Gus. 16 MR. HARRIS: My sense is that at issue here is not 17 whether we will compel people to display orders, but rather 18 whether we will aggregate orders that are -- or aggregate 19 displays across markets. So, while the comments are well 20 taken, I don't think that's the market data issue that 21 presently presents us. 22 Rich, I -- did you want to speak to the question of 23 independence of pricing? I thought I saw a glimmer in your 24 eye. 25 MR. BERNARD: Well, I've got many items. I just 1 wanted to speak on this issue of displaying away from the -- 2 the inside quote. I completely agree with Rick, and -- and 3 it's where the Seligman committee came out, which is that the 4 market is doing a pretty good job of having people come 5 forward with depth of market information. We have to 6 remember that the institutional market is not spread all over 7 the score, so markets that are -- are competing here, by and 8 large, the -- that interest is represented in New York, 9 NASDAQ, Instinet, and a few other places. And those markets, 10 in fact, are and in some cases have been for many years 11 displaying their -- their depth of market information. So I 12 don't -- I don't see a case for Commission regulation beyond 13 the -- the sort of light once-over that Rick was talking 14 about. 15 If I could come a little bit to pricing and -- and 16 a comment that -- that Larry said, you know, we -- when we 17 got to the Seligman committee we had a -- it was a 15 to 8 18 vote for getting rid of consortia. And if you had gotten -- 19 and there's a lot of reasons for that. 20 Secondly, which we really haven't gotten to yet, 21 but we need to. The current system of giving a franchise to 22 any exchange in the country to jump into the consolidated 23 data stream, that is what's causing the -- the subsidies, 24 that's what's causing the rebates, and that's what's funding 25 much of the payment for order flow, now that the spread 1 doesn't support it anymore. And I hope we're going to get to 2 those issues. 3 MR. HARRIS: Well, we most certainly will. Let's 4 -- why don't we turn to them now. There are about $400 5 million that are being collected through the consortia right 6 now. If anybody had any doubt as to the importance of the 7 national best bid or offer, despite the criticisms displayed 8 here, they certainly are worth at least $400 million to 9 participants. Many folks, though, believe that that data is 10 really theirs, and they wonder why they have to pay for it. 11 And I was wondering whether anyone wanted to speak to -- to 12 that point of view. 13 Larry? 14 MR. LEIBOWITZ: Yeah. I mean, I think it's clear, 15 when you -- when you look at the arbitrage that's going on, 16 particularly prior to the abrogation of -- of tape revenue 17 sharing, that there's an awful lot of money there to be 18 spread around. And what we wonder is why that isn't going 19 back to the investors. 20 Now, at least with tape revenue sharing, that was 21 an indirect way to get money back to people who were trading. 22 So, in some senses, maybe there's some fairness there. But 23 the reality is that clearly there's a lot of money there, and 24 -- and how do we ensure competition to either make that 25 distributed to the proper people, or make that pot smaller? 1 Because I think what we've done is, we've intermingled the 2 different reasons and uses of that money, whether it be 3 incentives to people to trade on their exchanges, or paying 4 for operating costs for exchanges, it's not doing what it was 5 intended to do. 6 MR. HARRIS: Michael? 7 MR. RYAN: Yeah. I think this is probably the most 8 important issue when it comes to market data, in terms of 9 what we -- what we do with that -- the revenue that we 10 receive. And for a long time, the system, I think, has 11 probably worked fairly well, although increasingly that's 12 been less true in terms of the way that revenue is 13 distributed among market participants in simply -- in the 14 listed side on -- based on trades. 15 And we just submitted to the CTA participants on 16 Friday, and began distributing to the Commission yesterday, a 17 proposal to amend the CTA plan that would really shift in a 18 significant way and say that market data revenue really 19 should be distributed among the exchanges that are providing 20 quality markets, so that there would be -- the proposal we're 21 putting forth would say that revenue should be distributed in 22 part on trades, in part on share volume, and also in part on 23 the national best bid and offer. And we've put some quality 24 parameters around the way that would work so that just 25 putting up a quote doesn't -- doesn't get you credit. It 1 would have to be a quote that would be up for some time, and 2 the greater the size, we give you even more credit. And we 3 believe that that's the -- that those types of incentives 4 should be put in the marketplace. 5 A little bit off the topic, but we -- to get to the 6 ownership issue, we think that the Commission should look 7 real hard at the current environment for exchange regulation, 8 and abolish Regulation ATS, relook at the -- and modernize 9 the way exchanges are -- are regulated, so that it is easier 10 for other markets to become exchanges, but once you cross 11 over from a broker-dealer into an exchange world, your -- 12 there's similar fair regulation, even regulation, that the 13 regulatory environment is focused on what is important, not 14 -- not the solutions. I think Rick mentioned earlier today 15 that solutions 20 years ago may have been good and worked 16 then, but they don't necessarily matter anymore. And the 17 Commission should take a real hard look at what -- what 18 they're focusing on, and that the exchanges should -- the 19 market data is really the critical output of the operation of 20 a marketplace, operation of an exchange. And that I think 21 that the exchanges and markets should be compensated for 22 that, and they should be compensated on the quality of the 23 job that they do. 24 COMMISSIONER PITT: Let me ask a very sort of basic 25 question, which is: Why should there be any revenues at all? 1 If IBM, AT&T, and others have to report on what the results 2 of their operations are, they don't charge anyone for that 3 information. Why is there a fee associated with market data 4 for transactions that are executed in public securities in a 5 public market? 6 MR. RYAN: Oh, I think, quite simply, that is -- I 7 mean, IBM is compensated for the output of what they do as an 8 organization, and that -- and any organization should be 9 compensated for their output. And that is one of the outputs 10 of operating a marketplace. And we -- and we, the markets, 11 should be compensated fairly and reasonably for that. And we 12 think that the process, by and large, has worked very well. 13 We think that the process of having the rate set by the 14 exchanges initially, which are required to have public 15 members upstairs and floor members on their boards in 16 analyzing that, it goes out -- then out for public comment 17 before the rates are approved, so there's plenty of 18 opportunity for the -- for the public in a variety of arenas 19 to address. And then the Commission has to approve it, 20 finding that it's consistent with the Act. And that the -- 21 the -- going back to what I said earlier, the distribution of 22 the revenues should be based on market quality. 23 The other two components that I would add to that 24 is that -- that no -- no market center should get credit for 25 trades or quotes coming from a market participant when that 1 market participant isn't fully participating in the National 2 Market System. That we really should focus on what it means 3 to be a National Market System, and -- and create incentives 4 to -- to add value to that process. 5 And the other piece, going back to what a number of 6 people have said, is that ending markets paying for order 7 flow, whether they're using market data revenues or 8 otherwise, and focus on providing quality markets and 9 efficient markets. 10 COMMISSIONER PITT: Yeah. With -- with due 11 respect, I don't know whether that answers the fundamental 12 issue. 13 COMMISSIONER GLASSMAN: Can I -- can I just follow 14 up on that question? 15 MR. HARRIS: Sure. 16 COMMISSIONER GLASSMAN: What would happen if there 17 were no payment? How would the market evolve? What would 18 change? 19 MR. RYAN: Well, I think if you -- if you go and 20 put into place the process that we're proposing in terms of 21 market quality, market centers would evolve, exchanges would 22 evolve that are providing value added services, better 23 quotes, you know, better transaction reports, because that's 24 what they're -- that's the way they get credit for providing 25 market data. And there would be incentives to develop 1 innovative trading systems, quotation methodologies, and 2 products that would allow them to earn greater market data 3 revenue. 4 MR. HARRIS: It would appear that we've touched 5 some interesting controversies. Let's go -- Bernie, Rich, 6 and Rick, and then David, and we'll see how we -- let's see 7 if we run out of new ideas. 8 MR. MADOFF: I think the Chairman is -- is breaking 9 the rule. You've asked a simple question and you want a 10 simple answer. 11 CHAIRMAN PITT: No, I'll take any answer. I just 12 want it to be direct. 13 (Laughter.) 14 MR. MADOFF: I'll try to give you a direct answer. 15 That's the way it was. I think that there was this -- going 16 back as long as I can remember, there was this revenue pool 17 that was generated from the information. The exchanges 18 needed a funding process, whether it be for regulation or to 19 -- to run their businesses, and this was a convenient pool 20 that they were the only ones at that time that were able to 21 -- to put it -- put it together, and then said this was a 22 valuable -- a valuable commodity. 23 I think the problem that we all faced in -- on the 24 Seligman committee that I had the misfortune of participating 25 in... 1 (Laughter.) 2 MR. MADOFF: ...was that it really boils down to a 3 -- an allocation process. And there were people who felt 4 that there was no way of reconciling a checkbook, so to 5 speak. That there were these -- these funds were being used 6 for -- for all valid purposes, but nobody could really get 7 their arms around saying, "Give me a reconciliation. Tell me 8 how much money is spent for regulation, how much is being 9 spent for other issues." 10 So I -- my -- I would answer your question, is that 11 theoretically you could stop charging or -- or taking this 12 market data and giving it to the exchanges, and come up with 13 some other way of funding the -- their various operations. 14 It's just something that no one has even -- in my -- in my 15 memory, has attempted to do it. But I don't think there is a 16 -- I think the question you're searching for is -- is not out 17 there, because it's just never been attempted in the past. 18 MR. BERNARD: Let me build on Bernie's comments and 19 -- and also Michael's. Remember Merrill Lynch or Schwab 20 could sell their orders for their data content, and -- and I 21 would have no objection to doing that, but they're not very 22 interesting until they interact with other orders, other 23 interests, either to generate the quote and know where that 24 order stands in the queue, to generate the book, or to 25 generate an execution. That's all work done by the stock 1 exchanges. 2 So that the first answer to your question, and I'm 3 not -- you know, this isn't a legal answer, but I can give -- 4 I can cite you to the Supreme Court cases. But the point is 5 that there is work done when you produce executions and -- 6 and market data jointly, which is what exchanges do. 7 The second part of your answer which Bernie touched 8 on, and I'll use the New York Stock Exchange as -- as the 9 example. Round numbers, it costs $800 million a year to run 10 the New York Stock Exchange. Most of that goes into markets 11 operations, data, that sort of thing. But we also have to 12 fund [the regulatory department}. And we do that by asking 13 the people who are the beneficiaries or the constituents of 14 that process to pay up. It's the broker-dealers, it's the 15 listed companies, and to a lesser extent, it's the 16 institutional investors. And we -- we go at that in about 17 five major parts: listing fees, transaction fees, regulatory 18 fees, market data fees, and the facility fees. And I've 19 pretty much given those in the order of their size. 20 Market data, and we -- we were able to look back to 21 about 1934, has never contributed more than 20% of the NYSE 22 revenue, and it's been about 17 or 18% since -- since 1975. 23 That seems -- and the people who are paying that particular 24 pool, listed companies don't pay much. The broker-dealers 25 pay, round numbers, about 45%. And the institutional 1 investors pay, round numbers, about 55%. If you ran that 2 through the total revenue stream, what it would tell you is 3 that the institutional investors end up paying about 10% of 4 the overall costs of running the New York Stock Exchange, 5 which I submit's a good deal. 6 So the way to answer your question, therefore, is 7 that you do have to fund the stock exchanges. These -- as 8 Bernie says, these are the traditional places we've gone 9 after our constituents in getting them to pony up, and it's 10 based on value service propositions. I don't think anyone in 11 here thinks that our whatever it is, $160 million out of 12 market data revenue is not a bargain. I mean, everybody 13 needs that data. So there's no issue, when the New York 14 Stock Exchange sells its data, that you've got a bargain. 15 And -- and, Larry, you can disagree with that, but I think 16 most people would say that's a -- it is a bargain. 17 And then the question is: What's stopping us? Why 18 don't we charge -- why don't we load 50% of our revenue on 19 market data? I submit that the value of that data would 20 support it, if it were a pure value proposition. And the 21 answer is that we have a -- the same constituents who are 22 sending us the orders, sending us the listings, and -- and 23 paying -- and paying our fees are the ones sitting on our 24 board. And that's why you've seen that 17, 18% rock along 25 all these years. 1 So we think that system works. Evidently about 20 2 out of 23 members of the Seligman committee thought that was 3 a pretty good system. And the real problem, as Bernie said, 4 was what happens when you take the NYSE cost structure and 5 generate a revenue stream, and then give it to somebody who 6 doesn't have that structure. 7 MR. HARRIS: Rick, you wanted to speak, too? 8 MR. KETCHUM: And I will try to remember your 9 stricture and not repeat everything, although I certainly do 10 agree with what Rich said. But if you -- I'll try from a 11 balanced standpoint to acknowledge there are multiple values 12 here. I think Rich beautifully said that why markets provide 13 value added, and indeed, why they have consistently been the 14 only ones able to sell their data. Obviously part of it's 15 regulatory, but I think it's a great deal more than that, as 16 demonstrated by loads of people taking it that don't have a 17 regulatory responsibility directly with respect to it. 18 At the same time, I'll freely acknowledge that Ed, 19 and Larry, and lots of other firms around here, and Bernie, 20 provide value with respect to their -- the orders that they 21 submit that we aggregate. And, indeed, particularly in a 22 world of decimalization and wide varieties of competitors 23 encouraging people to provide liquidity and to aggregate 24 orders or to, God forbid, actually put some of their own 25 money online is a good thing, also. 1 I think that, to me, leaves the Commission with the 2 fundamental responsibility here. I think it would be -- 3 scare me very much if the Commission decided that something 4 that had been demonstrated to have market value should 5 suddenly get reallocated outside of that. I think there's -- 6 there's danger to that, that I -- I think, generally 7 speaking, regulatory agencies haven't gotten a great track 8 record on. 9 But you do have issues you have to decide. And 10 with respect to consolidated best bid and offer information, 11 you've got to decide what the right price is, if, indeed, 12 you're going to look at it for price with respect to what's 13 delivered from it. And that's not fun, but it's something 14 the Commission's been involved with and I think has generally 15 done a good job with. 16 With respect to questions ranging from liquidity 17 rebates to tape sharing, I think the Commission was 18 absolutely right to step into tape sharing and stop it at the 19 moment, while it stepped back and looked at some of the 20 abuses that occurred. But the end product has to acknowledge 21 one way or another that other people were producing value. 22 Whether that be liquidity providers, whether that go to 23 investors with respect to lower prices, I don't pretend that 24 there's one answer. My preference probably is carefully 25 regulated liquidity providers. But -- but, nevertheless, 1 it's something that it's fair for you to consider. 2 The last piece that goes to some of Rich's points, 3 but -- but I think builds in a different way, that it's 4 absolutely the Commission's responsibility, or I think raises 5 a terrific risk of running things off the cliff here, is that 6 there are things that, whether it be market information or 7 whatever that fund, that are critical to investor protection, 8 and that is the basic self regulatory structure that operates 9 in the marketplace today. And the present environment, with 10 respect to tape information, does -- and I sit on both sides 11 of it with respect to NASDAQ and NASDAQ securities and 12 InterMarket, does create risks of regulatory arbitrage that 13 are greater and greater as time goes on. 14 And it is absolutely critical that if the 15 Commission feels that there are minimum standards for primary 16 markets to provide, and feels that regulation must look at 17 multiple levels as to an order is -- is generated from the 18 marketplace, run through an intermediary, handled by a 19 discovery mechanism or whatever else, then it must find a way 20 to ensure that people share in those responsibilities. Or 21 you can take Gary's choice and decide to all to do yourself 22 (sic). But one way or another, the -- a means that does not 23 ensure that those responsibilities are fairly shared is -- is 24 a token to guarantee that all of us will try to find ways to 25 no longer do it in the same way we did it before. 1 MR. HARRIS: Larry? 2 MR. LEIBOWITZ: Just two other issues that I 3 haven't heard brought up here, is that one is that going to 4 decimals, in addition to all the other issues we've 5 discussed, quote flicker, resulted in a larger number of 6 trades. And as a result, as -- as you have sort of regulated 7 pricing that's trade-based, all of a sudden the trade count 8 explodes, and what ends up happening is that you have 9 windfall profits in terms of -- of those sorts of things. 10 And maybe we can hear some more facts on that. 11 But more -- more importantly, one thing we have to 12 be really careful about is letting exchanges, which have this 13 -- this pot of gold, compete unfairly against other market 14 centers that don't. Because they sort of can use that to 15 subsidize their operations at the expense of others like ECNs 16 who really don't have that. And you saw that, and that's why 17 that Instinet or Island or other people have been driven to 18 try to claim some piece of that, because they felt that they 19 couldn't compete fairly against exchanges. 20 MR. HARRIS: Now, would the solution be to make it 21 easier for them to get part of that pot of gold, or to 22 decrease the size of the pot? I'm troubled by the -- the 23 fact that we do have a number of folks who are offering 24 rebates of market data fees. And there were proposals to 25 offer greater rebates that the Commission abrogated. Did the 1 Commission make the right decision when it did so? It 2 stopped the development of a market in -- in market data 3 fees. And the question is: Was that appropriate? 4 Gary? 5 MR. GASTINEAU: Well, on that point I think it 6 definitely was appropriate because, while a number of the 7 ECNs and other markets which have been created have been very 8 innovative and have pushed the exchanges in desirable 9 directions, we probably have far more ECNs and other 10 acquirers of market data revenue than we need today, simply 11 because of the way the system worked. 12 I'd like to address the Chairman's question more 13 directly than it's been addressed yet. And that's because 14 I'm a more direct person, I guess, than -- than Rick, 15 although I think we're coming more or less from the same 16 point. You have a system out there that is reliant in 17 significant ways on market data revenue. You also, if I read 18 the papers correctly, have a budget situation which is not 19 particularly pleasant at the Commission. You're going to 20 have to have more help from SROs in the period ahead than 21 you've had in the past. And those SROs have to have means of 22 financing that activity. 23 The only SRO revenue that has any monopoly elements 24 of it in it at all that I can see, given the way the market 25 is structured today, is tape revenue. And, while there may 1 be all kinds of things that you would want to do to affect 2 the way it's charged for and so forth, you're not going to be 3 able to eliminate the dependence on tape revenue anytime 4 soon. So even if you feel in your heart of hearts that this 5 belongs to the world and should be given out free, I don't 6 think that's practical in the current environment. 7 MR. HARRIS: Gus? 8 MR. SAUTER: I would say that if -- there are 9 several people that actually provide value when it comes to 10 aggregating information. Part of the people would certainly 11 be the exchanges who do the -- the work of the aggregation. 12 The other people would be the executers or the -- the 13 investors who are using---a bad analogy---but say 14 intellectual property. They're giving up their intellectual 15 property, and should be compensated for that. 16 Now, who's going to -- to pay for the aggregation 17 process and who's going to pay for the intellectual property 18 process? In many instances, the same people that create the 19 value, from an intellectual property standpoint, are the 20 users of the data. If, in fact, they were exactly identical, 21 you probably could argue that the tape revenues should just 22 equal whatever the exchanges are providing insofar as a 23 service. Whether or not they're getting the right amount of 24 revenue from the various sources of revenue is debatable, and 25 I certainly wouldn't pretend to know that. But it seems to 1 me that the providers of this intellectual property are not 2 being compensated, except through potential rebates. And I 3 don't believe it should be universal. One way to do it would 4 be just to reduce the -- the fees. But I think another way 5 to do it is to keep them where they are, and allow the rebate 6 to the appropriate people providing that intellectual 7 property, who may be different from the people using it. 8 So I -- I do believe it can be appropriate to -- to 9 have lower fees using different venues, because you're 10 getting a rebate of part of the tape revenue. And I -- I do 11 believe that -- that can be appropriate, given the fact that 12 you have different providers of information and different 13 users of information. 14 MR. HARRIS: Joel? 15 MR. HASBROUCK: Thank you. I'd like to speak to 16 the distinction between value and cost. That this 17 information has terrific value is beyond debate. But in a 18 competitive market, the cost of something declines not to its 19 value, but really to the -- to the cost of producing it. And 20 I would argue that when you're trying to socially regulate 21 something like an information fee, like a data fee, the guide 22 should not be sort of -- the guide to allocation should not 23 be value, but the cost of producing it. 24 And this relates to the print rebates. Is a print 25 rebate a cost? Well, if you view your orders as an input, 1 it's a cost. But that's really stretching things quite a 2 bit. So I'd have to argue that they're -- tape rebates are 3 more distortionary than they are a valid cost. 4 MR. HARRIS: Rich? 5 MR. BERNARD: Yeah, if I could maybe sort of pull 6 some of these together. I feel I -- just a point of 7 information, market data revenues don't go up just because 8 the number of trades go up. Actually, the market data 9 revenues have been fairly flat for a long period of time, so 10 the per transaction cost of market data revenue was about a 11 dollar on the exchange in '85, and it's down to about 11 12 cents today. So that's actually been a declining -- your 13 point, it's a cost issue that the exchanges are interested 14 in. 15 Secondly, we completely agree that cost of 16 production is what has to gauge this, because the value of 17 this data is -- is, you know, very high. And the problem 18 that we -- we've mentioned is that the -- the primary 19 markets, the -- as I like to say, the Ellis Islands of the 20 New York Stock Exchange takes 88% of the share -- or 88% of 21 the trades and 84% of the shares day in and day out. And, 22 unlike Bernie, who picks his spots, we take everything that 23 comes through the door. We also have to put up with Eddie to 24 make sure that the game is fair. We also do listing and -- 25 and a number of other things. 1 MR. COLBY: Could I just say, for the people 2 listening on the web, that Eddie is their -- the head of 3 regulation at the New York Stock Exchange. 4 MR. BERNARD: I had understood that everybody knew 5 who Eddie was, but that's -- that's right. 6 So the system, to -- to your cost allocation point, 7 the way it's been done is the constituents to the cost 8 allocation. And -- and we -- we went through a lot of pain 9 and killed a lot of trees in the concept release looking at 10 other ways of doing that, the market data concept release of 11 -- of '98 or '99. And by the time it got to the Seligman 12 committee, I don't think you could find a vote for setting up 13 another rate making agency. 14 The -- really, letting the constituents do it is 15 the way to go. The problem and the source of the rebates, 16 which I agree are a distortion, and I know Larry can make a 17 market for rebates, but I think that's a hell of a way to run 18 a railroad. Harvey, in his first run at the Commission, got 19 away from rebates and -- because of -- of fixed commissions, 20 and I don't think we ought to be doing that in market data 21 now. 22 The problem is that you've got the cost umbrella of 23 the major markets setting the revenue of the periphery 24 markets, and that's what's dumb. And that's what Seligman 25 said get rid of. Let every markets come forward with its 1 value proposition and with its cost proposition. I submit 2 that much of the -- the value proposition of many of the 3 periphery markets won't be very high. But so be it. That'll 4 be what they get for -- for the data. 5 And I understand it's a discussion around whether 6 display rule creates an additional subsidy which we can have. 7 But if -- if you broke the consortium, go back to where we 8 were in -- in '73, where each market set its value 9 proposition out and each market, through its constituent 10 board, tapped that through a cost allocation, the subsidies 11 that create the rebates, that create the payment for flow go 12 away. 13 MR. HARRIS: Commissioner Atkins, you have a 14 question? 15 COMMISSIONER ATKINS: Yeah, I just -- I just had a 16 quick question. When we were talking about costs of 17 regulation and what-not, how do we know what we're paying for 18 as far as efficiency goes? I mean, what -- what sort of, you 19 know, judgement is there as to what that cost is, what -- 20 what's -- 21 MR. BERNARD: Can I -- can I help answer that? 22 MR. HARRIS: Let me -- let me paraphrase the 23 question because Commissioner Atkins' mic wasn't on. The 24 question is, is how do you -- is it correct, how do you draw 25 a link between what's being paid for regulation and -- and 1 what's being delivered? Is that fair? 2 COMMISSIONER ATKINS: Yeah. I was just wondering, 3 as far as, you know, from an efficiency standpoint, you know, 4 obviously we know what we're doing here at the SEC, but as 5 far as SROs go, you know, what are -- what are we getting for 6 the dollar that's being paid? 7 MR. BERNARD: If I could try to answer that, 8 remember the -- the very same group of constituents that are 9 deciding how much money we get are also deciding the level of 10 services we get. The exchange is a cooperative. And so we 11 can't get very out-of-whack with what our constituents want 12 without hearing about it. 13 In the case of regulation, there is an overlay on 14 that, though, of course, which is that the Commission asks us 15 to do a lot of things. And recently the Chairman has been 16 asking us to really do a lot of things. And -- and so the -- 17 the -- in effect, the Commission, in how it oversees the 18 regulatory process and how it allocates what it cares to keep 19 in-house and what it cares to go into the self regulatory 20 process is actually decide -- and then it sends in the folks 21 from OC to make sure we're doing whatever it is that the 22 Commission asks us to do. And that's how you actually get 23 the balance on how much money is devoted to regulation. 24 MR. HARRIS: David? 25 MR. WHITCOMB: Could I make a quick point? I think 1 Commissioner Atkins' question illustrates the -- probably the 2 most difficult but solvable problem with the markets today. 3 We have overlapping multiple layers of regulation and 4 regulatory arbitrage, and we're being forced to pay for it 5 through things like tape revenue fees. If we would eliminate 6 SROs and turn them from cooperatives into private businesses 7 whose product is producing trading, fundamental product is 8 producing trading services, the need to cross-subsidize their 9 regulation would go away. The SEC should perform all rule 10 making and all enforcement. And if that were done, you 11 wouldn't need these huge tape revenues to support an 12 unnecessary service. 13 MR. BERNARD: Just one point. There is no cross- 14 subsidy most years between our regulatory services and our 15 market data. By and large, Ed Kwalwasser is -- is self- 16 supporting most years because of our discrete regulatory 17 fees. And secondly, if you remember that the production of 18 -- of executions and -- and market data is a joint product, 19 you're also throwing the listing fees into the pool in the 20 creation of that joint product. So there's -- there's no 21 cross-subsidy coming from market data regulation. 22 MR. HARRIS: Ed? 23 MR. NICOLL: Well, just a couple of points, and -- 24 one, you know, we generally support the New York Stock 25 Exchange's position on this, as surprising as that may be to 1 some. And, you know, it shouldn't be lost on everybody that 2 there's a tremendous disparity, on a per transaction basis, 3 between what some of the -- the exchanges are charging or 4 receiving on market data, and what others don't. I mean, 5 there's not much opportunity for rebating and for market 6 distortions when you're talking about 11 cents a share or 11 7 cents a transaction, I should say. But there -- there is 8 when you're talking about market data fees of 5 to $10 a 9 transaction. And -- and this is really the -- you know, 10 causing much of the concern that we have here. 11 So I think, you know, fundamentally we should -- we 12 should be aware of that. I don't think rebating market data 13 would be much of an issue, one way or another, if we were 14 talking about pennies a transaction in the first place. And 15 I think that's -- that's a very important distinction to -- 16 to make, and people should keep in mind. 17 MR. RYAN: Can I -- can I just make a follow-up 18 point on that? 19 MR. HARRIS: Sure. Go ahead. 20 MR. RYAN: I think it's a little -- it's a little 21 bit of a misnomer to use the -- the way revenue is 22 distributed, which has been going on for 25 years, as a kind 23 of a -- a measure of the value or the cost of creating it. 24 So, you can have the infrastructure, and it's going to cost 25 the same regardless of how many transactions go across that 1 infrastructure. So it's -- it's kind of an artificial device 2 is the way we -- we distribute that revenue. But the 3 infrastructure, whether it's regulatory or technology or 4 otherwise, is -- is very significant. 5 And so the cost -- if you look at it from a cost 6 perspective, the -- to launch a marketplace, and especially a 7 primary marketplace, has significant cost structures. And 8 the -- finding revenues to support that is -- is very 9 significant. And with the limitations on regulatory -- on 10 the exchanges in terms of the regulatory handcuffs, there's 11 very few opportunities to -- to generate revenue sources. 12 MS. NAZARETH: Could I ask something? 13 MR. HARRIS: Go ahead. 14 MS. NAZARETH: I'd like to ask Rich a question, 15 because you talked about, you know, if you did away with the 16 consortium you wouldn't, you know, have these problems. I 17 don't understand how you can say that when you clearly still, 18 particularly in the case of a market like the New York Stock 19 Exchange, have -- have dominant market power. So the fact of 20 the matter is that, to some extent, the oxygen on which our 21 markets breathe would include New York Stock Exchange data. 22 It's not something that the markets could function without. 23 So you have -- you're in a much different position from a 24 smaller market whose -- whose information may be tangential 25 to traders. So I'd like you to address that. 1 MR. BERNARD: For sure. But we have, as far back 2 as we could count, almost 70 years of the New York Stock 3 Exchange not exercising whatever that market power might be 4 by pricing our market data way below not only its cost, 5 because of its joint production, but below its value. And as 6 I've said, the reason that that happens is because the very 7 people who are sending us the orders, which, by the way, the 8 market data magnetizes, so that, in and of itself, is a 9 governor on how much we charge for market data. But the very 10 people sending us those orders are the very people who are 11 sitting on our board and who are deciding what kind of 12 institution we are. And that has been the resolution. 13 And, you know, I could sit here and hold up this 14 pen and say, you know, I'm keeping the elephants away by 15 having this pen, and I wouldn't believe it. But I can tell 16 you that we're keeping the market datas at about 70% of our 17 revenue because of that board of directors and that 18 constituent process. And the way I can tell you is there was 19 no pen, there was no CTA, there was no Commission regulation 20 to speak of in the market data arena from 1860 or whatever it 21 was, when we started sending it out, until 1975. And yet the 22 constituents, themselves, taxing themselves, chose to hit up 23 market data for only -- for less than 20% of their revenue. 24 MR. HARRIS: Rich, during most of that period, the 25 period prior to the '70s, New York Stock Exchange would not 1 sell data to the public as a whole, only sold data to its 2 members. And so I understand that, given that the members 3 run the exchange and they're buying the data, that there 4 might be a very strong control on prices. In the present 5 environment, we could argue that -- that in that former 6 environment the price to the public was infinite. In the 7 present environment, the public's buying a lot of that data 8 and... 9 MR. BERNARD: No, it's -- it's not. Remember, the 10 institutional investors are paying their fair share, but -- 11 but the retail investor pays almost nothing. It's the -- of 12 course, it's Schwab and the other broker-dealers who pay -- 13 pay those fees. But, secondly... 14 MR. HARRIS: Well, what would be the -- who would 15 represent the institutional investors as the New York Stock 16 Exchange decided how much to charge for data in an 17 environment such as you and -- and Cathy Kinney proposed 18 earlier today we... 19 MR. BERNARD: Oh, the same people who have 20 represented institutional investors forever. For -- for -- 21 we actually literally, in our constitution, require that an 22 institutional investor be on the board. In fact, we've had 23 several people at any one time who are neither broker-dealers 24 nor listed companies. Listed companies themselves, of 25 course, have an interest in -- in the best interest of the 1 institutional investors, since those are their shareholders. 2 So you've got only half of our board -- less than half of our 3 board, if you include the management directors, are -- are 4 broker-dealers. And so that's how that works out. And you 5 -- if you think that the CTA process is what is governing how 6 much the New York Stock Exchange charges, you are simply 7 wrong. 8 MR. HARRIS: Wayne? 9 MR. WAGNER: Just an irreverent comment here. You 10 know, I bought the newspaper and it was filled with 11 advertising -- advertisements of things that I could buy and 12 the prices that people were willing to sell them at. And it 13 seems to me that the price tapes and the price information is 14 the advertisement component of the exchanges here. So I kind 15 of -- you know, I go with what Chairman Pitt said here, you 16 know. Why should we pay for that? You know, why -- why is 17 it not something given away for free in order to entice 18 people to... 19 MR. BERNARD: Do you charge for your reports? Your 20 -- your business that you do? 21 MR. WAGNER: Oh, yes. 22 MR. BERNARD: But that's advertising. You're 23 telling the people about what a good job you do. Why -- why 24 would -- should you be able to charge for... 25 MR. WAGNER: Well, I'm here sort of advertising, I 1 suppose, and I didn't charge for this. 2 MR. BERNARD: And I'm not charging, either. My 3 point being that, without making a -- an intellectual 4 property argument---although anybody who wants to play lawyer 5 with me, I can certainly do that---is that there is costs 6 that are done in getting that, quote, "advertising" out, and 7 those costs have to be shared by the industry. We don't have 8 any money that doesn't come from the industry. 9 MR. WAGNER: But it's an administered cost. You 10 know, when you -- when you hear of things like rebates you 11 say, "Well, maybe that -- maybe the cost setting mechanism 12 isn't quite adequate to what -- what we need here." 13 MR. BERNARD: Exactly right. 14 MR. WAGNER: You know, if we had market -- you 15 know, in 1975 we decommissioned the fixed -- excuse me -- 16 well, decommissioned the fixed commission. Maybe this is 17 another parallel to that. 18 MR. BERNARD: Couldn't agree with you more. And if 19 you broke the consortium, you would stop the extra money 20 going to the markets that aren't carrying the cost of the New 21 York Stock Exchange, and they would have to set -- put their 22 own value proposition out, as well as their own constituent 23 governance, to decide what they got for their -- their data. 24 MR. NICOLL: Larry, just very quickly. 25 MR. HARRIS: Go ahead. 1 MR. NICOLL: I mean, that -- everybody should 2 realize that that basically is the position that we staked 3 out in the Seligman commission in the minority with -- with 4 Schwab, which was that -- that we should use a free market 5 approach here. 6 I would -- one of the things that I would -- a 7 total free market approach. I think what -- what the New 8 York Stock Exchange wants to do is a -- is a half-step rather 9 than a full step, but we like progress and we're for 10 incremental progress, so in that respect... 11 But I think one of the answers, Larry and -- and 12 Annette, to your -- to your question, we made this last case, 13 we don't have any hard data one way or another, and I think 14 it's counterfactual. But, in fact, you're right, you know. 15 Quotes are one of the elements of advertising that gets 16 people to come to an exchange. And if an exchange charges 17 too much for its quotes, then people are not going to sign up 18 and are not going to know what the trading interest is down 19 there. And they're going to be at -- at risk of losing, you 20 know, their business to a -- to a competing exchange. So 21 there is a tension that any exchange has in terms of 22 publishing its quotes and it... 23 MR. HARRIS: Ed,... 24 MR. NICOLL: You know. 25 MR. HARRIS: ...at the beginning of our discussion, 1 we -- I believe we established that the national best bid and 2 offer was extremely important. If we break up CTA and allow 3 each exchange to -- and market center to sell its data, will 4 we still require that people buy the data? And if we require 5 people to buy the data, how will we price it for people who 6 have governmental structures -- government structures that 7 are not as -- as progressive as the New York Stock Exchange 8 is? 9 MR. NICOLL: Yeah. I think our -- our position 10 then, and -- and still is, that -- you know, that we should 11 do away with the Vendor Display Rule, and that we should not 12 require people to buy the data. I mean, it should be -- one 13 of the things that I bring, for whatever it's worth, to this 14 debate is a unique perspective. I ran two very large retail 15 brokerage firms, and for a long time. And I do have an 16 understanding of what the retail investor wants. And -- and 17 -- you know, and before a retail investor trades, he'd like 18 to see the best bid and offer to the best of -- of our 19 ability to deliver that to him. 20 I can't -- I don't think there is a retail broker 21 in the world today that serves its customers without 22 delivering a best bid and offer; yet, it's not obligated to 23 do so. There's no rule, prior to accepting a trade, that a 24 broker delivers the best bid and offer. It's -- the reason 25 that he does so is because he has best execution 1 responsibilities. And we think the Commission can rely to a 2 greater extent than it does now on those best execution 3 responsibilities, and put those duties on the brokers. And 4 the mechanisms are now in place, like they weren't in the 5 past, technology has improved to the point where -- where the 6 private marketplace can deliver this information to people 7 and -- and the people that have that best execution 8 responsibility can gather the data that they need in order to 9 perform their brokerage responsibilities. And we don't need 10 this sort of what I would regard as -- as a cumbersome and 11 out-of-date regulatory structure, that has unintended 12 consequences, to create that -- that mandate to deliver the 13 best bid and offer to investors before they trade. 14 MR. HARRIS: In an environment where there is no 15 vendor display rule, so that vendors do not have to buy from, 16 say, a new exchange, how would new exchanges establish 17 themselves? 18 MR. NICOLL: Well, I think -- you know, I think new 19 exchanges will establish themselves by coming up with 20 innovative and new ideas and new services. I don't think 21 that there's much of a hope of anybody starting, for 22 instance, a traditional floor-based exchange for listed 23 securities to compete with the New York Stock Exchange these 24 days. But I do think that there are different types of -- of 25 trading mechanism, different types of trading networks that 1 do have a chance of competing with the New York Stock 2 Exchange, and -- and we don't need to subsidize them. 3 The New York Stock Exchange is an institution that 4 has a lot of competing interests that it has to deal with, 5 and there are a lot of -- you know, of interstices among 6 those competing interests that I would argue create 7 inefficiencies and allow new and innovative marketplaces to 8 compete with them. I think, to a large extent, that 9 competition has been actually disincentivized by a regulatory 10 structure that in many cases creates barriers to competition. 11 MR. HARRIS: Back to the point, the question of 12 rebates. The premise -- one of the premises of competition 13 among market centers is that new technologies or new ways of 14 doing business may emerge that have lower costs. And if 15 that, indeed, is the case, wouldn't it be appropriate, then, 16 that those centers be allowed to rebate the excess of their 17 revenues over their costs? If we assume that, indeed, it is 18 a very competitive market, that would explain what's going on 19 with the market data rebates. 20 MR. RYAN: Well, I'll take a crack at it. One -- 21 one part of the problem with the market data rebates right 22 now is you have significant market participants that are not 23 fully participating in the apparatus, the National Market 24 System, that generates the market data. So they are -- it's 25 -- they are able to go out and differentiate themselves, and 1 the way that the traditional marketplaces operate, by 2 ignoring the rules and the system for -- of the National 3 Market System. And that -- that allows them to operate at a 4 much lower cost. And then that system is funding their 5 operations. So it's -- you have a very artificial 6 environment here where -- where markets are being saddled 7 with the burdens and responsibilities of operating a National 8 Market System, and using the -- the results of that to -- to 9 fund those who don't play in that environment. 10 MR. HARRIS: Rick? 11 MR. KETCHUM: Yeah. I think there is -- there is 12 something of what Mike says, but I think it -- it goes too 13 far. It seems to me the Commission has two choices with 14 respect to this, and neither of them are perfect, and -- and 15 both probably should be looked at. You could go down Ed's 16 analysis, conclude that -- that the marketplace has evolved, 17 from a -- from a cost standpoint and a technology efficiency 18 standpoint, that -- that valuable information will be 19 delivered and you don't have to worry about the consolidated 20 best bid and offer. And -- and we're certainly closer to 21 that than ever before. I am a product of my past. I do 22 remember when -- that vendors often chose not to carry that 23 information. I do remember why. Because those smaller 24 markets tended to generate more message traffic from the 25 standpoint, and -- and impose costs that -- that vendors, 1 freely choosing, are not going to pick up. 2 Now, one can simply say, okay, then that 3 information isn't valuable enough to pick up and, cool, let's 4 move from there. That worries me a little bit from a market 5 entrance standpoint. I think it's closer than it used to be, 6 but it still worries me a lot from -- from that standpoint. 7 On -- on the other hand, it does -- it does make 8 sense to at least get us out of a planned environment and 9 give people a choice there. The Commission can require 10 consolidated information, with still an ability for each 11 marketplace to have a separate negotiation over the value of 12 their information. And I think that does make sense. Gets 13 us out of a lot of anti-competitive things and -- and 14 inability to make progress or even react that is caused by 15 plans which drive to the lowest common denominator. And I 16 think that's kind of good. 17 But if you -- if you have consolidated information 18 as a requirement, it will require the Commission to make 19 judgements with respect to the -- the appropriate price of 20 the data. And I think that's -- that's sort of inevitable. 21 I think that, in the end, has been okay over the last 30 22 years. And -- and it is better for the Commission to wrestle 23 with that than to -- to take the risk that -- that 24 competitors basically can't enter into the marketplace. 25 Well, again, I -- I could imagine thinking that'd be okay for 1 my particular position. I'm not sure it's the right public 2 policy position. 3 MR. BERNARD: If I could just extend -- Rich 4 Bernard for the stock -- New York Stock Exchange. If I could 5 just extend Ed -- Rick's point, the -- when you -- if you get 6 rid of the display run, and I've got to tell you that the New 7 York Stock Exchange was on both sides of that issue, so 8 depending on whether Bob Britz or I were sitting in the 9 chair, we had a different answer on that issue. 10 But if you do get rid of the display rule, the 11 concern that you have is that the -- if the mechanism I've 12 described for the New York Stock Exchange works, and indeed 13 has been working, and the CTA really has nothing to do with 14 it, your issue is -- of course, is not the New York Stock 15 Exchange or the primary markets, it's the -- it's the 16 periphery markets. 17 But there are a number of factors that are going to 18 tend to prevent that from being the standoff that -- that you 19 fear. First of all, there will be the major markets, who, to 20 be sure, are allocating their costs. But it's hard to 21 imagine that a periphery market is going to be able to do 22 better than -- and have market data prices that are higher 23 than what the major markets are doing on some units, a share 24 or a quote or a -- or a trade basis. So that -- that's going 25 to tend to govern it. 1 Secondly, let's not forget that all these markets 2 are still people looking for orders. And -- and while there 3 are some more difficult issues presented when you move from a 4 cooperative mutual model to a profit maximizing model, the 5 fact remains that your law, Section 6, says that stock 6 exchanges have got to have representatives of all of their 7 constituents in their governance function; and moreover, 8 those representatives continue to have the ability to give or 9 withhold orders. And they especially had that ability in 10 regard to the peripheral markets whose competition you're 11 trying to foster to keep the large markets honest. 12 So, other than some, you know, sort of theoretical 13 examples that Larry and I -- you've done, where we worry 14 about two guys in -- in Denver setting up an exchange, solely 15 to take advantage of the display -- I think these are not 16 terrible real concerns. But if they are, Rick has given you 17 the answer, which is, again, there is a place for Commission 18 intervention. 19 MR. HARRIS: Well, I think we're pressing up 20 against lunchtime, so perhaps we should close here. I want 21 to give Commission Glassman the last word. 22 COMMISSIONER GLASSMAN: Well, despite the fact that 23 I said I didn't think we'd get any solutions, I believe I 24 heard a consensus that the market data revenue rebates should 25 come to the SEC; was that right? I think we'll take an hour 1 break for lunch and reconvene at 1:15. Thank you. 2 (Lunch break.) 3 AFTERNOON SESSION 4 MR. COLBY: Well, good afternoon. This afternoon's 5 session is focused on best execution. Viewed broadly, 6 obtaining the best possible execution of orders in the market 7 is the reason we're all here together today, and for many of 8 us, the reason we're in the business at all. This goal's so 9 important, that Congress said it was indued with the public 10 interest, providing high quality executions, and thus 11 reducing friction in the secondary markets, encourages 12 investor's participation in the securities markets, and makes 13 possible the capital-raising function. 14 The best execution goal can be viewed from the 15 standpoint of an individual order, or collectively from the 16 standpoint of all orders in the market. From the standpoint 17 of an individual order, this goal is bolstered by the legal 18 obligation of best execution owed by a broker-dealer to its 19 customer. This obligation, which is grounded in agency and 20 fiduciary law, requires a broker-dealer to obtain the most 21 favorable terms for a customer's transaction available under 22 the circumstances. Easier said than done. 23 In a world where a security is traded in only one 24 market, and all orders meet and are executed within that 25 market, best execution would be relatively simple both for 1 individual orders and orders collectively. It would be 2 primarily a matter of determining how best to present each 3 order to the market. The competition between orders on that 4 market would produce the best prices available for those 5 orders at that time. 6 On the other hand, a single market for a security 7 would preclude competition between market centers, which 8 could encourage the development of cheaper, more innovative 9 forms of trading services. And for that reason, the Congress 10 and the Commission have long encouraged competition between 11 market centers. Unfortunately, in a world of competing 12 market centers, assuring best execution becomes much more 13 difficult, both for individual orders and orders 14 collectively. There are both agency problems and structural 15 problems. 16 First, the agency problems in executing individual 17 orders. In a world of competing market centers, it's very 18 difficult for investors, especially small investors, to 19 monitor and enforce best execution. The amount at stake 20 individually is often small, and detecting, much less proving 21 the availability of a better execution is arduous. 22 Therefore, retail customers put relatively little pressure on 23 brokers over best execution. 24 At the same time, brokers have incentives to not 25 strenuously seek the most optimal execution for customer 1 orders, both to avoid the search costs involved, and because 2 of profit opportunities from trading with those orders. The 3 broker could either trade directly with customer orders as a 4 dealer, or sell them to another dealer in return for payment 5 for order flow. 6 There are also structural problems in executing 7 individual orders. First, if -- if best execution is to be 8 achieved in a world where multiple market centers are 9 competing with each other, broker-dealers must be able to 10 identify the location of the best available prices, as we 11 talked about this morning, and access those prices routinely 12 and efficiently. One of Congress's principal goals in 13 enacting the 1975 act amendments, was to address poor 14 customer executions resulting from the trading of securities 15 in separate, unconnected markets. In addition to 16 establishing National Market System transparency mechanisms, 17 Congress envisioned a framework in which broker-dealers had 18 access to all markets, and competing markets would be linked 19 together in ways that would produce efficient executions at 20 the best available prices. 21 To this end, the Commission has striven to remove 22 discriminatory barriers to direct access erected by 23 marketplaces. The Commission has also sought to integrate 24 new trading systems, such as ECNs, alternative trading 25 systems, and other routing services into intermarket 1 mechanisms, and the Commission's encouraged efficient 2 linkages between market centers, at least for some 3 securities. These intermarket linkages proved a backstop 4 mechanism to help assure orders that are routed to one market 5 can reach a better price available in a competing market. 6 Some smaller and newer markets have argued that this sort of 7 intermarket linkage helps them compete with dominant markets 8 by reducing the need for each broker-dealer to build separate 9 linkages to each competing market. 10 By way of example, on the listed market today 11 members of exchanges and the NASD can access each other's 12 quotes in exchange listed equity securities through a now 13 notorious order routing system called the ITS. In contrast, 14 there is no intermarket mechanism like ITS linking competing 15 markets in NASDAQ securities. Rather, the competing markets 16 provide member access to their markets, and firms that have 17 best execution obligations use those access methods to access 18 each individual market. 19 Now, in addition to the issues for executing 20 individual orders, there are also best execution issues for 21 the orders collectively, the whole -- the whole set of orders 22 in the market. How should we structure a system to produce 23 the best prices for all orders? A National Market System 24 must thus reward orders for contributing to price -- a public 25 price discovery. As we discussed this morning, otherwise, 1 the existence of multiple market centers competing for order 2 flow may reduce the interaction of all buying and selling 3 interests, and thus discourage vigorous limit order 4 competition. This arises in particular because most markets 5 are structured in order to encourage competition within their 6 own market, but not across markets. 7 Only -- to my knowledge, only in the ITS system 8 must participants accord -- accord a level, and not a 9 complete, but a level of price priority to quotes of other 10 markets displayed at a superior price. And even there, as in 11 the NASDAQ market, dealers trading with their customers can 12 freely match a superior price displayed by another market 13 center. These sort of issues give rise to the question of 14 best execution for orders across the market. So, with that 15 -- that picture, let's start with some introductory questions 16 about best execution. 17 Currently today it's common practice for brokers, 18 particularly routing retail orders, to decide to choose a 19 default market at which to route them. It's where they -- 20 they decide, sort of on a collective basis, we're going to 21 route our orders to a default market, without attempting on 22 an individual basis to look at the desires or the needs of 23 the individual order, or necessarily whatever quote is being 24 displayed by that market at that time. And as I just say, 25 that's -- that's a typical means that retail orders are 1 displayed. 2 How does this compare to how institutional orders 3 are instituted? Gus, could I pass that one to you? 4 MR. SAUTER: Yes. Thank you, Bob. 5 We have access to bid and -- and offers for every 6 type of trading venue, whether it's ECNs, whether it's 7 primary exchanges, whether it's other exchanges. We actually 8 route our orders ourselves, so we do not just give up an 9 order to a broker and ask them to execute it for us. We -- 10 we take charge of that process and determine our own venue. 11 We -- we do use the brokerage community, but not for 12 direction or routing for us. 13 MR. COLBY: Wayne, would you like to amplify this 14 at all? 15 MR. WAGNER: Yes, I would. The best execution, as 16 you've defined it here, is really a -- an exchange and broker 17 obligation. But that operates with inside another obligation 18 for best execution, which Gene Goldkey will talk about, which 19 the AIMR is, I understand, about to pass some standards on. 20 And that duty is the duty of a fiduciary, in effect, to 21 secure the most favorable terms for the trades that need to 22 be maintained to -- to operate that. 23 That leads to the Guses of the world giving 24 instructions to broker-dealers for them to follow out, which 25 may or may not have some discretion with them, that allows 1 that to do that here. And -- and, again, the -- you know, I 2 was at a conference where Anthony Neuberger from the London 3 School of Business said that best execution begets market 4 design. In other words, the whole purpose of thinking about 5 market design is kind of empty unless you have this idea of 6 best execution with respect to the -- the actual investors, 7 who are investing in the marketplace, getting the best 8 possible prices and conditions for the orders that they want 9 here. 10 And I keep coming back to the -- the numbers that I 11 look at, which says that, you know, 60% of the dollars -- 12 institutional dollars being traded in our databases is in 13 orders of over 250,000 shares. Well, mostly the broker 14 doesn't ever see an order coming to the broker of 250,000 15 shares. That is sliced and diced and chopped down in what I 16 call a meat grinder effect, where the manager says, "I want 17 250,000 shares," gives this piece by piece to his own trader, 18 who gives it piece by piece to his own broker, who gives it 19 piece by piece, each piece level being smaller pieces, until 20 it's a size that it can get through that marketplace. And in 21 the process, of course, the time to complete that order 22 lengthens out significantly. 23 MR. COLBY: David, how does this form of default 24 order routing to one particular market compare to how active 25 traders route their orders? 1 MR. WHITCOMB: Yeah. Active traders, like -- like 2 some institutions, like Gus, make their own decisions, and 3 frequently access -- and clearly need to look at all markets, 4 all venues at the same time, and rarely would send -- if they 5 had a large order to do, would rarely send it to a single 6 venue, or -- or slice and dice it and keep sending it to a 7 single venue. 8 One of the wonderful things about technology these 9 days is that it's relatively easy to -- to be connected to 10 multiple venues. And active traders are; most brokerage 11 firms are; most large brokerage firms are. And there are 12 firms who provide that service, that smart order routing 13 service. So that a lot of what was intended to be achieved 14 by the National Market System, and originally most of us 15 academics thought it needed to be done by -- by design and by 16 centralizing, is now being done by, for lack of a better 17 term, network technology, where it's quite cheap and quite 18 quick to do the slicing and dicing, and to -- to split up the 19 order into multiple venues. 20 MR. COLBY: Which to me raises the question is -- 21 is this default form of order routing for retail orders still 22 appropriate, given the technology available? Larry, I'm 23 going to -- certainly. You want to take that one? 24 MR. LEIBOWITZ: Sure. I would say it's probably a 25 misnomer to assume that there is default of routing. I think 1 any destination certainly that Schwab uses has an obligation 2 to either provide the best bid or offer that is -- that is 3 conglomerated among all markets, or send it to that 4 destination that has it. And it has to do it on a trade-by- 5 trade basis, and also show, over the large numbers by way of 6 its SEC statistics, that that has been done for -- for those 7 customers. So it's not like there are -- there are disparate 8 destinations, and it's just send it to one destination, send 9 it to another, and so on. I think things have really 10 progressed quite a -- quite a ways from -- from what were 11 probably past practices. 12 MR. COLBY: How does Schwab route a customer order, 13 if you could split it between ones that they make a market 14 and ones they don't make a market in. What are the factors 15 that go into how it's routed? 16 MR. LEIBOWITZ: And this probably is different 17 versus -- listed versus -- versus over-the-counter. But -- 18 but the way we view internalization is -- is, in other words, 19 routing to ourselves, is we have to be able to provide more 20 liquidity than the inside of the market. And if we can't do 21 that, then we have to give it to the inside of the market. 22 So we view internalization as enhancing the best bid and 23 offer. And, in fact, our execution statistics show that. 24 Otherwise, we have intelligent order routers that 25 determine the best bid and offer at any point, and will 1 parcel an order up to continue to get that best execution, 2 much in the way that -- that an active trading intelligent 3 order router would do. 4 MR. COLBY: And they parcel it up to get that best 5 execution by routing to the best bid and offer, or by some 6 other means? 7 MR. LEIBOWITZ: Yeah. No, to -- to -- in fact, we 8 -- we probably route out two-thirds of our order flow between 9 listed and OTC, either to alternative exchange destinations 10 like New York, to third market, like -- like someone like 11 Bernie Madoff, or even to other ECNs. I mean, we're the 12 biggest ECN customers, or certainly among them. 13 MR. COLBY: Maureen? 14 MS. O'HARA: Well, I'd like to just follow up a 15 little bit on the point about trying to make the collective 16 liquidity better. I think the challenge with best execution 17 is there sometimes seems to be a tension between best 18 execution for individual orders and best execution for the 19 market. I think, as we look at best execution, it has to be 20 done on an order-by-order basis, or you really can't get 21 anywhere. But you also have to go beyond that. When you 22 take an order and you match the price that's in the -- in the 23 market, you're not actually improving the collective best 24 execution for the market, because if you'd allowed that order 25 to be out there and having the dealer competing for that 1 order by improving upon the market, you'd have both made 2 better execution for the trade and better execution for the 3 market. And that's the piece that I think is challenging 4 when we look at current practices such as payment for order 5 flow. You can justify them as giving best execution for the 6 order, but they undermine best execution collectively for all 7 of them. 8 MR. COLBY: Bernie, would you like to address that? 9 MR. MADOFF: I knew that. Does not surprise me. 10 Even though we stopped paying, we always have to answer the 11 questions about paying for order flow. 12 First of all, I think that there -- there are many, 13 many misconceptions out there. And I -- they keep 14 resurfacing by comments made about payment for order flow 15 today. The -- there is absolutely no reason why payment for 16 order flow, all its various forms -- and we won't bore 17 everybody here with the various forms because everybody's 18 familiar with them. There's no reason why payment for order 19 flow, itself, in and of itself, has to come at the expense of 20 best execution. And I would venture to say that certainly in 21 the equities markets today I could prove to you that the -- 22 the venues that are paying for order flow, for the most part, 23 are giving better executions than those that are not paying 24 for order flow. 25 The problem with payment for order flow, and I -- 1 it pays to go into this for a little bit in background, is 2 that most of the competing markets use payment for order flow 3 as a way of disparaging another marketplace. It's very easy 4 to do that. If you want to -- if you want to have an 5 inflammatory conversation, just turn it into a rebate. And 6 you could see that happening with market data. You call it a 7 rebate, it becomes a problem. If you say I'm passing on part 8 of my profits to my customer and he's sharing those profits 9 in -- in the form of lower commissions to his customer, all 10 of a sudden payment for order flow doesn't seem that bad. 11 Now, the SEC dealt with the perception of payment 12 for order flow very effectively, in my mind, with the 11Ac1-5 13 and 11Ac1-6 stats. It took years for that to come about, but 14 by doing that, they took the perception away from -- or the 15 negative perception away from payment for order flow in the 16 equities market. There is still a problem in the options 17 markets because there are no linkages, there are no 11Ac1-5 18 stats, for all intents and purposes. And that's a different 19 issue. But there's no reason why that can't be addressed the 20 same way it has in -- in the equities market. 21 If you look at the equities markets today, in both 22 listed and in NASDAQ markets, the spreads are -- are 23 virtually non-existent. The -- the payment for order flow is 24 not a problem because it's very easy for a firm, whether it 25 be a Charles Schwab or for a Merrill Lynch to look at the 1 execution stats that are now uniform, that everybody, for the 2 most part---I'm sure my -- my colleague at the left will take 3 issue with this---but for the most part everybody feels that 4 the stats are valid stats for what they were designed to do 5 and measure small orders. When you talk about large orders, 6 that's a different issue. 7 But firms can today look at the -- the -- very 8 easily the 11Ac1-5 stats that measure execution quality, and 9 they can clearly make a determination where they should be 10 routing their orders. And, for the most part, that is what 11 has been happening. It's somewhat slower than firms like 12 ours would like to see happening, but I think that's a -- an 13 ongoing -- an ongoing process. So I think it's very clear 14 that -- I think it's very important that we don't keep on 15 characterizing payment for order flow as being this evil 16 child, because that is not the case. And the statistics are 17 clearly out there to demonstrate that today. 18 MR. COLBY: Joel? 19 MR. HASBROUCK: Thank you. I -- I notice the 20 discussion to this point has sort of gone along two lines. 21 First of all, it seems pretty clear that the institutions can 22 take care of themselves, at least in figuring out where to 23 route their orders and how to best achieve their executions. 24 Whereas retail investors, being dependent on the national 25 best bid and offer, enjoy, what Bernie has pointed out, is an 1 extremely narrow spread. This may suggest that, at least as 2 far as retail traders are concerned, best execution of market 3 orders is not a particularly large issue. 4 But then this suggests that we broaden it beyond 5 best execution, to best order representation. And this would 6 include limit orders as well. And here it's not clear what 7 best execution criterion would rely on. Furthermore, the 8 criterion of best execution, almost by definition, refers to 9 the trades that get done. It ignores the trades that do not 10 get done, and sometimes trades that get done at bad prices 11 may be preferable to no trades at all. But I think that 12 limit order representation is probably the elephant in the 13 room here. 14 MR. COLBY: Sorry, Brett. Do you want to pick up 15 on Bernie's comments... 16 MR. REDFEARN: I mean,... 17 MR. COLBY: ...regarding... 18 MR. REDFEARN: ...you know, the thought -- the 19 thought that I had on this was that -- and part of my job has 20 been looking at ways in which we could bring order flow to 21 the -- to the floor of the American Stock Exchange. And we 22 were very pleased, actually, when the 11Ac1-5 reports came 23 out, because it was an opportunity to sort of really quantify 24 apples to apples comparisons, and everybody knew prior to 25 that there really wasn't sort of the apples to apples 1 comparisons. And it was an interesting experience having 2 discussions with a lot of different firms, because first of 3 all, the notion of best execution, even with the data, was 4 somewhat, you know, anomalous. Some people said, you know, 5 it's about speed; some people said it's effective spread; 6 some people were looking at something else. 7 So even if you could come in and say, in certain 8 cases, you know, effective spreads have been superior for, 9 you know, three consecutive months or for two consecutive 10 quarters, a lot of times we saw that that actually did not -- 11 did not change switches in various order flow routing 12 practices. And usually that was in environments where 13 payment for order flow was part of the equation. 14 So what our initial hopes were, in terms of some of 15 the data in some cases, was something that -- that didn't 16 completely pan out, because -- because, on the one hand, even 17 with really solid stats, it was -- the definition did remain 18 somewhat anomalous, and there still were, you know, 19 coutnervailing forces with -- with payment for order flow out 20 there, which still, in many cases, did determine some of 21 these decisions. 22 MR. COLBY: Ed? 23 MR. KWALWASSER: Yeah. 24 MR. COLBY: This is the famous Ed Kwalwasser, now 25 joined us at the table. 1 MR. KWALWASSER: Yeah. The expensive one. 2 (Laughter.) 3 MR. COLBY: Or should I say notorious, maybe. 4 MR. KWALWASSER: I think we have to look at, from a 5 market perspective, and when we talk about payment for order 6 flow and internalization. The last panel -- first panel 7 agreed that one of the characteristics of a National Market 8 System that we all said -- who were sitting here said was 9 really good, was that customer orders, should be able to 10 interact without the benefit of a dealer. Both 11 internalization and payment for order flow are distortions of 12 that public good that we all agree was there. And, to the 13 extent that those orders aren't in the marketplace, they're 14 not helping create the best price for all orders in the 15 marketplace; and therefore, I think... 16 I never thought Bernie was a bad guy. I started 17 out, however, out of law school representing people in the 18 record industry and in the rock-and-roll part of the 19 business. And to think that the securities industry has a 20 lower code of conduct than the record industry is really, I 21 think, something that I -- I never wanted to believe would be 22 part of an industry that I was working at. 23 I think also, when you look at the dash five 24 statistics, that they are so flawed, both technically and 25 internally, that it's hard for people to use them as a guide. 1 And I'll point out that we -- we took our figures and we 2 brought them to each of the two institutions that most other 3 providers use to put out their dash five statistics. They 4 were so much better than our internal view of what we did 5 that we determined that we couldn't use them, because we 6 thought that they didn't present what actually was taking 7 place in -- in our marketplace, and we -- we did alert the 8 Commission, and we hope that some -- some staffers are going 9 to go out and look at that to make a determination as to 10 whether, in fact, they all are equal, or whether there's some 11 difference based on what the provider is -- is putting out. 12 And there are also substantial disincentives to an 13 active and large market in the dash five statistics. If we 14 have a -- an offer for 5000 shares at a dollar or $10, and 15 50,000 shares come in, and 5000 -- and 20,000 shares are 16 given the price of $10, which is much better than the market 17 would show, that would show that 50 -- 30,000 shares did 18 worse than was traded outside the market, even though we gave 19 the market much better than they would have had -- expected. 20 MR. COLBY: Ed, you've got ten more seconds to take 21 apart the statistics, and then we got to go on to another 22 point. 23 MR. KWALWASSER: Sure. And quote exhaustion, where 24 there's 1000 shares offered and ten 1000 shares come in, once 25 that first 1000 shares is used up and was traded at the 1 market, the next 9000 shares would be -- if they traded at a 2 different price, would be considered traded outside the 3 market, even though the market was only offering 1000 shares 4 at the time they came in. 5 MR. COLBY: Just to deal with some terminology, 6 could I ask Maureen or Joel or both, is there a difference 7 between payment for order flow and internalization, from an 8 economic standpoint or... 9 MS. O'HARA: I think, just from the academic usage 10 of it, would be that, you know, payment for order flow would 11 be external to the firm, obviously; internalization would be 12 the same. It should be the same in terms of economic impact, 13 would be my response to that. 14 MR. HASBROUCK: Yes. By -- presumably, if a broker 15 were to acquire a dealer, what was formerly -- what was 16 formerly payment for order flow could be internalized. So 17 the distinction is really one of, you know, organizational 18 form, rather than substantive economic. 19 MS. O'HARA: Right. 20 MR. COLBY: There are -- I'm sorry, Rick, did -- 21 were you signaling? No? 22 MR. KETCHUM: No. 23 MR. COLBY: Well, then, Rick, could I ask you, is 24 -- what are -- what are the... 25 (Laughter.) 1 MR. COLBY: ...what are the arguments in favor of 2 -- internalization and payment for order flow have clear and 3 explicit conflicts of interest; right? The -- the broker's 4 making a choice of where to route the customer, the order, 5 and getting some sort of economic remuneration. 6 Two part question. First is, would you or anyone 7 else like to explain what the -- what the basis, what the -- 8 the benefits are? And then, second... 9 MR. KETCHUM: I'd be glad to start. 10 MR. COLBY: ...and then, the second part of the 11 question is: Should there be more done to address the 12 conflicts of interest that are inherent in that function? 13 MR. KETCHUM: Yeah, I -- in a lot of ways, it seems 14 to me this is a -- a little bit like Bernie, this is a 15 particularly curious time to be terribly concerned about the 16 conflict of interest built into internalization and -- and 17 routing orders, whether it be for payment for order flow or 18 other incentives to dealers that provide executions or -- 19 but, having said -- for exactly the reasons Bernie said, that 20 the spreads are narrower, orders can participate and -- and 21 create the quote. And decimalization has not only 22 dramatically changed those spreads, but dramatically changed 23 the effect of spreads, as well. 24 If I were particularly within that -- that set of 25 facts situations, recognizing that those Commission concerns 1 over -- over internalization and payment for order flow, in 2 different fact situations, had, I think, greater merit and 3 strength, and it's great to hear Ed say the things about 4 11Ac1-5. I can remember arguing them on the other side 5 before. I now find those statistics much more compelling 6 than I used to. 7 (Laughter.) 8 MR. KETCHUM: But to -- in those fact situations, 9 with those facts, our primary concern is that somebody's 10 still willing to trade with lumpy order flow on a regular 11 basis. And I think a great deal more attention should be 12 focused, not on the folks that can manage their own order 13 flow, because they do have a variety of exceptional tools to 14 be able to determine where they -- they're most likely to get 15 quality executions or not. Incidentally, usually those tools 16 are telling them where they may be able to find depth that 17 isn't shown in the marketplace. The only way a retailer and 18 investor realistically is going to find depth that he doesn't 19 find in the marketplace is through guarantees of a market 20 maker, whether that market maker be upstairs or be a 21 specialist. 22 And I think in this environment, the huge advantage 23 of internalization is that somebody's willing to -- to 24 provide guaranteed executions for retail investors. If you 25 look at the rest of the world, in an environment with much 1 wider spreads than in the United States, what you find is 2 exactly the opposite. In places like Germany they're trying 3 to figure out how to reincorporate sponsorship into the 4 market, so somebody's willing to stand there and provide 5 retail orders an execution. In the United Kingdom, they 6 solve it by having that function totally separate and totally 7 unprotected and totally uninteracted with regard to the 8 institutional order flow with respect to it. 9 I think right now the Commission should be focused 10 on finding as many ways as possible to give people incentives 11 to trade against small retail order flow, remembering that in 12 -- in better times that order flow gets very lumpy in the 13 millions of shares. And I -- I think the -- if there is a 14 concern of conflicts and the legitimate concern of orders 15 meeting orders, which of course, Ed, is only if it doesn't 16 interfere with best execution and efficient execution 17 transactions, it would be that -- that you need an 18 environment where limit orders are incentivized to show up in 19 the marketplace. I'd suggest that the best way to incent 20 limit orders to be in the marketplace is to have them 21 automatically executable. And if you have that environment, 22 they get taken care of very quickly and there's plenty of 23 incentive to show, and that's why people use them today. 24 MR. COLBY: Okay, Larry, did you want to address 25 this point, too? 1 MR. LEIBOWITZ: Sure. Just to -- to sort of add to 2 what Rich said, I think Ed's point about, you know, the fifth 3 tenet of the National Market System being violated here, 4 investor opportunity be executed without the interplay of a 5 dealer, I think what we really discussed was that those five 6 tenets can cross against each other. They're not always 7 mutually exclusive. The reason that internalization is good 8 for customers is it helps them get the best execution 9 possible. If you look at our numbers, in particular, and 10 Bernie and other people who execute in this way, more often 11 than not we execute well inside the spread. The reality is, 12 what we're doing is providing better executions for customers 13 and faster executions for customers when compared to a lot of 14 the other marketplaces. And so, what it does for us is it 15 allows us to provide more consistent executions for 16 customers. 17 I think in the old days these -- these execution 18 destinations were huge profit centers for -- for the dealers. 19 But in this day of transparency, with the SEC rules, 20 transparency and disclosure, if there is payment for flow, 21 and -- and full disclosure of all of these things, what it 22 really is for us is, is it's a way for us to provide the best 23 possible execution, better than you can see on your screen, 24 for the customer. 25 MR. COLBY: Ed? 1 MR. KWALWASSER: That might be a good argument if 2 most of the internalization and payment for order flow was in 3 the bottom 2000 stocks that traded both either in NASDAQ or 4 in our market. But, in fact, most of the payment in order 5 flow and internalization are in the top 500 or so stocks. 6 And those stocks trade very easily in the market, 7 particularly for retail order flow. So I don't think that 8 taking those shares out of the market is really a benefit to 9 the market or a benefit to the customer. 10 MR. COLBY: Gus? 11 MR. SAUTER: Yeah. I'd like to take a little bit 12 of an opposing view here. I'd like to draw a distinction 13 between payment for order flow and internalization. I think 14 that in Bernie's old case he used to pay for order flow, and 15 he did internalize. You can internalize without paying for 16 order flow just the same. 17 I personally don't have a problem with payment for 18 order flow if the investor can protect themselves. In our 19 case, we take on the fiduciary responsibility of protecting 20 ourselves and making sure that we're going to the venue that 21 is providing us best execution. And if that venue wants to 22 also pay us to execute there, then we're even that much 23 better off. 24 I will take a strong exception to internalization, 25 though. There we believe that internalization is very 1 destructive. And I would define it more broadly than we've 2 defined it so far. I would not say it's just necessarily 3 crossing internally what happens to come across Schwab's own 4 book. I would say that ECNs are internalization. In other 5 words, trading gets trapped inside of a -- a market, and it 6 -- it trades there before going outside. So whether it's in 7 one broker's book before going to the outside, or whether 8 it's in an ECN, or, quite honestly, whether it's in a primary 9 market, it's isolated in that marketplace before it goes 10 outside. 11 Maureen talked earlier about what is best 12 execution. And what may be best execution for an order may 13 not be best execution for the market. And I think that's a 14 -- an extremely important point. Where we talk about, well, 15 this order went in and they -- they were executed internally 16 at no worse price than they would have gotten if they had 17 gone outside. Who is hurt in that situation? It's the -- 18 it's the limit order on the outside. 19 Larry asked me earlier if we placed limit orders. 20 Ed talked about how I mentioned sweeping the book. I -- I 21 think the best market is one that has a tremendous number of 22 limit orders. Best execution is really in a market that has 23 penny-wide spreads, infinitely deep on both sides. So how do 24 we get to that marketplace? I think we need to incent limit 25 orders, and I think internalization does not. Because the 1 best bid or offer out there can be matched internally, 2 without ever taking any risk of providing a limit order. By 3 providing a limit order, you're -- you're providing 4 transparency, you're giving a tremendous amount of 5 information to the marketplace, you're providing for price 6 discovery, you're also providing a free option to the 7 marketplace. Someone can execute just a penny above your 8 order, and know that you're a stop-loss for them. They can 9 always turn around and sell it to you, because you -- you're 10 there with your limit order. 11 So I think we need to have protections for limit 12 orders. I think internalization goes against that. Now, 13 internalization inside the spread does not go against that. 14 But internalization at the NBBO does hurt the limit order. 15 And to us, what happens is, it makes us not place limit 16 orders. There's a disincentive to place limit orders, 17 because you're just providing free information to people, and 18 -- and really, the time you get executed is when you don't 19 want to be executed. 20 MR. COLBY: Ed Nicoll, you run two limit order 21 markets, one of which is an island, and so it -- it doesn't 22 route out, as Gus had talked about. Is there more that 23 should be done in order to encourage limit orders to be 24 placed in the market? 25 MR. NICOLL: Well, yeah, I think I agree that if -- 1 that one of the focuses of -- of the SEC should be to try and 2 figure out a market structure and a set of incentives that, 3 to the greatest extent possible, gets people to reveal their 4 -- their intentions to buy or sell. And when they place a 5 limit order, they are putting themselves at a disadvantage to 6 somebody who doesn't place a limit order. And there ought to 7 be incentives to the person that is giving out a free option 8 to somebody else. 9 This gets back to -- interestingly, to the issue of 10 rebating liquidity providers. One of the benefits of 11 offering somebody who places a limit order a rebate, and 12 charging the person who takes that limit order, is you are, 13 in effect, paying the liquidity provider, the person who 14 places a limit order, for the free option that he -- he or 15 she is giving to the person who takes that limit order. And 16 right now you have, through -- I think through competition, 17 mostly through competition in the NASDAQ marketplace, we seem 18 to have settled on a rate structure which provides "X" level 19 of rebate to -- to the liquidity providers, and "X" level of 20 charge to the people who take the -- that liquidity. And I 21 do think that reflects the dynamics of the market forces at 22 work here. 23 I would, you know, obviously strongly disagree that 24 one can view an ECN as a form of internalization, unless we 25 -- we went to some sort of system of a national consolidated 1 limit order book. And that kind of -- which -- which Gus 2 started out advocating. In that kind of a super regulatory 3 environment, then one could view any interaction outside of 4 the national -- as being internalization. 5 But, short of that, ECNs, especially ECNs which 6 have significant markets, are no more internalizations than 7 any other destination execution venue like the New York Stock 8 Exchange or the American Stock Exchange, or any of the 9 regional stock exchanges. 10 MR. COLBY: Unbeknownst probably to some of us, Ed 11 just took off his red cape and waved it in front of the faces 12 of a number of people on the panel. I'm going to hold the 13 access fee discussion for a little bit, if we could. 14 (Laughter.) 15 MR. COLBY: Bernie, are you doing access fees or... 16 MR. MADOFF: It's not on access fees. 17 MR. COLBY: Okay, go ahead. 18 MR. MADOFF: Earlier, Chairman Pitt asked a simple 19 question regarding market data, and I responded by saying -- 20 by answering, "That's the way it was." And he was trying to 21 find out why do we continue to -- to use this process. Maybe 22 we should look at other -- you know, other methods. 23 And my own view is that just because that's the way 24 it was in the past does not necessarily mean that's the way 25 it should go forward in the future. And that -- that could 1 speak to market data, it could speak to the way transactions 2 are handled in the marketplace today. 3 I think the challenge, you know, for the 4 marketplace today is to address the issues of how do you 5 allow firms to do their business in the most efficient 6 manner, possibly, and capture the greatest amount of revenue 7 from that business, while providing the best execution for -- 8 for the investor. And the concerns that people have with 9 internalization have always been are you fragmenting the 10 markets, and are you not allowing market orders to interact 11 with limit orders. And that clearly, to all of us who are 12 sitting around this table, because they're all familiar 13 faces, you know, to -- to me, have been struggling with these 14 issues for a lot of years (sic). And we -- our firm, as well 15 as, I think, many others, have said, okay, that is the 16 challenge in the future. How do we allow us to do our 17 business profitably, give investors the best -- the best 18 service and execution quality, and allow orders to interact, 19 both market orders and limit orders? 20 Now, we came up with a -- with a system, and joined 21 in partnership with -- with four of the largest firms that 22 exist today in -- in the United States and NASDAQ, to build a 23 system that -- that achieves that. I won't mention the name 24 of the system because this is not to be a marketing effort. 25 MR. COLBY: Primex System. 1 MR. MADOFF: But -- but, basically, the concern 2 that all of these firms had was that they wanted to 3 internalize their order flow, but they all very fully 4 realized that the danger of having these various pockets of 5 liquidity not visible to everybody and not allowing orders to 6 interact, this was a problem unto itself. And even though 7 that may be what a lot of people would like to do, if you -- 8 if you degrade the market in such a way that you really don't 9 do what you're supposed to do for the investor, then you 10 really have no marketplace, anyhow. 11 So we went about building this type of system that 12 would allow firms to internalize their order flow, allow 13 orders to interact with limit orders, and to do it 14 anonymously. Because one of the problems that you have with 15 transparency today is because the technology exists that 16 allows people to interact with limit orders very, very 17 quickly, you have a practice where you have lots of people 18 sitting up in -- in trading rooms all over the world seeing 19 the same news at the same time that can get into the market 20 much faster than certainly the average investor can get into 21 the market, and you have everybody stripping limit orders, 22 and you have the limit orders buying stock at a time when 23 really is not, you know, the opportune time to buy it. 24 Now, people would say, well, if you put a price on 25 an order and you -- you displayed it, you really wanted to 1 buy that stock. But that's the way it works in theory. The 2 reality of it is that a lot of times, after people buy the 3 stock, if they're buying into a declining market, they're not 4 happy that they're buying that stock. 5 So the problems that we all have in market 6 structure is to somehow or other develop a system that 7 accomplishes all the goals that -- that I tried to address. 8 That is not an easy situation, but it is doable. And I think 9 that's what I think everybody has to focus on. I think that 10 these are not problems that are not, you know, capable of 11 being solved. Everyone has to work at them. But there are 12 systems out there that will accomplish them. 13 And I think that just because people are losing 14 market share or because things are happening differently than 15 they've done in the past, that's no reason for people to, you 16 know, have any great alarm and say that we have to now, you 17 know, ban certain practices, whether they be internalization, 18 payment for order flow, or any other new device that comes 19 along. 20 MR. COLBY: All right, Dave, Bernie might say that 21 you want to strip your operation. What do you -- what do you 22 think of this -- what needs to be done to encourage limit 23 order trading? 24 MR. WHITCOMB: Yeah. Actually, quite the opposite. 25 It's possible that I used to, but I don't anymore. 1 (Laughter.) 2 MR. WHITCOMB: Basically, we -- we are among those 3 that Ed was talking about, that provide limit orders to the 4 market through -- through ECNs. But I think the -- the thing 5 that I heard today that made me the happiest is the comments 6 in favor of incentivizing limit orders. Because, indeed, if 7 you incentivize limit orders, you will narrow the spread. We 8 saw that happening mainly through ECNs in the NASDAQ market, 9 and we saw how the spread collapsed. And then, if the best 10 execution obligation takes place at the NBBO, the -- the 11 retail market order trader is not disadvantaged even by 12 practices of -- of internalization and capture, if not 13 payment for order flow. 14 What we need to do is to make sure that we do not - 15 - not only that we incentivize limit orders, but that we 16 don't create disincentives for limit orders. 17 There are a number of potential disincentives, 18 either actual or ones that are being discussed. For example, 19 charges for entering and canceling limit orders that exceed 20 the marginal cost of band width, of adding band width, have 21 the very serious consequence of disincentivizing limit 22 orders. And that would unravel the desirable changes that 23 have been made in the marketplace that would allow the 24 spreads to widen back up and the liquidity to diminish. Some 25 practices in the listed market where specialists have an 1 ability to delay the posting of limit orders and to delay the 2 execution of market orders against limit orders that have 3 come into their book. Those are actual practices that tend 4 to disincentivize limit orders in those markets. And one 5 reason you haven't seen the spread collapsing in the listed 6 market to the extent that it has in the NASDAQ market is 7 precisely because of those disincentives -- difficulty of 8 canceling, the long time it takes to cancel, and again, ten 9 seconds is an eternity if you're an active limit order trader 10 trying to protect yourself against what we academics call 11 adverse selection, trading only when you don't want to. 12 We really have to take a close look at how the 13 market system can be made real by incentivizing limit orders 14 to provide the liquidity and the narrow spread that will then 15 give us best execution. 16 MR. COLBY: Joel, you had indicated -- do you still 17 want to speak? 18 MR. HASBROUCK: Thank you, yes. 19 The best internalization there -- well, to the 20 principle that you want to encourage direct interaction of 21 customers, I can think of two concerns here, really only one 22 of which has been addressed to this point, and that is if you 23 allow dealers to internalize, presumably they will be in a 24 position to take advantage of the customers. Now this has 25 arisen in these discussions and it's clear that the SEC and 1 other regulatory organizations have safeguards in place. But 2 there's another effect; namely, although we have not reached 3 this point, there is a chance that enough internalization 4 would divert the orders away from the primary market, thereby 5 making the primary market thinner, presumably less 6 informative and perhaps, going down the road, more subject to 7 manipulation. And this is one of those externalities, I 8 believe, that may want us to, and take steps to. encourage 9 some kind of consolidation. 10 MR. COLBY: I'm going to point the next question 11 towards Rick because he operates a distributed multi-market 12 center, but I'd like to ask, given that Bernie and Gus and 13 Wayne and Joel and a number of others have said there is this 14 tension between an efficiency of a dealer executing the 15 orders internally and the interaction between customer 16 orders, I ask you how severe are the current problems and 17 what are the trend lines. Think about that one. 18 Rick, do you want to address the question about 19 what happens if there's so much internalization and will it 20 affect the primary market, given that NASDAQ's primary market 21 is a distributed market? 22 VOICE: I think the answer is it won't. 23 MR. KETCHUM: There -- it's going both ways. It's 24 always good to have an intermediary near you. 25 (Laughter.) 1 MR. KETCHUM: It won't if the orders are easily 2 accessible. If you're in an environment in which orders 3 can't be represented into the market, which the Commission 4 saw with the order handling rules; if you're in an 5 environment where you can't immediately reach out and access 6 those orders, then you do have a concern as to whether or not 7 those orders will need to be somehow consolidated in a way in 8 which you can find them and in which they can interact. 9 On the other hand, what we've seen happen in NASDAQ 10 once the order handling rules both required orders to be 11 displayed at the best price and the order handling rules 12 encouraged at least in the NASDAQ standpoint, a determination 13 to link all of the liquidity providers together; i.e., 14 marketmakers and ECNs, is that you get a rough breakout 15 between those who want to use -- who want to access liquidity 16 and get immediacy and those who want to use limit orders and 17 they choose to do that through ECNs and now in the future 18 they have a choice of choosing to do it through ECNs, through 19 marketmakers or through SuperMontage, with a wide degree of 20 transparency and being able to see where those orders are 21 residing. 22 So I guess my answer would be as long as you can 23 see the orders, particularly if you can see them in depth, 24 recognizing we're now a decimal although there isn't so much 25 in one place, as long as you can immediately access those 1 orders, you don't have a centralization problem and within 2 that environment, you are going to have very narrow spreads 3 and what you're going to want to do is encourage people to 4 trade against those retail orders to make sure they get 5 executions. 6 So I'll stay away from determining what that leads 7 you to conclude in a different floor-based environment where 8 you have different dynamics, but in the NASDAQ environment, I 9 don't think those are concerns any more in the same way, 10 because I think the conflicts and the accessibility has 11 changed dramatically. 12 MR. COLBY: Do we have a centralization problem in 13 the listed market and in the NASDAQ market? Would anyone 14 like to address that? Bernie. 15 MR. MADOFF: It's -- certainly I don't think you 16 have that problem in the listed market at this stage and I 17 think the way you -- because you still have the New York 18 doing 85 or some odd percent of the business -- so I think 19 the competition that the markets, whether they be third 20 markets like ourselves or the regional exchanges, what they 21 brought to the table is the competitive pressures that the 22 New York has responded very well to in lowering their costs 23 for their members, to prevent, you know, people like 24 ourselves from capturing even more order flow than we have. 25 So the competitive pressures I think benefit everyone. 1 I think that Rick is absolutely correct in that the 2 key to the future as to whether or not you have problems with 3 the listed market being fragmented is if in fact you have 4 linkages. And although everybody blames ITS -- if people 5 that don't know better look at ITS, you think it's this 6 archaic sort of like Rube Goldberg type of structure that 7 nobody can possibly get an order through. We've been 8 operating in ITS since its inception and I can tell you we do 9 huge amounts of business, more than most firms do in their 10 various other models, all through ITS. 11 Now there's no problem with the technology of ITS 12 as a linkage, the problem is with the rule structures of ITS. 13 So as long as you have linkages, whether they be ITS or 14 whether they be linkages like SuperMontage that allows orders 15 to interact and for dealers to be able to access quotes on an 16 immediate basis, not necessarily automatic, because although 17 there are marketplaces like ourselves and like NASDAQ and 18 some of the regional exchange that do give automatic 19 execution, there are marketplaces that choose not to give 20 automatic execution like an ECN, but they do for all intents 21 and purposes give an immediate execution. 22 All I care about is that if I see a quote 23 somewhere, whether it be on the floor of the New York Stock 24 Exchange or whether it be in NASDAQ or whether it be in ECN, 25 that I can get an immediate response to my indication that I 1 wanted to trade. Now it doesn't have to be automatic, I'm 2 perfectly happy to send an order to an ECN which does not 3 have an automatic execution and get an immediate response. I 4 am not happy, as are others, if I send an order down to a 5 floor that chooses to expose that order to a crowd, because 6 that is their business model, and says that they're doing 7 that in order to let that order interact with hidden orders 8 that are in the crowd. If they want to do that, that is 9 perfectly fine, providing the person that's sending the order 10 to that floor allows them to do that. If I am perfectly 11 happy to just have my orders interact with the best bid and 12 offer that is displayed in that marketplace, then I should 13 have the right to have my order treated that way and get an 14 immediate execution, if not an automatic execution. 15 And that I think is the rub that everybody is 16 dealing with with the floor-based environments. 17 MR. COLBY: Well, Bernie and Rick would take us to 18 the access issues and linkage issues. But before we get 19 there, does anyone want to speak to whether we have a 20 centralization problem, listed or NASDAQ? Ed. 21 MR. KWALWASSER: Yeah, I think it's not so much 22 whether we have a centralization problem, as Bernie 23 accurately reports, we do about 84 percent of all of the 24 volume in our stocks. What I think is troubling, 25 particularly with internalization, is that most of the stock 1 that's internalized are retail orders that have less 2 information and that makes it more difficult in the whole 3 market and it also puts a pale I think over whether the 4 market is treating everybody fairly and whether all people 5 are given the chance to get the best price, because you don't 6 see institutional orders being internalized by marketmakers 7 and that should say something to retail customers whose 8 orders are dealt with that way. 9 MR. COLBY: Gus. 10 MR. SAUTER: A lot of comments really refer to 11 whether or not the market order is getting best execution. I 12 personally do think there is a centralization problem. If 13 you're placing limit orders and you place your limit order on 14 New York, you have access to one liquidity pool. If you 15 place that same limit order in Chicago, it could be at the 16 exact same price and the stock can be trading at that same 17 price all day long and you have no access to trading, even 18 though you're still providing a limit order that actually is 19 support for the New York. There's exchange arbitrage that 20 can happen very simply. The same is true with an ECN, if you 21 just happen to pick the wrong ECN to place your limit order 22 on, the stock can trade all day long around you. 23 So I think there is a centralization problem and 24 this is what leads us to conclude that a central limit order 25 book is the optimal solution. In that environment, then the 1 exchanges become entry points into the central limit order 2 book and have to compete for order flow based on services 3 they provide giving access to the central limit order book. 4 MR. COLBY: We're going to pick up the pace. 5 David, one minute -- did you want to speak? 6 Brett, one minute. 7 MR. REDFEARN: I think in the listed market, there 8 really isn't in our equities business any more, we're doing 9 82 percent or so of the trades and a substantial amount of 10 the volume as well. I think where you have the 11 centralization problem is in the ETF business, and the ETF 12 business is very different. And one thing here and perfect 13 for the discussion of market structure is there are so many 14 different market structures operating in the ETF world, aside 15 from our floor where we have brokers and specialists and 16 multiple marketmakers, New York is trading them with brokers 17 and specialists. And you have Island, what Island has been 18 doing in the marketplace, which is very much not linked into 19 ITS and not accessible and not to belabor the point which has 20 been raised earlier today, and then you have upstairs markets 21 where you have a lot of the firms who are making upstairs 22 markets and operating sort of outside of the ITS/CAES 23 framework. 24 So in that particular area, there is I think a 25 substantial lack of centralization and a substantial amount 1 of fragmentation in that marketplace and I think that's one 2 area where centralization and linkage and ultimately how that 3 market works most efficiently really does need to be 4 addressed. 5 MR. COLBY: Commissioner. 6 COMMISSIONER GOLDSCHMID: I have a question. 7 Stepping back, how efficient are our markets compared to 8 other markets today around the world? 9 MR. COLBY: Wayne, do you want to take that? 10 MR. WAGNER: Yes. From our data, our markets are 11 the lowest cost markets that we measure, by a significant 12 amount. The only one that comes close is the U.K. market and 13 they go up from there. And that's including not only the 14 cost of executing in it, but the cost of getting to the 15 marketplace and getting the orders to the size that they can 16 be executed. 17 COMMISSIONER GOLDSCHMID: Thank you. 18 MR. COLBY: Bernie. 19 MR. MADOFF: I'd just like to respond to that as 20 well. 21 We actively participate in the global marketplace 22 and have for many years. I don't think there's any 23 comparison between the efficiencies and the core structure 24 both in the execution and in the settlement side of the U.S. 25 markets compared to any marketplace that exists. They're the 1 envy -- as a matter of fact, most of the world markets have 2 tried to copy the dealer market environment that NASDAQ has 3 created over the years. Some have been able to do it better 4 than others. 5 I just want to correct one issue that my associate 6 Ed Kwalwasser said, where he said that large orders are not 7 internalized. I would venture to say that most people would 8 say that 90 some odd percent of the block trades that are 9 done in listed securities are internalized in one form or 10 another off the floor of the exchange, they just get executed 11 on the floor, they get printed on the floor. 12 So to say that when somebody wants to sell 100,000 13 or 200,000 shares of IBM and that trade is sent up to the 14 block desk of Merrill Lynch and they, you know, internalize 15 part of it and cross the other, I would say that that is an 16 internalized order by anyone's imagination. 17 MR. COLBY: No cross rebuttals here. Sorry, we're 18 running out of time. 19 A question from this side of the table about 20 internalization and payment for order flow. To what extent 21 would conflicts be better resolved if internalization had to 22 take place at a price better than NBBO or if -- and/if 23 payment for order flow would have to flow back to the 24 ultimate customer? 25 MS. O'HARA: I think that certainly goes a long way 1 to dealing with some of the concerns. I mean certainly my 2 main concern when we deal with internalization is that I do 3 think it degrades the overall market. So a rule that you 4 have to better the market does an awful lot to deal with that 5 concern. It doesn't totally deal with the concerns that I 6 think are quite legitimate, that to the extent that the 7 orders are not being out there and serving as a mechanism to 8 induce limit orders, that you're sort of getting better but 9 maybe not best. You don't know what you could have gotten in 10 the absence of a rule that said, you know, you have to make 11 everything visible, but as Joel pointed out earlier today, 12 you can never make, you know, visible orders the rule. 13 People will or will not put them in. So I would find it a 14 lot better if the rules were changed such that you had to 15 improve. That would certainly go a long way to getting rid 16 of the concerns that I have. 17 MR. COLBY: Larry. 18 MR. LEIBOWITZ: Yeah, I think in practicality, 19 order-by-order improvement is very difficult, particularly 20 given the way the NBBO flips around the way it does today. 21 The reality is if you look at the statistics of the 22 internalizers, they do improve on average. And so the 23 question is well, what is the burden of enforcement on an 24 order-by-order basis. I think that that would be a real 25 problem. The second thing is it doesn't reflect the fact 1 that a lot of times internalizers are enhancing liquidity. So 2 putting up more volume applies it at the inside market. 3 I think one of the things that we saw when we went 4 to a decimal world is that we have substantially less 5 liquidity at the inside of the market. A lot of times the 6 market is being set by an ECN for 100 or 200 shares or 7 something like that and the reality is that that doesn't 8 serve the investor best. 9 MR. COLBY: Joel. 10 MR. HASBROUCK: Lawrence, you mentioned the problem 11 of flipping the "back and forth" as the instability of 12 flickering bids and offers or flickering limit orders, but 13 presumably that's already being monitored closely enough to 14 prevent trade throughs. So presumably there would be no 15 barrier to adding say an additional premium on top of those 16 prices to better the execution. 17 MR. KWALWASSER: I think that that's better, but 18 it's not good enough. I think that customer order 19 interaction is what Congress said, it's what we all agreed 20 and what I actually think is the best thing for customers, 21 because if you get one cent better, maybe the market would 22 have given you two cents better. If you got two cents 23 better, maybe the market would have given you three cents 24 better. And it's impossible to know until the orders 25 interact. And it also creates a better marketplace because 1 trading begets trading and liquidity begets liquidity and if 2 people see that orders are interacting, they will send more 3 orders to the marketplace. 4 MR. COLBY: Unless there's other comments on this 5 topic, let's -- Rick and Bernie, shortly before this part of 6 the conversation said that they thought that the problem was 7 access and that if you had good access to prices, that 8 centralization would be largely mitigated. 9 That comes in two forms, if you will. One form of 10 access is access directly from the broker or from the 11 institution to a market center and then the other form is 12 access across competing market centers. 13 Let's start with the first. To what extent are 14 there problems in access to individual market centers today? 15 Now let me put a little color on this. One thing that we've 16 historically observed is that with some notable exceptions, 17 which I think New York is one, there seems to be a tendency 18 that the bigger you get, the less you want to trade with 19 informed order flow, that there's access to uninformed order 20 flow quite freely but informed order flow is not as welcome. 21 What sort of difficulty -- how many problems are 22 there with access from a broker, say handling customer orders 23 to the many market centers that may be trading a particular 24 stock? Have there been problems? 25 MR. REDFEARN: I mean I can jump in on this because 1 I mean we have been subject to a certain amount of criticism 2 about access to the AMEX market in the context of our 3 participating in the NASDAQ stock trading business. And it's 4 a very interesting one because now it's definitely overlapped 5 with the issue of best execution where increasingly markets 6 have sort of indicated that perhaps you don't have to pay 7 attention to our quotes in the concept of best execution in 8 terms of the difficulty of accessing the market. 9 Just in terms of the context of our background on 10 this, I mean, you know, there's a few things that are 11 critical to keep in mind in the history here. The first is 12 that as soon as the AMEX entered into the discussion about 13 trading in the NASDAQ world -- 14 MR. COLBY: Can I interrupt? Let me frame this a 15 little bit more for the people that aren't familiar with it. 16 The criticisms are that the NASDAQ market -- as new exchanges 17 began trading NASDAQ securities, they entered a world in 18 which NASDAQ had -- was operating an automatic execution 19 system. 20 MR. REDFEARN: Correct. 21 MR. COLBY: Whereas, some markets, including the 22 AMEX had an interaction model that was not automatic 23 execution, it involved a manual component. 24 MR. REDFEARN: This is correct. 25 MR. COLBY: Okay. 1 MR. REDFEARN: Okay, so I mean there are two sides 2 of that coin. The first one is simply access to the market. 3 And the way we look at this is that when we first came to the 4 table, we tried to put a discussion on the table of the UTP 5 operating committee and said let's discuss intermarket 6 linkage, and basically -- 7 MR. COLBY: No histories, let's just talk about the 8 issue, okay? 9 MR. REDFEARN: Well, I mean the bottom line is we 10 were rejected in several attempts to try to figure out -- 11 MR. COLBY: But access, is there a problem with 12 access to markets today? You explained how you got to where 13 you were, but is there a problem today? I'm sorry, I'm just 14 being ruthless because I've got 15 minutes. 15 MR. REDFEARN: No, that's okay. The way I look at 16 it is to a certain extent there's a linkage lag. We talked 17 about -- for over a year about how people could access our 18 marketplace and there was a lag in terms of a lot of vendors 19 and a lot of people establishing connectivity to our market. 20 There were a number of people who could do it right away and 21 they were sending us business on day one and it wasn't a 22 problem. There were other people who said we've got some 23 work to do, we need to figure out how to get into the AMEX 24 market, and so they did a lot of work to try to set that up. 25 And some of them were very close, I mean we have BRASS and 1 UMA and other vendors out there that say they're within two 2 weeks now of actually having those lines in place. And so we 3 think that as that has happened, there's sort of a catch up 4 in terms of the electronic access in the market. That's the 5 first part of this thing. 6 The second part is that increasingly, access is 7 being equated with the issue of automatic execution, and you 8 know, I have a problem with saying that access equals auto-ex 9 because ultimately in the access issue, we have not tried to 10 say when we entered into the NASDAQ marketplace that we are 11 going to be the fastest place to execute an order for 1000 12 shares, you know, if that's what you want to do. 13 However, we have recognized some of the problems in 14 the NASDAQ market. The size has dried up, it's difficult to 15 see where you can get size orders done within a decimalized 16 environment, so our offering is really in a slightly 17 different model. We're willing to say we're going to have 18 orders out there for 10,000 shares, 20,000 shares or up and 19 if people want to try to find liquidity again in that market, 20 that's something that we think that we can offer to the 21 marketplace. 22 MR. COLBY: Well, that's interesting. 23 David was up first, I think. I see four others. 24 MR. WHITCOMB: There is a problem with -- 25 MR. COLBY: Let me just say we ought to talk about 1 whether there's a problem with access directly to markets, 2 all right? 3 MR. WHITCOMB: Yeah. There is a problem with 4 access. It's not a matter of lack of lines, it's the fact 5 that your limit orders cannot get to the market 6 instantaneously and be displayed instantaneously and included 7 in the NBBO instantaneously. You can send a limit order to a 8 listed market and it won't affect the NBBO because the 9 specialist has discretion whether to display it, even if it's 10 a good sized order that's inside. 11 There's a problem with access of market orders. 12 Market orders coming in through certain systems do not 13 necessarily interact with the best limit orders that exist in 14 those marketplaces. That's the access problem. 15 MR. COLBY: Rick. 16 MR. KETCHUM: There are essentially two different 17 types of access and Brett and David have come to different 18 conclusions on one of them, which is the question of a firm 19 that chooses, with respect to whatever order flow it has, to 20 send some of it down to the floor of an exchange. One can 21 argue one way or another on whether that access is optimal 22 but not operating a floor exchange, I'm going to stay out of 23 that one. 24 But I think Brett misses the whole point of where 25 the problem comes when you move that type of environment into 1 an environment that is predominantly characterized by 2 automatic executions. And that is the fact that most of the 3 firms have chosen not to send orders down to the American 4 Stock Exchange. Instead they are executing those orders with 5 their own guarantees against the best bid and offer upstairs. 6 That creates a very difficult and different problem because 7 we have a tendency in thinking of best execution to think of 8 it in freeze frame and think of it as one order with respect 9 to two quotes that don't change. But of course, that never 10 was true and in the world of decimalization, it's completely 11 false. 12 The reality with respect to a firm trying to 13 execute orders upstairs at the present time is that it has to 14 deal with a constantly changing quote and if one of those 15 sides of the quote does not change as quickly. Not assuming 16 whether it's better or worse, but assuming -- even assuming 17 on average 50 percent of the time it's wrong, although 18 usually when it's a cross market, it's the side that crosses 19 that's usually right. 20 But that marketmaker has a problem in executing his 21 customers' orders because the offer has just moved to that 22 bid by four or five cents, even if an order was sent down to 23 it waiting for the fairly interminable time -- seconds, 24 minutes -- before it gets executed. Then that marketmaker 25 has to decide, not with one order but tens of orders, 1 hundreds of orders, however many orders may come in during 2 that time period, what to do. He can do one of two things. 3 He can stop executing, that will ensure that he has not 4 violated any best execution problems, it will just simply 5 completely frustrate what retail investors generally want, 6 which is to lock in a price and to get a price with respect 7 to their market orders. 8 Or he can continue to execute. Then he has to 9 decide one of two places to execute against. He can execute 10 against the crossed bid, which experience would tell you most 11 of the time no longer reflects the market and that the offer 12 that has crossed it now does reflect the market. He can do 13 that and provide a subsidy to his customers, which is good 14 for as long as he does it, but he'll probably stop doing it 15 fairly quickly, at which point you figure out somebody else 16 to provide liquidity. Or he can do that in what he believes 17 as best he can imagine is the new best market, which reflects 18 the primary market and the way that marketplace has worked 19 with respect to every one operating on automatic execution, 20 which would include everybody in NASDAQ, everybody who 21 operates out of an ECN where again they're accessible from an 22 automated basis or not. 23 That would make a lot of sense in a lot of ways, 24 but might raise best execution problems. All of that is what 25 created the mess when AMEX entered, not that they didn't 1 provide, for somebody who freely chose to send their orders 2 there, an accessible marketplace. I'm sure that for those 3 that chose, they did. 4 And that is the problem that I focus the Commission 5 on and the difficulty that needs to be defined, that 6 marketmaker that has to execute tens of orders while that 7 market is crossed, while that market is moving with a 8 marketplace that cannot be reached immediately. And that's 9 why we suggest that you have to move -- at least when the 10 primary market provides automated executions, markets should 11 provide automated executions with respect to other 12 professionals trying to access and impact the price. If not, 13 you create chaos. 14 MR. COLBY: Let me just point out that what you're 15 saying is this is not technically a question of access, it's 16 a question of the coexistence of two different market trading 17 systems, trading the same sort of security. 18 MR. KETCHUM: Yes, it is access between 19 participants in order to be able to impact and create price 20 formation in a reasonably effective way. 21 MR. COLBY: One could get the order there. The 22 problem is what happens once it gets there, correct? 23 MR. KETCHUM: It is problems of the time to get the 24 order there and when you have chosen a different way to 25 provide guaranteed executions and instead you're trying to 1 decide what price to provide those guaranteed executions at. 2 MR. COLBY: Ed, you've struggled with this 3 situation. Did you want to speak to this? 4 MR. NICOLL: I do, I do. And I want to do it in 5 the spirit of my opening statement, which is that what we 6 have here is a struggle between two competing types of order 7 execution. And just let me make sure everybody understands, 8 I mean what -- the argument that the floors have is that 9 we're better off waiting and taking some time for the price 10 discovery that occurs down on the floor and that a better 11 result will occur while the electronic markets take the 12 position that the immediacy that they offer by and large 13 provides better execution over the course of, you know, any 14 reasonable period of time. 15 You say it's not an issue of access, but it is 16 fundamentally an issue of access because the very model of 17 the floor creates a preference, a geographic preference for 18 the people standing in the crowd. And by definition, if you 19 are not standing in that crowd, you do not have access to 20 this basic, you know, superior execution model. 21 The floors can't have it both ways. They can't say 22 that by definition, you know, this is a superior execution 23 model and at the same time we provide universal access to us, 24 because by definition, they're saying that their model 25 requires exclusivity. I mean if there was true access to the 1 floor of the New York Stock Exchange, a price on the seat -- 2 and if ITS, as Bernie said, really worked, there'd be no 3 difference between a price of a seat of the Boston Stock 4 Exchange and a price of a seat on the New York Stock 5 Exchange. 6 And in terms of barriers to execution and barriers 7 to access, I can tell you what the barrier is to get on the 8 floor of the New York Stock Exchange, it's about two million 9 bucks, because if you're not a member, you're not standing in 10 that crowd. 11 Now I'm not saying that that's not -- that there's 12 not value creation there and I'm not saying that the members 13 of the New York Stock Exchange don't have valuable private 14 property expectations. But what I am saying is that we have 15 to find a way for this market to exist the way it has and 16 make its argument about the fact that this is a superior 17 execution methodology and the competing marketplaces which 18 take a different point of view. And it is about access, 19 because by definition, the very model that the New York Stock 20 Exchange practices means that if you don't have access to 21 that floor, then you're not getting as good an execution as 22 you could elsewhere, and ultimately if you're not standing in 23 that crowd, you're not getting as good an execution as you 24 would otherwise. So that's access. 25 MR. COLBY: Now Larry, I know you exposed your 1 interest in speaking, but we don't have time priority here 2 and I can't promise your interest is going to get executed 3 before this session. 4 I'm going to go to Ed. 5 MR. KWALWASSER: I think there are two things. 6 One, David, I don't know what country you're trading in, but 7 our specialists have the same obligations for getting limit 8 orders up as anybody else in any of the other markets, and in 9 fact they do. 10 MR. COLBY: Go ahead, Ed. 11 MR. KWALWASSER: And with respect to access, 12 everybody has access. You don't need to have a seat to have 13 direct access. Although we'd love to have Bernie as a member 14 of the New York Stock Exchange, he has direct access to the 15 floor of the Exchange and executes a lot of trades on the 16 floor. 17 Every specialist on every regional exchange that we 18 know can come through ITS, but most of the time or a lot of 19 the time, they come through links that they have through our 20 member firms. It's clear that membership has some privileges 21 and that's why there's some cost, but the members don't get 22 to trade ahead of non-members and member orders don't get to 23 trade ahead of non-member orders. All orders have the same 24 standing in our marketplace and if you want to come in 25 electronically, you can do that and you can use our NX 1 product; if you want to trade 1000 shares against our bid or 2 offer automatically, you can do that. If you want to send an 3 order to the crowd because you think you could get a better 4 execution, you can do that. 5 We're not trying to tell anybody how to set up 6 their market structure for their own market and we think we 7 have a pretty damned good market structure and I think our 8 customers are telling us that we do. And we don't think that 9 other people should be telling us how to set up our market 10 structure. They may decide not to come to us because of it, 11 but that's their privilege. 12 MR. COLBY: But ultimately, we're talking about how 13 different market structures interact. 14 MR. KWALWASSER: I can do that one, too. 15 MR. COLBY: I know you can, Ed, but Larry had 16 expressed an earlier interest. 17 MR. LEIBOWITZ: Just a quick point. We may not 18 tell you how to run your market, but we're the one on the 19 hook for best execution. Forgetting about the internalizer 20 for a second, when we make a choice of where to route an 21 order, if we then have to calculate into it that there's a 10 22 second delay or a 15 second delay, it may make a big 23 difference, and it's not a known solution, there are multiple 24 problems here and then when the regulators come in, they say 25 well what did you do here. It's a tough one for us to 1 justify. 2 MR. KWALWASSER: Yeah, but almost 85 percent of the 3 orders that come in through ITS get executed in our market. 4 The automated markets are only 15 percent of the orders that 5 get executed. So it's not only a question of time, it's a 6 question of when you send an order in, is it going to get 7 executed or is it not going to get executed. 8 MR. COLBY: Hang on. It's not a completely open 9 interaction. Gus expressed a -- 10 MR. SAUTER: Yeah, I would say there are electronic 11 linkages and accesses and an issue from that standpoint. But 12 there is a problem when you enter through one venue and you 13 end up trying to get to another venue that has a totally 14 different model. I would say that on the institutional side 15 of the business and I think I can state this for many members 16 of our mutual fund trade organization, that they would prefer 17 auto-ex, strongly prefer auto-ex because of the immediacy and 18 the certainty of the trade. And there are problems when we 19 end up defaulting -- moving out of a certain box, say an ECN, 20 and having it executed elsewhere, because of the different 21 model structures. But we would prefer auto-ex. 22 MR. COLBY: I think Brett and then -- 23 MR. REDFEARN: I mean our model is to be able to 24 get size orders done at one price, and if you look at even 25 the 11Ac1-5 data, if you're looking at order sizes over 5000 1 shares relative to other markets, our speeds of executions 2 are equally as fast. And we have always said one size doesn't 3 fit all. If you want an execution of 100,000 shares in a 4 subsecond, then this might not be the place for you, but if 5 you're in a situation, as Wayne mentioned earlier, you know a 6 lot of people are having a hard time seeing where it is 7 displayed size, where can I get size done, what are the 8 market impact questions. And if we offer a model where we say 9 you can get size done at one price, I think that's an 10 important thing and it's an important thing eventually for 11 institutions that are out there. 12 And the other thing I just wanted to throw out here 13 is that, you know, there seems to be a view that the auto-ex 14 market is the ideal market structure here when you look at 15 sort of the NASDAQ framework. And I'm not necessarily sure 16 that's true. I don't think that that's been empirically 17 proven. If you look at the inter-day volatility in a NASDAQ 18 market compared to the New York Stock Exchange market and you 19 realize that inter-day volatility is created from a lot of 20 the speedy, speedy executions going across one another, you 21 could make an argument that it's harder for customer orders 22 to meet. Everyone doesn't necessarily want instant speed. 23 Why do you have ITG, why do you have places where they wait 24 for liquidity to pool over the course of an hour, why do 25 people have VWOP trades? Because they know they can't get 1 the exact order at one point in time. And why do people do 2 dollar cost averaging? There's a number of reasons why 3 people look at the best price they can get in a little bit of 4 a different time frame. 5 And so to say that the auto-ex model is the be-all, 6 end-all model for the world, I'm not necessarily sure if 7 that's been empirically proven. 8 MR. COLBY: This disagreement is flaring, and I 9 should point out in the context of best execution, and we 10 haven't yet gotten to whether there should be more stringent 11 honoring of other people's quotes in a market rule 12 environment. But Bernie, you were I think up next. 13 MR. MADOFF: There seems to be a lot of concern 14 about limit orders not being incented to be displayed. I 15 would say that the best method to incent limit orders to be 16 displayed is to make sure that they get executed against. 17 The process of allowing people to stand in the crowd, knowing 18 they're going to get a second look at an order that comes 19 down, just encourages them to not display the order. If you 20 went to an auto-ex process on the floor and people that were 21 standing in the crowd with these orders in their pockets saw 22 executions taking place on the best bid and offer, it 23 wouldn't take long before those people stepped forward and 24 put their orders on the book. 25 So I think that if we want to incentivize limit 1 orders, make sure that they have the ability to get executed 2 immediately. And I think when you talk about access, just 3 having access to get to a location is worse if you can't get 4 executed against. And that's what the problem was in the 5 AMEX. You know, it's like telling someone that you can get 6 into the movie by just getting in line and then getting there 7 and them saying that they're out of seats doesn't encourage 8 anybody from showing up. 9 So if I'm going to send my order to a marketplace, 10 I want to make sure that they're going to do business with 11 that order and that they're going to execute against. So 12 access means two things. Getting there and getting an 13 execution. Just getting there is meaningless to me, it 14 creates more problems than it's worth. 15 MR. COLBY: Gus, you said that you are increasingly 16 using limit orders -- not using limit orders, but using floor 17 brokers, but you also said what the institutional community 18 wants is automatic execution. How do those reconcile? 19 MR. SAUTER: Well, you have to play by the rules. 20 If we're trading over-the-counter stocks we're highly likely 21 to pursue an ECN or now probably the SuperMontage, but if 22 we're going to a floor, would we place a limit order knowing 23 that the floor can participate in front of us or with us, or 24 would we use -- and that's because there are certain 25 advantages that a floor broker has -- or would we use those 1 very advantages that the floor broker has. Since the floor 2 broker is just ordained with those advantages, they're almost 3 crazy not to use them. So that incents us not to place limit 4 orders and actually use people on the floor. 5 MR. COLBY: Larry, have you withdrawn your order or 6 do you still -- 7 MR. LEIBOWITZ: No. All I would say is in 8 listening to all the arguments around the table, I think that 9 this is the best argument you can find that says one hard 10 link market that trades things all the same way isn't the 11 best thing. You have different investors with different 12 types of trading and different valuations of what's good and 13 what's bad. And what works is how they -- we have to figure 14 out better ways to make them interact better, but suddenly 15 legislating that they're all together would cause all sorts 16 of problems. 17 MR. COLBY: Rick. 18 MR. KETCHUM: I would just briefly add while 19 obviously I wouldn't necessarily agree with everything Brett 20 said, let me take forward an assumption that he's right and 21 the AMEX is providing a competitive, effective alternative 22 for institutional investors of some sort to go to -- great, 23 that's what competition is, it's a perfectly logical niche 24 for them to attempt to compete with, and they should, along 25 with everyone else. 1 It is pretty much though beside the point. The 2 point being what has that done with respect to a market that 3 was characterized with automatic execution, with regard to 4 those firms that provide the executions for their customers' 5 order that they have come to expect on an immediate basis 6 when the rest of the market is moving much more quickly than 7 the market on the floor with respect to the quotes adjusting 8 and finding price discovery. It is that that causes me to 9 say that it is just not the same as 20 years ago and the 10 Commission cannot elucidate best execution principles without 11 explaining to a firm what they want to do in that situation. 12 And I don't think the answer is either to stop executing 13 against retail orders or to execute them at a price that is 14 either disadvantageous to them or whatever. I think there 15 has to be some cost for a market when dealing with retail 16 executions for not providing automatic execution. Doesn't 17 mean they can't compete, doesn't mean they shouldn't compete, 18 but there should be some cost and recognition that the rest 19 of the marketplace, if automatic execution is the standard 20 name of the game, should be able to continue to go on. 21 MR. COLBY: Ed, did you still want to speak? 22 MR. NICOLL: I completely agree with Rick. Just 23 would like to make a point that this is -- what we're talking 24 about here with this particular set of examples is the same 25 problem that we find in terms of interacting with, you know, 1 with the floor on listed securities. Just like we're having 2 this issue on what happens when NASDAQ executes, it's the 3 same issue, which is, you know, that you have a marketplace 4 in which you have, you know, a slower marketplace in which 5 the market itself has a particular model and it says its 6 model is superior, and you have another model which cannot 7 function with its value proposition and integrate seamlessly 8 into this slower marketplace. 9 It seems to us that the only rational solution is 10 to let the markets compete with each other and let the 11 investors, who are directing their order flow to those two 12 marketplaces, choose which model they believe will execute 13 their order better and that what we ought to be evaluating is 14 not a set of criteria which is the result of a regulatory 15 structure which was set up in 1975 and which in terms of ITS 16 is largely, you know, a private institution. But what we 17 ought to do is to recognize that when the majority of 18 investors and the majority of people in that marketplace 19 choose one model over the other, that there's a reason why 20 that's happening and that they are informed themselves and 21 that the competition in the marketplace is telling us which 22 is the model that is really creating the value. 23 MR. COLBY: Let's give Ed the -- 24 MR. KWALWASSER: Just two seconds. I just would 25 say I think the characterization of the New York Stock 1 Exchange, if that's what you're characterizing as a slower 2 market, is just a mischaracterization. 3 For almost any order size other than the smallest 4 order size you can find, we actually have executions much 5 faster than ECNs and other markets in those securities. 6 MR. NICOLL: I didn't mean to characterize the New 7 York Stock Exchange -- 8 MR. COLBY: You're burning my time here. 9 MR. NICOLL: But the thing is, they're different. 10 MR. COLBY: It's now time to take the break and 11 move on. There are several classes of issues we never got to 12 today and so maybe in the next round, we can make sure we get 13 there. 14 A 15-minute break. Thanks. 15 (A short recess was taken.) 16 MS. NAZARETH: Well, here we are in the final 17 round, you've all done really well, with no food fights -- 18 terrific. I think some people would have paid extra for 19 that. 20 I'll try to limit my remarks so that we can really 21 get on with the discussion with our panelists. This last 22 segment is on exchange registration and our self-regulatory 23 system. I thought I'd start off by talking very briefly 24 about the self-regulatory system and then secondly, exchange 25 registration. 1 Recent discussions of the current self-regulatory 2 model have focused on several issues, as you know. First, 3 the growth of ATSs and ECNs and electronic trading by SROs 4 have heightened the competition between traditional SROs and 5 their members who sponsor electronic trading systems. 6 Demutualization plans may further aggravate these 7 conflicts as SROs compete against their members as for-profit 8 entities. 9 Some market participants have advocated a single 10 independent SRO model, arguing that a separate SRO that is 11 independent of any particular market or market participant 12 could lead to fewer conflicts and greater economic and 13 regulatory efficiencies. Others have argued in favor of a 14 hybrid model in which the regulation of member activity is 15 conducted by a single independent SRO, but the oversight of 16 trading remains with each individual market center. And of 17 course, there are those who advocate the status quo. 18 So we'll be discussing the SRO model in the second 19 part of this segment. 20 As to the issue of what is an exchange, the current 21 definition focuses on the bringing together of orders for 22 execution under a common set of trading rules. Under this 23 definition, traditional floor-based auctions, newer ATSs that 24 provide systemwide price and time priority, options exchanges 25 that allow dealers to trade on parity with customers, and 1 even an organized over-the-counter market like NASDAQ that 2 does not provide systemwide price and time priority would be 3 considered exchanges. 4 A threshold question that is presented by the 5 NASDAQ exchange application is whether a market that allows 6 dealers to trade against their own orders, away from a 7 voluntary central order interaction facility and that does 8 not, therefore, provide either systemwide price and time 9 priority or customer order priority over dealer trading 10 should be allowed to be registered as an exchange. 11 What will be the ramifications for our existing 12 exchanges if such a marketplace obtains this designation? 13 Would the national market system become a series of 14 fragmented, internalizing exchange models? 15 So that's where we are now. Let's start where I 16 just ended, with the question of exchange registration. And 17 we'll call on someone who is totally objective and unbiased, 18 Rick Ketchum -- 19 (Laughter.) 20 MS. NAZARETH: -- and ask him why should the 21 Commission permit NASDAQ to be registered as an exchange if 22 it does not require interaction between orders or provide 23 systemwide price priority. And to go even further, isn't the 24 analysis further complicated if you become even more purely 25 driven by the profit motive by going public? Won't this lead 1 both to more internalization by the existing exchanges and to 2 a dilution of regulatory efforts? 3 Rick. 4 MR. KETCHUM: Thanks, Annette. 5 (Laughter.) 6 MS. NAZARETH: Answer that very valid question. 7 MR. KETCHUM: I think I liked the setup of the last 8 hour and a half better. 9 Actually though I guess I would refer to the last 10 hour and a half. I think what that discussion showed is I 11 think a credit to the Commission, which is that while there 12 continue to be problems to resolve, that the Commission has 13 allowed to evolve, as you would expect with the type of broad 14 definitions you have in the Act, two types of market 15 environments that can both compete against each other and 16 compete with other entities which adopt both of their models. 17 And the Commission has evolved rules that have reflected both 18 that electronic market environment and an environment where 19 the primary market is a floor. 20 Now I think that's good and it strikes me that the 21 reasons why the SEC should allow NASDAQ to be an exchange is 22 exactly the same reasons why the SEC has allowed NASDAQ, and 23 indeed pushed NASDAQ, in the evolution direction that it has. 24 It would seem odd that there was something magic about the 25 definition of exchange that somehow NASDAQ was a good place 1 to trade and operate a competitive, if it wasn't called an 2 exchange, but if it was, suddenly it became bad. At least, 3 one would think that there would need to be a very specific 4 definition that would tell you that, because otherwise, you'd 5 have to assume that the Commission operated in the public 6 interest and made determinations that the evolution of NASDAQ 7 made sense. 8 In a more substantive response to your question, I 9 think NASDAQ should be an exchange because it provides, 10 particularly after the approval of SuperMontage, a display of 11 the complete depth of the market, that that market display is 12 absolutely accessible, that it integrates dealers with best 13 execution responsibilities and high levels of disclosure and 14 that the result is statistically, from the Commission's very 15 framework that it has chosen to have this basic disclosure, a 16 very competitive marketplace. 17 For all those reasons, and given what the exchange 18 definition says, which is a place that pulls investors 19 together to buy and sell securities, I think it both meets 20 the definition and meets the public policy needs. 21 As to the question of for-profit, I guess it's 22 useful to try to think of why NASDAQ chose to move down this 23 direction initially. Sometimes it's hard for me even to 24 remember. But the reasons were basically two-fold. One was 25 to split off what we thought was one of the fundamental 1 conflicts in what has been mentioned today of having the 2 self-regulatory/regulatory function be essentially performed 3 by the marketplace. We don't think in a world -- because I 4 would differ with some things that have been said -- I do 5 believe that self-regulation is terribly important from both 6 a knowledge standpoint and an industry buy-in standpoint. We 7 don't think that self-regulation should be in any way 8 compromised or create cynicism over whether the decisions of 9 the NASD in any way were driven by NASDAQ's feeling about a 10 particular member firm and how they were participating in the 11 NASDAQ market. That was reason one. 12 Reason two was from a for-profit position, we were 13 able to sell stock, raise capital, invest in technology and 14 allow us to compete more in again the environment the 15 Commission set, which was a very competitive environment, 16 both with exchanges and in the double level both with our 17 customers and working for our customers, along with the 18 potential internationally of trying to compete in a steadily 19 more global world. 20 Those are the good things. The only reason one 21 would not allow for-profit exchanges then I would think would 22 be if one would conclude there were terrible things. And I 23 don't think there are. And I think while there are conflicts 24 here, they are not different in dramatic note from the 25 conflicts of a membership organization. And we avoid the 1 regulatory one. 2 If the thought is that my concern about Schwab to 3 my left or Island and Instinet across from me relates to 4 whether they're a member, that hardly impacts their leverage 5 on me as a marketplace. Their leverage on me in the 6 marketplace is that they have orders and both our position as 7 a market and our ability to profit as a marketplace only 8 succeeds if they participate with us. And God knows, I have 9 enough trouble convincing Ed of that on a regular basis. 10 So I think the leverage, whatever leverage members 11 have, still exists. Fair representation requirements require 12 them to continue to be represented on our board and they are 13 represented in pretty much the exact same numbers, albeit a 14 minority because we believe in a public board, as do you, as 15 before. 16 And the final difference is that there's SEC 17 oversight. And the SEC oversight from every bit of what we 18 operate as well as all of our rules, I think also ensures 19 that whatever conflicts exist and are inevitable in a 20 marketplace environment can be addressed. 21 So I think for all those reasons, I guess I'd 22 probably say that the real question is, from the standpoint 23 of NASDAQ being an exchange, why not? I think it delivers a 24 quality marketplace, it delivered that -- the Commission has 25 developed it and helped it evolve and I think there's no 1 reason to view -- there to be different standards for 2 exchanges than for affiliates of associations. 3 MS. NAZARETH: Gus, as an institution 4 representative here, do you have a certain expectation when 5 you transact business on an exchange that's different from in 6 the dealer market? 7 SAUTER: No, I would say we do not. I would say 8 that if we transact in an ECN or perhaps directly with the 9 dealer, we still expect best execution. We have an ability 10 to ensure that it's best execution, so we perform perhaps our 11 own little regulatory service, but we don't think we're 12 getting any less service from those providers than we are 13 from an exchange. 14 MS. NAZARETH: Maureen, you talked earlier about 15 the importance for pricing in the marketplace of order 16 interaction. 17 MS. O'HARA: Uh-huh. 18 MS. NAZARETH: Can you address what you think would 19 happen in a model where the level of order interaction now is 20 sort of defined as something different for purposes of 21 exchange registration, to the extent that obviously in the 22 dealer market, you have substantially more internalization, 23 customers don't come first, no time-price priority -- do you 24 have a sense of what that would do to the model? 25 MS. O'HARA: I think it raises the issue of 1 customers not coming first, I think is an increasingly 2 challenging question, because I think when we think about 3 exchanges and the access for public orders, there is some 4 sort of presumption that customers trade before dealers and 5 that customers' orders go first. So I think that's an 6 interesting challenge for the Commission to sort out. I 7 mean, I am sympathetic to the notion that the distinctions 8 any more have really become quite moot in many ways. As 9 we've talked today about the various ways that people trade, 10 the distinctions are becoming increasingly slimmer in some 11 ways. I mean calling something one thing as opposed to 12 another now blends more and more as we observe 13 internalization for payment for order flow and everything 14 else. So I am actually sympathetic to the notion that -- 15 sort of putting a name on it -- is less of an issue than it 16 was in the past, but I remain concerned with the issue of the 17 customer order and time priority and I think that's an issue 18 that is not easily dismissed. 19 MS. NAZARETH: Well, do you view this as, to some 20 extent, a labeling issue and something where -- a situation 21 where if there are certain underlying principles that should 22 apply, then they should be applied cross markets regardless 23 of the label that attaches to the marketplace? In other 24 words, if we're saying the time-price priority is important 25 to the marketplace, it should be applied whether you have the 1 label of an exchange or not. 2 MS. O'HARA: I think to some extent these are 3 labeling issues. The notion -- I think one of the things 4 that's become very clear today is that part of the challenge 5 that faces the Commission and anyone trying to sort out "a 6 single national market" is that there isn't one. And you 7 can't have one because there are different needs from 8 different customers. 9 So in some sense, I am sympathetic to the idea that 10 these are labels. I think we do have to go back to what is 11 it that we think is important for the national markets to 12 exhibit. Time priority is something that I would find more 13 attractive than not, but again, an interesting question how 14 you get there from where you're at now at the NASDAQ market. 15 MR. SAUTER: Annette, I'm not sure if I interpreted 16 your question correctly. We would certainly favor a price- 17 time priority system as opposed to the old dealer market. 18 Our view of NASDAQ is a little bit different from that 19 nowadays, given the SuperMontage structure and given the fact 20 that a lot of NASDAQ trading is being done through ECNs. But 21 insofar as whether or not NASDAQ is an exchange or not an 22 exchange, we think we'll probably get the same level of 23 service. We certainly believe there should be appropriate 24 regulation regardless of whether it's an exchange or not. And 25 we also monitor the situation ourselves to the extent we can. 1 MS. NAZARETH: Larry, did you have something? 2 MR. LEIBOWITZ: Yeah, I mean I think there's 3 nothing inherent in an exchange that says time-price is 4 critical to being an exchange. In fact, the New York Stock 5 Exchange itself has various rules under which time-price does 6 not take precedence like clean crosses or sizing the book and 7 situations like that. Curiously it's the larger orders that 8 take precedence over the retail investor in that case, but 9 the point being there's nothing inherent that says time-price 10 is the requirement for being an exchange. 11 I think when we really look at it, it's the whole 12 process by which they perform their regulatory functions and 13 the whole rigor which the exchange subjects them to to become 14 registered and to do their ongoing operations. And I think 15 that that's got to be separated from the actual operation of 16 the market, which could be done as a separate entity, meaning 17 that all of the exchange -- the things that we think of as an 18 exchange could very easily be done in a super regulatory body 19 outside of that letting the marketplace develop in different 20 ways as long as they are consistent with various principles 21 that the exchange sets out. 22 MS. NAZARETH: John Markese, do you have a view on 23 what the expectations of retail investors would be to -- what 24 do they think they're getting when their order goes to an 25 exchange? 1 MR. MARKESE: I think what they're assuming is 2 regulation, I go back to that. I think they're assuming a 3 certain set of rules, requirements, regulations and 4 oversight. And I think the exchange has symbolically over 5 the years conjured that up in investor minds, and I think it 6 does with NASDAQ also. 7 MS. NAZARETH: Do they have an expectation that the 8 customer is coming before the specialist? 9 MR. MARKESE: You know, in most cases, I think they 10 would. The question is size, liquidity issues. I don't know 11 if they're that conversant, but if they were head-to-head 12 with a 100 share order, I would assume they would think they 13 would come first if it were size and other issues. I think 14 if you gave them that, they wouldn't understand it's not 15 always that issue. 16 MS. NAZARETH: Joel, you've studied these issues, 17 do you have a view on whether there's something magical to 18 the definition of exchange or are there certain attributes 19 that should be embedded in it? 20 MR. HASBROUCK: I think in the public mind there 21 very well may be. And also, as the SEC, when the SEC confers 22 the designation of exchange on an entity, it also puts its 23 own reputation on the line. And for that reason, it needs to 24 ensure that the regulatory functions are up to what the 25 public expects. 1 But I can also see benefits from labeling something 2 an exchange, you give it a collective identity and you put 3 some responsibility on the members that they have a 4 collective reputation at stake and that may help them 5 regulate each other. 6 As far as price-time priority, it's an important 7 issue. I don't think it's tied in the public mind with the 8 notion of an exchange or not. 9 MS. NAZARETH: Rich. 10 MR. BERNARD: Thank you, Annette, I think you know 11 I have some thoughts on this. 12 (Laughter.) 13 MR. BERNARD: New York Stock Exchange, for those 14 who are listening in. First of all, I would be the first to 15 say to Rick and everyone from the OTC community what a great 16 improvement the over-the-counter market is today from what it 17 was 20 years ago. So this discussion is not about NASDAQ's 18 failing to move in the right direction, they certainly have. 19 Secondly, Rick, you're probably surprised that 20 there's no question in my mind that NASDAQ falls into the 21 definition of an exchange, but that turns out to be your 22 problem. 23 But before I get to that, I think that there are 24 some profound issues around a market, particularly a primary 25 market, being profit maximizing. I also think that there are 1 some profound issues about separating the so-called upstairs 2 regulator, the member firm regulation, from the market and 3 the market regulation, as much as Ed and I, Ed Kwalwasser -- 4 that same famous Ed Kwalwasser and I -- fight about all kinds 5 of things all the time, I will tell you that what he brings 6 to the table with Cathy and Bob Richardson, Dick Grasso and 7 I, makes us a better market and a better self-regulatory 8 organization. 9 That said, I'm not here to say that NASDAQ and the 10 NASD shouldn't be free to pursue their own genius and their 11 own direction in these areas. They and we have the right to 12 fail, they and we have the right to be proven correct or 13 wrong. 14 Rick's problem is what I call the rock and the hard 15 place problem. And I'm going to make reference to a statute 16 and a rule, but I don't mean to make a lawyer's argument 17 about this, I want to expose the public policy that guided 18 Congress and the early SEC when they grappled with the issue, 19 not just a price-time priority but at a price whether the 20 customer always had to go ahead of the dealer. That is the 21 discussion that the application for registration by the 22 NASDAQ creates. 23 And to illustrate that public policy issue, I want 24 to remind you that there was no 15A in the 1934 Act in 1934. 25 So you went -- you had Section 6 -- you had Section 3 that 1 said, as Rick said, that if you organize trading, you are an 2 exchange. And Section 5 says that if you organize trading 3 and you are an exchange, you must register that exchange. 4 And then here's the rub, when you get to Section 6, it 5 imposes certain obligations on national securities exchanges 6 and they are different from 15A because 15A was written to 7 regulate the unorganized OTC market. That was what the 8 Maloney Act in 1938 did. 9 The problem that Congress had in 1934, coming off 10 of the 1928 parallels to the 1990s was floor trading on the 11 New York Stock Exchange, where people with no obligations to 12 the market had used their time and place advantage to trade 13 ahead of customers and to manipulate the market and do a lot 14 of things. And Congress came this close to saying that there 15 had to be a complete separation of the dealer and the agency 16 function on the stock exchange -- this close. And in fact, 17 there was a report mandated by the original 1934 Act that the 18 SEC fulfilled in 1936, called the segregation report, that 19 was all about would we segregate, as the statute almost had 20 done, the role of the agent from the role of the dealer. 21 In the end, the SEC chose not to do that and there 22 was a fellow by the name of David Saperstein said no, we're 23 going to let the specialists -- and by the way, this is a 24 time of competing specialists, so what I'm describing here is 25 not about a competing dealer system versus a unitary 1 specialist system. What David Saperstein said was no, we're 2 going to let the dealer, the specialist, be both an agent and 3 a marketmaker, but two things that agent/marketmaker has to 4 do. He's got some affirmative obligations and they all 5 freely concede that if you have a multiple marketmaking 6 system, the nature of the affirmative obligation can be and 7 is collective and doesn't reside with a particular 8 marketmaker. 9 But the negative obligation is another matter. The 10 negative obligation says that a marketmaker, registered 11 marketmaker on an exchange cannot trade ahead of a customer 12 order at the same price on the same side -- absolute flat out 13 prohibition. That was carried forward, that became an NYSE 14 rule. And in a 1964 special study which was looking at 15 problems, particularly on the AMEX but also on the New York 16 Stock Exchange, those rules became hard-wired rules and you 17 can read that rule in the NYSE rule book today. 18 Why do I say that, because Rick has built a better 19 market, but the public policy behind the separation of the 20 broker-dealer role goes back to a very simple idea. What 21 Congress said in 1934 that exchanges had to do was maximize 22 order competition. It also said, with the help of David 23 Saperstein, that we're going to resolve conflicts of interest 24 between agents and dealers in a particular way and that is 25 that if you're going to be both an agent and a dealer, the 1 customer always goes ahead of your dealer interests. And 2 that's how we end up with the negative obligation. 3 So what I think Rick's problem is that while he has 4 done better, he's got to do better still. 5 And I listened to the best-ex discussion when Ed 6 was sitting in this seat and I heard arguments for why the 7 dealer ought to have parity with customer orders and ought to 8 be able to trade ahead of customer orders. And I understand 9 that that's better than what's going on before, but I don't 10 understand why the fundamental public policy that says that 11 in an organized market the dealer goes last should give way. 12 And I think that's the fundamental question that faces the 13 five Commissioners here -- has something happened between 14 1938 or 1964 and 2002 that changes that fundamental public 15 policy approach. And I submit it hasn't and I won't take any 16 more time, but I can give you a very nice scenario. I could 17 read you a description of what was going on in the New York 18 Stock Exchange in the 1920s and I could read you a 19 description of what was going on in our secondary and primary 20 markets in the 1990s and you wouldn't be able to tell the 21 difference. 22 MR. COLBY: I can't let a historical statutory 23 discussion go without just making a few comments. 24 The first was that the segregation study was for 25 all brokers and dealers, it wasn't strictly speaking for -- 1 the second was that if I remember correctly the multiple 2 dealer market on the New York Stock Exchange at this time had 3 basically boiled down to two and that they were cooperating. 4 They were alternating when they took the orders. And the 5 third is that Section 6 and Section 15 statutorily read 6 almost identical. 7 My point in saying this is that the Saperstein 8 interpretation that you're pointing to was in effect a staff 9 interpretation looking at the New York Stock Exchange which 10 was not so different then as it is today. 11 MR. BERNARD: Just one quick point. I think the 12 competing specialists lasted longer on the New York Stock 13 Exchange than 1938 and, although you can read the special 14 study, you'll see that they're still talking about competing 15 specialists. That said, I will not argue the point that the 16 negative obligation may not work well with the competing 17 specialist system. That may be intentional. It's no 18 accident that there are not competing specialists on the New 19 York Stock Exchange, because when you put those affirmative 20 obligations and you put that negative obligation on someone, 21 it may tend to drive you to a unitary specialist model. 22 MS. NAZARETH: Bernie. 23 MR. MADOFF: A couple of points. First of all, the 24 term exchange today, I'm not sure that it has the same 25 meaning as it did years ago. I mean if you look at the rest 1 of the world, every marketplace is considered an exchange and 2 none of them look like the New York Stock Exchange, which for 3 the most part, most people look at as an exchange. 4 So I'm not sure that I really agree with the fact 5 that someone is called an exchange unto itself changes 6 anything or requires different responsibilities. 7 It's also I think important to understand that 8 customers do come first at each dealer market. NASDAQ is a 9 dealer market and to say that customers do not come first I 10 think is just flat out wrong. Every dealer is required to 11 put his customers' order first over his own proprietary 12 orders. And since most of the orders take place at various 13 dealer markets, I would argue to say that I never trade ahead 14 of any customer order that I have in my marketplace. 15 So to say that the customer does not come first in 16 NASDAQ is not the case. If you look at it as a global 17 marketplace and you say that there's no time priority, that 18 is correct, but within each dealer's market, the customer 19 does come first. 20 I also think it's hard to define who is a customer 21 today. You know, years ago, most people looked at customers 22 differently than they are today. Today, the typical customer 23 is not the, you know, the little old lady from Pasadena who 24 everyone used to talk about that was buying 200 shares of 25 General Motors. The customers that most dealers are dealing 1 with not only in the NASDAQ market, but in the exchange 2 market, are professionals. It'd be very hard to say that the 3 typical customer, professional trader, whether they be 4 program trader, hedge fund, day trader or whatever, 5 institution, is any different than a dealer. And I think 6 there is a big danger in saying that dealers should be the 7 market of last resort, particularly in a competing dealer 8 market like NASDAQ. 9 The New York Stock Exchange is a franchise and the 10 dealer has a monopoly basically and gets compensated entirely 11 differently than a marketmaker in a dealer environment. 12 So I think you have to be very sort of mindful 13 there are differences in that marketplace and you cannot 14 necessarily fit them all into the same cubicle. 15 MR. BERNARD: If I could just make a few comments 16 in response. First, I'm not talking about labeling 17 exchanges. I'm saying that if you organize markets with the 18 statutory scheme and the public policy says that you must 19 register under Section 6 and follow the Section 6 20 requirements, which are different in some material respects, 21 particularly this one, from a 15A. 22 Second, Bernie, you're on the wrong side of the 23 market. I understand that the theory and other agency 24 notions in the over-the-counter markets say you can't trade 25 ahead of your own customer. I'm talking about trading ahead 1 of the customer on the other side. I'm talking about 2 stepping ahead of the public customer who could have traded 3 with your customer but for the fact that you interposed 4 yourself on the other side from your customer at the same 5 price as limit orders or latent interests in the market. 6 And thirdly, Bernie, I just want you to know that 7 you're my customer on the New York Stock Exchange and that's 8 one of the outputs of a system like this. You come to the 9 New York Stock Exchange the same way those investors standing 10 behind the institutions come to the New York Stock Exchange 11 collectively or individually and our specialist goes last 12 after you. 13 MS. NAZARETH: Wayne. I'll get to you, don't 14 worry. 15 MR. WAGNER: I hate to impose on this argument 16 that's been going on for 60 years here, but -- 17 (Laughter.) 18 MR. WAGNER: -- you know, one of the questions that 19 always comes up to anyone who is evaluating the quality of 20 executions is is there any difference between what goes on on 21 the New York Stock Exchange listed stocks and in what goes on 22 on the NASDAQ listed stocks. I said that very carefully 23 because we don't really look so much at the individual 24 exchanges, because we don't have that data, but we can tell 25 by the registration of the company where it primarily trades 1 here. 2 If you go back say five years ago, there was a very 3 significant difference between that in the favor of the New 4 York Stock Exchange. What we have witnessed in the last five 5 years is a coming together of the results of trading on 6 NASDAQ and on trading on the New York Stock Exchange much 7 like, and I think in large part attributable to, the fact 8 that the rules have gotten closer together. You know, the 9 exchange has incorporated some dealer like functions and the 10 dealer market has incorporated some exchange like functions, 11 particularly the limit order being able to compete in that 12 here. 13 So from an institutional standpoint, you know, with 14 the array of facilities that are available here and not all 15 trades go to the NASDAQ directly, not all trades go to New 16 York directly, they can be done in various places, but 17 institutional investors who have all of these choices have 18 pretty much equalized out the difference between the cost of 19 executing in one exchange or the other, even considering the 20 vastly different nature of the companies and the volatilities 21 and liquidity that's available on the different stocks that 22 tend to be NASDAQ listed versus exchange listed. 23 MR. COLBY: So what you're saying is not just at 24 the top of the market but all the way down the spectrum of 25 stocks the execution quality has become more comparable 1 across NASDAQ. 2 MR. WAGNER: I believe that, yes. Now it went kind 3 of crazy when we were trying to look at the effect of penny 4 spreads and we found that we couldn't even look at them 5 because there was such a massive shift going on from the high 6 tech NASDAQ companies into safer securities here, the bad 7 effect -- you know, that was the tide overwhelming the little 8 lapping waves on the shore of the penny spreads. 9 MS. NAZARETH: Rick, you've been very patient. 10 MR. KETCHUM: Thanks. I'm glad I was patient, it 11 was good to hear Wayne's comments. 12 I would start at the same point Rich starts at, 13 which is I'm not here to say anything or denigrate or 14 question the operation of the New York Stock Exchange because 15 the New York Stock Exchange has consistently made us a better 16 market for years and that's sort of part of the point, that 17 we do tend to make each other better markets. 18 I said earlier that it was a lot more fun thinking 19 about the flexibility of Section 11A when I was at the 20 Commission than when I'm on the other side. But I would find 21 that much less terrifying than to think that the Commission 22 was frozen on interpretations in 1937. 23 I don't know if what Rich said is correct, that his 24 market hasn't changed and there is no difference in then or 25 2000, but there is not. From -- and I appreciate Rich's kind 1 words on noting those changes. 2 I won't repeat because I think it's been discussed 3 across the table, there are benefits to many models, I 4 obviously passionately believe in the benefits of the NASDAQ 5 model from an open standpoint of integrating marketmakers and 6 ECNs together and providing accessibility with automatic 7 execution when that's certainly combined with the requirement 8 that customer orders are always shown, and as Bernie 9 indicates, a requirement that each dealer and each 10 marketmaker must not trade in front of its customers' orders. 11 There's a different environment on a floor 12 environment, which tries to pull all of this together. It 13 depends -- and I think history will show even in the most 14 competing specialist of times -- depends fundamentally in a 15 monopoly which the Saperstein interpretation I believe refers 16 to about four times in two pages. That requires different 17 regulatory treatment and negative obligations makes sense 18 there and different affirmative obligations than NASDAQ 19 imposes there. 20 It's not right nor wrong, but it's very different, 21 the definition of exchange interestingly enough doesn't 22 choose one of those environments or another, it describes 23 pulling together investors and pulling together orders in 24 order to encourage executions. And that's what both 25 environments do. I think the Commission is perfectly capable 1 of interpreting both in areas whether we're called it an 2 exchange or not, certainly has been able to before in two 3 different regulatory setups, and I'm confident you'll 4 continue to be able to, but I think the way you look at 5 markets is within the way they're set up and decide whether 6 they're consistent with the public interest or not, and you 7 ensure that the regulations encourage that. 8 MR. BERNARD: Surrebuttal or whatever we're into 9 now. A couple of thoughts that Rick didn't really come back 10 to. First of all, of course, it's not 1936, it was brought 11 forward in 1964, it remains on the rule books of the New York 12 Stock Exchange and every other stock exchange, including 13 Cincinnati to this day. So the negative obligation is alive 14 and well and living in America. 15 Secondly, I challenge you to do better, I challenge 16 you to do best. You have to answer the question that was 17 posed by policymakers, you know, throughout generations in 18 the SEC of why you would ever have the dealer go ahead of the 19 customer. You have not articulated why your market model 20 requires letting firms in effect internalize their order 21 flow, and again, I don't mean to -- I understand it's a whole 22 lot better than what it used to be, don't misunderstand me. 23 But the core question is and it's in that Section 11A that 24 you used to like a lot better than you do now, that we wanted 25 to let orders interact, that was going to be a priority. 11A 1 remember, was talking about exchanges facing off against the 2 over-the-counter market. 3 You have chosen not to be the over-the-counter 4 market any more, Rick, you've left that to the NASD and the 5 ADF. You want to come into the organized market regime and 6 there are a lot of smart people over a lot of years that have 7 said that the customer goes ahead of the dealer. Why? 8 Because it promotes order interaction and it negates the 9 conflict of interests that have been the elephant in this 10 room, if there can be two elephants, since the discussion 11 started today. And it seems to me that if you want to be 12 registered as an exchange, you have to answer that public 13 policy question, and then unfortunately you've got another 14 problem because you don't have a choice not to register as an 15 exchange because you have crossed the line from being a 16 regulator of an unorganized market into creating an organized 17 market by putting up at least a SuperMontage, if not things 18 earlier than that. 19 MR. KETCHUM: I promise I won't do this more than 20 one more time today. I have a couple of things, I think I'll 21 choose to let the Commission determine when NASDAQ has to 22 register as an exchange or not, since last I checked, they 23 disagree with you. But in the end, I'll go with their 24 judgment. 25 But let me make a couple of things clear. One, we 1 do reach for the best and we think we are. And we think we 2 are because we're transparent, we provide multiple levels in 3 which people can actually execute against. And why dealers 4 have the ability to trade with their customers without 5 exposing the orders is because, in our judgment, the result 6 of that is better liquidity, competition, the avoidance of a 7 monopoly or choosing a club environment which time after time 8 in Europe has been demonstrated as not working very well for 9 retail orders, as the New York Stock Exchange has indicated. 10 So with respect to the three models, we believe 11 this model operates and proves the best level of liquidity, 12 the best prices and indeed, as I say, I will freely admit 13 these numbers have looked different at different times, when 14 you compare S&P 500 companies right now with respect to the 15 Commission's own measurements, which the New York Stock 16 Exchange rabidly endorsed at the time, from speed quotes and 17 effective quotes, that shows up as not a market, but the best 18 market. 19 So for all those reasons, while we don't pretend 20 that necessarily everybody would agree with that all the 21 time, we do think that it has rather resulted in a better 22 market environment to have that choice, and we don't think 23 it's anywhere in the definition of exchange. But again, I'm 24 more than willing to defer to the Commission on that. 25 MS. NAZARETH: Even putting aside for a few minutes 1 what the definition of an exchange is, I think it's pretty 2 clear that one of the challenges that the Commission will 3 face is if it defines a market that does not have time-price 4 priority as an exchange, then undoubtedly you'll see, as 5 we've already seen with some of the filings that we have such 6 as the voluntary book filing, a number of the other 7 exchanges, existing exchanges, will come in with proposals 8 that basically blow through time-price priority and would 9 permit those exchanges to just be print facilities or 10 internalization models. And obviously, you know, in the 11 interest of fairness, and in avoiding regulatory arbitrage, 12 it's going to be difficult for the Commission to distinguish 13 between the existing NASDAQ model and these other things 14 which will also have the same label. 15 I'm interested if anyone has any comments on that 16 or thoughts in terms of how the Commission could address 17 those issues in a way that doesn't basically result -- to 18 some extent, I think our fear would be in the race to the 19 bottom and a degradation of the pricing mechanism by virtue 20 of there being less order interaction in the marketplace. 21 Mike, do you have any thoughts? 22 MR. RYAN: Well, as you know, we filed with the 23 Commission and have had pending for awhile the ability to do 24 what would likely would come -- would have to result if 25 NASDAQ became an exchange, and that is being able to 1 basically sell prints. We don't believe the Commission 2 should approve our rule any more than they should approve the 3 ability for NASDAQ to operate that way, but we think that 4 you're absolutely right, if the Commission does approve that 5 NASDAQ is an exchange, all this gets thrown into flux and you 6 end up with markets just becoming internalization models and 7 print facilities. And I think given what we have experienced 8 at the American Stock Exchange over the last year, it 9 demonstrates the unhealthiness of that. 10 The one thing that we haven't focused on in this 11 conversation is -- and we talk about best execution and 12 principles that we articulated this morning -- is the notion 13 of -- we talk about price-time priority, but also the trade 14 protection in the listed market and ensuring best execution 15 of markets -- of customer orders in that environment. And I 16 think that's an important piece of this conversation and we 17 think very -- feel very strongly that that notion should 18 continue to exist in ensuring that customers do get the best 19 prices available in the marketplace. 20 MS. NAZARETH: Ed. 21 MR. NICOLL: I just want to address this issue, if 22 I can. 23 The race to the bottom is an interesting 24 conception. Effectively, some people call it sort of 25 pathological competition, which results in a bad result for 1 all. A classic example of this, in another regulatory 2 setting is the competition between the states for corporate 3 charters and Delaware supposedly raced to the bottom and got 4 all the corporate charter business. But in fact, there have 5 been many academic studies which have shown that Delaware 6 actually -- that the so-called race to the bottom for 7 Delaware resulted in an optimum set of legal standards that 8 benefit everybody. 9 I'm a little confused too in this setting about why 10 we have this potential for race to the bottom since we have 11 very much entities which are regulated by the SEC. And the 12 SEC can stop any so-called pathological competition by 13 setting certain standards and by creating certain mandates 14 for exchanges. And you know, if we don't want exchanges to 15 be "merely print facilities," then you know, we can make sure 16 that they meet certain standards of, you know, activity and 17 of compliance that the exchange mandates. So we can actually 18 at any time stop the race to the bottom simply by requiring 19 exchanges to have certain facilities and to invest in their 20 compliance departments and make sure they perform certain 21 functions. 22 You know, a lot of this -- I feel a little bit like 23 the 900-pound gorilla here -- a lot of this is really the 24 result of the fact that at one point, Island quoted in one 25 market and reported its trades in another. And many people 1 feel that somehow this is bad. But it in fact resulted in a 2 lot of price competition which benefitted not only Island but 3 everybody in the community, in our opinion. And we dealt 4 with an entity which was a regulated exchange and which had 5 to meet certain requirements under the regulatory regime. 6 We thought that we were dealing with a national 7 securities exchange and we thought that that exchange met 8 certain minimum standards and we thought that it was 9 perfectly permissible for us to report our trades and now to 10 quote in that exchange because it met those standards. Now 11 either it does or it doesn't but if it does, I think we 12 should stop talking about a race to the bottom and talk about 13 the benefits of competition. If it doesn't meet those 14 standards, then I think we should say why it doesn't and we 15 should make sure that it does. 16 MR. RYAN: But you make that comment and you fail 17 to completely recognize that Island at the same time was 18 failing to meet the standards of regulation which said that 19 Island needed to -- had met a level of activity that required 20 it to become part of a national market system. And so -- and 21 then it was using its process and its ignoring of that new 22 regulation -- this wasn't something back in 1934, this was 23 something that was adopted in 1998 -- and tried to give ECNs 24 a pass on some of the burdens of an exchange until you 25 reached a certain level of activity. And you ignored that 1 and then you used the revenues from that national market 2 system that are there to ensure customers get the best 3 execution available and it also tried to incent or provide 4 opportunities for markets to compete for customer orders. 5 MS. NAZARETH: Well, this is a little off topic 6 from where we were, but I think that Ed sort of circled back 7 to a question that I asked earlier, which was are there some 8 principles that are fundamental, that regardless of your 9 market structure should be applied and then let markets 10 compete on whatever their models are. And I guess one of the 11 questions we have is, to the extent we believe that the time- 12 price priority within a marketplace has at least encouraged 13 some interaction of orders in those marketplaces, you know, 14 is there some way that you can reconcile that in the auction 15 model with the dealer model and use the same label. 16 Bernie. And by the way, the Chairman rightly said 17 we've been talking about print facilities and it would be 18 helpful to explain what do we mean when we say that something 19 is a print facility. You want to start with that, Bernie? 20 (Laughter.) 21 MR. MADOFF: Well, I don't use any of these print 22 facilities. 23 MS. NAZARETH: I know, but you know the definition 24 of all these things, Bernie. 25 MR. MADOFF: Well, you know, I think what people 1 refer to as a print facility is somebody using the structure 2 of an exchange that has the ability to capture market data -- 3 revenue from these prints but does not actually operate as an 4 exchange itself, as some people define an exchange. 5 MS. NAZARETH: Good job. 6 MR. BERNARD: It's pretty good for somebody who 7 never does any. 8 MR. MADOFF: My brother told me that. 9 (Laughter.) 10 MR. RYAN: Can I add one piece to that because they 11 are printing on an exchange, but they are not providing any 12 orders or quotes into that market that are meaningful, that 13 can be accessed through the national market system. So 14 that's a critical piece of it. These markets end up with 15 very wide quotes and then they just print at or between the 16 NBBO and the order is never displayed. So there's no ability 17 to access anything meaningful other than through the primary 18 markets. 19 MS. NAZARETH: I'll let Bernie make his comment and 20 then I'm going to let Rich tell us why he thinks -- because 21 I've heard this song before -- why he thinks that it's a 22 problem to have a market's data reflect transactions that did 23 not actually transact in that marketplace and then we'll -- 24 MR. MADOFF: Yeah, I was going to address a 25 different issue, which I thought was the question. 1 (Laughter.) 2 MR. MADOFF: But I think that it's important to 3 understand that we have not come to a conclusion as to which 4 type of exchange or market structure is better. In other 5 words, there are people that feel that the New York is a 6 better market structure and there are people that think that 7 the NASDAQ market is a better structure. I think we've 8 listened to three or four hours so far of the problems that 9 exist certainly with the floor type system. So I think you'd 10 be hard pressed to say that the exchanges that use a strict 11 price-time priority have demonstrated to the marketplace in 12 general -- meaning investors -- that their model is 13 definitely without its problems, particularly in a rapidly 14 changing environment. 15 So I don't think it's necessary to get hung up -- 16 as our Chairman had said earlier that just because that's the 17 way it was, that that's the way it should be. I think the 18 jury is still out and I think that, you know, Ed makes a 19 valid point, that all of these entities are regulated very 20 closely and if in fact problems exist or there is this so- 21 called race to the bottom that people worry about occurs, 22 then the Commission will deal with it at that point, but I 23 think there's more danger in not letter NASDAQ go forward as 24 an exchange than holding it back from being an exchange, 25 because you're never going to see whether these so-called 1 perceived problems occur until it goes down the path. 2 Again, I think that this whole issue of price-time 3 priority is almost sort of academic at this stage. In a 4 decimal environment, I never hear anybody complaining -- and 5 God knows we do as much volume as pretty much anyone around 6 this table other than New York and NASDAQ, but we do hundreds 7 of thousands of trades a day and I never hear any of my 8 customers complaining that they have limit orders that are 9 not getting executed or they're putting orders in of any type 10 that are not being interacted with. I mean in a decimal 11 environment, that just doesn't exist. These markets move so 12 quickly, are so volatile, that if anything, people complain 13 that they got an execution sometimes. 14 MR. BERNARD: I think there was a question in there 15 for me about print facilities and if you'll just let me say a 16 few other things about some other things that have been said 17 -- and this does tie back to our view of NASDAQ. 18 I guess the place to start is we've talked about a 19 race to the bottom. If you're going to let our major order 20 providers internalize their order flow on the NASDAQ exchange 21 and the NASDAQ exchange competes with the New York Stock 22 Exchange and listed stocks, as it has announced to do, I want 23 you to think about what the competitive response to the New 24 York Stock Exchange is going to have to be. 25 We have said that internalization is not a 1 particularly good idea. We have said that dealer/agency 2 conflicts are not a particularly good idea. They should be 3 left in the unorganized over-the-counter market. I would 4 welcome competing with NASDAQ if it took it the rest of the 5 way, took out its internalization features, put up a 6 competing marketmaker system and had a -- and their non- 7 physically convened auction goes up against our physically 8 convened auction, and let's let the best market model win. 9 But the idea -- and by the way, I'm not telling the 10 Commission to do anything, this is a public policy discussion 11 and I am speaking of public policy here. The idea that 12 something that's called a national securities exchange, which 13 does as some of these folks have said, has some meaning. If 14 it means nothing else, it means that these conflicts of 15 interest get resolved by letting the customer on the other 16 side of the market -- Bernie, you've got to get on the other 17 side of the market -- letting the customer go ahead of the 18 dealer. And the tape print issue ties into this because 23 19 percent of the executions in the over-the-counter market, 20 according to Rick's report yesterday, occur inside of NASDAQ 21 facilities. Said another way, 77 percent are either 22 occurring on Ed's or other ECNs or are marketmakers 23 internalizing their order flow. 24 This is a profound issue because NASDAQ is going to 25 say that those upstairs executions took place on the NASDAQ 1 stock exchange, and they're going to report their volume as 2 if those trades took place in the exchange facilities, and 3 they will not. 4 MS. NAZARETH: Bernie. 5 MR. MADOFF: Can I just say, there are very few 6 people that would ever dream of saying that the marketmakers 7 that trade NASDAQ securities, you know, every day, that those 8 transactions are not taking place in the NASDAQ marketplace. 9 It's one thing to say that the ECNs' trades do not take place 10 in the NASDAQ marketplace, but if you're going to tell me 11 that my trades in NASDAQ securities or Merrill Lynch's or 12 Goldman or Morgan or whoever else, that are done the way we 13 normally do our business and have been since the beginning of 14 time, should not be counted as NASDAQ trades because they're 15 not flowing through SuperMontage or through some system -- I 16 mean to me that just seems silly. 17 So you know, the statistic that's thrown out there 18 that some people have jumped on of only 20 percent of the 19 trades -- when I heard that statistic, I was speaking at a 20 conference three days ago, I called up NASDAQ and I said I 21 see this statistic, is this true. And then I understood 22 where the statistic was coming from. And I said no, no, this 23 is absolutely ridiculous. If NASDAQ trades are taking place 24 in NASDAQ marketmaking trading rooms, those have always been 25 and always should be considered NASDAQ trades. 1 MR. RYAN: Bernie, we just started trading NASDAQ 2 stocks. If you become an AMEX member, we'll make a deal with 3 you and you can print them, you can bring those trades to the 4 American Stock Exchange. If you're a AMEX member, you're 5 trading those securities on a listed basis, why is that -- 6 why aren't they AMEX prints? 7 MR. MADOFF: Well, they would be if you provided 8 the same type of facility, of efficiency, that NASDAQ 9 provides us today and you could convince me that I should 10 take my trades to -- 11 MR. RYAN: So you can -- 12 MR. MADOFF: I would do it, I'd be happy, as would 13 everybody else that if you can build a better mouse trap in 14 trading NASDAQ securities, then NASDAQ -- 15 MR. RYAN: But those trades don't have to happen on 16 our facility though. They're happening in your facility. 17 You can take your -- you can use the NASDAQ facility and 18 execute in SuperMontage all you want, but the trades you're 19 internalizing in your shop, using your own trading facility, 20 come to us, we'll buy them from you. 21 MR. MADOFF: Well, first of all, I happen to 22 believe -- which is a different question but we're going to 23 get into it hopefully -- that everybody should operate within 24 the national market system structure. I mean I think it's 25 very important and one of the problems that's occurring today 1 is that there are people that are trading outside of the 2 national market system and that to me is detrimental to the 3 organized market. 4 MR. RYAN: In NASDAQ stocks, we're not trading 5 outside the national market system for NASDAQ stocks, we're 6 trading within it. So we'll buy your prints that you 7 internalize in NASDAQ stocks in your shop. It's not going 8 through a NASDAQ system. 9 MR. MADOFF: Well, we'll talk when you -- 10 (Laughter.) 11 MR. RYAN: No, I'm serious. 12 MR. COLBY: Is this a private interview? 13 MR. RYAN: I'm trying to prove a point that once 14 you get -- if it's an exchange and it's not happening through 15 the facility of an exchange, then everything is up for grabs. 16 I mean if he registers an exchange and it gets approved and 17 it's happening within Bernie's shop and it's not hitting any 18 of your systems, why is that any more a NASDAQ print than it 19 is an AMEX print? 20 CHAIRMAN PITT: At the risk of sort of localizing 21 debate, let me ask both Rick and Rich a question. 22 If NASDAQ exists today as it does and trades are 23 taking place on NASDAQ today and if Rick is correct in his 24 statement that NASDAQ is not required to be registered as an 25 exchange, then the question is what is the principal benefit 1 to investors that will be realized by NASDAQ being registered 2 as an exchange. And for Rich, it's the obverse of that, 3 which is to say if NASDAQ is already in existence and 4 trading, what is the principal disadvantage to investors if 5 NASDAQ is allowed to call itself an exchange instead of 6 whatever it is today? 7 MR. KETCHUM: Well, let me try to just answer your 8 question and assume that someone will note that there are 9 numerous exchanges where you can presently print, starting 10 with the Cincinnati Stock Exchange, as Ed indicated. I think 11 the answer to the principal advantage that should give NASDAQ 12 of the choice of whether it chooses to be an exchange or not 13 is the ability to separate from the NASD. It is to provide 14 an environment where NASDAQ remains responsible and indeed 15 accepts fundamental responsibility for its self-regulation 16 but has the ability to separate the operation of that through 17 the NASD in effective form in which the NASD does not remain 18 the majority owner of NASDAQ which is what is required today. 19 And thus, the ownership conflict that we cannot eliminate 20 simply by privately placing part of NASDAQ would go away. 21 I think that conflict, while again agreeing with 22 Rich that I don't think -- it is not our suggestion that 23 there should be one means of the Commission dictating self- 24 regulation in all marketplaces. I think that's a benefit, 25 certainly one that NASDAQ should have the ability to do. 1 Miraculously enough, NASDAQ operates today 2 reporting all of its trades as the NASDAQ stock market 3 competing with the New York Stock Exchange and the American 4 Stock Exchange. Miraculously enough, New York continues to 5 survive, even though all stocks listed after April 1979 6 already have the ability for Rich's member firms to provide 7 internalized executions. And notwithstanding that, the New 8 York Stock Exchange remains a formidable -- 9 MR. COLBY: He's gone now. 10 MR. KETCHUM: So I think that the benefit, given 11 the fact that I see no cost, is precisely that, to allow 12 NASDAQ to choose, as it chooses, to separate and eliminate 13 that conflict. 14 MS. NAZARETH: Rich. 15 MR. BERNARD: Just coming back to first principles, 16 internalization is something that is, in the thought of a lot 17 of people, -- we found a London Stock Exchange report in 18 1880, I think, didn't we, Bob -- people have said that's not 19 what stock exchanges are about. Internalization is something 20 that belongs in the over-the-counter market and in the world 21 I would envision, the NASD would continue to provide the 22 facilities for people to have the transparency. But once 23 NASDAQ put up SuperMontage and once it put up Primex and 24 created facilities for order interaction, you can't go 25 halfway. And again, while it's in the statute and in the 1 rules that have been adopted on the statute, it's a public 2 policy point. You don't want the Securities and Exchange 3 Commission designating as an exchange something that does not 4 resolve conflicts of interest between the agency function and 5 the dealer function by having the dealer go last as regards 6 other orders on the same side of the market as the dealer. I 7 think that's a fundamental public policy issue and I just 8 think it's bad public policy. The competitive environment 9 will still be very compelling with NASDAQ operating an option 10 market, with registered dealers that go behind the customers. 11 And if the member firms continue to want to internalize, then 12 they can do that in the over-the-counter market. They can 13 report those prints to the NASD. By the way, the NASD can 14 use the market data revenue to cover the regulatory costs. I 15 think that's a better world and that's the world that I would 16 recommend to the Commission. 17 MS. NAZARETH: Gus. 18 MR. SAUTER: In terms of market structure, we are 19 very strongly in favor of price-time priority, we're very 20 strongly in favor of auto-ex capabilities. None of the 21 exchanges or other marketplaces really has the exact same 22 model. Nevertheless, they all provide us a place to execute 23 our trades. None of them are necessarily exactly the same 24 model we would prefer. Consequently, since we view them as a 25 marketplace on which we can execute, we think they should be 1 regulated. 2 Whether they're called an exchange or not, they do 3 need an appropriate type of regulation commensurate with the 4 fact that we're executing there. So although you can argue 5 whether or not a model exists that's perfect, I think to me 6 would be somewhat less relevant than the fact that there 7 should be regulation. 8 MS. NAZARETH: Anyone else? 9 Well, maybe we'll move on to the next topic and 10 whiz through the SRO issue. I know this is something that 11 David Whitcomb has expressed some interest in earlier in our 12 program. 13 Let me start out by asking, whose interests are the 14 SROs supposed to represent, and should exchanges or SROs be 15 operated exclusively in the public interest or in the 16 interest of their members and/or shareholders. David. 17 MR. WHITCOMB: Absolutely. SROs should, of course, 18 operate in the public interest, but I think the truth is that 19 they can't. We're all human beings, we all have economic 20 self-interest and when that conflicts with the public 21 interest, the public interest will not necessarily come 22 first. 23 The whole concept of having businesses regulate -- 24 set the rules and regulate themselves dates back to the New 25 Deal. At that time, an early part of the New Deal experiment 1 was to have all kinds of businesses, industrial firms set 2 prices and regulate themselves. That was dismantled. 3 Only in the securities industry did it remain part 4 of stated public policy to have the fox run the chicken coop. 5 It just doesn't work, we don't need it. We have the 6 Securities and Exchange Commission which is the ultimate 7 regulator. It should really be the primary regulator. 8 Yes, people should have to obey rules, brokerage 9 firms should have to have regulation departments to make sure 10 that their people are satisfying the spirit and the letter of 11 the rules. But the ultimate regulator should be the SEC. If 12 we had a single regulator, we wouldn't have regulatory 13 arbitrage, we wouldn't even have long discussions like we 14 just sat through about whether an exchange is an exchange, 15 because the fundamental economic purpose of an exchange is a 16 place where trading occurs. The regulator would then set the 17 rules and if the rules weren't right, the regulator would 18 change them. 19 So life would be a lot simpler. We wouldn't have 20 all of these issues if we would just get rid of the concept 21 of an SRO and let the SEC be the regulator, and obviously 22 give it more resources. 23 (Laughter.) 24 MS. NAZARETH: Thank you. 25 MR. WHITCOMB: Higher salaries, higher budget. 1 COMMISSIONER GLASSMAN: Thank you. 2 COMMISSIONER GOLDSCHMID: Could I ask a question? 3 Could you be more specific as to what's wrong with the SRO 4 system? 5 MR. WHITCOMB: Yes. I think the fundamental thing 6 that's wrong is a conflict of interest, a huge conflict of 7 interest where if you let, let's say the dealers or let's say 8 the specialists set the rules for trading, they will set 9 rules which benefit them and don't benefit the public. If on 10 the other hand, you have a profit-making enterprise run the 11 trading system, in order to make a profit, it will have to 12 attract a lot of trading volume. And it will try to set 13 rules, which obviously are subject to the oversight of the 14 SEC, that will bring in trading volume, that will cause 15 orders to interact with each other. 16 So there's just an inherent conflict of interest 17 when the rulemaker is controlled by a small segment, a 18 constituency, be it the specialist constituency or the dealer 19 constituency. You can solve a lot of problems that way and a 20 lot of the issues that we then worry about will take care of 21 themselves. 22 MS. NAZARETH: Rich. 23 MR. BERNARD: I just -- it's late to do a whole 24 discussion of the theory of self-regulation, but it starts 25 from the idea of the inter-dependency of broker-dealers in 1 clearing and settling trades. And that's where it sprung 2 from and the exchange of course was engaged in self- 3 regulation for about 140 years before anyone thought about 4 inventing the SEC. 5 Secondly, there's 500 odd people who work for Ed 6 who I think most people would agree do one hell of a job and 7 the idea that that system is failing because of the self- 8 regulatory model makes no sense. The basic idea is that it's 9 okay for me to be a thief but I don't want Bernie to be a 10 thief because he'll bring down the public confidence. You 11 harness that in a structure where the broker-dealers self- 12 regulate, it changes the organization, it's the broker-dealer 13 self-regulating, but you have to understand at the New York 14 Stock Exchange, four out of 24 non-management directors are 15 from the floor, another eight are from the upstairs. That 16 means 12 public directors and three management directors are 17 making these decisions about the rules. So the idea that the 18 specialists are somehow setting the rules for this outfit is 19 simply absurd and of course the whole thing is done in a lot 20 of sunlight. 21 So I don't see the case for the theory of self- 22 regulation at all. 23 MS. NAZARETH: Larry, you're with a regulated 24 entity. Do you have a view on whether the current system of 25 self-regulation is working and whether we should move to a 1 hybrid model, a single SRO model or keep the status quo? 2 MR. LEIBOWITZ: Yeah, I think we would clearly be 3 in favor of either a hybrid or a single SRO model. I mean 4 it's clear that SROs do have conflicts of interest, they are 5 often in competition with their very own members. I mean 6 people who represent ECNs would say that NASDAQ is in 7 competition with them right now. And these exchanges or 8 bodies, SROs, can make rules, have access to tape revenues 9 and other ammunition that the people being regulated who are 10 trying to compete with them don't have. 11 I think that would sort of level the playing field, 12 would get rid of a lot of overlapping regulation and 13 confusing regulation about different rules corresponding 14 between exchanges or different markets. 15 MS. NAZARETH: Ed, do you have a view? 16 MR. NICOLL: I don't -- you know, I think generally 17 that we're probably better off with a single SRO than the 18 multiple models that we have now. But if I were the New York 19 Stock Exchange, I certainly wouldn't give up on my self- 20 regulatory organization because, you know, it's all about the 21 integrity of your marketplace and what the New York Stock 22 Exchange wants to do is to ensure that integrity, and has a 23 real incentive to do that. 24 So you know, I don't think that of all of the 25 things that face us right now, that this is something that 1 requires our immediate attention. I think that we obviously 2 are always concerned, as ECNs, with the conflict that we have 3 with both being a customer of and competing with NASDAQ at 4 the same time and being subject to NASD, you know, 5 jurisdiction. And we're always and have been at times 6 fearful that the conflicts that are inherent in that, you 7 know, were perhaps driving at least some of the perceptions 8 of our marketplace by our competitor. 9 But we think that the SEC is perfectly capable of 10 sorting that out and understanding those issues. And I think 11 more or less, we think the SEC has gotten it right. 12 Sometimes that's gone in our favor and sometimes it's gone 13 against us, but I think all in all, it's not -- you know, 14 we're not sort of advocating the kind of radical change that 15 others are in this arena. 16 MS. NAZARETH: Rick. 17 MR. KETCHUM: I'd just like to address the somewhat 18 separate issue from a NASDAQ market only situation because it 19 doesn't reflect how the floor exchange markets have evolved 20 over a long period of time and has handled self-regulation. 21 And that is the fact that the Commission is really 22 on the cusp of a very new environment with respect to NASDAQ. 23 For better or worse, 100 percent of activity in NASDAQ stocks 24 occurred and was reported through NASDAQ and through NASDAQ 25 systems until relatively recently. It now isn't, it now -- 1 the Commission has encouraged competition in a variety of 2 different venues and we've talked painfully about that today. 3 The issue that I do think that deserves a great 4 deal of attention is that the single threaded environment 5 with a single self-regulatory organization allowed for a 6 single order audit trail, a single transaction audit trail, a 7 single review of all activity that occurred and ability to 8 view that through the order entry firm to the marketmaker to 9 the ECN when the order was handled. 10 You are now in an environment where that can't 11 happen and I would say the experience of ISG would not 12 necessarily suggest that it happens as well. It reflects a 13 good means of dealing with the market that is already 14 separate from a regulatory standpoint, but doesn't seem a 15 desirable place. 16 Whatever the right decision is with respect to the 17 existing floor exchange marketplace, I think it's a very good 18 thing for the Commission to consider whether there should be 19 a single self-regulator for NASDAQ securities while you still 20 have the yarn not yet unwound fully and while you still have 21 the ability to maintain a single view. I think that would be 22 highly, highly valuable. If not -- I'll say only briefly, 23 because I only want to repeat myself briefly -- if not, 24 you've got to worry a lot about regulatory arbitrage and 25 about whether each entity is paying for that activity or not. 1 I think Ed said it fairly well, he has the right to assume 2 that. But I think it's a critical issue for the SEC to 3 consider in that case. 4 MS. NAZARETH: Thank you. Well, before we wrap up, 5 I'd like to give anybody here a chance, who feels there's 6 something they just were dying to say but we didn't recognize 7 them at the appropriate moment. 8 (No response.) 9 MS. NAZARETH: Otherwise, I think I'll ask 10 Commissioner Glassman to make some closing remarks. 11 COMMISSIONER GLASSMAN: Again, on behalf of my 12 colleagues on the Commission and the staff, I want to thank 13 everybody for their participation. I for one have found this 14 very enlightening. 15 As I said at the beginning, I didn't expect to get 16 answers and I don't think we got answers, but we certainly 17 understand the issues better and we will have a repeat 18 session on November 12 in New York, a similar platform, a 19 different set of panelists. It'll be interesting to see if we 20 hear the same thing. I expect we will, but maybe not. 21 Again, thank you very much, and we hope to see some 22 of you in New York. 23 (Whereupon, the hearing was concluded at 4:30 24 p.m.) 25 * * * * *