1 1 BEFORE THE SECURITIES AND EXCHANGE COMMISSION 2 -----------------------------------x 3 HEARING RE: REVISED: 5/03/04 4 PROPOSED REGULATION NMS File No. S7-10-04 5 -----------------------------------x 6 InterContinental The Barclay New York 7 111 East 48th Street 8 New York, New York 9 April 21, 2004 10 9:01 a.m. 11 COMMISSIONERS: 12 WILLIAM H. DONALDSON, Chairman 13 PAUL ATKINS 14 CYNTHIA A. GLASSMAN 15 HARVEY GOLDSCHMID 16 ROEL CAMPOS 17 ANNETTE L. NAZARETH, Director 18 Division of Market Regulation 19 ROBERT L.D. COLBY, Deputy Director 20 Division of Market Regulation 21 JONATHAN G. KATZ, Secretary 22 23 24 Diversified Reporting Services, Inc. 25 (202) 467-9200 2 1 P-R-O-C-E-E-D-I-N-G-S 2 MR. DONALDSON: Good morning. There's a ring to 3 that "good morning," everyone. We are missing Commissioner 4 Campos so we'll wait for just a second while we're waiting 5 for him to arrive. Let me just introduce my fellow 6 commissioners, on my right, Commissioner Glassman, on my 7 left, Commissioner Goldschmid, Commissioner Atkins, and 8 Commissioner Campos hopefully to arrive momentarily. 9 Let me begin by saying, on behalf of the Securities 10 and Exchange Commission, I'd like to welcome our panelists 11 and I particularly want to thank you all in advance for your 12 participation at today's hearing on Regulation NMS. Reg NMS 13 was proposed by the Commission for comment in February and 14 the aim is to modernize the regulatory framework of the 15 National Market System. 16 I also want to welcome members of the public and 17 other interested parties to today's hearing. Although the 18 format of the hearing restricts participation and discussion 19 to the panelists, it is our hope that the proceedings will 20 inform and stimulate observation and questions by all who are 21 attending today. We hope, too, we'll have the full benefit 22 of your thoughts via written submissions during the formal 23 comment period and through ongoing discussions and 24 consultations. 25 The National Market System was mandated, as many of 3 1 you know, by Congress in 1975 and laid the foundation for 2 competition among markets trading the same securities, while 3 preserving wherever possible and practical the benefits of 4 orderly interaction and price competition. The National 5 Market System relies on, one, public display of firm 6 accessible quotations in a consistent format; two, the 7 immediate display of orders to trade at or better than a 8 specified price, commonly known as a limit order; three, 9 consolidated publication of market quotation and trade data 10 in a form that is both reliable and widely available; four, 11 linkages between markets trading the same securities in order 12 to foster efficiency and competition; and, five, in the 13 exchange markets, rules that are designed to ensure that the 14 best quote is honored before trades are executed at inferior 15 prices. 16 In general, our system of multiple competing 17 markets has worked remarkably well. Our markets are among 18 the world's most competitive and efficient. Competition 19 among markets has fostered technological innovation and the 20 creation of trading platforms and order routing systems that 21 address the needs of all types of investors, regardless of 22 size and sophistication. Investor participation in the 23 markets has exploded in the last decade. 24 Over the past decade, however, there have been 25 significant developments in our markets that threaten to 4 1 erode the efficient functioning of our National Market System 2 regulatory structure. Advances in technology and trading 3 have led to the development of new alternative trading 4 venues, particularly ECNs. Their instantaneous electronic 5 capabilities often do not interact seamlessly with 6 floor-based exchanges and vice-versa. 7 The number of trading venues has increased, and it 8 has become more and more complicated and costly to access 9 them, as some market centers charge fees to access liquidity. 10 These fees not only raise trading costs, but also cause 11 identical published quotations to reflect significantly 12 different true prices. 13 Revenue from the dissemination of market data has 14 become a significant funding source for SROs, giving them 15 powerful incentives to maximize reported trade volume. In 16 turn, the structure of these incentives has produced 17 problematic practices, such as so-called "wash trades" and 18 "shredding," as ways to generate more "tape prints" and thus 19 more market data revenue. 20 The growth in subpenny trading has increased the 21 risk that trivial quote increments will be used to step ahead 22 of limit orders. Limit orders have already been discouraged 23 by decimalization, and the opportunities to "step ahead" of 24 the trade increase geometrically in a subpenny environment. 25 Regulation NMS as currently proposed grew out of 5 1 considerable preliminary fact-finding efforts and extensive 2 discussions with industry participants and represents the 3 critical next step toward the resolution of these issues. In 4 brief, the regulation in its current form is designed to 5 encourage honoring the best price between markets by 6 establishing a uniform trade-through rule for both exchange 7 and NASDAQ listed securities, with proposed exceptions for 8 slow markets and informed investor opt-outs. It seeks to 9 address the issues of efficient access between markets and 10 inconsistent quote prices by establishing a uniform market 11 center access rule with a de minimis fee standard. It 12 proposes to ban the display of subpenny quotes in most 13 stocks. And it would alter the rules concerning how market 14 data is disseminated and priced. 15 Regulation NMS was intended to advance the 16 discussion by setting forth concrete proposals and then 17 subjecting them to wide-ranging questions and critique. 18 While its publication was a critical step, focusing public 19 comments on these critical market structures, I want to 20 emphasize that the Commission has in no way reached any final 21 decision on any of the proposals' provisions. We are fully 22 aware that these rules are complex and may have consequences 23 not contemplated by the drafters of the proposals. 24 The purpose of this hearing is to broaden the 25 dialogue with respect to our proposals. We intend to use 6 1 this opportunity to probe your responses and ensure that we 2 understand the informed, analytical and factual underpinnings 3 of your observation. We will use this hearing, the written 4 comments that we have already received, the comments we will 5 receive subsequent to the hearing, and the results of a 6 continuing process of consultation, to obtain insights to 7 help us determine whether to adopt or modify the Regulation 8 NMS proposal. 9 To say that market structure issues are complex and 10 controversial is an understatement, to say the least. So the 11 perspectives of the market participants here today and those 12 who submit comments to the Commission are vital to assess the 13 impact of the choices implicit in these proposals. The 14 efficient functioning of our capital markets has intense 15 public policy aspects. For example, the advantage obtained 16 by some market participants at the expense of another subset 17 of participants must be balanced against public policy 18 mandates of fairness to all and not just those directly 19 involved in a particular transaction. 20 So I would ask that you put on your public policy 21 hats for today's hearing. You can help the Commission by 22 putting yourself in our shoes. You can help us by 23 identifying public policy goods that can be achieved by 24 enacting the right and appropriate market structure rule 25 proposal. 7 1 With the reminder that there will be plenty of 2 opportunity during the comment period to assess the fairness 3 of the economic impact of these proposals on certain industry 4 participants, I would ask that in your responses today, you 5 try to put aside your institutional interests, your biases, 6 and even your current business models. I would ask that you 7 set time-worn and all-too-familiar rhetoric to the side and, 8 instead, give us the benefit of your frank and succinct 9 insights about how the proposals and the alternatives to the 10 proposals would serve our capital markets and the investing 11 public from a public policy perspective. 12 With that opening plea or prayer, I believe the 13 time for considering market structure issues is now, and I 14 very much look forward to delving deeply into the issues 15 during today's hearing. 16 Commissioner Campos has arrived. Welcome. So why 17 don't we begin. 18 Annette Nazareth, who heads our Division of Market 19 Regulation, will have a few words to say. 20 MS. NAZARETH: Thank you, Bill. I would like to 21 thank all of our panelists, who so generously agreed to 22 participate today in these proceedings. As you know, we have 23 a very ambitious agenda. We have seven panels and we expect 24 to go to about five-thirty. Each panel is designed to 25 address a specific element of proposed Regulation NMS. 8 1 The first two panels will address the trade-through 2 proposal which, as you undoubtedly know, has engendered 3 perhaps the greatest amount of controversy of the proposals. 4 We look forward to a very constructive dialogue on those 5 issues. 6 I thought I'd very briefly summarize the proposal 7 for you and then introduce our panel. The trade-through 8 proposal would require any market center that executes orders 9 to establish policies and procedures reasonably designed to 10 prevent a trade-through, which is the execution of any 11 incoming order for a security at a price that is inferior to 12 the best price displayed in another market. The 13 trade-through proposal would apply to both exchange listed 14 and NASDAQ securities. 15 At the same time, the proposal would include two 16 important exceptions that are designed to accommodate market 17 centers with different market structures. First, a customer, 18 or a broker/dealer acting for its own account, would be able 19 to consent on an order-by-order basis to having its order 20 executed in one market, without regard to the prices in 21 another market. This exception is designed to provide 22 greater flexibility to informed traders while preserving the 23 average customer's expectation of having his or her orders 24 executed at the best price. 25 Second, an automated market, one that provides for 9 1 an immediate automated response to incoming orders for up to 2 the full size of the best displayed bid or offer without 3 restriction, will be able to trade through a better displayed 4 price on a non-automated market up to a certain amount that 5 would reflect to the greatest extent possible the cost of 6 attempting to access the other market. 7 This exception was designed to reflect the 8 comparative difficulty of accessing market quotes from 9 non-automated markets and to enhance the ability of 10 individual markets to compete fairly with each other. 11 Overall, the proposal is designed to be a practical 12 response to the current criticisms and changes in the 13 marketplace while still maintaining and indeed enhancing the 14 important customer and market integrity goals of best 15 execution and the protection of limit orders. 16 So with that brief introduction, I would like to 17 introduce our panel. I can assure you that they look much 18 better from this angle than the angle that you all are 19 seeing. 20 First on my far left is John Thain, who is the 21 chief executive officer of the New York Stock Exchange; next 22 to John is Bob Greifeld, who is president and CEO of The 23 NASDAQ Stock Market. Next, we have Scott DeSano, who is the 24 head of the equity trading department at Fidelity 25 Investments, then Ed Nicoll, the CEO of Instinet Group; next 10 1 Bob Steel, the vice-chairman of Goldman Sachs and who serves 2 as co-chair of the market structure committee of SIA and we 3 have Ivan Freeman, who is the chief operating officer, 4 institutional equities of Morgan Stanley and, last but not 5 least, Gus Sauter, who is chief investment officer of 6 Vanguard Group. 7 We've agreed prior to these proceedings that in the 8 interests of time, we were not -- we were going to dispense 9 with prepared statements. If any of the panelists have 10 prepared statements that they would like to introduce in the 11 record, we would be happy to do that but we do want to get 12 right on with the debate. 13 Generally, the discussion of the trade-through rule 14 and the questions that we want to address will fall into 15 three general categories, the first being, is there really a 16 need for a trade-through rule; the second being, is there a 17 need for an opt-out and the third, would you need an opt-out 18 if all markets were sufficiently -- provided for automated 19 execution. 20 So I wanted to start with the first group of 21 questions that go to the need for the trade-through rule, and 22 we'll go in order. Again, in the interests of time, I guess 23 if we could limit our initial remarks to two or three 24 minutes, that would be very helpful. 25 So the first questions are, are there any benefits 11 1 from an intermarket trade-through rule beyond those that are 2 produced by the duty of best execution and investor interest, 3 would an intermarket trade-through rule encourage the display 4 of limit orders and would it make it more likely that such 5 orders receive the best price? And we'll start with John 6 Thain. 7 MR. THAIN: Thank you. Good morning. Mr. 8 Chairman, Commissioners, Annette, Bob, and other members of 9 the Commission. Thank you for the opportunity to be here 10 today. 11 The New York Stock Exchange is the largest and most 12 liquid equity market in the world. And one of its 13 fundamental tenets is that orders are executed at the best 14 price. A trade-through rule ensures that all investors, 15 minnows or sharks, get that best price. And investors who 16 provide liquidity to the market by entering limit orders are 17 protected from their orders not being ignored. 18 Individual investors, who unfortunately are not 19 represented here today, overwhelmingly view price as the most 20 important factor when they are executing trades in the 21 marketplace. The trade-through rule encourages markets to 22 compete by offering the best prices. 23 Weakening the trade-through rule, particularly 24 allowing certain investors to opt out of the trade-through 25 rule, will undermine investor confidence and reduce liquidity 12 1 in the marketplace. The investor who provides liquidity to 2 the market by entering the best bid or best offer is not 3 being asked if he or she wants to opt out, and that's the 4 person who the trade-through rule was originally designed to 5 protect. The prices at which trades are reported if the 6 trade-through rule was weakened or people were allowed to opt 7 out, would not reflect the true best prices available in the 8 marketplace, undermining confidence in the reported prices. 9 And allowing trade-throughs will increase internalization of 10 order flow and fragmentation of the marketplace. 11 I believe, Mr. Chairman, as you said, that the 12 interests of America's investing public must come first 13 before the interests of certain large institutional investors 14 or certain proprietary trading venues. Weakening the 15 trade-through rule, particularly the opt-out provision, is 16 not in the best interests of U.S. and global investors, be 17 they minnows or sharks. Thank you. 18 MS. NAZARETH: Thank you. Bob Greifeld? 19 MR. GREIFELD: This trade-through proposal has 20 certainly received more publicity and created more 21 self-appointed experts than those of us who work in the 22 industry could reasonably imagine. But if we attempt to 23 reduce this issue to its essence, we realize that the heart 24 of the matter is, what is the best protection for investors? 25 The best protection for investors is best execution. 13 1 I remember reading in 1993 the Commission's 2 payment-for-order-flow release that defined best execution in 3 terms in addition to price. The Commission specified that 4 liquidity, cost, price impact, accessibility and certainty 5 were other important factors in determining best execution. 6 At that time, I was attempting to build a business 7 with a trade order management system called Brass. And I 8 recognized that incorporating best execution features into 9 that trade order management system was a good business 10 opportunity. It was good for my business, but it was most 11 importantly good for investors. 12 Now, NASDAQ has operated without a trade-through 13 rule for its entire history. The NASDAQ Stock Market is 14 governed by the principle of best execution. 15 If we look at The NASDAQ Stock Market in its two 16 hundred most actively traded stocks, we have a spread of one 17 cent. Our minimum price variation is one cent. We cannot do 18 any better. These narrow spreads mean that we have vigorous 19 limit order competition in our market. The market orders in 20 The NASDAQ Stock Market are the beneficiaries of this 21 vigorous limit order competition. 22 Reg NMS represents a bold step forward by the 23 Commission and as we contemplate the impact, we should be 24 guided by the principles that first, we should not fix what 25 is not broken, and two, let's ensure that we do no harm. We 14 1 understand the desirability for uniformity in the U.S. equity 2 market and we support the Commission's effort with respect to 3 Reg NMS. We believe that careful implementation of a fast 4 market definition with an opt-out provision will enable the 5 NASDAQ market to suffer no harm. We do state, however, that 6 the failure of this Commission to adopt the opt-out rule is a 7 denial of the full benefits of best execution as the 8 governing force in our markets. 9 We believe that the government should only be 10 involved where the government must be involved. The 11 definition of a fast market by the Commission should be a 12 two-paragraph description that mandates automated response 13 and automated update. The investors should determine the 14 finer points of fast. How do we allow the investors to 15 determine the finer points of fast and how do we allow the 16 investors to achieve best execution according to their 17 investment objectives? 18 The opt-out clause allows investors to opt in, to 19 choose which attributes of best execution are important to 20 them for that transaction. The opt-out clause allows 21 investors to opt in to their definition of fast. 22 Without the opt-out clause, best execution is 23 reduced to a single dimension of a concept of price. Our 24 industry's concept of best execution is derived from common 25 law agency principles and fiduciary obligations. Best 15 1 execution is truly a flexible concept that meets the needs of 2 customers under diverse circumstances. Reg NMS is about 3 modernizing our equity markets. Reg NMS is not about rolling 4 back carefully evolved principles that govern many diverse 5 industries in our country. I urge the Commission not to put 6 the concept of best execution in a straitjacket. Thank you. 7 MS. NAZARETH: Scott? 8 MR. DeSANO: It's better to be called a shark than 9 an eight-hundred-pound gorilla. 10 Well, our view on trade-through is a very simple 11 one, I think, and I would contend that we don't have an 12 effective trade-through rule today and would echo a lot of 13 the things that Bob just mentioned with respect to the fact, 14 how NASDAQ has worked without a trade-through rule for all 15 time. 16 There's concern about limit orders out there if we 17 don't have trade-through rules. And I will assure you that 18 we as an institution are more apt to put limit orders on 19 NASDAQ than we are on other exchanges and we are not 20 concerned whatsoever with there not being a trade-through 21 rule. We're a fiduciary and we are charged with doing the 22 best job possible for our shareholders each and every day. 23 That is the best all-in price that we can achieve 24 for our shareholders, not the best individual price. Walking 25 a stock up or down is a painful process, trying to achieve 16 1 execution of large orders. If we have the opportunity to see 2 and capture liquidity at a price two, three, four cents away, 3 but it completes our order, that's a much more effective way 4 for us to do our transacting. 5 Our fiduciary responsibility should allow us to 6 have the ability to make decisions to go where we want, when 7 we want, how we choose to. Investors act in their own 8 economic self-interest all the time. They don't need to be 9 instructed to do so. We were going to get to the opt-out on 10 the next go-around, but the simple fact is we don't see the 11 need for a trade-through rule, we think that it works just 12 fine without one on NASDAQ now and we're about choice. 13 MS. NAZARETH: Ed? 14 MR. NICOLL: In trying to answer the question 15 without referring to our other rhetoric, I think the -- one 16 of the things that I think the Commissioners have to be very 17 cognizant of is that the trade-through rule is not -- does 18 not give people priority for their limit orders, in the sense 19 that it allows -- the trade-through rule allows 20 internalization by all parties cap the limit order. 21 So if the New York Stock Exchange, for instance, an 22 order goes down there and establishes -- a limit order is 23 displayed on the New York Stock Exchange, the rule is that 24 people can trade through it. The rule is not that people 25 have to hit that limit order, because there's no time 17 1 priority. There's price priority but not intermarket time 2 priority. 3 So what we are allowing is parasitic behavior at 4 the price. We are allowing competition at the price. But 5 somehow, we're concerned about the very few exceptions where 6 one market center might want to trade through another by a 7 few pennies. And it seems to me that the notion that some 8 put forth in which they conflate the duty of best execution 9 with the trade-through rule, that somehow, supporting the 10 trade-through rule is about protecting people's limit orders, 11 is disingenuous unless they take the view, which very few 12 people do, that those limit orders have to be respected both 13 from a time-priority basis and a price-priority basis. 14 And there's, I think, as far as I know, only one 15 person on this panel who actually advocates that, at least 16 it's a consistent, in my view, philosophical point of view. 17 If you are concerned -- the real issue here, and some imply 18 that what the trade-through rule protects is the person who 19 placed the limit order, then we ought not allow people to 20 match that limit order in competing venues. Yet we do all 21 the time, and we allow that kind of internalization to take 22 place between competing venues. 23 In our view, you get very little benefit from the 24 trade-through rule and yet you get a lot of harm. You are 25 disincentivizing competition. You are allowing markets that 18 1 can manipulate their best advertised price to effectively 2 freeze out other markets that are pure agency marketplaces. 3 There's a lot of cost to the trade-through rule, and we 4 believe there's extremely little benefit. 5 And for those who advocate the rhetoric that the 6 trade-through rule is about protecting limit orders, let them 7 totally protect limit orders and advocate a time priority as 8 well as price priority. Yet they don't do that. 9 I think that, you know, our view is that the 10 opt-out is a reasonable compromise between the two views. 11 It's a moderate and very modest exception to the 12 trade-through rule. We've seen this kind of modest proposal 13 work in the past with respect to the three cent de minimis 14 exception for the ETFs that has been very successful. Of 15 course, I think it's well known that we would favor total 16 competition between market venues and let investors and 17 traders choose how to execute their orders. 18 But given where we come from and the market 19 structure, we think that the opt out is an important but 20 modest improvement to the current regulatory environment. 21 MS. NAZARETH: Bob? 22 MR. STEEL: Thank you, and it's a pleasure to be 23 here. I speak to you today on behalf of the Securities 24 Industry Association so, as I've tried to develop a 25 perspective that represents all 600 members of the SIA, I 19 1 have some sympathy for the task that you've assigned yourself 2 to developing a single view. 3 I think in reality, though, while I certainly will 4 disappoint some members of the SIA, that the consensus is in 5 favor of a trade-through rule for the reasons that have been 6 articulated by the far right of my panel here today; that 7 basically, and we'll get into this in the next question, that 8 there should be some exceptions for choosing a different 9 alternative or opting out, that in general the SIA favors 10 that the trade-through rule protect the book as it is. 11 MS. NAZARETH: Ivan? 12 MR. FREEMAN: Thank you. Thanks for having us 13 here. We operate as a broker/dealer in the U.S. equity 14 markets and as a broker, we represent both individual and 15 institutional investors and as a dealer, we can do capital 16 trading in virtually every market in the U.S. and in every 17 instrument. 18 So against that backdrop, and really in the spirit 19 of what Chairman Donaldson asked us to do, we put on both a 20 public policy hat and in our interest in evaluating efforts 21 to enhance the efficiency and fairness of the U.S. equity 22 markets. And we approached this without a bias toward a 23 particular exchange or market center. Instead, we really 24 tried to evaluate what is the impact on the market itself, 25 and on the ability for investors to accomplish what they want 20 1 to accomplish in the marketplace. 2 Short of forcing interaction of orders to a single 3 venue, we believe that a National Market System in which 4 participants are economically induced to compete vigorously 5 for market share on a level playing field is an attractive 6 objective. I think the overarching objective of an enhanced 7 national market still should be to leverage these 8 state-of-the-art meetings that we have at our disposal today 9 to put in higher standards of best execution, so the National 10 Market System should ensure that there is immediate 11 accessibility of better prices on an away market. But that 12 also requires that we recognize that speed is an important 13 element in obtaining the best execution. 14 Ultimately, the informed customer's freedom of 15 choice needs to be respected, and I'll defer the comments 16 with respect to the opt-out in specifics to the next question 17 round. But I think that it is clear that from our 18 perspective, the proposed Regulation NMS does move us closer 19 to a National Market System objective that is attractive. 20 A certain level of trade-through protection should 21 be available to those placing market and marketable limit 22 orders so long as you have clear standards of accessibility 23 and that those standards are strictly enforced. 24 We think that, as we read the rule, we don't view 25 this as a safe harbor. We don't view it as a substitute for 21 1 best execution. It doesn't alleviate the burden on somebody 2 handling an order on behalf of a customer to meet its best 3 execution requirements. So we think that standard is still 4 going to hold. But in the context of the question that's 5 asked relative to whether the intermarket trade-through rule 6 would produce benefits beyond those that due to the best 7 execution and investor self-interest already do, we would say 8 it does. It quantifies something that creates a minimum 9 standard and creates competitive behavior among market 10 centers that allows market orders to be obtained in a manner 11 in which they may not have been in the absence of the rule. 12 MR. SAUTER: I'd like to thank for you having me 13 here today. When Ed Nicoll referred to one person in the 14 crowd who believed in price/time priority, I hope he was 15 referring to me. We have been consistent supporters of the 16 concept of price/time priority. In fact, we have been 17 consistent supporters to the idea of a central limit order 18 book; but I have no delusion that we're going to be able to 19 argue that one today so I'll leave that one alone. 20 Quite naturally, all investors want to minimize 21 transaction costs and how do we do that? By having the most 22 liquid market possible. So our view is, try to create a 23 marketplace that is -- that does provide the maximum 24 liquidity; in other words, provide the incentive to put limit 25 orders on the book. Price/time priority does create that 22 1 incentive. At the same time, we think that a trade-through 2 rule is necessary to preserve price/time priority. 3 I would note that there's a natural tendency for 4 investors to not place limit orders. Who wants to convey 5 that much information to the marketplace? It's like playing 6 a poker game where you're revealing your hand. At the same 7 time, you're also providing a free option to the marketplace. 8 You're willing to allow someone to put an order to you. If 9 they want to exercise against you, they have that right. 10 You're standing there always willing to be traded against. 11 So that there is a natural tendency not to place 12 limit orders, a natural tendency not to create liquidity in 13 the marketplace which is, I think we all agree the optimal 14 benefit to investors is to have a totally liquid market. 15 Therefore, we think that we need to create some 16 sort of incentive to place limit orders to get around that 17 natural tendency to not placing orders. We need to have some 18 form of price priority, time priority ideally as well. But 19 we think a trade-through rule is necessary to ensure that the 20 markets are not totally fragmented, to ensure that an 21 investor that does place the best limit offer or bid is 22 executed first. We think it would be a tremendous 23 disincentive to place limit orders if in fact other orders 24 can be executed around you. 25 We think that frequently the new entrant into the 23 1 marketplace, the entrant who is taking liquidity out of the 2 market, is favored. We think that the entrant that puts 3 liquidity into the market, that places a limit order, should 4 be favored. So while we would say that there are many times 5 where we would like to access all of the liquidity, say, 6 perhaps away from the NBBO, we believe that would be 7 detrimental to the marketplace. We might be best served at 8 that instant in time taking a price away from the NBBO by 9 being able to fill a hundred-thousand-share order or 10 half-a-million-share order. But we think in the long run 11 we're creating a disservice to ourselves if we do skip the 12 NBBO and hit the maximum liquidity provided at a different 13 price, an inferior price. 14 We've created a disincentive to put limit orders on 15 the books. If the person who established the NBBO is 16 ignored, they won't come back. They won't continue to place 17 limit orders. 18 What we -- we, ourselves, place fewer limit orders 19 than we used to because of some of the recent changes in the 20 marketplace which we actually view as being positive, that 21 being decimalization; but at the same time, it showed many of 22 the problems in the marketplace. So we do place fewer limit 23 orders now than we used to. 24 I would note some of the comments about limit 25 orders not needing a trade-through rule in the NASDAQ 24 1 marketplace. I would point out that in fact, NASDAQ market 2 makers have a best-execution requirement. And I believe if 3 you ask most of them, they think that means that they have to 4 honor the NBBO. So in effect, you do have a quasi limit 5 order requirement in the NASDAQ marketplace. 6 And I would also note that ECNs are designed to 7 honor price-time priority; at least price-time priority is in 8 their own system but they also seek best price outside of 9 their system. So I don't think that comparing the NASDAQ 10 marketplace, that, quote-unquote, does not have a limit 11 order -- I'm sorry, a trade-through requirement, with the 12 listed markets, and saying, "Well, one market functions fine 13 without trade-through rule, therefore, the other market 14 doesn't require it as well," I think that in effect, there is 15 a quasi trade-through rule being honored in the NASDAQ 16 marketplace. 17 So again, we think that limit orders should be 18 encouraged and we think the best way to encourage them is to 19 make sure that they are protected and given price priority if 20 not time priority as well. 21 MR. DONALDSON: I'm sure that my fellow 22 Commissioners have got questions they would like to ask. I 23 suspect maybe all of you have got questions you'd like to ask 24 of each other. Let's start and see if the Commissioners have 25 anything. 25 1 MS. GLASSMAN: Yes, I have a couple of questions. 2 First, on our proposal to except the trade-through rule for a 3 slow market so the trade-through would only be for fast 4 markets, what is fast enough? What does "fast" mean? 5 MR. NICOLL: If I could jump in, I think that's one 6 of the problems. At Instinet, we've worked to reduce our 7 response time from milliseconds to microseconds. We now have 8 total turnaround times under a millisecond. For those 9 uninformed a millisecond is a thousandth of a second. 10 We're not -- we're not, you know, we're leery of a 11 rigid definition of what is and what is not a fast 12 marketplace. We think, for instance, if you were to say that 13 a fast marketplace is one that responds within seconds, that 14 would create not a, you know, not a ceiling, but in effect a 15 floor. And, you know, and you would have all markets 16 basically responding to a second, because there would be, you 17 know, it would give everybody the ability to be part of this 18 large trade-through regime and yet have the slowest 19 marketplace. This is essentially the problem. You are 20 looking at one simple criteria of competition, which is 21 price. 22 Now, without -- without, you know, preaching, I 23 mean, I think it is -- I agree with the New York Stock 24 Exchange. At the end of the day, it's all about price. 25 Customers want to get the best price. 26 1 The question is, how do they achieve that? Do they 2 achieve that simply by chasing the best advertised price or 3 by interacting in the marketplace according to various 4 criteria that they believe will get them the best net price 5 in terms of executing their orders? And if we are to simply 6 reduce competition to the best advertised price, as we do in 7 the trade-through rule, then we have a -- we have a set of -- 8 of perverse incentives to actually go to the slowest, to 9 create the slowest market within the rules. 10 So whatever you define as -- as an automated market 11 is effectively going to be the floor and it's going to, in my 12 view, it's going to stop innovation from people trying to 13 offer better, faster, different kinds of marketplaces. And 14 this is the -- my worry about the trade-through agreement. 15 MR. DONALDSON: If I could just jump in on part of 16 your question, that is, a one-second market, two-second 17 market, what's the need for a micro or millisecond execution 18 capability? How much of the marketplace is affected by that? 19 MR. NICOLL: Let me just answer that and then I'll 20 let others answer. We are working on microseconds because we 21 want our system to be completely responsive and real-time, in 22 the most contentious environment -- 23 MR. DONALDSON: My question is, responsive to whom? 24 MR. NICOLL: Everybody. 25 MR. DONALDSON: How much of the overall market is 27 1 interested in micro or millisecond trading? 2 MR. NICOLL: I think it's important to everybody 3 because when you reduce response times, and you have massive 4 spikes in demand into the system, those reduced response 5 times give everybody what amounts to a seamless interaction 6 into the marketplace, even in the fastest time. 7 You're right, a trader right now cannot even tell 8 the difference between twenty milliseconds where we used to 9 be, and three hundred microseconds where we are now. 10 But when the Fed chairman has something to say, and 11 everybody wants to enter the market at the same time, 12 reducing response times to microseconds is important to 13 eliminating latencies in the system due to overall demand 14 into the system. 15 MR. DONALDSON: Do you have any idea how much of 16 the marketplace is affected by that? That's my only 17 question. 18 MR. NICOLL: I think it's most important at the 19 time when the most amount of people want to access the market 20 immediately. And I don't know -- I don't know. I don't know 21 if we are actually getting a bang for a buck in terms of 22 trying to reduce the response time as much as we have. I do 23 believe that we ought to be able to try give kinds of systems 24 and compete on different criterias of value and let 25 traders -- I'm trying to get Scott's business. I'm not quite 28 1 sure what it is that I've got to do to get Scott's business, 2 but I think you ought to let competition play out and let 3 Scott and Gus -- let us know what it is that we have to do to 4 meet the demands. And the demands of the Ameritrades and the 5 customers around the world. 6 Just one last thing. I made a lot of money 7 offering a system that guaranteed retail investors 60-second 8 response times. Last, I think, two weeks ago, the 9 competition in terms of response time is now down to two 10 seconds. One of the on-line brokers is guaranteeing a 11 two-second execution. Those people are responding to the 12 needs of individual investors and the demands for speed by 13 individual investors. I don't know how well it's going to 14 go, Mr. Chairman, I honestly don't. 15 MR. FREEMAN: I think this question actually goes 16 to the heart of one of the question that outlines the opt-out 17 issue. Because the concept of fast or the concept of 18 autoexecution is to some degree in the eyes of the beholder. 19 What does that constitute? And we could be talking about the 20 turn-around time from the time an order actually hits a 21 market center, is processed and turned around, but from an 22 investor's perspective, you probably also have to consider 23 the trip from the investor's point of decision to the market 24 center. So really you need to look at the time between, and 25 understand where the access can accommodate peak volume or 29 1 whether or not in fact there's some latency as a result of 2 that. 3 And I think to a large degree different investors 4 will have different priorities in terms of how they look at 5 what is fast and in particular, how fast impacts their 6 meeting their overall objective of what they would consider 7 best execution to be in the context of their overall rules in 8 the marketplace. 9 MR. SAUTER: Our view of fast is immediate. It's 10 faster than humans can do. It's an automated process. We 11 would like to be able to access all of the liquidity of the 12 marketplace to fill an order. That might mean that we would 13 have to jump from venue to venue and perhaps hit ten 14 different venues in order to get our entire order filled. In 15 a volatile marketplace, if there's a second delay or even a 16 half a second delay, at a given exchange, that may mean that 17 our order goes unexecuted. We believe that best execution 18 has two components. I mentioned earlier best price and 19 certainty. 20 And with speed, you get certainty. We can look and 21 see what the national book, quote-unquote "national book" 22 looks like at this instant in time. If we had linkages in 23 place, we could hit every -- every offer out there at a very 24 instant in time. 25 That is, for us, an ideal. So a fast market is an 30 1 instantaneous market to us. 2 MS. GLASSMAN: I have a lot of questions but I know 3 we have limited time so I will just ask just one that follows 4 up on what George just said. 5 The debate in some sense has been characterized as 6 speed vs. price, speed vs. best price. But isn't it really, 7 speed is certainty, certainty now, vs. uncertainty later, and 8 so the slower market means less certainty that you'll 9 actually get that advertised price? Isn't that the issue? 10 MR. SAUTER: Yes, it is, in our view. 11 MR. DeSANO: Let me also add one more component to 12 that. Speed, certainty and volume. Your buying a couple of 13 hundred shares here and there is not a lot of use to a lot of 14 institutional investors. So we look at it in three 15 components; the all-in price, and buying is a very key 16 component. Speed and certainty are equivalent. 17 MR. GREIFELD: I think the global issue is again 18 about choice. But we believe the best execution allows the 19 investor to have choice of what's important for them. 20 I tie back to Gus' comment. He said that in the 21 NASDAQ market, we essentially operated with a trade-through 22 rule in place, even though none exists. And that is a direct 23 result of the fact that we're governed by best execution. So 24 we talk about standards of best execution, we ourselves 25 include price as first among equals. It's not the only one, 31 1 the only standard. But it's first among equals, and that's 2 how the NASDAQ market operates today without a trade-through 3 rule. 4 In the rules contemplated here with Reg NMS, to get 5 us back to the competitive dynamic that exists, we think we 6 need to have the opt-out. The opt-out represents the opt-in 7 opportunity for investors to choose; choose based upon 8 various criteria that are outlined under best execution that 9 have been supported by this Commission for a long period of 10 time. 11 MR. DONALDSON: Anyone else? 12 MS. NAZARETH: Why don't we -- 13 MR. GOLDSCHMID: Annette, let me lead you into 14 where I think we've got to go. On the opt-out, will it 15 fragment the market, will it give too much advantage to, 16 whether you want to call them sharks or whales, gorillas, 17 pick your own. But tell me what's really going to happen. 18 MR. THAIN: Commissioner Goldschmid, let me add on 19 to some of the prior panelists and get your answer to the 20 question. First of all, price is not the only criteria, and 21 one of the reasons why we support the Commission's position 22 on fast market/slow market is that we understand that if 23 there is a dramatically different execution time, that 24 markets do move very quickly. And so speed does make a 25 difference to people. 32 1 And I also agree, though, that price should be 2 first among equals. So the fast market/slow market 3 definition, and the differentiation between those markets, 4 and I do think there's issues about exactly how you define 5 automatic fast markets, I think the simplest way is that they 6 provide automatic execution, which basically means 7 computer-to-computer execution, and I frankly don't think it 8 makes very much difference whether we argue about 9 milliseconds or microseconds or anything like that, as long 10 as you get a direct computer-to-computer execution which, by 11 the way, also gives you the certainty. 12 But once you have fast markets, once you have 13 computer-to-computer certain automatic execution, there is no 14 reason why you shouldn't go to where the best price is. And 15 if you allow people to opt out after you get comparably fast 16 markets and comparable certainty, and by the way, the other 17 thing the NASDAQ would like, which I'm sure he'll say 18 eventually, is anonymity, which he also gets with an 19 electronic execution, a computer-to-computer execution. Once 20 you get to that stage, there is no good reason and no good 21 public policy reason why you should not execute at the best 22 price. 23 And the ability for institutions or "informed 24 investors" to opt out after they have that same execution 25 capability will lead to internalization of order flow, which 33 1 is bad for the market, it will not protect the limit orders, 2 which you've heard from a number of people, is a very 3 important thing, and ultimately, will lead to fragmentation 4 in our marketplace. 5 MR. DONALDSON: I'd like to ask just a general 6 question of, Scott, you and Bob have talked about, I think 7 your words were, "Executing an order in terms of what's 8 important for that order." And I guess my question is, is 9 there an obligation, a fiduciary obligation if you want to 10 call it that, an obligation for a marketplace itself to look 11 beyond your interests in a particular transaction toward the 12 interests of the marketplace itself? 13 In other words, is there any obligation to think 14 that this may be best for me at this particular moment to opt 15 out and hit the best execution, and in so doing, perhaps I 16 damage other parts of the market that aren't involved in my 17 transaction to get a limit order that gets passed? 18 MR. GREIFELD: A very good question. I think the 19 question results from the concept of an existing limit order 20 that gets opted out. And the thought seems to originate from 21 the fact that an existing limit order that's opted out is 22 frozen in time and is not, in itself, portable. 23 The way I look at this discussion is, we have 24 competing self-interest. And the competing self-interest is 25 governed by best execution. So that limit order that is sent 34 1 to a market is sent by a broker who has an obligation of best 2 execution for that order. And that broker has to evaluate, 3 dash 5, dash 6 status, to make sure he's sending it to the 4 proper market center. And I think the fundamental fear of 5 the New York Stock Exchange and others is that an existing 6 limit order will choose the best outcome for itself and has 7 the ability to move in the new world. 8 So you have the order that can be opted out, and 9 you have the new order coming in. But they both have their 10 self-interest under best execution, and they both will come 11 to arrive at what's best for that order. So the fact is, if 12 I send an order to an exchange and I'm not getting a fill, I 13 have the ability to move that order. And I was going to move 14 that, I was going to say picoseconds, which is billionths of 15 a second, but we're not there yet, but I can move that very 16 quickly. So that order that is there is portable and can 17 move the milliseconds and get to the place where you want to 18 go. 19 MS. NAZARETH: Bob, you could answer Commissioner 20 Goldschmid's questions, maybe everybody can answer both sets 21 of questions, and I think John, if you want to jump in on the 22 last part -- 23 MR. GREIFELD: Make sure I answered it properly. I 24 wanted to make sure I got the full benefit. 25 MR. GOLDSCHMID: Fundamentally, Bob, it goes to the 35 1 heart of what opt-out is going to be, will it fragment, will 2 it help out the big guy and not the small guy, what's really 3 effected? 4 MR. GREIFELD: I think opt-out will be used less 5 and less in time as the market centers learn to compete, and 6 we saw this happen in the NASDAQ marketplace after Reg ATS. 7 Certain ECNs were actually faster to respond. And we had to 8 learn that customers were measuring milliseconds and making 9 intelligent routing decisions. And there was a big gap of 10 performance. Ed mentioned 22 milliseconds. They might -- 11 somebody's at 22 milliseconds, everybody else was at, you 12 know, a hundred, 150. 13 Within twelve months, it wasn't days and it wasn't 14 weeks, but within twelve months all the competitors were in 15 around 22 milliseconds and at a point it doesn't matter, the 16 speed difference, and I would say once you're into 17 milliseconds, the difference between 20, 25 and 30 really 18 doesn't matter to very many investors at all. So what you 19 saw is investors responded to what the customer wanted and 20 what they did is improve their systems. And the incidence of 21 opt-out in the NASDAQ marketplace declined as competitors had 22 to respond to the market forces. 23 We're saying what the opt-out is, is the market 24 forces will be released to make sure that we all have to 25 compete and compete effectively. 36 1 MR. GOLDSCHMID: You say it as basically a 2 corrective to keep flexibility and pressure on, if I 3 understand. 4 MR. GREIFELD: The way I look at it is, right now 5 you can define a fast market literally in two paragraphs. If 6 you don't do it that way, if you don't have an opt-out, you 7 have to write a book. You have to write a book of rules, and 8 writing a book of rules is difficult but it can be done. But 9 the really insurmountable task is to follow and track who is 10 following that rulebook, and you have to follow that not in 11 an audit once a week, once a month. You have to follow it 12 every second of every single trading day. And Ed points out 13 that there are certain points in the trading day where the 14 volume just explodes. You have to have a whole set of rules 15 and a whole set of tracking for that. The opt-out will let 16 the market determine who is actually providing the fast 17 market. 18 MR. GOLDSCHMID: After yesterday, you think there 19 can be an explosion? 20 MR. DeSANO: I would add in on the fast market 21 notion that we would like to be able to access multiple 22 levels of the book, in order to be considered a fast market. 23 Also with respect to where this fragment, I mean, 24 SuperMontage has about 18 percent market share and Mr. Thain 25 has a much higher market share and we see very little 37 1 difference in our day with respect to trying to get done what 2 we need to get done given that disparate market share 3 difference. 4 What else do we have? 5 MS. NAZARETH: Opt-out. What does the opt-out do? 6 MR. DeSANO: Well, the opt-out, there are different 7 constituencies who obviously have different purposes and 8 needs. And you spoke to, you know, if we need to do 9 something, do we worry about hurting other people. That 10 would be really difficult for me to go in to our board of 11 trustees and say, "I really didn't take care of our 12 shareholders because I was afraid I was going to damage the 13 marketplace." 14 That's an unacceptable explanation from me to the 15 board of trustees. And that's why we get involved in this 16 process, to try to worry about those issues in this kind of 17 setting so that we work in a viable market structure that's 18 good for all. 19 But the opt-out is about choice. And it's about 20 having informed investors go where they want to go, when they 21 want to go, in a manner that suits their needs. 22 To have this dictated to me precludes you from 23 being able to achieve best execution. Best execution is best 24 all-in price to us. That's what it is. And if I have to do 25 a hundred trades at the best price at that moment to get a 38 1 certain order done as opposed to ten trades, I guarantee it's 2 going to cost a lot more money for our shareholders. And so 3 we need to be able to exercise our judgement to go where we 4 want and when we want using our fiduciary responsibility and 5 our understanding of best execution. And that's why I 6 believe the opt-out is absolutely necessary in this process. 7 MR. DONALDSON: Let me just push on that now. You 8 know, in terms of your trustees. Assume for a minute what -- 9 an extension of what you're saying is, we ought to be able to 10 trade anywhere, any time that suits your particular purposes 11 at that moment. 12 Supposing that the result of that was people's 13 unwillingness to place limit orders, and supposing the day 14 after tomorrow, one of your trustees asks you, "What's 15 happened to the marketplace? There's no limit order, there's 16 no market that I can refer to; has your doing anything you 17 want for a period of time done this to our market?" 18 MR. DeSANO: Essentially that's where we are now. 19 We don't place limit orders because there's no real 20 protection out there. We're going through this in the hopes 21 that by having each venue have to compete, you know, head to 22 head, just pure competition, that we get to a world where he 23 will be comfortable placing limit orders. The situation is 24 not a comfortable one now and I think this is a means. 25 Allowing the opt-out is a means to create and foster 39 1 competition between venues to get order flow and in the end, 2 I think that you'll see far more limit orders show up. 3 MR. DONALDSON: What change, now, would make you 4 more comfortable placing limit orders? 5 MR. DeSANO: Everybody gets to trade by the same 6 rules. You don't have a geographic advantage. Our orders 7 don't get broken up. We don't get price improved. There's a 8 lot of -- 9 MR. DONALDSON: You're talking about access? 10 MR. DeSANO: Access. I think time priority by 11 venue is important. But not across venues. 12 MR. NICOLL: Can I try and really touch on your 13 concern? Really, what you're talking about is that if we 14 allow people to trade through limit orders, that people will 15 stop placing limit orders. And while it might be in the -- 16 in an individual's interest to trade through an order, that 17 the -- at the Commission, there's a public policy issue here 18 that we want to protect the incentives market-wide to 19 place -- limit orders, getting people -- I mean, placing 20 limit orders is getting people to reveal their reference 21 points, you know? I mean, that's what really market 22 structure ought to encourage, it's getting people to openly 23 say what they want to try and sell. And we need to create a 24 set of policies to the greatest extent possible that 25 incentivizes people to reveal their reference price in the 40 1 marketplace, to, in effect, place limit orders. 2 And you're worried that an opt-out provision to the 3 trade-through rule will create disincentives for people to 4 place limit orders and yet, what we're really talking about 5 here now, okay, is a world in which stocks are trading in 6 increments of a penny, okay? Not eighths and quarters and 7 when the trade-through rule was passed thirty years ago, but 8 a penny. 9 We are allowing people, and there's no talk, other 10 than Gus, there's no talk of a CLOB anymore, there's no talk 11 of time priority. So although the New York Stock Exchange 12 talks about protecting limit orders, they are quite happy to 13 compete with the Chicago Stock Exchange by matching the 14 Chicago Stock Exchange's limit orders and not sending their 15 order to the Chicago Stock Exchange. So really what you're 16 saying -- so we don't have that now, largely. We have an 17 occasional -- an occasional trade-through and there are 18 literally hundreds of traders on the streets now, as we've 19 reduced the cost of trading, who will arb markets for a 20 penny. 21 So do we want to create this structure with all the 22 disincentives for competition to worry about that, the rare 23 occurrence when markets trade through each other, while at 24 the same time we allow them to ignore each others' limit 25 orders simply by matching them? One of the reasons why 41 1 people don't place limit orders now in particular venues, 2 because another venue can simply match those orders and 3 internalize against that. 4 So the fact of the matter is, we don't have a lot 5 of protection for limit orders. We have a situation in which 6 venues are competing. The people who place limit orders are 7 the people who are the power. They are -- they are the 8 people that venues compete for. And I think the best from a 9 public policy position is, you ought not to worry about the 10 occasional trade-through for a penny. What you ought to 11 worry about is incentivizing these various venues to compete 12 with each other so that you know, we -- the competition 13 settles in on that venue which meets the needs of most 14 investors. 15 There's -- this is not a world in which people are 16 trading through each other for twelve-and-a-half cents 17 anymore and I submit to you that there's rare -- we rarely 18 see trade-throughs that -- that the Commission would need to 19 be concerned about in the NASDAQ marketplace, where there is 20 no trade-through rule 21 MR. STEEL: I think there are two questions. Let 22 me try to address them specifically. One, I think from 23 Commissioner Goldschmid was, the issue, would you end up with 24 a fractured marketplace, and I think it's that consternation 25 that the SIA interprets that we should have a narrowly 42 1 available opt-out provision and we should do this in a more 2 patient way as opposed to an opening where there's more, a 3 greater degree of flexibility with regard to opting out. So 4 I think that addresses that issue and I think that is a 5 consternation, and so the fast-slow is our first step in 6 trying to think about that, and so that I think is a public 7 policy issue. 8 I think the second issue that the Chairman raises 9 is one that really has to be the -- the big issue for all of 10 us, and that is that it's quite imaginable to me that some of 11 us are going to have to have a reduction in our ambition 12 because it's for the public policy good. And I think that to 13 the extent that you help proscribe that and say that these 14 are the rules of the road that are required for the benefit 15 of the marketplace, it will be helpful to all because Scott 16 does have a fiduciary duty. And to have him be unclear as to 17 how to do that is a little bit unfair to him, and I think the 18 dialog with his trustees is a challenging one for him to 19 have. And so I think the conclusion would be that a 20 relatively proscribed opt-out, and then with clarity about it 21 so then people can begin to play within the defined rules 22 would be I think a policy perspective that addresses both of 23 your questions, I think. 24 MR. FREEMAN: I think the point that was made about 25 freedom of choice is an important one and that we support the 43 1 view that -- 2 MR. DONALDSON: Get closer to your microphone. 3 MR. FREEMAN: I'm sorry. That the customer should 4 be a final arbiter of what constitutes best execution for him 5 or her in the context of the markets and to resolve the issue 6 of how fast is fast, et cetera. But I think we should also 7 look at the existence of antitrade-through rules or 8 trade-through protection and the behavior modification that 9 that engenders on the same aspect of behavior modification 10 with respect to the placement of limit orders, and I'll 11 elaborate on -- for those who aren't familiar, that means 12 that they -- they move around in time and space and price and 13 react to the market as well. 14 They are moving. And you have to presume that the 15 impact of a trade-through rule on the execution of market 16 orders, and marketable limit orders, is also going to have an 17 impact on the behavior of how limit orders get routed, and 18 that this concept that there's a best execution obligation 19 for how to route limit orders is quite valid, that if 20 somebody is getting traded through routinely because maybe 21 that was perceived as a slower venue, there will be a 22 decision to be made, is that the right place to place limit 23 orders. So I think we have to be cognizant of both the 24 interaction of the two as well as the dynamics of limit 25 orders. 44 1 One other point that I would make is that we should 2 also recognize that even in the absence of an opt-out rule, 3 just the dynamics of the marketplace are going to lead to 4 things that will be trade-throughs on the tape. There's lots 5 of market dynamics today including capital commitments if you 6 traded intraday and need to get put up on the tape because we 7 want to see all trades that take place when they get 8 reported -- by the way, there's a 90-second requirement for 9 posting trades if the first place, so you can see latency in 10 trades that are not as late, but could look to the limit 11 order placement as trade-throughs. 12 And there's also lots of activities that take place 13 in the cash markets that is reflective of activity that was 14 going on in the derivative market. So you can see that 15 executions are taking place where they are at the market, but 16 they are against, not cash, but against derivative 17 transactions, and may not look like they are meeting a best 18 execution standard from the perspective of a trade-through 19 rule but are quite valid trades and may be perceived by a 20 limit order placer as a trade-through. 21 So I think we need to tolerate that even in the 22 absence of any opt-out protection, or even in the other 23 exemptions, and given the other exemptions that are in the 24 rule with respect to contemporaneously trading, there will be 25 appearances of trade-throughs. And that's a factor in the 45 1 market today and you'll see that continues as the 2 trade-through rule change goes through as proposed. 3 MR. SAUTER: To answer Chairman Donaldson's 4 question first about fiduciary obligation and the requirement 5 to minimize transaction cost, we would -- we believe there's 6 a very different set of circumstances in the short run vs. 7 the long run. If you try to minimize your transaction costs 8 on one trade, you can do that by perhaps opting out and 9 leapfrogging the NBBO. 10 We think that that perhaps does not minimize -- we 11 believe strongly it does not minimize transaction costs in 12 the long run and our goal is to minimize our total 13 transaction costs. So we do worry that the opt-out rule 14 really does lead to fragmentation in the marketplace as well. 15 We would like to see a marketplace that ultimately is 16 somewhat of an integrated National Market System that would 17 enable investors to literally ping back and forth across the 18 National Market System hitting limit orders as they go on 19 their way to getting complete execution. 20 If you had that instantaneous type of system set 21 up, there would be no logical need for an opt-out rule. Who 22 would chose to opt-out if you could take everything on the 23 way up? 24 So we strongly oppose an opt-out rule because we 25 believe that in the long run, that does disincent limit 46 1 orders, even though it might enable us to get a better fill 2 in the very short run. 3 We would support the de minimis exception in the 4 short run because we believe that actually sunsets on itself. 5 The de minimis exception really would ultimately require all 6 markets to become automatic, fast, if you will, and once that 7 happens, it does create a national market book, if you will, 8 and therefore, would sunset on itself. 9 MR. THAIN: May I just add one thing? Mr. 10 Chairman, there are markets that don't have these kinds of 11 rules. Many of the fixed income markets don't have these 12 kinds of rules. And those markets are characterized by lack 13 of transparency, by much wider bid-ask spreads, by no 14 centralized source of liquidity, and frankly, by 15 broker/dealers being able to make a lot of money in them. 16 And so I think that there are examples of what 17 happens when you don't have these types of rules but I don't 18 think that's good for the U.S. equity markets and I don't 19 think that's good public policy. 20 MR. CAMPOS: Can I ask what the status is of this 21 ideal world of interconnectivity, is that something that can 22 be accomplished relatively soon, are we -- I mean, I keep 23 hearing about this pinging back and forth from market to 24 market. But yet, when we get down to details, it doesn't 25 seem to really be there. 47 1 MR. SAUTER: It's pretty well there in the 2 ECN/NASDAQ marketplace. 3 MR. THAIN: I think, Commissioner, probably -- I 4 don't -- there aren't many things probably all of us would 5 agree with, but hopefully this is one, and we'll see. 6 I think most of us will agree that ITS in its 7 current construct doesn't work. And so it needs to be fixed 8 so that you do in fact get good interconnections between the 9 marketplaces. 10 MR. ATKINS: Well, actually that was one question I 11 wanted to give back because I have a lot of questions about a 12 lot of this. But to get to a practical question, with 13 respect to what is it right now, and we're talking about the 14 trade-through rule, that at least when we were looking at 15 this with respect to the ETFs last year, year-and-a-half ago, 16 whenever it was, and then what was seen through some of our 17 enforcement actions, I really wonder what does exist right 18 now with respect to the trade-through rule and with respect 19 to ITS, you know, a lot of people say that ITS was meant from 20 the beginning to not work. And so we're sort of seeing that 21 now. 22 So I'm wondering what, as far as each of your 23 perspectives go, whether or not the trade-through rule is 24 effective right now in the markets. 25 MR. NICOLL: I think it's ironic, I agree with -- 48 1 this is where I just have to say how ironic it is, with my 2 good friend Gus, we're saying that we had an effective 3 national market system in NASDAQ where there is no 4 trade-through rule, which has created competition where there 5 are mechanisms in which we are accessing each other's markets 6 at lightning speed and an effective de facto CLOB as a result 7 of that and I would argue, Gus, and I would argue to the 8 Commission, that that's because of competition that is played 9 out without the rigid hand of regulation requiring people to 10 go to one venue vs. another. 11 And I don't think, I think there are thousands and 12 thousands of trade-throughs within the listed -- between the 13 markets, I mean, I think, you know, there are -- there are 14 trade-throughs in the ETFs all day long and the markets are 15 apparently locked all day long, yet they shouldn't occur 16 because things -- because when you trade a hundred million 17 shares a day, which is what the QQQs trade, every day, 18 whether we like it or not, it's trading a hundred million 19 shares a day, it's so fast that it's almost impossible to 20 have an orderly National Market System where everybody goes 21 in absolute order, even with computers, even with a 22 trade-through rule overlying it. 23 And really what we have is a kind of chaos of 24 activity which has in fact, you know, lowered transaction 25 costs. The QQQs, since we created a de minimis exception in 49 1 the QQQs, we've seen the spread go from an average of six 2 cents where, today on Inet, the spread is three tenths of a 3 penny. That is effectively trading with no -- with no cost 4 between the bid and the offer. And that kind of -- of 5 competition and the interplay of technology -- 6 MS. NAZARETH: Ed, I have to add I have no 7 recollection of the spread on that -- 8 MR. NICOLL: Well, I -- 9 MS. NAZARETH: -- I hadn't finished -- extremely 10 narrow all along and the three cent de minimis with a market 11 that normally trades at a tenth of a cent is enormous. So -- 12 MR. NICOLL: Well, I can go back and we can 13 question the statistics that were given to me and we can go 14 through that. Clearly, the spread is narrowed and the spread 15 today is three tenths of a penny and we can certainly 16 demonstrate that that's the case. And it's not around a 17 penny increment. It's a true subpenny spread. But we -- but 18 there are lots of trade-throughs. There always have been 19 lots of trade-throughs, and certainly it's very difficult to 20 keep an orderly marketplace of, you know, when stocks trade 21 with that kind of velocity and that kind of volume. 22 MR. DONALDSON: Well, there's a flashing signal 23 over here that we're coming to the end of this allotted 24 period. I want to say, we're going to see some of this panel 25 back on a couple of the other panels. I'm sure that there 50 1 are just thousands of questions sitting behind this table and 2 over there, so this is just a beginning. I hope we're going 3 to have a chance to talk to you all in a different forum. I 4 do want to thank you for coming here and -- thank you. 5 MR. NICOLL: Thank you, Chairman. Annette, my 6 notes -- I misspoke. It was -- our records show that it was 7 four cents when we traded on the American, not six cents. I 8 apologize. 9 MR. DONALDSON: Okay. We're going to take a little 10 break here, a ten-minute break and then we're back at 10:30. 11 (Recess taken.) 12 MR. DONALDSON: Okay. Why don't we get going here. 13 Welcome to our second panel. Since you have your backs to 14 the audience, maybe you would all, starting with you John, 15 say who you are and who you represent so the people can hear 16 you. 17 MR. WHEELER: I'm John Wheeler with American 18 Century Investments and I'm here representing the Investment 19 Company Institute, the trade group for the mutual fund 20 industry. 21 MR. HERRON: Dave Herron, Chicago Stock Exchange. 22 MR. FAGENSON: Robert Fagenson, VDM Specialists on 23 the New York Stock Exchange, and Stock Exchange Specialists 24 Association. 25 MR. MADOFF: Bernard Madoff of Madoff Securities. 51 1 MR. CRONIN: Kevin Cronin from AIM Management, 2 actually now AIM Investments. Under the Act of full 3 disclosure, I'm also the chairman of the New York Stock 4 Exchange Institutional Creditors Advisory Committee. 5 MR. PETERFFY: Thomas Peterffy, Interactive 6 Brokers. 7 MR. PUTNAM: I'm Gerry Putnam, CEO of Archipelago. 8 MS. NAZARETH: Since this group has a little unfair 9 advantage, it's like having a quiz after having seen somebody 10 just ahead of you take it. Perhaps we'll even be able to get 11 into some more in-depth questions because we do have some 12 sense of what's going to be discussed. But I thought maybe 13 we could start with the same basic questions, which are, you 14 know, are there benefits to the trade-through rule beyond 15 best execution and investor -- beyond the duty of best 16 execution and you do have views on that, Gerry? 17 MR. PUTNAM: Thanks for having me. I think, Mr. 18 Chairman, you mentioned in your opening remarks that this was 19 a complex rule, and we agree with you, and in many ways, we 20 think it's actually too complex, which would make it 21 difficult to actually enforce in the context of a 22 trade-through situation. Today, we do have a trade-through 23 rule for New York listed stocks, and that rule is routinely 24 violated. It's a very simple rule. It just says, "Don't 25 trade through." 52 1 The rule that's proposed here obviously has more 2 components to it with exceptions and opt-outs and all of 3 that. 4 I had one of our folks take a look yesterday at -- 5 in light of doing this today, what's going on right now on 6 our trade-through situation, and we actually automated the 7 process of dealing with trade-throughs, complaining about 8 them and trying to get them resolved. And yesterday, in one 9 four-minute period between 1:52 and 1:56 in the afternoon, we 10 actually had 37 outstanding complaints with no resolution -- 11 again, a simple rule, just don't do it -- in 32 different 12 stocks. 13 Not surprisingly, 70 percent of them were with the 14 New York Stock Exchange, and they do have the largest market 15 share so that's where you would expect it would happen the 16 most often. But our concern is that the rule is too complex 17 to enforce when you consider the rule today, which is very, 18 very simple. 19 I also have concern with, we've come up with this 20 new definition of fast vs. slow markets. And I think that's 21 going to be difficult to define, and I think it's dangerous 22 to go in that direction. 23 There's another standard that we've worked with in 24 the over-the-counter marketplace, which is really the issue 25 of firm quotes vs. soft or what we like to call "maybe" 53 1 quotes. 2 And we think that's really where the distinction 3 should lie and that's where investors are asking for a 4 choice. In the case of what you'd call or the analogy would 5 be a slow market, we would say that is a market that has 6 "maybe" quotes or indications. 7 At times, investors would like to choose to avoid 8 the "maybe" market, in exchange for trading with a certain 9 market or a market that had a firm quote, one that they knew 10 would not be manually interfered with. 11 I think that's where you'll find the -- if you look 12 back to what happened in the NASDAQ world, and Bob Greifeld 13 mentioned it earlier, NASDAQ had a different structure than 14 the ECNs did when they came in. ECNs came in with firm 15 quotes. Today all quotes for the most part are firm in 16 NASDAQ and we don't have a trade-through problem. 17 I think finally, just touch on NASDAQ, another 18 thing that you mentioned was the possibility of unintended 19 consequences. And I think there are serious unintended 20 consequences for the NASDAQ world if we impose a 21 trade-through rule. 22 Today, through competition, a large number of very 23 sophisticated routing systems have been made available to the 24 marketplace. And the way they operate is, they operate on 25 the basis that all quotes in NASDAQ are firm, and I'll just 54 1 make a quick example, say we have a quote on Instinet at five 2 cents, one in NASDAQ at six and one in ours at seven. One of 3 these smart order routers will say, take the five, take the 4 six, take the seven, all at the same time, so they have 5 imposed a trade-through always seeking best price. But if we 6 put a trade-through rule in the NASDAQ world, a quote coming 7 into your system, to take the seven stock, the seven cent 8 stock, we'd either have to hold it up and not display a 9 customer limit order, which is exactly the opposite of what 10 we're trying to get at here today, which is to display 11 customer limit orders, or we'd have to send that order back 12 to that participant saying, "It's going to trade through, we 13 can't handle it," at which point we interfere with time 14 priority within our own system. So -- 15 MR. COLBY: Before you stop, could I ask the other 16 panelists, could you address the fundamental question that 17 the prior panel was addressing, which was, they all said that 18 we need to protect limit orders in order to encourage price 19 discovery. Do we need a trade-through rule to help protect 20 limit orders? 21 MR. PUTNAM: In the NASDAQ world we don't have one, 22 and all protected competition has done it. The ultimate 23 reward, by giving people a choice, if you place your order, 24 if you make the distinction between firm quotes and soft 25 quotes, and then give customers the opportunity to opt out of 55 1 trading with the soft ones, what's going to happen is, you're 2 going to get rewarded for making a firm quote. 3 At that point, Scott's got what he wants. He puts 4 an order out there, it will be interacted with. It's exactly 5 what happened in the NASDAQ world as it went from being 6 semiautomatic with market makers and totally automated with 7 ECNs to one that is a totally automated marketplace today. I 8 think it will -- it's self policing, best execution, Gus 9 mentioned best execution. It applies in both the listed 10 world and the NASDAQ world. Leave it there. 11 Let the markets compete and they will automate like 12 they did in the OTC world and I think we'll get to our best 13 result that way, but you do have to give customers the 14 opt-out, I mean, the opt-out being the choice between trading 15 with the best quote, which is just the best price, and being 16 able to say, "On this particular order, I want to override 17 the system. I don't want to trade with the 'maybe' quotes. 18 I only want to interact with the ones that are firm." 19 MR. DONALDSON: How do you define a firm quote? 20 MR. PUTNAM: It's -- 21 MR. DONALDSON: Is there a time that it's firm? 22 MR. PUTNAM: The fast flow thing, because that it's 23 a question of time. So you say, "Well, one second, geez, 24 that's really fast, but not necessarily firm." Firm means 25 the order is available on the touch. If another order comes 56 1 in, it will execute against that order without any human 2 intervention. The order doesn't stop for someone to look at 3 it and think about what they might do with -- they might do 4 that really quickly. 5 MR. DONALDSON: How long does it have to remain 6 firm? 7 MR. PUTNAM: It just has to be firm to begin with. 8 Today's world, in electronic marketplace, all of our quotes 9 are firm, if they are displayed and if an order comes in and 10 that order still exists, it will execute. In the listed 11 world, where, if you look at the competing -- competing 12 models like the floor-based specialist model on New York, 13 those quotes aren't necessarily firm. You can see it. 14 And Scott says, you know, in the interests of his 15 shareholders, he wants to take that quote? Well, he gets 16 there to find out that that ten thousand shares at a nickel 17 isn't really there. There's ten thousand shares at a nickel 18 but there are 17 other people that are going to take that 19 order. So to him, that's not firm. Or there's some price 20 improvement that takes place, or it traded ahead because 21 there's other people standing there. So they are not -- the 22 distinction needs to be made between quotes that are 23 interfered with or could be, and quotes that are purely 24 sitting will with an electronic trigger on them that says, 25 "If you touch it and you're the first one in line, you get 57 1 it." And that's where the market wants, I believe would like 2 to be able to make a choice. 3 MR. PETERFFY: Well, we enthusiastically support 4 the Commission's proposed trade-through with some 5 modification. We believe that instead of focusing on whether 6 market center is automated, we should focus on whether any 7 particular quote disseminated by that market center is 8 automated. 9 We propose that automated quotes be followed by one 10 corrector symbol that signifies that the quote is 11 automatically traded when a matching order meets. 12 Automatically executed quotes, whether they are on the top of 13 the book or up and down the book, should be protected by the 14 trade-through rule, and manual quotes should not be. This is 15 a simple and technically easy idea to implement, focusing on 16 whether a particular disseminated quote by market center is 17 electronically executable, on a hybrid market such as the New 18 York Stock Exchange, to get the benefit of the trade-through 19 rule protection for those quotes that deserve it, and would 20 provide these market centers with the flexibility to evolve 21 towards a more automated system in any way they see fit. 22 In addition, this approach would deal with the 23 question of whether quotes away from the NBBO should be 24 protected or not, and how those trades should be done. 25 Accordingly, a rule should allow automated markets 58 1 to trade through manual markets by any amount. Limiting the 2 trade-through, trading through of manual markets to the de 3 minimis requirements would merely entice manual exchanges to 4 adjust the quotes lower so as to be off the market by more 5 than the trade-through requirement and thereby force orders 6 to come to them and to continue to take advantage of these 7 three options. 8 Conversely, we see no justification not to give 9 best execution against automatically available quotes. We do 10 not see an up tick as a valid option and we do not believe 11 that we should go along with that. Lastly, there was a 12 proposal that brokers should disclose to their customers the 13 amount by which they trade through manual quotes. We do not 14 believe that you should require this because this would 15 merely create a false impression that those prices were 16 actually attainable where we all know that most of the time 17 they are not. 18 MS. NAZARETH: Thomas, can you address whether you 19 think there are benefits to protecting the best price limit 20 order in the market, at least not -- 21 MR. PETERFFY: There certainly are benefits to 22 liquidity providers. If you like our firm is a large 23 liquidity provider. If we were told that anybody can trade 24 through our quotes, we would stop providing such quotes. 25 MR. GOLDSCHMID: I just want to ask a question. I 59 1 want to make sure if I understand both of you. Instead of 2 using "fast" terminology, you're using automated quotes that 3 can be taken as -- 4 MR. PUTNAM: That's correct. 5 MR. PETERFFY: And by the way, we also think that 6 there should be a clear standard as to what an automated 7 quote is. And in today's market, most internally tradable 8 orders by ECNs are executed in less than a tenth of a second. 9 So we believe that an automated quote should be one that 10 either executes or books or declines or routes an order to a 11 better market in less than a quarter of a second, at least 98 12 percent of the time. 13 MR. GOLDSCHMID: Gerry, do you agree with that? 14 MR. PUTNAM: I think that it could actually be, and 15 that's sort of where the marketplace is today, that's the 16 standard for it, it could be simpler than that. It is simply 17 what we all do today in the electronic world is, there is no 18 manual intervention in that quote. Tom's point is well 19 taken. If we sat there with a two-second quote, it took us 20 to automatically fill an order in two seconds, we'd have no 21 customers. So the marketplace takes care of -- takes care of 22 that. 23 MR. PETERFFY: But that's not true for other 24 dominant market centers, right? If you give them two 25 seconds, they will take two seconds. 60 1 MR. CRONIN: Thank you for giving me the 2 opportunity to today present our views on the proposed Reg 3 NMS. As we look at the proposal, I think it's clear in that 4 general concept, this proposal promotes some of the most 5 valuable tenets we think of what will ultimately forge the 6 most efficient capital market structure in the world and that 7 clearly is innovation, transparency, speed and efficiency, 8 and ultimately we believe that this proposal would provide us 9 as institutions, and investors in general, the competition 10 that's necessary with the investor protections that are 11 necessary and ultimately with the choice that's necessary for 12 us to pursue best execution. 13 You hear a lot of talk about these terms, 14 competition and choice and protection. And it's unfortunate 15 in the earlier panel that protection, it didn't seem to me, 16 it was dismissed as rhetoric. I don't believe that 17 protection of limit orders is rhetoric. If you look at the 18 absence of protection today, I think many of us have come to 19 the conclusion that the marketplace isn't nearly as efficient 20 as it should be or could be. And part of our process here 21 today should be to collectively figure out what will lead to 22 a greater benefit for investors. What will lead to greater 23 incentives for investors to post their limit orders? 24 And I think clearly the trade-through proposal and 25 uniform application of it is part of that protection. If 61 1 necessary, I know that there might be some unintended 2 consequences I believe that Gerry specifies that are 3 tangible. But I also believe that the technological 4 advancements that are made and the capabilities that many of 5 these systems provide will overcome that problem. 6 In theory, it's hard for us to imagine any concept, 7 public policy that would not advance the protection of the 8 limit orders. So again, in theory, I support Reg NMS's 9 proposal for the uniform application of the trade-through 10 rule. 11 Opt-out is a vexing issue for institutions. 12 Opt-out would allow us in theory to trade at, quote-unquote, 13 inferior prices and I think Scott did a good job earlier of 14 describing how his institutions' displayed prices aren't 15 necessarily either certain or the best prices and are pursued 16 in terms of best execution. 17 And yet it is troubling to advance any policy that 18 would promote our ability to pursue these inferior prices at 19 basically the exception of those who are willing to display 20 their limit orders. 21 In the end, I think opt-out does give investors 22 needed flexibility and I do believe that it creates an amount 23 of competition that's necessary to facilitate the advancement 24 of the marketplace, but I'm not sure that the end justifies 25 the means. And that is to say, I believe that competition 62 1 will ultimately get us to where we need to be in terms of 2 efficiencies, in terms of automatic execution capabilities 3 that many of us want. But it's very hard to establish a 4 public doctrine that, in order to facilitate that end, 5 basically takes away some of the protections that we feel are 6 very important to this whole process. 7 I will probably bring my comments to a close there 8 in the interests of letting others speak, but the issue is a 9 very difficult one for institutions, as I know it is for the 10 Commission. I believe that if we properly define what fast 11 markets are, and I think you have to very specifically define 12 what a fast market is, and I believe it's under a second. 13 I'm not going to get into picoseconds or whatever that term 14 was earlier for billionths of seconds. I think that's 15 nonsensical to most investors, Mr. Chairman. 16 However, I think it is clearly in the best 17 interests of investors in general that that time frame be 18 very specific, that it be flexible over the course of time, 19 because our fear is that the lowest common denominators would 20 be all that we would get. 21 But it's critically important to define that, to 22 define the standards of linkage, and I know others will 23 advance that all levels of the book need to be electronically 24 accessible for it to be deemed automatic or fast. I don't 25 view that as appropriate. I think it infringes on the form 63 1 of structure. I think it's asking for example an option 2 market to be something more than it is. 3 I think it is reasonable for the top of the book 4 initially. I think competitive forces will ultimately lead 5 us to a better sweeping capability at the exchanges, for 6 example, and I think the advancement, again, is by and large 7 to make sure that we all understand what a fast market is, to 8 make sure that competitive forces are there, to make sure all 9 the market centers have the ability to get to that point, and 10 do get to that point, and in the absence of that, I think 11 unfortunately, an opt-out may be necessary. 12 MR. MADOFF: Our firm has made a, you know, fairly 13 decent living as a fast market competing with a slow market, 14 so I'm not sure that it's in our own best interests to have 15 everyone on a fast market but, since our good Chairman has 16 asked us all to take off our selfish hats and speak for the 17 public good, I'm going to try to do that. 18 I agree a little bit with what everybody has said 19 here. And I think that really this debate comes down to fast 20 markets vs. slow markets. If you were in a situation just 21 looking at the NASDAQ world today, which is operating without 22 a trade-through rule, which is a fast market, I don't think 23 this discussion would be taking place for the most part. 24 I think what's triggered it is that you now have a 25 lot of fast markets represented primarily by ECNs that are 64 1 trying to compete with what is a slow market. We have 2 accomplished that by putting our own capital at risk and 3 being willing to protect limit orders and trade-throughs from 4 taking place because we do operate in an environment that has 5 a trade-through rule. In the NASDAQ market, we don't have 6 that issue. 7 But the ECNs who don't really have their own 8 capital that they can put into play, they are just 9 representing customer orders, they cannot make capital 10 commitments and therefore, do not protect orders in a 11 marketplace, and that's why it makes it very difficult for 12 them to compete in an area where there's a trade-through 13 rule. 14 And if you look at the fact that they were able to 15 successfully compete in the NASDAQ market where there isn't a 16 trade-through rule and it's fast market against fast market, 17 you're now asking them to compete with a slow market that has 18 a trade-through that doesn't allow that to take place. So 19 the ECNs, you had to give them the de minimis exception to 20 enable them to function there. 21 If you want a clear example of the difficulty of 22 having a fast market compete with a slow market and not 23 having chaos with a rule that says they can't trade through 24 it, you just have to look at the AMEX trading of ETFs with 25 the NASDAQ marketplace. When it first happened, and I can 65 1 tell you as a very active player in the ETF world, it was 2 sheer chaos the first few days when you allowed trading of 3 ETFs on both the AMEX and in the NASDAQ marketplace where you 4 had a clear example of a slow market competing with a fast 5 market. 6 In reality, if you made all markets become fast 7 markets and -- up to certain share size, this issue goes 8 away. I know of no reason why slow markets today, and let's 9 talk about the New York because that's basically what we're 10 here to talk about, if they automated their marketplace, 11 which they certainly have the capability to do, and probably 12 will do, then, and you -- you said that there was a 13 trade-through rule that you could not, fast markets could not 14 trade through other fast markets, we would support that. I 15 can see no justification for one fast market trading through 16 another fast market. 17 You asked for a definition of a fast market, Mr. 18 Chairman. My definition would be, which other people I think 19 have already echoed, is that an immediately accessible or an 20 automatically accessible market. It doesn't have to be under 21 a second. Our average execution speed is clearly under one 22 second for almost 99 percent of our executions. I don't need 23 to go down to milliseconds, picoseconds and so on. I just 24 want to know that when I see a quote, that that quote is 25 immediately accessible and the only way that that quote will 66 1 not be accessible to me would be if another automatic 2 execution takes place before mine. 3 If I'm going to send an order to a marketplace and 4 someone is going to be able to take that order, hold it up 5 for 30 seconds or even five seconds, show it to other orders 6 that are in the crowd who are going to get a second look, 7 that's not fair. 8 If you want to encourage people to put orders into 9 quotes, which is really what this is all about, you cannot 10 allow them to get a second look at an order. 11 I'm not going to get into the data of CLOB because 12 that's not why we're here, and I don't think that's 13 necessary. In any event, I think competition levels that 14 playing field. But you cannot have an order go down to a 15 marketplace in an automatic fashion and then once it arrives 16 at that marketplace, manual intervention takes that order and 17 then delays the execution of that order because, while that 18 process is going on, even if it might afford a better price 19 to the order that is currently shown in that marketplace as a 20 quote, I may very likely and probably will miss executing 21 that order, and then the order that is below that price will 22 no longer be there because those orders come and go, you 23 know, literally before you can blink your eye. 24 So the way to solve this issue is force all market 25 centers up to a certain share size to become automated. 67 1 Otherwise, you can opt out of that market. 2 Now, to deal with the opt-out issue, I believe that 3 there should be an unfettered opt-out of a slow marketplace 4 because, as a practitioner in the marketplace, it's very 5 important for me to have a rule that I can comply with. I 6 don't want a rule that I cannot properly comply with. And I 7 cannot comply with a rule unless there are linkages in place, 8 whether those markets are accessible, and they are accessible 9 at a fair price which I guess we'll discuss at the next -- 10 the next panel. 11 So as far as opting out is concerned, you cannot 12 force fast markets, you know, to not be able to opt out of a 13 slow market. And the opt-out has to be unfettered because I 14 have concerns that even with this one, two and three cents 15 ban that you are proposing, because in a marketplace that 16 takes place today, these prices flicker and I don't know 17 whether a spread that is three cents, when I decide to opt 18 out, may be two cents, you know, by the time I press that 19 button. 20 So it sounds simple to -- and I'm not sure how much 21 compliance there is, even with the three cent de minimis ban, 22 and -- I mean, somebody here, I guess Bob Greifeld, said the 23 problem is going to be how do you surveil, you know, all of 24 these -- all of these issues. That's a very difficult 25 situation and I genuinely believe that most practitioners in 68 1 the marketplace want to comply with the rules. They are not 2 trying to disadvantage anybody. 3 I would like to start with the premise that most of 4 the practitioners in our industry come in to their firms in 5 the morning saying they want to comply with the rules, they 6 want to do what is best for their client. I don't think 7 anybody is trying to opt out of an execution that they think 8 they could really get. 9 So those are the issues that I think are important 10 to dealing with in any rule proposal that you may make. 11 Thank you. 12 MR. FAGENSON: Mr. Chairman, Commissioners, 13 Annette, Bob, and Jonathan Katz, I finally get to see the man 14 who is behind all those letters. 15 In many ways, at certain times, today, it sounded 16 like a Congressional hearing. Some of the panelists find no 17 need to get a clear question that's asked, but simply get 18 their comments into the record. And I will try and adopt the 19 Chairman's opening comment that we should try and keep a 20 public policy hat firmly on our heads, because that's really 21 what we're here to talk about. 22 The good news is that because we have the deepest, 23 most liquid markets in the world, and are in fact the envy of 24 the world, we're not talking about something that's damaged 25 or broken, we're simply talking about something that's 69 1 terrific and trying to make it better. That makes our job 2 easier but also more challenging because we have to be 3 certain not to upset that which has built over so many years 4 and served us and the people of this country and the world so 5 well. 6 So the question is, a trade-through rule, is it 7 necessary? Is it going to incent people to put limit orders 8 in the book? And my answer to that is, absolutely yes. 9 If we turn the clock back to the confusion that 10 existed in the old NASDAQ pricing model where people would 11 enter a limited buy order at 15 and see the stock trading 12 that was at 14 and seven eighths and not understand why they 13 didn't buy that stock at 15 and the broker would calmly take 14 their hand and say, "We can't buy it at 15 unless it's 15 offered at 15," and it can be trading at 14 and seven eighths 16 or 14 and three quarters forever. And unless it's offered at 17 15, you cannot get it. 18 That was extremely confusing and I think some of 19 the elements of what we're talking about, if it's not 20 perfectly clear in terms of absolute limit order protection, 21 will be extremely confusing and will disincentivize investors 22 large and small to enter their limit orders if they don't 23 feel that those limit orders are going to be afforded price 24 protection. 25 Now, I testified to the Common Sense Pricing Act, 70 1 which made it very clear from what was stated at the time, 2 that every penny saved investors over a year would be a 3 billion dollars in their pockets. We have to be careful, if 4 we eliminated a trade-through rule, and have a de minimis 5 exemption or an opt-out, we're talking about three cents. 6 Well, it may not be every share. 7 But if we do the multiplication, if the Common 8 Sense Pricing Act put five billion dollars in investors' 9 pockets, this has the potential to take nine billion out of 10 their pockets because if everyone were pure and all motives 11 were perfect, and as one of the earlier panelists said, it 12 only happened occasionally, and people didn't set up systems 13 to take advantage of what the system offers, then everything 14 would be fine. But I think history has proven that not to be 15 the case. 16 Where an exemption exists, people are going to find 17 a way, I think, to take intermarket/intramarket advantage of 18 it. And that poses a tremendous danger because the opt-out, 19 as well-intentioned as it may be, creates to a certain extent 20 an elitist two-tiered market, two classes of citizens, and I 21 can't find anything within that that speaks of good public 22 policy. 23 I think we have to be careful as we take a look at 24 the benefits that the trade-through rule provides to 25 investors that we don't start designing a system for 71 1 practitioners and forget about the customers. 2 Now, I heard it's no longer politically correct to 3 talk about the eight-hundred-pound gorillas, they have been 4 on the Atkins diet, no pun intended. So now they are five 5 hundred pound gorillas. Pretend they are not sitting in the 6 corner of the anteroom and that large institutional investor 7 does not have a certain weight in the marketplace. It would 8 be disingenuous at the very least. 9 In fact, what we haven't talked about today is the 10 medium sized institution, the small Midwest bank, the 11 individual investors, and what the effect on public 12 confidence will be if we do away with the trade-through rule. 13 And that is a huge danger here. In some ways, before talking 14 about the SEC, I feel like I'm in front of FAA or should be, 15 because if it's not price, then it would be safety, and 16 certainly there are other issues that are important. And 17 whether it's the food quality, or on-time performance, if you 18 don't get there, nothing else really matters much. 19 In this case, the best price and the ability to 20 have confidence that you're going to be protected are the key 21 elements to what makes our markets the deepest and most 22 liquid in the world. The liquidity provider has to be 23 rewarded and has to be protected, and I'll get back to what 24 the Chairman's question was before. 25 Just because you can do it doesn't mean you ought 72 1 to. And just as we have systems that can trade in 2 nanoseconds, microseconds, or whatever the phrase of the day 3 is, doesn't mean that that is relevant to the most -- to most 4 investors who we seek to protect. 5 Intermarket competition by making markets fast and 6 making quotes automated and automatic and accessible is the 7 key, and it's rare over our careers that I have agreed with 8 so much of what Bernie Madoff has said, but in fact, the man 9 did build a business competing with me, and he has to be 10 congratulated. And he didn't come here to whine his way into 11 rules changes or Congressional mandates to do it. He did it 12 by offering a better price or a better execution system and 13 built a business on that. And that's to be applauded and 14 rewarded. 15 On the other hand, as I see the Commission moving 16 towards protecting customers and investors by giving more 17 disclosure and better disclosure and better protection of 18 mutual funds, I would hate to see us do away with the 19 trade-through rule and take a tremendous step backwards in 20 the markets. 21 MS. NAZARETH: Mr. Herron? 22 MR. HERRON: Thank you for the opportunity to talk 23 today. At this point, I will no doubt be echoing many of the 24 thoughts that have come before. In one sense, it's a little 25 bit like Phil Nicholas in the lineup as -- on the 18th of the 73 1 Masters, but we will be repeating some things. 2 One thing that has not been stated is really that 3 the -- the trade-through rule of the last 25 years has acted 4 as a cornerstone that I feel has unified the markets. I did 5 trade in the markets prior to the trade-through rule, and 6 believe me, we did not have liquidity, we did not have the 7 depth, we did not have the tightness and spreads that have 8 come to exist over the last 25 years. 9 The -- even the proponents of doing away with the 10 trade-through rule have benefitted from the price discovery 11 process that limit orders have put into the marketplace and 12 without the trade-through rule, many of those limit orders, 13 much of the incentive to place those limit orders by 14 professionals, by dealers and by customers might not have 15 occurred. 16 The -- any action to weaken the trade-through rule 17 will only serve to take away investor trust in the 18 marketplace. And this industry has not done much to earn 19 that trust lately. I think extending the trade-through rule 20 to other markets, especially as the markets become more 21 linked together and look more alike, will only serve to win 22 back the trust of those investors that is so key to the 23 marketplace. 24 We believe that the Commission has a framework to 25 adopt a trade-through rule with requirement for automatic 74 1 execution at all venues that will necessitate the need for 2 an -- that will eliminate the need for an opt-out provision 3 completely. If markets are truly automatically accessible 4 and no one has to wait for exposure to an auction to receive 5 their answer if they choose to not expose that order to the 6 auction, there is no reason to pick a price worse than the 7 one displayed in the marketplace. 8 In reality, for 25-odd years, we have had 9 trade-throughs occurring in the marketplace but those 10 trade-throughs have occurred with a requirement that the 11 offending party satisfy the traded-through party. 12 So even though there have been trade-throughs, the 13 investor at other exchanges, in the third market, has been 14 protected, if not in direct interaction, by the offending 15 party coming out to satisfy that quote, by that dealer's 16 ability to protect that order and satisfy that -- satisfy the 17 order that's been traded through, knowing that there is some 18 protection that exists for him or her. 19 As far as an opt-out provision, we get a little 20 ahead of ourselves, but we believe firmly, as I said, that 21 the opt-out is not needed and indeed, that the opt-out is 22 offered to the wrong party. The person posting the limit 23 order should be rewarded and the professional trader or the 24 large institutional trader should not be allowed to ignore 25 that price to trade for their -- to speed up their execution 75 1 or simply to fill an order more quickly. Thank you. 2 MR. WHEELER: Good morning. On behalf of the 3 Investment Company Institute, I'd like to thank the Chairman 4 and the Commissioners and the directors for giving us this 5 opportunity to speak before you today. 6 Again, I'm with American Century Investments and 7 I'll take Kevin's lead in the interests of full disclosure, 8 American Century does have a very small minority stake in 9 Gerry Putnam's technology, the Archipelago Stock Exchange. 10 But again, I'm here speaking on behalf of the Investment 11 Company Institute this morning. 12 I think first of all, on this distinction between 13 fast and slow markets, the Institute would agree that 14 investors need to be provided choice. And to me this is all 15 about choice. For far too long, I think our equity markets 16 have been structured and regulated in a way that rewards 17 passive behavior in price discovery in the transparency of 18 our markets. Whether it's floor brokers who stand in a crowd 19 and price-improve an incoming order, or specialists who have 20 the discretion of their capital to inject into the 21 marketplace at their own discretion. 22 I think we have a unique opportunity in front of us 23 with Reg NMS, and the N representing national, we have an 24 opportunity for all market centers within that National 25 Market System to compete on a level playing field and allow 76 1 investors the option to place their orders wherever they deem 2 fit. 3 And whether we as institutions choose to operate 4 electronically and anonymously on a system like Gerry 5 Putnam's or if we choose to have a manual floor-broker-based 6 system like Bob Fagenson's, we should be afforded that 7 choice. We have a responsibility to our shareholders, but 8 the Institute does feel like there is a bigger common 9 responsibility and that responsibility is to protect any 10 investor within the National Market System who chooses to 11 display their limit order. 12 In our view, limit orders are the building blocks 13 that provide the foundation for an efficient market, and 14 again, I think Reg NMS provides us with a unique opportunity 15 to put the faults of the structures of the marketplaces in 16 the past where they belong, in the past, and allow the 17 injection of technology that's been developed by the private 18 sector, for the most part, over the past couple of decades, 19 to give investors every opportunity to reduce their 20 frictional trading costs and enjoy automatic execution trade 21 if they -- if they so choose. 22 So the Institute is in favor of this distinction 23 between fast and slow or maybe vs. firm, you know; however, 24 we want to label these market centers but again, our focus is 25 going to be on protecting that limit order. And as many of 77 1 the panelists before me have kind of kicked around this idea 2 of, let's focus on the people that opt in, and put them ahead 3 of the people who choose to opt out, force the people who 4 want to opt out and circumvent display of limit orders to 5 honor all those people who have opted in. 6 And I've made this argument to Bob and Annette in 7 the past. The limit order has a value to it. Whether it's 8 the ability to buy a stock $20 higher than where it's 9 currently trading, if you open up The Wall Street Journal you 10 will see a whole list of options that are out of the market, 11 that are out of the money if you will, but they will have a 12 price attached to them. And in my view, what our markets 13 have done as -- is the intermediaries in our market have 14 extracted that option value from those limit orders and it 15 goes right into their pockets. 16 Trade-through will put that trade through back in 17 the option money of the investors who display those limit 18 orders. So again, there may be trades that we all, the 19 Institute and all of our various mutual fund accounts, feel 20 like on a given trade we may be better suited to trade up a 21 dime and get a block of stock done, we feel a common 22 responsibility to each other within the industry, with a 23 feeling that if we honor everybody's limit order along the 24 way, the markets will be more transparent, more robust, and 25 all investors, small, large, whales, gorillas, minnows, 78 1 whatever you want to call them, will all have lower 2 frictional trading costs along the way. Thank you. 3 MS. NAZARETH: John, could I ask you, if we could 4 start with you and move this way, it was implicit in what you 5 said, but do you think that we can sufficiently define the 6 terms of what a fast market is such that we can mandate that 7 everybody be a fast market, so that an opt-out is unnecessary 8 to the rule? Is that what you suggest? And I think some of 9 the folks on the panel I think earlier addressed this 10 question, but not everyone, so that -- 11 MR. WHEELER: Absolutely. I think Gerry brought up 12 a point, if he wants to be autoex and he gets a seven-cent 13 offer coming at him but there's a five- or six-cent offer 14 coming ahead of that bid, he's kind of stuck as to what to do 15 with that order, give it back, hold onto it, whatever the 16 case may be. 17 I guess in my view the liability in that situation 18 resides in the person disseminating that order. And if you 19 all lay out regulation that requires our exchanges to code 20 into their order distribution process that they, at the time, 21 nanosecond, whatever it may be, that that order hits the 22 system, that it just reads what's there and ships Gerry his 23 order at seven cents, Gerry should be free to trade that with 24 the understanding that whoever sent him that has the 25 responsibility of meeting those orders along the way so that 79 1 that five-hundred-share offering at five cents that may be, 2 you know, Gerry or Bernie Madoff, or the six cent offering 3 that may be one of you on a home PC, you know, interacting 4 through a discount broker and sitting on the New York Stock 5 Exchange, whatever the case may be, the responsibility I 6 think lies on who's disseminating that order and then we all 7 as participants would, I guess, in my view, compete on where 8 that order resides. 9 We will -- there's always going to be, unless 10 everything matches up perfectly, there's always going to be a 11 residual behind that offer, if there's a ten-thousand-share 12 offer, four or five cents in a bid for twenty thousand 13 shares, you're going to kind of step up and take all of those 14 limit orders along the way but then there's still going to be 15 a residual and it's going to reside wherever I choose to let 16 it reside, whether it's the New York Stock Exchange, 17 Archipelago, Chicago Stock Exchange or wherever. 18 So yes, to answer your question, we can operate I 19 think without a trade-through and I think you, we get a peek 20 of it with the ETFs but I think again, we have a bigger 21 responsibility to come back to the trade-through and protect 22 all those putting up orders and I think we can do that 23 technologically. 24 At least, I think at this point, with NMS, I think 25 we have to identify that as the ultimate goal rather than 80 1 dilute this down to say, "Gee, the technology just doesn't 2 really allow us to do what we wanted to do." We need to set 3 that out, even if it's further down the road, as the ultimate 4 goal of what we decide here, that protection of limit orders 5 is crucial. They are the building blocks for an efficient 6 marketplace, and as technology develops, to get better and 7 better along the way. We need to do that. 8 MS. GLASSMAN: I'd like to ask a question about the 9 opt-out. I don't think it's been raised before. Assuming 10 that we adopt an opt-out proposal as a rule, is the proposal 11 that we put out for comment practical, is it too complex, 12 what are your views on that? 13 MR. WHEELER: I'll speak if nobody else does. In 14 our view you should only opt out to circumvent a manual 15 marketplace, and let the marketplaces determine whether they 16 want to be autoex, whether they want to be autoex at their 17 displayed price or all through their order book and let the 18 competitive forces determine who wins this. 19 And as investors, we're agnostic. We don't really 20 care who wins this, whether New York wins this or Gerry 21 Putnam wins this. Our only goal is to save our investors as 22 much money as we possibly can and to protect the people that 23 display those limit orders along the way. I think there will 24 be unintended consequences of an opt-out. We didn't know 25 what SOES would do twenty years ago. We didn't know what 81 1 Direct Plus -- I think Bob, maybe he didn't know what Direct 2 Plus would do to the New York Stock Exchange. Not then. 3 These are all well intended, you know, innovations 4 in the marketplace with some very negative unintended 5 consequences, and I view opt-out as the same way. Who is 6 going to step in to a trade that opted out and determine 7 whether the person who made the determination was truly 8 acting in a best execution interest or if they were paying a 9 bill with that particular trade? 10 And, you know, again, have mentioned, we haven't 11 done great job in this business of late. Let's close some of 12 these loopholes and focus back on what's really paramount to 13 an effective market and that's protecting the people who 14 display limit orders. 15 MR. GOLDSCHMID: Want to leave room for me to trade 16 on an auction market if I want to. 17 MR. WHEELER: Absolutely. When I think to this, in 18 discussions that we've had, you know, the views in no way can 19 put the auction market out of business. I think there's 20 still a place for the New York Stock Exchange to transact 21 their business here. But in my view, I guess those blocks of 22 stock are going to get priced in more of a relative manner 23 than an absolute manner. It's not going to be, hey, there's 24 a twenty cent offering and an order comes in, I'll pay 25 twenty, turn to the guy, 19 cents, offered at twenty, oh, 82 1 trading at twenty, somebody else got the stock benefit. 2 It's going to be, I think a floor broker that's a 3 seller meeting a floor broker and a buyer and these markets 4 moving really fast, a penny comes back -- they just say, 5 "Hey, in the middle of the spread," when they walk over to 6 the specialist when they push the button, in the middle of 7 that spread, "I'll meet you and we'll trade a hundred 8 thousand shares." So whether it's 18 cents, 19 cents, twenty 9 cents, we don't care, as investors, about, you know, the 10 incremental movement of that midpoint, if you will, which 11 does get into subpennies. We really feel like that's the 12 only use for the subpenny. But, you know, if the floor 13 brokers meet and they agree on a relative price and they walk 14 over and push the button and it trades, and you're not again 15 compromising anybody's display order in the process, I have 16 no problem with that whatsoever. 17 MR. CRONIN: Commissioner, I'd like to come in on 18 your question as it pertains to how onerous the process is or 19 should be relative to opting out. If it's the Commission's 20 objective to offer an opt-out provision, and again, I think 21 I'm on record as saying I don't believe that's necessarily in 22 the best interests of all investors, but if it is the course 23 of direction that you take, I would hope that you would make 24 it a very onerous process to undertake for investors. An 25 opt-out should only occur in very specific instances. 83 1 We shouldn't say that particular investors opt out 2 25 percent of the time. It should not be done on a global 3 basis. I think those are the kinds of things, if there is an 4 opt-out, and again, there are some, I think competitive 5 advantages to having an opt-out, but if it is offered, it 6 should be very definitely offered on the pretext of having it 7 happen very rarely as opposed to a commonplace occurrence. 8 MS. GLASSMAN: Bernie, did you have anything to 9 say? 10 MR. MADOFF: Yes. If the question is the opt-out 11 provision, is it -- is it too onerous to comply with, I would 12 say absolutely yes. Based upon conversations that we have 13 had with our clients, both broker/dealer clients as well as 14 institutional clients, have all said that they would not be 15 able to effectively opt out based upon provisions that you 16 have. 17 And what would happen, they claim, is for the most 18 part, they would not opt out. Therefore, I think their 19 clients would be suffering. So if you put brokerage firms in 20 a position where this, in order for them to opt out, they 21 have to comply with it as you people have laid out, they will 22 choose not to opt out and that will probably be to the 23 detriment of their customer. And I know that that's not what 24 you're really trying to accomplish. 25 So if you have to on an order by order basis, and 84 1 then provide the information that you would be required to 2 provide them so that the investor could make the choice, 3 which I know is what you're trying to do, you're trying to 4 have the investor make the choice, not the broker make the 5 choice. 6 The problem is that most investors are willing to 7 let their broker make the choice and if they find that their 8 broker is not making the choice, you know, in their best 9 interests, they are going to move their account to another 10 broker that will do that. 11 Again, you start -- have to start with, and I know 12 this -- the modern times today does not support, you know, 13 this amount of good faith in your brokerage firm. But you 14 really have to start with the assumption that most of us in 15 this industry really have their clients' interests, you know, 16 coming first. Not necessarily the firm's self-interest. 17 So if you're going to have an opt-out provision, 18 which I believe you have to have in certain instances, then 19 it has to be an opt-out that people will use. Otherwise, 20 you're not accomplishing anything. And again, I go back to 21 that, our position is that an automated market should not, 22 you know, be able to opt out of another automated market. If 23 I know that I'm going to send the order to that marketplace 24 and that I have every reason to believe it's going to be 25 executed without any manual interaction delaying that order, 85 1 then I have no interest in opting out of a price that I see. 2 MR. FAGENSON: Following on on Bernie's comment, if 3 you have an all fast market, an opt-out becomes irrelevant 4 from our perspective. To think that opt-out is actually 5 going to reach through to the individual investor and be 6 something where they are going to issue informed consent, 7 what they are doing is almost as if thinking that payment for 8 order flow was sanitized by printing some disclaimer on a 9 customer form. And most customers tend to expect that their 10 brokers are going to act in their interest and the brokers 11 then have to live up to that. I don't think there's any 12 accountability with letting an individual investor know. 13 An opt-out in an automated computer-to-computer 14 marketplace, fast market, I think is absolutely unnecessary, 15 and the next question is, can we do it? Absolutely. I think 16 that it is viable and the later panels will discuss how this 17 integrates with the auction market principles. But NYSE is 18 committed to it and if we're the ones that everyone is 19 looking at as the benchmark of anything from recidivism to 20 activism, depending on your perspective, we will get it done. 21 MR. PETERFFY: So we all seem to be in agreement 22 here. I would like to -- everyone is up -- now, I would like 23 to address the issue of time and why it is important. If you 24 imagine a five-minute period, you see that in the five-minute 25 period, the stock price wanders up and down and the bids and 86 1 offers wander up and down. 2 If you give somebody five minutes to execute the 3 trade, he can within that period of time, among the various 4 prices that occur, he can pick the one he wants, right? This 5 is also true for a second. There are less changes in price, 6 but we do know that there are changes in price on active 7 stocks within a second. 8 This is why I'm suggesting that we go with the time 9 that is technologically feasible that works today, and that's 10 why I'm suggesting that we should stick to a quarter of a 11 second, because otherwise, if we do not have a limited amount 12 of time, then everything else we say, up dot, not up dot, et 13 cetera, doesn't mean anything. 14 MR. PUTNAM: Our belief is that the opt-out is too 15 complex. Simple automated markets vs. non-automated markets 16 is where the distinction should be made. To the extent that 17 a manual market has value, people will use it and they will 18 use it. Problem is, you don't have a choice today under the 19 current rule not to use it. If the -- the thing is, we 20 wouldn't even be sitting here talking about this today if 21 people felt that the dominant manual market that we all deal 22 with was the best choice one hundred percent of the time. We 23 wouldn't even be having the debate. 24 The fact is, the marketplace wants a choice when it 25 doesn't want to be in manual mode, and I think that's why we 87 1 do need the option, but it has to be something that's 2 enforceable because it's clearly defined. 3 MR. ATKINS: Well, I guess I'm going to follow up 4 on all this, especially with respect to the specialist 5 system, because I was intrigued by the comments as to the 6 executability of the quotes, and I was wondering, is that 7 concept inimicable to a specialist system? We talk about 8 fast and slow, if we were to go down that road, and I was 9 just wondering how we would structure that in an auction 10 system, or is that impossible? 11 MR. FAGENSON: Well, it's definitely possible and I 12 note the next panel is going to deal with that in depth so I 13 won't try to take too much time. But I think what John said 14 earlier is true, that negotiated transactions, wholesale 15 large transactions will take place and, so long as there is a 16 trade-through rule in the context of the quote, limit orders 17 are protected. This is a matter of negotiating price and 18 quantity, and is not going to be unduly influenced by small 19 transactions that are taking place at the same time. So the 20 integration is possible. 21 I think starting back with the issue of fast 22 markets having autoexecution capability improves investor 23 choice, improves market execution, improves price protection, 24 and how we ultimately integrate the auction market principles 25 and protect the ability of people to use judgement in 88 1 achieving prices in transactions that are not executable in 2 the context of the posted inside market is a challenge that 3 we're going to have to rise to. But I think it's certainly 4 something that we can and will do. I think to a certain 5 extent it's going on now. Just, the system is somewhat 6 arcane from point of view of ITS and the times that people 7 have that it creates a tremendous amount of animus. 8 We have to get that out of the market. We have to 9 have all people feel good about how their orders are 10 executed, how their orders are treated and how they 11 understand the system and make it as simple as possible. And 12 protecting and not having the Commission come out in favor of 13 one market system vs. another, and I think that's absolutely 14 possible. 15 MR. DONALDSON: We've reached the witching hour of 16 11:30, sadly. We've -- we'll take just a couple of minutes 17 here for everybody to stand up and turn around in their seat 18 and we'll bring the next panel in. Two minutes. Thank you 19 very much. 20 (Recess taken.) 21 MR. DONALDSON: I want to ask the panel to identify 22 themselves for the benefit of those who are sitting behind 23 you, just starting with you, Bob, if you will. 24 MR. FAGENSON: Robert Fagenson, same person, 25 different seat, vice-chairman, Van Der Moolen Specialists, 89 1 member of American Stock Exchange and also a member of the 2 New York Stock Exchange Specialist Association. 3 MS. WILLIAMS: I'm Jennifer Williams, I'm with the 4 Griswold Company. I'm also a member of the New York Stock 5 Exchange and an active executing floor broker on that 6 exchange. 7 MR. LaBRANCHE: Good morning, Michael LaBranche, 8 CEO of LaBranche & Company. We're a specialist firm on the 9 New York Stock Exchange. 10 MR. TILLY: I'm Ed Tilly. I'm vice-chairman of the 11 Chicago Board Options Exchange and serve as a market maker 12 and specialist. 13 MR. BRODSKY: I'm Bill Brodsky. I'm chairman and 14 CEO of the Chicago Board Options Exchange. 15 MR. SODANO: Good morning, I'm Sal Sodano, Chairman 16 and CEO of the American Stock Exchange and vice-chairman of 17 the NASD. 18 MR. THAIN: And I'm John Thain, I'm the chief 19 executive officer of the New York Stock Exchange. 20 MS. NAZARETH: This panel was teed up very well by 21 Commissioner Atkins, who asked the exact correct question as 22 to what will the challenge be of combining a traditional 23 trading floor with an automatic execution facility of the 24 type we've been talking about being required by the 25 trade-through rule. 90 1 So the question that we'd like you to address, and 2 we can start with John Thain, who no doubt has done a great 3 deal of thinking about this question in the recent past is, 4 how feasible is it for a market to combine an active trading 5 floor with an automatic execution facility? Is it possible 6 to achieve the twin goals of preserving the trading floor's 7 potential for improved prices and liquidity beyond the 8 market's displayed quotes and also operating an automatic 9 execution facility that provides fast and efficient access to 10 the market's displayed quotes? 11 So undoubtedly, many folks here are very interested 12 in hearing about what the New York's thinking is on this and 13 we'll start with John Thain. 14 MR. THAIN: Thank you, Annette. First of all, I'm 15 happy to have on the panel someone who has actually done 16 this. And so there's proof that it can be done. But it 17 certainly is one of our challenges. 18 I am committed to making the New York Stock 19 Exchange a fast market. Full stop, we're going to do that. 20 I'm committed to do that primarily because it's what an 21 important part of our customer base wants. And if we don't 22 offer our customer base the ability to trade electronically, 23 quickly, with certainty, and with anonymity, they are going 24 to move their order flow away from the exchange, and that's 25 bad for our business. So our fundamental motivation is, a 91 1 subset of our customer base wants to be able to trade this 2 way so we will give them the capability to do that. 3 But I'm also committed to maintaining the 4 advantages of the continuous auction system that goes on on 5 the floor of the exchange. 6 It's clear to me that the auction model with the 7 specialists and the floor brokers has advantages in certain 8 circumstances over a purely electronic market. The most 9 important advantage is reduced volatility. And there's a lot 10 of ways and a lot of ways to look at this, but the easiest 11 way to look at it is, over the last two years, 48 companies, 12 so the exact same companies, moved from the NASDAQ, a totally 13 electronic market, to the New York Stock Exchange. And the 14 intraday volatility of those 48 companies on average fell by 15 50 percent. Not by a little bit, by a lot. 16 So it's clear that the value of the auction process 17 and the specialists and the floor brokers does dramatically 18 reduce the intraday volatility and ultimately the cost of 19 capital to companies who trade in that marketplace. 20 The auction process also provides better opens and 21 closes, is better able to deal with order imbalances, it's 22 better able to deal with unusual events like earnings 23 surprises, either positive or negative, or corporate 24 takeovers, and it also provides ultimately better prices, 25 because the New York Stock Exchange does have the best price 92 1 in its security 90-plus percent of the time, and lower 2 transaction cost. 3 So the key here, the key challenge is to marry the 4 best of both worlds. Now, I understand that there's a 5 question about, will the Stock Exchange actually get this 6 done, will they offer the customers the ability to trade 7 electronically, and in particular, there's been this question 8 raised about, will the execution be automatic. 9 And the answer is yes, it will. We will allow -- 10 and as a matter of fact, Direct Plus, which is our version of 11 this, already exists. We will allow customers to trade 12 against the best bid or the best offer. They will be able to 13 trade electronically. It will be a computer-to-computer 14 link, and, you know, subject to the -- to the constraints of 15 the speed of electrons, it will basically allow very fast 16 execution against the inside market. 17 We, though, we still, though, want to keep very 18 much the benefits of the auction process. We want to be able 19 to deal with order imbalances. And this I think is a very 20 important thing. We do not want the trading in New York 21 listed stocks to turn out to have the same kind of volatility 22 that you've seen in the electronic marketplace. So that 23 benefit of moving from NASDAQ to the New York Stock Exchange, 24 we don't want that to go away. 25 So we have to make sure that, as we provide people 93 1 with the ability to trade electronically, that if there are 2 significant order imbalances, we can deal with those in a way 3 that doesn't cause dramatic ups and downs in the prices. 4 We also have to make sure we don't in any way upset 5 the value of the closings and if you've seen that S&P has 6 decided to use the American Stock Exchange closes in certain 7 of its stocks that are traded on NASDAQ and its index, 8 because of the better process of getting closes done through 9 an auction process on the quoted exchange. 10 So we have to keep the good things about the 11 auction process and marry them with the fast execution 12 capabilities that some of our customers had asked to have. 13 It's also important to us, and I'm sure some of my 14 fellow panelists will voice this, that the members on the 15 floor are able to interact with the electronic market. And I 16 made this comment, I was actually at one of the other 17 marketplaces that also is electronic and manual, and there, 18 the floor, inside the pits they are actually able to interact 19 and trade electronically as well as on the floor, and we very 20 much want to make sure that we give our floor brokers and our 21 specialists the ability to interact with the electronic 22 marketplace. So we're not going to run this as two separate 23 markets. We want to integrate the markets, allow customers 24 the choice to be able to trade electronically if that's what 25 they want to do, but also to be able to keep all the benefits 94 1 of the floor and ultimately, you know, assuming we get this 2 to work properly, which I am confident we will do, we'll have 3 to best of both worlds. 4 So there are issues, there are concerns, there are 5 challenges, but I'm very much committed to making this work. 6 MR. SODANO: Thank you. Good morning. The AMEX 7 would like to thank the Commission for allowing us to express 8 our views this morning. Combining the traditional auction 9 trading floor with that of automatic autoexecution is clearly 10 something we know we need to do, even from our own 11 perspective being able to be more nimble in dealing with the 12 ECN environment, which basically was built primarily for 13 speed. 14 But rather than go through a lot of the points that 15 John just went through with which I agree, one of the points 16 I do want to raise here is, we want to make sure that the 17 auction marketplace roll, as it pertains to innovative 18 securities, such as the types of securities that we bring to 19 market at the AMEX, new things to market, in the smaller and 20 mid cap marketplace, act differently than they do in the 21 larger markets. And it's very clear what benefits the 22 traditional auction marketplace provides for those types of 23 securities. 24 Let me say up front, the AMEX is clearly committed 25 to moving forward and quickly and deliberately to provide an 95 1 automatic execution functionality for the products listed and 2 traded on the exchange. We have rule filings pending before 3 the Commission. And we expect to, hopefully, get our autoex 4 capability in place and not too far down the track for the 5 exchange traded fund products such as the larger traded QQQ 6 ETF, SPDR, and DIA products that in and of themselves create 7 their own liquidity, we want to get that done as soon as we 8 can. 9 And as John mentioned before our NASDAQ UTP 10 program, where S&P did come to us and ask us on a prior 11 program to try to create some more stability on a test basis 12 for closing prices in NASDAQ, get an automatic execution 13 program in place as soon as we can for the NASDAQ UTP problem 14 products as well. 15 The difficult challenge remains in terms of how do 16 we effectively and responsively integrate the autoex 17 functionality into our auction model while continuing to 18 maintain the unique benefits that our market provides to 19 investors, listed companies, and as I just mentioned a second 20 ago, the innovative products that we come up with at the 21 American Stock Exchange. As John just mentioned, I think it 22 is very clear that a specialist environment does present some 23 very clear improvements to the marketplace as opposed to a 24 NASDAQ environment. 25 Issuers choose the American Stock Exchange just as 96 1 they would the New York Stock Exchange, for the decreased 2 volatility. Particularly, again, I want to underscore the 3 differentiation we'd like to make in the types of securities, 4 rather than focusing on just a fast market/slow market, the 5 types of products that are being traded. Companies that 6 trade less volume can use, clearly, the advantages that an 7 auction marketplace provides. 8 I'll be specific. Based on the statistics that are 9 available to us, at least we see at the AMEX, over 60 percent 10 of the equities right now on all the markets trade under a 11 hundred thousand shares a day. So when you go through the 12 statistics in looking at how is volatility effected in the 13 dealer market vs. an auction market, I would agree with John 14 and we see the same statistics, even more so at the AMEX, 15 concerning middle market companies are the core of our equity 16 business. We believe the marketplace and therefore, 17 investors, clearly receive a benefit by utilizing and having 18 available to it the traditional auction marketplace. 19 So our job is to come up with a model whereby we 20 can provide autoex functionality for products that are -- a 21 product that should be in that type of an environment as 22 opposed to just a straight autoex marketplace or being 23 defined as a fast market or a slow market. We're looking at 24 an environment where we should look at the products 25 themselves rather than just the overall marketplace. 97 1 When determining the appropriate balance between 2 speedy execution through autoex and liquidity, stability and 3 potential for price improvement offered by the auction 4 marketplace, we clearly believe one size does not fit all. 5 The balance for the most liquid securities may not 6 be the appropriate balance for the less liquid securities or 7 any innovative securities that the AMEX has been known for. 8 Simply said, securities that lack natural liquidity can and 9 do get the help of a marketplace that's afforded in a place 10 like the American Stock Exchange. I do believe there is a 11 solution or business model that works. We are working to 12 implement a new order type that will allow autoex executions 13 to take place, or simply be cancelled with immediate advice 14 back to the customer if that's what happens. 15 In addition, we will continue to work with our 16 traditional auction marketplace as we have, and continue to 17 put improvements into that traditional system such as 18 increasing speed of access through the systems as we've 19 traditionally offered. 20 So in summary, what we're saying at the AMEX is 21 clearly, we will look to, at some point, an automatic 22 execution for all the securities that trade on the AMEX but 23 we believe, in view of the debate that you've been hearing 24 today, the Commission should not lose sight of the fact that 25 it's not really a fast market or a slow market. It's really 98 1 the types of products that are being traded. 2 When you look at the AMEX in particular, things 3 that are new to the market, there is not a natural source of 4 liquidity. When you look at very mature products like, 5 whether it's Microsoft that trades on NASDAQ, or GE on the 6 New York Stock Exchange, or QQQs, they provide their own 7 liquidity. When something is new, it needs help. That's 8 what we do. We don't think there should be a mandate that 9 those types of securities get automatically executed, but 10 over time we will develop a system that does both. 11 MR. GOLDSCHMID: A question for you and John, a 12 balance between simplifying it, customers on auction floor on 13 setting up these new markets, should it be me as the customer 14 that says, I want the floor or I want -- 15 MR. SODANO: The system that we're working on on 16 the AMEX is we want to create a new order type -- I guess the 17 old term would be fill or kill. I guess that's the closest 18 analogy that I could make where an order would be sent to the 19 AMEX for automatic execution, either it will be filled or it 20 won't, immediately, and there should be immediate advice back 21 to the customer that that order was either executed or not. 22 And then the customer will have to determine where that order 23 goes from there. So we believe that is very possible. 24 MR. GOLDSCHMID: And why wouldn't it be executed? 25 MR. SODANO: I'm not prepared to say why it would 99 1 not. There's a lot, -- well, it gets into debate on, when 2 should an automatic execution take priority over a price 3 improvement order? Should it be that clear? Steadfast in a 4 rule, should there be other opportunities where that may not 5 be the case, and quite frankly, internally we're still 6 looking at what the appropriate answers to those -- 7 MR. GOLDSCHMID: John, can you address that? 8 MR. THAIN: Yes. It's a little bit complicated 9 because it depends on the order type and it depends whether 10 or not it's a -- whether it's a market order or whether it's 11 a marketable limit order. 12 But the basic answer will be customers first of all 13 will be able to choose so they will be able to send an order 14 in via Super Dot, which they can today, or they will be able 15 to sent an order in through Direct Plus, which is give the 16 automatic execution. 17 But we will probably also automate market orders 18 and/or marketable limit orders to allow what, in essence, 19 happens today to happen automatically where the specialist is 20 pairing off the trades. 21 So if you think about it, if you have five hundred 22 shares bid at a quarter and you have a market order come in 23 to sell that same security, that trade can be paired off. 24 And even if it came in as a Dot order, will probably still 25 autopair it off as opposed to have anyone interact with it 100 1 because in essence, that's what happens today. 2 MR. COLBY: I think one of the questions we were 3 interested in hearing from this panel is, how can it actually 4 be done? How can you retain a floor presence with the price 5 improvement features and also have these electronic orders 6 going off? And we heard you say that you want to do it and 7 that you're working on it. 8 Bill and Ed from the CBOE, you've been struggling 9 with this, which is why we're very interested in you, but the 10 rest of the panelists, if you can talk about what can be 11 preserved and what will be lost if you have this, that will 12 be very helpful. 13 MR. BRODSKY: I guess that's my segue in. I'm Bill 14 Brodsky, chairman of the CBOE. The CBOE applauds the 15 Commission for Reg NMS and for holding this hearing today. 16 Many of the issues that are discussed here ultimately will 17 affect the options markets which, as you know, are intensely 18 competitive. CBOE is not only the largest options exchange 19 but also in terms of floor brokers and market makers is the 20 largest securities exchange in the country with vigorous 21 intramarket competition as well. Options specialists, DPMs 22 and market makers are all considered specialists under the 23 Act, represent a significant amount of stock trading in the 24 equity markets. 25 All the markets that you heard from today are 101 1 places where our market makers, specialists and DPMs send 2 their order flow for hedging the option orders that they do 3 on our exchanges. 4 Often, that represents hundreds of millions of 5 shares a day in the equity markets. Therefore, we not only 6 have an interest in today's hearing, we have a vital stake in 7 the outcome. 8 We also wish to compliment the Commission and the 9 staff overseeing the implementation over a year ago of a 10 state-of-the-art electronic linkage system among the options 11 exchanges. 12 Sitting next to me is Ed Tilly, who is not only 13 vice-chairman of our exchange, but was the person from our 14 membership who chaired the committee that helped develop the 15 system that we're talking about, a system that doesn't throw 16 out the baby with the bath water, as John says, is the best 17 of both worlds, and we think we have gotten to a point where 18 we're now reaching the benefits of that, of achieving a 19 combination of instantaneous execution along with retaining 20 the benefit of price discovery of the trading floor. 21 I'll just touch on the high points of the hybrid 22 system and Ed will go into some of the specifics for you. 23 Up until several years ago, CBOE employed a 24 traditional trading floor-based auction system. Last June, 25 we introduced a new trading system that we named the Hybrid, 102 1 which combines the feature of open outcry with electronic 2 competing dealer model. The hybrid platform enables market 3 makers and designated primary market makers, which is our 4 equivalent of specialists, to electronically stream their 5 individual quotes and for floor brokers to submit their 6 customer orders into an electronic system. Ed will go into 7 the mechanics of the system. But let me give you some 8 statistics. 9 We have reduced our average quote width in half 10 while tripling the average size quoted. Our quotes now are 11 on the NBBO about 70 percent of the time while the 12 floor-based exchanges that have not done what we have done 13 have quotes on the NBBO about a third of the time. And we 14 were not much better than that. So the difference is, when 15 we imposed the electronics with firm quotes, our quotes 16 narrowed and the size got bigger and I expect that to 17 increase over time. 18 Statistically, about 83 percent of our orders come 19 in on the Hybrid system and 70 percent is done on the trading 20 floor. So we see the larger amount of orders being done 21 electronically, as opposed to open outcry, but there is still 22 a substantial amount of volume being done in open outcry. 23 For those market participants who desire an immediate 24 execution against display quotes or size, the Hybrid trading 25 system provides such an electronic means to access the best 103 1 quotes of competing market makers on the exchange. 2 For those participants who have orders such as 3 complex or large orders, they can benefit from the trading 4 crowd information and interaction. The exchange maintains 5 and continues to maintain this trading floor with numerous 6 competing market makers where such orders can be worked. 7 And our floor brokers, by the way, add value in 8 working those orders and getting price improvement. Our 9 exchange is unique in that we have many competing market 10 makers, both on the electronic system and on the trading 11 floor, along with an integrated limit order book between the 12 two. 13 We're now in the process of obtaining SEC approval 14 for an enhanced Hybrid system which we call Hybrid 2.0. The 15 enhanced Hybrid would create competing and for the first time 16 remote electronic DPMs or specialists and competing remote 17 market makers who would stream quotes into the exchange's 18 Hybrid system. Up until this point, by the way, if you 19 wanted to be a market maker or specialist DPM, you had to be 20 physically on the floor. So this is a marked departure from 21 our 31-year history. 22 In-crowd market makers would be able to submit 23 quotes on the Hybrid system as well, so they could be in the 24 crowd in an open outcry fashion or at the same time, submit 25 electronically. 104 1 We've had a tremendous response from firms desiring 2 to be electronic DPMs and market makers including firms that 3 have never acted as market makers on the floor. We hope with 4 SEC approval to begin Hybrid 2.0 next month. 5 Although we are very proud of our Hybrid system, we 6 still see great value in offering trading floor liquidity. 7 Part of the genius of our securities markets is the ability 8 to offer true liquidity on a trading floor. While we suggest 9 that the SEC try to accommodate the needs of automated and 10 manual markets, we implore the Commission not to force 11 markets into a completely electronic model. Let the 12 customers decide large and small what form of price discovery 13 they want. Thank you. 14 MR. TILLY: Hi, I'm Ed Tilly, vice-chairman of the 15 CBOE, and I'll start by answering your question yes, hybrid, 16 or a hybrid system can be done or can be implemented. And 17 before I touch on a few of the key aspects of our current 18 hybrid trading platform, I'd say that this was a very complex 19 process, and the encouragement we received from the 20 Commission's department of regulation was extremely valuable 21 in coming to our launch of the Hybrid platform. 22 Let me focus on four key aspects of our hybrid 23 platform, what we've done and the obstacles that we've had to 24 overcome. 25 We've eliminated the disincentive on our floor to 105 1 competitively quote and actually provided ample incentive for 2 our market makers to competitively quote by rewarding those 3 that are providing liquidity on the NBBO. We promoted 4 greater access to our markets, which enables greater order 5 interaction. It's very important to us to open up access to 6 our marketplace while we're adding this technology. 7 We're disseminating a real time firm quote 8 comprised of customers, firms, broker/dealers and our 9 on-floor liquidity providers who stream in electronic quotes. 10 We are maintaining an open outcry trading pit where price 11 discovery added liquidity beyond disseminated electronic size 12 with trade. And we provide a competitive environment where 13 complex orders can be handled. 14 Mr. Thain and Mr. Sodano are correct. When we find 15 high volatility days, our transactions come off of the Hybrid 16 trading system are represented by either the customers' 17 choice or their order handler's choice, their broker's 18 choice, and into the open outcry environment more and more as 19 volatility increases. 20 We've had the luxury in kicking off this system at 21 the same time our industry was embracing the intramarket 22 linkage plan. It was very important to us that we're able to 23 protect our customers and the customers of other exchanges 24 through trade-through's in these fast markets. So all these 25 things together for us came together at the right time. So 106 1 we had an intermarket linkage plan being kicked off, we were 2 able to protect our customers, we're incenting our market 3 makers to competitively quote, all while maintaining the 4 benefit of open outcry when customers or their 5 representatives choose to route their order that way. 6 We don't get into the business of choosing how to 7 route an order. The customer does that or his representative 8 does that. 9 MR. DONALDSON: What's happened to the economics of 10 the open outcry portion that's now getting only 30 percent of 11 the volume? 12 MR. TILLY: Well, let's be very clear, it's 30 13 percent of the volume but about fifty percent of -- I'm 14 sorry, 30 percent of the orders but fifty percent of the 15 volume. That's very distinctive point. Complex orders, 16 spreads, rolls for us, on expiration, in particular, are 17 still handled in open outcry. The majority of small retail 18 paper is routed electronically -- 19 MR. DONALDSON: What about the economics of the 20 people involved in -- 21 MR. TILLY: It's interesting that the economics 22 haven't changed because the electronic NBBO, which we've held 23 ourselves to for quite sometime prior to hybrid, the 24 effective electronic bid/ask, see, we've had a no 25 trade-through policy for sometime. It came before linkage, 107 1 it came before our Hybrid platform. The effect of NBBO is 2 really unchanged. We're just finding it on our own exchange 3 now. 4 So rather than having to reroute orders to the AMEX 5 or to the Philly, we keep them in house and actually reading 6 electronically on the NBBO as found on the CBOE. So the 7 effective market nationally hasn't changed. It's only 8 changed on our floor. It's a big distinction. The economics 9 are truly unchanged because the bid/ask is unchanged. It's 10 all on our trading floor. So the opportunity is increasing 11 to trade on the CBOE for our members, but the economics are 12 really unchanged. 13 MR. BRODSKY: Mr. Chairman, one other aspect to 14 this is that the economics in our business have changed in 15 the macro. And that is that the competition that has been 16 fostered by the electronics, by intermarket linkage and all 17 the things that have happened over the last three or four 18 years have been nothing less than breathtaking and the spread 19 and the margins have narrowed. As the margins have narrowed, 20 volume has increased quite dramatically. 21 The option industry is now trading around five 22 million contracts a day. That's equivalent of about a half 23 billion shares a day. But it's on a per trade basis less 24 profit annually than it was before. 25 MR. TILLY: Certainly, but not to imply that the 108 1 electronics at home or the Hybrid system was -- 2 MR. SODANO: Chairman Donaldson, I think that's a 3 really important point that you bring up. It's one of the 4 points that we are grappling with at the AMEX as we do bring 5 in an automatic system. 6 Clearly, the people who risk their capital every 7 day are going to have less opportunity to get return on that 8 capital that they employ every day. But we need to, and 9 marketplace is going to have to come up with some way of 10 incentivization to that capital to stay on the trading 11 floors. So it's a very important point that we need to come 12 to grips with and come up with a solution. 13 MS. GLASSMAN: I just want to clarify what you mean 14 by, when the customers are choosing the open outcry system 15 over the autoex system, is that essentially an opt-in? Is 16 that what they are doing? 17 MR. TILLY: I can speak to how the routing 18 parameters are set up. Each firm is able to customize their 19 routing parameters, how their orders will be handled on the 20 floor of the CBOE once they hit our front door. Some firms 21 route to open outcry or broker interaction based on size of 22 the order, based on a day's, maybe volatility in a stock. 23 They want a broker handling their order. We allow them to 24 choose. They are not really opting in or out, meaning we're 25 still protected by NBBO. 109 1 We're forced through intramarket linkage to honor 2 all of the other options exchanges' markets. So they are not 3 opting out of any price protection. They are choosing how to 4 represent their orders once that order has hit the floor of 5 the CBOE, whether it remains electronic and is autoexed or is 6 handled by one of their representatives on open outcry. 7 MR. BRODSKY: One specific example of is that, 8 Commissioner, would be, a customer wants to do five thousand 9 contracts and the aggregate offer on the exchanges is two 10 thousand contracts. They may automatically execute the two 11 thousand by literally pressing one button on the system that 12 we've introduced but then they will go to the floor to do the 13 balance. So it's not even an either/or. You can choose 14 either order by order, or you can have a large order be part 15 of it electronically because that's what the offering allows 16 and then go down to the floor to negotiate the balance. 17 MS. NAZARETH: You have an integrated book. Nobody 18 is -- 19 MR. BRODSKY: Yes, this is not parallel systems, 20 and to Ed's credit and the credit of his practitioners that 21 designed this with our computer people, what they did is, 22 they said, we don't want two separate markets, we don't want 23 to know what tracks market for ID and options and an open 24 outcry market. We have one integrated book. If you're on 25 the book, you're on the book and doesn't matter whether it 110 1 was put in by a broker from the crowd or from the firm from 2 upstairs or a broker in the booth. Once it's on the book 3 it's protected. 4 MR. TILLY: If you can envision, it's one giant 5 catcher's mitt. It takes nice uniform quotes, a customer's 6 resting order, a firm or a fellow professional's order and 7 holds them all together in one giant book. 8 MR. ATKINS: So you're relying more on an NBBO 9 concept rather than a trade-through, sounds like a post hypo 10 than a -- an analysis, is that what you're saying? 11 MR. TILLY: I'm sorry? 12 MR. COLBY: Can you explain how -- there is an 13 intermarket trade-through rule. Can you explain how that 14 fits into this picture? 15 MR. TILLY: As part of the interactive linkage 16 program, we are -- no system, no floor-based or electronic or 17 screen-based options system exchange can trade through 18 electronically any other existing exchange. So the national 19 best bid/best offer, is sacred for electronic execution. 20 On the floor-based exchange, those orders are 21 kicked out for linkage and we will automatically link an 22 order to another exchange where the customer can receive the 23 NBBO price. And the screen based -- the one screen based 24 trading -- we're up to two now -- options exchanges, automate 25 that process and send them to either the CBOE or the AMEX or 111 1 another options floor. 2 MS. NAZARETH: I think the question is also how 3 does this work. It's because through the linkage, each 4 participant in the linkage is offering automatic -- 5 MR. TILLY: Absolutely. 6 MS. NAZARETH: Everybody is essentially a fast 7 market, offering orders up to the book size through the 8 linkage. 9 MS. GLASSMAN: Everything is firm -- 10 MR. TILLY: Firm. 11 MR. DONALDSON: Let's move to the other half of 12 table. Do you want to go? 13 MR. LaBRANCHE: Good morning, afternoon, actually. 14 One of the points I'd like to make is the auction in my mind 15 is really the most important thing that the New York Stock 16 Exchange represents to the public. In my opinion, the 17 auction functions like a bill of rights for investors, 18 because it means that all participants in the marketplace, 19 whether they have a hundred-share order or an eight-share 20 order, get to interact in the market in a fair way. No one 21 gets a special advantage because they have a larger order and 22 an individual that chooses the New York Stock Exchange as a 23 destination gets included in that auction at all times. 24 So the other thing about our market is, we have a 25 continuous auction, which means that there's price discovery 112 1 occurring at all times during the trading day. 2 Take, for example, some commodities markets which 3 have limit up or limit down. When markets are limit down, 4 there's price discovery because the markets simply stop 5 trading. Our markets don't do that. They trade all the 6 time. So I think what we need to do is see how we can 7 integrate an automated system when we have displayed 8 liquidity because when displayed liquidity is there, price 9 discovery is not occurring. 10 So what we're finding, and this gets back to the 11 question that Mr. Colby asked earlier, how do we have both 12 markets simultaneously? And what, in my mind, is a practical 13 way to look at it is, when there is displayed liquidity, then 14 you can add speed and certainty of execution in those 15 situations but when an order enters a marketplace that's 16 bigger than what's seen as displayed liquidity, then price 17 discovery starts to occur. 18 Let's say, for example, there's five thousand 19 shares offered at a certain price and fifty thousand comes in 20 a bind. We know that five thousand will trade at that price. 21 The question is, what happens to the other 45,000? The 22 specialist's job is to correct imbalances in that situation, 23 provide price discovery. And what I think the practical way 24 to do it is to have an automatic execution for that five 25 thousand so people will know, if it's there, they can press a 113 1 button and get it with certainty. 2 Now, our job would be to find the right price for 3 that other 45,000 in a way that makes the market efficient, 4 and that's what we want to do. So we have we want to have 5 that capability for our customers, that's what we hear. We 6 hear that they want this and we want to offer that but we 7 also still want to preserve the auction, the price discovery 8 process in a way that gives the customer the best price. 9 One of the great things about our market is that 10 people have an inherent confidence that their orders are 11 going to be handled in a good way, that makes our capital 12 market in the United States better than any other capital 13 markets in the world such as (unintelligible), which means 14 that the 45,000 shares that were looking for the right price, 15 they get the right price. 16 So there's going to be a moment when we're finding 17 that price discovery new price for the stock and it won't be 18 automated. But as soon as we find that new price discovery, 19 then we go back to an automation. That's how I see it 20 happening. 21 MS. WILLIAMS: Hi, I'm Jennifer Williams. I'm a 22 member of the New York Stock Exchange and I work with the 23 Griswold Company. I'm also an active broker on the floor of 24 that exchange. 25 I'm going to take a bold step here and say 114 1 something outright that none of my colleagues have said thus 2 far, is that the New York Stock Exchange has already begun 3 its transformation to a hybrid market and I think that's 4 important to understand. And many of the investing public 5 are not aware of that. 6 Through the Direct Plus system, orders, small 7 retail orders as well as large institutional orders have the 8 option of bypassing a floor broker and going straight to 9 automatic execution to the best bid or offer displayed on the 10 screen so that an investor who values their speed of 11 execution over the efficiency of the price discovery model of 12 the auction market has the option to do so. 13 Obviously, the system is not perfect. If it were, 14 we would not be sitting here today. But I know that the New 15 York Stock Exchange is working tirelessly to perfect that 16 system so that we can create a truly hybrid market. I 17 suspect over the next few months, we'll really be able to 18 fine-tune the details of the linkages. 19 I'm hoping to explain to you what my role in all of 20 this is. I'm not a specialist, I'm not the chairman of the 21 firm. I'm just a floor broker but I value that title, and it 22 means a lot to me and I'm not here to talk to you because I'm 23 hoping to have a job in five years. I'm here to talk to you 24 because I do what I do because I believe in what I do and I 25 believe that we have the best market pricing model in all the 115 1 world. 2 I am a broker, meaning that I am an agent. I do 3 not at any time effect proprietary transactions, so that my 4 incentive to go into a trading crowd is solely my fiduciary 5 responsibility on behalf of my customers and it is to get 6 them the best price. As I said before, those customers that 7 are more interested in their speed of execution can go autoex 8 through the system and by-pass me. 9 My fiduciary responsibility is to achieve the most 10 efficient execution possible. I would like to make the point 11 that I believe efficiency used in this term is not synonymous 12 or exclusive to speed. And efficiency of execution must also 13 take into account the cost of transaction and the price of 14 transaction as well as the investment objectives of the 15 particular customer. 16 That's my job. Somebody calls down and wants to 17 buy fifty thousand shares of IBM, it's my job using my 18 professional judgement with the market indices and the 19 current market conditions at my disposal, to use my 20 professional judgement to get that customer the best price. 21 The Griswold Company represents customers that are 22 some of the largest mutual funds and pension funds in the 23 country. And I've heard a lot of market buzzwords thrown 24 around at this table this morning. I'm going to bring it 25 down a notch. I'm talking about your pension funds. I'm 116 1 talking about your parents' retirement accounts. I'm talking 2 about your children's college education funds. 3 As the former panel mentioned about the 4 trade-through truly that every penny that is saved by virtue 5 of an auction market pricing system on an average of a 6 one-and-a-half billion share trading day times 250 days a 7 year, is $3.75 billion in savings every year to the investor, 8 all over a penny. That's a lot of college educations. 9 So I think it's important to remember that while we 10 can offer automatic execution and, like I said, the New York 11 Stock Exchange is working nonstop to really perfect the 12 system so that we can get instant executions for all size 13 orders if that is what the customer so pleases, that the 14 auction model is still the proven, most effective pricing 15 mechanism and, after all, we're here because of the investing 16 public. 17 Any market that favors solely speed over the 18 pricing mechanism that we offer invariably favors the market 19 practitioners and not the market investors, and we should be 20 here representing the interests of those investors and how we 21 can best make the market work for them. 22 MR. FAGENSON: An efficient market is one that 23 advocates liquidity and brings buyers and sellers together 24 and the challenge of how we integrate that in terms of an 25 electronic autoex environment with the use of judgement in 117 1 terms of how to get the best price, what velocity, what price 2 points to enter and exit the market and how do you achieve 3 that best execution and the customer's objectives is clearly 4 something that's going to present the greatest challenge as 5 we move ahead. 6 Jennifer is right, we've got experience with this 7 now, and have for some time, the integration of our own 8 internal NYSE Direct Order type which allows an autoex at the 9 inside quote albeit only up to 1,099 shares for limited 10 orders now, has allowed us to see how an autoex environment 11 can cohabitate the same trading as an auction marketplace. 12 And nuances that will have to be brought to bear in terms of 13 expanding this order type so it can be without limit to size, 14 without limit to order type, is going to be where we're going 15 to have to focus. 16 But in terms of taking the judgement that people 17 want to exercise in their order, and the speed that they want 18 to achieve, and eliminating some of the volatility we've seen 19 in very unusual situations in other markets that don't have 20 that judgement I think is something that we have to make sure 21 that we focus on. 22 Now, clearly there have been some examples. One 23 stock I think was very much in the news as we saw the price 24 move 40 percent, $19 whatever it was, in a short period of 25 time, and then rebound. Clearly, this is not the stuff of 118 1 which customer confidence is built. 2 So how do we prevent a situation that looked more 3 like a day care center where no adults showed up that day 4 from being the prevailing market system, and yet integrate 5 the ability to autoex into a marketplace that has always had 6 as its benchmark customer protection and the ability to 7 exercise from the information one gets in the marketplace, 8 one's judgement in how to achieve the customers' goals. 9 New York Stock Exchange operates at its finest when 10 there are imbalances. Otherwise, things tend to happen very 11 much automatically. When suddenly the half a million for 12 sale or the 800,000 to buy shows up in a stock as average 13 daily volume is the same as the order with which one is 14 faced, that is where you have to rise to the challenge. 15 Now, years and years ago, reaction was simple. You 16 halted your trading. But in these days, we want to enhance 17 and protect the essence of free markets, how to we make sure 18 that doesn't happen? What Michael described before was that 19 short period of time when customers are brought together, 20 buyers and sellers, to achieve a fair, reasonable and 21 accurate best price as close to the last sale as possible is 22 where the essence of our auction and the integration into an 23 autoex environment is going to really create the greatest 24 challenge but also one that inevitably and in my mind 25 absolutely we are going to be able to achieve. 119 1 The key here is not to destroy the essence of all 2 that is good in trying to back us into a model that may 3 appear on its face to be more efficient but in fact, has 4 tremendous weaknesses. We can marry the two. We're doing it 5 now. We will expand the capabilities of autoex and we'll 6 integrate it with the issues of judgement and auction 7 pricing. And I -- I, as everyone else at the exchange, are 8 committed to that, and we know we can get it done. 9 MR. DONALDSON: Can I just ask you, Bob, or John, 10 how long it will take to autoex all the way? What are the 11 innovations in terms of time? 12 MR. THAIN: Well, let me start. First of all, we 13 need to refile with the Commission which we would hope to do 14 in the next three or four weeks. And we've had discussion 15 with Annette and Bob about what the nature of the refiling 16 would be, so I think there's an understanding of where we're 17 going. 18 We are, as you've heard, I think in a very good 19 position in that Direct Plus exists, and if Direct Plus 20 didn't exist, this would take a lot longer. So most of the 21 changes we have to do are relatively straightforward, 22 relatively easy. So we need to remove the 1,099 share 23 restriction, we need to remove the 30-second restriction. We 24 have to remove a few other things that were coded into how 25 Direct Plus works, for instance, the hundred-share order 120 1 turns it off on either side. So we're going to take that out 2 as well. 3 And I think that inside of the review period, that 4 we would expect from the exchange, you know, we can be up and 5 operating, you know, within the next several months, whatever 6 the review process, whatever time the review process actually 7 takes. 8 The technicality of getting this to work I think 9 can be done inside the normal SEC review process. We're 10 already in the process of putting the change into the floor 11 brokers' wireless handhelds so that they can also interact 12 with the electronic market and that is already being 13 implemented as we speak. 14 MR. DONALDSON: The technology is there? 15 MR. THAIN: Yes, so technology is not going to be 16 the hangup. 17 MS. GLASSMAN: Can I just, on that question the 18 Chairman asked on how long it would take you to go autoex all 19 the way, are you planning to be autoex all the way, or is 20 there the other floor component? 21 MR. THAIN: No, we absolutely intend to run a 22 hybrid, so no, we do not intend to be autoex all the way in 23 terms of not having all of the benefits of the floor. So we 24 absolutely intend to run a hybrid market and to have the 25 markets integrate with each other. 121 1 MR. ATKINS: Are you contemplating something like 2 in Chicago with the CBOE, which almost sounds like an opt-in, 3 like Commissioner Glassman stated, an opt-in to the manual 4 side, or is there a bifurcation there? 5 MR. THAIN: No, it is, I think, similar to what the 6 CBOE has done. I don't really think it's an opt-in. It's a 7 choice, a customer choice, and so a customer can choose, and 8 you guys correct me if I don't say this properly, but a 9 customer can choose to enter an order such that it goes to 10 the floor and it gets the benefit of price improvement or 11 they can choose to enter an order that gets automatically 12 executed, so the customer is deciding which way they want to 13 trade. And by the way, the same thing is true when they give 14 an order to a floor broker. 15 So a customer can give an order to a floor broker. 16 The floor broker can also decide whether to interact with the 17 floor and put the order in as the counterside comes available 18 with the specialist in the auction process, or the floor 19 broker can choose to electronically execute it in the 20 electronic market and they can do that with pieces of it. 21 So as you heard before, they can choose to 22 electronically execute a piece of it, they can choose to work 23 the auction process for a piece of it, they can decide how to 24 do that and frankly, that is in fact how they represent their 25 customers, figuring out what's the best way to do that. 122 1 MR. ATKINS: May I ask another question here? I 2 guess one thing that was brought up with respect to CBOE as 3 well, the chairman asked about the economics of the 4 situation. We've seen obviously, with respect to New York, a 5 huge consolidation the last number of years of the specialist 6 firms. And then we've also, with respect to decimalization, 7 and other strains on the system, you know, we've seen some 8 other things that should not have happened that, you know, 9 we've obviously taken action with respect to. 10 So I'm just wondering, hypothetically, you know, 11 trying to get over to this sort of system obviously is very 12 important in order to achieve something new. How is that 13 figuring into your rearranging of your business model, have 14 you contemplated this yet? Obviously with Direct Plus, there 15 are a lot of things that you hadn't contemplated, but I was 16 wondering how it is getting into your rejiggering of the 17 system. 18 MR. FAGENSON: Value added specialist systems not 19 executing transactions that can be manually paired off. It's 20 integrating that capital to reduce volatility an maintain 21 public confidence, those instances are still going to be 22 there. The issue is these -- I think earlier, it was asked 23 what happens, the challenge is, what happens to an order that 24 reaches an execution point and doesn't get executed, and the 25 question is why. 123 1 Well, if someone executed ahead and liquidity was 2 no longer there, what does that order convert to, does it 3 become a limit order on the book, does it seek the next best 4 price, how do you reset in an auction hybrid with an autoex, 5 how does the next price get set, is autoex immediately 6 available. Those are going to be the keys. But the 7 specialist's real value in terms of helping to bring buyers 8 and sellers together is not pairing off a hundred shares at a 9 price that's almost automatically set by rote. 10 John has said, and obviously we will make that even 11 more efficient than it is today. The issue is how do we then 12 fairly price that imbalance and do that efficiently and 13 effectively without interrupting an autoex capability. And 14 that's going to be the nuance that's going to take the 15 greatest work. But the technology is there. The desire to 16 do its is clearly there. As Michael said, the customer is 17 demanding it and we have no choice but to do it and we will. 18 MR. DONALDSON: How long does it take to resolve 19 that? I mean, can you give us a feel for it? 20 MR. FAGENSON: In terms of some time -- 21 MR. DONALDSON: I understand what you're saying, 22 the technology is there, and now it's the nuance of writing 23 internal rules, if you will, how this works, right? 24 MR. GOLDSCHMID: When do we see plans? 25 MR. FAGENSON: I think the give-and-take between 124 1 the Commission and the exchange has been ongoing already. I 2 think they are far past the first chapter. I am not the one 3 who determines the timing of this. It's not as if we're 4 starting from square one. I think it will happen sooner than 5 later. I'd rather John, Annette and Bob spoke to that since 6 I'm sort of on the outside looking in. 7 MR. COLBY: We're certainly not going to speak to 8 it. 9 MR. THAIN: Mr. Chairman, I think it is -- it is a 10 matter of months. I think it's not a matter of weeks, but I 11 think it's a matter of months. 12 MR. COLBY: Many of you have said that you think 13 the floor's finest hour is when there's a large influx coming 14 down. Do you anticipate that you'll need to shut off the 15 automatic execution capability at times to handle those large 16 imbalances? 17 MR. LaBRANCHE: Can I answer that question? 18 Because there's a very important point here. You wouldn't be 19 able to add value if it's necessary; in other words, you talk 20 about -- delays and that's an automated market. If every 21 quote was automated, and the quote you saw was twenty 22 dollars, it would go down twenty dollars. That's the thing 23 that we do that is different in our market. 24 So trying to answer your question, I think, and 25 it's an important question, what we envision is a matter of 125 1 seconds with pricing the stock and then when a displayed 2 liquidity is available, when a bid is there, then you go to 3 an autoex capability. But you can't autoex something that's 4 not there, and if there is no bid or offer in the market, and 5 there's a bid/sell imbalance or buy imbalance, that's where 6 we come in, and then as soon as there is displayed liquidity 7 and we give our customers the ability to access that 8 displayed liquidity with greater speed and certainty, and 9 that's really what we want to do here. 10 MR. GOLDSCHMID: Sal, how about your timing? 11 MR. SODANO: We're also working with the Commission 12 staff in terms of the rule filings. We were working on 13 providing a new autoex technology. Our time frame is the 14 latter part of this summer, early part of the fourth quarter. 15 MR. DONALDSON: Okay. 16 MR. CAMPOS: Assuming that this hybrid model is 17 something that's beneficial, how does the proposal we have 18 right now with the variations of the opt-out play into this, 19 does that impede it, does it allow it to work within our 20 proposal, need adjustment? 21 MR. SODANO: I listened to some of the earlier 22 panels. The view of the AMEX, my view is, we do not believe 23 an opt-out solution is a good thing. For someone who elects 24 to opt out means someone is involuntarily being opted out, 25 and we don't believe it's a good thing. Limit order 126 1 protection we think is still a good thing in the marketplace 2 so we don't think the opt-out provision is something we'd 3 like to see. 4 MR. DONALDSON: We're at 12:30. We will adjourn 5 and resume at 1:30. 6 (Luncheon recess: 12:30 p.m.) 127 1 A-F-T-E-R-N-O-O-N S-E-S-S-I-O-N 2 (12:34 p.m.) 3 MR. DONALDSON: Okay, if we could get started, if 4 we could close the doors and -- 5 Good afternoon, everyone. Once again, I'm going to 6 ask the panel to identify themselves, John would you raise 7 your hand and say who you are, so we don't have to figure out 8 who is talking. 9 MR. GIESEA: My name is John Giesea, president and 10 CEO of the Security Traders Association, and I'm pleased and 11 proud to be able to participate in this process and commend 12 the Commission for both the proposal and the process by which 13 you intend to move forward and I can assure you that the 14 Security Traders Association is aggressively working today 15 towards a document that will add to the commentary as you 16 make decisions fulfilling the promise of the National Market 17 System. 18 We were proud and pleased to have created our white 19 paper in August of last year, in which we had three points 20 that we made that made it into your proposal -- 21 MR. DONALDSON: I'm going to let you come back to 22 your statement. I just wanted you to say who -- 23 MR. GIESEA: I am John. 24 MR. DONALDSON: I'll give you a chance -- 25 MR. GIESEA: No problem. 128 1 MR. LEIBOWITZ: Larry Leibowitz, executive vice 2 president, Schwab Capital Markets. 3 MR. FREEMAN: I'm Ivan Freeman, back again, last I 4 checked I'm still the chief operating officer for 5 institutional equities at Morgan Stanley. 6 MR. MADOFF: Bernie Madoff of Madoff Securities. 7 MR. PETERFFY: Thomas Peterffy, the chairman of 8 Interactive Brokers Group. 9 MR. CONCANNON: I'm Chris Concannon, vice-president 10 of NASDAQ. I'm not Robert Greifeld. If I misstate anything, 11 you can attribute it to Robert Greifeld. 12 MS. NAZARETH: I thought I'd tee up the issue and 13 then pose the questions. 14 The -- what we're talking about in this panel is 15 the linkage portion of the market access proposal. The 16 market access proposal would address intermarket access to 17 quotes and order executions in the equity markets of the 18 National Market System. A market center such as an exchange 19 that offers execution functionality will be considered a 20 "quoting market center," while a market participant that 21 enters quotations -- solely on a quotation facility that does 22 not offer order execution functionality such as the NASD's 23 ADF, will be considered a "quoting market participant." 24 The market access proposal seeks to ensure access 25 not through government imposed linkages but rather, through 129 1 linkages established by the marketplace. At its core, the 2 proposal would prohibit a quoting market center or quoting 3 market participant from imposing unfairly discriminatory 4 terms that prevent or inhibit any person from accessing its 5 quotes indirectly through a member, customer or a subscriber. 6 The standard is intended to ensure that a member, 7 customer or subscriber of a quoting market center or quoting 8 market participant can sponsor access to quotes and order 9 execution without receiving disparate treatment in the 10 handling of that order with respect to fees, speed or other 11 terms. Thus, the quoting market center or quoting market 12 participant would not be permitted to treat orders from 13 non-members, non-customers or non-subscribers that are 14 communicated indirectly through a member, customer or 15 subscriber any differently than from the way it treats the 16 orders of that member, customer or subscriber. 17 Consequently, securities market participants would 18 not need to establish direct relationships with every quoting 19 market center or quoting market participant in order to 20 access the quotes in all markets. Rather, these participants 21 would need only have relationships with a member to obtain 22 effective access to these quotes. The proposed rule also 23 would require each quoting market participant to make its 24 quotations available for the purpose of order execution for 25 all quoting market centers and all other quoting market 130 1 participants on terms as favorable as those it grants to the 2 most preferred customer, member or subscriber. 3 So to tee up sort of what the fundamental questions 4 are with this linkage proposal, is access to markets through 5 the members of an SRO or through customers or subscribers of 6 ECNs or market makers sufficient to assure fair and effective 7 access to displayed quotations? And then, are there barriers 8 to access that must be removed for this indirect access to be 9 feasible? 10 Now, I guess we could start with John Giesea. 11 MR. GIESEA: Part 2? 12 MS. NAZARETH: Part 2. 13 MR. GIESEA: I was about to make reference to our 14 submission last August where we talked about three primary 15 concerns; one being linkage and access for the purposes of 16 this conversation; two being regulation, common regulation, 17 and I would just single out in this particular context what 18 you've proposed for locked and crossed markets. We certainly 19 endorse the elimination of those types of markets. 20 And thirdly, we have been -- long been a proponent 21 of the elimination of access fees and the unfairness of one 22 segment being able to access and another not. 23 So we're pleased with the progress in terms of the 24 de minimis cap that's embodied in your proposal of one mil, 25 and the fact that all participants are able to make that 131 1 charge, and we would still prefer that to be zero, but we 2 recognize the fairness issue as having been addressed as well 3 as the de minimis fee which is not totally unreasonable. 4 With respect to the current access proposal that 5 we're discussing today, my opinion on the first part of the 6 question is, yes, those linkages are sufficient, and then the 7 second part relative to barriers, we have a concern in the 8 trading community where we have been 15 years involved with 9 fighting rogues, so to speak, that have a purpose of 10 disrupting the market rather than helping the market, so that 11 preventing the unintended here, in the access portion, we 12 believe access is the absolute cornerstone to the National 13 Market System. 14 And that any misuse of that access or quoting, 15 putting quotes in that you cannot access, that breakdown of 16 the system, is the unintended consequence, we would like to 17 see avoided. Thank you. 18 MR. LEIBOWITZ: I think that we would say that the 19 principle of open and equal access to all markets is 20 absolutely critical to the functioning of the markets. I 21 think a fair level playing field is exactly what the market 22 needs. The problem as we see it is that this is necessary 23 but not sufficient to gaining these -- to having these 24 markets function properly. 25 The problem in market access isn't a technology 132 1 problem and it really isn't even a fragmentation problem. 2 Even the much maligned ITS system is a failure not because of 3 technical limitations but because of the lack of automated 4 execution and because of a monopolistic governing system 5 where a single vote can prevent innovation in markets that 6 have opened it up and become more active. Brokers and venues 7 have built a highly efficient network, practicing markets, 8 and third party linkage providers such as Lava and other 9 vendors have thrived by providing these services for people 10 who choose not to provide it themselves or can't afford to. 11 So this principle in the marketplace has worked 12 very efficiently. The other pieces towards making the 13 markets efficient, we believe that the Commission is working 14 towards addressing; that is this access fee roulette where 15 basically rogue trading locks markets and trades solely for 16 the purposes of rebates and -- and creates market distortions 17 and unautomated markets which again, lock markets and prevent 18 the fair distribution of quotes and access to trading venues. 19 We believe that if the markets are open and fair in 20 the quotations, that the automated execution is implemented, 21 and that access fees are taken care of, then a lot of 22 problems that we see in today's markets will go away by 23 competition. Thank you. 24 MR. FREEMAN: I'm going to give you a trading desk, 25 institutional trading desk perspective on this topic, and I 133 1 also think that this is a topic, while it's isolated for 2 purposes of this panel discussion really does have a lot of 3 overlap with some of discussions that have either taken place 4 or will take place later today. 5 As we look at the entire rule, and in particular, 6 the linkage concept, we like the approach the Commission has 7 taken with respect to establishing a trade-through rule and 8 the recognition that if you're going to comply, you need to 9 have access in order to comply, and we agree with the notion 10 of having direct access or nondiscriminatory indirect access 11 to a member and customer. 12 But one of the things that we would focus on is 13 that we need to look at access to quotes as a package; that 14 is to say, if you're going to have autoex, you really need to 15 look at end-to-end connectivity, so it really is everything 16 from making sure that you have access to efficiently deliver 17 an order, and that it is processed, and that's really dealing 18 with the speed issues that we're talking about, in real time, 19 and that the response from the execution or the non-execution 20 on that order is also sent back to the original sender of 21 that order in real time. 22 The whole package is what really hangs together and 23 what we are talking about, nondiscriminatory access or direct 24 access, we think all of those end-to-end from sort of my door 25 to the market center's door and back, really have to be 134 1 included in that. 2 Moreover, as we look at providing access to the 3 quotes, we also think about having access to the information 4 about the quotes being a very necessary component to a market 5 participant that wants to play, say, with the trade-through 6 rule. So one of the things that we would ask to spend more 7 time looking at is, what is the actual minimum data 8 requirement with respect to being able to comply with the 9 rule, in addition to the actual access to the quote itself? 10 So what's the -- you need to know where to go, not 11 just what the inside market is, you need to know what is the 12 venue that's going to get me that quote and have access to 13 the data at a reasonable price in order to do that, but I 14 know there's a market data discussion later in the day that's 15 focused on a different aspect of this but we can look at 16 access to the data in tandem with access to actually 17 actionable quotes. 18 MS. NAZARETH: Bernie? 19 MR. MADOFF: I think that we can support the 20 Commission's proposal on this second piece, the access. It 21 seems to be very well thought out. You've addressed the -- 22 the issues that we were concerned about to bring it down to a 23 minimum -- a minimum amount of a fee. I think it's something 24 that we probably can all -- we can all live with. 25 As to the connectivity and the linkages, that is 135 1 something that, on the face of it, looks fine. I guess the 2 only concern that we have is that anybody that is a quoting 3 market center, and whose quote that we cannot trade through, 4 that we have to honor and that we want to honor, they must be 5 available -- those quotes must be accessible. We don't -- 6 there's some concern in the industry that someone who does 7 not have a significant amount of market share and that 8 doesn't have a linkage in place or has the cost of that 9 linkage would be something that would be -- make no sense to 10 acquire unless we could get to that. 11 Now, that shows up to a very small degree in the 12 marketplace today. But, you know, that doesn't mean that it 13 cannot become a real problem in the future, and I think 14 that's very easy to address and just to make sure that the 15 quotes that we're going to honor are going to be accessible 16 and that they will either go through an SRO that we'll have 17 to -- that those quotes have to be posted on an SRO or a 18 universal hub that's acceptable to everyone, or that we'll be 19 able to ignore that quote. 20 So I think that all of that is very doable. But 21 that's the caveat that I would mention. 22 MR. PETERFFY: I think that what Bernie said would 23 also go for market centers that do have significant market 24 share. 25 MR. COLBY: Feel like expanding on that, Tom? 136 1 MR. PETERFFY: Obviously, we are in favor of 2 efficient access because who would not be? But open and 3 efficient access does not really matter when quotes are not 4 automatically trading. As we heard from the previous panel, 5 they are working on a hybrid system but they will turn off 6 the automated market when the book is traded away. What will 7 at that time be the price that we cannot trade through? 8 Whatever it will be, it will not be automated and we'll be 9 paralyzed, as we are today. 10 This is the reason we need an exception to be able 11 to trade through the non-automated quotes and we need it as 12 soon as possible. If we do not get this, there is no 13 incentive for open outcry markets to proceed with automation. 14 They have all the financial incentive to drag their feet. 15 You must approve the exception for automated 16 markets to trade through non-automated quotes if you want 17 serious action on the part of open outcry exchanges. 18 MR. CONCANNON: The NASDAQ actually supports the 19 Commission's approach here to what I would call soft 20 linkages. We believe that, given the trade-through proposal 21 in Reg NMS, this topic of standards, while none is exciting 22 as the trade-through debate, it's critically linked to the 23 trade-through debate and you clearly point that out in your 24 proposal. Obviously, a market center that is obligated to 25 provide order interaction with other market centers under the 137 1 trade-through proposal has to be able to efficiently access 2 quotes of those other market centers. 3 We do agree with the approach used where access can 4 be done through a member or subscriber of the other market 5 center as opposed to a hard linkage, or an ITS version. 6 NASDAQ today does not have a hard linkage. It in 7 fact -- members are free to link, markets are free to link. 8 And we do have some linkages in place. I will tell you, 9 bilateral linkages between two markets are very difficult to 10 negotiate and to complete, given the competitive interests on 11 both sides. The approach that the Commission is using is 12 actually a very good approach where market centers can use 13 other members of the other market center to access the quotes 14 of that market center. 15 We will ask for further clarification on how you 16 actually achieve that. There is a market center, 17 Archipelago, that has an affiliate or a sub that is a 18 broker/dealer that accesses similar to the proposal or the 19 model contemplated by the proposal. When market centers 20 actually own broker/dealers or contract with broker/dealers 21 to access other market centers, there are questions about 22 whether it's a facility of that market center, what are the 23 boundaries of the broker/dealer that is working with the 24 market center to interact with the different market centers. 25 So we agree with the approach. We'd ask for 138 1 further clarification of how you actually effect that. They 2 are clearly complex, when a market center owns a 3 broker/dealer, what are the boundaries around those 4 complexities? And I do agree with Bernie's point about new 5 entrants. 6 What the trade-through rule will potentially create 7 is an incentive to put up a quote, knowing that I will always 8 have someone to interact with my quote, without having 9 customers that are subscribers of my entity. So I confirm an 10 ECN and have one customer posting bids and offers, knowing 11 that the trade-through rule guarantees that interaction, yet 12 I am a market share participant that's less than one percent. 13 Will the industry incur major costs of linking to 14 that one market center because of the guaranteed execution 15 they get from the trade-through rule? That is an issue that 16 does have to be dealt with. We do support the Commission's 17 approach with respect to changing from 20 percent to 5 18 percent the ATS requirements for fair access. I think five 19 percent is a very reasonable number. If you've achieved five 20 percent of market share, you have to meet certain standards 21 of fair access. You are clearly a market center, and you're 22 obligated to interact like a market center. 23 MS. NAZARETH: You and Bernie both raised the 24 interesting issue of what you do with a smaller market 25 center. You've used the one percent number. Is it fair to 139 1 require that smaller market, and what would that definition 2 be, basically, to make their quotes accessible to an SRO 3 automated execution facility because of the costs involved in 4 expecting the industry to connect to smaller players? 5 MR. CONCANNON: Yes, I think that would be a fair 6 approach to force them into an SRO execution facility, 7 clearly biased in my answer. But I do think it does 8 alleviate this cost burden to the members on how they achieve 9 best execution when there's truly a cost in communicating 10 with that sole destination despite its size. So I do agree 11 that one approach is to force them to use an SRO structure. 12 MS. NAZARETH: Bernie? 13 MR. MADOFF: Yes, I think that's fair. Look, I 14 spent most of my business life being forced into different 15 venues in order to compete with that venue. And somehow or 16 other, we were able to figure it out. I think that you, as 17 Chairman and Commissioners have said, there is a certain 18 amount of thinking that has to be done for the public good. 19 So while you would like to allow everybody to be able to 20 compete, no matter how small they are, and I, better than 21 most, can appreciate that philosophy, that being said, you 22 cannot disrupt the market. 23 So there's nothing wrong with forcing someone that 24 has not -- not significant enough to make the amount of order 25 flow or volume to -- you can't force people into his game but 140 1 you can allow them to play in that game by putting them into 2 an independent marketplace like a NASDAQ, or, as we did, into 3 ITS. 4 So I think there's no problem with setting some 5 sort of a limit and saying that somebody has to have a 6 certain percent of market share before they will be required 7 to go to -- through their linkage. It's not a great 8 imposition to go through an SRO's linkage. 9 You know, I just want to explain one thing about 10 ITS, now, maybe because 25 years ago, I was -- I was on the 11 committee to put ITS together, so I feel I have to defend 12 that to a certain extent. And everyone, when you speak about 13 ITS, people seem to feel, "Well, it doesn't work, it was 14 designed not to work." And I was a -- that's true and not 15 true. 16 Number one, it was designed not to work by certain 17 interested parties, unfortunately. But there is nothing 18 technologically wrong with the ITS system. 19 I use it all the time. You can deliver an order to 20 another exchange electronically, as fast as I certainly want 21 to trade, and we trade as fast as pretty much everyone. 22 It's what happens when that order arrives at that 23 marketplace, which is what Tom was addressing. So having an 24 automated delivery into a marketplace is only part of the 25 question. There has to be an automatic response in order for 141 1 it to truly be an automatic market. And since I have the 2 microphone, I just want to address one other issue that came 3 up this morning, but is relevant as it comes to access. 4 Don't believe for a minute that the displayed size 5 that you see in an electronic marketplace is the total size. 6 That is only the tip of the iceberg. Below the quote, the 7 displayed size that appears in most electronic markets, is 8 what we call in the industry a reserve size. 9 Once that displayed quote is exhausted 10 electronically, almost instantly, more size shows up, you 11 know, to replace that size. There is no single specialist on 12 any floor of an exchange, nor any upstairs market maker that 13 themselves today in this environment can provide enough 14 liquidity to offset an imbalance. Imbalances are offset by 15 customer orders interacting into that marketplace. It is not 16 the specialist, it is not the market maker like myself that 17 can stand there and provide the liquidity, particularly 18 during extreme moments. 19 The best way to provide that liquidity is to allow 20 customer orders to get into the marketplace as soon as 21 possible. The easiest way to do that, and certainly the most 22 efficient way to do that is through some sort of an 23 electronic capability to them to do that. 24 MS. GLASSMAN: Can I just jump in with a request 25 which a number of you have already done on this panel. But 142 1 as we get through this panel and the others later today, 2 you're talking about different components of the rule. If 3 you would just address briefly how important -- how dependent 4 whatever component you're talking about is relative to the 5 other components of the rule, because ultimately we will 6 probably be dealing with them individually in some sense, so 7 it will be important for me to understand how linked they are 8 or are not. 9 MR. PETERFFY: I think the ultimate issue is 10 automate it. As long as automated orders are executable, we 11 are fine. Once we allow manual orders to hold up the 12 process, we are all paralyzed. So I think, unless we agree 13 on trading through manual markets, this entire conversation 14 is meaningless. 15 MR. GOLDSCHMID: That is to say, where there's a 16 hybrid system, the minute they break from automatic, they 17 ought to be able to be traded through -- 18 MR. PETERFFY: Right. And that is why I'm 19 proposing that ultimately tradable limit orders be followed 20 by a character, a unique character that indicates that the 21 matching order that's accessing the system to trade with that 22 quote will be automatically executed. That's a simple 23 matter. 24 MR. MADOFF: I don't think that anybody is 25 suggesting elimination of an auction marketplace. I think 143 1 that we all understand the importance of that type of a 2 marketplace, the role that it plays, and that there is no 3 reason that it can't exist side by side with an electronic 4 marketplace. And I think you heard in the first panel, and 5 will probably continue to hear that there is a role for both 6 to play and they both can complement each other. 7 So it is, what we're focusing in on now is that 8 automated orders must be delivered automatically and 9 responded to automatically. If that displayed interest does 10 not appear, and you then have to rely upon a crowd, and some 11 people may decide they want to go into the crowd initially, 12 and I think as what you heard this morning was that the 13 challenge of the exchanges is going to be, how do they allow 14 that to happen, how do you bring the two of them together? 15 I don't doubt for a minute that the technology is 16 there to do that. But there are plenty of orders that 17 deserve to be handled by a specialist or by a market maker in 18 a manual mode. That's not the problem. 19 The problem is, do not take someone who wants to 20 trade in an automatic mode and have his order taken out of 21 the automatic mode and put into a manual mode, you know, when 22 it's not his choice to do that. 23 MR. COLBY: Could I just talk on that? Tom's 24 suggestion was that it's all right to connect the floor with 25 the automatic by designating when a quote was automatically 144 1 executable and when it wasn't. Is that a condition that the 2 rest of the market participants could live with, that they 3 could identify on a quote-by-quote basis when something is 4 tradable through -- 5 MR. PETERFFY: They currently do. Whenever we get 6 a quote, we read where it is coming from. So that when we 7 or -- because each quote is accompanied by a letter that 8 signifies what exchange it is coming from. So when we are 9 going to trade with that quote, we are going to send the 10 order to that exchange. 11 So all we are asking for is an additional letter 12 for each of the manual exchanges so that we could tell them 13 that -- whether we are sending this order to trade with the 14 manual part of the exchange or the automated system of the 15 exchange. 16 MR. COLBY: You're on a very high end of the 17 automated world. Can some of the other participants address 18 what worked for them? 19 MR. FREEMAN: Well, I think that one of the issues 20 here, and I that I we can link this to Commissioner 21 Glassman's question about what other parts are implicated 22 here, I'm talking about the market data implication before, 23 but I think that in a practical sense, there is a concept 24 that in some ways, to comply with the trade-through rule, at 25 least one market participant and probably many need to have 145 1 the network of connectivity to all the market centers. 2 And if you think about the actual mechanics of 3 executing a trade where you see shares are offered at two 4 cents, three cents and four cents in three different venues 5 and you want to take that liquidity from the street, the most 6 efficient way to do that may well be to go and send orders to 7 the two-cent venue, the three-cent venue and the four-cent 8 venue. 9 An issue could arise where each of those market 10 centers themselves having obligations under the trade-through 11 rule to receive the order, say the three-cent order comes to 12 the three-cent venue, looks at the marketplace and says, "Oh, 13 wait a minute, I can't execute this order," something I 14 discussed before, "I either have to ship it to the two cent 15 place or send it back." 16 And I think one of the things we need to think 17 about in addition to this concept that a quote might go from 18 automatically executable to not automatically executable on 19 the same -- at the same venue, we'd also have to think about, 20 can an order placer send an order to a venue and say, "I'm 21 taking, in effect, I'm taking care of the trade-through rule 22 on the two cents, please execute in your book." 23 We need to be able to have this distinction made so 24 that the trade-through rule doesn't work to your detriment 25 because everybody's trying to comply with it. So there's a 146 1 concept of a new order type that's not an opt-out, but it's, 2 I'm complying, you can trade at three cents, you -- 3 MR. COLBY: I'll just point out, that's half there 4 in the rule. The concept right now is that if the other 5 orders have been accessed during the process of being 6 accessed, the -- the rule doesn't currently have a concept of 7 you telling the -- 8 MR. FREEMAN: That's sort of my point. In other 9 words, know there's an exception that you can 10 contemporaneously do a trade-through if I'm taking out the 11 top of book in the other centers. But what I'm saying is 12 what is the behavior of the other centers? I mean they're 13 not -- they don't know unless I've told them I've taken out 14 the top of the book. Trading through would not be a 15 violation of the trade-through rule. 16 MR. GOLDSCHMID: But you want to take 17 responsibility for that, too. I take it -- 18 MR. FREEMAN: Well, what I'm saying is that if I'm 19 executing an order and I'm trying to comply with best 20 execution and trade-through requirements, and I think it's 21 better for me, instead of going serially, to go 22 simultaneously. I want to ensure that the venues I'm sending 23 the order to don't concern themselves with the compliance 24 because I've taken on that responsibility. 25 MR. GOLDSCHMID: Yes. That's what I'm -- but you 147 1 have taken on the responsibility. 2 MR. FREEMAN: They need to know that I have. 3 MR. DONALDSON: In the instance of an order that's 4 in the automatic -- let's say that you've exhausted the bid, 5 and let's say there is no big reserve there, and there's 6 going to be a gap. How do you handle the timing of filling 7 that gap? 8 In other words, how do you gather together 9 liquidity if it's not there immediately in order to take care 10 of a big block that's unexecuted that threatens to drop the 11 market in an unacceptable way? 12 MR. FREEMAN: Well, to some degree, I'll take a 13 shot at that, I think that one needs to be concerned that if 14 you have a large order, that is satisfied in part ahead of 15 you, and that you have an overage, that you need -- you need 16 an ability to have an immediate cancel. I'll take out the 17 top of the book. If it doesn't fit in, it's gone by the time 18 I get there, or it can only be partially filled, I'd like to 19 get my order -- 20 MR. DONALDSON: Let's say it's a market order not 21 held. You get part of it executed and there's nothing there 22 immediately, and -- in the -- in the manual market? How much 23 time -- how do you handle that? How much time should be 24 allowed to put together some sort of a bid or some sort of a 25 price that's closer to the last sale as opposed to going back 148 1 into the automatic market and dropping the price by an 2 unacceptable amount? 3 MR. MADOFF: We deal with that all day long today. 4 Now, but the people that deal with that, that's a competitive 5 issue. So if someone is trying to attract market share into 6 their -- into their marketplace, then they are going to have 7 to put up their capital to complete that order after they 8 have exhausted, you know, whatever is available in other 9 marketplaces. 10 So most firms like ours for sure have what they 11 call smart rallies. So we can instantaneously take out every 12 bid or offer that is displayed and then clean up the balance 13 of that order by providing liquidity ourselves. 14 If we choose not to do that, then we will just keep 15 on exhausting all of the order flow, all of the bids that 16 reappear. That's a competitive issue that you're dealing 17 with today in the NASDAQ marketplace. But people, as I -- 18 it's important to deal with what I had just brought up, that 19 the person that's controlling the order has to have the 20 ability -- has to be an -- to say, listen, once that order -- 21 and it exists today for the most part -- it's a -- it's an 22 immediate or cancel order, or fill or kill order. You put 23 the order into the marketplace. If it doesn't get executed, 24 it comes back. Then it's up to the person who is controlling 25 the order, certainly if it's a not-held order, to decide 149 1 where do I route that order, do I provide liquidity myself, 2 do I send it to another venue? That happens. 3 And that what you will see, for the most part, 4 there will be extensions in some inactive stocks, but for the 5 most part, you will see liquidity develop because the way, if 6 you look at the way the NASDAQ market operates today, there 7 is tremendous reserve size. So people don't want to show 8 their total hand by displaying all of their order at one time 9 because they are afraid that people will disadvantage them if 10 they do that. 11 The way we accommodated that in the NASDAQ 12 marketplace is giving people the ability to put reserve size. 13 That automatically rises, once the displayed size disappears. 14 So I do not think that's a real problem. 15 That does not mean that there will be instances 16 where the order will have to be held manually, and that's 17 when the specialists on the exchanges will take their role 18 and that's when market makers like myself will fulfill their 19 role. And that happens pretty much today. 20 MR. LEIBOWITZ: I'd like to echo what Bernie sees. 21 I think if you look -- the question was, what happens if we 22 blow through the book? And through a sort of trial and error 23 of what's the right amount of time to wait and seek the 24 reserve in there, there's much more liquidity than people 25 ever expected. And then brokers, through, as a service for 150 1 their customers, start to provide capital to make sure that 2 stocks don't move too much, and all of that is really 3 reflected in our execution quality statistics so that our 4 customers can say, "Oh, you're doing a good job executing 5 large orders for us," or if we were constantly blowing 6 through orders, that would show up relatively quickly. 7 MR. PETERFFY: There is always -- there is always a 8 block that is so large that it cannot just be dropped into 9 the market. So this is where the broker/dealers execution 10 comes in to play. But what is important for the broker, 11 after he has dealt with the block or reverses the block, that 12 he is able to go to the various market centers and get 13 automated executions for the pieces in which he is digesting 14 the block. 15 MR. ATKINS: To make this -- maybe to make this 16 even more like a college exam question or something like 17 that, I'll throw a couple of variables in like Commissioner 18 Glassman was saying, when we're talking about access, if we 19 adopt adequate access, and there is in the rule, we might 20 talk about that, there are two other provisions that we're 21 talking about extending to NASDAQ that I wonder, with respect 22 to all of the discussion that we had earlier on the 23 importance of limit orders and encouraging that, I wonder 24 what you think is going to be the effect of that, and that is 25 the provision of locked and crossed markets. We have no 151 1 rule. And then also the extension of the trade-through rules 2 to NASDAQ. 3 The combination of these factors, what does that 4 point to do to the marketplace, and is there something with 5 respect to access that might obviate those two provisions of 6 our rule? 7 MR. PETERFFY: We see nothing wrong with locked 8 markets. A locked market is as tight as a market can get, 9 namely, it's as tight as the Commission -- the problem with 10 locked markets is that some market centers have programs to 11 turn off their automated system and they do not work. So you 12 should ask these market centers to change their programs and 13 let the locked markets live and let customers have the tight 14 market. 15 MR. GIESEA: The STA opposed locked and crossed 16 markets for a number of reasons. The basic principle is that 17 if you accept an order, the terms under which it could be met 18 in the marketplace, it seems like a rational thing to do, to 19 seek out the bid or the offer the customer has obtained, that 20 he desires to have an execution, as opposed to entering a 21 locking or a crossing quotation, essentially, to collect 22 liquidity rebate, perhaps, as opposed to serving the 23 customer, and also giving the existing bid or offer its 24 rightful execution. 25 MR. PETERFFY: You are correct when you're talking 152 1 about customer orders but not when you're talking about 2 market maker orders. 3 MR. LEIBOWITZ: I don't see a useful purpose to 4 locked and crossed markets. I think a lot of proposals in 5 other areas of regulation actually will solve most of this 6 problem, will be access fee problems, and the automated 7 execution. 8 I think a lot of this will go away. But -- but I 9 think it's disingenuous to say that if somebody is showing an 10 indication they want to trade, a bottom line is, that's an 11 inefficient use of the market when there is actual indication 12 on the other side. 13 MR. PETERFFY: Well, as long as the market is 14 automated, and I'm quoting eleven of eleven in the hopes of 15 getting then to within one hundredths of a cent of rebate, 16 I'm doing a public service and I'm doing it all by myself. 17 MR. ATKINS: I guess that's what I was hoping to 18 get in discussion, because if you do iron out these 19 extraneous impediments like access fees or whatever, then do 20 we really need it and then are we going to create unintended 21 consequences through an artificial prohibition, I believe, so 22 that's the response -- 23 MR. CONCANNON: Yes, I agree with Larry that some 24 of the other components of the proposal could actually 25 alleviate something like locked and crossed markets. The 153 1 access fee is an issue. People do lock and cross to obtain a 2 rebate. That is a trading strategy today. 3 Changing the pricing of access could actually 4 change the incentives of lock and cross. Access terms are 5 another issue, why there are locks and crosses. People that 6 don't provide automated execution or linkages, whether 7 bilateral or customer electronic linkage, are going to lock 8 the market. They have stale quotes. They aren't being 9 updated, aren't being accessed. 10 So you will have a situation where a broker has a 11 valid reason to post a quote against a slow market quote, so 12 if you can trade through that quote, arguably, you should be 13 able to lock it. 14 There are similar issues in both decisions. But I 15 do think locks and crosses do provide confusion for investing 16 public. What is the true market when it's locked? More so 17 when it's crossed? So I think discouraging locks and 18 crosses, providing other parts of the rule so that it 19 actually provides enhancements that avoid the incidents of 20 locks and crosses, is actually a very good approach. 21 Your question on trade-through and NASDAQ, I know 22 this is access standards, but I did want to try and address 23 it. When we look at, we've had this discussion about what 24 will happen and what will happen with an opt-out. 25 NASDAQ is the test case. The market is an opt-out. 154 1 It's governed by best execution, and in fact, limit order 2 display is happening every day and it's very aggressive limit 3 order display. Our quotes, our spreads are extremely narrow, 4 and they are more narrow in these stocks than the market that 5 has a trade-through rule in it today. 6 So I would argue that the opt-out, while we support 7 Reg NMS and the trade-through component of it, we do believe 8 the opt-out allows for this aggressive limit order display. 9 It won't discourage it. It's actually the access portion of 10 it, the automated features will encourage the use of limit 11 orders and that's what happened on NASDAQ. As you introduced 12 more automation, and first it was ECNs, and NASDAQ responded 13 to that competition, people are using limit orders on NASDAQ. 14 And I think we heard that from the prior panels, especially 15 from the buy side, that they do like trading in the NASDAQ 16 market as opposed to the market that has the trade-through 17 rule. 18 MR. LEIBOWITZ: I was just going to say, it seems 19 hard to argue with the principle of not being able to trade 20 through an automated market. Just in principle, that seems 21 to make sense, whether it's NASDAQ or anywhere else. But 22 what it's really going to come done to at the end of the game 23 is, what's the practicality of enforcing it? Some of the 24 examples that Ivan gave of, you know, sort of ending up in a 25 circle where orders get shipped from place to place because 155 1 limit orders keep getting cancelled or changed, and so that 2 you never really know where the market is. 3 The question is in the electronic world, how will 4 we enforce that in real time, whereas if it becomes something 5 that's part of an auditable thing, because right now, when 6 there's a trade-through, the trade-through rule is, call the 7 New York Stock Exchange and complain to professionals. 8 Well, that can't happen in an electronic market. 9 You can't get satisfaction that way. So how will we actually 10 do it? 11 MR. COLBY: Given the interest in moving towards 12 automatic markets, should we be going farther than just 13 having a trade-through rule with a fast market exception, 14 should we be mandating automatic execution or saying that 15 non-automated quotes are not part of the best bid and offer? 16 MR. CONCANNON: Well, if I could address that, I do 17 think the fast or slow market approach could have a dramatic 18 effect. Markets -- and we're hearing it today -- markets are 19 committing to being fast for fear of being deemed slow. It 20 will dramatically affect a broker's decision of where to 21 place his limit orders. So there is this race before you 22 even actually approve the rule, to be fast. 23 I am concerned that there is this option of slow 24 out there. And that under the trade-through rule, you have 25 the de minimis standards of -- even though they are slow, you 156 1 still have to respect that quote. I do believe at some point 2 there is a business model in being deemed a slow market. In 3 fact, by sitting out on a quote, you obtain tremendous amount 4 of information, trading information because people are 5 required to try and hit you that there could be a business 6 model in being deemed a slow market or participating in a 7 slow quote. 8 And I was a little bit concerned about Tom's 9 proposal to just deem a quote slow or fast, in real time. 10 There are times, I like the concept where you better be fast 11 or else no one's going to put a limit order on you. That's a 12 very good competitive environment. 13 MR. PETERFFY: I think you absolutely must have 14 standards. Otherwise, everybody will go to the lowest common 15 denominator, namely, we'll take as long as possible. If you 16 give me two seconds, then I can wait for two seconds and see 17 what is the lowest bid that has occurred during those two 18 seconds, and that's where I'm going to buy my customer's 19 order. So I don't think that's fair. If you do that, you 20 will create a circumstance in which the bad guys will make 21 all the money and the good guys will go out of business. 22 MR. MADOFF: I have a hard time, you know, 23 addressing the issue as to what's a fast market or a slow 24 market. And certainly, I don't think that we have to get -- 25 we don't have to -- do not have to define it down to the 157 1 nanoseconds or whatever we were talking about this morning. 2 You're going to have to -- first of all, I think 3 that it's going to be very difficult for the Commission to 4 say everyone has to be fast or that everyone has to be 5 automated. I do not think that's the direction we want to 6 go. I think that letting the competitive forces work will 7 accomplish what you really want to accomplish anyhow and I 8 think that if you say that automatic execution involves order 9 delivery, and automatic response, and what you want to say 10 that's one second, whatever, everyone will be able to live 11 with that, I think in the real -- in the real world. 12 I do not -- I do not see a problem with the access 13 fee issue as you've addressed it, as I said earlier. I think 14 that whether or not you need a trade-through rule is a 15 different issue. I think that if you have the linkages, you 16 deal with the access issues. People are not going to trade 17 through markets. And clearly, they are not going to trade 18 through automated markets. 19 So you know, I don't think we have to deal with, do 20 we force everyone to be automated or not. I think what 21 you've witnessed this morning was the so-called slow markets 22 are going to become fast markets because of the competitive 23 reasons that exist in this marketplace. 24 So, and to answer the question, I guess, that Bob 25 asked before, can markets read an identifier, which is an 158 1 important issue to those that operate in an automated market, 2 the answer is yes. Our system -- an automated system is an 3 automated system. So everyone's automated system will be 4 able to read whether they're sending an order to a 5 non-automated system market or not. And I think that 6 clearly, that's what you want to happen. 7 I want to know that the order is going to an 8 automated market, and that that may not be executed 9 automatically. That happens all the time in today's world. 10 I can send orders to other automated markets. I do not get 11 an execution if in fact, you know, there's a limit on that -- 12 on that order. And that's fine. As long as I get -- I get 13 it back immediately and I know that I didn't get that 14 execution, that's as much as I can ask you. 15 MR. FREEMAN: Bob, I'd like to address your 16 question by expanding it a bit, which is to say, are there 17 other things that we should be mandating in terms of the 18 operation of a market participant when we're having all these 19 linkages, and I'd like to raise two things: 20 Maybe we should be thinking about uniformity across 21 the market centers and trading halts and how to trade those. 22 And a potentially systemic issue also is the disparity of 23 different market centers and how they handle erroneous 24 trades. And as market participants go and dispute through 25 various means, end up in places they may not know where they 159 1 are actually ending up, if they are not looking at a uniform 2 set of erroneous trade procedures, it might pose some sort of 3 risk to the system. 4 MR. DONALDSON: I think we're -- 5 MR. GOLDSCHMID: One last quick one, or maybe not 6 quick. What will this do to ITS, and does it matter if we 7 take this route, Bernie, you -- 8 MR. MADOFF: I don't think it matters. I think 9 that ITS, there's going to have to be some sort of a linkage. 10 The problem with ITS, which was, you know, discussed this 11 morning, is a governance issue. It's not a technology issue. 12 It's the fact that in order to make enhancements to ITS, you 13 have to have a unanimous vote, which anyone who has 14 involvement with this industry knows is impossible. And the 15 problem is that people confuse that ITS is an automatic 16 linkage. It isn't. I deliver my orders automatically, then 17 they turn into manual orders down at the exchanges. 18 Now, that also, if you just -- if the New York 19 becomes a fast market, my guess would be, you could use ITS 20 as it presently exists with no problem at all. Now, maybe 21 somebody wants to build a better technology platform than 22 exists in the pipe. But I have to tell you I'm as large a 23 user of ITS as exists today, and we probably do ten percent 24 of the transactions in New York Stock Exchange issues. I 25 have no problem with ITS as a technology platform. 160 1 I have a problem with it once it gets to the 2 ultimate destination. 3 MR. FREEMAN: I think one of the things we also 4 need to consider with respect to any hard linkage is, to the 5 extent that it becomes a market standard, should we be 6 concerned about single point of failure. If everyone is 7 grabbing onto it because -- maybe we have to think about that 8 as well. 9 MR. LEIBOWITZ: I think you have a perfect example 10 of that in NASDAQ, where there was a time where people really 11 felt that NASDAQ couldn't handle the stress of the system and 12 what ended up happening is, ECNs created direct connections, 13 brokers created -- so this sort of third party network sprang 14 up around NASDAQ because it wasn't capable of handling it and 15 I think that's all happened here. 16 MR. COLBY: ITS gives a guaranteed way to get into 17 an exchange. Now, the execution process when you get there 18 may not be as good. If it's not there, are the other routes 19 into these exchanges adequate? 20 MR. CONCANNON: I would say yes. People have built 21 these linkages as need -- 22 MR. COLBY: But Chris, they haven't built them 23 in -- we're assuming they would build -- 24 MR. PETERFFY: That will be applicable to the 25 listed stocks as soon as the listed stocks trade somewhere 161 1 where it's worthwhile to access it. 2 MR. CONCANNON: And actually today building around 3 ITS because of the speed and the response time issues, so 4 people do have -- brokers have hard wires into multiple, you 5 know, regional exchanges of trade listed and they do sweep 6 through the listed markets. And avoid ITS for the reasons 7 like Bernie mentioned, not because of the technology, but 8 because of the response time. 9 But there is a big issue on ITS and it's the hard 10 linkage issue. It's a free access into the markets. It 11 devalues the -- the value of membership. Why should I become 12 a member of an exchange or an SRO when I can be a member of 13 one SRO and just access all the other SROs? So the approach 14 that he -- has used is actually a very good approach using 15 the member to access the market. The value of the seats on 16 New York will dramatically fall when ITS is enhanced and give 17 automated execution. 18 That same rule applies to NASDAQ. Why would anyone 19 be a member of NASDAQ if they can go through a hard linkage 20 and be a member of an exchange that has lower standards and 21 less regulatory issues for them? Those -- 22 MR. LEIBOWITZ: Of course what that would say is 23 that ITS should really go to work on access speed methodology 24 more like what applies in the rest of the access speed 25 portion and that might make the whole problem wash out. 162 1 MR. CONCANNON: Yes, potentially that could be one 2 way to do it. I do like the approach that the Commission set 3 forth the membership approach. 4 MR. PETERFFY: I would like to correct myself if I 5 may. About the DOT routing system being terrific and 6 obviously, it works very well. All we would like to see is 7 that being extended. 8 MR. DONALDSON: Thank you all very much. 9 (Recess taken.) 10 MR. DONALDSON: All right, we'll ask this panel to 11 do a hand up in the air so people in the back of the room 12 know who you are. Bob? 13 MR. BRITZ: Bob Britz, I'm president of the New 14 York Stock Exchange. 15 MR. PUTNAM: Gerry Putnam, CEO of Archipelago. 16 MR. O'BRIEN: I'm Bill O'Brien, chief operating 17 officer for Brut ECN. 18 MR. NICOLL: Ed Nicoll, CEO of Instinet Group. 19 MR. LAVICKA: Matt Lavicka, managing director at 20 Goldman, Sachs. 21 MR. JOYCE: Tom Joyce, CEO of Knight Trading Group. 22 MS. NAZARETH: Likewise, we're going to continue 23 our discussion of the market access proposal, but this panel 24 will be focused on the access fee portion of the proposal. 25 As you know, the market access proposal would also establish 163 1 a de minimis standard for access fees. It would assure 2 access to public quotes for a fee that could not accumulate 3 to more than two tenths of a cent per share in any 4 transaction. 5 The application of the fee standard to different 6 entities would depend on how a quote was accessed by an order 7 router. Such access can be divided into three categories: 8 First, quotes could be accessed through an SRO 9 order execution facility defined in the proposal as a quoting 10 market center, which would be authorized to charge a fee of 11 no more than one tenth of a cent per share. 12 In addition, if a quote was attributable to a 13 particular broker/dealer, the broker/dealer would also be 14 authorized to charge a fee of no more than a tenth of a cent 15 per share. 16 Second, quotes could be accessed through a 17 broker/dealer whose quotes were not accessible in any SRO 18 order execution facility. Such a broker/dealer is defined in 19 the proposal as a quoting market participant and he's 20 authorized to charge a fee of no more than a tenth of a cent 21 per share. 22 And third, quotes could be accessed directly 23 through a broker/dealer whose quotes were also accessible 24 through an SRO execution facility. Except for the 25 accumulated fee cap of two tenths of a cent per share, this 164 1 type of direct access is not specifically covered by the 2 proposal and would continue to be subject to the existing 3 rules with respect to access fees. 4 The market access proposal also would require every 5 SRO to establish and enforce rules requiring its members to 6 avoid locking and crossing quotations and enabling market 7 participants to reconcile locked and crossed quotations 8 before effecting a trade. The SRO rules also would prohibit 9 members from initiating a pattern or practice of locking and 10 crossing quotations. 11 I guess we'll start with the basic questions which 12 are, at what level do fees associated with access to a 13 market's quotes become a problem that hampers the fair and 14 efficient access to quotes and are competitive forces alone 15 sufficient to address the level of access fees. We'll start 16 with Tom. 17 MR. JOYCE: I don't think competitive forces alone 18 are enough to control access fees. In point of fact, a 19 little history, access fees were introduced in the Order 20 Handling Rules that the SEC endorsed in 1996. ECNs at the 21 time, ATSs in general, were given the ability to charge a fee 22 to members, other broker/dealers, who were forced because of 23 the best execution rules to go to that ECN. So even though 24 you were a non-subscriber, you were forced to go there and 25 ATSs were then given the right and no other broker/dealer had 165 1 that right to assess a fee to you. 2 So this was introduced by regulatory fiat in 1996. 3 While it was clearly not a level playing field, which was to 4 say no other broker/dealers had the right to do that, this 5 was established and got credence in the market and the next 6 thing you knew there were ten ECNs or ATSs, or both. So 7 clearly, with the proliferation of ECNs, the proliferation of 8 ATSs, all the broker/dealers that were out there on the other 9 side of the equation, competition did not do almost anything 10 to knock down access fees. 11 In point of fact, ATSs assumed as their given right 12 to charge access fees while the rest of the market 13 participants could not. So I believe that they were 14 artificially imposed on the market and they should be 15 eliminated entirely. I don't know if there is a minimum 16 amount of an access fee that is appropriate vs. not 17 appropriate. 18 Another bit of history, one of the reasons access 19 fees were introduced in 1996 was because it was the view of 20 some people at the Commission that market makers have the 21 ability to capture spread and as a result, ATSs were, quote 22 unquote, disadvantaged. 23 Well, over the competitive environment that's 24 occurred since then, the -- because of decimalization, and 25 now, again, another regulation seems to be coming our way, 166 1 the ability to capture spread that supposedly market makers 2 had has been virtually eliminated yet access fees are still 3 here. 4 So I am of the firm belief that competitive forces 5 have done very little to eliminate access fees, that access 6 fees were artificially introduced into the markets, and 7 frankly, the only good access fee is zero. 8 MR. LAVICKA: We believe that markets should be 9 fair, open and transparent. However this transparency can 10 only be achieved if quotes comprising SEC mandated national 11 best bid and offer reflect a true and complete value of the 12 transaction. Best execution obligations, the 11Ac1-5 rule, 13 and the proposed trade-through rule, have elevated NBBO to a 14 key role as the benchmark for measuring price and execution 15 quality in the equity markets and for this reason, when it 16 comes to the NBBO, we believe that what you see should be 17 what you get. Access fees that hide behind a quote means 18 that the quote is not what it says it is. If we're going to 19 continue to use the NBBO as a benchmark for price discovery 20 and best execution, then the SEC must ensure that the NBBO 21 quote is an accurate representation of price. 22 Apart from these transparency concerns, we also 23 believe that, for a quote to be part of the NBBO, it should 24 be possible to execute against that quote in an efficient 25 manner. In other words, in addition to being what it says it 167 1 is, the NBBO should be available for immediate execution. 2 Subjecting the quote to this requirement will help ensure the 3 validity of using the NBBO as benchmark. 4 Also, it should be noted that access fees may 5 create an alternative way of quoting in subpennies, something 6 that Regulation NMS proposes to eliminate. And for this 7 reason as well as the transparency and accessibility reasons 8 already mentioned, it is imperative that the SEC ensure that 9 access fees do not distort the quote. 10 We believe that our markets are robust precisely 11 due to the extent of the lively competition that exists. 12 However, we think that competition should be based on the 13 value of services provided to clients sending orders to a 14 particular venue, services such as providing capital, deeper 15 liquidity, flexible and sophisticated order types, and smart 16 order routing and handling. It should not be based on hidden 17 fees or burdens on intermarket access to a venue's published 18 quote in the NBBO. 19 In summary, if the NBBO is based on transparent and 20 accessible prices, then broker/dealers like Goldman Sachs 21 will be able to fulfill their duty to seek best execution for 22 their customers' orders and market participants will be 23 assured of an accurate, meaningful and reliable benchmark. 24 Thank you. 25 MR. NICOLL: I'll try and answer your question. I 168 1 don't know -- I don't know where there -- whether anybody has 2 done a study or a credible analysis on where access fees 3 begin to disincentivize people from entering the marketplace. 4 I don't know where that is. That's why I think it should be 5 established by competition between market venues. Just like 6 I don't think anybody knows what the right price of many 7 things are, except as it's established by forces of supply 8 and demand. 9 Let me be absolutely clear. For agency markets to 10 exist, they have to charge agency fees. Instinet, by the 11 way, has been from the very beginning and still is prepared 12 to represent its access fees within the quote itself, but 13 that would require quoting in pennies. So I hardly think 14 that they are hidden. They are published, they are uniform, 15 because we are a fair access ECN, everybody knows what they 16 are. 17 And they are necessary to create an agency market. 18 Otherwise, we will have dealer marketplaces. Can't have it 19 both ways. So the question is not should there be agency 20 fees. The question is, is there a need for the Commission to 21 cap them and to take a position about what the appropriate 22 level of agency fees is for agency marketplaces. 23 MR. O'BRIEN: Thanks for the opportunity to speak 24 to you all today. I think in deference and in partial 25 apology for speakers on this panel that represent large 169 1 customers of Brut, I'm going to, in answering your question, 2 I think counterpoint to some of the statements they have 3 already made today. 4 It was interesting. At the end of the last panel, 5 it talked about the problem of ITSs and zero cost structure 6 devaluing the member, the value of exchange membership. 7 ECNs, prior to the Order Handling Rules, really, Instinet was 8 the only one, did charge access fees at their locations going 9 back to their start in 1969. I think to kind of balance the 10 interests of getting their order prices into the public quote 11 and to improve market transparency, they allowed, they didn't 12 impose by fiat, but they allowed continuation of charging for 13 access to those quotations via the public quote system, 14 notwithstanding the fact that a firm had not entered into a 15 contractual relationship with the ATS operator, and I think 16 it was kind of validating, an example of the value of 17 maintaining the value of membership on a customer basis, does 18 make sense. Has competition driven those fees down? 19 Absolutely. 20 Before 1996, the average fee captured by an ATS was 21 three cents a share. Remember, Instinet at the time charged 22 liquidity provider and charged liquidity taker. Today net 23 fees for ECNs are under a tenth of a cent. Talking about a 24 95 percent reduction of those fees on that basis. So it's 25 competition by and large is working to keep those fees as a 170 1 reasonable level, and as Ed pointed out, there is a reason 2 for those fees. We talked a lot earlier today about how 3 limit orders have value and promoting them, encouraging their 4 display, rewarding the people that display them is something 5 that needs to be furthered by a market structure. 6 Current market center fee structures do just that, 7 and you need to be able to preserve that. And you know, in 8 the previous panel, a couple of speakers, talked about rogue 9 traders, that people providing limit orders are some way, you 10 know, not worthy of our respect. It's a good economic 11 outcome, it's a good public policy outcome. 12 On the way to decimalization, there has been a 13 significant, you know, influx of liquidity into the market by 14 non-traditional firms providing net liquidity to the 15 marketplace, taking advantage of the economic model that 16 agency markets offer. You disregard liquidity and the value 17 of those net liquidity to the average investor really at your 18 peril. 19 So it is a valuable fee structuring, and 20 competition by and large is driving it down. The problems 21 that some of the panel has talked about, distortions, I just 22 flatly don't agree with it. And the idea of NBBO somehow 23 being distorted is just not true. The average investor that 24 places an order on an on-line platform at Schwab or 25 Ameritrade or Fidelity, sees price. They have no knowledge 171 1 of these fees, it's netted out to them in the form of the 2 commissions that they pay. In terms of intermediaries, they 3 are very highly aware of these fees. They are driving these 4 fees down to historically low levels. And they can -- they 5 can operate their business model successfully in ignoring 6 them. 7 So I think by and large, competition is working to 8 keep these fees at reasonable levels. I think where 9 competition may fail is the instance that was spoken about in 10 the last panel, kind of a de minimis market center that has a 11 right to put an order into the public quotation and cause a 12 lot of disruption with a very small amount of order flow that 13 can attempt to, you know, extort rates that are not 14 consistent with, you know, a disciplined market with market 15 rates, and that creates a tension for Matt and for Tom, given 16 their duty of best execution to, A, build the linkage and pay 17 for it, whether it's on their own or through a third party 18 vendor that is inefficient relative to the liquidity that 19 market destination may have; and two, accept fee structures 20 that are just not competitive. 21 So some of the ideas that were bandied about a 22 little earlier about making de minimis markets participate 23 through SROs or execution facilities really would do two 24 things: One, it provides efficient linkages, and two, it 25 subjects those markets to the rule sets of SROs like NASDAQ 172 1 has today to keep their rate structures at what the market 2 views as reasonable levels. 3 MR. PUTNAM: We strongly support the 4 nondiscrimination policy that is part of the section of Reg 5 NMS. ArcaEx has been over the years a strong proponent of 6 best execution, always finding the best price regardless of 7 where it traded. We actually pioneered that aspect of the 8 business. And as a result, we've also been on the leading 9 edge of the unfair access fee issue. 10 We've actually gone through a period of time where 11 we had competitors who came up with rates specific to us, a 12 competitor rate, and it was two to five times what they 13 charged everybody else. 14 When we fought it, and refused to pay in some 15 cases, as a way of fighting those fees, our access was 16 threatened. So threats of, "We'll pull your wires out of the 17 wall if you don't pay this extortion fee because we know your 18 model's best ex, and there we can take advantage of you." 19 That said, we are entirely against the idea of the 20 government setting rates for us. Access fees have dropped 21 dramatically and you heard earlier from Tom, and when it 22 started, it was a relatively high fee. Fees have gone almost 23 to zero at this point. It's been the result of 24 hypercompetition. There is no problem in the marketplace. 25 Market makers in NASDAQ today are actually paid for access to 173 1 their quotes. NASDAQ pays a two tenths of a cent per share 2 rebate to them when someone accesses their quote. 3 I would say if you're going to go down this path, I 4 mean why stop at market access? And many of you know, we 5 filed an S-1 to go public. And as a result we have to use an 6 investment banker. And I know there's been a lot of debate 7 over the fact that investment bankers tend to charge a 8 relatively high, very similar fee. So why not go to the next 9 step and lower the fee that they can charge us for doing our 10 investment banking business? 11 And then there's the matter of my on-line broker. 12 They all tend to charge the same fee today, so why not take a 13 whack at them? 14 I think that some of what we're heard earlier from 15 the CBOE, I don't know that it was Bill Brodsky or Ed Tilly, 16 they are talking about how they are getting people into their 17 electronic quotes and the comment was, "We reward 18 participants for adding liquidity." 19 Well, we reward participants for adding liquidity 20 and I think John Wheeler earlier mentioned that you look up 21 in the option pages and you see those option prices, well, 22 there is option value in those limit orders. And the takers 23 of liquidity are willing to pay a fee for unfettered access 24 to those quotes. So the market has set a fair rate. 25 Those who provide or are willing to stick their 174 1 neck out and provide quotes get paid in exchange for letting 2 people access them. We've talked a lot about protecting 3 limit orders and the value of limit orders here. You take 4 the reward away for displaying limit orders, all of a sudden 5 we're back at some policy that's going to threaten limit 6 orders again. So again, I think that is exactly the wrong 7 place to go. 8 We do not like the idea of somebody telling us what 9 we can charge for our service. That's the issue which Bill 10 brought up about a few ECNs that set up these toll booths in 11 the middle of the execution highway and charge exorbitant 12 fees. Well, it's over with. That's done in the marketplace. 13 We, they have been there. They are not there today. NASDAQ 14 solved its problem by passing a rule, which is where we think 15 this should be addressed. SROs passed rules like NASDAQ did 16 that said, "We're going to limit what you can charge in our 17 marketplace. If you don't like it, go someplace else." 18 We can choose to do the same thing, so let the SROs 19 determine price, let competition deal with the issue of 20 getting price gouged like you all liked to do with us, 21 because we did get gouged for years. We suffered through it. 22 The marketplace determined access to the best price was the 23 best thing. So now those guys that were gouging us all do 24 what we used to do, which is seek the best price regardless 25 of where it is, and that took care of the price tabs problem, 175 1 as they obviously don't want it done back to them. 2 MS. GLASSMAN: Your putting the fee in the quote, 3 even though it would involve some fees? 4 MR. PUTNAM: Another way to do it would be to put 5 the fee in the quote. Everybody charges the same thing 6 today, so we all know what the fee is anyway. So, you know, 7 I mean, a hundred percent of the marketplace is charging 8 roughly the same price. So everyone can do the math. It's 9 sub one penny. It's sub a half a penny. It's actually 10 closer to a quarter of a cent or three tenths of a cent per 11 share. Everyone can do that math. 12 MR. O'BRIEN: I would just ask the question, what 13 would the value of doing that be to the individual investor, 14 would the commission rate charged by an Ameritrade or Goldman 15 Sachs or Knight go down because they explicitly absorbed 16 these fees in the price of transaction? Why not do that with 17 SEC fees or the NASD's trading activity fee or clearance and 18 settlement fees? 19 Those fees don't necessarily need to be transparent 20 to the end user provided that they are transparent to their 21 intermediaries and there's competition among participants 22 regarding them. 23 MR. PUTNAM: What if you chase off the really 24 efficient modern market maker, the one who comes in and views 25 that two tenths of a cent that they receive for providing 176 1 liquidity or displaying limit orders in the system, it's part 2 of their -- a big part, in many cases a big part of their 3 economics. They are so nimble and so quick, but enormous 4 liquidity providers, if you take that away and they just 5 disappear, what to we do with limit orders? We just shrunk 6 the supply of limit orders in the marketplace. Spreads are 7 widened if they go away. We had a lot wider spreads before 8 those folks showed up than we do today. 9 MR. BRITZ: I think I'm probably going to end this 10 particular rotation where T.J. began it, but I'd like to 11 start by saying, while I'm aware that this has been a hotly 12 debated issue in the over-the-counter market for quite some 13 number of years now, it's not been an issue in the listed 14 market. 15 So I think we come to this issue in a somewhat 16 different way than many of the other people around this table 17 who have been living and breathing this issue for X number of 18 years. If nothing else, we've come to it with fair, fresh 19 eyes. In fact, very fresh eyes. It wasn't until the 20 publication of Reg NMS that we realized maybe this really was 21 an NYSE listed issue. In fact, even after the publication, 22 and having read Reg NMS, we respectfully disagree that this 23 actually does border on to the NYSE domain. 24 And the simple reason is, the NYSE has never 25 induced anyone to send it an order via a cash payment. And 177 1 so we have never had any need for access fees to fund those 2 kinds of schemes. 3 But Reg NMS does something that did get our 4 attention and it says, SRO, or where it suggests that SRO 5 transaction charges ought to be categorized in the same way 6 as access fees, and it says that notwithstanding that ECN 7 access fees are in addition to SRO transaction charges and no 8 such additional fee exists at the New York Stock Exchange. 9 ECN X's fees are true variable costs and are not 10 transparent in real time whereas NYSE transaction charges are 11 completely transparent, known before the trade, and because 12 of the very structure of those transaction charges, they act 13 much more like the fixed charge and a variable cost that can 14 be tagged to any particular trade. 15 And lastly, the NYSE levied these transaction 16 charges on its members, its customers, and those members, and 17 not incidentally, the SEC, approved those charges. ECNs levy 18 access fees on non-customers. There was no privity. Those 19 customers have not agreed to the charge and indeed, at the 20 time of the trade, may not know that they are going to get a 21 bill from an ECN on behalf of the counterparty to that trade. 22 So I want to be on record first and foremost saying 23 we don't see the comparability yet between ECN access fees 24 and SRO transaction charges. 25 I'll take a shot at answering straightforwardly 178 1 some of the questions, and here's why I said I think I'm 2 going to end up where T.J. began. Having said all of that, 3 bids and offers that have access fees attached should always 4 be differentiated from those that don't. There should be no 5 de minimis exceptions, de minimis exceptions or an attempt to 6 rationalize that there is no difference between an offer that 7 has a two mil access fee tagged to it and an offer that has 8 no access fee. Well, they are fundamentally different and no 9 amount of rationalization is going to change that. 10 So first point is that there always ought to be 11 differentiation, and so the question as to whether or not the 12 market forces and competition would work. I'm not sure I 13 have the definitive answer to that, but I know they can't 14 work unless there's full transparency and for differentiation 15 of those orders that have a tag and an access fee and those 16 that do not. 17 As far as its form, differentiation can take any 18 number of forms. I suspect probably we'd have a unanimous 19 vote in this room that I'm the last person that you'll want 20 to architect the format here. And there's nothing new. It 21 can yield, an order that has an access fee can yield to those 22 that do not in terms of priority. You can round up or round 23 down, an order that's got an access fee to the nearest penny, 24 and probably lots of different other schemes that can get you 25 to the same result. 179 1 There was another question as to whether or not 2 allowing everyone to charge access fees is somehow a remedy 3 to this problem. And I'll just give you one man's reaction 4 to that. 5 I think allowing everyone to charge access fees to 6 validate the practice of the few that today do charge access 7 fees smacks of regressing to the lowest common denominator 8 and I haven't seen any analysis that the Commission has done 9 in this regard, but I would be concerned that it would add to 10 trading costs rather than diminish trading costs. 11 There was another question about locked markets, 12 and I heard a little bit of the discussion on the prior 13 panel. I can't get my mind around those who think that 14 locking markets is somehow a public service. That may just 15 be me. 16 Locked markets are, without a question, a system of 17 an inefficient market. If somebody wants to buy for fifty 18 bucks and somebody wants to sell for fifty bucks in an 19 efficient market, they trade, they don't lock. I think they 20 should be dealt with in terms of regulatory action. I 21 would -- you haven't asked for my advice, but I'll give it to 22 you. If you ban the practice of requiring liquidity rebates, 23 you get rid of contrived locked markets very, very quickly. 24 And liquidity rebates are payment for order flow by a more 25 delicate name. So I think that would be a good place to 180 1 start. 2 And I guess I'd wrap up by saying, I think the 3 current proposal, A, pretends that access fee orders and 4 non-access fee orders are the same thing when they are not, 5 gets the Commission into the rate making business and I 6 always understood you were loathe to do that. It's an 7 indirect and likely an ineffective way of dealing with locked 8 markets and it imposes a solution on markets where the 9 problem doesn't exist. Thank you. 10 MR. COLBY: We've had a very sharp difference of 11 opinion, I can say here. But in the middle they've talking 12 about rogues and outliers and how much they exist, and the 13 problem there is that there have been some, maybe, smaller 14 ATSs that charge considerably more than other ATSs, all 15 right? So there's been some consolidation of rates down to 16 the three tenth of a cent level, but then there have been 17 outliers. And that's been in the non-trade-through rule, 18 best execution environment. Yet somehow they've managed to 19 survive at that. 20 If you have a trade-through rule, what happens in 21 that context? Can you operate this partial access fee, 22 partial non-access fee, different rates, does everyone then 23 have an incentive to cut an outlier if you've got a 24 requirement to go to a quote but you can attach a fee to it? 25 MR. JOYCE: I guess I'm starting. Just as a point 181 1 of clarity, though, I just want the Commission to understand 2 that in some way with all due respect, disingenuous to 3 compare commissions attached to working an order as a 4 subscriber on behalf as an agent, and an access fee which is 5 charged without debate to a non-subscriber. 6 So to say that Instinet, and there was only one ECN 7 before 1996, saying Instinet's commission charges that we at 8 Merrill Lynch -- at the time I was at Merrill Lynch, paid 9 happily to access liquidity, as a subscriber, as a willing 10 client of Instinet, to compare that with an access fee which 11 we are forced to pay whether we want to or not as a 12 non-subscriber is frankly two completely different things and 13 should not, probably be permitted. 14 MR. PUTNAM: Tom, you and Bob will have to go back 15 and read this, because this isn't just nonmember access fees. 16 This is a fee that we can charge a subscriber. So the 17 specialist on the New York Stock Exchange changes his 18 business model to become more electronic, he may want to 19 decide to change on how he charges for access to his 20 liquidity, because you can't enter position trades anymore, 21 you can't take advantage of things that you took advantage of 22 in the past, as a way of making money. 23 You may decide to charge an access fee for access 24 to this brilliant marketplace. This is a limit on that fee. 25 It is not just non-member access. 182 1 MR. JOYCE: I'm not saying it is, with all due 2 respect, Gerry. I'm saying comparing the commissions that we 3 paid to Instinet before 1996 and access fee today, whether a 4 subscriber or non-subscriber, they are not the same thing. 5 They are just fundamentally not the same thing, having 6 written checks for both kinds of bills. 7 MR. PUTNAM: That's over with now. 8 MR. JOYCE: I agree, but -- 9 MR. PUTNAM: It's over with now. We're trying to 10 solve a problem that doesn't exist anymore with government 11 mandated rate making. 12 MR. LAVICKA: It's not over yet. You still have 13 the fact that the SEC has created this NBBO. It's holding us 14 to it because of best execution, 11Ac1-5. I think you have 15 the trade-through rule, you're just giving that much more 16 weight to what the NBBO that we're going to have to go and 17 access. So you've already mandated that, you're already in 18 the game, so it's a public good that needs to be protected. 19 We need to access it, and I don't know how you can avoid 20 having to regulate the fees for accessing that. 21 MR. O'BRIEN: I would agree with Matt that a 22 trade-through rule could exacerbate the problem that exists 23 today because there are some firms with the courage that 24 despite the duty of best execution can stand up and make a 25 decision that a high-rate inaccessible market can be avoided 183 1 consistent with the duty of best execution. Not all firms 2 have the courage to make that decision. So the trade-through 3 rule exists, and a market says I'm accessible on an automated 4 base, provided you put a direct line to me in for a thousand 5 or two thousand dollars a month, despite if that is correct 6 that I have a thousand shares of liquidity a day on my 7 system, it's going to create a real problem, but I think to 8 reconcile that with Gerry's point is that to the extent that 9 clearly identifiable problems exist in the current access 10 structure, the way to deal with them is direct regulation of 11 them, not a broad-ranging effort to fix prices across the 12 industry for subscriber and non-subscriber access. 13 So with respect to the outliers that the trade rule 14 might further encourage, I think the idea of requiring them 15 to responsibly participate in the public quote through an 16 order execution facility of an SRO and subject them to the 17 wide usage of that SRO linkage and the discipline of an SRO 18 rule set would address that problem and then with respect to 19 locking and crossing markets, and I think -- I think 20 competing public policy concerns lead me personally to a 21 tipping point that they are on a net basis not good for the 22 industry and the individual investor. 23 But the way to deal with that is not through 24 killing a bee with a bazooka. It's prohibiting the conduct, 25 not prohibiting or trying to prohibit the alleged underlying 184 1 motivation, because there are several motivations for locked 2 and crossed markets, not just fees. 3 All the reasons why someone would argue that they 4 would want to trade through a price of an inaccessible market 5 would justify wanting to lock or cross the quote of an 6 inaccessible market. So you need to understand that fees are 7 not the only motivation for wanting to lock and cross 8 markets. That being said, they may be a problem but the way 9 to deal with it is prohibition of the conduct, not indirect 10 regulation of potential underlying motivation. 11 MR. JOYCE: Locked and crossed markets actually 12 exploded after the introduction of the alternative display 13 facility, which was another little bit of history, which was 14 created because the SuperMontage proposal that NASDAQ had was 15 viewed by many people to disadvantage the current ECNs, that 16 so they wanted a different place to execute in which, you 17 know, whether you argue that point or not, they did get a 18 different place to execute. It was called the alternative 19 site. It started sort of in full form, I guess in early 20 December of '02, and locked and crossed markets went from a 21 nuisance to over 500,000 of them a week, and that was pure 22 fee gamesmanship because you could lock the ADF and wait till 23 I, I have to go to you for my best execution 24 responsibilities, would then route to the ADF, take your 25 offer and then pay the two mils that you were waiting to 185 1 collect. 2 So the fees that have been embedded in this market 3 have in fact created an enormous amount of locked and crossed 4 markets and you can directly trace them back to the 5 introduction of the ADF. So there is germane data that point 6 directly to people using fees and creating locked and crossed 7 markets which, and I completely agree with Bob Britz, locked 8 and crossed markets are trading already and they do not in 9 any -- there is no benefit to the market to have a locked and 10 crossed market out there. 11 MR. COLBY: So I heard several people say that the 12 trade-through rule exacerbates, and Bill said you can deal 13 with it by making sure outliers are part of an SRO that 14 charges, puts limits on the fees. I think there are four 15 SROs that accept quotes from ECNs of which one has a limit on 16 the fees and that's very recent, so it's not a no-brainer 17 that SROs are going to limit the fee sets. But if I go back 18 to this price/time rule, if you have a trade-through rule, 19 another way to deal with it would be to say that quotes that 20 have fees or that have substantial fees can be traded 21 through. Like non-automated quotes, they can be traded 22 through. What do you think about that possibility? 23 MR. PUTNAM: We've considered that and we think 24 it's a great way to address the problem. Each SRO files its 25 rules. And we feel a trade-through rule that says if we deem 186 1 another market to charge an excessive fee, then we'll choose 2 not the trade with it, and we can name that marketplace. It 3 is certainly one way to deal with it. I would say though, 4 this is going on, and then you could look at it and say, you 5 know what, that's not an unreasonable fee. The better way to 6 deal with it, though, is to let the marketplaces, like we've 7 seen happen in NASDAQ, where they have limited what a -- what 8 a participant in their marketplace can -- can charge. 9 This problem has been solved. I mean, we've talked 10 about it for a really long time and while we've talked about 11 it, there's a couple -- and there's not even a couple of 12 little anything left anymore. I don't think there's a single 13 ECN out there any longer that we consider to be a price 14 gouger. They've all come into line because of competition. 15 The small ones have to fall inside of an SRO. They won't be 16 their own, so they can't -- you're not going to let someone 17 create an exchange and say, "Okay, we're an exchange, our fee 18 is three cents a share." We'll just say no. You guys can 19 dictate that. Today, this is a problem that is gone. And 20 again, no one suffered more than we did. I mean, tens of 21 millions of dollars in price gouges. 22 So we should be the first one here that way say, 23 "Hey, guys, that's it, please don't let them do this to us 24 anymore." We're completely against. We've had a couple of 25 people on this panel that have suggested that you should not 187 1 allow us to charge for our service. I don't know how those 2 two hundred people that come to work every day are going to 3 get paid if you decide I can't charge anything unless you 4 decided that the service we're providing to the marketplace 5 is worthless, and you want to ban us from the industry. 6 But if you don't let me charge for my service, then 7 it's going to be very difficult to -- for us to exist. 8 MS. GLASSMAN: Can I just jump in a minute? What 9 difference does it make if the SROs impose the cap or the 10 exchanges impose a cap or we impose a cap? And 11 alternatively, if they impose a cap, why do we need to impose 12 a cap? So it's kind of two sides of the same question. 13 MR. PUTNAM: That's what I would say, if they 14 impose a cap, in fact you do not need to, and the SROs are 15 not going to settle up, I don't believe that an SRO would 16 ever get through this Commission whose business model was to 17 set up a toll booth in the middle of the best execution 18 highway that says, "If you want to pass through here, you've 19 got to give me three bucks." I mean, it's just not going to 20 happen. And we actually had that environment since 1997 for 21 a period of time, and it competed itself out of the 22 marketplace. 23 And when we did fight, and the Commission did step 24 in at times when someone told us, "We're going to pull the 25 wire out of the wall as a means of forcing you to pay this 188 1 exorbitant fee," the Commission gave us help. So the 2 Commission gives guidance and gives help to the marketplace 3 when it sees something unfair. 4 That said, leave it to the SROs to determine what 5 they want to charge. If someone, you know, it's a real 6 business, and has a real market share, and throws a massive 7 fee out there, they are going to lose all their customers. 8 That will take care of it. The problem is when you have 9 these little ones, when I will tell you the problem with the 10 small market centers is nonexistent today. 11 MR. O'BRIEN: I think the reason is that this goes 12 back to the prior panel on linkage. That not all market 13 centers necessarily need to provide access to an SRO order 14 execution facility. So to the extent that they require minor 15 market centers to be accessible via an SRO but allow major 16 market centers is to utilize private direct linkages, SRO 17 regulation kind of provides more flexibility to the industry 18 over all than, say, an SRO-by-SRO rate structure. 19 In terms of also why the SRO is better to act, it's 20 a general belief from a public policy perspective on the 21 value of a self-regulatory system, that SROs are closer to 22 their membership than the Commission as a whole, on an 23 ongoing, sustainable basis, and to provide that type of 24 service to their membership of access to minor market 25 centers, they are best equipment to -- to cap rate structures 189 1 for access to minor markets. 2 MR. COLBY: May I just add to this debate two 3 points. One is that fortunately in Gerry Putnam's situation 4 where there was this gouged rate, the entity that was 5 involved in charging the rate was at a level that they were 6 subject to fair access requirements, and so there was a fair 7 access consideration. I think we probably should point out, 8 that at least two of the major ECNs are quoting to SROs where 9 they constitute the vast, vast predominance of the SRO's 10 business in that market. So to expect that SRO to then curb 11 the fees of the entity that's providing them almost all it's 12 business is perhaps not as realistic -- 13 MR. O'BRIEN: But those SROs are sufficiently large 14 enough that competitive market forces are going to help -- 15 MR. COLBY: Maybe you can explain that because it's 16 not clear to me why the market is not disciplining the rate 17 structure generally, why they would -- 18 MR. O'BRIEN: I will explain it, because Ed is the 19 CEO of a public company, Gerry is about to become CEO of a 20 public company. They are in a hypercompetitive market. They 21 cannot, consistent with the duties to their current or 22 ultimate investors in their company, try to impose 23 significant rate hikes in that type of environment without 24 jeopardizing their business franchise. That's not true for 25 Bill O'Brien starting his own ECN that has no duty and just 190 1 is able to acquire order flow from one provider, and has a 2 right to voluntarily provide that to the public quotation 3 system and then dictate access to that system, A, privately 4 and B, whatever other rate structure I feel like charging. 5 MR. LAVICKA: I just want to agree with Bob. The 6 SROs have had plenty of time to limit fees and the only 7 reason the one that did it did it was because of Reg NMS 8 coming out with the input that it was going to cap the fees. 9 That's if only reason you got any action. 10 MR. O'BRIEN: I would disagree. I think the 11 existence of alternatives, whether by ADF or by regional 12 exchanges entering the market from a competitive perspective 13 to provide quote services, gave NASDAQ what they felt was the 14 flexibility to cap fees whereas a historical monopoly 15 provider quote display service felt like they couldn't. 16 MR. NICOLL: Can I make one other point here, I 17 think we all have to, again, focus on -- I think it's curious 18 that we call these things access fees. These are agency fees 19 being charged by an agency marketplace. And -- 20 MS. GLASSMAN: Could I ask it, who are you the 21 agent for, isn't that the question? 22 MR. NICOLL: We create a set of services whereby we 23 allow -- we create a network in which broker/dealers come in 24 to our market in accordance with Reg ATS and they display 25 their bids and offers and other people access those bids and 191 1 offers. If I can't charge a fee for that service, I can't be 2 in business. 3 The New York Stock Exchange has a curious set of 4 fees in which you can access the fees initially because 5 the -- the brokers on the floor, if you come in in the first 6 five minutes you access to quote, it's free, but if you put 7 up your orders on the floor of the New York Stock Exchange, 8 and your orders sit down there for over five minutes, then 9 you get charged. You get charged a lot. In fact, you're 10 not -- to me, the distinction between the agency fees charged 11 by ECNs and the agency fees charged by in cooperative effort 12 between the specialists in the New York Stock Exchange are a 13 distinction without a difference. 14 And yet we, you know, we are a huge router of order 15 flow down to the floor of the New York Stock Exchange. We 16 get charged down there, we get charged on average .8 cents. 17 Some specialists charge us more than a penny a share to leave 18 orders down in their books. They need to do that or they 19 need to trade on a principal basis to make a living. It's 20 the same kind of fee structure. And we are in the business 21 of being an agency market. We don't take positions, we don't 22 trade on our own account. We never act as a dealer and our 23 customers value that position in the marketplace and we must 24 charge some kind of agency fee above and beyond the quotes 25 that you see in order to exist. 192 1 MR. JOYCE: You know, I think it's interesting that 2 a couple of individuals today have talked about, well, if we 3 can't charge the fee, we can't be in business. Of course, 4 the fees are mandated courtesy of the SEC. But I thought our 5 business model was supposed to add value to the market. 6 And if my business model is, I pay for an order, I 7 pay two mils for an order, and then I charge somebody else 8 three mils for an order, and that's my reason to exist, 9 that's an interesting business model. I will say for sure 10 that before the NASDAQ market introduced, and actually, 11 Instinet effectively introduced the central limit order 12 board, it was a pure example of commission 101. 13 There was a need in the NASDAQ market for a central 14 limit order book and Instinet very cleverly filled it. But 15 now, that need doesn't nearly exist to the extent it did 16 before if it exists at all, and if my entire business model 17 is to pay you something and charge you, who has to come here, 18 you don't get a vote, you have to come here, more, well, 19 that's just an interesting business model. And I personally 20 believe if, without that, that access fee that is imposed 21 upon people who don't subscribe to your system, if that goes 22 away, the business model goes away with it, then really what 23 kind of business model was it in the first place? 24 MR. NICOLL: Let me respond if I could. I'd be 25 happy -- it's interesting -- it's interesting that you're 193 1 our -- questioning our business model given the fact that 2 your business model is simply is to internalize a lot of 3 other people's quotes and pay people for order flow to come 4 in to your -- in order to -- in order to free-ride off of 5 publicly created quotes. 6 So I mean, we can sit here and we can spit over 7 each other's business models. But we are not internalizers, 8 we are not dealers, we are pure agents. We are not, you 9 know, we think actually rebates make sense because they came 10 about as a result of market competition. 11 We'd love to be able to go back to a time when 12 Instinet was able to charge for both posting and taking 13 liquidity, but it was market forces that forced Instinet to 14 pay rebates, and just -- and I -- and let's not forget that 15 the trade-off that Instinet made at the time when it went 16 from being a voluntary central limit order book for NASDAQ 17 dealers to one where it incorporated into the quote, into the 18 Order Handling Rules, that it no longer had a choice of who 19 could and couldn't access its quote. 20 Instinet then could cherry pick who it wanted to do 21 business with. After it -- after the Order Handling Rules, 22 it no longer could. It was a fair access ECN. And I 23 think -- and I think that a perfect example of what the 24 Chairman asked when we started out, it was not in Instinet's 25 interest to do that. Instinet has seen a price decline of 194 1 over 90 percent in terms of its revenue capture for a hundred 2 shares since that happened. 3 But the Commission was right, the SEC was right. 4 Fair access is absolutely fundamental to the market structure 5 system and if I have to put on my public policy hat, I agree 6 that that was a fair tradeoff for the Commission to demand of 7 ECNs. 8 But I think it's quite unfair, then, to turn around 9 and say that ECNs are not -- should not be able to charge an 10 agency fee for their services. 11 MR. JOYCE: Just to clarify. We don't internalize 12 if we don't have order flow. People choose a route that's 13 their -- I understand internalization well. We did it at 14 Merrill Lynch, it may or may not be an issue, but just to 15 clarify, we do not internalize because we don't have any 16 capital. People choose to send us their order flow for 17 execution. 18 MR. ATKINS: As a follow-up to that, just about 19 value added of the ECNs, I'm assuming that because you do 20 send some of your orders there, that trades there, that there 21 is value added. But I was -- I was wondering if you sort of 22 implied that perhaps you were only doing it because of your 23 best execution obligation. 24 MR. LAVICKA: Absolutely, and I think you have to 25 distinguish between when we see a quote from Archipelago or 195 1 Instinet, and that's the best bid or offer in the country, 2 and as a market center we want to access that quote, so we 3 can -- we can fill fulfill our best execution obligation, I 4 think that's different than when we want to take advantage of 5 the facilities and the service that an Instinet or 6 Archipelago provide, namely, anonymity, smart order handling, 7 there's a lot of unique order types or discretionary order 8 types that they have, and these are all value added services 9 that they can charge fees to, and I think if the rule doesn't 10 have it, it should make a distinction between those two types 11 of fees. 12 MR. JOYCE: Agree. We do subscribe to several of 13 them because we do think in many cases they are value added. 14 It's the issue around where we are forced to go to some of 15 them because the best execution requirements that the 16 subscriber or non-subscriber, commission vs. access fee 17 assessed gets blurred. Subscriber route is clear. 18 MR. O'BRIEN: In fairness to the broker, I think 19 that's a conundrum that many brokers face, that -- Bob had 20 asked the point, if there's a trade-through exception, would 21 that alleviate the problem, and really, there's no 22 trade-through rule in NASDAQ today and brokers are still 23 struggling with the problem, making clear that to the extent 24 a market center is charging nonmarket or excessive 25 transaction costs, that that may be a order routing 196 1 determination as part of their best execution decision. I 2 think it would give brokers a lot more comfort in making the 3 decision about what market centers add value and which ones 4 do not. 5 MR. COLBY: One of the proposals in the release is 6 to reduce the fair access standard which currently apply to 7 ATS above twenty percent market share level to five percent, 8 but there's a question in the release about what are fair 9 access requirements -- and this is a question of who can 10 become a subscriber and on what terms, the gouge rate, the 11 other issue, whether this should apply to any ATS that has a 12 quote published in the NBBO. 13 MR. PUTNAM: Easily apply it to any model 14 whatever -- you want to be in the marketplace, you have to 15 allow access to your market. 16 MR. O'BRIEN: Well, I mean, yes and no. Brut is in 17 that five to twenty percent category today and I think we de 18 facto comply with fair access requirements to our publicly 19 displayed quotations. The problem is even if you apply that 20 to all ATSs, they can still decide, I'm going to charge nine 21 tenths of a cent for access for -- 22 MR. PUTNAM: That's not the question, though. The 23 question is whether they can deny access or they can provide 24 access on special terms to competitors and what I think Bob 25 is suggesting or what I'm suggesting is, no, you can't do 197 1 that. You can't single someone out and say, "We don't like 2 you, so you can't trade with us," or, you can't single 3 someone out and say, "We don't like you but if you want to 4 trade here, you can pay extortion rates for trading here," so 5 that would not be allowed. 6 MR. O'BRIEN: I would agree with that. I'm just 7 saying it wouldn't solve a problem that Matt and Tom struggle 8 with that for an ECN that is universally charging a 9 non-market rate to all of its customers even on -- albeit on 10 a nondiscriminatory basis and struggling with how to, you 11 know, link to that center for the best execution respect. 12 MS. NAZARETH: But it does sound like you think it 13 would help in situations where you think there have been 14 discriminatory rates charged to competitors in the way -- 15 MR. PUTNAM: But it's not saying we're going to 16 force you to connect to a marketplace that charges an unfair 17 rate. What it's saying is, if you choose to connect to a 18 marketplace, you have the freedom to connect to it. I mean, 19 I can think of an outlier on the ADF where we might not want 20 to be a subscriber of that system. That said, we should be 21 allowed to be a subscriber of that system but if we don't 22 want to be a subscriber to that system and there's no public 23 linkage to it, there's an outlier and it solves itself 24 because they find out and people don't trade with them, 25 because they charge a gouge rate. 198 1 And this quote's no longer accessible through the 2 NASDAQ linkage, so you can't use the SuperSOES get in the way 3 of the method of charging this fee. So guess where are they 4 now. They're out on the ADF and they charge the same like 5 everybody else does. It solved itself without any help. 6 MR. LAVICKA: It has solved itself, I think once 7 they cross that line of putting a quote in the NBBO, we need 8 to protect the integrity of if NBBO, and what that means. I 9 don't think you just void it. I don't think that's enough. 10 I think this you have to regulate and make sure there's fair 11 access to that quote. 12 MS. NAZARETH: I think we're done. 13 MR. ATKINS: I guess I can ask another one. I have 14 a lot on the subject I could ask. But I mean, Bob, you were 15 talking about, you know, with respect to some of the other, 16 comparisons with transaction fees or regulatory fees and 17 things like that. And I don't know if other people have 18 this, but certainly, I'm not one to defend government trying 19 to come in and regulate, but with respect to compare the 20 access fee issue here, to, say, the market data area, which, 21 you know, as things change in the next few years, I think 22 things will get more comparable as far as some of these 23 issues go, but we don't -- we basically provide, they will 24 tell, just rubber-stamp the fees that are charged by various 25 marketplaces. So I was wondering what implications are from 199 1 that to this discussion here. 2 MR. BRITZ: My experience in the market data space 3 is very different than the SEC rubber stamping CTA and NYSE 4 market data filings. But I do take your point. Listen, I 5 think it's a slippery slope when the government gets involved 6 in rate making. We're going to have a panel here on market 7 data in the not too distant future. 8 I think it's a slippery slope when the government 9 designs formulas for revenue allocations. The way this has 10 been expressed to me in terms of why access fees have somehow 11 migrated into the listed space and sort of grabbed ahold of 12 the transaction charges is that in one way or another, they 13 are all involved in reaching a quote in a particular market. 14 But to your point, some are market data fees, some 15 are membership fees, some are real estate fees on our trading 16 floor and lots of fees that others charge as well. So why 17 the Commission would zero in on transaction charges, I agree 18 completely with what Matt said and with what T.J. said. If 19 ECNs add value to their customers, they ought to be able to 20 charge subscriber fees, which I, roughly, I guess, are akin 21 to the NYSE transaction charges. 22 I would not argue one way or another that those 23 fees ought to be capped and on balance, I would argue that 24 they ought not to be capped. To the extent that Gerry's firm 25 provides valuable order types, he ought to get what the 200 1 market thinks the value of that is. 2 Access fees are in a very different bucket than 3 those kinds of fees and SRO transaction fees. And I think 4 your point, once you go down this road, I don't know how you 5 distinguish the various other fees that are used to fund 6 marketplaces or ECNs from transaction charges or access fees. 7 MR. GOLDSCHMID: I have a question. We'll think 8 hard about what we've heard from this panel. But assume we 9 stay on course with what you propose. What would the world 10 look like in three years? 11 MR. NICOLL: Just in this particular area? 12 MR. GOLDSCHMID: Just in this area. 13 MR. NICOLL: I didn't -- you know, I think one of 14 the -- I mean, I think one of the unfortunate, or one of the 15 possible unfortunate side effects of this proposal might be 16 that it would discourage new entrants, and so I think it 17 might actually benefit the larger players in the space, 18 because there's not, certainly going to need enormous scale 19 in order to make a buck and you're not going to have any new 20 entrants coming in, and that may, you know, and there would 21 be less pressure from innovation and less competition. 22 So I think what's likely to happen is consolidation 23 in the space and no new entrants in the space. I -- I'm sure 24 there will be other things. I'm just not smart enough to 25 figure it out. 201 1 MR. BRITZ: I think the staff at the SEC that 2 handle investor complaints will grow exponentially. People 3 will be getting bills from counterparties they do not 4 recognize after the trade has taken place. 5 MR. DONALDSON: Why don't we draw this panel to a 6 close and reconvene in ten minutes. 7 (Recess taken.) 8 MR. DONALDSON: I won't make you introduce 9 yourselves since everybody knows the back of your heads by 10 now, except for you, Dan. Why don't we get Annette, you want 11 to start us off? 12 MS. NAZARETH: Sure. I think the substance of the 13 subpenny quoting proposal is pretty clear. The proposal 14 would prohibit market participants from accepting, ranking or 15 displaying orders, quotes or indications of interest in 16 subpennies. The proposal would not, however, prohibit a 17 trade from being printed in a subpenny increment, so that 18 firms could continue to offer average price trades and 19 subpenny price improvement. So I think we should launch 20 right into a discussion of whether the people think this 21 proposal has merit, are competitive forces alone sufficient 22 to address the problems discussed in Reg NMS with respect to 23 subpenny quoting or do you support the subpenny quoting 24 proposal? Bob? 25 MR. GREIFELD: Thank you, Annette. With permission 202 1 of Commission I'd just like to back up one second to a panel 2 that I was on earlier today in response to a point, and I say 3 that in light of the fact that I have very few words to say 4 about subpenny, so I'll be under my allotment. There was a 5 quote made with respect to volatility in NASDAQ stocks and a 6 study was referenced, and what I want to say is, the NASDAQ 7 market has a trade reporting responsibility of 90 seconds. 8 So we do not receive trades in sequence. And I think Mr. 9 Peterffy referenced within a given time frame, it can -- he 10 talked about seconds versus volatility. So as we do our 11 studies of volatility in The NASDAQ Stock Market, we realize 12 that trade reporting is not the way to look at volatility 13 studies in that the trades are routinely out of consequence. 14 We think the best way to look at volatility across all 15 markets, especially a market with a trade-through rule, is 16 with quoted spreads. The quoted spreads reflects where 17 stocks can trade at each instant of time. And when you take 18 that study of the 40 some-odd stocks, and you quote spreads, 19 you actually find that the volatility stayed roughly the 20 same, and the largest stocks that were referenced in that 21 study of volatility actually increased on a marginal basis, 22 and on the smaller stocks they decreased. 23 And the basic theory that we supported at NASDAQ 24 and supported by a lot of economists is that volatility is 25 almost exclusively company-specific, it's not 203 1 market-specific. And we'd like to ask permission from the 2 Commission to commit to the record a more detailed study on 3 volatility on the NASDAQ market. Thank you. 4 With respect to subpenny quoting, it just -- a 5 thought just hit my mind and I think it's a quote from Yogi 6 Berra, "I feel it's deja vu all over again." We leave that 7 limited order display and before limit order display we had a 8 bifurcated market where retail investors were treated 9 differently than professional investors. And we solved that 10 problem. But the problem reared its ugly head through the 11 form of subpenny quoting. Retail investors in this country 12 today cannot see or trade with subpennies. Professional 13 investors can. This has to stop. It has to change. It is 14 not right for these investors. 15 I don't think there's anything else you can say 16 about it. We also believe this proposal is fairly 17 straightforward and should be put on a different review track 18 than the other proposals of Reg NMS. I think it's incumbent 19 upon all the people at this panel and in this room to make 20 sure we get this approved sooner rather than later. Thank 21 you. 22 MR. JOYCE: I believe that -- I also believe, and 23 we at Knight firmly believe that there is no benefit to the 24 market to have subpenny trading. It is my humble opinion a 25 race to the bottom. Just because we can trade in subpennies 204 1 doesn't meet we trade. We can trade in sub-subpennies. We 2 can do a lot of things. It's certainly hard to find a retail 3 product in any part of the economy short of gasoline, and I 4 think that line nine in gasoline prices never actually moved 5 very much. It's hard to find anything that you trades in 6 subpennies. 7 I completely endorse what Mr. Greifeld said about 8 the fact that the retail investor has been excluded from the 9 user group of people who are very adept and have audit 10 computing power. I don't think there's one thing wrong with 11 the U.S. market having certain standards of excellence and I 12 think a penny minimum trade quote is one standard that we 13 should adhere to so I firmly endorse the abolition of 14 subpenny trading. 15 MR. NICOLL: I think just think markets should be 16 allowed to go to equilibrium. I think once again, we 17 recently moved from subpennies to pennies because our 18 competitors who were trading in pennies were actually taking 19 market share away from us. Our customers really prefer that 20 we trade stocks in pennies. So we went from trading in 21 pennies to all stocks down to a dollar, with the exception, 22 as I mentioned before, of the QQQs and the ETFs. There's an 23 active, a very active market in the ETFs in subpennies. Much 24 of that is arbiting against the derivative marketplaces. And 25 as I mentioned before, the average spread in, for instance in 205 1 QQQs on Inet is 4.3 cents. It trades on average a hundred 2 million shares a day. If, you know, if that were the spread 3 to the entire marketplace and we were to enforce that to a 4 penny, we would be imposing additional transaction costs of 5 close to two hundred million dollars a year on people buying 6 and selling QQQs and I just don't see why that is. 7 I understand that there is a cost/benefit trade-off 8 here, that, you know, we have to think about the cost of 9 vendors, in terms of delivering all the various price points, 10 and I think that's a reasonable concern. I do have some 11 experience in serving retail investors and the last company 12 that I had a significant role in creating delivered subpenny 13 pricing to its investors over the web. And there's no reason 14 why others can't. 15 But you know, this seems to be, again, almost all 16 the ECNs that traded in subpennies have now gone to pennies 17 because the vast majority of clients want to trade in 18 pennies. One exception I think is Brut, which trades in 19 subpennies under five dollars. We only trade in under a 20 dollar. And, you know, this is just, again, to me, I just 21 don't think that this is a -- a problem that needs to be 22 solved by regulation, and I think the markets have actually 23 already solved the problem for the most part. 24 PROF. WEAVER: I testified before Congress in 25 support of decimalizing stock prices. It was contrary. At 206 1 that time, markets were allowed to choose the pick signs they 2 wanted. They chose a penny. That was a mistake. Liquidity 3 dried up. Limit orders are shock absorbers for liquidity 4 events. We need depth at prices. I believe that now is not 5 the time to allow markets to set the tick. I think now is 6 the time for regulatory intervention. I support the 7 Commission's proposal but I don't think it goes far enough. 8 I think not only should subpenny quoting be banned, I think 9 subpenny trading should be banned. In fact, I urged the 10 Commission to consider going to a nickel tick for some higher 11 priced stocks. 12 MR. COLBY: Is this where we're supposed to cheer? 13 MR. GIESEA: I just have a few comments. Since 14 2001, the Security Traders Association has opposed subpenny 15 quoting. And I would like to make sure we distinguish 16 between quoting and trading and my comments are specific to 17 quoting. Because there are some valid uses for subpenny 18 transactions to be reported via average pricing across the 19 network for fee block kinds of transactions. 20 With regard to quoting and with regard to the 21 question are competitive forces enough, I think competitive 22 forces as it relates to the field of athletics brings to mind 23 that in each sport, there are rules, and in this case, rules 24 are better -- the proposal here by the Commission is better 25 than allowing competitive forces to allow to come to its own 207 1 decision. 2 So we clearly and unanimously support the one penny 3 increment. 4 MR. DONALDSON: I don't assume you have too much 5 support for subpenny pricing. 6 MR. JOYCE: The only thing I'd add, and it may not 7 surprise anybody here that Ed and i slightly disagree on 8 this, too, is that I think if you go from subpenny pricing of 9 QQQs, I assume the entire market is not a carve up for some 10 of the ETFs. I think -- I personally believe that it's 11 overstated that we did a transaction opposite to the market, 12 because largely, as pointed out earlier, the trading that 13 takes place in many, many cases, in really liquid ETFs like 14 the QQQs, are largely agency trades. It's not mom and pop 15 buying from a dealer. It's mom buying from pop. 16 So if a lot of transactions are taking place at 1.3 17 cents, and I am a buyer, and it's a penny market, and it's 18 bid one offered at two, instead of paying two, I pay 1.3 19 cents. Well, I did a little better. But the seller did a 20 little worse. 21 So in my opinion, in an agency situation like that, 22 it's simply a zero sum game where mom might do better 23 sometime but pop doesn't do as well, i.e., mom buys it better 24 but pop sells it worse. So I think in an agency market, I 25 think you run the risk of overstating the -- the added cost 208 1 of widening the spread in an instrument like the QQQs that I 2 just don't think it's there. Again, I think it's a zero sum 3 game in a transactional -- added transactional costs are 4 probably overstated. 5 MR. DONALDSON: Professor Weaver, since you brought 6 it up, have you got any empirical work done or that you know 7 of that leads you to conclude that we put to put it at a 8 nickel rather than a penny? 9 PROF. WEAVER: Well, spreads have gone down quite a 10 bit. And so has debt. And I think that the short answer is, 11 I don't think there's anything out there that shows that we 12 should go to a nickel. It's my personal opinion that by 13 going to a nickel, we would bring more depth back to the 14 marketplace, and we seem to be focused way too much on 15 narrowing the spreads. But the other part of that is depth 16 in the market. Another very important measure of market 17 quality. 18 MR. ATKINS: Can I ask a question, I mean, as far 19 as depth goes, isn't the depth still there, it's just spread 20 out over more price points and does that have to do, like the 21 stock exchanges' liquidity quote product that they've been 22 talking about, I mean, doesn't that sort of get back at what 23 you're talking about with respect to -- 24 MR. WEAVER: I think that a lot of limit order 25 traders have switched to being market order traders and just 209 1 monitoring the market more rather than putting limit orders 2 in there. Limit orders are smaller sizes now. And I've seen 3 some studies, it's not conclusive whether or not they -- if 4 add up all the depth between certain price levels, you'll get 5 the same depth you had before. 6 So, no, I don't think that's the case. I think 7 that a lot of limit traders have switched to being market 8 order traders. And I'm worried about that. Liquidity 9 belongs in the place where shocks are going to occur, not 10 away from the market. And I believe in rules that will 11 encourage as much liquidity at the source of shocks as 12 possible. 13 MR. ATKINS: So basically what you're saying is, 14 you sort of talked about that before with one of the other 15 panels about adding another context on the floor of an 16 exchange, where there has to be a return of capital, and a 17 lot of this for decimalization was built into the spreads, 18 the vigorish there in what market participants were able to 19 wring out of the system. So basically what you're arguing is 20 we went too far, we need to go back and restore some of that 21 non-transparency -- 22 MR. WEAVER: Yes. Yes. Now, transparency in all 23 areas of trading is not a good idea. And there are some 24 areas, for example, it was a mistake opening up the limit 25 order book to the public. I think was a mistake. 210 1 MR. COLBY: Can I bring up the question of the QQQs 2 or other ETFs, I heard Tom address it, but are there some 3 instruments that should be traded possibly would benefit to 4 trade in subpennies, other than ones that are very low 5 priced, because of their derivative price nature and the 6 like? 7 MR. NICOLL: Let me just, it's interesting because 8 we've had -- we're interested in cascades, because we went 9 from subpennies to pennies over the last six months. I'm not 10 quite sure -- I should know this, I don't know exactly when 11 it happened, but I should think it was about three months 12 ago. And we gained overall, we gained market share as a 13 result of doing it. We did lose significant market share in 14 one stock under five dollars, which went to -- which 15 essentially went to -- is continuing to trade in subpennies 16 and for the life of us we can't figure out why. It's a 17 particularly volatile stock but there are a lot of these four 18 or five dollar stocks that are volatile, and it's not clear 19 to us why the market seems to prefer to trade that stock in 20 subpennies. 21 We're talking about Brut having like a sixty 22 percent market share in all of the trading in that particular 23 stock. So we're a little bit scratching our heads about it, 24 and we -- this is what we do all the time, and this is what 25 we -- we think we ought to know why these kinds of effects 211 1 are occurring, and quite honestly we don't. 2 I do think in ETFs that there is a tremendous 3 amount -- there is a good reason to have subpennies because 4 of the arb between the futures. I think if we go to one, the 5 spreads will lose activity to other markets, and there's a 6 good reason to have as finally priced marketplace in these 7 ETFs as we can. Other than that, I can't make an argument 8 for what the spread should be in a particular equity, where 9 there is no other derivative price involved. 10 MR. COLBY: If you give us the name of the stock 11 we'll have our crack market surveillance team look at it. 12 MR. ATKINS: I guess that's what I wanted to go 13 back to them to help goad Ed along here on this question. 14 But -- and get back to what Bob was saying with 15 respect to subpennies. I mean, maybe it's fine and well to 16 say that we shouldn't have subpennies on anything except ETFs 17 now, today, in, you know, April 2004. But just like we 18 couldn't predict what went on with respect to SPDRs and QQQs 19 and all that ten years ago, looking from here on out, how can 20 we predict what's going to be going on at some other product 21 and if we limit things artificially to a penny now, how will 22 we -- how will that market ever be able to figure out what 23 was needed later on? 24 MR. GREIFELD: Well, certainly from our 25 perspective, the Commission has a continuing right to change 212 1 the market when they see the need to do so. So to the extent 2 that we ban subpenny trading today, and some time in the 3 distant or not to distant future, that is looked upon as a 4 correction that is no longer proper, then we address, you 5 know, tomorrow's problems tomorrow. I they we have to deal 6 with what exists today. 7 MR. ATKINS: It's taken ten years to get NMS -- 8 MR. GREIFELD: To the extent we're going with 9 subpennies at some time in the future, that the proper way to 10 hold it out is to give the retail distribution mechanism time 11 to implement the support of subpennies. So we went from 12 fractions to decimals and decimals to subpennies very quickly 13 and nowhere in the process was there a comment period and a 14 period of adjustment for the retail distribution network in 15 this country. 16 So today, retail investors are not able to trade in 17 subpennies and professional investors are. And every day 18 that goes by, that's not a proper situation for any of us in 19 this room. 20 MR. NICOLL: Since we have a little bit of time, 21 Chairman, if we can just make one other comment, because you 22 asked me about speed, and I just want to point out something 23 that might be a little counterintuitive. Which is that the 24 bigger you make, you know, the increment and the more 25 chunkier you are, the more you're rewarding the first party 213 1 to get at that wider increment. And the more -- and the more 2 benefit accrues to very, very fast, sophisticated traders who 3 specialize in getting to the market before anybody else. If 4 I'm a super -- and lots of these guys out there, electronic 5 market maker now, and I have -- and I'm running models, I 6 make more money like Tom makes money as a market maker, and 7 he should because he provides capital to the marketplace -- I 8 make more money with spreads if I can get to the head of the 9 line at that penny increment or, if we were to go to five 10 cents, there would be that much more important to get there 11 really, really fast. 12 So time priority as we widen the spreads become 13 even more important, and remember, even with trade-through, 14 we are -- nobody here is proposing intermarket time priority. 15 So the wider the spread, the more you have people line up at 16 each individual increment, the more important time priority 17 becomes and the more that we reward really sophisticated 18 investors who can get to the front of the line quicker. 19 The absolute, you know, best example of that are 20 the guys down on the floor who can see, who can get there 21 faster than anybody else on a floor based market. On a 22 market like mine, that would be a very, very sophisticated 23 investor who is investing an enormous amount in throughput 24 down to the matching entry. And it just is a factor. So I 25 mean when we talk about this dichotomy between very 214 1 sophisticated traders and non-sophisticated traders just 2 realize that when we widen the increment, we are getting a 3 benefit to people who can get to the head of the line first. 4 MR. JOYCE: What I'd like to add is that it is 5 happening now in the subpenny world and again it's 6 sophisticated traders because they will look at that quote 7 and they will sniff the quote. You of course can't see a 8 subpenny quote. It's a computer quote that is sort of a 9 private closed user group. And they'll sniff the quote, then 10 see it's not 1.2, it's 1 and 1 to 1.9 so then they 11 immediately offer 1.8. The only people doing that are people 12 with a lot of computer power and professional traders. It's 13 definite my folks who have quickness -- 14 MR. NICOLL: I don't disagree with that. I don't 15 disagree with that. I'm just saying that there are 16 countervailing issues the other way also, is all that I'm 17 saying. It's not one or the other. 18 MR. ATKINS: Well, are there any stocks now that 19 are really being traded in subpennies, because I thought Brut 20 has changed over -- 21 MR. NICOLL: Well, anything under a dollar anywhere 22 is still trading under subpennies, and as I said, Brut is 23 still trading five dollars stocks and under. So -- not on 24 any of the major ECNs other than Brut is any subpenny trading 25 occurring over a dollar. 215 1 MR. ATKINS: So that's -- 2 MR. JOYCE: And a lot of -- 3 MR. ATKINS: So your points at the market -- okay. 4 MR. DONALDSON: If there are no objections from 5 either side of the table, we'll steal a few minutes from the 6 subpenny panel and donate them to the next panel. 7 MR. NICOLL: I'm sorry -- 8 MR. DONALDSON: Unless there are other comments 9 that people want to make. Thank you all very much. And why 10 don't we get our last panel. 11 (Recess taken.) 12 MR. DONALDSON: Starting on my right, will you just 13 raise your hand and just tell the people in back of the room 14 who you are? 15 MR. LYNCH: I'm Phil Lynch, I'm the CEO of Reuters 16 America. 17 MR. McBLAIN: I'm Leo McBlain, chairman of the 18 Financial Information Forum. 19 MR. HALVORSON: Kurt Halvorson, chief 20 administrative officer for Ameritrade Holding Corporation. 21 MR. BROWN: Jeff Brown, general counsel of Schwab 22 Capital Markets. 23 MR. COLKER: David Colker, CEO and President of the 24 National Stock Exchange. 25 MR. GREIFELD: Bob Greifeld of The NASDAQ Stock Market. 216 1 MR. BRITZ: Bob Britz. I'm President of the Stock 2 Exchange. 3 MR. DONALDSON: Okay. Annette -- 4 MS. NAZARETH: Sure, I think I'll give this one in 5 particular an introduction because it's one of the more 6 complicated proposals. The market data proposal sets forth 7 four subproposals that would be, one, update the formulas for 8 allocating more than four hundred million dollars in market 9 data revenues to the SROs; two, broaden participation in the 10 governance of the joint SRO plans to help assure that 11 interested parties other than SROs have an opportunity to be 12 heard. 13 Three, authorize individual market centers to 14 disseminate their own data independently in addition to the 15 consolidated data that is disseminated through the joint SRO 16 plans; and four, rewrite current Rule 11Ac1-2, which requires 17 the consolidated display of data to streamline its 18 requirements and to ease its burden of compliance. 19 The proposed market data formulas are very complex 20 and I'm sure we'll give that to everyone here, but in 21 essence, they would divide market data revenues equally 22 between trading and quoting activity. 23 With respect to quoting, markets would be 24 compensated both for quoting at the NBBO and improving the 25 NBBO. Modifying the net income allocation formulas would not 217 1 alter the total fees charged by the three existing data 2 Networks but would address the most pressing and serious 3 problem with current market data rules, namely the economic 4 and regulatory distortions caused by the other formulas. The 5 other three subproposals set forth in the market data 6 proposals was recommendations of the Commission's Advisory 7 Committee on Market Information including permitting market 8 centers to distribute additional information such as depth of 9 book, independently from the jointly administered plans, 10 broadening governance of the plans to a non-voting advisory 11 committee and finally, generally easing the burden of 12 compliance with market data rules. 13 MR. COLBY: In addition to the exact proposal, the 14 release tried to give fair treatment to a number of other 15 proposals and ideas that have been routed by the Commission's 16 Advisory Committee on market data and other areas and we set 17 them out with criticisms of them, but I know that members of 18 the -- have views, so I'd like to just mention them and get 19 them on the table here. One was the deconsolidation model 20 which would say that brokers and vendors no longer have to 21 provide consolidated data, they can choose whose data to make 22 available. And so there wouldn't obviously need to be any 23 consolidation of -- the overall data could be disseminated 24 separately. 25 The second model was one that was called the 218 1 multiple consolidators model which would be, each market 2 would put out its data, then there would still need to be 3 consolidation by the brokers and the vendors, but the vendors 4 could pull it together and make it available. And the third 5 was a hybrid model which would say, we could deregulate 6 trades that could be put out on an individual market basis, 7 deregulate the full quote stream. Quote stream could be put 8 out by each individual market, but there would still be a 9 consolidated national best bid and offer which would be sold 10 at a much substantially reduced rate. It would be considered 11 public good. 12 That's what would be provided to the customers as 13 their baseline trading unit and it would be sold at a very, 14 very substantially reduced rate, perhaps at cost-plus. 15 And so there would be a similar variety of 16 consolidation models. So I wanted to get those out so in 17 case you wanted to stray beyond the pure formula that was set 18 out. 19 MS. NAZARETH: Bob? 20 MR. BRITZ: I think I'd start by agreeing that the 21 SEC is right to look to address some of the, my term, 22 objectionable practices that have crept into the marketplace 23 in the last five or ten years, whether they be payment for 24 order flow, or wash sales or shredded tape prints to the 25 highest bidder and a variety of others. They are at best 219 1 inconsistent with investors' need for transparency and 2 credibility in the market data space in a well functioning 3 national market. And secondly, they have the potential to 4 compromise the broker's fiduciary obligation to the client. 5 We believe, rather than deal with these issues 6 indirectly through market data revenue allocations, the SEC 7 should either enforce existing rules that deal with these 8 practices, or to the extent you think you need to promulgate 9 new rules to ban them, we think you ought to do that. 10 What's striking to me in this whole discussion is 11 that there is no natural link between the practices that you 12 cite in your questions and market data revenues. The actions 13 of a couple of markets and ECNs have caused them to be linked 14 by those markets' paying for orders on the one hand, or, as I 15 believe even worse yet, finished tape prints. 16 On the other hand, and doing so, even though money 17 mingles, doing so extensively out of market data revenues. 18 Through Reg NMS, the Commission proposes to deal with these 19 issues, but not by banning them, but by modifying the market 20 data revenue allocation formula. 21 The problem is that the practices you cite don't 22 spring from the market data formula or any particular 23 formula, but they drop out of the joint action on the part of 24 market participants in either CTA in the listed space or the 25 OTC/UTP plan in the over-the-counter market. So I know the 220 1 question that you put before the house is, do you need 2 additional regulatory tools to deal with these issues, and if 3 you'll allow me a little poetic license, I would suggest you 4 need a deregulatory tool by abolishing the consortia that 5 give rise to the cross-subsidies that in turn give rise to 6 what you characterize as the distortive incentives. 7 They don't drop out of any particular formula, and 8 I want to be clear, New York is neither wed to the existing 9 allocation formula nor is it at this point, we're still 10 trying to figure it out, for or against the proposed formula. 11 We simply think you're missing the markets if on the one hand 12 you're trying to get at the source of money that gives rise 13 to these objectionable practices. It's not any particular 14 formula. It's the joint action in general and particularly 15 the joint pricing action on the part of what are otherwise 16 natural competitors. 17 We all know that some today have figured out a way 18 to make the current trade-based formula work to their 19 financial advantage without necessarily adding any value into 20 NMS price discovery. And so my question would be, is there 21 anyone who thinks that, by modifying the formula, and as best 22 I can tell, it continues to cling, roughly half, to the 23 existing formula, by adding a few more metrics and not 24 incidentally, a great deal of complexity, does anyone think 25 that the new formula will not be exploited in the same way 221 1 that the existing formula is currently exploited? 2 So I think what -- I repeat, I think you're missing 3 the market. I'm neither for nor against any particular 4 formula. I do think that the stated objectives are A, to see 5 if you can sort of cut off the oxygen -- these are my words, 6 not yours -- cut off the oxygen for some of these 7 objectionable practices, I don't think you're going to get it 8 because I don't think the formula is the issue, and B, 9 implicitly, it seems as though you're trying to encourage 10 better quotations, improving the NBBO and so on. And I will 11 tell you, having modelled this as best we have been able to 12 do at New York, I think it's indisputable that New York 13 quotes better than any other market in the listed space. 14 Intuitively, you would think we would do somewhat 15 better under a formula that rewards quoting. In point of 16 fact, we do somewhat worse, as best we can tell. 17 I don't mean to make an individual point of that, 18 only to suggest the law of unintended consequences. If you 19 think that the current formula is instilling incentives for 20 better quotations, we have an opinion that maybe it isn't. 21 The other thing I would say is that failing to 22 listen to me, and the Commission has a wonderful track record 23 of doing that, probably rightfully so, if you are determined 24 to try and cure this via tinkering with the formula, I would 25 just ask you to do two things. Put it through what I'd call 222 1 an exploitability filter and then do your best to satisfy 2 that it will not be damned in the same way that the current 3 formula is damned. And the other thing I would mention is 4 try to put it through a cost/benefit analysis, and I think 5 that's part of your process anyway, so I apologize for being 6 redundant. But I'll give you an example. 7 Dropping out of the middle of -- what I call the 8 middle bucket, there's the trades, there is the -- if there 9 are three markets that each display the NBBO for 20 seconds 10 and the cumulative size is ten thousand shares, that results, 11 as we understand it, in total quote credits for this minute 12 event of 20 million quote credits for that one single event. 13 Our folks have estimated that the total quote credits only 14 for 2003 on a pro forma historical basis would have been 1.3 15 quadrillion quote credits. I don't know what a quadrillion 16 is but I know it's a very large number. 17 And so I ask you to think about, and I tell you, as 18 we have run the numbers, we lose a little bit but not a whole 19 lot worth mentioning but I guess my point would be we largely 20 come out at the same place from this, now, I can't speak for 21 others, from this complex formula, and in what we end up 22 doing on account of having a formula that is so difficult to 23 administer and the law of large numbers kicking in here, if 24 this ends up being a transfer of wealth from the market 25 centers to the data processing vendors and the data storage 223 1 vendors, I think we will have not accomplished what we're 2 trying to accomplish. So it's a roundabout way of saying, I 3 think -- I would question whether or not you're actually 4 going to meet what I think are your objectives, because I 5 don't think you're going to the heart of the problem 6 practices and I don't think at the end of the day we've got a 7 formula just yet that's going to incent better quotation. 8 Thank you. 9 MS. NAZARETH: Bob? 10 MR. GREIFELD: I think I agree with a lot of what 11 Bob said. So that's hardly specializing. But our context is 12 a little different and I think our recommendations probably 13 go a little further. First, we recognize it's in the public 14 interest to have a basic amount of market data available to 15 investors. We believe that market data is the NBBO and 16 nothing more. If this market data is a part of the public 17 good, then a government mandated utility is a natural 18 consequence. 19 But we believe the government should only be 20 involved where the government must be involved. So we must 21 limit the monopoly to the data that is part of the public 22 good, and provide it at a low cost. Currently that cost for 23 professional investors is around $20. That cost in a NASDAQ 24 market was established by NASDAQ over 20 years ago. It was 25 about $17 twenty years ago. There was no great wisdom in 224 1 that number, and we look at the number today, that number is 2 too high. 3 We agree with that concept. The number probably 4 should be somewhere around five to seven dollars. And that's 5 after some thoughtful analysis on our side, and trying to 6 imply a rough cost-plus basis to the analysis. 7 With the current structure, then, data is not 8 provided at a low enough cost and it does create, as Bob 9 mentions, unintended results and distortions in our market. 10 The market centers today are the beneficiaries of that 11 excessive rent and are utilizing this money to buy prints 12 independent of market center value. 13 If the utility rate was five to seven dollars, then 14 the market centers would have a very difficult time buying 15 prints. And I think that is the problem we're trying to 16 solve. 17 Now, understand that NASDAQ can sit here with great 18 credibility and say we want to endorse this, because we also 19 are a beneficiary of that $20 price range, and I have to 20 admit I get nervous advancing this proposal because what it 21 means is, I have to go to the competitive market and develop 22 data products that provide value beyond the monopoly product. 23 So with this approach, what changes is, first and 24 foremost, investors will win. They will have lower cost for 25 basic market data. In addition, they will have a choice to 225 1 buy additional data products from people like NASDAQ. 2 The brokerage firms also will win. Their customers 3 will trade at a lower cost, lower net effective cost, and the 4 brokerage firms now control customers' order and trade 5 information and they will be in a position to be a value 6 added consolidator to an organization such as NASDAQ. 7 So the dynamic changes from the market center to 8 the brokerage firm that has retail CLOB. The retail 9 exchanges, they certainly will lose in this proposal. NASDAQ 10 starts out losing, we go from $20 to five or seven. We take 11 our risk that we come up with products that bridge the gap 12 and maybe do better in time. Where we win is, we remove some 13 aspects of regulatory arbitrage. What's unresolved is the 14 concept of market data funding regulation. This concept, if 15 we go with the $5 rate will be revealed as false. 16 Money is fungible. I run NASDAQ under this basic 17 truth. I don't assign individual dollars to regulation. But 18 where is the money for regulation going to come from? 19 If we've learned anything today, it is, there is 20 intensifying competition between markets. Regulation cannot 21 be a victim of that competition. Regulation is the 22 cornerstone of the quality of our markets. Regulation 23 requires its own direct and completely attributable and 24 secure source of funding. 25 I've had such a good time here today, as a panelist 226 1 and a spectator, that I look forward to coming before this 2 Commission and testifying on the future of the regulatory 3 function. Thank you. 4 MR. COLKER: First of all, I'd like to thank you 5 for the opportunity to participate today. I'd like to 6 challenge the assumption that the current market data revenue 7 distribution methodology needs to be changed and I'd also 8 like to caution the Commission against making a change which 9 could prove to be far worse than any problem that exists 10 today. 11 I think it's important to put the market data issue 12 in historical perspective. In the last 25 years, the trade 13 was quoting methodology. That system has been in place over 14 25 years and for most of that time it's been perceived to be 15 fair and effective and easy to administer. It really was 16 only first questioned a couple of years ago and that was 17 because, and it was brought up by NASDAQ, because of the 18 National Stock Exchange's success in attracting a lot of 19 order flow from NASDAQ. 20 What we did in brief was combine our old electronic 21 model with a commitment to operate like a utility which 22 allowed us to meet the brokerage community's longstanding 23 need and request for lower market data costs by sharing that 24 data with our members just like any mutual company would do. 25 And what's happened is, the other exchanges in order to 227 1 compete with us have provided similar programs and what 2 that's done is, we estimate it's now saving the brokerage 3 community about $65 million a year. 4 So I think if the SEC is not going to address the 5 explicit cost of market data, which Reg NMS implies it's not 6 going to do, it's very important that you preserve a 7 competitive environment so that the exchanges, through 8 competition, can indirectly continue to push down the cost of 9 market data. 10 Very quickly also, we believe in the premise that a 11 trade-based formula and trade allocation -- applies to 12 behavior has not been sufficiently proven to justify the 13 formula that's being proposed. We also think the formula is 14 unnecessarily complex and while well meaning, it's -- price 15 discovery value junk. Seems to be flawed and very expensive 16 to administer. Just to -- if you understand that today, 17 the -- the rate markets are generating 18 million NBBO 18 changes, and imagine administering cost of trying to 19 determine annually how many thousands of quote credits each 20 quote is due for each of the 23,400 seconds in every trading 21 day. 22 I certainly respect the Commission's desire to try 23 and incorporate -- try and encourage quote competition but I 24 think the benefit to doing so under this proposed formula 25 don't come close to outweighing the formula's administrative 228 1 costs. 2 Finally, I'd like to say that whatever action the 3 Commission ultimately takes, I think it's important to 4 combine such action with changes that require the NBBO and 5 trade reports to become real time. In an electronic world, 6 that means automatic executions, automatic quote updating and 7 automatic trade reporting. So I would offer that maybe no 8 quotes or trades should be given credit for market data 9 revenue unless they are going to go past unrelated markets. 10 Thank you. 11 MR. BROWN: Thank you Annette. I want to applaud 12 the Commission for it's creativity in coming up with a 13 formula for allocation. I don't remember the last time I 14 really tried to think about square roots and things like 15 that. So it really got me, you know, to refresh a lot of my 16 old mathematics. 17 But the point that Schwab believes in is that the 18 market data reform proposal you put forth simply misses the 19 point, and that point is is that market data revenue is too 20 expensive. And I'm really pleased to hear today for the 21 first time that, you know, NASDAQ, you know, seems to have 22 recognized that fact. We at Schwab have been arguing and 23 advocating market data reform for many years and we've heard 24 all the stories and all the rationale about why the market 25 data rates are what they are and frankly, we don't believe 229 1 them. 2 You know, last year, the market data cartels took 3 in $424 million in revenue and had expenses of $38 million. 4 Well, that's a profit margin of over a thousand percent and I 5 mean, we at Schwab wish we had that profit margin, but we 6 think it really reflects that there's too much revenue 7 washing around in this system. 8 And clearly, that excess revenue manifests itself 9 in the types of practices that you're concerned with, you 10 know, tape shredding, market data rebates, excessive pay to 11 executives. And there's clearly a link to, between market 12 data revenue and these practices. We don't think you can 13 ignore that. 14 So we would advocate that, before you have us, you 15 know, or the exchanges buy Cray computers to start figuring 16 out what their market data revenue share is, we deal with the 17 problem; reduce the cost of market data because it's the 18 investors who pay that cost and Commissioner Atkins, you 19 raised it in -- a good point, I thought, in the last, or the 20 panel before the last, where you asked about, you were, the 21 relationship between capping access fees for brokers and 22 capping or not capping the market data revenue structure that 23 we have today. 24 And we certainly think that's a real telling point, 25 that, you know, the Commission is willing to get into rate 230 1 making or proposed rate making for private businesses because 2 you see a harm, but you haven't proposed a rate making -- any 3 action for monopoly priced data that we pay for and investors 4 pay for. 5 So we would strongly advocate that you review this 6 proposal and work with NASDAQ to reduce the cost of market 7 data to investors. Thank you. 8 MR. HALVORSON: Thank you, and again, I appreciate 9 the opportunity to participate this afternoon on behalf of 10 Ameritrade, and the retail investor, we think that we bring a 11 unique perspective to this panel and to the proposed Reg NMS 12 and we're pleased to comment on that. 13 I want to start with the premise that market data 14 is really the fuel, I think, that runs the market. And four 15 years have passed since the Commission issued its concept 16 release concerning market data structure. Ameritrade in this 17 regard is disappointed in a couple of things; one, that Reg 18 NMS does not address market data and related revenues in a 19 comprehensive fashion, and two, that the Commission has not 20 taken the step of requiring transparency by requiring SROs 21 and SIPs to publicly disclose the costs, true costs of 22 providing market data to the public. By comparison, recent 23 Rules 11Ac1-5 and -6 have contributed greatly to the 24 transparency and competition and in the order flow arena. 25 Similar market data transparency would increase competition 231 1 and potentially reduce costs for all users. 2 Accordingly, Ameritrade is not today interested in 3 necessarily debating the merits of the proposed allocation 4 model as we're not a SIP, as we're not an SRO, and we're not 5 interested in debating the merits of necessarily the hybrid 6 model. What we are interested in gaining an understanding of 7 the costs associated with providing the data and determining 8 the appropriate structure to allow for either return of these 9 excess revenues back to investors, or a model in which market 10 data revenues simply equal the costs of providing such data 11 to the investing public. 12 Market participants like Ameritrade are forced to 13 pay the costs of the very data they provide. Any broker or 14 vendor, as those before me have indicated, can conduct 15 business in the current environment, will tell that you the 16 process is overly costly, complicated and burdensome. 17 SROs require a tremendous amount of detailed 18 information about how a firm like Ameritrade may use market 19 data, the type of services that we provide, our use of 20 technology and how we monitor the users of the information. 21 Ultimately, brokers like Ameritrade must share confidential 22 and competitively sensitive material with the SROs and SIPs. 23 Further, the SROs require individual investors to consent to 24 an agreement which requires the payment of sometimes 25 discriminatory fees, and is a contract replete with legalese 232 1 and confusing terms or conditions which many retail investors 2 don't understand. 3 We spend an inordinate amount of time and money 4 simply complying with the administrative burdens of tracking 5 market data used by our customers in maintaining two separate 6 systems, one for real time data and one for delayed data. 7 The current proposal we believe does not address these 8 issues. 9 Again, to Jeff's point, we believe that the SROs 10 and SIPs are really granted monopoly powers in terms of the 11 fees charged and the control over the dissemination of that 12 data. We believe market data fees are imposed in a 13 discriminatory fashion. For example, investors accessing 14 realtime quotes through an account executive by phone, for 15 example, or from media distributors, do not incur market data 16 fees. That same investor who is accessing quotes on-line are 17 charged a market data fee each time a realtime quote is 18 accessed, and whether it's charged directly to the end 19 investor or the retail customer, or to the firm, ultimately, 20 that customer is ending up paying higher costs for the use of 21 that data. 22 The current definitions around professional and 23 unprofessional users are inconsistent and costly to the 24 retail public as well. As Jeff pointed out, the release 25 notes that out of the $424 million in revenue derived from 233 1 market data fees, we'll take a little bit different look at 2 this. 386 million or the corollary is distributed back to 3 SRO participants. The Commission previously has said, "The 4 total amount of market information revenues should remain 5 reasonably related to the cost of market information." And 6 we believe there is no transparency to determine whether it 7 actually costs anywhere near 386 million in order to provide 8 data back to investors. 9 So we submit that the way to determine whether 10 there has been an equitable and reasonable allocation of 11 costs is to require the SRO or the SIP to have publicly 12 audited financials regarding the true cost of providing 13 market data to the end users, including the retail investing 14 public. 15 Ameritrade recommends these financial statements 16 similar to the disclosures under 11Ac1-5 and -6 be made 17 available to the public through the SEC's website. Given 18 that investors ultimately pay these fees, either directly or 19 indirectly, we clearly believe that requiring the 20 transparency of such information is in the public's best 21 interest and only then can the true cost of such data be 22 analyzed and act as the basis and direction for what we 23 believe should be future market data reform, both in terms of 24 pricing and ultimately the distribution of these revenues. 25 Thank you. 234 1 MR. McBLAIN: I should explain that the Financial 2 Information Forum is an organization that focuses on the 3 industry's information systems. Our participants come from 4 the markets, from broker/dealers, and from the vendors who 5 provide these services to the community, the market data 6 vendors, service brokers, so when we look at any of these 7 issues, we typically do it from the perspective, can the 8 information systems handle it, and how might it impact the 9 systems. 10 We've heard this morning and all day today, of 11 course, the NBBO, really has changed over the years. When 12 the National Market System rules were first promulgated, it 13 was pretty much a system of people acting with systems 14 linking them. We've heard today that today, it's very 15 difficult for people to be taking the trading actions and 16 monitoring the information. The NBBO doesn't have a duration 17 of several seconds. It's down to milliseconds. And so the 18 character of the market to a great degree is changing to 19 where it is computer black boxes interacting. 20 And so the duration of the NBBO is much shorter and 21 I do think that the proposed regulations are good in that it 22 has reduced the requirements for the display of consolidated 23 information. Data vendors today are required to maintain a 24 montage of quotes for all markets and that looks like that 25 would not be under the proposed information regulation. So I 235 1 think it is good that the -- we have to realize now that much 2 as we had real time information and delayed information in 3 the past, as a basis of segregating market data, to some 4 degree, there is now a different tiering of the information, 5 and that the information as it is used by the black boxes vs. 6 the information that is displayed an a terminal and watched 7 by the human eyes. 8 And in many cases, on these fast market conditions 9 that we're talking about, we can really see that it's very 10 difficult for a person to be analyzing what they see on a 11 screen and taking action with the same degree that the 12 trading systems do. 13 I'd also like to point out for clarification, I 14 know we're talking about the formula, but I think it's hard 15 to answer that question without addressing some of these 16 other areas. 17 I would not walk away with the idea that technology 18 is not a problem. I think we have heard that from the 19 participants in these panels, but remember, again, they are 20 representing large organizations where computer power 21 comes -- they have the dollars to do that. 22 We might believe as we look at the Internet and 23 information today to get the impression that things are real 24 time but the fact of the matter is that there still are 25 communications systems and there are propagation delay times 236 1 and there are processing cycles. And when you look and try 2 to draw up a page on the website, for example, there truly is 3 a perceivable display at some time. 4 So I guess the question, the formula is, the first 5 question that we had is, who would do this calculation? 6 Since the whole matter of who might be the consolidator has 7 been put up for question, that gives rise, now, then 8 presumably whoever it is that calculates the NBBO would also 9 be the calculator of the allocation of these revenues. 10 And we've heard the previous speakers say, it is 11 very complex. When the NBBO, the components that go into the 12 NBBO are changing in milliseconds to somehow capture that and 13 put that data together is a significant burden, perhaps it's 14 a burden that the party that might be interested in being a 15 consolidator may not, given the fact that he becomes a part 16 of the task. 17 So that's basically, as I say, as we look at it, we 18 don't make a judgement as to what this does to the markets 19 but what we view here is that it indeed seems to be something 20 that's going to require much greater capacity in the 21 distribution, the downstream people who do not necessarily 22 have the dollars to spend on high powered computers. 23 So that is our concern as we look into it, is that 24 the formula appears to be very complex, probably adding a 25 cost burden, and then finally what I would look at is to say, 237 1 if in fact investors do realize that it takes a black box to 2 really get the best price, will they perceive that the data 3 that they see on the screen now has less value if it is not 4 data that they can act upon for their own investment 5 decisions. 6 So thank you. 7 MR. LYNCH: On behalf of Reuters, I appreciate the 8 opportunity to share our views on the proposed Regulation 9 NMS. They are not really focused on the -- on the structure, 10 but the effects of the ultimately formulated approach as to 11 how the fees are distributed. Although Reuters is best known 12 as the worlds largest news agency more than 90 percent of our 13 revenue actually derives from our financial services 14 business, providing professionals in equities, fixed income, 15 foreign exchanges, et cetera, with services. 16 Our open technology, which is based on industry 17 standards, enables our customers to integrate our information 18 with content from other sources, and we provide institutions 19 with the ability to reduce risks, distribute and manage an 20 ever increasing volume of market data. 21 We think that as a general principle, we support 22 that the -- the introduction of greater competition and 23 flexibility into the consolidation and distribution of the 24 U.S. securities information. We believe that investors 25 should have greater freedom to decide which information they 238 1 need to conduct their business and make investment decisions. 2 It is our view that market forces should play an even greater 3 role determining the value of information and sharing that is 4 provided for investors. 5 Turning to the specific proposals, we will be 6 filing a more detailed comment but I wanted to offer a couple 7 of things. One is that we are encouraged by the proposal to 8 eliminate the montage portion of the consolidated display 9 requirements. We think the Commission has correctly 10 identified the data which is most important to investors; 11 namely, price, size and the market center of the NBBO. 12 We agree that market forces and investor choices 13 rather than regulatory requirements should determine what 14 additional data should be displayed. We also see merit in 15 the proposal to require consolidated display only in the 16 context in which a trading or order decision could be 17 limited. The language of the proposed rule needs to be 18 studied more carefully. We think the proposal would promote 19 the availability of more information at a lower cost, while 20 ensuring that investment decisions are made upon the basis of 21 complete and accurate information. Being the last speaker, 22 on the last panel, I wanted to keep my remarks brief. I 23 appreciate the opportunity. 24 MR. GOLDSCHMID: We do, too. I do have a question 25 or two, I'm not trying to hold everybody. You suggested the 239 1 Commission identified the right data. And Bob, I take it you 2 were saying the same thing. 3 Is there anyone who doubts, one, that that data is 4 important, and two, that someone at the Commission has got to 5 take a role in making sure it gets out, without worrying 6 about the price or allocation of it? 7 Did I hear everybody in effect saying the 8 Commission's got to do this if we're going to have a national 9 market system, with a kind of sensible transparency? That's 10 fair? 11 MR. LYNCH: We think it should be available on 12 similar commercial terms to everybody, so it should be 13 available equally. 14 MR. GOLDSCHMID: And then I take it the second 15 basic question is, what kind of price should be involved in 16 it? How much cost? And I think fairly, people have said, if 17 I understand it correctly, it's four hundred some-odd million 18 dollars involved and we're not sure why it's that number, or 19 whether it ought to stay, with Bob and several others saying 20 it's got to be way too high; is that what I'm hearing? 21 MR. GREIFELD: Yes. 22 MR. BROWN: Yes, sir. 23 MS. NAZARETH: Bob, I'm sure, had a view on this. 24 I heard a bunch of different things. I heard cost, whatever 25 that is, and it sounded like the cost for -- technology cost, 240 1 putting it together, I heard something less than $20, and 2 then I heard let the market -- let market forces determine 3 it. So I think we heard many different formulations. 4 MR. GOLDSCHMID: Well, Annette, once you conclude 5 that someone -- there has to be one central source of it, 6 then you're not talking about the market doing it. 7 MS. NAZARETH: I agree. 8 MR. GOLDSCHMID: Bob, where are you on -- 9 MR. BRITZ: Commissioner, there are different 10 economics depending upon which securities we're talking 11 about. I'll give you but one example. The last time I 12 looked, the priciest market data, at least in the equities 13 space, on a per print basis of all markets was American Stock 14 Exchange data by a significant factor vs. New York. 15 NASDAQ's, I don't know them in a granular way, but I'm 16 guessing their economics are very different than New York, 17 and National is very different as well. 18 I would share with you a couple of data points, 19 because I think if I may say so, I think there was a little 20 bit of fun with numbers taking place a short while ago. 21 There was a number about $28 million in market data cost. 22 The Commissioners should understand, those are the marginal 23 distribution-only costs. They are not the costs to actually 24 produce the data. There is no distribution and therefore no 25 distribution cost if there is no production of the data. So 241 1 you should just understand the space that that number refers 2 to. 3 In the New York's case, and I only have visibility 4 obviously into the New York's cost structure, I understand 5 the CTA overall revenue structure, in New York's case, it 6 costs us $488 million last year to produce market data. 7 Now, let me bring a little clarity to that number 8 because people like Commissioner Glassman would suggest that 9 when we execute a trade, we're doing two things. There's 10 essentially a joint product there. Within one hand, we're 11 executing two or more investors' orders, so that's the sort 12 of transaction processing piece of the puzzle, on the other 13 hand, we're manufacturing market data. And people like me 14 come down on the side that at the end of the day, transaction 15 charges are -- and the agency execution fees are about as 16 interesting to the counterparties to the trade, the price of 17 the data that dances across the tape, or the actual price 18 itself, is interesting to millions of people around the 19 world, and so you get into interesting allocations of cost as 20 between what are you doing, executing a trade or producing 21 market data. 22 I'm a little rusty in my cost accounting skills, 23 but the good news is, you don't actually have to go there. 24 In the New York's case, against $488 million of cost to 25 produce the data, 172 -- 488 -- excuse me. $178 million 242 1 in -- $172 million in market data revenue, including twelve 2 from proprietary data, so 160 from the consolidated regulated 3 data and $151 million in terms of transaction revenues. 4 So you don't even have to make the distinction. If 5 you combine the 151 in transaction revenues and the 160 or, 6 depending upon your point of view, the 172 in market data 7 revenues, you get $323 million worth of revenues against $488 8 million worth of costs. 9 I've been on the other side of the Schwab debate 10 for about ten years now, and in the very first meeting, with 11 Mr. Brown's predecessors, I said that if Schwab could get the 12 New York to the point where it could recover its costs of 13 market data, I would pay their legal bills for the campaign. 14 So I emphasize and I throw those numbers at you because I 15 think each market has a cost structure that's very different 16 and the economics of market data are very different as well. 17 MR. GOLDSCHMID: Go ahead. 18 MR. BROWN: Well, as one of Schwab's lawyers, I 19 particularly appreciate the thought that someone else may pay 20 some money to us to raise our salaries. But the issue, then, 21 becomes, everything that goes into manufacturing and 22 execution and disseminating and apparently adding up to this 23 488 million. We would question that. First of all, we don't 24 know, because the -- there's no real accounting to anyone 25 about this. 243 1 Secondly -- 2 MR. BRITZ: On the contrary, those numbers are 3 published in the New York Stock Exchange's annual report. 4 MR. BROWN: Yes, but it's hard to understand 5 exactly what relates to the production of market data. 6 Second, and more importantly, in my mind, there is a 7 distinction about, you know, the order flow that comes in 8 isn't necessarily their product. You know, that the formula 9 you've come up with and the basis for it that you talked 10 about was rewording SROs for price discovery. 11 Well, in my mind, SROs don't discover prices. It's 12 the order flow within those SROs that discover prices, and 13 the fact has to be that that order flow is what creates the 14 market data. And it's customer orders doing that. So to say 15 that, you know, it's all a cost to the New York Stock 16 Exchange to produce that data is not quite true. Brokers 17 have their own expenses to produce that information, to 18 transport it from the customer to the execution venues. 19 Well, all of that has to go into the calculation. But what's 20 coming out is the investors are paying for the entire panoply 21 of cost that relate to market data. 22 MR. GREIFELD: We've spent a lot of time in the 23 eleven months since I've been at NASDAQ understanding our 24 cost structure. And we are saying to this commission today 25 that, at a five to seven dollar rate for basic data, we have 244 1 the ability to provide a decent return to our shareholders. 2 So I don't sit here today saying we have the lowest 3 cost structure in the world and we're certainly working to 4 reduce our cost base, but it is inconceivable to me as 5 someone now who is responsible for the SIP in the NASDAQ 6 world, who has the technology background, to figure out any 7 way to spend anywhere near the $488 million, and certainly, 8 we would welcome the opportunity to take a contract from New 9 York Stock Exchange to maybe run it a little more 10 efficiently. But, you know, investors have to be the 11 beneficiary here. There is just too much money in the 12 system. It's been allowed to become inefficient and we're 13 saying that at five to seven dollars we can make a decent 14 return on that. And I think that's the right answer for 15 investors. 16 MR. HALVORSON: I can't speak to Mr. Britz' numbers 17 from the NYSE. I know that in doing the mental math when 18 Jeff was speaking, he and I used the same basis of numbers in 19 the marketplace that the Commission has provided. And I 20 would simply suggest that if the costs are indeed real, then 21 let's just open the books and take a look at them and give a 22 true line-by-line display of what it actually costs to 23 produce and distribute the data. 24 To Jeff's point, from a retail inventory 25 standpoint, and really the retail order flow, it's the 245 1 institutional order flow that's feeding the quotes that are 2 displayed. Nobody's making them up. They are based upon 3 real orders coming to the marketplace from the end investors. 4 So they are the ones that really own the data that's being 5 produced. And that's why we simply make the statement and 6 have the position that those market data revenues, whatever 7 is in excess of costs, if it's not a revenue equal cost 8 model, should come back to the individual investor. Those 9 that ultimately provide the quotes. 10 MR. GOLDSCHMID: Now, let's say we did that, would 11 you be concerned at all about any other data being the 12 product of the SRO? 13 MR. HALVORSON: I think that that does raise some 14 interesting questions. There is a lot of other data that's 15 provided and within the proposal, there's questions about how 16 we segregate that, do we put that into this thought of an 17 independent model vs. consolidated model, and I think there 18 is other market data that perhaps has a different pricing 19 scheme or a different model to it. 20 MR. BROWN: To your point, sir, obviously in 21 today's environment, depth of book data is becoming more and 22 more important to everyone. And we thought that permitting 23 SROs to sell that data independently of the monopoly system 24 is a good thing. Let's see how it works, and I think there 25 are entities today, we've heard them, ECNs say, that put 246 1 their data out on websites for free, their full book. So 2 there are going to be cost pressures to on -- on the entire 3 system to drive down and provide competition for data. So 4 that might work. 5 But we're also concerned at Schwab about creating a 6 two-tiered system of data where the investor will only 7 receive the NBBO and institutional investors and those who 8 can afford it will be seeing the full book and really be able 9 to judge where the market may be going or where is the best 10 trading opportunities. 11 That concerns us and we really would encourage the 12 Commission to review that and think carefully about whether 13 the NBBO is sufficient for making trading decisions by a 14 retail investor. 15 MR. COLBY: Can you explain why, if you're 16 concerned -- what you're concerned about whether the NBBO 17 prices are -- what's the market pressure going to be on the 18 depth of the book and that's going to keep those prices low? 19 Because depth of book currently for market makers is in the 20 hands of NASDAQ, the limit order display is a separate 21 product and the -- and New York has the entire New York book, 22 nobody else has it. So where is the market pressure to -- 23 MR. BROWN: Well, one, I guess I'd look at what's 24 going on currently. The full book, and I'm not sure of the 25 actual names that the markets have for them, but they are 247 1 available now and they are not, people aren't rushing to buy 2 them. They are expensive, but the -- the institutions are 3 buying, but I think that if we can permit competitive forces 4 to, I mean, you're right, the one concern is, if you're the 5 sole -- if you're 90 percent of the volume, and you have 90 6 percent of the order flow, what does anyone else have to show 7 to compete? I mean, that's a fair point. 8 But if, like in the NASDAQ market, where there is a 9 greater diversity of centers of orders, then there can be 10 competition for that information. 11 MR. GREIFELD: I think that two competitive 12 disciplines will keep the price in check and I think the 13 basic question is if you have 80 percent basic market share 14 beyond the NBBO, can you charge anything. One is, the rules 15 contemplated with Reg NMS, you will not see an 80 percent 16 market share player. That tends not to be a natural outcome 17 of competitive markets. So you have that one factor. 18 But the second factor is the contributors to this 19 flow have the ability to form their own consortium and 20 develop their own product. So I look at it as NASDAQ in the 21 world that we would advocate, you could have Merrill, Morgan, 22 Goldman get their information together and come up with a 23 trade theme. They are perfectly able to do that. 24 So you will have the ability for alliances to shift 25 based upon, you know, deals that are offered, and that's the 248 1 competitive market that we want to have. You don't have that 2 ability today when the market centers are long. So to the 3 extent we put it towards the retail investors, close to the 4 retail investor, one, they benefit, and two, it should create 5 competitive dynamics in terms of how it's going to work. 6 MR. ATKINS: I, as far as I'm glad we have both 7 Schwab and Ameritrade here. I was wondering from your 8 perspective what do your customers, your retail customers 9 want as far as services via the order execution, the 10 information that they are not able to get or that they don't 11 receive at a reasonable fee? 12 MR. HALVORSON: Well, from the Ameritrade client 13 perspective, they basically want, you know, some fairly 14 simple things. They want a tradable quote, right, they want 15 to be able to get the price they see or better, as quickly as 16 possible. And whether it's a -- whether it's an active 17 trader or whether it's a longer-term investor, they both seek 18 the same things. 19 And we've actually got written -- written comments 20 on the rest of the proposal which we will be pleased to 21 submit to the Commission that addresses our viewpoints on the 22 trade-through aspects and the other aspects as well. But 23 Commissioner Atkins, it really is pretty simple in terms of 24 what our clients want. And they want the quote that they see 25 to be executable in a very quick time frame. 249 1 MR. BROWN: And we shouldn't really forget the 2 other piece that is really very important, too. Trade people 3 who are trading or contemplating a trade, and that is the 4 series of previous trades. Having access to the stream of 5 trades gives you an idea of where the next trade is going to 6 be. So that's a service that we do provide and they will 7 want. 8 I think one of the -- and Kurt raised a very good 9 point earlier, about what our customers don't want is when 10 they want data, they get sent a ten-page contract that they 11 have to sign as an individual and they can't even begin to 12 understand what's in that document. And they call us and we 13 have to work them through it and try to help them understand 14 before they sign. 15 That process, this onerous administrative process 16 that we have three networks to deal with, has to change. I 17 mean, it's become so time-consuming for the brokerage 18 community, the man-hours that applies to this has to -- has 19 to change. 20 MR. HALVORSON: I also just to follow on -- with 21 Jeff's comment, I would also just suggest that when we look 22 at the costs associated with market data fees, we really 23 aren't including costs that broker/dealers are incurring to 24 manage the process that Jeff and I just alluded to, that 25 process of organizing and tracking and distributing these 250 1 agreements to our clients and also assisting them in 2 understanding what's required of them before they can gain 3 access to the very data that we believe they are providing. 4 MR. BRITZ: I speak only for New York. It's on the 5 Internet. It's a one-page agreement. It's click on, so I 6 don't know what it is Ameritrade and Schwab are organizing 7 and distributing to their customers, and there are lots of 8 other market data vendors and agreements out there, so 9 perhaps they are, but it's not the New York. 10 MR. HALVORSON: And, Mr. Britz, I would suggest 11 that you're right, ours is a click through, it's an 12 electronic document, and there are a number of different 13 agreements out there. And there, you know, the complexity of 14 those is what we have stated. They are written largely in 15 legalese and they are difficult for a retail investor, even a 16 sophisticated investor, to understand. 17 MR. CAMPOS: Given that -- if we can go back a few 18 steps, we need a requirement in government to make sure that 19 this data is available, begin there, and then you know, we 20 can see that there's some debate about what's the right 21 price, and it's clearly, you know, some are saying it's too 22 expensive to achieve, or the more you look into it, the more 23 discussion there is about really whether the price is right 24 or not. So that's a key point. 25 And given that there's clear distortion today given 251 1 the way the formula is allocated, given what's going on, why 2 not try to do something like what the proposal puts out which 3 is essentially a, you know, we don't know totally what the 4 right numbers are, but let's allocate it fairly and let's try 5 to promote it even if it's imperfect, you know, quotes and 6 being the NBBO price, how bad is that if we go forward with 7 the proposal? 8 MR. BRITZ: Commissioner, can I suggest the 9 construct to your first point, which is not quite sure 10 whether -- whether it's the right price, and -- 11 MR. CAMPOS: I'm just reflecting what I'm hearing. 12 MR. BRITZ: And I'd just like to offer a comment on 13 that and perhaps one way to think about it, as proposed to 14 the Commission having to grapple with providing investors 15 access to this information, they can hardly avoid it. It's 16 on television, it's on the Internet, it's in airports, in 17 train stations. It's in restaurants and bars. And I would 18 offer the suggestion that the very pervasiveness of this data 19 is suggestive that the cost is absolutely at the margin. One 20 way to think about it. 21 MR. COLKER: I would question whether distortions 22 existed. I would take exception to that. I think the 23 exchanges today are competing very strongly for order flow; 24 we're doing it through a combination of costs, technology and 25 service innovations and our success is measured in terms of 252 1 the number of trades we're able to record, and that, I think, 2 that's appropriate. And the competition, at least on the 3 NASDAQ and AMEX side has successfully pushed down the cost of 4 the data to the tune of 65 million a year on a NASDAQ side, 5 about fifty million on the AMEX side, and if the one 6 remaining barrier to competition on the New York side, which 7 is trading -- that comes down, I think we'll see enormous 8 pressure on market data on the other side, if not through 9 market forces. So I would argue strenuously to let the 10 market do its thing. You're going to see, in order for the 11 exchanges to respond to the major retailers, you're going to 12 see the cost of market data come down significantly, it 13 already has. 14 MR. CAMPOS: You're just saying leave it alone? 15 MR. COLKER: I think it's working pretty well. 16 MR. GREIFELD: I would disagree with that, but I 17 think the current proposal trying -- we think it's just a 18 second opportunity for gamesmanship and we've already 19 conceived different ways that people can develop programs 20 that can basically arbitrage that and we don't think it 21 solves the issue. 22 With respect to the allocation beyond trades and 23 quotes, we think that proper allocation should be about 24 shares. It seems the easy way to go and less subject to 25 gaming. But again, our overall point is, those reallocation 253 1 formulas do not get at the heart of the issue. And the heart 2 of the issue is that the investors need an NBBO. It should 3 be at a very low cost, and vendors such as ourselves should 4 be able to sell above and beyond that on true value. 5 MR. CAMPOS: So you want us to set ceilings? 6 What -- 7 MR. GREIFELD: We believe when you agree with the 8 concept, you've got certain data that's required by all 9 investors, that's part of the public good, then the natural 10 corollary to that is that the government should set the rate 11 for that data. And it does pain me to say that but it's the 12 right answer. They are tied to together. And we are 13 revealing our hand. Twenty dollars is too high by any sane 14 measure of the costs associated with it. 15 MR. BROWN: And, you know, it's the government that 16 has mandated that everyone purchase this data. So you've 17 gone that step. You've created a monopoly price for this 18 data. So now it's in your hands to regulate that price. 19 MR. GREIFELD: And that monopoly price has not been 20 thought of for a very long time. NASDAQ itself set that 21 price twenty years ago. So we have a monopoly price that was 22 set almost twenty years ago without any active review of how 23 that relates. And any time you're dealing with a monopoly 24 price, then the participants have the right to understand how 25 that price is derived. 254 1 MR. BRITZ: In case I was too subtle earlier on, 2 there another way to get at so called monopoly price and that 3 is to bust up the consortium. 4 MR. DONALDSON: Okay, I think we're coming to a 5 close. I want to first of all thank this panel, I want to 6 thank those members of the other panels who are still here. 7 I also want for say to all of you who are sitting behind this 8 desk, I fear that you have been sitting on your hands and 9 biting your tongue and not saying things. I hope we'll hear 10 from you one way or another by e-mail or by direct visitation 11 to the SEC or by carrier pigeon or any way, help us with your 12 thoughts and arguments. Thanks to the staff for the work in 13 putting this together, and thank you for bearing with us all 14 day long. 15 (Time noted: 5:09 p.m.) 16 * * * * * 17 18 19 20 21 22 23 24 25