1 MEETING 2 of the 3 MARKET STRUCTURE HEARINGS 4 New York University 5 Tisch Hall 6 New York, N.Y. 7 November 12, 2002 8 9:00 a.m. 9 ANNETTE L. NAZARETH, Chairperson, Presiding 10 DIVERSIFIED REPORTING COMPANY, INC., 11 1025 Vermont Avenue, 12 Washington, D.C. 20005 13 (202) 296-9626 14 15 16 17 18 19 20 21 22 23 24 Diversified Reporting Services, Inc. 25 (202) 467-9200 1 P-R-O-C-E-E-D-I-N-G-S 2 COMMISSIONER GLASSMAN: Good morning. I'm 3 Commissioner Cynthia Glassman, and on behalf of my 4 colleagues, Chairman Pitt, Commissioners Goldschmid, Atkins 5 and Campos, who will be here in a second, I'd like to welcome 6 all of you today to our market structure hearing, and thank 7 you all for your participation. 8 Our first interactive session in Washington, on 9 October 29th, was very informative, and I know I speak for 10 all of my colleagues when I say that we have been looking 11 forward to the second session today. 12 As I said at the last session, the purpose of 13 today's session is to educate the Commission on issues 14 relating to the structure of the U.S. equities markets, with 15 emphasis on how the public interest and protection of 16 investors can best be served. 17 A key subtheme is whether our well-intended rules 18 to protect the public have unintended consequences. We want 19 the benefits of your views on the topics we've chosen for 20 discussion today. Even more useful to us, will be the 21 opportunity to listen to you, the panelists, discuss and 22 debate this array of issues. 23 In my ten months as Commissioner, I've had meetings 24 with representatives of many different markets and market 25 participants, and each meeting has been useful. But what I 1 really wanted to do was to get all of the differing 2 viewpoints on a particular topic represented in the same 3 room, and let the people advocating one position respond to 4 those advocating a contrary position. No holds barred. 5 We had some lively discussion last time, and this 6 is what we hope will happen today. 7 As far as our expectations, it's obvious that two 8 days of hearings, no matter how productive they are, are not 9 going to answer all the questions. But today's discussion 10 will enhance our understanding of current market structure 11 policy concerns, as well as possible tradeoffs that we may 12 have to make as we attempt to resolve the important issues 13 before us. 14 Before we get started, I'd like to again thank the 15 key members of our market structure working group. Annette 16 Nazareth, Bob Colby and Alden Atkins and their staff in 17 market reg. Larry Harris, our chief economist and his 18 colleagues; Giovano Presioso, and Meridith Mitchell, who are 19 not here today, but from our general counsel's office. Steve 20 Jung on the Chairman's staff and last but not least, Mary 21 Head and Brian Stern on my staff. 22 Now, let's get started. I'll turn this over to 23 Annette. Thanks. 24 MS. NAZARETH: Thank you, Commissioner Glassman. 25 I, too, would like to welcome the Commission, our 1 distinguished panelists and members of the public to this 2 second market structure hearing. Today the Commission 3 continues its more than thirty-year tradition of critically 4 examining issues relating to the structure of our equity 5 markets. Prior Commissions have faced many complex and novel 6 issues in this particularly challenging area of securities 7 regulation, both before and after the Securities Act 8 Amendments of 1975 that gave the Commission a mandate to 9 build a better national market system. 10 To its credit, the Commission took actions that 11 have resulted in a vibrant, healthy and efficient 12 marketplace, a marketplace indeed that is the envy of the 13 world. As technology continues to transform our lives and 14 the way we do business, once again we stand at a crossroads 15 with respect to various market structure issues, and the 16 solutions we arrive at will have far-reaching consequences, 17 especially in this highly competitive marketplace. 18 We hope that these hearings will achieve several 19 important objectives. First, as Commissioner Glassman 20 mentioned, we hope that these hearings will provide a 21 comprehensive picture of the structure of the U.S. equities 22 markets. Our marketplace consists of a sweeping variety of 23 market centers and trading methods that cater to different 24 types of securities and different categories of investors. 25 Some markets are centralized, others are decentralized. Some 1 markets are driven by the interaction of customer orders, 2 others are propelled by the willingness of designated market 3 intermediaries to committed capital. Some markets are 4 grounded in the physical trading floor where bids and 5 offerings meet face-to-face. Others float in the anonymous 6 realm of cyberspace. 7 Another goal of these hearings is to discuss the 8 complex tradeoffs that we have to make in regulating market 9 structure. How should we reconcile the investor's need to 10 obtain current information about market activity with each 11 market center's desire to exploit the commercial value of the 12 data that it generates? How should we reconcile the 13 investor's desire to have access to the best prices with the 14 right of individual market participants to employ the 15 business model of their choice? 16 When should a market be required to act in the 17 public interest, rather than in the interest of its members 18 or shareholders? 19 These questions do not have simple answers, and 20 there are many difficult decisions to be made. We are 21 nevertheless committed to resolving these competing interests 22 in the fairest and most balanced manner possible. 23 A third goal of these hearings is to use what we 24 learn to maintain consistent policy positions on the key 25 issues facing the Commission. In recent years, the 1 Commission has undertaken a number of market structure 2 initiatives. Just to name a few, the Commission adopted 3 order handling rules for market makers and ECNs in 1996, 4 Regulation ATS governing alternative trading systems in 1998, 5 and order execution and order routing quality disclosure 6 obligations for market centers and retail brokers in the year 7 2000. 8 While some may differ, most would say that these 9 rules have significantly improved the transparency and 10 accessibility of our markets. 11 The Commission has also solicited comment on a 12 number of specific areas of market regulation. Among others, 13 the Commission has issued concept releases on the regulation 14 of market data fees and revenues, the fragmentation of the 15 U.S. securities markets, and the regulation of exchanges. 16 The Commission also convened an advisory committee in 2000, 17 chaired by Dean Joel Seligman, to consider issues related to 18 the collection and dissemination of market information. 19 These efforts to develop sound policies for market 20 regulation must continue if we are to make our markets 21 fairer, more efficient, and more transparent. To this end 22 the Commission seeks to ensure that the decisions it makes in 23 the coming months and years will benefit from the wise 24 counsel of the industry, the academic community, and the 25 public. 1 Now, before I introduce our participants today, 2 I'll briefly mention just a few housekeeping matters. This 3 morning we will have two sessions. In the opening session, 4 we'll touch on the overarching principles of market structure 5 and discuss whether there is a consensus on those guiding 6 principles. This session will last only an hour, and will be 7 followed by a fifteen-minute break from approximately 10:30 8 to 10:45. We may run a little bit late. 9 We will then follow with our first major session on 10 market data to be moderated by Bob Colby of the Division of 11 Market Regulation. At 12:15 we'll break for lunch for one 12 hour and then reconvene at 1:15 for our afternoon program. 13 As to the general groundrules, in order to keep the 14 dialogue moving and to afford everyone an opportunity to 15 speak, and I am very optimistic, I'm told we have a group of 16 good chatters here, I'd like to ask each of you to attempt to 17 limit your responses to two or three minutes, and given the 18 limitations on time, we also ask that you not pipe in just to 19 repeat a point that's already been made. 20 The next groundrules are just as tough. Try to 21 leave your ego at the door, your own personal business 22 perspectives and biases, and keep in mind that the 23 Commission's ultimate goal is the protection of investors; 24 that this is a policy discussion after all. And I'd also 25 like to note that just outside this room, we have a black 1 box, which you can put note cards with questions that you may 2 have for the panel that, time permitting, we will attempt to 3 address later in this program. 4 So with that underway, I'd like to introduce each 5 of the participants briefly and then circle back and give 6 everyone the opportunity, if you wish, to make a brief 7 statement, although it's not required, but you're certainly 8 welcome to. 9 Before I do that, I would like to introduce our 10 Commission which is present today. We have Chairman Pitt, 11 Commissioners Goldschmid, Atkins and Campos and of course 12 Commissioner Glassman. I'd especially like to thank Chairman 13 Pitt for calling for these hearings and Commissioner Glassman 14 for her very substantial efforts in putting this program 15 together. 16 We also have at the table Bob Colby, who is the 17 Deputy Director of Market Regulation, and Larry Harris, who 18 is our chief economist. 19 Going around the table, in order, we have Robert 20 Murphy, who is here from LaBranche. We have Commissioner 21 Campos. We have Albert Kyle, who is a professor from Duke 22 University; Minder Cheng, who is with Barclay Global 23 Investors, Ian Domowitz, with Investment Technology Group; 24 Janet Angstadt with Archipelago; Matt DeSalvo with Morgan 25 Stanley; Scott DeSano from Fidelity Management and Research, 1 I think the staff is pulling a stunt on me, DeSalvo and 2 DeSano next to each other. Larry Harris, Chairman Pitt, Bob 3 Colby, Commissioner Glassman, we have Tom Peterffy from 4 Interactive Brokers Group, Tom Gardner, who you may not 5 recognize without his hat, here from Motley Fool, Paul 6 O'Kelly from the Chicago Stock Exchange, Robert Steel from 7 Goldman Sachs, Gunner Burckhart with Deutsche Asset 8 Management; Commissioner Goldschmid, Commissioner Atkins and 9 Rubin Lee from Oxford Financial Group. 10 I'd like again to give each of you the opportunity 11 to make a brief statement. We could start, Bob, with you, if 12 you'd like to. 13 MR. MURPHY: Be happy to. Good morning, I am Bob 14 Murphy, New York Stock Exchange specialist with LaBranche & 15 Company as well as New York Stock Exchange Director and Vice 16 Chairman of the board. I wish to thank the Commission for 17 affording us this opportunity to exchange views on issues 18 relating to the structure of our security markets. 19 Our markets can only benefit from these hearings. 20 Specialists on the New York Stock Exchange are accountable to 21 the investing public, for the quality of NYSE markets and 22 their specialty stocks. This means that the specialist is 23 responsible for fostering and maintaining liquidity and 24 continuing two-sided auction markets on the NYSE floor. The 25 specialist plays a pivotal role in insuring that investor 1 orders receive the best possible execution. To accomplish 2 this, specialists act both as agent and principal in their 3 specialty stocks, to help insure that markets for those 4 stocks are fair and orderly, and that they operate 5 efficiently in the public interest. 6 The NYSE specialist is the focal point of the 7 auction price discovery process. The specialist's primary 8 job is to bring buyers and sellers together in an efficient 9 and orderly manner to provide transparency and liquidity and 10 to maximize the opportunity for public buy and sell orders to 11 interact without getting between them. 12 The NYSE specialist has an affirmative obligation 13 to step in and step up when no one else wants to. The 14 specialist buys or sells as principal when necessary to 15 minimize short-term imbalances between supply and demand. 16 In this way, the specialist provides depth and 17 price continuity, minimizes volume at this time in the 18 auction market. In addition, the specialist has a negative 19 obligation. The specialist may not trade for a proprietary 20 account unless reasonably necessary to maintain a fair and 21 orderly market. 22 The NYSE specialist places customers on both sides 23 of the market first. In my view, all of us in the industry; 24 brokers, dealers, market makers and specialists, should 25 adhere to that fundamental principle. 1 The industry has changed quite a bit since I first 2 came to Wall Street and not all for the better. I am 3 concerned about a number of anticustomer practices that I see 4 in the industry today. Dealers in other markets cherry pick 5 the easiest orders of the least informed customers and 6 execute those orders at prices discovered at the NYSE while 7 NYSE specialists stand ready to accept all orders of all 8 customers, including the offsetting orders of the cherry 9 pickers. 10 Agents for customers receive payment for their 11 customer orders. Agents for customers trade against orders 12 for their customers, without giving those orders a chance to 13 interact with other orders of public customers. Dealers in 14 other markets disseminate non-competitive quotes, thereby 15 shunning the responsibility to make markets for all, while 16 still guaranteeing significant order flow. An exchange buys 17 reports of off-exchange trades from OTC dealers and ECNs and 18 fraudulently prints those off exchange trades as though they 19 have actually taken place on that exchange. An ECN hides 20 quotes and its participants ignore better priced orders on 21 other markets. 22 Regaining investor confidence is crucial to the 23 continued preeminence of our secondary markets, yet we have 24 allowed to creep into our markets practices in which dealers 25 and even agents put their own interests ahead of the 1 interests of their customers. 2 The NYSE trading model precludes these 3 anti-investor practices by the very way that it has evolved. 4 We integrate the trading of institutional and retail orders 5 into a single auction and we apply automation to the extent 6 consistent with that integration. This provides investors 7 with optimal price discovery and maximum flexibility. Other 8 markets take different approaches, but whatever model a 9 market may choose, the Commission should not permit it to 10 embody those anti-investor practices -- and that requires the 11 Commission to attack the root causes of those practices. 12 Finally we believe Congress got it right in 1975 13 when it determined that transparency was the heart of 14 effective inter-market competition and when it bounded 15 intermarket competition by the principle that dealers should 16 only intervene as a last resort, thus, we support all markets 17 disseminating as much as is possible, trading interests 18 disclosed at the point of sale. We support maximizing the 19 opportunities for customer orders to interact directly within 20 a market so that the particular participants to a trade 21 receive best execution and all market participants receive 22 the benefits of optimal price discovery. 23 Let's place the customer first. I thank you for 24 this opportunity to present my views. 25 MS. NAZARETH: Thank you, Bob. Pete? 1 PROFESSOR KYLE: Thank you. I just want to make a 2 couple of points. Three points. First, computers have 3 become very commonplace in trading and we need to think about 4 what the role of computers is. On the one hand, computers 5 allow us to implement very complicated trading strategies and 6 keep track of very large amounts of information. On the 7 other hand, computers are very honest. If they do something 8 wrong you can generally catch them and they generally do 9 whatever you tell them to do and we shouldn't lose sight of 10 the fact that computers will allow us to have a better audit 11 trail and more continuity in the markets. 12 Second of all, I want to make a few points of 13 organized exchanges. Organized exchanges have sophisticated 14 members and sophisticated customers. They also have 15 unsophisticated customers who are usually retail type people. 16 The value of an organized exchange's franchise largely lies 17 with its reputation among the least sophisticated customers. 18 And the ability of an exchange to build this reputation 19 largely consists of how transparent the exchange is, so the 20 idea of building transparency is very important for organized 21 exchanges to maintain their reputation, which is their 22 capital, with the least sophisticated customers. 23 The third point I want to make, we have with the 24 national market system a ticker, and that ticker could easily 25 be augmented to have added to it various pieces of 1 information that would make the market transparent and also 2 make it easier to audit what's going on in the market. Among 3 other things, you could have a unique identifier attached to 4 a ticker that would allow it to be traced back to a 5 customer's trade so a customer can easily identify his trade 6 on the ticker. Whether you're a large customer or a small 7 customer, that type of information would be very valuable and 8 would not be all that costly to implement. 9 Thanks. 10 MS. NAZARETH: Thank you. Minder? 11 MR. CHENG: Good morning. I also would like to 12 thank the Commission for giving us the opportunity to comment 13 on these important issues that will be important not only to 14 us, but also to the market and to the investors. 15 Let me just start by saying I think that trading, 16 especially on the buy side for the large institutional asset 17 managers has evolved from pure execution to an element of -- 18 a cost element of the investment return to now account for a 19 large portion of the additional return we can provide to our 20 investors. 21 Now, with the technology and with regulatory 22 environment changes, it has opened the way for buy side 23 traders to directly access market liquidity, but what it also 24 has introduced is it created an environment where it becomes 25 a challenge for the buy side institution to find a way to 1 additional return, minimize transaction cost, but then 2 without really going through some of the hurdles that 3 certainly did not exist before. 4 And I would just highlight one major challenge, I 5 think, which I think will be discussed later, for a buy side 6 trading operation, generally speaking, the most difficulty 7 challenge is to move large blocks of orders through the 8 market over multiple days. So we're going to get into a 9 discussion of fairness and transparency in the market. From 10 our point of view, meaning for at least for a buy side shop, 11 transparency generally is a good thing, but may not be too 12 good a thing, given what we're trying to do. 13 In other words, this may certainly introduce some 14 bit of a controversy, so I was hoping to see an environment 15 where there's enough transparency, enough fairness, but 16 without compromising buy side trading's capability to get our 17 trades done without paying excessive market impact cost. 18 Thank you. 19 MR. DOMOWITZ: Thank you for inviting me here 20 today. I join with everyone else in the most interesting 21 times in market structure and possibly all history. 22 I'm the managing director of products at an 23 institutional agency brokerage, and as overseer of three 24 alternative trading systems, I obviously believe that ATS 25 systems provide value and contribute to best execution for 1 our clients. Our own system is non-price discovery in 2 nature, attests to a deep interest in a market structure that 3 promotes meaningful price discovery in the national market. 4 Now, as such, I therefore promote consolidation, 5 but I don't believe that it ought to be mandated or 6 necessarily all encompassing, and in thinking about market 7 driven solutions, I think that we all must also recognize 8 that exchanges now compete with brokers along with a variety 9 of dimensions. This complication has implications which I 10 believe should be acknowledged and I'm sure will come up here 11 today. 12 Now, as I say, ITG is an agency brokerage and we do 13 not commit capital. Nevertheless, such commitment is very 14 important and should be fostered in a competitive 15 marketplace. In today's environment, however, I don't 16 believe that capital provision need be tied to any particular 17 form of market microstructure. In other words, if immediacy 18 is indeed demanded, it will be supplied at some price under a 19 variety of conditions. 20 Finally, I view best execution which I believe is a 21 topic that will come up today, not just as a set of reporting 22 requirements, but as a process. And these are broker 23 responsibilities, by and large, and they include measurement 24 analysis and control of transaction costs on the part of 25 clients. 1 I believe in development of market structure that 2 contributes to and perhaps more importantly, may depend on 3 best execution obligations, and I hope that some of our 4 discussion focuses on that today. 5 Thank you again. 6 MS. NAZARETH: Janet. 7 MS. ANGSTADT: Good morning, I'm Janet Angstadt 8 here for Jerry Putnam. I am with Archipelago Holdings and I 9 think we're uniquely positioned to talk about market 10 structure from the ECN and the market exchange perspective. 11 We've been an ECN for five years now and an exchange since 12 March, and we've really benefited from, a lot of innovative 13 mechanisms that the SEC put in place; the order handling 14 rules that got us to the quote, the Reg ATS that gave us 15 some certainty as to our status as an ATS and also invited us 16 to become an exchange. 17 I think from our perspective, we found that in fact 18 the state of the union is very good. We've seen a lot of 19 good initiatives from the Commission, again the order 20 handling rules, Reg ATS, decimals 1-5. There are just a 21 couple of areas that we should see some attention should be 22 given, and in fact those things are happening now. I think 23 with respect to access among markets and making sure that 24 access between these competing markets is fair. If that's in 25 the ATS arena, let's make sure that access fees are in fact 1 fair between these competing markets, and with respect to 2 linkage, between the reform that the Commission has urged 3 with the ITS linkage we think that's a good area for these 4 market areas to really push to bring some innovation and some 5 technology, so we would ask the Commission to watch closely 6 what the ITS operating committee is doing there and make sure 7 there's enough momentum for that group to continue, but we're 8 pleased to be part of the debate this morning. 9 MS. NAZARETH: Matt DeSalvo? 10 MR. DeSALVO: Matt DeSalvo, Morgan Stanley. I come 11 here as a broker-dealer that represents order flow that 12 originates from both the retail and institutional side on 13 both an agency and principal basis; that transact in the 14 public marketplace on both the auction, electronic and ATS 15 level in both cash and derivatives. 16 One of the biggest structural issues facing the 17 market today is relevant inaccessibility of prices in the 18 system often driving the NBBO. The concept of fairness is 19 paramount in our mind also. We need to end up with either an 20 access standard for all markets that require quick execution 21 against split interests. We offer a flexible approach to 22 best execution, recognizing the fact that there are actual 23 and opportunity costs associated with trying to reach prices 24 in non-automated and unlinked markets. Whatever the SEC does 25 in terms of market structure at the end of the day, its 1 paramount objective should be levelling the playing field. 2 That means eliminating or rationalizing access fees. In 3 other words, this issue right now where market makers and 4 ECNs have one ability to charge and one not to charge needs 5 to be addressed. That means answering the question as to why 6 broker-dealers have to pay exorbitant rates to the monopoly 7 for market data that it needs to satisfy, -- to meet best 8 execution requirements. That means access standards and best 9 execution obligations that could be fairly and realistically 10 administered. 11 MS. NAZARETH: Scott. 12 MR. DeSANO: Good morning, thanks for the 13 opportunity to be here today. I'm Scott DeSano from Fidelity 14 Investments. Fidelity along with the fund industry generally 15 serves individual investors. Mutual funds as a whole have 16 over 6 trillion in assets under management, which it 17 reinvests for over 90 million individuals. As of October, 18 Fidelity had over 400 billion in equity assets in over 200 19 equity funds domestically. 20 We approach market structure issues with only one 21 purpose in mind and that's to better the returns for our 22 funds and in turn for the individual investors who invest in 23 our funds. For this reason, distinctions between retail 24 investors and institutional investors may not always be the 25 most useful. Through full and fair competition, investors 1 are more likely to achieve best execution for their trades, 2 but best execution is in the eye of the beholder. 3 The goal of best execution is not limited to 4 obtaining the best price. As important, if not more 5 important, is the speed and certainty of execution. This is 6 especially the case for trading the large blocks of stock 7 given the volatility of the markets and the demands of 8 liquidity involved in trading large blocks of stock. 9 Liquidity is very key. 10 Market centers should disclose their entire limit 11 order book. The securities markets have begun to make 12 improvements in this area with Super Montage and the New 13 York's Open Book, but these steps fall short of full and 14 timely transparency. Disclosing the size and price levels of 15 orders on the limit order book will lead to greater liquidity 16 and depth of trading. 17 Mutual funds and other institutional investors will 18 add to liquidity by placing their orders on the limit order 19 books if strict time and price priority are given to those 20 orders over orders entered later on the book or made in the 21 trading platform. The lack of strict time and price 22 priorities effectively deters institutional investors 23 including Fidelity from placing limit orders on the 24 specialist books. This contributes to fragmentation and a 25 lack of transparency in the market and detracts from the 1 depths and liquidity in the trading of stocks. The lack of 2 strict time and price of trading of orders has undermined the 3 appeal of the New York institutional express system for the 4 trading of large blocks of securities. The practice of being 5 pennied by traders on the Floor adds to the disincentives 6 against placing limit orders on the specialist books. 7 Decimalization was intended to foster better 8 execution for customer trades by narrowing spreads and adding 9 liquidity to the markets. The unintended consequence of 10 decimalization has in fact made best execution problematic 11 for investors. Traders and specialists are now able to jump 12 ahead of earlier entered customer orders with much less 13 downside risk. They can offer minimal price improvement of 14 one cent to the ordering sides of the trade with the 15 knowledge that if the market turns against them they are 16 likely to lay up their position with little loss while able 17 to capture far greater upside gain. This forced us as an 18 institution to trade predominantly upstairs or through ECNs. 19 It is open to serious question whether on balance investors 20 obtain better execution within an environment that provides 21 floor traders with heightened trading advantages over 22 customer limit orders. 23 Thank you. 24 MS. NAZARETH: Tom Peterffy. 25 MR. PETERFFY: I'm Tom Peterffy, Chairman of 1 Interact Group. We are an electronic broker-dealer who 2 provide liquidity and market makers and execute for our 3 direct access profession accounts. We do this fully 4 electronically in Asia and Europe and to the extent possible 5 in the United States. 6 We appreciate this opportunity to tell you our 7 views. We think that broker-dealers should be able to 8 interact with electronically accessible limit orders. We 9 should provide strong economic incentives to traditional 10 exchanges throughout the world. Approximately 100 of our 11 employees work as market makers on the floors of the four 12 traditional option exchanges and two oversee our market 13 making activities on the fully electronic ISE option 14 exchange. These two people transact more volume every day 15 than the other 100 combined. We need electronically 16 accessible limit order books with automated matching to 17 efficiently fulfill our duty of best execution and provide 18 access to the markets. Thank you. 19 MR. GARDNER: I'm Tom Gardner, co-founder of Motley 20 Fool. I'd like to thank the Commission for including me and 21 the Fool today in the proceedings. 22 We are an organization that reaches more than 20 23 million investors each month through our on line service, 24 newspaper column and our radio program on NPR. I am not 25 someone who works at a firm that handles orders and we are 1 not an organization that speaks to active traders that are 2 using many of the trading systems that we'll speak of today, 3 so I am largely irrelevant to today's proceedings and in the 4 interests of full disclosure I want to get that out there. 5 I can speak about general philosophy, and that is 6 that we believe that healthy markets, a rise in liquidity, in 7 trading activity rely on equal access to material information 8 and presently available prices in the national market system 9 for all investors, regardless of the size of their account 10 and regardless of the size of their order. We do not believe 11 that distinctions between retail institutional investors are 12 useful or should be relevant to the work, those regulating 13 the marketplace. However, we should not lose sight of the 14 fact that our markets in large part because of the SEC over 15 the last decade are increasingly transparent, efficient, more 16 competitive and more accessible to all investors, but 17 progress requires vigilance. At this point, henceforth, 18 today I am the quintessential proxy for Dr. Kyle's least 19 sophisticated customer. 20 MS. NAZARETH: And proud of it. Paul. 21 MR. O'KELLY: Thank you, Annette. First of all, 22 I'd like to commend the SEC for undertaking this review of 23 market structure and thank them for allowing the Chicago 24 Stock Exchange to participate today. 25 More than 25 years ago, Congress directed the SEC 1 to facilitate the creation of a national market system. When 2 it met the objectives that Congress identified as important 3 for protection of public investors, one built on a set of 4 principles that Congress also identified in the 1975 Act 5 amendments. To those who listened to the market structure 6 hearings in Washington a couple of weeks back, you heard 7 Annette ask the various participants whether those principles 8 continue to be relevant today when determining how to 9 structure a market. It was one of the few areas where there 10 was a consensus among the participants. 11 I mention this fact, because when the Commission 12 market participants back in 1975 built the national market 13 structure system, it contained two principal attributes. One 14 was the consolidated mandatory display of quote and trade 15 information to promote market transparency, and the other was 16 intermarket access to prices which promote the best execution 17 of orders. I think that if the Commission again examines 18 market structure in light of these principles, it will reach 19 the same conclusion, and the market will look as it does 20 today at the current national market structure. The 21 challenge for the Commission is not to find an alternative 22 market structure, the challenge for the Commission is going 23 to be to try to bring new participants with new trading 24 models into the existing market structure in a manner which 25 both promotes competition, promotes investor confidence in 1 the market and is fair to the other market participants. 2 MS. NAZARETH: Thank you, Paul. Bob? 3 MR. STEEL: My name is Bob Steel of Goldman Sachs 4 and I thank the Commission for the opportunity to be here 5 today and I assure Tom he's not the only one who feels 6 foolish on a regular occasion. I think a lot of what I was 7 going to say by introduction has been said so maybe I can 8 shorten the comments and say from our perspective at our firm 9 we see the equities marketplaces through multiple lenses and 10 from multiple perspectives. We act as a traditional 11 brokerage firm handling orders for clients. We are a major 12 block trading firm often using our own capital to provide 13 liquidity to institutions and issuers. We make markets in 14 over 5,000 over the counter stocks. One of our affiliates is 15 the second largest specialist on the floor of the New York 16 Stock Exchange. Another affiliate provides sophisticated 17 electronic order technology to institutions and broker- 18 dealers. We are among the largest options market makers and 19 we also own a substantial minority interest in Archipelago, 20 an alternative trading system. 21 In summary, from our vantage point decimalization, 22 technological advances in order handling execution and the 23 conversions of roles of various market participants have 24 exposed frictions in the cash and derivative markets that 25 should be addressed. Examples that have been subject to 1 recent debate include ECN access fees, the use and submission 2 of market data revenues and the effectiveness of market 3 linkages particularly the connection between auction and 4 electronic markets. Although extensive Commission 5 involvement in structured markets is often unnecessary and 6 potentially harmful due to the swift and dynamic nature of 7 change, we believe in this case it is incumbent upon the 8 Commission not only to remedy some of the existing ailments 9 in the U.S. equity markets, but also put a forward looking 10 framework in place that enables the rules and macro 11 structures of markets to evolve as the underlying 12 participants both compete and develop. Thank you. 13 MS. NAZARETH: Gunner? 14 MR. BURKHART: My name is Gunner Burkhart. I'm the 15 global head of trading at Deutsche Asset Management based in 16 London. We are an 11th hour addition to this gathering and 17 we appreciate being invited to it. We manage about $800 18 billion in assets in all the asset classes around the world 19 and our constituents are retail investors, be they in the 20 United States, Asia, Europe, as well as corporate pension 21 funds, institutional funds, et cetera. We think the timing 22 of these discussions, both in October and today and hopefully 23 the ongoing discussions, is highly relevant. We are 24 currently engaged in similar discussions with stock exchange 25 or Bourses, as they're known in Europe, and also in Europe 1 with regulators in Europe and Asia as well -- Given the 2 turbulence in our industry collectively as well as in the 3 world, there's certainly been no better any time my twenty- 4 year career to sit down with the main and major participants 5 and constituents in our industry to discuss these topics. 6 Our concerns around the world and certainly very 7 relevant here in the United States boil down into three main 8 categories and I echo some of my competitors Scott and Minder 9 this respect. They are liquidity, efficiency and 10 transparency, and I hope and I am sure that throughout the 11 remainder of the day we will enter into fairly lively 12 discussion about how we attain and regulate all three of 13 these, not just for the institutional asset managers in the 14 world, but for the investor on the street. That's not just 15 Main Street, USA but also Main Street, Frankfurt or Main 16 Street, Taipei, for that matter, who are also very large 17 investors in the United States. 18 Thank you for allowing us to participate. 19 MR. LEE: Good morning, my name is Ruben Lee and 20 I'm founder and managing director of the Oxford Finance 21 Group. May I echo everybody's previous statements that it's 22 a pleasure and honor to be here and thank you very much, and 23 may I also say that I think that these issues are very, very 24 important, notwithstanding the fact that some other issues in 25 the private markets seem to be dominating the newspapers, I 1 think these are very important and should not be eclipsed by 2 them. 3 The Oxford Finance Group is a research, consulting 4 and implementation firm and I do all of three of those 5 things. I have clients in special financial organizations, 6 government regulators, exchanges; a whole series of 7 commercial activities. I'm also an advisor to the European 8 Parliament on financial services. 9 Notwithstanding all of that, I view my role here 10 today to be to propose to the Commission to further its own 11 goals, namely, to look at the American securities markets, to 12 look at those issues which are most important, to find 13 optimal answers and most relevantly, to find answers that are 14 not so radical that they can't be considered, but yet are 15 still not so conservative that they should not be considered. 16 So in particular, I'm interested in the feasibility of 17 different types of questions, and by that I mean, the extent 18 to which different sorts of solutions may be feasible within 19 the current legal structure, so I don't think we should seek 20 to change the current legal structure, but I think that there 21 is much that is open to interpretation, that many people do 22 not yet appreciate. 23 I also am interested in the feasibility of 24 solutions, that is to say, the political feasibility, within 25 the market and I'm also interested in the extent to which 1 current regulatory structures are, in my view, in the 2 American context, captured by a range of different political 3 interests, which restrict change. We will be looking at a 4 whole series of issues to do with transparency, best 5 executions, exchange regulations, and I would be happy to 6 participate in that. Thank you again. 7 MS. NAZARETH: I think it's clear from these 8 participants we have an incredibly knowledgeable group and 9 clearly everyone comes with their own perspective, but that's 10 exactly why you're here. We're interested in carrying forth 11 with these concepts and you'll see that well over the course 12 of the day. 13 As Paul O'Kelly mentioned, we're going to briefly, 14 as we did in the last market structure hearing, discuss the 15 basic principles of the national market system and see 16 whether, again, although no one should feel compelled, 17 whether again we have a broad consensus that those core 18 principles that characterize our national market system 19 remain valid today. 20 Let me go over again briefly what they are. The 21 first is transparency of information. That is, the 22 availability to brokers, dealers and investors of information 23 with respect to quotes and trades. The next is economically 24 efficient execution of securities transactions; ease of 25 obtaining best execution; fair competition among markets and 1 intermediaries within marks and lastly, the opportunity 2 consistent with the aforementioned goals for investor orders 3 to be executed without the participation of a dealer. 4 So again I'd like to ask the group before we do go 5 on, because it's important to know whether we're all working 6 off the same principles, whether these are the right 7 precepts, and where there's conflicts, how those conflicts 8 should be resolved. It would be helpful to establish, again, 9 before we go ahead, whether there's consensus on these 10 principles. 11 This is somewhat theoretical, although obviously it 12 has very pragmatic implications for our discussion today, but 13 since it's a somewhat academic excise to begin, I thought I'd 14 begin by calling on one of our academics, it's sort of like 15 getting back at some of these academics like Professor 16 Goldschmid who called on us in Columbia, but I thought I 17 would ask Professor Kyle if you have a view that these are 18 the correct goals. 19 PROFESSOR KYLE: I think they're the correct goals, 20 except I want to think a little bit about what we mean by 21 "fair competition." To some people, fair competition means 22 if the other guy is more efficient than I am, then the other 23 guy should be punished so he and I have equal market share 24 and equal equilibrium. To others fair competition means the 25 rules of the game should be the same for everybody so the 1 person who provides the most efficient service captures the 2 greater market share. To me it means the latter, that is 3 the more efficient should capture the larger market share. 4 MS. NAZARETH: Tom? What do you think? 5 MR. PETERFFY: Well, legally, these are the correct 6 principles and we have made a lot of progress towards them in 7 the last three or four years, especially decimalization was a 8 very large step. These objectives are completely within our 9 reach, if only we could just get along. 10 With technology and sufficient economic incentives, 11 I think that the participants and the Commission can achieve 12 them in a very short period of time. 13 MS. NAZARETH: Bob Steel? 14 MR. STEEL: I was just laughing. I'm sure "get 15 along" in Tom's vocabulary means we do it his way. 16 MR. PETERFFY: Or yours. 17 MR. STEEL: I've got lots of ways. 18 You know, Annette, I think that it's probably 19 pretty easy to pledge allegiance to these ideals, but really 20 the challenge is the nuance at the edges and we heard a 21 dichotomy, I think on this side of the table, about 22 transparency is a good ideal, but at some point too much 23 transparency carries with it a burden that people who are 24 representing large numbers of individuals with a larger order 25 feel disadvantaged and so I think the real issue is threading 1 the needle on these three or four issues with greater 2 specificity and we can all kind of, as I said, pledge 3 allegiance to the ideals, but then it gets down to really 4 determining how you get down to this dynamic of Bob's 5 complete transparency and Scott and Minder's idea of 6 transparency which has a different idea and cost. I think 7 this is the tricky part, especially in the area of 8 transparency. 9 MS. NAZARETH: You alluded to this also, Ruben that 10 to some extent it's in the eye of the beholder and people's 11 obviously own business models or political inclinations can 12 factor into this. Do you have more comments on this? 13 MR. LEE: I have a couple of comments; one on 14 transparency, one on fair competition. On transparency, I 15 would make three comments. The first is that what you're 16 getting now is not what you think you're getting, so you 17 think that you're getting the best bid and offer, but you're 18 not actually getting that, but most people don't appreciate 19 that, I think in the American context retail. The second is 20 that I very much agree and I think there's a fair amount of 21 academic evidence to support Minder's viewpoint that optimal 22 transparency is not full transparency. There are a range of 23 different tradeoffs, but in my mind, I'm very much on that 24 point, and we can go into the subtleties of that. 25 The third is and this is going back to the '75 1 amendments which is very important, is that, Pete alluded to 2 this, which is that market structures of many different types 3 have strong incentives to provide transparency, number one. 4 The second is that even where there are no mandated 5 links between markets, one can point to Europe, we have seen 6 consolidators of information, so the idea that we need to 7 have some sort of institutional consolidator of information 8 here in the U.S. I think is much less relevant than it was 9 when it was initiated and in part because it's so much 10 cheaper to do that. So I think one needs to be careful about 11 requiring a need for mandatory transparency when there's so 12 much private incentive to produce appropriate levels. 13 Of course there are exceptions to that, where 14 transparency must be structured and mandated in some way, but 15 I view that to be an exception. 16 Let me now make a comment on the fair competition. 17 In my view, in most contexts, there are very strong 18 incentives for orders to centralize on particular trading 19 systems and in fact what we see here in America is actually 20 very unusual, that we have such an abundance of trading 21 systems, notwithstanding the fact that they are shrinking in 22 number and in part I think that abundance does arise as a 23 result of various SEC rules which allowed the continuation of 24 a range of different systems which in fact are being 25 subsidized in various ways by these rules. 1 I think there's a counter balance to that, which is 2 if you think as I do that orders may tend to consolidate that 3 where you have institutions such as the New York Stock 4 Exchange where they do consolidate, I think there are very 5 good grounds for examining the actions of such monopolistic 6 providers very, very carefully and a whole range of different 7 aspects and while in many contexts it is appropriate to say, 8 yes, leave people to decide what sort of market structure 9 they want to adopt and I agree with that, once we start 10 approaching the monopolistic provider, I think there we need 11 to be much more careful. 12 MS. GLASSMAN: I have a question for anybody. 13 Given, there seems to be general agreement on what the ideals 14 are, perhaps not the specifics, can we meet all of these 15 ideals simultaneously, are there tradeoffs we have to make 16 among them? 17 MR. CHENG: Maybe can I start? I think the 18 question, maybe if I can rephrase it, is from the 19 Commission's point of view, are you in a position to design 20 or promote a market structure that basically is one size fits 21 all. The question we have here is the market is full of 22 retail and institutional investors. Going back I think to 23 the five principles behind the NMS, the third being the ease 24 of obtaining best execution, we already heard earlier that 25 best execution means different things to different people. 1 To the retail investors it may mean the best price, 2 but for institutional investors, it may not mean the best 3 price at the moment of trade, because again, going back to 4 the fact that we often trade large blocks over time, so the 5 question is, how do we create the right balance between the 6 two? And back to the transparency. I think we can achieve 7 it, although the question in our current market, I think 8 Scott mentioned as well, that transparency cannot be achieved 9 without protection, because if there's no protection, in 10 other words, there's no, say, price, time, size priority, 11 then I think the market participants will not be incentivized 12 to really show their hands so they could get picked up. 13 So I think on the one hand we are basically dealing 14 with competing goals here. The five elements are all good on 15 paper. The question is, can we somehow find a way to piece 16 them together without conflict. Thank you. 17 MR. DeSANO: I just wanted to add, I don't believe 18 there's a perfect market structure, period. And that we need 19 to think about that as we think about all these things. 20 There's a little give-up on everybody's part because of all 21 the constituencies. Competition and fairness lends itself to 22 continuous improvement in the market structures, but we'll 23 never get to a perfect structure. 24 MR. DeSALVO: I think what would help some of these 25 issues that we're talking about, would be from our point of 1 view a clear definition of best execution. Fragmentation and 2 competition has drawn away the retail orders into shades of 3 gray with regards to best execution away from price. Market 4 accessibility between, where we have auto ex in certain 5 marketplaces and not other marketplaces have led even retail 6 order flow to be subject to shades of gray when it should 7 clearly be price subject to best execution at a certain 8 level. We don't even have that in this case and what we 9 would like to see coming out of this is a clear definition of 10 best execution with getting back to price and the only way 11 that you can get back to price with regards to best execution 12 is markets that efficiently transact between each other as 13 opposed to the current market structure where there is 14 inefficient linkages and inefficient ways that markets access 15 each other. 16 MR. DOMOWITZ: Getting at the original question, 17 which is whether or not you can satisfy all these at once, I 18 mean, there's a small preamble to that. We can all agree, 19 I'm with Bob, it's easy to pledge allegiance. This is like 20 motherhood. Frankly, it's easy to agree. Following up on 21 Ruben, if we didn't agree, we'd have to change legislation. 22 This is not something we can actually agree on at this table 23 by ourselves in some sense, but it's worth saying, first of 24 all, market structure is a term used by a variety of people 25 to support a variety of positions. It is essentially in my 1 view, the rules and institutions that contribute to 2 competition in the marketplace for transaction services. 3 Now, viewed that way, what's feasible now, all 4 right, again, following up on something Ruben said, is indeed 5 I think the reconciliation of these five points. Should we 6 indeed get a little bit deeper to these rules that are 7 promoting competition in the first place and to me that is 8 the right place to start. 9 Absent that, were I to go back to the legislation, 10 I would actually turn to this point about best execution and 11 I would say let us create best execution obligations and the 12 rest will follow. 13 MS. NAZARETH: Janet? 14 MS. ANGSTADT: I was going to say, I think all 15 these questions go to best execution. I would urge that we 16 not try to define best execution any more than we have. I 17 think the 1-5 rules have really offered the disclosure that's 18 necessary for people to make their choices. Again, too, I 19 think it was Minder's point that best execution to an 20 institution may be very different than a retail investor and 21 perhaps some of the suitability rules that we have are 22 already in place to help protect retail investors and I think 23 the diversity that we have in the market structure today, 24 really allows people to choose speed over price if that's 25 what's important, so I think actually these principles help 1 lead us to that kind of a best execution very strongly. 2 MR. DeSALVO: I hate to go back and repeat this, 3 but again, we don't even have a situation right now where we 4 have an agreement on smaller, for the retail institutions, of 5 what is best execution, because they're opposing views of 6 speed or price. I think if the Commission can come out and 7 say what is the definitive goal, at least that would allow 8 all of us to be on the same playing field rather than the 9 current shade of gray. 10 MS. NAZARETH: Well, I think, as you know, Matt, 11 that the analysis is so colored by other principal/agency 12 conflicts, like payment for order flow and the like, that 13 it's not as simple as just speed or price. But we'll get 14 into that later. 15 Does anybody else have any points before we take a 16 break? Yes, Paul. 17 MR. O'KELLY: I'd like to make one point. While 18 these principles are very noble, if they suffer from 19 anything, they suffer from a little vagueness. But if we 20 look at the '75 Act amendments where these principles were 21 articulated, you'll see in the very next section where 22 Congress makes the point that markets that are linked 23 together can promote these five principles. Now, how we 24 interact with each other is a very difficult issue to 25 address, but it seems clear to me that Congress back in 1975 1 didn't like the idea of market fragmentation, didn't think 2 that it would promote these five principles and I have no 3 reason to believe that that's changed in the subsequent 27 4 years. 5 MS. NAZARETH: Thank you. Yes? 6 MR. MURPHY: I'd just like to, a couple of points. 7 As far as best execution, I think as long as the ultimate 8 customer is at the forefront and his outlook is what's 9 paramount, the markets in general will acquiesce and should 10 be encouraged to just keep the ultimate customer primary, 11 primary concern. 12 When it basically gets to a point where best 13 execution is defined by economic terms or being able to take 14 advantage of a situation, that's where we get to that gray 15 area that Matt talked about, and right now, the 1-5 data that 16 we've seen so far I think is pretty skewed and it's not 17 talking about apples and oranges. I think we can do a lot 18 better or maybe get some more independent input on best 19 execution statistics, so you know, I basically agree with all 20 these principles, but giving the customer choices, because 21 definitely, there are different venues that benefit the 22 different participants in the markets, so as long as there 23 are choices and they're transparent choices, we will do okay. 24 MS. NAZARETH: Yes? 25 MR. LEE: Yes, I've got one more point. To be very 1 explicit, I think that there are tradeoffs between these five 2 different goals, and I think that the choices that the 3 Commission has made have implicitly showed their choices for 4 these different goals. So in particular, if you look at the 5 fifth one, which is investor access, subject to the first 6 four, one could debate whether that one has been promoted 7 given membership structures of exchanges and so on and so 8 forth, and in fact one could strongly argue that it has not 9 been promoted, so these goals have not been equal in the eyes 10 of the Commission historically. 11 MS. NAZARETH: It sounds like yet again we have 12 violent agreement on the principles, and we'll spend the rest 13 of the day arguing over the details. 14 We'll take a fifteen-minute break and reconvene at 15 five of eleven. 16 (Brief recess.) 17 MR. COLBY: Welcome back, everyone. We're going to 18 turn to the issue of market data as one of the fundamental 19 building blocks. I want to make sure everybody talks into 20 the mike, because the audience has had trouble hearing. 21 We also have two players replace people who were 22 injured in the prior panel. Duncan Neiderauer is replacing 23 Bob Steel. Actually, he wanted to do it, but Duncan pushed 24 him out of the way. 25 U.S. investors today have access to a consolidated 1 real-time stream of market information for each of the 2 thousands of stocks traded in the U.S. markets. This price 3 transparency is a cornerstone of the U.S. national market 4 system. It facilitates the best execution of customer 5 orders, promotes investor protection and mitigates the 6 fragmentation of buying and selling interest among different 7 market centers. The most basic form of market information is 8 the best quotation and last sale data with respect to a 9 particular security at a given time. The best quotation is 10 the highest bid and lowest offer price for a security 11 currently made available by self-regulatory organization and 12 is commonly referred to as a national best bid and offer or 13 the NBBO. Last sale data generally identifies the price at 14 which the most recent trade in a particular security 15 occurred, the size of that trade, and the market in which the 16 trade took place. 17 Price transparency in the United States, as we 18 discussed earlier, was mandated by Congress in 1975. 19 Congress directed the SEC to use its new authority to adopt 20 rules to integrate the market through the collection and 21 consolidation of market information. Under these Commission 22 rules, quotes, trade price and volumes are collected by the 23 stock exchanges and NASD for all exchange listed and NASDAQ 24 stocks within seconds of the quote or the trade taking place. 25 The current NBBO quotation, as well as individual market 1 center, over the counter market maker and ECN quotes and 2 reports of all trades are disseminated for each stock on a 3 nearly real-time basis. Vendors and brokers that make this 4 market information available from--vendors that make it 5 available from any individual market are required by 6 Commission rules to make equally available consolidated 7 information from all participating markets. 8 Furthermore, to implement this quote collection and 9 trade reporting mechanism, the SRO's, the self-regulatory 10 organizations, have acted jointly under several national 11 market system plans to disseminate consolidated market 12 information. These plans govern all aspects of the 13 arrangements for collecting and distributing market 14 information. Among other things, they require the individual 15 SRO's to transmit market information to a central processor, 16 which then consolidates the information into a single stream 17 for dissemination of vendors and some large end users. In 18 turn, the vendors, brokers, and large end users may 19 disseminate the information to the public. 20 These plans also establish fees for market data. 21 The resulting net revenues are divided among the 22 self-regulatory organizations that provide the data and 23 represent a significant percentage of the SRO's revenue. The 24 plans also establish non-fee standards such as reporting 25 obligations, hours of operation and regulatory halt 1 procedures. 2 The implementation of decimal pricing and point 3 increments in 2001 has increased the demand for deeper or 4 different levels of market information, as was touched on 5 earlier in the discussion. For example, the New York Stock 6 Exchange recently began to provide real-time semi-dynamically 7 updated view of the limit order books for all New York Stock 8 Exchange traded issues. NASDAQ Super Montage, which began to 9 operate on October 14th, provides NASDAQ stand alone 10 quotation and ordering system by a means of aggregating 11 quotes at a multiple of price levels to create an integrated 12 trading system. Some ECNs such as Island and Archipelago 13 make their limit order books available on their websites at 14 no charge and some Internet portals also make this 15 information available for free for just users on the web. 16 Nonetheless, despite these steps taken by the 17 markets, the impact of decimalization led some market 18 participants, as was evidenced earlier in the discussion, to 19 believe that more limit order information is needed in 20 standard formats that can be integrated. There have also 21 been repeated questions raised regarding the market data fees 22 that are charged by the consolidated systems disseminators. 23 In particular, the fact that market centers are rebating or 24 seeking to rebate a significant percentage of these market 25 data revenues to their members, has led some to suggest the 1 market data fees are too high, and could be reduced 2 substantially. 3 In light of the wide range of views and comments 4 that are received on a concept release that was put out by 5 the Commission on market data revenues, the SEC formed an 6 advisory committee in August, 2000, to assist it in 7 evaluating issues relating to market value data available. 8 In a final report in September of 2001, the committee 9 reaffirmed the importance of price transparency and 10 consolidated information, the importance of these for the 11 health and vitality of the securities markets and then it 12 recommended with varying degrees of consensus a number of 13 measures to increase flexibility in market centers and data 14 vendors to make data available to the public. 15 It also suggested that technological and 16 competitive developments may have lessened the need for 17 market centers to act jointly to consolidate data under 18 existing national market system plans. 19 So the purpose of this session is to consider the 20 ways that market data, in particular real-time market data, 21 is collected, processed, and made available and the many 22 questions that have arisen regarding this data. 23 These include the utility of the data for 24 investors, and we need to talk about for which investors, the 25 way that the data is made available, the fees that are 1 charged and inherent in talking about the fees that are 2 charged, I think is the topic of how self-regulatory 3 organizations get funded, even though that's a major issue 4 that comes up later in the day. 5 So, with that information and introduction, let's 6 start in on the questions. I wanted to start first in 7 talking about the information that's available, is it useful, 8 what is its nature, is it useful, then go to should other 9 information be available and then after we have talked about 10 that, start talking about fees, if that's all right with the 11 panel. Fees are going to be an interesting topic in and of 12 itself. Let's start with the national best bid or offer. 13 The national best bid or offer is in a sense a product built 14 and maintained for retail investors, and so I think, first 15 question should be how useful is it? How important is it to 16 have some sort of a measure of the existing quote and how 17 good a measure is the NBBO? Anyone want to address this from 18 a retail investor standpoint? Ruben, you expressed some 19 views about the nature and utility of the NBBO. Do you want 20 to start? 21 MR. LEE: I think several points. The first is, I 22 agree with Pete's point, that if you are relatively 23 uninformed, that as much needs to be done, that should be a 24 focus of the Commission and it clearly is not, given your 25 historical mandate to investors out there. So retail 1 investors should be a primary concern. 2 The best bid and offer does play a very, very 3 important role in showing what may be available, but I think 4 that it's, as I indicated beforehand, it doesn't show the 5 full bids and offers that are available throughout the 6 market, and there's a question in my mind, both as to whether 7 it could do that or whether it should do that. That's the 8 first point I'd like to make. 9 The second is that, and this comes to the issue of 10 competition between different markets, and both as trading 11 institutions and in terms of their ability to gain revenues 12 from data. 13 If you considered the value of data from competing 14 markets, you would probably say that the New York Stock 15 Exchange data is much, much more valuable than anybody 16 else's, and on that basis-- 17 MR. COLBY: You're speaking about quotes or trades 18 as well? 19 MR. LEE: Let's talk about quotes initially. 20 On that basis, there's a question as to whether the 21 SEC's rule, which says that if you carry any particular data, 22 you need to carry all data, okay, is necessarily appropriate, 23 indeed, whether it follows from the Act, that's to say, you 24 might choose, even if you decided you accepted that 25 transparency was an appropriate goal because you're mandated 1 to, whether you needed the rule to deliver that goal, and I 2 think there are questions about that, which lead to questions 3 specifically about the options market, where, and I know that 4 you're familiar with this, where there's an excess of data 5 with all these different dates and so on and so forth. There 6 the vendors are complaining that you may mandate it because 7 you have a regulatory interest in getting that data, but 8 actually the cost of getting all that piping needs to be 9 borne in mind. 10 MR. COLBY: Several issues raised and they're all 11 important. Pete, did you want to address this NBBO issue? 12 PROFESSOR KYLE: If you go back to the five points 13 from the previous session; transparency, economic access to 14 the market, best execution, fair competition and facilitating 15 investors order crossing, it seems like having a national 16 best bid best offer is necessary to achieve at least four of 17 those, because the basic idea is that you want a small 18 investor who closes an order with one securities firm to buy 19 at 4, another investor places an order in another securities 20 firm to sell at 3, or sell three or four. A transaction 21 ought to occur and at a very minimum in order to achieve four 22 of those objectives, you need some mechanism for keeping 23 track of the fact that a transaction ought to be occurring, 24 and that requires having national best bid/best offer. 25 Now, if you want it to be set up in such a way that 1 an economically relevant sized transaction can also be 2 monitored, then you want to show a certain amount of depth to 3 transactions, because it's possible the best bid might be for 4 a tiny quantity. Even for a retail investor it might be 5 relatively smaller than his order so you want to show at 6 least enough depth to show a reasonable sized quantity, be 7 able to be monitored with respect to the prices offered in 8 the market. So it seems to be a requirement that that be the 9 case and probably a requirement that it be augmented enough 10 to have relevant sized quantities in it. 11 MR. COLBY: Bob, would you like to talk about what 12 the New York Stock Exchange is doing to respond to that? 13 MR. MURPHY: Sure, just how I feel about it, and 14 again, there is a little bit of a split at the stock 15 exchange, but I do feel that there is a need for the NBBO, 16 and it's -- definitely in a decimal environment it's been 17 diminished, but, you know, the issue of not needing it any 18 more, it's almost we're going too far the other way. We 19 would be taking transparency out of the market at the expense 20 of the little guy, so I think we definitely need it. 21 But depth, as the Professor said, is paramount. Now 22 because of decimals and we have our Open Book, that's been 23 very well received as a data product. I think we have over 24 7500 subscribers now since it's been launched, and we also 25 have, I believe before the Commission, Dual Quote, which will 1 include NBBO, include the Open Book, and will also force the 2 specialist to put up a size held market that all participants 3 can interact with as need be. So-- 4 MR. COLBY: Meaning that it's a market that's below 5 the NBBO or above the NBBO where there's a greater size 6 available? 7 MR. MURPHY: Absolutely, and should and must be 8 disseminated to all participants. So you know, that's our 9 answer to the lack of transparency in a decimal environment 10 and I think the more--that also, it includes the crowd, it 11 includes all orders entrusted to the specialist and the 12 specialist's interests as well, to give institutions a better 13 look at what kind of liquidity is down there. 14 MR. COLBY: Matt, you have trading prices and a 15 large retail brokerage operation. What role does the NBBO 16 play? 17 MR. DeSALVO: I think the NBBO has been diluted 18 with the introduction of decimals. It had much more 19 importance at the steep level than it does at the penny 20 level. The introduction of Super Montage and Dual Quotes and 21 Open Book I think are representations that the marketplace 22 recognizes that the NBBO needs to add some depth, so you've 23 had exchanges respond to it. I think the important thing 24 from our point of view is that the NBBO is the basis point 25 with which you start making some of your decisions, but 1 accessing the NBBO I think is where we come out on whether it 2 is important or not. 3 In other words, an NBBO that is represented by a 4 market that is easily and fairly accessible is meaningful. 5 An NBBO that is representative of underlying markets that are 6 difficult to get to, slow in response time, et cetera, where 7 you start to grade away from an auto-ex environment, leads to 8 NBBO's that are different. An NBBO in one stock versus an 9 NBBO in another stock with equal one penny bid offer spreads 10 can have entirely different meanings with regards to response 11 time and accessibility, so I think that a marketplace that 12 has an NBBO as a minimum requirement for market data 13 dissemination you have to also step back, the NBBO goes up in 14 importance if there's equal accessibility between markets. 15 MR. COLBY: So the accessibility is a key question 16 with the NBBO. We're going to leave actual detailed 17 questions about access to the later session. But I note the 18 point if you have an NBBO, and it can't be accessed, it's a 19 very difficult question. 20 MR. COLBY: Could I ask Minder what role does the 21 NBBO play and what do you look at, I'm going to ask Gunner to 22 address the same thing. 23 MR. BURKHART: I think we probably have the same 24 answer, I wouldn't want to speak for Minder, but the NBBO to 25 us is nothing more than a guideline or an indication. Matt's 1 point is well taken, in that it's the depth and the 2 accessibility that's the real issue for institutional 3 investors. You have to have an NBBO I think for the retail 4 world out there, and even for us to have a guideline that 5 says that even if it's a hundred share market on the bid and 6 the offer, that's the best bid and best offer out there. 7 It's really the depth beyond that and the accessibility, 8 whether it's a hundred share market or the hundred thousand 9 share market which Bob is referring to, the capital 10 commitment market outside of that, the specialist is now 11 willing to put up, it's of more interest to us. 12 Yes, the NBBO is an essential item. However, it's 13 largely irrelevant for an institutional investor. Sorry, 14 Minder, I didn't want to speak for you. 15 MR. CHENG: I think we're in competition across the 16 table but I think our views pretty often are similar so no 17 further comment. 18 MR. LEE: Could I make one comment about the NBBO? 19 I'm wondering what is the incentive for markets not to 20 provide their quotes. If there are incentives not to provide 21 their quotes, to what extent do you guys need to be involved? 22 That's what I'm wondering. Not to put the questions to the 23 listeners today, but-- 24 MR. COLBY: I would just say that we went through 25 an era where we needed to mandate quotes to be made available 1 because there are mixed incentives from different trading 2 systems. It's worth, can I pick up on one of your--unless 3 someone wants to rise to that, could I pick up on one of the 4 questions that you posed, Ruben, which is that we currently 5 have at SEC a rule which says that if vendors or brokers give 6 information about one individual market to customers, they 7 must provide equally the information from all markets that 8 are participants in the NBBO. There were a number of 9 historical reasons for that. You posed the question whether 10 it's still necessary. Recognizing that some markets have the 11 larger percentage of the best quote than others. 12 Does anyone, Paul, is this something you would be 13 interested in talking about? 14 MR. O'KELLY: Well, the requirement for the Vendor 15 Display Rule is one that I think it's important from public 16 investor protection reasons, because if you do not require 17 all folks to be part of a consolidated quote and vendor 18 display rule, some market participants will only buy a New 19 York quote or only buy the NASDAQ quote. And that's going to 20 have a spiraling effect over time. If fewer people buy other 21 market competitors quotes, market competitors are going to be 22 less able to provide markets that compete with the markets 23 who are selling their quotes and eventually you're going to 24 lead to one, a single market selling its quotes and there 25 will be no alternatives. 1 If you do not allow other markets to display their 2 quotes, even a consolidated quote, to give customers the 3 opportunity to choose among them, you're going eventually 4 lead to a single both on the listed side and NASDAQ side. I 5 don't believe people are anxious to have that. 6 MR. COLBY: So there's a limited ability to get 7 your quotes disseminated. I'm done. 8 MR. O'KELLY: You earlier talked about tradeoffs in 9 these principles that Annette articulated and in this NBBO 10 area, I think this is one of the areas where a trade-off 11 occurs, is that you have the NBBO where people have to act 12 jointly, because we believe price transparency is a good 13 thing, it will make everyone join together to provide 14 transparency to everyone for the NBBO. Beyond the NBBO there 15 can be a relaxation of that requirement and allow markets to 16 compete with each other. If the New York Stock Exchange 17 wants to provide its open order book to its customers because 18 they believe it will give them a competitive edge, they 19 should be free to do so. If NASDAQ wants the Super Montage 20 to show the market three quotes deep, they should be free to 21 do so. 22 This is an area where I think the Commission has 23 acted responsibly in allowing the goals of transparency and 24 competition to co-exist, requiring vendor display, 25 consolidated NBBO, on the other hand, allowing markets to act 1 on their own competitively to sell products beyond the NBBO. 2 MR. LEE: Can I take issue with that, in the 3 following sense, that if I'm an investor, clearly I want to 4 know where I can execute my trades. Because I want to know, 5 someone's going to fulfill that demand for me. A vendor is 6 going to package together exchange X, exchange Y and exchange 7 Z. Suppose that the vendor decides that, or rather, even the 8 investor decides that actually exchange Z is not really worth 9 knowing about because I'm not going to undertake much 10 business on that. Nobody's saying that they should not be 11 allowed to package the data together, but the question is 12 should any be required to package the data together and 13 indeed if they are required to package the data together, 14 does that have implications of supporting inefficient trading 15 systems. That's where I think you might want to think about 16 it. 17 MR. GOLDSCHMID: I want to get a question in. The 18 problem of entry barrier in your analysis. 19 MR. LEE: Entry barrier is a real problem in which 20 there are market externalities, no question and it may be 21 that you want to effectively subsidize entry barrier, which 22 is what is going on here, but on the other hand, by doing 23 that, it's important to recognize that you are subsidizing 24 it, and that's what I'm saying it. 25 MR. COLBY: To what extent also is there an 1 agency-principal problem, because, it sounds like what you're 2 setting out is a model for sophisticated investors choice of 3 information, but to the extent this is as, people said a bare 4 minimum that's used for retail investors, to what extent do 5 you think that the choice would be in the hands of the 6 investor as opposed to the purveyor of the information to the 7 investor? 8 MR. LEE: People go to Tom Gardner and he says he's 9 a fool, but he's not a fool, right? And they go to him 10 because he's built up a reputation, and the issue is, if I'm 11 a vendor and I'm going to be offering something which retail 12 investors who know that they're relatively unsophisticated, 13 but can trust him, that is something that I would want to 14 invest in, so it doesn't seem to me, there's still a private 15 incentive to be promoting. 16 MR. COLBY: Let's turn to another issue that was a 17 very big part of the advisory committee consideration of 18 market data. Let me just put it out, whether it's of any 19 interest to the people in this room, and that is the question 20 of how the data gets pulled together. Is it important to 21 have a central consolidator that pulls the data together, or 22 is it sufficient to require information to be made available 23 and either leave it open to people to pull it together or to 24 say it needs to be consolidated without a central 25 consolidator? My first question, is that of anything that's 1 of any real interest to anyone at this table and if so, what 2 are your views on it? 3 It's a real interest to many market participants, 4 Ruben, I'm going to hold you for a minute, because we see all 5 these things are of interest to you. Anyone else? 6 MR. DeSALVO: I would say with regard to SIP's 7 work, whether it's one SIP or multiple SIPs competing-- 8 MR. COLBY: By SIP you mean a central processor? 9 MR. DeSALVO: Yes. The actual process of 10 consolidating disseminating is not broken. What is broken 11 and we hope that you get to it, is the cost that we pay for 12 it. 13 PROFESSOR KYLE: I have some views on this. 14 MR. COLBY: Pull the mike up. 15 PROFESSOR KYLE: I think that having consolidated 16 data going out through one feed has potential advantages, 17 especially if you augment it to put more audit trail type 18 data in it, because then you automatically get a fair time 19 sequence, assuming all the different endings that are feeding 20 data into it are feeding data into it in approximately the 21 same way and what comes out at the end is going to have a 22 certain kind of time sequence that's very integral to 23 monitoring whether customer orders that are sitting there in 24 the market are respected or not respected, there's a one- 25 second difference in execution time. 1 MR. BURKHART: I would say also that in respect to 2 the principle or ideal of fairness, if you like, in 3 transparency, if you have a consolidated data source at least 4 you know or you could feel somewhat comfortable that all the 5 participants, be they retail, institutional or otherwise, 6 with brokerage are receiving the same level of data, and what 7 they do with it and how they handle it, or how they repackage 8 it and analyze it to their customers or internally is really 9 where their efficiencies or competitive advantages should be 10 borne, as opposed to who receives what. 11 If we all received the same data from a single 12 source -- if Matt at Morgan Stanley gets the same data as 13 Duncan at Goldman Sachs, what they do with it will determine 14 their competitive edge over one another. Ditto if Minder and 15 I received or Scott and I received the same data. How we 16 analyze it, restructure it, retool it for our own purposes 17 really should be the different shading factor amongst us, not 18 multiple data feeds that we have to pay different amounts 19 for. 20 MS. ANGSTADT: It seems, too, that it becomes such 21 a tool of the regulators to, in the examination for best 22 execution and it's really been a measurement of best 23 execution that, even if somebody wants to buy unconsolidated 24 data, they really do need for regulatory purposes the 25 benchmark of the NBBO just to measure where they are. I 1 think one aspect of the Vendor Display Rule that I don't know 2 if we'll get into is the display of the last sale data, what 3 evil will befall us if we're able to display on a real-time 4 basis and unconsolidated basis the last sale as opposed to 5 just the NBBO. 6 MR. COLBY: What evil will befall us? 7 MS. ANGSTADT: I don't know that any will. Is 8 there a reason that we should not, you know, revisit the 9 Vendor Display Rule and relax some of those requirements so 10 that while we've been able to provide real-time limit order 11 books and again on an unconsolidated basis so people can take 12 a look at this information, what about providing the last 13 sale as well, if that's something that a marketplace chooses 14 to do. 15 MR. COLBY: Since not everyone has memorized the 16 display rule, what it requires is that if you make the quotes 17 from any one market available, you have to make the quotes 18 for all markets, and if you make the last sale for any one 19 market available, you have to make the last sale for all on 20 an equally accessible manner, but it doesn't require that you 21 make available the quotes of any market, it doesn't link the 22 two, the quotes and the trades together. So, Ruben, I know 23 you want to talk on these two issues. 24 MR. LEE: A couple of points. The first is, to use 25 an example in Europe, we have competing markets and private 1 market participants, either vendors have consolidated without 2 any mandatory requirement to do so. That's the first. The 3 second is I would take issue with two things which Gunner 4 said as follows: I think that his notion that if we have it 5 all together in a single consolidated fee, we all work from 6 the same information base, is wrong. I think that in fact, 7 market participants do work from very, very different 8 information bases. They see all different sorts of order 9 flows. In fact, not only is it wrong, but it gives the wrong 10 impression in that having a BB, a best bid and offer makes it 11 seem as though we all have the same information, but in fact 12 we do not. 13 And the second point I would make, which is a 14 classic economist dilemma, thank goodness I'm no longer an 15 economist, which is by requiring everyone to use the same 16 structure you destroy incentives to provide better and faster 17 structures and in fact, that is part of what makes and 18 enhances the market, so I think that it gives the wrong 19 incentive structure. 20 MR. BURKHART: Perhaps our disagreement on this 21 point is founded over our definition of the data that's 22 provided. I was referring really to the basic standard trade 23 data that should be available in a consolidated fashion to 24 all participants. Obviously, each of the participants, 25 whether it's at the brokerage end or at our end, have lots of 1 different internal information that we have gathered from our 2 own proprietary sources which may or may not give us an 3 advantage, and I wouldn't want to imply that we should either 4 make that data wholly available to everybody, thereby giving 5 up our proprietary advantage, or in fact that would be of 6 interest to anybody other than ourselves. I think our point 7 perhaps differs on the data itself more than anything else. 8 MR. COLBY: Maybe we should talk--a number of the 9 panel participants said that after decimals, that there was a 10 crying need for more and different information for trading 11 purposes. It would be worth talking, and I know that Gunner 12 and Minder took it to another level in terms of what the 13 terms on which it should be available would be, but what is 14 needed, sort of public information at this stage. Tom, you 15 represent some very active trading. What are your thoughts 16 on this? 17 MR. PETERFFY: We pull the data together ourselves. 18 If we did not do that, we couldn't provide best execution. 19 Our problem is that that data means different things to 20 different people, because some of it is electronically 21 accessible and some of it is not. Now, what do you do when 22 you see a quote that is a penny apart, and the one that's a 23 penny better is not accessible. This is the problem. So I 24 think that the market data question basically depends on 25 market structure. Once we have a market where all quotes are 1 electronically accessible, the best solution automatically 2 falls out. 3 MR. DOMOWITZ: I think the market data question is 4 not just a question of market structure, it's a question 5 from, let's say, my perspective of also quality service. 6 Now, in the same way as Tom mentioned, we also pull the data 7 together. I find it very interesting a discussion of what 8 you might call size adjusted quotes, okay, size adjusted 9 spreads, you like to know what's there at a penny, what's 10 there at a nickel. When we talk like that, what we're really 11 just saying is we'd like to see some notion of an order book, 12 that's all that really is, that's basically saying there's a 13 certain amount of shares available at certain prices as you 14 step down the price panel. Now, we put that type of data 15 together ourselves for the benefit of our clients. In other 16 words, we pull data from a variety of sources, we use a 17 variety of vendors to do that, and, frankly, we don't even 18 find that in and of itself completely sufficient to meet what 19 we believe are the needs for execution. 20 We also need, if you like, scorecards for how well 21 the markets are doing. So when we talk about what do you 22 actually need to get the job done, it's actually not even 23 restricted to prices and clients, we actually need a lot of 24 things. But we can do it, we're pretty happy that the market 25 does it, we happen to think that the market does it rather 1 widely right now. It's not a question of having to mandate 2 the notion for this quasi order book for this information 3 that's provided to all participants, the market is basically 4 providing this and with means to access, although I would 5 agree with some of the comments around that access is often a 6 trifle difficult. 7 MR. COLBY: The limit order book display, other 8 than the top, is a voluntary market driven system and some 9 systems like Archipelago and Island makes its entire book, 10 New York is making its book available, NASDAQ is working 11 towards. Is this producing what the market traders need and 12 is it having a negative effect on reducing people's 13 willingness to put information on the book, and Scott, maybe 14 I can, you said I think earlier that post pennies that you're 15 doing a lot more trading upstairs and on ECNs. Is that 16 trading visible or is it less visible than it was before or 17 what's the thinking behind that? 18 MR. DeSANO: The reason we go upstairs is to take 19 some of the problems of market structure off of our plate and 20 put it on the Street, to be honest. It's an unpleasant thing 21 for them. It's more motivated by splitting trades on the 22 Floor that we don't want to split. It has nothing to do with 23 really what we're talking about here. 24 MR. COLBY: Minder? 25 MR. CHENG: I think just one more comment. I think 1 taking a step back, what is the reason for NBBO? Before even 2 getting to that, I think Gunner mentioned, for the buy side, 3 it used to be, when the buy side didn't really have much 4 direct access to the market, when we traded just about 5 everything through the broker-dealers, so which means that 6 even if I could see there was a better quote on this exchange 7 as opposed to the other exchange, the relevance was very 8 little, because we still traded through a broker-dealer, so 9 we basically imposed that responsibility on the broker- 10 dealers. Not these days. These days with this ongoing 11 intermediation from the sell side and also with the 12 technology that offers the sell side the ability to directly 13 access the market. 14 Now, here comes the question that I raised earlier. 15 If we put, what we are trying to do is basically trying to 16 figure out the order book or supply/demand. But post decimal 17 I think lack of protection, as Scott put it, basically, it 18 has forced the buy side in a way to trade more upstairs and 19 that is indeed true. And, so, in other words, there's an 20 order book, there's an NBBO and we can even address the 21 depths, but there's also that liquidity that are not showing. 22 We often go through ATSs and ECNs and we place our order 23 through the reserve book. We don't show the sizes of the 24 market because, again, because of the lack of protection by 25 showing it we will get picked off, most likely. So I think, 1 again, going back to Ian's point, the data structure or 2 whether we want to create the right data that, again, that 3 would sort of fit the retail need and the institutional need 4 for them, meanwhile, there's no protection that has been 5 offered to guarantee that whoever is providing that 6 transparency to the market can be protected and rewarded. 7 So I think that as long as we can't, I mean, to the 8 extent that we cannot resolve that issue, how much do you 9 want to provide through NBBO, how much depth, I think will be 10 ongoing. 11 MR. COLBY: Can we follow this up? Earlier I think 12 Minder and Gunner both said they need to be rewarded for 13 putting limit orders on the book with some kind of a price- 14 time. 15 MR. BURKHART: I don't know if I would use the word 16 "rewarded " in front of the regulators. I certainly am not 17 looking for any financial reward. 18 MR. COLBY: Incentive? Well, you're probably the 19 only person. Incentive for putting the orders. Could we 20 proceed on this for just a while? How much price/time 21 priority would it require to encourage institutions to put, 22 and ultimately, is there any amount that's going to encourage 23 you to put the full size of a major order on the book or is 24 that something that has to be worked, that ultimately has to 25 be worked upstairs? 1 MR. DeSANO: It's not just price/time priority, 2 it's I don't want to be improved. If in order for me to put 3 an order on the book, I want to reach two or three cents and 4 take that stock from that person who's provided that 5 liquidity on the book, I don't want to be improved. I don't 6 want somebody to step ahead and sell it to me a penny 7 cheaper, I don't want it. If I get that, I'm not putting an 8 order on the book myself, all I'm providing is a way to get 9 picked off. So it's beyond price/time priority. Does that 10 make sense to you? 11 MR. COLBY: Are you saying when there's an order in 12 the market you want to be able to access it? 13 MR. DeSANO: I want to be able to access it, okay, 14 but in order to make it work, in order to get that limit 15 order there so you have size on the book, you have to protect 16 the person who has provided the liquidity and if I want to 17 reach through and grab that liquidity by definition I should 18 say I forego wanting price improvement, I have made the 19 decision I want to buy that stock. I don't want somebody to 20 step ahead of that and free ride that person's order so. By 21 definition I don't want them to sell it to me cheaper and I 22 don't want them to go along with that purchase or I won't put 23 an order down there myself. 24 MR. BURKHART: I agree with Scott and to invert the 25 perspective, what we would, we, being the institutional asset 1 management community, would be used for when we put our large 2 and show large size orders on the limit order book is a free 3 option. It's a free option for the specialist in the New 4 York Stock Exchange, it's a free option for the floor 5 brokers, it's a free option for anybody that wants to step in 6 front of us for a penny or a smaller increment if we ever get 7 to that, to trade ineffectively and to dilute or shut us out 8 effectively. It's one reason why Scott is trading upstairs 9 and why we do and Minder does more today than perhaps in the 10 days of old, is we don't want to give that free option to all 11 the market participants. That's a structure issue. I'd be 12 more than happy to put 100,000 shares of ABC stock down on 13 the order book. I don't want to show it and I want to know 14 that I've got price/time priority with that order down there. 15 MR. DeSANO: I just want to add, the concept of 16 price improvement in a market environment in the past was 17 widespread I think made more sense today with decimals, I go 18 back to best execution is in the eye of the beholder. We 19 need liquidity, we need access, too. I think price 20 improvements should go to those who have orders out loud. If 21 somebody's on the book at 20.03, 20.04 and there's a block 22 out at 20.05, I want to reach out to 20.05, I'll pay the 23 person at 2, 3 and 4 cents at 20.05 and I want that liquidity 24 on the book. That's actual price improvement, not what we 25 have today. 1 MR. COLBY: Gunner, if I heard you correctly, you 2 said you'd be perfectly happy to put a 100,000 share order on 3 the book, but you don't want it to be shown, and you want 4 price/time priority. 5 MR. BURKHART: Does that sound contradictory? Is 6 that what you're asking me? 7 MR. COLBY: It's what everybody wants, to have your 8 cake and eat it, too. The question I'd ask is right now, we 9 have individual markets set their own price/time priority, 10 and forgive me if I'm encroaching on your panel here, Larry, 11 but there's no market price/time priority, and even within 12 markets the amount of price/time priorities vary. So my 13 question to you is how much payback would there be if there 14 was greater price/time priority within a market, and someone 15 mentioned the penny versus the larger increment, and also 16 there's a question that even within a market you might have 17 price/time priority but not across a market. Janet runs I 18 think a pure price/time priority system and that's been a 19 selling point of yours. 20 MS. ANGSTADT: Yes, it has been. But we certainly 21 have people who want to be able to access quotes with the 22 price/time priority. We have some order types that help 23 reach out to people who are very fast if we don't have 24 liquidity on the system, but it is a strict price/time 25 priority system within Archipelago. 1 MR. NIEDERAUER: I think what Scott and Gunner and 2 Minder are all saying, would be, I think the genie is out of 3 the bottle on this one a little bit. I think now you're in a 4 decimalized environment, we might aspire to have a pure 5 price/time environment, but under these conditions there's no 6 such thing as a pure price/time environment now. 7 When you listen to Scott's depiction of what a 8 larger institution would like to do, that's what I think the 9 NYSE Institutional Express order was designed to do, and we 10 could argue in practice whether it has or hasn't worked. I 11 think what you're seeing from our seat is competition is 12 kicking in. I think six months or a year ago we would say 13 let's mandate that since the NBBO is becoming less relevant, 14 but it's still important, let's just mandate that people have 15 to show more than one level. Well, it turns out you really 16 didn't have to mandate that, because everybody is showing 17 multiple levels. Folks like us are consolidating everything 18 that people are contributing and putting in front of 19 institutions like Minder, Scott and Gunner represent. On top 20 of that the competition kicks in with these various order 21 types. It may sound like these guys want to have their cake 22 and eat it too, I'm not sure in a decimalized environment 23 what else we're to do, other than to display a small quantity 24 display, have a lot on research and count on the competitive 25 forces again to take over where intelligent order routers 1 begin to understand there are certain pools of liquidity that 2 there's more than meets the eye behind there. I'm not sure 3 how we can solve that. 4 I think short of turning the clock back to a 5 smaller minimum price variation, I think these are natural 6 outgrowths of a competitive market structure that maybe 7 aren't as dysfunctional as we think they are. 8 MR. MURPHY: Could I weigh in a little bit, another 9 observation with Gunner and Scott in a decimal environment. 10 I still think price improvement is a valid option. In the 11 opening statement, Scott, when you said that shareholder 12 value for your fund holders is your primary concern, why 13 wouldn't you, if it was offered to you, want price 14 improvement? That liquidity is still there, the perception 15 is that someone jumped ahead of that published bid or 16 offerings whatever it is, for liquidity. One misnomer and I 17 know NYSE staff can back this up. The specialist as far as 18 pennying is extremely insignificant, almost nil, as far as 19 penny jumping. Really where the pennying comes in on the 20 Floor is agents that represent the institutions, who like 21 Minder said earlier, you know, the transparency is a double- 22 edged sword. A lot of people that have some size to do are 23 careful about what they divulge and advertise on the national 24 best bid or best offer, but if a large institution has a 25 further interest, and gets price improved on a block of stock 1 while that advertised, say up a penny or two, is still there, 2 if he has a further interest, it's still available to him. 3 So why not take advantage of that price improvement, even if 4 it is a penny or two? I don't think that's a bad thing. 5 MR. COLBY: Let's now shift directions and talk 6 about market data revenues. Currently, the public market 7 data system is very closely linked to self-regulation. SRO's 8 collect the quotes, the trade information from their members, 9 and they make this information publicly available, and 10 historically the SRO's have charged for this information. So 11 let me start with what was our Chairman's question at the 12 last hearing which is, is market data a public good that 13 should be available for free to investors and market 14 participants? Start with that one. 15 PROFESSOR KYLE: I'll answer that. 16 MR. DeSALVO: I don't have a problem with that. 17 PROFESSOR KYLE: There are a lot of externalities 18 that exist with respect to markets, and if you think about 19 market data, market data is providing a positive externality 20 to the people who use it. It has characteristics of a public 21 good in that once it's out of the bag it's a little hard to 22 prevent it from circulating around, but it can be prevented. 23 If you think of market data as providing an externality and 24 if you thinking about providing a limit order as hanging 25 yourself out there to get picked off, that's providing an 1 externality. That suggests that market data should be 2 expensive because the people who are benefiting from it as a 3 positive externality should in some sense maybe pay for it, 4 like a little tax on them having to pay for it I think is 5 probably reasonable. If you think about an efficient funding 6 for exchanges, market data fees, that is payments for a 7 positive externality that they're creating for everybody else 8 should be a large component of their revenues and that's kind 9 of an efficient way of creating that positive externality. 10 COMMISSIONER PITT: Now, let me follow up with 11 that, because there are different types of market data and 12 maybe the problem of coming up with an answer to the question 13 exists when you try to conflate the different types of data. 14 Let's just deal with transactions and stated bids 15 and offers. If you assume the best execution, whatever it 16 means, is a legal obligation and if you assume that the only 17 way you can fulfill that obligation is to be able to have 18 sufficient data to note where best execution will take place, 19 again, whatever best execution is, then if you just limit 20 market data to that kind of information, people who are 21 willing to have limit orders on the books in size, et cetera, 22 why isn't that akin to IBM's results of operations, for which 23 we do not charge; that's available to investors for free, 24 because it's thought to be relevant information about making 25 a judgment. Why isn't that equally true of market data about 1 bids and offers and actual transactions? 2 PROFESSOR KYLE: I think because in the case of IBM 3 providing set financial information, the shareholders of IBM 4 can be located as the individuals who ultimately will benefit 5 from having that information provided, and it's fair enough 6 to make them pay for it, so in some sense, maybe it's IBM 7 that bears the cost of creating its own financial information 8 and that's a reasonable way to finance the provision of that 9 information. But when you're talking about a trading system, 10 there are going to be people out there who would like to 11 benefit from the price information that is created by the 12 trading system, and the people that are benefiting from the 13 price information created by the trading system are not 14 necessarily using the trading system, they could be offshore 15 or maybe trading outside the United States. Those persons, 16 it seems to me it would be fair to ask them to pay for that 17 information because they're actually benefiting from it in 18 realtime. 19 MR. HARRIS: Who is creating the price information? 20 PROFESSOR KYLE: The exchanges. 21 MR. HARRIS: Is it the exchanges or the traders 22 that act at the exchanges, and if it's the traders who act at 23 the exchanges what mechanism takes the data from the market 24 revenue back to the people who are actually producing the 25 information, which is what's necessary for best execution. 1 PROFESSOR KYLE: That was your question for the 2 session this afternoon, I think. 3 MR. HARRIS: There is the issue that the incentives 4 for passing along the information is too great or too small. 5 So if traders can shop around for exchanges and exchanges are 6 charging for data in some manner that's proportionate to the 7 amount of business they do, then exchanges are going to 8 rebate the fees they get from traders ultimately to the 9 traders, and traders shopping around for exchanges are going 10 to ultimately play one exchange off the other. How far that 11 can go is an issue. 12 MR. NIEDERAUER: I don't think it's the same today 13 and I think the Chairman's question was at the first hearing, 14 think about what the exchange's product is, and I think we'd 15 all agree that it's fairly non-controversial that it's fair 16 that an exchange charges a transactional fee for bringing 17 together a buyer and seller to transact with one another. 18 That's their product, their core competency, they're 19 delighted to provide that facility and I think we're in all 20 within reason willing to pay. I think where it becomes more 21 confusing for us is, think about what we've been talking 22 about in this session so far the provision of market data to 23 help you decide where to trade. One could argue it's nothing 24 more than the exchange facility trying to attract liquidity 25 to come to it's place so it could charge that transaction 1 fee. I think as we thought about this, we felt the same way 2 about the reporting of the trades that took place. That's at 3 best a byproduct of what we've already paid for on the Floor. 4 To Larry's question, one could argue that you could 5 trace its ownership back to the sell side or the buy side, 6 not the exchange, so I think we all feel like we've paid for 7 the exchange's key product, which is bringing buyer and 8 seller together and the provision of data to help us make the 9 decision before and after that transaction challenges whether 10 we should be paying for that. 11 MR. DOMOWITZ: There are a couple of things, here. 12 I actually don't completely disagree, but the first point is 13 that I think it's misleading to say that IBM provides its 14 corporate data for free. The shareholders pay that cost. 15 The army of accountants, lawyers, technology systems and 16 whatnot that's required to produce that data impacts the 17 bottom line of IBM. 18 MR. PITT: I think that's right. 19 MR. DOMOWITZ: The provision of data impacts the 20 bottom line of the exchange. 21 MR. PITT: But they provide it to the public, the 22 investing public for free. That's the distinction. I 23 certainly agree with you that there is a cost for the 24 provision of that data and somebody is paying for it. 25 MR. DOMOWITZ: Someone is paying for it. So the 1 issue is perhaps also how much, right, which is something 2 they could come to, I'm sure, something which I disagree a 3 little bit, with Duncan, we haven't raised yet what the 4 monies are really used to pay for. In other words, if I am 5 an ATS and I provide share transaction services, and I charge 6 for that, no question, if I'm an exchange, I also charge for 7 that. 8 If I'm an exchange, I have to at least under the 9 current system fund regulatory oversight. In fact, that 10 exchange may even oversee my ATS, as things are, these days. 11 So there is a question-- 12 MR. COLBY: Hope they do. 13 MR. DOMOWITZ: --of who is going to pay for that. 14 We could say, you could make an argument, that the share 15 transaction charge should pay for the regulation, but this is 16 basically talking about cross-subsidization across different 17 areas of the business. There's a data area of the business, 18 there's an execution area of the business, they do complement 19 each other, where the monies come from and what are they 20 paying. 21 MR. NIEDERAUER: I think that's fair. I think I 22 actually do agree with you. I think the corollary to that is 23 there are some cross-subsidies that this creates and I think 24 one our guidelines for this session was to just keep it 25 simple. I think it's gotten overly confusing and I know 1 we're going to get into probably market data rebates at some 2 point, which I think is an additional distortion. Nobody 3 wants to go to a place to trade that is underfunding its 4 regulation. I think our view is make the cost of that 5 regulation more transparent. We'll decide collectively 6 whether we think that's fair or adequate, because we don't 7 want anyone taking shortcuts there, but don't confuse the 8 issue by saying we're going use this other source of revenue 9 to fund this regulatory cost. Just put it out there. 10 MR. DOMOWITZ: I'm a hundred percent for that. 11 When I was thinking about my own opening remarks, the one 12 that I really wanted to come out with -- perhaps now is the 13 right time -- is that the morass of both implicit and 14 explicit rules and the cross-subsidies that are created is 15 more complicated than the U.S. Tax Code. If I stacked all 16 the books summarizing the U.S. Tax Code on the podium I 17 guarantee given the rickety nature of it, this podium would 18 just collapse. The thing, is we don't even have all these 19 market structure intricacies written down. They are 20 byproducts of how we said we are going to operate the 21 markets. I think Duncan's point was extraordinarily correct 22 -- if we're able to clear away some of that, we can get a 23 better view of the underlying economics of the issue. 24 MR. COLBY: As the lone registrant at the table, 25 Paul, do you want to respond to this? 1 MR. O'KELLY: Nothing is free. The notion we 2 should put out market data for free is often said by people 3 who don't have to put together exchange operating budgets for 4 the coming year. This is simply an allocation of the cost of 5 running the exchange. Who is going to bear what costs. 6 Exchanges go through these exercises regularly. The exchange 7 has to adequately fund itself, the SRO's responsibilities, 8 they have to get the money from some source, try to allocate 9 those reasonably across all the people who use their markets. 10 We're not here today to debate the cost of annual membership 11 dues, we're not debating the cost of transaction fees, we're 12 not debating the cost of clearing fees, we're not debating 13 the cost of listing fees. These are all fees an exchange 14 assesses so it can fund itself as a marketplace, as a 15 self-regulatory organization. There is nothing special, 16 nothing magic about market data fees. They are one of the 17 sources of income for the exchanges to fund themselves and we 18 should not lose sight of the fact that exchanges are mutual 19 organizations run by the broker-dealers providing those 20 orders and if they believe that the allocations of the costs 21 across the various lines of business in the exchange are 22 unfairly discriminating against one segment, they have the 23 opportunity within those organizational structures to make 24 those points and get changes made. 25 MR. DeSALVO: Could I just weigh in? I think it 1 would be helpful for a minimum standard -- yes, they'll sit 2 on a monopoly of the market data and as such they have their 3 privilege to disseminate that information that we end up 4 receiving. There should be a minimum standard that allows us 5 to trade that we get. Could it be NBBO plus three, last 6 sale, and size, at which point the pricing of that minimum 7 data be at one level and then any of the other data that the 8 SRO decides to put out on a competitive market basis they 9 should be able to get whatever they want. 10 But we pay in excess of a hundred million dollars a 11 year for market data. We are not getting a hundred million 12 dollars plus worth of best execution analysis on that data, 13 and there's no way that the current market structure is going 14 to allow that to change. A strong end minimum low-priced 15 NBBO should be priced at one level, anything else should be 16 reflected into what the competitive market prices should 17 bear. 18 MR. COLBY: Some folks mentioned that recently 19 there was a development where some folks rebate market data 20 revenues back to people that report trades or sometimes 21 submit limit orders and that raises a number of questions. 22 One, is this an appropriate use of market data revenues? Is 23 it somehow undercutting, some folks have said there's a 24 self-regulatory function and is this rebate or rebates by 25 some undercutting the self-regulatory function? Does it 1 suggest the rates are too high or is this an appropriate way 2 to allocate out the value of market data revenues to the 3 people that are producing it? Anyone have views on that? 4 MR. O'KELLY: May I respond? 5 MR. COLBY: Paul would like to respond. 6 MR. O'KELLY: To me this is a classic case of 7 damned if you do, damned if you don't. You have half of your 8 members saying market data fees are too high. So you say, 9 okay, let me look and see how I can run my exchange 10 efficiently and use the market data fees that are beyond 11 those required to run the exchange, try to incent my members 12 to try to make better markets. I use those monies, return 13 them to the members rather than keeping them. I advocated 14 internally a long time market data fees shouldn't be 15 returned, they should go to officer's bonuses, but we passed 16 that over many times. 17 (Laughter.) 18 MR. O'KELLY: So we rebate them to members and we 19 give them to members who demonstrate they can increase their 20 market share, provide better markets. We think that's the 21 most competition among markets and it's a valid use of market 22 data fees. The alternative would be to keep the market data 23 fees. My bonus would go up. We don't think that's the right 24 thing to do. We see nothing wrong with rebating market fees. 25 The issue isn't rebating, it's to make sure that whatever 1 rebate program you put in place is not one that's susceptible 2 to manipulation or misuse to the people to whom you give the 3 money. 4 MR. DeSALVO: That assumes that the current level 5 of market data fees is correct and we're saying it's 6 incorrect, let alone whether you should keep it for your 7 officers or send it out to attract order flow as a form of 8 payment for order flow. What we're saying is that the 9 general notion for market data fees is too high and we should 10 not have to pay the same for the general, to place an order 11 on a trade, the minimum level strong man that I talked about, 12 that level fees is too high. 13 MR. O'KELLY: Not all exchanges rebate the fees, so 14 for those exchanges who are not rebating any fees to the 15 members, perhaps they could answer for them they need all 16 that market data to run their exchange. Perhaps because 17 there are revenues from other sources, not sufficient to run 18 the rest of the responsibilities or perhaps they run their 19 SRO responsibilities and other mechanisms in a more robust 20 fashion. 21 MR. DeSALVO: Why should we pay market data fees 22 for best execution purposes for you to rebate to your 23 participants to attract order flow to your exchange. I don't 24 follow that logic. 25 MR. COLBY: Bob, as vice chair of the New York 1 Stock Exchange, I think you would have some insights to this. 2 MR. MURPHY: Remarkably, in doing my homework to 3 get ready for the panel I realized nearly 70 years market 4 data fees have stayed constant as about 17 percent of the New 5 York Stock Exchange's revenue stream. 6 The other point I would like to make, as far as the 7 Board goes, in our structure, we have nearly half our board 8 is made up of industry leaders, and they get, I think it's 9 twelve out of 27 board members sit on the board, and they get 10 a look at the budgets and the data pricing schedules, and I 11 think that the challenge is to keep it fair and to keep it 12 reasonable, but I agree with Paul on one point, that there is 13 some heavy lifting in, especially as far as the New York is 14 concerned, because there are so many moving parts and we have 15 to put it all together and then the product is produced. We 16 have the floor brokers, the specialists, we have the trades 17 and the quoting and there's some heavy lifting there. 18 Where I totally disagree, and where Paul and I 19 disconnect is the data rebates for buying tape prints. That 20 I think is maybe an indication that market data fees could be 21 a little on the high side, because I think it's fictitious 22 data anyway, because these are internalized trades, there's 23 no opportunity for any interaction by any other market 24 participants, and as far as this consortium goes, it's almost 25 like we're subsidizing, and I have to agree with Ruben on 1 this, we're subsidizing a lot of what's going on, because 2 they piggyback the consolidated tape and consolidated quote 3 consortium. It's a tough issue, but it's one that should be 4 dealt with. 5 MR. PITT: Bob, let me just ask one question. 6 Which way does it cut if you've been able to maintain market 7 data fees at a set percentage for 70 years? What conclusion 8 should we draw from that? 9 MR. MURPHY: I think it's right now, in this 10 environment, we need every penny we can get, but it funds a 11 lot of things that go on there, and 17 percent of our revenue 12 stream I don't think is exorbitant, but again, I want to 13 point out that our board, made up of twelve industry leaders, 14 can react if they feel at the table that these are, we are 15 overcharged, and I think we're okay as long as we're not 16 maximizing profits, but just making it sufficient to fund 17 what we have to do as an SRO. 18 MR. PITT: I guess one model of pricing tends to be 19 what's your cost for the production of either the product or 20 the service, and then what's a reasonable return. 21 Presumptively, if the costs were being set that way, it would 22 be highly unusual if it came out to be 17 percent of total 23 self-regulatory costs over 70 years, which suggest that the 24 costs are being set some other way, which then leads to the 25 question that I think some of the people who pay the fees are 1 asking, which is how are the costs set, it's not just a 2 question of what they're funding, but how are they set and 3 why is it appropriate to pay that amount of money and 4 potentially, as was raised earlier, are there other costs of 5 regulation which are obviously critical, can be recovered 6 without tying into market data. 7 MR. MURPHY: Chairman, I don't know, I think it's a 8 really stretch to know if you can identify exactly what the 9 expenses are of putting all the data together to form one 10 product or to come up with our tape and our quote line, but 11 you know, I don't have the answer for you. I think that's 12 something staff could work on and maybe have an answer to, 13 but it's a good question. No doubt about it. But right now, 14 I think it's fair. 15 MR. DeSALVO: I think the cost of processing data 16 has not been reflected in the cost of disseminating the fees 17 to the buyers of that data. That Moore's law has not been 18 implemented into our bills that we receive every month with 19 regards to data. That the cost of producing a billion and a 20 half shares worth of data is not is cheaper than it was ten 21 years ago when it was a billion shares of data. That's not 22 being reflected in the market. 23 MR. COLBY: Did you want to speak to this? 24 MR. LEE: I had a couple of points. One is 25 practical, two is conceptual. The practical point is to 1 agree with the stock exchange down there, which is saying if 2 you actually look at the different sources of revenues of the 3 different exchanges, in fact almost all those sources are 4 under threat, apart from market data, so actually, I predict 5 that we will see market data fees going up, much to the 6 chagrin of the Morgan Stanley friends over there, that's the 7 first point I would make. 8 The second point is a conceptual point and here, 9 while I very much agree with Duncan in his statement that we 10 need to identify clearly what the costs of self regulation 11 are, and have a debate appropriately about how to fund them, 12 I think that's absolutely vital, I agree with Ian. What I 13 disagree with him, which is a general conception, is that an 14 exchange is a trading system and a byproduct of that is data. 15 To the contrary, a completely appropriate view of an exchange 16 could be it's a media company and what it produces is data, 17 and I think what you're saying is exchanges and other forms 18 of trading systems are reducing their trading costs to get 19 them closer and closer to zero, so it doesn't surprise me 20 that data fees are going up, and I think that's the wrong 21 model in which to decide what is appropriate. 22 I want to do a couple of more conceptual points 23 which Ian raises and follow them up about the appropriate 24 price for data. If something is a public good does not mean 25 that it's the appropriate cost, the appropriate cost for it 1 should be zero, and in fact, again, as a very ex-economist, 2 one would normally talk about that the appropriate cost 3 should be its marginal cost. In this context, the marginal 4 cost of producing and disseminating data is actually falling 5 very, very close to zero, okay, and what I would argue is 6 that that modern economic model for deciding what is an 7 appropriate cost for data is actually inappropriate and then 8 we can go into the details of that subsequently. 9 MR. NIEDERAUER: I think I'll try to have a 10 practical one and a conceptual one, too. I think the first 11 thing I would say you may just be having a conversation about 12 different kinds of business models. The sell side's business 13 model, for example, tends to be a bundled matter. If we 14 trade with BGI, Fidelity or Deutsche Asset, I think the 15 equivalent we're talking about the exchange is doing, is 16 every time I send Gunner an indication, I send him a bill 17 with that indication because here's my attempt to say I've 18 got some stuff going on, here's what I'm competitively 19 putting in front of you to express a willingness to transact 20 with me and I would charge him for that. I would charge him 21 to transact with me, I would then report the trade and then 22 charge him for that also. We think that our transaction fee 23 is certainly fair, it's negotiated with our customers every 24 time, it may be as simple as a difference between a bundled 25 pricing model and an unbundled pricing model, it may be that 1 simple. 2 I think to get back to the rebate question, I think 3 the way Paul explained it, it sounds great. It sounds if 4 that's how everybody administered it, maybe we wouldn't have 5 a problem with it, I think it's easier to just reduce the 6 fees instead of talking about charging here and rebate it. 7 One of the buzz words for today is unintended consequences. 8 I think one of you mentioned that in your opening remarks. I 9 think one of the unintended consequences of these market data 10 rebates, it led to strange behavior. It led to people 11 trading in one place, but then printing the trade somewhere 12 that just was going to write them the biggest check. I think 13 with all the themes we're trying to discuss today, that's 14 just terribly distorting and disappointing, because it 15 misleads the very people that we're trying to put better 16 information in front of. Suggesting that some exchange 17 simply because they pay the most money is a real venue for 18 liquidity, it's just terribly misleading if that's what we're 19 trying to avoid. 20 MR. COLBY: Unless the Commissioners, if any 21 Commissioners have any specific questions, then I think we're 22 going to call a halt for this one. I just say, from a 23 practical standpoint of we are going to look at the costs of 24 production and then setting the fees from that, we're very 25 glad that we have a very able chief economist. 1 (Whereupon, at 12:15 p.m., a luncheon recess was 2 taken.) 3 A-F-T-E-R-N-O-O-N S-E-S-S-I-O-N 4 (1:25 p.m.) 5 MR. HARRIS: Welcome back from lunch. 6 My name is Larry Harris, I am Chief Economist at 7 the SEC. We are now going to start the session on best 8 execution. 9 Viewed broadly, obtaining the best possible 10 execution of orders is the reason we are all together here in 11 this room. For some of us, it is the reason why we are in 12 this business. High quality executions encourage investor 13 participation in the securities markets and thereby reduce 14 the cost of raising capital. 15 Best execution is a concept ground in the agency of 16 fiduciary law. The legal obligation of best execution owed 17 by a broker-dealer to its customer requires that the 18 broker-dealer obtain the most favorable terms for a 19 customer's transaction available under the circumstances. 20 When a security is traded in only one market and all orders 21 meet and are executed within that market, best execution is 22 relatively simple. Broker-dealers merely have to determine 23 how best to present each order to the market. 24 The competition among orders produces the best 25 prices available for those orders at that time. The 1 simplicity of such fully consolidated markets is attractive 2 to many people. Many securities, however, trade in multiple 3 markets. The competition for order flow among market centers 4 encourage the development of cheaper and more innovative 5 trading services. 6 For this reason, Congress and the Commission long 7 encourage competition among market centers. Unfortunately, 8 assuring best execution may be quite difficult when competing 9 market centers trade the same security. Investors, 10 especially small ones, often cannot easily monitor and 11 enforce best execution. The amount at stake is often small 12 and detecting, much less proving, the availability of a 13 better execution is difficult. 14 Brokers, therefore, have relatively little risk of 15 liability from investors complaining of poor execution. In 16 these markets, some brokers may not aggressively seek the 17 most optimal execution for their customer orders either to 18 avoid the search costs involved or to profit from trading 19 with those orders. They may trade directly with customer 20 orders as a dealer or route their orders to another dealer in 21 exchange for payments for order flow. 22 When multiple market centers compete for order 23 flow, broker-dealers who seek to provide best execution must 24 identify the location of the best available prices and access 25 those prices routinely and efficiently. If they do not have 1 facilities to accomplish these tasks, investor orders may not 2 interact at the best available prices and efficient price 3 discovery across the marketplace may be impaired. 4 One of Congress's principal goals in enacting the 5 75 Act Amendments was to address pure customer executions 6 resulting from trading securities in separate unconnected 7 markets. In addition to establishing national market system 8 mechanisms for collecting and disseminating market data, 9 Congress envisioned the framework in which broker-dealers 10 have access to all markets, competing markets would be linked 11 together in a way to produce efficient executions at the best 12 available prices. 13 The Commission has striven to remove discriminatory 14 barriers to direct access directed by exchanges and the NASD. 15 The Commission has also sought to integrate new trading 16 systems, such as ECNs, alternative trading systems and other 17 routing services into intermarket mechanisms. The 18 Commission's Regulation ATSs requires display of customer 19 orders of certain high volume ATSs be integrated into the 20 public flow, and the major ATSs provide fair access on 21 neutral nondiscriminatory terms. 22 The Commission has also encouraged sufficient 23 linkages between market centers. These linkages provide a 24 back-stop mechanism to help assure that orders routed to one 25 market benefit when better prices are available in a 1 competing market. These linkages may also help smaller and 2 newer market centers compete with the dominant markets by 3 reducing the need for each broker-dealer to build separate 4 links to each competing market. 5 These linkages and the anti-trade through 6 regulations that often accompany them, have been quite 7 controversial. Slow linkages favor slow markets; fast 8 linkages favor fast markets; and no linkages favor 9 well-established markets in arbitrage orders. Not 10 surprisingly, the SEC has heard every possible preference 11 expressed. I expect we will hear that again today. 12 Best execution involves more than just the 13 execution of individual orders between the given market 14 structure. Market structure affects the incentives to quote 15 vigorously and offer aggressive limit orders. Aggressive 16 quotes and limit order prices produce the best prices for all 17 orders. Many people believe that the national market system 18 should be structured to award market participants for 19 contributing to price discovery. 20 The existence of multiple market centers competing 21 for order flow can reduce opportunities for order interaction 22 and, thus, discourage vigorous quote and limit order 23 competition. This problem arises because most markets 24 encourage competition among orders within their markets but 25 not across markets, and because dealers trading with their 1 customers can freely ask superior prices displayed in other 2 market centers. 3 In this session on best execution, we will now 4 consider these questions, among many others: 5 Should markets be linked? 6 What does best execution mean? 7 If the markets are linked, how fast should the 8 linkages be? 9 Do we still need trade-through rules to assure best 10 execution? 11 How should trade-through rules be applied? 12 How should we assure the best representation of 13 limit orders? 14 Should market centers be allowed to charge access 15 fees for market orders? 16 Should market centers be free to offer liquidity 17 rebates for limit orders? 18 Should brokers be allowed to take such rebates? 19 And should brokers be allowed to take payments for 20 order flow? 21 Anyone of these questions is a question that could 22 easily occupy us for the two hours that we have in front of 23 us. We will try to try cover as many of these as we can, and 24 other questions of interest, as we debate the market 25 structuring. I will note very quickly before we start that 1 one thing that I found fascinating at the SEC in this market 2 structure area, nobody ever comes to us saying that they are 3 in favor of less competition. Everybody is in favor of more 4 competition, but there are two types of competition that they 5 speak for, usually not carefully identifying the fact that 6 there is another type out there. 7 They either speak for best competition or 8 competition for best price. That's the traders competing 9 with each other for best price. And that competition, of 10 course, works best when everybody comes together in one 11 place. Or they speak for competition among market centers, 12 which is wonderful for producing low-cost execution services 13 and innovative services, but it is fundamentally at odds with 14 the first type of competitions. 15 So we have two competitions that are, in some 16 sense, at odds with each other. I venture to say that just 17 about everybody in this room is, indeed, in favor of 18 competition and we usually hear arguments for one side or the 19 other. The difficulty that the SEC faces is trying to trade 20 off between those two types of competition. 21 So let's start off with just a question that we 22 have touched upon a few times this morning already: What 23 does best execution mean? Who is responsible for assuring 24 best execution? Should it be the broker who handles the 25 order, or should it be the national market system that 1 insures that orders get routed to the best prices? 2 Why don't we start with Pete; do you want to offer 3 an opinion about that? 4 PROFESSOR KYLE: I don't know if I want to offer an 5 opinion on what best execution means exactly, but I want to 6 offer something that's related to that. And that is that the 7 key ingredient is allowing the customer limit orders to meet 8 each other in the marketplace. That is for having a 9 mechanism in place for if I place a limit order to buy and 10 you place a limit order to sell, they can meet. 11 The second ingredient is tick size. If tick size 12 is really large, there is a good chance that we will not meet 13 even if we are at the same price or some area in between. If 14 tick size is very small and we have the opportunity to meet, 15 then we probably will meet. To think about the consequences 16 of having a larger tick size limit order not meeting, you 17 just have to go back and figure out the way NASDAQ worked 18 five years ago. When the tick size of NASDAQ was reduced, it 19 simultaneously allowed customer limit orders to meet each 20 other more efficiently, my understanding is there was a 21 dramatic drop in trading costs for the small customers on 22 NASDAQ. 23 So I think those two ingredients, customer limits 24 orders meeting each other and small tick size, kind of solve 25 most of the problem regardless of how the rest of 1 institutions with margins work. 2 MR. HARRIS: So how would you have customer orders 3 meet each other in the national market system? How would you 4 interpret the national market system? 5 PROFESSOR KYLE: I would interpret it that there is 6 a mechanism where a customer can place an order with a 7 broker, that order gets displayed nationally, and there is a 8 mechanism where if I place a limit order with your broker, 9 automatically the mechanism can find me. 10 MR. HARRIS: And the broker would be responsible 11 for insuring that that automatically happens or would you 12 have some consolidated mechanisms to do that? 13 PROFESSOR KYLE: Then it gets down to details, but 14 the basic detail is that if you tell your broker, I would 15 like to get the best price in the market, the broker has an 16 obligation to find me if I am also advertised on the national 17 market system, which I would be. 18 How that gets implemented is kind of details, 19 important details, but the big picture is that mechanism is 20 in place. 21 MR. HARRIS: I neglected to mention that Chris 22 Quick from Fleet Securities has now joined our panel 23 representing himself, of course, but more broadly, I presume 24 NYSE Specialists. 25 Chris, what does best execution mean to you? 1 MR. QUICK: I think best execution is when the 2 broker looks after the customer's best interests, regardless 3 of his own interest. We have to, again, think about who is 4 driving this market? It is the customers. It is not the 5 other end of the spectrum. It is not the dealer, it is 6 always the customer first. And I think that's why the New 7 York Stock Exchange has in its history put in the forefront 8 its long tradition of serving investors first. 9 MR. HARRIS: Chris, is the customer well served by 10 having an ITS system? 11 MR. QUICK: The customer is served by, again, 12 getting the best price for that, regardless of where it comes 13 from. If it comes from ITS, if I have a customer order that 14 gets delivered here on the floor to me, I have an obligation 15 to go to that ITS marketplace where that best price is, 16 whether it is a penny, whether it is two cents, whatever it 17 is, it is the best price. 18 MR. HARRIS: Is it your opinion that the ITS 19 systems insures and helps produce the best price? 20 MR. QUICK: I think when you set out in 1975, that 21 it did probably accomplish that. I am not sure today that 22 ITS is the vehicle to continue on in the future. 23 MR. PETERFFY: I would like to say something about 24 ITS. 25 The ITS rules practically stated prohibit any 1 market to trade through the quotes of any other market. This 2 would make sense if all the quotes would be electronically 3 instantaneously reachable. But they are not. 4 These rules provide traditional exchanges to have 5 somewhere between 30 and 90 seconds to decide whether they 6 want to trade with an order or not. This is a free option 7 which is very expensive to the order provider. The 8 Commission should examine whether the exchanges and ECNs with 9 electronically accessible venues would be cut free from the 10 ITS quotes. 11 If the answer to the question would be affirmative, 12 these electronically accessible venues would form their own 13 ITS rules, and there are a sufficient number of them now that 14 if there would be a no trade through -- a trade through 15 prohibition among them -- they would certainly come up with a 16 fair price. 17 MR. HARRIS: Thank you, Tom. 18 We will undoubtedly return to ITS and the question 19 of speed of linkages a number of times today, but I want to 20 stay just a little further on best execution and return to 21 that question very quickly. Though, I am sorely tempted to 22 jump in. We will be in interesting territory very quickly 23 with best execution also. 24 Gunner, let me ask you, for institutional size, 25 does the Commission have any business worrying about best 1 execution or is this really just a retail issue? 2 MR. BURKHART: First of all, I found it rather 3 curious that when you asked initially did the responsibility 4 for best execution rest with the market structure and 5 exchange or did it rest with the broker-dealer, that, in 6 fact, I think that the responsibility for best execution most 7 squarely falls on the shoulders of the institutional asset 8 manager. And that's where the responsibility, I think, lies 9 today. 10 And the fact, as I see it, the plain truth of the 11 matter is, the current market structure in the United States 12 is in direct conflict with our fiduciary responsibility of 13 best execution. And that makes it very difficult for us. 14 MR. HARRIS: Your institution, undoubtedly, has a 15 very large buy side desk and you refer to the responsibility 16 of the trader that you employ to obtain best execution for 17 the firm? 18 MR. BURKHART: Correct. 19 MR. HARRIS: What about smaller institutions that 20 don't have buy side desks? Should they have them or can they 21 contract for that by going through a broker? 22 MR. BURKHART: Whether you have a centralized 23 trading desk or dealing desk or not, somebody at your end is 24 responsible for the execution of the order, whether it is the 25 portfolio manager or some type of administrative person, 1 somebody obviously licensed by the authorities. 2 By and large over history, most buy side firms have 3 outsourced the responsibility of execution to the 4 broker-dealer. I think that at least today, and the trend 5 going forward, the responsibility for and the control of the 6 execution and, therefore, by and by best execution however 7 you define it, is going to come firmly into the lap of the 8 buy side house. Whether it is with a centralized trading 9 desk or portfolio manager, I think the shift of the 10 responsibility is very apparent. 11 MR. HARRIS: Tom Gardner? 12 MR. PETERFFY: He is not here. 13 MR. HARRIS: I'm sorry, I can't see that way. I am 14 sure he will be returning shortly. 15 Minder? 16 MR. CHENG: Yes, I think I would just add to 17 Gunner's point. In terms of where the responsibilities 18 reside, the responsibilities associated with best execution, 19 there are some residing with the institutional trader, there 20 are some residing with the broker-dealers, and there are also 21 some residing with the exchanges. 22 I think that when there is a better quote that is 23 away from the NYSE, it is up to ITS, which I know we are 24 going to get into, I think perhaps up to the specialist or to 25 the broker-dealer at least to make sure that you do route 1 your order up to a certain size to a market where it has the 2 best quote. But buy side doesn't have control over that 3 unless we have direct access to the market, which we do in 4 some other countries, but not in the U.S. 5 I think the other point I would make is in terms of 6 incentives you can regulate. Among the three layers here, I 7 think you can regulate the stock exchanges and the 8 broker-dealer to do best execution, getting the best price, 9 but you can regulate that, but the incentive, though, is 10 mostly on the buy side because every penny we overpay for 11 what we have to pay normally for that order, it will end up 12 costing us performance, and it will end up really coming back 13 and biting us on performance. 14 So I think our incentive is perfectly aligned, with 15 or without regulations, we are going to do our best because 16 otherwise we are going to lose our investors. Any costs we 17 have to pay, we end up paying more, that will come back and 18 bite us. And the second point -- and I think if you don't 19 get the incentives to carry out best execution, I think it is 20 more natural on the buy side than probably more so than on 21 the sell side with the stock exchanges. Not that they won't 22 do it, I think that they will do it, but suppose you don't 23 get the price best price on an agency basis on the sell side. 24 At least you just pass the cost on to the buy side. Again, 25 not that that's the case, but I think that if you look at the 1 design of the game, it is much more in line with the buy side 2 interest. 3 The second point is the difference between retail 4 investors versus institutional traders, it is the size, 5 again, we talked about this morning. Getting the best 6 execution for the buy side entity, it means getting the best 7 price for the overall basket, for a million shares as opposed 8 to getting that for a hundred shares at that moment. So we 9 are happy to -- Scott mentioned it earlier, Gunner, too -- 10 that we are happy to give up price improvement so that we can 11 actually get a large size done. To us, that's more important 12 than getting best execution for that hundred or two hundred 13 shares. 14 Coming down to it, if you look at the makeup of the 15 transaction cost we end up paying, for retail investors, 16 typically the spread, which is why it tends to help retail 17 investors more than to us. Because a large part of our key 18 cost comes from market impact, which cannot be really 19 displayed through the NBBO, so I think that's the difference 20 I would highlight. 21 MR. HARRIS: Paul, the Chicago Stock Exchange, if I 22 am correct, represents a fair amount of the retail order 23 flow. Listening to Gunner and Minder it sounds like they 24 would, at least for the institutions, place a fair amount of 25 responsibility for best execution in the hands of the buy 1 side. 2 Is the retail able to handle that and should they? 3 MR. HERRON: Dave Herron sitting in for Paul in 4 this session. 5 MR. HARRIS: Sorry. 6 MR. HERRON: That's okay. 7 Certainly Chicago, like New York is a blend of 8 institutional and retail flow. We do represent and provide a 9 lot of liquidity for a lot of internet brokers and a lot of 10 smaller customers. 11 I think from the point of view of best execution, a 12 broker-dealer has to be responsible for best execution for 13 the customer's order. Best execution can be different things 14 to different people. It's a combination of facets including 15 speed, effect and spread, price improvement, liquidity and 16 cost of execution. 17 And I think as each broker-dealer fills their 18 business model and chooses which markets to go to, they are 19 able to compete for that customer business and the customers 20 will vote with their commission dollars as to what is the 21 best model for them. 22 The exchanges also are required, I think, to 23 provide for best execution, making sure that the specialist 24 community and the broker community are, indeed, looking out 25 for the best interest of their customers. I would agree with 1 Chris, that the marketplaces have to be designed to put those 2 customer orders together. And if they don't, the system 3 breaks down. 4 MR. HARRIS: Janet, your Archipelago has an 5 extraordinary rich set of linkages. 6 Can you describe to us the importance of those 7 linkages in terms of how they assure best execution, the role 8 that they play? 9 MS. ANGSTADT: Sure. 10 I think one of the most important aspects of our 11 linkage is that it is choice. There is a way to reach just 12 our book with an IOC or a P&P order time, so if you do wish 13 to reach outside of Archipelago, you can do so. We will 14 forego an execution on our system to get you a better price 15 outside our system. 16 I think one of the points from this morning was 17 that we talked about how important the NBBO was, and access 18 to that NBBO is often a difficult thing. One thing we have 19 tried to build into our system is direct access through 20 direct phone lines, if possible, to avoid those things that 21 create delays. 22 MR. HARRIS: Should all brokers have the same sort 23 of linkages that you have to everybody? Or is that necessary 24 or perhaps maybe not surprising that you think it is 25 sufficient that they come to you? 1 MS. ANGSTADT: Absolutely. Very good point, Larry. 2 MR. HARRIS: I didn't think through my question. 3 There are a couple of ways to do linkages. Go 4 ahead, Matt. 5 MR. DeSALVO: I think the simplistic way to go is 6 not as simplistic as it seems when you have markets that have 7 different access standards and, in addition, within those 8 markets you have certain ATSs that have access fee changes, 9 you can look at a $10 bid and because it may take 30 seconds 10 to respond versus less than a second, or there is an embedded 11 access cost to that $10 bid. 12 So when we have unequal playing fields with regards 13 to the access standards and access fees, you cannot determine 14 price as being a standard for best execution. And I think 15 what a lot of us are trying to say is that if we are going to 16 fall back and answer the question of best execution if we 17 don't even address equal access standards and $10 means $10 18 to everybody, than you cannot get to a determination of price 19 as a standard for best execution. 20 MR. HARRIS: Should price be a standard for best 21 execution? For what sort of customer should it be? 22 MR. DeSALVO: Price at NBBO can be one standard for 23 one type of order. But, again, we don't have agreement. You 24 know, a million share order at best execution at an inside 25 market looses its meaning, but if we can't even determine 1 that $10 in one marketplace is equal to $10 in another 2 marketplace, it is very difficult to even go beyond the 3 meaning of best execution at that point. 4 MR. HARRIS: Let's suppose that we are talking only 5 about small orders. 6 How would you construct the market system so that 7 price would be meaningful across exchanges or across centers? 8 MR. DeSALVO: I would say that at a minimum, a 9 price in one market should be equal to a price in another 10 market and for participants price should be the same. When I 11 access a $10 bid, it should be $10 to me, not $10 plus an 12 access fee. And that should be true for all participants 13 whether it is direct access on the buy side or the sell side. 14 In addition, there should be a minimum level of 15 access standards between markets. That if the minimum is 16 less than two seconds or less than five seconds for some 17 level, that is equal and measurable between markets, that 18 there is equal access on price and equal access on turn 19 around response times. At that point, at least you have a 20 skeleton in which to draw markets. 21 MR. HARRIS: Specifically, what plumbing would you 22 install to do that? Would you require high speed linkages to 23 market centers? 24 MR. DeSALVO: I would allow the market participants 25 that they can either do it directly themselves or if they 1 want to go through an ATS or if they want to build a network 2 themselves. 3 We are indifferent as to which approach, but the 4 standard when you get there, should be the same. How you get 5 there I don't worry about. 6 MR. COLBY: Morgan Stanley retail customer orders, 7 are you bringing them out to other markets or are you 8 typically trading against them as principal and matching the 9 prices? 10 MR. DeSALVO: Both. 11 MR. COLBY: Pete? 12 PROFESSOR KYLE: I want to say that this issue has 13 already been kind of dealt with in a way that I suppose has 14 been partially successful. After the 1987 stock market crash 15 remember that NASDAQ was totally criticized because of the 16 bids and offers that they were posting. 17 It turned out it appeared to be completely 18 meaningless, so the SOES system was introduced after that. 19 And the SOES system is a type of system which is your bid and 20 your offer has to mean something and there is also a 21 mechanism in place where somebody can punch a button and the 22 person who punches the button first actually gets an 23 execution with a reasonably quick turn around time. Those 24 bids and offers actually makes them equal. 25 And if you have a ticker that has different numbers 1 displayed on it that aren't meaningful, you are just opening 2 yourself up for the same types of criticism that NASDAQ got 3 after the '87 crash. 4 MR. HARRIS: So you would have high speed linkages, 5 then? 6 PROFESSOR KYLE: Yes, basically, a high speed 7 linkage. It doesn't have to be -- we are talking about a 8 hundredth of a second, a tenth of a second or one second. 9 Those are the kinds of time frames I am talking about. But 10 one minute on the stock market is far, far too slow. 11 It basically is something where you punch a button, 12 it is waiting for the next person to punch the button and 13 then you have execution and very prompt confirmation that I 14 punched the button first and you punched it about the same 15 time. 16 MR. DELLA ROSA: When the question of best 17 execution came up, the definition of best execution, there is 18 always a tendency when talking about best execution to try to 19 have a simplistic view of how to think about it. We all 20 strive toward that. 21 It is obviously a very complex issue, but in the 22 small orders, the orders that are low in size that are 23 displayed in the marketplace within the NBBO, that there is a 24 place where if you become the closest to the black line is 25 what is best execution. However you go to where the market 1 looks today and you go to where the functionality of the NBBO 2 is, there are two perspectives. 3 Number one, a standard of connectivity; access, if 4 you will. Number two, is equivalency of price. Is the price 5 that is displayed the actual price that one would see. If a 6 person in small orders could actually have a definition by 7 what is posted and how quickly you try to achieve that within 8 the NBBO, then you need to have those two things in place. 9 Equivalency in price where you can actually see 10 what exists is number one. And number two, is access equal 11 at the price that is shown? If it is not, then you sacrifice 12 that. So on the small orders, orders below the sizes that 13 are actually displayed on the NBBO, we need to accomplish 14 that. 15 MR. HARRIS: So, Joe, once again, what would the 16 plumbing look like? How do you guarantee access and how do 17 you guarantee price? 18 Would it be high speed linkages again? 19 MR. DELLA ROSA: It is certainly high speed 20 linkages. The question is who provides them or what is the 21 process for their provision? 22 And if, in fact, the NBBO is that black line 23 standard for execution; if, in fact, it is that, because if 24 it doesn't have those two ingredients, again, it doesn't 25 deserve to be that. But if that's what our goal is for a 1 standard, I think that the NBBO is really important to have 2 reference in the marketplace. So that if, in fact, we hold 3 to that standard on the small orders, and I think that those 4 who post in that are required to post in that, there has to 5 be a level of standard of access that is equal so the prices 6 that are displayed in that are equal also. 7 MR. HARRIS: Chris, so there have been a couple of 8 arguments for high speed linkages. We don't have to accept 9 those arguments, but let's assume that we were to go to that, 10 what would we sacrifice at the New York Stock Exchange by 11 imposing high speed auto-ex response on the New York Stock 12 Exchange? 13 MR. QUICK: I think with respect to the NBBO, the 14 exchange will Dual Quote, we are going to automatically NBBO 15 going forward after the Commission decides on whether or not 16 the Dual Quote will exist. 17 So the inside quote will be automatic, will be 18 accessible to everyone. The outside quote will be a binding 19 quote for the specialists who all have interests aggregated 20 on the price point on the NBBO to that point where you are 21 bidding on depth. 22 In terms of linkage, linkage in the exchange, you 23 can come through the front door of the New York Stock 24 Exchange right now for about $2 million as a member. But you 25 can come through a lot of other ways to the New York 1 Exchange, as a clearing agent, as a clearing customer of the 2 numerous people who offer clearing. And we do all know how 3 easy access is. 4 You talk about direct access to the floor of the 5 exchange not being a member of the exchange, so linkage is 6 not a problem getting into the New York Stock Exchange. 7 MR. HARRIS: Assuming that the current linkages, 8 which for the listed side is primarily ITS, assuming that 9 that's not changed, how vigorously should the Commission 10 pursue enforcing trade through rules in the decimal 11 environment? 12 Any opinions about that? 13 MR. DeSALVO: I think you answered it already by 14 allowing the trade through option. 15 MR. COLBY: The nine month ticker, right? 16 MR. HARRIS: In any event, everything at these 17 hearings is on a blank piece of paper, so did the Commission 18 make a mistake when they did that? 19 MR. DeSALVO: Yes. 20 MR. HARRIS: What was the mistake? 21 MR. PETERFFY: Twofold. 22 MR. HARRIS: Tom? 23 MR. PETERFFY: I understand that somebody made a 24 study as to how much the free option would cost, but the fact 25 of the matter is that listed stocks have much higher trade 1 through than ATSs. So if you made a similar study for listed 2 stocks of the free option, you would find a higher number. 3 MR. HARRIS: So we should, if we are going to grant 4 -- you would suggest that we grant -- if we maintain the 5 existing ITS system, you would suggest that we grant a larger 6 de minimis exemption for listed stocks as opposed to ETFs? 7 I want to make sure I understand what you are 8 saying. 9 MR. PETERFFY: I do not think there should be any 10 exemption among ECNs or electronically accessible markets. 11 In other words, you should not allow any trade through 12 whatsoever in a market that provides immediate electronic 13 executions and immediate access. 14 The problem is that the manual markets change the 15 electronic markets into a situation where they simply come to 16 a grinding halt when they want to trade through and the 17 traditional market has its quote in the way. 18 MR. DELLA ROSA: I don't know how you can address 19 trade throughs until you get to price equivalency and equal 20 access. I don't know how you address trade throughs before 21 that. 22 MS. ANGSTADT: How do you have equal access when 23 you have floor-based models and electronic models, and I 24 think that one of the issues that the ITS is struggling with 25 right now is bridging the gap between the two. 1 One of the proposals that they have is to allow the 2 automated market to trade through the manual markets. And 3 with respect to the ECN electronic markets, the trade through 4 rule will still stand, although probably with an exemption 5 that it be sized for that particular stock. 6 So I think part of it is it is not just the 7 linkage. What is access, and when you are an automated 8 market and you are trying to access some manual market, what 9 does that mean? 10 MR. DELLA ROSA: I am not disagreeing with that, 11 actually. 12 MS. ANGSTADT: But I think that's a good next step 13 in terms of linkage. And then the next step is access, what 14 does access mean for each individual market? 15 MR. COLBY: In the earlier discussion, Gunner and 16 Minder said that as a precondition for them putting up limit 17 orders there would have to be some sort of protection, some 18 sort of price and time protection. And there is, obviously, 19 price -- I mean, time protection, and now we are talking 20 about whether there should be price protection. 21 So what would be the consequences for the people 22 that are displaying a limit order in one market if other 23 markets can freely or semi-freely trade through that price, 24 trade through an inferior price to the limit order that they 25 are displaying? 1 MR. QUICK: Does that go back to a problem we had a 2 few years ago with the NASDAQ market? 3 MR. COLBY: Well, which of the many? 4 MR. QUICK: Trade throughs. 5 You are, again, putting the dealer in front of the 6 customer. And, you know, in another life I was part of Quick 7 & Reilly Group, which is a large retail complex. We never 8 had a customer who would call us and say they didn't care 9 about the best price. 10 You are, again, putting the dealer in front of the 11 customer with de minimis ETFs exception, and it is to the 12 advantage of the dealer. 13 MR. DELLA ROSA: Trade through at what level? 14 Trade through at the level of the New York Stock 15 Exchange or the NBBO? Is it's at the level of the NBBO, then 16 that's where -- what level are we talking about? 17 If it is the NBBO, then these other pieces have to 18 be in place at the level of the NBBO? 19 MR. LEE: Can I interrupt? 20 This problem seems to me, accepting the price is 21 the key criteria for what determines best execution and then 22 ensuring that price is the price, whether it is an access fee 23 or whether it is accessibility to it. 24 And I want to step back a little and question 25 whether the price should be the sole criteria. And if it is 1 not, then I want to know how should brokers be held liable 2 for their actions, to what benchmark? 3 And one way for it, and this hasn't been addressed 4 properly, but I think the Commission may want to think about 5 this, is to have it much more contractual. Maybe I, as a 6 client, I go to my broker and I say, "I want speed," so then 7 that broker will be held accountable to speed. Whereas I, 8 another client, go to my broker and say, "I want price." So 9 that is very different from the current situation where you 10 have really a single benchmark. 11 And I am wondering what people feel about that, 12 because I certainly feel that there is no single criteria for 13 best execution. 14 MR. HARRIS: For a small trader, is there any 15 reason that the small market order trader couldn't have both? 16 That's the issue we are debating here I think, can 17 a small market order trader have both speed and price? 18 MR. LEE: It depends upon the market structure. I 19 can envision a situation where I might want to sacrifice 20 speed for price. 21 MR. HARRIS: And, hence, these hearings. The 22 question is, what do we trade off in doing that? 23 MR. HERRON: I think in the current market 24 environment the small customer does have both. They do have 25 speed, prices and executions with price improvement being 1 offered at many venues. 2 So a manual market doesn't necessarily mean a 3 slower execution for any manual market at the table or any 4 industry, so the customer does get speed and they do get best 5 price. 6 MS. GLASSMAN: From a practical perspective, if the 7 customer had a choice and it was contractual, how would one 8 monitor that, that they were complying with that? 9 MR. LEE: That's your problem. 10 (Laughter.) 11 MR. LEE: But my point being that if you want a 12 simple answer, as you currently have now, then you don't 13 allow for multiple criteria. 14 And by not answering that question, you are 15 answering it. You are not allowing multiple criteria. You 16 can't say best execution is only one at a time. 17 MR. DeSALVO: Ruben, you haven't been on the other 18 end of the phone of someone who selected speed and didn't get 19 price. So when they choose speed, they still mean price. 20 MR. PETERFFY: I think that there must be some 21 confusion here speaking about speed and price. The only 22 reason we go on speed is because they worry that the price 23 will move away from them, so it is price they are into; only 24 price, not speed. 25 Best execution certainly means the best aggregate 1 price for the order. There are small orders, and it is easy 2 to see that it matters for small orders, yes. Large orders 3 are basically hedged by the dealer. The dealer makes the 4 large deal and then assembles the other side of that trade 5 from many small orders. And he can only do that efficiently 6 if he has electronically accessible venues to do it. 7 MR. HARRIS: Tom, I wanted to follow-up with a 8 comment you made earlier when you said that you believe it 9 would be sensible to enforce a trade through rule on all 10 systems that have high speed access. 11 Would you extend that to require that all market 12 centers have high speed access? 13 MR. PETERFFY: I think market centers should be 14 free to have any access they wish to provide. 15 MR. HARRIS: Does anybody else want to comment now 16 on the question of speed with best execution as it relates to 17 the market orders? 18 We can return to this later. 19 MR. HERRON: Just to point out that any attempt to 20 weaken the trade through prohibitions would seriously weaken 21 the national market system. I think the ability of the 22 customer or dealer to display an order in one marketplace and 23 know that another marketplace is obligated to come to that 24 order or match that order is key to a strong national market 25 system. 1 MR. HARRIS: What are the strengths of the current 2 market system that you are afraid that would weaken? 3 MR. HERRON: The ability of various participants in 4 the national market system now to continue to compete in that 5 environment. 6 The national market system is a combination, I 7 think the listed side is a combination of the best of the 8 auction and dealer markets in that price discovery is 9 permitted; yet, there is also the competition provided by 10 multiple dealers. 11 MR. HARRIS: How does the trade through rule 12 support your position? 13 MR. HERRON: If a trader in our marketplace is not 14 confident that by displaying his bid or offer, or if the 15 customer is not confident that by coming to our marketplace 16 that our bid or offer didn't have some price priority, you 17 might see all that, it would weaken the ability of the 18 competing market. It would weaken our ability to attract 19 order flow if they thought that displaying in our marketplace 20 would not allow them to get an execution. 21 MR. HARRIS: In order to preserve the trade through 22 rule, what linkages do you need? 23 MR. HERRON: I think we are in a position where we 24 have to take a look at the rules of the ITS trading system. 25 I think the trading system itself is very dynamic, it works 1 very well. I think the rules perhaps allow for too much time 2 to go by before execution have to be addressed. 3 I am not sure we can get immediately to where we 4 need to go, but perhaps the idea of auto execution, for 5 markets to provide auto execution is as far as it goes. 6 MR. DeSALVO: Larry, I think in general we ask a 7 lot of questions with regards to linkages, and I think what 8 we are trying to attempt to say is don't worry about 9 linkages, but worry about standards and let us worry about 10 the linkages. Standards are more important than linkages. 11 MR. HARRIS: So what standards should we be looking 12 for? 13 MR. DeSALVO: We addressed it before when I think I 14 said and Joe said, and we have beaten it around a lot, price 15 is price and turn around times have at least a minimum 16 threshold with which to participate to warrant your price 17 being presented in the NBBO. 18 If I am a market center and it takes two minutes to 19 get to me and back, my market center bid or offer should not 20 be represented as the NBBO. If you want your bid or offer to 21 be represented in the NBBO, you should be subject to a 22 minimum standard, access standard, and not worry about 23 linkages required to get there. 24 MR. HARRIS: Chris, when Matt speaks about an 25 access standard here, and we are certainly not talking about 1 membership at the New York Stock Exchange, he is saying I 2 have an order, I see your quote, I want it and I want it now. 3 How quickly should he be -- I mean, if he goes to 4 you, is he going to get it? 5 MR. QUICK: We have a lot of venues for access and 6 quote. You have institutional express, which you can debate 7 whether it is a success or not, but, unfortunately, the 8 participants don't know when their quote becomes accessible 9 because the vendors have not pointed that out on their 10 screens. So, yes, the quote is accessible right now only if 11 they go back and find it accessible. 12 MR. DeSANO: That's after what time frame? 13 MR. QUICK: That's after 15 seconds, Scott? 14 MR. DeSANO: That's right. 15 MR. HARRIS: Let me follow up. 16 How do traders find if a market is displayed 17 because it is not there quite often? 18 MR. DeSANO: I understand the concept of the 19 institutional express, but there has been 15 trades in it, I 20 think, this year. And that's the reality of it, and I know 21 that there have been problems with respect to that, but I 22 don't even know why we talk about it. 23 But very often we do try to go after quotes. You 24 can see them but you can't touch them, it is very 25 frustrating. 1 MR. HARRIS: Gunner, you have something to add? 2 MR. BURKHART: Absolutely. 3 And I find the discussion interesting from a lot of 4 angles but I think the primary angle is that we are talking 5 about linkages. We are talking about the justification, if 6 you like, of the NBBO. We are talking about access; yet, we 7 are doing so in the context of trying to justify the existing 8 very fragmented market structure. 9 I think we have fairly unanimous opinions of buy 10 side participants, large and small, that what we need is a 11 single marketplace where you know what the best bid and best 12 offer is that you can access immediately. Trying to link 43 13 different platforms around the United States and trying to 14 figure out who should pay for that and how do I access that 15 best bid and offer, is that actually going to be there when I 16 finally find that ultimate player is not the real issue. 17 It is the issue if we are accepting that the 18 current market structure is the correct one. In other parts 19 of the world, the different market structures apply and they 20 are far more efficient and you get a single point of 21 liquidity for all the players, single and institutional 22 investors as well as retail players, but that's a far more 23 profound discussion than I think you are willing to have 24 today. 25 MR. HARRIS: I think we are willing to discuss 1 anything today, and I would like to pursue that possibility. 2 Clearly, the history of that proposal is not good. We won't 3 spend a lot of time on it, but I would like to ask one 4 further question. 5 Is it your expectation that Archipelago would 6 survive in that sort of environment or how would you get from 7 here to there? I know what you just described would either 8 put Archipelago or the New York Stock Exchange out of 9 business or both of them, and the question is would that be 10 desirable, and, if so, how would it be done? 11 MR. BURKHART: First of all, I am not looking to 12 put anybody out of business. I am looking to help construct 13 or aid in the construction of the most efficient marketplace 14 for the preeminent economy in the world. 15 And I think what we have right now is far from 16 that. There are a lot of, I would say, highly incumbent, 17 very politically strong structures currently in place which 18 are not necessarily for the benefit of the end user, either 19 the retail investor or the institutional investor. And I 20 think we need to seriously consider whose interests are we 21 really looking after in this respect. 22 When we talk about all these different angles in 23 the discussion, let's get back to the primary principles and 24 the five ideals we laid out initially. And I would say this 25 again, getting back to my earlier point, the current market 1 structure in the United States is in conflict with trying to 2 obtain these five ideals or principles. It is not about 3 putting somebody in or out of business; it is developing a 4 common goal. 5 It might mean changes to the Archipelago model, to 6 the New York Stock Exchange model, to the NASDAQ model. 7 MS. ANGSTADT: I think our model is just a response 8 to what makes life easier for people who are trying to seek 9 that best execution. Whether or not it is Archipelago 10 providing a routing platform or we could simply be a limit 11 order book and just be one destination, so I think you can 12 either be a member or a subscriber to all of the pools of 13 liquidity or you can go to one marketplace or a couple of 14 marketplaces that are reaching out for that liquidity, 15 because I think the alternative is you are building that 16 order routing system to actually get to those places that 17 have better prices. 18 MR. CHENG: I think -- just one more point here -- 19 Gunner and Scott and myself, we all have one thing in common. 20 We all look after trading globally. And one observation that 21 has been shared, although not explicitly and not officially, 22 but by a lot of us in the market on the buy side, from just 23 looking at how each market has really been structured, one 24 observation we can make about the United States is that it 25 has very good liquidity, but the market structure does not 1 let that liquidity somehow find its way to interact with one 2 another. 3 And the thing about -- and I think Ruben's and 4 Gunner's comments were based in London -- with European 5 markets, or the London market and the European, they have 6 perhaps better market structure, although liquidity may not 7 be as good as in the U.S. market. So there you have it. 8 I think you need the liquidity to drive the market 9 structure, but also you need market structure to give you the 10 way to have liquidity with one another. Whether the NYSE or 11 Archipelago or other ECNs will survive, we don't know. But I 12 think what was clear from this morning and the previous 13 points, I think that if there is a way to somehow have a 14 centralized book that retail investors and the institutional 15 investors can access, I think that that appears to be a 16 better structure. 17 Whether it is going to be through one ECN or one 18 stock exchange, I think that that's probably secondary. 19 MR. HERRON: Isn't it possible that liquidity is 20 driven by the market structure we have and that by changing 21 that market structure, we take that liquidity out? 22 MR. DOMOWITZ: No, I think a variety of studies 23 have shown that across various market structures, the 24 liquidity doesn't really change, all things equal. 25 When you get right down to it, investors provide 1 liquidity; markets do not provide liquidity. That's a core 2 concept. Now, getting back and linking that to what Gunner 3 and Minder and I think others are talking about, on the one 4 hand it is corporate dictum, you can't manage or you can't 5 control, and everybody here is talking about gaining control 6 one way or another. The issue is control for whom, all 7 right, and on what basis. 8 With that in mind, I want to echo I think what Matt 9 said, first of all, as a policy prescription, talking about 10 the precise linkages, even talking about the notion of 11 complete consolidation as though that's where you should 12 actually set policy, I think that that's incorrect. 13 I think that the issue is standards. I buy 100 14 percent into Joseph's comments about what are two equivalent 15 prices. We are talking about access in terms of time and we 16 may be talking about access in terms of money. We may find 17 that competition, okay, amongst the market centers basically 18 gets rid of this issue about a price is a price depending on 19 the access. That may or may not be true, but timing can be 20 helped by setting standards. 21 Now, I know that in the end we all operate 22 according to a set of standards. Starting with the SEC, I 23 know when, viewing Gunner as a client, viewing Minder as a 24 client, when they outsource to ITG as a broker, they are 25 setting a particular standard, and basically, they will come 1 back and ask us to analyze ourselves if they don't do it for 2 us. In fact, we are asked to provide the tools to analyze 3 ourselves with respect to performance. And since they have 4 access to the tools, we can't exactly give them the rigged 5 deck. 6 So it is the standards that are important here. I 7 am not trying to avoid the devil in the details, and I think 8 that is the wrong way to respond to this, but getting too far 9 down so you are actually talking about individual elements of 10 market structure is not the way. The market has actually 11 been very good in certain respects. When we talk about trade 12 through rules, you know, the speed within the electronic 13 marketplace actually has been fast enough, leaving the manual 14 markets aside for a second, they have done pretty well. 15 People are actually, you know, it is not even 16 distinctive anymore to claim you have a so-called smart order 17 router that figures out where the price is, where the 18 liquidity is and how to get there. That's just taken as 19 given. I mean, if you are a decent broker, that's what 20 you've got. So the market solves these problems. At that 21 level, the market solves the problems. 22 Will the market actually produce complete 23 consolidation? If it doesn't, it will probably get close 24 enough to satisfy you if, indeed, the standards are set and 25 we operate accordingly. Because given those, Gunner will set 1 his, I will set mine and the retail investor will have a say 2 as well. 3 MR. HARRIS: The complete consolidation model has 4 been proposed by a couple here and whether it has a political 5 future or not is perhaps a different issue, but I do want to 6 give Chris an opportunity to tell us what is happening in 7 front of him on the floor of New York Stock Exchange that is 8 valuable. 9 And the reason I raise the issue is because that 10 complete consolidation model isn't inconsistent with your 11 market structure. If your market structure is to be 12 preserved, it's probably because it has some value and I 13 would like you to tell us what it is. 14 MR. QUICK: I agree with Ian that customers want 15 liquidity, and especially when you put customers trading with 16 customers. That creates liquidity. 17 On the New York Stock Exchange we put customers to 18 customers 85 percent of the time, no dealer intervention. 19 That's what creates frictionless trading on the New York 20 Stock Exchange, no displacements, volatility profiles are 21 lower than other markets, This is what we have been selling 22 all these years, continues -- it continues with innovation, 23 it continues with technology wrapped around it, but the New 24 York Stock Exchange is no different than any trading room on 25 Wall Street. 1 The four biggest trading rooms on Wall Street are 2 the New York Stock Exchange, Goldman, Merrill Lynch and 3 Solomon Smith Barney. Everybody is behind that technology. 4 But the New York Stock Exchange, by not only being 5 responsible for both sides of the transaction, the sell side 6 and the buy side; not just adhering to one side of the 7 transaction while discrediting the other side of the 8 transaction. 9 So this customer to customer, the constant flow is 10 begetting liquidity, and that's why we do 84 percent of the 11 volume listed trades. 12 MR. HARRIS: I would like to now change the 13 direction of our discussion just slightly, still -- 14 everything we are doing, of course, is very highly integral. 15 When we talk about best execution, implicitly we 16 seem to always assume that we are talking about the execution 17 of market orders and people want to trade and the speed and 18 so forth, and we frequently give, at best, lip service to the 19 execution of limit orders. I would like us to give far more 20 attention now to the question of execution of limit orders. 21 First of all, I would ask, is it important that 22 public investors have an opportunity to offer liquidity 23 through limit orders? 24 And then secondly, if they do, what are the 25 implications of that value for the construction of our 1 markets? So is it important that public traders, whether 2 they be institutional or retail, have an opportunity to offer 3 liquidity to limit orders? 4 I know Gunner's answer, so... 5 MR. BURKHART: I will expand upon it, okay. 6 I think it is very important, I think it is 7 important to everybody involved in the markets that 8 institutional players, as well as retail players, feel 9 comfortable and, therefore, place limit orders on the market. 10 I mean, you want to have that liquidity present there, 11 whether it is fully transparent, what type of priority or not 12 that those limit orders -- you get into rebates and things 13 like that, that's a different discussion -- are given is a 14 different issue. 15 But, yes, we certainly want to be able to do that 16 because it benefits all of us. 17 MR. HARRIS: So the opportunity to offer liquidity 18 is a value that probably everybody agrees with. 19 The question is: Will people offer liquidity if 20 there isn't a meaningful opportunity to get hit? I assume 21 that everybody at the table recognizes that offering 22 liquidity without an opportunity to trade is a foolish thing. 23 So then the difficult question is: If we want people to 24 offer liquidity, or allow them to, and if they are to offer 25 it in a meaningful way, what mechanism should there be, if 1 any, to ensure that their orders somehow get filled? 2 Anybody willing to take that? 3 MR. CHENG: I am afraid to say, as I mentioned 4 briefly in the morning, with the current structure I think 5 that it would make it pretty difficult for the buy side, at 6 least, for the buy side to place large size limit orders, 7 although we very much like to. 8 And, again, the point really comes down to the lack 9 of protection or the opportunity of being picked off by 10 placing a large size limit order, even though by doing so we 11 are supposed to provide liquidity and reduce our transaction 12 costs for both us and the other side, but I think the lack of 13 protection is really hindering us from doing that. Hence, 14 again, not just using up fair trading a lot more, but also I 15 think I mentioned reserve book. 16 We use quite a few ECNs. I think Gunner and Scott 17 may be doing something similar. The way we place our limit 18 orders is through the reserve book. It is out there but it 19 is not going to be seen by everyone. It will only be seen by 20 servers who apply, so, in other words, we are providing 21 hidden liquidity that only will be tapped into by the others 22 like us. 23 MR. HARRIS: And the reward to providing that 24 liquidity is that somebody will somehow discover you, right? 25 MR. CHENG: Yes, discover but without the negative 1 consequences. 2 MR. HARRIS: Exactly. 3 MR. BURKHART: We are not asking for special 4 dispensation or rewards or incentives. We want to put 5 liquidity on the order book, whatever form the order book 6 might take. But we also have to have fairly iron-clad 7 guarantees that we are not being used for free option, which 8 is exactly what is happening today. 9 MR. DeSALVO: I think there is an additional point 10 with regard to this, when a broker-dealer presents a limit 11 order, whether it is from an institutional or retail client, 12 and our market is accessed, there is no access fee to access 13 that market. 14 When a limit order from the same client is 15 presented to an ATS, there is access fee. I think that 16 disparity between the two representatives is wrong. We 17 should either both be allowed to charge or neither should be 18 allowed to charge. But to have the same client give the same 19 limit order to participant A and Participant B, and one gets 20 the charge and one doesn't get the charge, that bifurcation 21 in the marketplace is unfair. 22 MR. DELLA ROSA: I think one level of incentive is 23 price-time priority. In other words, if you are there you 24 get priority just by exposing that order. 25 At the same time, I don't think it is the behavior 1 of traders in the marketplace or investors trying to execute. 2 If you display an order and behavior will still stay in place 3 even if the time price priority was in place, that one would 4 not display an order under the circumstances because market 5 information would be out there. 6 MR. HARRIS: So price priority ensures that 7 displayed limit orders will get executed no later than when 8 they've gone completely stale. Time precedence actually 9 gives them some incentive, to enforce time precedence across 10 exchange systems, you will have a routing range and we are 11 back to linkages again. I know where Gunner stands. 12 How does the Commission resolve these issues? Is 13 it sufficient -- we are going to get to access fees, but 14 let's go on the back road to access fees. 15 Presently a number of ECNs pay brokers in the 16 following ways; rebates on limits orders to execute. ECNs, I 17 know that Archipelago does it, NASDAQ is doing it now, 18 basically everybody is doing it. These payments are, in a 19 sense, very much like the payments for market orders. The 20 question is: Who should hold, say, a retail broker 21 accountable for their limit order routing decisions in an 22 environment where the retail broker is being paid for the 23 limit order? 24 Pete, how about you? 25 PROFESSOR KYLE: I didn't want to answer that 1 question directly other than to say that what is displayed in 2 a national NBBO should be net of the fees so that you know 3 what it is. 4 MR. HARRIS: We will talk about access fees for 5 sure. That's kind of a different issue. 6 I want to talk about the agency problems that are 7 associated with payments for order flow. Now we are talking 8 about payments for limit orders. 9 MR. DELLA ROSA: Doesn't this go hand in hand, if 10 you got the ability to charge, then you get a rebate. But if 11 you take that out of the quote, then it no longer exists in 12 the marketplace. 13 MR. DOMOWITZ: That is part of an overall agency 14 business model. 15 MR. HARRIS: I understand you can't pick it apart. 16 One of my difficulties -- 17 MR. DeSALVO: I think the agency business model can 18 be changed. The toll is collected now at the time of the 19 access to the market. It should be reversed so that the toll 20 is collected up front for the person placing liquidity on it, 21 then you could have equal access standards. 22 MR. HARRIS: We will talk about access fees and we 23 will do it in the next five minutes. But I want an answer to 24 this question: Who is responsible for best execution, or in 25 this case, best representation of limit orders? 1 Is it sufficient for a broker, a retail broker to 2 route their limit order, or its limit orders, to the ECN that 3 offers the highest rebate? 4 MR. QUICK: No, it is in direct competition with 5 what the customer wants. You are again putting the dealer 6 ahead of the customer. 7 Can we talk about the customer here? The 8 broker-dealer is responsible for achieving the best price for 9 his customer, ex-internalization, ex-payment for order flow, 10 ex-rebates from a broad data base. 11 MR. HARRIS: So how would this then be enforced on 12 a practical basis in the present system? Is IIAc1-5 and 6 13 sufficient to deal with the best representation issue 14 associated with limit orders? 15 MR. HERRON: They don't actually deal with 16 representation of limit orders. The broker-dealer is 17 responsible not only for trying to lower their cost by 18 getting perhaps the rebate from an order, but also routing 19 the order to a venue that can expect execution. 20 And if that venue is an ECN, depends upon another 21 ECN customer coming in, without any price protection against 22 other exchanges, that may not happen. But, again, the 23 customers can vote with their business. 24 MR. HARRIS: Dave, how do you get limit orders? 25 MR. HERRON: We get limit orders directly and from 1 various brokerage firms. 2 MR. HARRIS: Are those linkages, the order flows 3 associated, are there payments or liquidity rebates or 4 anything like that? 5 MR. DAVIS: There occasionally are. The payment 6 for it goes up to the specialist community, and we give our 7 specialist community a rebate based on share of market date 8 revenue. 9 MR. HARRIS: Which is an issue we have already been 10 through. I want to focus on access fees and liquidity 11 rebates. 12 MR. COLBY: Just to follow up, Dave, you get limit 13 orders from your customers, about 84 percent or some odd 14 percent of the volume in those stocks that are going off on 15 the New York Stock Exchange. So what do you do to help 16 assure your broker customers or their customers, if anything, 17 that sending a limit order when the buy is mostly taking 18 place elsewhere is a safe thing to do? 19 MR. HERRON: In order to compete for that business, 20 our specialist community guarantees to the customer that they 21 will receive an execution in our venue. So we don't rely on 22 simply displaying the order and getting filled. 23 It does require that oftentimes our community must 24 interact and go out aggressively to other communities to make 25 sure they that don't get caught behind the eight ball when 1 they fill those orders, but our community, it is much as the 2 New York Stock Exchange, when we interact with our customer 3 orders it is a matter of risk and reward, and oftentimes we 4 are interacting with our customers with prices and at times 5 we don't particularly want to, but we need to do that to keep 6 our customer base happy. 7 MR. HARRIS: I intend to keep my promises -- 8 MR. BURKHART: Before we get into that, can I just 9 ask a question. 10 Why do we allow payment for order flow? 11 MR. DOMOWITZ: Before we answer that, can I just 12 sort out whether we are talking about the payment for order 13 flow in the sense that once upon a time we used to talk about 14 it, which is really payment for market orders, or whether we 15 are talking about limit order, you know, driving the about 16 quote rebate. 17 Which one are we talking about? 18 MR. BURKHART: When we answer that question, the 19 follow-up question, who actually gets the money? 20 MR. DOMOWITZ: Can I answer that question? 21 MR. BURKHART: Hypothetically, if I have an account 22 with Charles Schwab, let's say, and the Chicago Stock 23 Exchange is paying me to give them limit orders, is that 24 rebate somehow coming back to me in the form of a rebated 25 commission? 1 MR. HARRIS: As I understand it, presently it is 2 going to Schwab and Schwab, presumably, is giving you better 3 service for lower commissions as a consequence. 4 MR. BURKHART: Isn't that a very curious motivator? 5 MR. PETERFFY: Indirectly, it does go back to you 6 because brokers all have to compete, yes, all have to compete 7 on price -- 8 MR. BURKHART: Competing on price is one thing, 9 competing on how much rebate is different. 10 MR. PETERFFY: To the extent we receive a rebate, 11 we can provide a lower commission. 12 MR. COLBY: If there was equal visibility for every 13 element, that would probably work out fine. The concern, of 14 course, is that it is much more difficult for a normal 15 customer to observe whether they get a good execution than it 16 is to figure out the commission. 17 MR. HERRON: I think the customers are much more 18 sophisticated than you think. 19 MR. COLBY: Well, I hope they are. 20 MR. HERRON: Based on my experience at Charles 21 Schwab and in the regional exchange community, they are very 22 sophisticated. They know exactly what they are getting. 23 MR. COLBY: Here, here. 24 MS. ANGSTADT: One of the things I wanted to 25 mention about the ECN rebates is that the distinction is 1 pretty clear between the payment for order flow and the 2 rebate in that when people are paying for orders it is 3 because they believe they can make a lot of money on that 4 order flow, I presume. And with our rebates, one of the 5 reasons we are happy to give someone their money back on the 6 transaction fee is because we have actually saved by being 7 able to cross those orders internally. 8 So our costs for writing out are much higher and we 9 have to charge for that. But, obviously, when we can match 10 internally, it is a cost savings to us and we want to pass 11 that along to people as opposed to just make more money on 12 those orders. 13 MR. BURKHART: The net result is the same, though? 14 MS. ANGSTADT: It is, and I think the difference 15 is, though, if I sent my order as fiduciary -- if I sent my 16 order to an execution point that doesn't charge very much, 17 like an ECN, as opposed to calling somebody who may work the 18 order for me at whatever the desk prices are, who is getting 19 the cost savings for going to the electronic venue and the 20 prices are lower. So it is a difference in the cost. 21 MR. HARRIS: Gunner asks some great questions and 22 even in asking questions is making points -- his points. 23 MR. BURKHART: I go back to London tomorrow, don't 24 worry. 25 MR. DeSALVO: I would take exception to the ECNs 1 not charging much. Our bills, we pay tens of millions of 2 dollars in ECN fees annually. They may not be large on a per 3 transaction basis, but like everything else, they add up. 4 MR. QUICK: I have to question whether the better 5 price is on electronic venue. 6 You know, that's a statements that is fairly -- the 7 better price on electronic venue, I disagree with, and the 8 stock exchange takes exception to that. 9 MR. HARRIS: Let's talk a little about access -- 10 MR. GOLDSCHMID: Larry, let me follow that up a 11 second. 12 What are the data out there now in terms of the 13 price packages of fast exchanges at the New York Stock 14 Exchange? 15 MR. QUICK: What, I'm sorry? 16 MR. GOLDSCHMID: In terms of execution. 17 MR. QUICK: In terms of execution, we have a 18 product right now if the customer wants speed, it is called 19 NX. They can access either side of the NBBO with an 20 automatic execution that doesn't get exposed to the auction 21 process. 22 But in terms of the auction process, 57 percent of 23 the orders that do come through the New York Stock Exchange 24 do get price improvement, 37 percent of which get the price 25 improvement more than half of the spread of the quote. 1 MR. PETERFFY: The problem with NX is that it 2 basically limits you to a thousand shares. 3 MR. QUICK: Because it is a product designed for 4 the retail customer, not professionals, Tom. 5 MR. PETERFFY: When do professionals get best 6 execution? 7 MR. QUICK: You need to come to the auction through 8 the front door. 9 MR. PETERFFY: Easy to say. 10 MR. BURKHART: For $2 million, Tom, you, too, can 11 play. 12 (Laughter.) 13 MR. HARRIS: Access fees, okay, so that everybody 14 knows what we are talking about, Some market systems like 15 NASDAQ, ECNs, charge a fee for market orders that are 16 executed. Many of these systems are also rebate fees in the 17 form of a liquidity rebate. 18 So the question is: Is this sensible? What I will 19 do is paint a very quick picture and then ask our panelists 20 to respond to this picture. Is there anything wrong with the 21 following story: I am an ECN, unnamed, limit orders provide 22 liquidity. My job is to aggregate the liquidity. I will pay 23 people for aggregating -- for coming and offering their limit 24 orders. That's the liquidity rebate. And in doing so, I 25 will create a liquid market that benefits everybody. 1 But, of course, I can't pay people if I can't 2 recover my expenses; therefore, I have to charge an access 3 fee. I have aggregated the liquidity, I have paid for it, 4 anybody who wants to use it has to pay an access fee. 5 We will talk later about comparing prices. I just 6 want to know is there anybody uncomfortable with this 7 argument? 8 MR. DOMOWITZ: Yes, the first one is you left 9 something out, which is it doesn't have to be an access fee. 10 We certainly pay ECN fees for hitting a bid. In other words, 11 if we come in and we take the liquidity, we pay a fee. 12 MR. HARRIS: Right. 13 MR. DOMOWITZ: So we don't need to differentiate 14 the types of fees. 15 Did you mean the access fee just to include any 16 liquidity? 17 MR. HARRIS: Yes, any liquidity for taking access. 18 PROFESSOR KYLE: I just want to say if a ticker is 19 meaningful, you can't have too much flexibility in the fees 20 and commissions that are kind of built into the ticker. 21 If the best bid and best offer are 4 bid at 5, and 22 I want to put a price through at 7, I can just have a 23 negative commission paid by one party and a higher commission 24 paid by the other party and basically cheat the system by 25 having a lot of flexibility in the fees that I charge. So it 1 seems to me, for the ticker to have some integrity, there has 2 to be some kind of assumptions made about the nature of the 3 fees and the commissions positive and negative that went into 4 the tick and out of the system. 5 And the same is true with the bids and ask. If I 6 am posting a bid or an offer, they should be posted in some 7 standardized way to kind of pence out fees that are not 8 common to everybody posting a bid or an offer in the system. 9 That being said, you usually eliminate the incentive for 10 people to cheat on the ticker by using fees to distort 11 prices. 12 On the other hand, you want to look at what 13 actually goes on on the exchange floor with the current 14 trading rules. That is kind of another thing, if they want 15 to, let's say, make market orders pay transactions fees and 16 let the limit orders not pay the fees, that may be a good way 17 to create liquidity. Liquidity reflected on the ticker 18 should be uniform and standardized. 19 MR. HARRIS: So you would have the SEC set a 20 standard that the industry can't rule itself? 21 PROFESSOR KYLE: Well, the ticker should mean 22 something, so if I look and see the ticker went out at 23 $75.43, I don't want it to wind up being the case that maybe 24 someone is paying a minus $1 commission and somebody else is 25 paying plus $1.10. 1 MR. HARRIS: Which would require somebody saying, 2 "This is the way we price things. You determine what the 3 prices are, but this is the way we do it"? 4 MR. DeSALVO: Well, you start off by painting a 5 picture of an unnamed ECN being able to withdraw liquidity, 6 post an order and recoup some of the cost by adding an access 7 fee. That sounds remarkably familiar to what we do every 8 day. And we also have to track order flow, we also post 9 limit orders, but we don't have that luxury. And we also 10 post principal bids and offerings similar to what other 11 people do around here, and we don't have that. 12 So why is there an unlevel playing field for doing 13 the exact same business model? 14 MR. DELLA ROSA: That also takes us back to the 15 equivalent pricing question. Whereas, in the marketplace you 16 have 1 to 10 bid, 1 to 10, and one offers a liquidity bid. 17 Actually, that is a 10 net bid. 18 Why isn't that a better bid in the marketplace? 19 Why isn't that displayed as a better bid in the marketplace? 20 MS. ANGSTADT: I don't understand why you are 21 making a distinction between transaction fees and access 22 fees. If, as a service, I am providing liquidity and I am 23 able to charge a transaction fee, why should I have to give 24 that cost away if somebody wants to reach in who is not a 25 subscriber? 1 It is the argument that New York has about 2 nonmember access, which, as I said before, I don't really 3 understand why you have to give it away. 4 MR. DeSALVO: I agree with you. Why do we give it 5 away, though? 6 If you access, if you go outbound for your client 7 because you are linked in NASDAQ to us or any of the other 8 broker-dealers here, and you access our limit order, do you 9 pay an access fee on our bid or offer? 10 The answer is no. But when I access your limit 11 order, the same limit order potentially from the same client, 12 potentially in the same stock, I pay you an access fee. So 13 what you are saying is that because you are an agency model 14 you can charge a fee, but somebody that represents both 15 principal and agent cannot. And I find that hard to 16 understand, how you can charge and I cannot. 17 MS. ANGSTADT: I am not suggesting that you 18 shouldn't be able to charge an access fee. 19 MR. DeSALVO: Thank you. 20 MS. ANGSTADT: I am just saying that access should 21 not have to be given away for free. What has happened in the 22 NASDAQ market is a blending of the agency and dealer markets 23 and the way we make money is different. And I think that 24 cannot be reconciled just by saying one group cannot charge 25 an access fee. 1 MR. DELLA ROSA: I think that from the point of an 2 equal playing field, and I think that you just stated that 3 you thought it should be equal, but I think from the 4 standpoint of equal prices in the marketplace, that's the 5 second question, as to what is actually displayed in the 6 marketplace and are they equivalent? 7 And in that case, they would not be equivalent, so 8 it is either you take that fee outside of the displayed price 9 in the marketplace -- 10 MS. ANGSTADT: I think the Super Montage has come 11 up with a great a way of handling this, and that is you have 12 two different algorithms. 13 MR. HARRIS: We will talk about that also -- 14 MR. PETERFFY: Ultimately, liquidity takers must 15 pay liquidity providers for the liquidity. Now, it is either 16 done by paying a spread or it can be done by creating good 17 competition by rebates for limit orders. The question is, 18 which is cheaper for the end customer? 19 And if you look at what is happening at the ECNs 20 and market order, lately you see that spreads are extremely 21 tight, so it is probably a good deal for the customer. 22 MR. HARRIS: How would that be the case? The 23 market order trader -- is it a good deal on net or just a 24 good deal on gross? 25 MR. PETERFFY: It is a good deal net. 1 MR. HARRIS: I don't see why that's the case. 2 Perhaps I am missing something. The market order trader 3 faces a very tight spread, which is a good deal, but has to 4 pay the access fee. 5 MR. PETERFFY: Right, but just to turn the issue 6 around, it is the broker that pays the access fee. 7 MR. HARRIS: Not if the broker pays the access 8 fee. 9 MR. PETERFFY: So when we execute for a penny on a 10 market order, then we have to pay 3/10th of a penny to the 11 ECN. On the other hand, we route a limit order, we get 12 2/10th of a penny back. 13 Now, where do we route the limit order? Do we 14 route the limit order to the ECN which provides the quickest 15 way to cancel, because we think that that is part of best 16 execution, to allow the customer to cancel when he wants. 17 MR. BURKHART: Maybe I just don't understand it 18 properly, but aren't the access fees passed along, directly 19 or indirectly, to the ultimate investor; yet, the rebates 20 seem to stop short of being passed along to the ultimate 21 investors. 22 MR. PETERFFY: No, they are not. We charge a penny 23 a share. That's it. 24 MR. HARRIS: Here is my problem with the entire 25 system, and perhaps, Janet, you can explain this to me. It 1 is unclear to me what determines the liquidity fee in the 2 end. 3 So we have four ECNs or exchanges trading the same 4 securities, each one feeling that they can get more order 5 flow by offering larger liquidity rebate, and each one offers 6 a larger liquidity rebate and they end up playing a game of 7 leap frog. Where does it end? 8 MS. ANGSTADT: I don't know. I think it is largely 9 being driven by competition. 10 I think, again, for us to have orders crossed in 11 our system is very efficient for us cost-wise, so that's why 12 the rebates are available. And that's something we are 13 always striving to do is to increase that cross trade so that 14 we work to increase that rebate, so I don't know where it 15 ends. 16 MR. HARRIS: Was the Super Montage set up right in 17 the following sense: You have a situation where you have 18 dealers who offer liquidity, and you have ECNs that have 19 customers who offer liquidity. ECNs charge access fees, and 20 so the issue is, what is the price? And issue that seems to 21 keep reoccurring over and over again. 22 And as NASDAQ is set up, they allow the customer to 23 choose whether they want to have their marketable orders 24 allocated according to price-time priority or price -- let's 25 see, the fee price-time, the access fee. The problem being 1 that the traders who submit marketable orders do not want to 2 be in a situation where they were going after best price but 3 had to pay a large access fee? 4 Was that done correctly? 5 MR. DeSALVO: I think that the premise of being 6 able to execute against those that don't charge an access fee 7 is wonderful, but you have to remember, you only get to 8 exhaust the liquidity from noncharging access members at some 9 point if you look to change the price one penny. You are 10 going to eventually have to go through access fee charging 11 ECNs that are there. 12 So that Super Montage algorithm works wonderfully 13 provided that there is not an access fee charging liquidity 14 in the marketplace. And by the way, you have to change price 15 point, then you have to pay the fee at some point. 16 MR. HARRIS: Let's explore some other standards. 17 One other possibility would be that you simply do everything 18 on a net pricing basis. And to do that we would have to 19 break the penny and go down presumably to hundredths or 20 something like that. 21 Another possibility would be to pass through all of 22 these fees and rebates to the customer so that the customer 23 saw everything. What are the pros and cons associated with 24 these alternatives? 25 Does anybody want to take a bite out of that? 1 MR. DeSALVO: I would vote against sub penny. 2 MR. PETERFFY: I don't think you can break the 3 penny, but you also cannot pass it on to the customer because 4 it is extremely confusing. I think we just have to live with 5 what we got. 6 MR. HARRIS: Well, we live with what we got, what 7 we've got right now are large access fees which seem to be 8 generating a fair amount of controversy, and liquidity 9 rebates that are roughly commensurate. 10 We have had several people speak for standards. I 11 presume the standard here would be something to the effect 12 that if you are a broker -- and ECN is acting as an exchange 13 facilitator, or whether you are an exchange, either way, that 14 your intermediation fee -- which essentially presently is 15 the difference the between access fee and the liquidity 16 rebate -- that the intermediation fee can only be collected 17 on an equal basis between buy and sell, or only on the sell 18 side, or only this way or that way, but not as the difference 19 between an access fee and liquidity. 20 So you might say there is zero liquidity, equal fee 21 on either side. Is that the standard? We have a lot of 22 request for standards today, the question is, is that the 23 standard that you envision? 24 MR. BURKHART: I don't see why we are complicating 25 the issue with all these fees. Why don't we do away with the 1 fees and you can charge a commission for trading? 2 If ECN ABC has the most efficient platform, great, 3 they have the best price, we will go there. If the New York 4 Stock Exchange does, great, we will go there. If it is a 5 consolidated platform one day, even better, we will go there. 6 But to allow for rebates to effectively buy order flow 7 because it understandably higher margin/lower cost business 8 for you, but then because you are paying for that order flow, 9 you then need to charge an access fee to offset it, you are 10 really confusing the client. 11 Again, the retail guy, the institutional guy, just 12 let them display prices. If you want order flow on the basis 13 of efficiency of your platform, wonderful, charge people a 14 commission to buy and sell from you. 15 MS. ANGSTADT: Just to be clear, the access fee is 16 not to offset the rebates that we are offering. It is when 17 people access us, there is more money there too and it is 18 maintaining routes into us, so our routing out and having 19 people access us is a higher cost to us as opposed to just 20 internalizing order flow. 21 MR. BURKHART: Maybe Minder, Scott and I should 22 charge Morgan Stanley and Goldman Sachs for our order flow. 23 MR. DOMOWITZ: Yes, one has to be a little careful 24 about how these charges are stated. I haven't seen any 25 documents for order flow, or I should say financial records, 1 but I have taken a peek at some others, and I can tell you 2 that to set up a radianz network solution to bring people in, 3 it is about 5,000 a month. 4 This is not exactly a back-breaking proposition 5 given that the client does pay $200 to get hooked up. So 6 there is a point at which this can't be pushed too far. 7 MR. HARRIS: Dave, if the Commission set a standard 8 setting practice, not a rate setting practice, but a standard 9 setting practice that said that you are to -- you, as an 10 exchange intermediary, an exchange facilitating intermediary 11 -- as an exchange, cannot offer rebates for people who offer 12 liquidity, and we say the same to everybody else as well. 13 Where does that leave you? 14 Same question to Janet, if we say that there simply 15 will be no liquidity rebates, but we say that also to all 16 your competitors, are you worse off or better off? 17 MR. HERRON: On our part, I think the restructuring 18 fees -- the customer and the exchange have to pay 19 restructuring fees, and of course, right now the fees, as 20 structured, have set proper behavior from our dealer 21 community. 22 They are not going to effect order flow if they do 23 a bad job once they get it. I would be worrying about the 24 ability of our community to continue to compete unless we can 25 structure the fees so that incented to do what is good in the 1 marketplace. 2 MS. ANGSTADT: Yes, I don't know that I have an 3 opinion on that. I guess I do think that the ECN issue 4 should also be thought of in the context of member and 5 nonmember access to exchanges. 6 We talked a lot about linkages, and I think with 7 ITS we assume that there is a cost to ITS in that you get 8 what you pay for, it is free and it has been slow. So the 9 speed at which you can access the market, and people going to 10 charge for that, it does seem that we should able to charge 11 for the value that we are providing. It just seems that that 12 is what all of the exchange markets have done. 13 MR. COLBY: I hate to say anything about the ITS 14 system, but the electronic moves at the same speed as the ITS 15 wire and everybody else. It's when you get to market, that 16 execution -- 17 MS. ANGSTADT: Fair point. 18 MR. HARRIS: One last question, not last question, 19 really, another question on this issue, what do you expect is 20 happening to limit orders traders as their living 21 increasingly in environments where they are paid to offer 22 liquidity? 23 Do you expect that their limit orders should become 24 more aggressive or that this has had no effect upon them? 25 MR. DeSALVO: I think the limit order traders are 1 shopping for the best rebate. 2 MR. HARRIS: Are the limit order traders shopping 3 for the best rebate or are the brokers representing the limit 4 order trader shopping for the best rebate? 5 MR. DeSALVO: Both. 6 MR. HARRIS: And do you think that -- 7 MR. DeSALVO: What I would prefer is that you 8 eliminate access fees, you come down to the most efficient 9 and you don't have to worry about rebates. 10 MR. HARRIS: Which would be the better standard, to 11 eliminate access fees or rebates? 12 MR. DeSALVO: Access fees, because if I represent 13 different ten bids, when I access Joe's ten bids, it is ten 14 bids, when I access Chris on the floor, it is ten bids. When 15 I access Archipelago, it is ten plus the access fee. So the 16 access fees get eliminated first. 17 MR. HARRIS: If we eliminate the access fees, what 18 is the expectation, what happens to the limit order rebates? 19 MR. DeSANO: I think it just goes away naturally. 20 MR. DELLA ROSA: The benefit of limit orders in the 21 marketplace is it should be incentivized by some rebate given 22 to the marketplace by some unlevel playing field that's 23 really a benefit is why that is incentivized that way. The 24 benefit of the marketplace should be price-time priority, 25 that's why you put in the limit order in the first place. 1 MR. HARRIS: I am a clever guy here, if you 2 eliminate the access fee, what happens, and you don't control 3 the limit order rebate, wouldn't we just see the transaction 4 fee rise very quickly? 5 PROFESSOR KYLE: But it might be paid not by limit 6 orders but by market orders, you place an order to buy four 7 and I place an order to sell four. The criteria are we ought 8 to be able to actually meet in the market. 9 MR. DeSALVO: Gunner explained it quite well, if 10 Gunner gives Morgan Stanley a limit order and he pays me a 11 commission for getting the limit order done, I don't get to 12 charge an access fee so I collect on that side and collect on 13 his sides; otherwise, I should be paying. It works the other 14 way around on the ECN model. 15 MR. HARRIS: Let's change the topic for a moment 16 and talk about internalization. Internalization is when a 17 broker-dealer routes an agency order to that same 18 broker-dealer, usually in a separate unit of the firm, for 19 execution and, generally, at the best bid or offer. 20 Let's first ask a blank-sheet question: Should 21 internalization be allowed? 22 MR. HERRON: Yes. 23 MR. HARRIS: Why? 24 MR. HERRON: I think internalization or payment for 25 order flow, I think that is an incentive to order flow that 1 has not proven in any way to be bad for the customer. As a 2 matter of fact, I think that many studies have shown that 3 people subject to those practices are taking their best 4 execution obligation very seriously and making sure that the 5 customers receive very good execution. 6 MR. HARRIS: Taken as given what you said, and I am 7 sure that many disagree with that, but let's take as given 8 what you said, that the customer would be no worse off with 9 payment for order flow or internalization, might it still be 10 the case, though, that if there weren't such things that the 11 customer might be better off. 12 So what you have said is that the customer is no 13 worse off, in a world in which there is internalization, the 14 customer is no worse off having the order go to the 15 internalizing dealer versus going to a market. The question 16 is: If there were no internalization, might the customer be 17 better off because they would have a better chance reaching 18 their limit order? 19 MR. COLBY: Or to put it another way, is there 20 another customer order that would be better off? 21 MR. HARRIS: Yes, that's the easier way to say it. 22 MR. HERRON: I will deal with the reality, that the 23 evidence points to the customers are well served, and let the 24 economists deal with the hypothetical. 25 MR. COLBY: That knocks Ruben out. 1 MR. QUICK: I think it has been proven that when 2 customers meet customers, a better price is achieved by the 3 customer. 4 And by having time priority, the customer, again, 5 is served best because he is shown that liquidity gets 6 interacted with. 7 MR. COLBY: In most internalized models, customers 8 are also meeting customers as well. It is an opportunity 9 where broker-dealers do take full advantage of their own 10 order flow to have their own customers meet their own 11 customers as well as best execution when necessary. 12 MR. HARRIS: When a trader offers a limit order and 13 it is the first order in the market at that price, is it fair 14 that dealers can be trading at that same price subsequently 15 for the dealers own account? 16 Fairness is one of our objectives, of course, so we 17 have somebody who has offered to -- 18 MR. QUICK: So you are acting as an agent and you 19 are acting as a dealer? You are acting as an agent for the 20 customer's order -- 21 MR. HARRIS: Let it be somebody else's order. 22 MR. QUICK: You are disadvantaging somebody who is 23 shown time priority at that price, and you are self-dealing 24 ahead of the customer. You are, again, making the markets to 25 the advantage of the participants rather than the customer. 1 MR. HARRIS: So now let me ask you a question: I 2 offered that order in Dave's market, Chicago, it was the 3 better price. And the question is: Should traders at the 4 New York Stock Exchange be allowed to trade at that price 5 before he fills his orders? 6 MR. QUICK: The rules do permit that right now in 7 our present model. But if he is trading as a dealer, my 8 initial point was if he is acting as a dealer and an agent, 9 he should not be able to trade next to his customer's order. 10 MR. HARRIS: I understand. 11 But the principal you just articulated, though, 12 would suggest in a different world with linkages that the New 13 York Stock Exchange would go to Chicago because they have 14 time precedence. 15 MR. QUICK: The customer's preference is to where 16 they send an order. If they are sending an order to the New 17 York Stock Exchange, if I am acting as an agent for an order 18 at the New York Stock Exchange, I am disadvantaging somebody 19 who is delivering an order to my venue at that same price. 20 MR. HARRIS: And in a world in which there are, 21 say, very high speed connections, why would the customer care 22 whether he got the same price from you or the same price from 23 somebody else? 24 MR. QUICK: The order, if it was interacting with 25 us, is on our market, not on the Chicago market. 1 MR. HARRIS: So let's go back to the question of 2 national fairness, then. If somebody went to the Chicago 3 Exchange for whatever reason and displayed in the national 4 market that he was the first person willing to trade at that 5 price, should he get the trade? 6 MR. QUICK: Not in the present rules, no. 7 MR. HARRIS: That's not in the present rules. I 8 didn't ask what the rules were. I asked what should happen, 9 what's fair? 10 MR. QUICK: I think if somebody is willing to trade 11 at the same price on the New York Stock Exchange, they should 12 trade. 13 MR. HARRIS: How is that different -- 14 MR. QUICK: I am acting as a dealer and as an 15 agent. I step aside when an order comes in at that price. I 16 don't trade as a dealer at that price until the customer is 17 satisfied. 18 MR. DELLA ROSA: On the national market system, the 19 price that is displayed takes precedence wherever it is 20 displayed. In a perfect world, in the national market 21 system, the order that is displayed in the book should take 22 precedence. 23 PROFESSOR KYLE: In a perfect world, the tick size 24 would be even less than a penny. I think the issue that you 25 are talking about when you go to eighths of pennies, it 1 becomes twelve- and-a-half times less important because when 2 you have pennies, the ability to trade exactly that price 3 with a one penny difference is not worth that much. 4 So I think that needs to be recognized, along with 5 the fact that in pennies traders who want to trade size 6 should place scaled limit orders covering every single penny 7 so that it is actually quite difficult to front run them or 8 to violate some kind of price priority or time priority in 9 the national market system environment. 10 MR. DeSALVO: I think you have IIAc1-6 rules to 11 determine whether the internal version models are correct or 12 not, so you also have, as regulators, the last vote to 13 determine whether the broker-dealer made a wise choice or 14 not. 15 But I think when you talk about internalization, 16 you also ought to speak about some of the benefits. Some of 17 the benefits are liquidity, including enhancement, providing 18 liquidity that is outside what the published NBBO outside 19 bears is a benefit to the client. 20 In addition to liquidity enhancement, speed of 21 execution and turn around time is something that clients 22 enjoy. So they enjoy the outside NBBO executions and they 23 also enjoy rapid turn around times. Those two benefits of 24 internalization sometimes get lost in the menu. In addition 25 to that, if a broker-dealer, as I said, missteps on 1 internalization, the IIAc1-6 numbers are on public notice and 2 you can determine there whether they are making good or bad 3 decisions. 4 MR. HERRON: Within the national market system, 5 each participant has various levels for the most part of 6 strict price-time priority within their venue. I think if 7 you talk about trying to provide strict time price priority 8 between all the venues, you start to walk down a path towards 9 very noncompetitive central limit order environment, which I 10 think takes away choice for the customer. It takes away the 11 competition that has made our market so strong. 12 MR. HARRIS: Once again, you are speaking in favor 13 of competition among market centers, and other people arguing 14 the opposite point speaking in favor of competition on price. 15 I want to explore one last issue as we approach our 16 time of closing, and that is the imperfect world in which we 17 live. We have been talking about in a perfect world, blah, 18 blah, blah. In a perfect world, everybody is completely 19 honest and the best execution happens because people take a 20 lot of care to make sure that it does. But in a very 21 important respect that in some sense explains why the SEC 22 exists, we live in a world where not everybody is honest and 23 not everybody works as hard as we would like them to work to 24 obtain best execution for their clients. And many of the 25 clients, undoubtedly, the vast majority of the retail clients 1 are not in a position to adequately evaluate best execution. 2 The question is: How expensive are the regulatory 3 efforts -- and by "regulatory" doesn't mean SEC only, but the 4 brokers, the exchanges and so forth -- how expensive are the 5 efforts to ensure that we eliminate all the dark places where 6 bad things can happen? 7 So, for instance, internalization is maybe a good 8 thing, but it also raises the possibility that it can be a 9 dark place where somebody is misrepresented and nobody 10 discovers it. So what I would like to have is a very short 11 discussion on the expenses of ferreting out dishonest people 12 and perhaps some comments, if interested, on alternatives for 13 doing that. 14 We will start with Dave. Dave, what do you guys 15 spend to ensure that the market is operating fairly, if 16 anything? I want to get you to respond quickly. 17 MR. HERRON: I have been up and down for six days, 18 I might have to defer to Paul on the actual number of what we 19 spend. 20 Speaking for the broker-dealer community in my past 21 incarnation, there is a great deal of money being spent on 22 monitoring for best execution, there was a great deal of 23 money being spent before the requirement for IIAc1-5 and 24 IIAc1-6 data. I think the general sense that disclosure is 25 very important has solved a need. 1 I think customers are much -- are very savvy. 2 Customers know the execution costs, they know where they are 3 getting executed, they know what the NBBO was, they know what 4 trades are on either side of their execution, and I think 5 what the industry has done at the SRO level and the 6 broker-dealer level in providing data and looking at 7 execution has been extensive and is really fulfilling the 8 purpose of disclosure. 9 MR. HARRIS: Joe, can you give us a sense of the 10 money that is spent by Goldman to ensure best execution for a 11 client? 12 MR. DELLA ROSA: To quantify the exact dollars 13 would be difficult to do off the top of my head, but I can 14 tell you that from the standpoint of the technology systems 15 that are in place to monitor anything from limit order rules 16 and things associated with that to inside code and best 17 execution are significant. 18 From the standpoint of our institutional clients 19 and our own internal research groups that we have simply a 20 setup for examining best execution and discussion of it, 21 because, as I mentioned before, the bigger it gets, the 22 argument becomes much more complex. Understanding that, we 23 have spent a significant amount of money as relates to that. 24 So I think that it has been, quite frankly, meaningful 25 dollars that have been spent. 1 MR. HARRIS: Gunner, you have spoken for a 2 consolidated system. That may be pie in the sky, but just 3 for amusement now, the two previous folks, Dave and Joe, have 4 spoken to the expenses that their organizations spend on best 5 execution, or in Dave's case, past experience with 6 broker-dealers. 7 In a consolidated world, how much would they have 8 to spend? 9 MR. BURKHART: I think in a consolidated world they 10 would have to spend much less, and I think we, institutional 11 investors, would have to spend much less as well. 12 We currently spend a great deal of money increasing 13 and great deals of money on assessing our transactions, both 14 best execution obligations and responsibilities. We also 15 spend a lot of money on trying to figure out and access a 16 more perfect marketplace and new alternative providers that 17 get people like Scott, Minder and me together to trade where 18 we reduce our market impact. 19 So there is an auxiliary cost, unfortunately, to us 20 and, therefore, to the investor on the street due to the 21 structure of the marketplace. But I have no doubt that if 22 the pie in sky model were to come to the United States, I 23 would think that costs for all participants would be lower. 24 MR. DeSALVO: I don't necessarily think that a 25 consolidated model would represent the cheapest cost because 1 when you look across between Europe and the United States, 2 the more expensive place to trade is in Europe because the 3 competition with one exchange leads to the highest trade cost 4 and post-trade cost. It is clear they are more expensive in 5 a noncompetitive environment than it is here in the United 6 States where you have competitive miles between NASDAQ and 7 New York which leads to low transaction costs and then, most 8 likely, lead to lower post-transaction costs. 9 So the cheapest model is the fragmented model where 10 there is competition, the total pre-trade, trade and 11 post-trade costs here, not in the consolidated model. 12 What is interesting, you asked the other gentlemen 13 to quantify cost. I would say that we spend over $200 14 million in market data alone at our firm, some of which is 15 going to regulatory policy in addition to, as Joe said, our 16 internal IT costs, legal and compliance. So you are talking 17 about hundreds of millions of dollars that are going into 18 this effort. 19 MR. LEE: I think the data about European and USA 20 costs are much more conflicted than you are indicating. I 21 don't think they are as simple as that. 22 MR. GOLSCHMID: Isn't most of the cost here the 23 cost of different platforms as opposed to a regulatory cost? 24 I think those two things may be getting confused. 25 If you have multiple platforms, you are going to 1 look the best and you are going to spend to do that. And it 2 is not the regulatory costs that's driving that, it seems to 3 me. 4 MR. BURKHART: In fact, the regulatory cost goes up 5 because regulators have to look at more and more platforms. 6 MR. HARRIS: I am in an unpleasant position in 7 several respects today. I presently have to cut off the 8 discussion, but I am sure many of these topics, including 9 Commissioner Goldschmid's topic, will be discussed at our 10 next discussion. I also, as many of you probably recognize, 11 have been in the unpleasant position of only asking questions 12 and not being able to answer them. 13 MR. GOLDSCHMID: Larry, we will give you a minute. 14 MR. HARRIS: If you give me a minute, I will offer 15 just one perspective. It is completely off the topic of what 16 we were discussing. 17 I believe that theoreticians and purists in market 18 structure have pretty firmly established the single most 19 important determinant of liquidity is asymmetric information. 20 That is to say, the extent to which some people know more 21 than others do. And I would like to connect that to the 22 importance of the SEC. 23 The SEC, through its extraordinarily broad 24 disclosure programs, primarily organized through the Division 25 of Corporation Finance, have done a tremendous job of 1 leveling the field in the United States, so that information 2 is very broadly available to market participants of all 3 different stripes. And to the extent that the U.S. markets 4 are very liquid, and perhaps much more so than most other 5 markets, it has probably more to do with the disclosure 6 programs in corporate finance than it may have to do with the 7 specific market structures that we have. 8 But in any event, I thank everybody for their 9 questions, for their responses, for their patience and for 10 their interest. 11 MS. NAZARETH: We will take a break and reconvene 12 in 15 minutes. 13 (Brief Recess.) 14 MS. NAZARETH: We are ready for the final stretch. 15 This last segment is on exchange registration and the 16 self-regulatory system. 17 Recent discussions of the current self-regulatory 18 model, and there has been much discussion, has focused on 19 several issues. First, the growth of ATSs and ECNs, and 20 electronic trading by SROs have heightened the competition 21 between traditional SROs and their members who sponsor 22 electronic trading systems. 23 Demutualization plans may further aggravate these 24 conflicts as SROs compete against their members as for-profit 25 entities. Some market participants that advocate the single 1 independent SRO model are doing it as a separate SRO that is 2 independent of any particular market or market participants 3 can lead to fewer conflicts and greater economic and 4 regulatory efficiency. 5 Others have argued in favor of a hybrid model in 6 which the regulation of member activity is conducted by a 7 single independent SRO, but the oversight of trading remains 8 in each individual market center. We will discuss the SRO 9 model a bit later in this segment. 10 As to the issue of what is an exchange, the current 11 definition focuses on the bringing together of orders for 12 execution under a common set of trading rules. Under this 13 definition, traditional floor-based auctions, newer ATSs that 14 provide systemwide price and time priority, options exchanges 15 that allow the dealers to trade on parity with customers, and 16 even an organized over-the- counter like NASDAQ that does not 17 provide systemwide price and time priority would be 18 exchanges. 19 The threshold question that is presented by the 20 NASDAQ exchange application is whether a market that allows 21 the dealers to trade against their own orders, away from a 22 voluntary central order interaction facility, and that does 23 not, therefore, provide systemwide price and time priority or 24 customer order priority over dealer trading, should be 25 allowed to register as an exchange. 1 What would be the ramifications for our existing 2 exchanges in such a marketplace that changed its designation. 3 Would the national market system potentially become a synergy 4 of fragmented internalized marketplaces? 5 So that's where we are now. I think we can start 6 by discussing where I just ended with the question of 7 exchange registration. The first question is: Should the 8 Commission permit NASDAQ to register as an exchange if it 9 does not require interaction between orders but provides 10 systemwide price priority? 11 And isn't the analysis further complicated if 12 NASDAQ becomes even more profit driven, will this lead both 13 to more internalization by existing exchanges and to the 14 dilution of regulatory efforts? So this obviously brings 15 together a number of things we discussed earlier. While I 16 can't promise, I think this will still be a lively 17 discussion. I can't promise the blood and guts that we had 18 in Washington, the discussion was between Rich Bernard and 19 Rick Ketchum, but we do, nevertheless, have people here, I am 20 sure, who have strong views. 21 Should I start with a New York participant who may 22 have some views on this. 23 MR. QUICK: Rich, want to help me out on this? 24 (Laughter.) 25 MS. NAZARETH: He had his shot. 1 MR. QUICK: I think in any exchange, I think we 2 famously talk about becoming an entity, a public entity. I 3 think in terms of what you build, we have an infrastructure 4 that has a capacity of 10 billion shares. With our interest 5 in maintaining that, we do like two billion shares or 1.6 6 billion shares a day. 7 Are our interests in the same interests as the 8 owners of the exchange? I think there is a lot of competing. 9 Do you become slaves to your shareholders, do you become 10 slaves to your market participants? It is a question that I 11 think is up in the air. 12 I think Rich probably has a lot more opinions of it 13 than I do. 14 MS. NAZARETH: Ruben, you have written about 15 exchanges. Obviously, we are talking about here in the 16 context of our regulatory structure, but do you think that 17 there is a requisite level of order interaction that should 18 be required in the U.S. for you to be defined as an exchange? 19 If not, what are the ramifications. 20 MR. LEE: No, I don't. 21 I think that the most complex trading systems have 22 the best incentive to determine their own what I would call 23 market structure that is order execution nondriven. And I 24 think there is no need for the SEC to require price-time 25 priority, which is not to say that price-time priority is not 1 the best, but if it is the best, they will discover it. They 2 have exactly the right incentive to discover it. 3 So I don't think that price-time priority should be 4 a criterion by which to assess whether NASDAQ should be an 5 exchange or not. In this context, in considering -- as your 6 brief is to consider the American securities markets, I 7 think, and I am bound to say this, I think that you should 8 consider, given that you trade a wide range of non-U.S. 9 stocks here in America, I think it would be beneficial for 10 U.S. investors to have greater access to European exchanges. 11 And this is not an argument for Europe. This is an argument 12 if I am sitting in your chair promoting the American 13 securities market. 14 I think that there are a range of different ways 15 which would be harmful or insufficient by Europeans 16 exchanges, which would benefit your brief, that is to say, 17 U.S. investors. But in answer to your question about 18 price-time priority, I think that's relatively simple. 19 There is a whole range of other issues to do with 20 governance, and I think that one point I make about that is 21 that I am less convinced of the economic distinctions between 22 for-profit and not-for-profit institutions than other people 23 may be. In fact, I think non-profit institutions can, in 24 fact, obtain a monopoly in very much the same way as 25 for-profit institutions. So I don't think that the 1 demutualization should be necessarily be a criterion either. 2 I am sure we will be coming to conflicts of 3 interest, and maybe now is not the time, but I have a couple 4 of points about that as well. In particular, I am not 5 worried about the existence of conflicts of interest and I am 6 much more concerned about how those conflicts are managed. I 7 think the SEC has a long and fine history of managing in 8 terms of disclosures, due process, representation, and so on 9 and so forth. And I think those types of remedies would -- 10 can be continued in the future in terms of for-profit 11 institutions. 12 In fact, I think that the characterization of 13 for-profit exchanges is being more conflicted than 14 not-for-profit exchanges is simply wrong. But in terms of 15 the practical considerations, from the Commission's point of 16 view, I think there are two big ones, from my point of view. 17 One is to -- and I don't know that it fully 18 addresses the NASDAQ application, but not to determine on 19 price-time priority. And the second is for the benefit of 20 U.S. investors to allow greater access by European exchanges 21 in an appropriate manner. And there are a wide range of 22 manners which could lead regulators or mutual recognition as 23 we widely discussed and has been signaled by Harvey, and I 24 don't know how that's going. 25 MR. COLBY: Let me just say for the record, we 1 welcome the registration of any European exchange. 2 MR. LEE: But then the issue is, is that necessary? 3 Are you imposing undue costs which will in the end be borne 4 by U.S. investors by requiring such registration. And the 5 question is: To what extent should you be happy with 6 non-U.S. regulation of exchanges? 7 And I would contend that -- and once again, I am 8 not intending to be an advocate of Europe here. What I am 9 trying to be is an advocate of exactly your constituency, the 10 investors in America, what would be best for them. And in my 11 view, I think that one should look carefully at the extent to 12 which the regulatory environment of exchanges in Europe is 13 beneficial, and, of course, to the many contexts that I think 14 that it is not only beneficial, but equivalent to the U.S., 15 so requiring double regulation does not achieve anything. 16 MS. NAZARETH: So we got in the foreign market 17 access point, which is not on our agenda, but it is on the 18 Commission's agenda, but it is not on the agenda today. 19 Pete Kyle, earlier in our discussions this morning, 20 you mentioned that something to the effect of retail 21 investors may, in fact, have certain expectations about what 22 they get on an exchange and it may involve more than simply 23 transparency. 24 Do you think that retail investors or market 25 participants in general have an expectation about what it 1 means to have a transaction executed on an exchange that is 2 different from their expectations of transactions in the 3 over-the-counter market? 4 PROFESSOR KYLE: Certainly, I think that 5 historically until the last five years or so, retail 6 investors have gotten a better deal on organized exchanges 7 than they have in pure dealer markets. If you try to make a 8 transaction called a small market order on NASDAQ versus the 9 New York Stock Exchange, the New York Stock Exchange 10 historically has looked very good until the last few years 11 when NASDAQ started functioning more like an organized 12 exchange itself. 13 And, largely, the reputation of an exchange with 14 its retail investors is its capital or a large part of its 15 capital. A point that goes along with that is that the 16 governance structure has changed. It has a lot to do with 17 what objective they can maximize on. And the exchange that 18 is likely to be most friendly to retail investors dealing 19 with an exchange that is governed by smaller customers. 20 But the fact is exchanges are kind of owned and 21 controlled by a mixture of organizations with different types 22 of interest, including the kind of local floor traders on the 23 exchange and the larger securities firms whose interests 24 sometimes conflict and sometimes go hand in hand. And I 25 think that it is very important to think about the government 1 structure of exchanges influencing the incentive that the 2 exchanges have to represent the interests of retail investors 3 or any other group. 4 There is also, looking at commodity exchanges which 5 historically have had different government structures, 6 different ones in New York and Chicago, there are a whole 7 bunch of them. They have had different government structures 8 over the years and they have also behaved differently with 9 respect to the rules, so you can see that in the commodity 10 markets as well. 11 MS. NAZARETH: Tony? 12 MR. LEITNER: I guess the question would be why 13 not, because if we look at the convergence in the 14 marketplace, we see that, number one, the old dealer market 15 that NASDAQ was is certainly not only a different market 16 today, that is to say, many NASDAQ dealers, including 17 ourselves, are more and more representing customer orders 18 essentially as agents in that marketplace. 19 The prime member system within NASDAQ is an attempt 20 to create an auction style marketplace within the context of 21 the NASDAQ market. 22 Number three, you would eliminate from a policy 23 perspective and finally draw closure to the issue NASD and 24 NASDAQ by forcing NASDAQ to, in fact, become its own SRO, 25 although the thought of another rule book on my shelf is 1 itself almost a sufficient negative against that, but we 2 could get to what that structure should do later. So it 3 seems to me if one looks at those kinds of things, you would 4 say that these other issues are sort of red herring issues. 5 That what happens is the Commission then, from a policy 6 perspective, really gets an ability to force that marketplace 7 to, in fact, be governed having a majority of outside 8 directors. 9 That is not true of ATSs today. They can be 10 totally privately owned or publicly owned or whatever, have 11 their own board of directors with no requirement for any 12 public representation and so forth. So I think the real 13 issue is not NASDAQ becoming an exchange but what do you do 14 with ATSs that become primary markets. 15 MS. NAZARETH: That is a very good question. That 16 was actually going to be my next question, as matter of fact. 17 MR. DeSALVO: With regard to the NASDAQ exchange 18 status, NASDAQ and New York each make each other better. 19 There are elements about NASDAQ that we like and there are 20 things about New York that we clearly like, but the existence 21 of the two marketplaces, clearly, clearly, although they may 22 not admit it, at the end of the day, they do make each other 23 better. 24 And it is hard for us to see a marketplace that has 25 the most volume on a day-to-day basis not being granted 1 exchange status. And for you to worry about whether they are 2 a model for internalization is worrisome. You are imposing 3 something that the client base seems to be voting with their 4 feet, currently that they are okay with it. 5 MS. NAZARETH: I guess the question is are the 6 clients voting for or against? 7 MR. DeSALVO: I think at the end of the day, the 8 Ac1-6 number will ferret out and also that the listings and 9 the volume will be the economic result of where it goes. But 10 we clearly enjoy the benefits of a robust New York and a 11 robust NASDAQ. And we have the benefits and our clients 12 have the benefits of the best of New York and the best of 13 NASDAQ. But to have one now not be an exchange, and with the 14 volume here, it doesn't make any sense to us. 15 MS. NAZARETH: Tony was talking about Reg ATS and 16 the question is, do the principles in there still work? 17 Obviously, what Reg ATS did was take a very expansive view of 18 what was an exchange and, arguably, did not specifically have 19 these other conditions that we are discussing, and would 20 leave it to the Commission to determine whether marketwide 21 principle should be applied virtually regardless of whether 22 you were an exchange or an ATS. 23 You may have noticed in connection with our first 24 market structure hearing, that AMEX presented a paper that 25 basically said that they thought Reg ATS had failed and that 1 all entities, technically defined as exchanges including 2 ATSs, should be subject to the full exchange registration and 3 regulation. 4 I would imagine, Janet, you might have a view on 5 that? 6 MS. ANGSTADT: Actually, I think we are a good 7 example, we grew up with Reg ATS. What I thought was a 8 positive with Reg ATS was that you got the choice between 9 either being a member of NASDAQ, somebody who is competing 10 with us and trying to deal in our market structure and our 11 market structure choices, or being able to go off on our own. 12 And it is, really, that choice that is very important and 13 taking on the responsibilities that come with that choice. 14 MR. LEE: I am not fully conversant with U.S. law 15 and regulations, perhaps my ignorance will become evident 16 here, but I have one comment which is as follows: My 17 understanding is that effectively there is an encumbrance 18 being an exchange and being an SRO. That's effectively the 19 same thing. 20 And there I have a potential problem that trading 21 systems, all ATSs, should become SROs, precisely because -- 22 and this is a point that Pete alluded to -- if you have a 23 government structure which is required by law to have 24 representation and so on and so forth, and then it will find 25 it very difficult to establish a trading system which 1 disintermediates the people who were put into your government 2 structure. 3 So that if, for example, you are an ATS and you 4 want to have direct access by institutions, that may become 5 harder with a mandated government structure that says we need 6 financial intermediaries to be on our board. And if that's 7 the case, then I would worry about requiring all ATSs to 8 become exchanges. 9 MS. NAZARETH: But I do think the Commission has 10 addressed that by trying to balance the issues and determine 11 that there will be some representation of the members who are 12 U.S. traders but not the large majority. You are right, 13 though, otherwise it would be a conflict. 14 I guess another issue raised by the NASDAQ exchange 15 application is: What are the outer limits? Is any trade 16 that's printed on the exchange an exchange trade? Does that 17 potentially lead to the buying of prints? Obviously, we have 18 seen a lot of that lately, we have talked about the rebates. 19 And what are the implications of having one 20 potentially become print facilities? Are there problems with 21 having transactions that on their face appear to be an 22 indication of liquidity in a certain marketplace, where 23 actually those transactions were not effected anywhere near 24 that venue? 25 MR. LEITNER: But you have to deal with that 1 anyway, right? That's what the first two panels were about. 2 What does that have to do with whether NASDAQ 3 becomes an exchange? 4 MR. O'KELLY: I will take a slightly different 5 view. I think you have to deal with that question, in that 6 if NASDAQ becomes an exchange and they are allowed to avoid 7 intramarket price-time priority, it won't take a day, it 8 won't take 24 hours for some other exchange to come in and 9 ask for similar relief. 10 As a matter of fact, I think I have already seen 11 some of those requests for similar relief, not from us, but 12 from others, and you are going to have to deal with that 13 question. And I think it is a very important question for 14 the marketplace. I think we have seen at least one model of 15 the market already where it seems to be the member is the dog 16 and the exchange is the tail. I don't know that that's the 17 kind of market structure that you want to incent us to 18 develop. 19 MS. NAZARETH: Tony, I certainly agree with your 20 point, they are there already and one can't bifurcate that 21 question from the question of exchange registration. 22 Taking the question separately, though, apart from 23 the NASDAQ context, do you think it is problem? Do you think 24 it is something that the Commission need to address? 25 MR. LEITNER: Which issue? 1 MS. NAZARETH: The issue of trade and print 2 facilities. 3 MR. LEITNER: Yes, but I think we would say that 4 that's an issue across all of the markets. Even in New York, 5 people trade upstairs and bring their crosses to the floor 6 and they are printed there, so whether or not it is still an 7 issue because they are exposed, that you can say there is 8 some legitimacy to the execution, I think the point that was 9 made in the earlier panel was, are you misleading the public 10 by the degree to which you permit these practices to occur, 11 or is it clear by the nature of the reporting what is 12 happening on the market and what is happening off the market? 13 Technology would today would potentially allow you 14 to make those distinctions, so there is may be more than one 15 way to deal with the policy issue but you have to define 16 what's the problem first. 17 MS. NAZARETH: Well, I guess that's the question: 18 Is there a reason for us, from a regulatory standpoint, to 19 make that distinction? 20 Obviously, to make that quite simple, would there 21 be a benefit to the marketplace of having an indicator of 22 when a transaction was effected either on an exchange or 23 through an exchange system versus when it was effected 24 separate and apart from that internalized setting? 25 MS. ANGSTADT: I think that one of the questions 1 that we talked about this morning, too, in the national 2 market system, that there was this principle of the 3 opportunity for a transaction to be effected without the 4 intervention of a dealer. 5 Obviously, NASDAQ has done quite well in the dealer 6 market. I don't know if in an exchange application, you have 7 to consider that objective of the national market system in 8 considering their application for exchange status. 9 MR. LEITNER: Certainly in regard to 10 internalization of members of NASDAQ and their ability to 11 print on NASDAQ, as they do today, I don't see why you would 12 force those prints to go into the ADF if it is done by a 13 member. That's sort of internalization, and that, I assume, 14 is the question you are asking, is that internal cost, 15 because they print right now, they just print to ACT. 16 MR. COLBY: Let me just -- 17 MR. DeSALVO: The concept of internalization. You 18 speak of it as if it was a majority of the order flow 19 involved. It is not. It is a minority of the order flow. 20 So you seem to be hung up on this concept of internalization, 21 which would make sense if it was lion's share of the volume. 22 It is not, it is clearly the minority. 23 MR. COLBY: The reason that we are talking this 24 issue out is if you move from an exchange where there is some 25 sort of notion of a centralized trading system where the 1 orders are interacting and broadening out to a system where 2 there are rules governing how trades are done, but they don't 3 actually internalize, the question is: Is there any sort of a 4 lingering notion that when you have bulk trades reporting, 5 that those are all interacting? 6 If that's not really not how people think about it, 7 then trying to differentiate between trades that were done at 8 an interacting location and nexus, it probably is not worth 9 the effort. But if there is some lingering idea by looking 10 at the volume done by the Chicago Stock Exchange that tells 11 about how many orders are interacted in the Chicago Stock 12 Exchange environment, then knowing that they were printing 13 orders by somebody that was just internalizing in their -- 14 upstairs in an office and was just printing them there, that 15 could be viewed as inflating and misleading. 16 But if that's not the way people view it, then 17 trying to differentiate it may not be important. 18 Does anybody have any thoughts about that? 19 MR. QUICK: I think you are creating false 20 liquidity if you are printing NASDAQ trades in Cincinnati, 21 you are creating a sense of false liquidity in a market that 22 really doesn't have liquidity if the trades are taking place 23 somewhere in a closet and they are just getting printed on 24 that marketplace to generate tape revenue. But, again, it is 25 not transparent, customers are not interacting with each 1 other, prearranged trades. 2 MS. NAZARETH: Pete, do you have a view on this? 3 Obviously, one of the reasons we focus on this is because we 4 have a concern about the price setting mechanism and insuring 5 that it is as efficient and effective as possible. I think 6 today we obviously have a system, particularly with respect 7 to listed stocks, where there is so much activity in this 8 venue, that there is a tremendous amount of order 9 interaction. And the fact that you have a number of other 10 market centers that may mimic those prices really may not 11 have a serious negative impact on the price setting 12 mechanism. 13 I guess the question is, eventually, in a worst 14 case scenario, if we move to a totally fragmented market 15 without a significant marketplace where orders interacted, 16 will we have the same efficiency as we have today? 17 PROFESSOR KYLE: I am a strong believer in a 18 mechanism where if Larry places his bid at four and I place 19 my offer at four, we meet each other. And I don't know -- 20 NASDAQ is set up kind of the way it is now, but as an 21 exchange, would it deliver as well as the New York Stock 22 Exchange delivers it or as well as an exchange that has kind 23 of normal centralized marketplace rules will deliver it. 24 MR. LEITNER: Just look at it from one perspective, 25 and that is intermediary perspective and the costs associated 1 with it. 2 Right now a broker-dealer that executes a 3 transaction in the NASDAQ marketplace reports to ACT. If 4 NASDAQ becomes an exchange, to the extent that we have to 5 change the technology and reporting system and the audit 6 trail to bifurcate between where we execute orders on the 7 same stock, it drives a whole bunch of things. 8 It drives, for example, how exchange fees or 9 Section 31 fees are assessed and paid. It is a whole bunch 10 of stuff, and one thing I can tell you with almost a 11 certainty is that right now is not the time to ask 12 intermediaries with the pressure we are under to accept a 13 whole bunch of additional costs. So I would urge the 14 Commission in considering this issue not to think of it in a 15 one dimensional frame work. 16 MR. LEE: I have a couple of points, which are as 17 follows: The first is that calling NASDAQ an exchange or not 18 calling it an exchange is not going to change its market 19 structure. And so that if you think that price-time priority 20 is better for market structure, then mandate it anyway. 21 Don't use the exchange ticker, if you will, to mask that 22 decision. 23 Second, my view is that it may well be that 24 price-time priority does deliver better markets. But then, 25 that's fine, let me go ahead and do that and set it up, and 1 anybody who does that to compete with NASDAQ will win. 2 My third point is I actually think that this is -- 3 to come back to Matt's point in a different way to focus on 4 what is the key problem at the moment. And the problem is 5 not fragmentation. The problem is to the contrary, it is 6 centralization of a very, very big exchange and managing that 7 big exchange that will now be associated with that. So in 8 some sense, I think you are looking at a problem which has 9 arisen because of an application, and you need to resolve 10 that problem, but that doesn't mean that you should take your 11 eye off the other market structures which may be more 12 important. 13 MS. NAZARETH: I would like to return for a few 14 minutes to the SRO structure. I know that some of you may 15 have views on that as well. 16 As I mentioned earlier, one of the debates about 17 SRO structure is whether having multiple, competing SRO's is 18 inefficient in and of itself, adds cost to the system, and 19 whether we should go with a single SRO model or a hybrid 20 model in which member regulation is under the purview of a 21 single SRO market regulation, regulation of the trades and 22 transactions in the marketplace being done by each of the 23 individual markets. 24 Some of you here have spent some time thinking 25 about that. Paul? 1 MR. O'KELLY: We talked about that issue internally 2 in our exchange. I think there are some trade offs here to 3 having a single regulatory body. There are a good number of 4 efficiencies that can come from that, expertise created by 5 that, regulators that they are now doing that across all 6 markets so it is more likely to have uniformity of regulation 7 across markets, which you don't have to worry about being an 8 NASD member or becoming a member of the New York Stock 9 Exchange, so having a single regulator can solve lots of 10 those issues. 11 There is a trade off, though, and the trade off is 12 that to the extent that a regulator needs to understand its 13 market and what is going on with its market participants so 14 it can effectively do its job, is going to be lost if that 15 single regulator stands away from a particular market. I 16 know the Commission spends a great deal of time itself not 17 only regulating U.S. securities markets but trying to figure 18 out what is going on in the U.S. securities markets so it can 19 regulate it correctly. 20 That same kind of phenomena is going to occur if 21 you have a single regulator. That is not to say that it is a 22 bad idea, there are just trade offs and you should understand 23 what those trade offs are before we opt to do that. 24 MR. LEE: I think, unfortunately, I am going to 25 take the NYSE position here, which is, I think, that firstly, 1 exchanges have a very strong incentive to provide a strong 2 regulatory environment. And this is what the NYSE sells as 3 its banner, and I think that is good of them. I think that 4 demonstrative of something that we face in Europe where it is 5 very, very important to not be confused by the terminology. 6 In my mind, regulatory arbitrage is something 7 widely to be welcomed. Here in the states, it is often 8 thought that it would lead to a race to the bottom, all 9 right, in many contexts. It would lead to people seeking to 10 obey particular rules and so forth. And in my mind, should a 11 country where we have seen competition in regulatory 12 structures where it is not a race to the bottom, so I think 13 that that is, in as much as competition between market 14 centers as trading systems is beneficial, I think competition 15 between market centers as SROs is beneficial. 16 And I think the down side of it, notwithstanding 17 Dave Wilken's comments last time, I think the down side to 18 having a single central regulator as exactly as one might 19 expect, it will be much more slower, it will have to be much 20 more politicized and so forth. So I would be strongly in 21 favor of competition between SRO's. 22 MS. NAZARETH: Tony? 23 MR. LEITNER: I guess the mantra here would be cost 24 competency and consistency as themes. 25 First of all, we don't know how much -- let's 1 understand what the structure is today. The SEC has primary 2 responsibility for regulating exchanges, and there is the 3 whole issue, of course, of whether ATSs should boost 4 themselves up into exchange status. Certainly something that 5 ought to be given a look for those -- you know, at what level 6 you increase the scrutiny of the exchanges and the direct 7 regulation of ATSs. 8 But with regard to the intermediary regulation, 9 right now we have, certainly for large firms who are members 10 of multiple venues, cross subsidization and indirect costs 11 that are not quantifiable. These structures being proposed 12 that potentially levy a tax at the intermediary level on all 13 of the business of intermediary, even that is not necessarily 14 associated with the trading in that particular venue. So 15 from a cost point of view, firms like ours on whom all these 16 burdens you have discussed like best execution and the like 17 are being placed and the costs associated with them after we 18 have sorted out how much it costs us to be regulated and to 19 pay our regulators, whether or not they are the designated 20 examining authority. 21 Competency. Competency is extremely important. We 22 absolutely benefit by rigorous, rigorous surveillance and 23 regulation. Firms that are compliance conscious will always 24 benefit from having regulators who understand our business 25 and come in and make sure we are doing it right. 1 Consistency. It is imperative that the rules that 2 affect members, that is, the people who participate in the 3 markets and their structure is such. For example, the margin 4 regulations, how much credit you can extend to customers. 5 There is no reason that they should differ from marketplace 6 to marketplace. They ought to be the same, in my opinion. 7 The examination for exchange of SEC rules such as 8 the customer protection rules done by exchange staff, why do 9 they need to be done by the exchange staffs of every SRO? So 10 when these issues -- and this is aside from the broader 11 question of conflicts of interest that have been raised by 12 marketplaces potentially regulating competitors and the 13 possibility for inconsistent performance. And that may or 14 may not have been an issue anticipated by bifurcation of 15 NASDAQ and the NASD. 16 When the GAO looked at this a year ago, their 17 report urged the Commission to take a careful look at this. 18 I don't have a view about what the right answer is, but I 19 think that the themes that I suggested, cost competency, 20 consistency across markets, and looking functionally at what 21 is the appropriate role for, say, a regulator other than the 22 SEC to perform, because putting more burdens on you in this 23 environment doesn't seem to be a great idea. 24 On the other hand, combining the best of what the 25 SROs currently do, into an area where you can create 1 efficiencies, and let the marketplaces regulate their market 2 rule. It is in the area of market rules, the integrity of 3 the marketplace, that exchanges know best and where their 4 surveillance systems can be set up appropriately to deal 5 fundamentally with their rules. So they should do that. 6 And the real question is whether all of the 7 exchanges should do all of the other stuff. And where you 8 have a staff that has to deal with 196-4 filings for every 9 rule approval that comes down the pike eight or ten times. 10 It doesn't seem to be a great use of the Commission's 11 resources. 12 MS. NAZARETH: That sounds, like you said the devil 13 is in the definitely details, to combine member regulation 14 from market regulation, that you are generally in favor of 15 the hybrid approach. 16 Is that fair? 17 MR. LEITNER: Yes. 18 MS. NAZARETH: Anyone else have any comments? 19 MR. DOMOWITZ: I find it interesting that on the 20 one hand, the position that Reg ATS has backfired is 21 basically a restatement of the fact that exchanges are 22 worried about competitive pressure, all right, and are 23 saying, well the competition doesn't have to pay the cost for 24 that. 25 On the other hand, the other side of the coin is 1 that someone says, well, but the exchanges enjoy all this 2 special status and they can earn money in different ways, 3 and, by the way, we have to be members, therefore, we have 4 our competition regulating us. We have also spoken today 5 about the interrelation between the market data fees and 6 regulation, self-regulation in particular, and the potential 7 conflicts there. 8 Leaving aside the details of how you do it, if you 9 really wanted to establish a level playing field that 10 everybody talks about between exchanges, ECN, ATS, and 11 whatever we want to call it these days, and leaving the issue 12 of NASDAQ aside, you simply divorce the SRO function. You 13 get it out, figure out how to fund it. There are lots of 14 ways to do it. 15 Once upon a time when there might have been a 16 hundred people on the floor and you really had to know 17 everybody else to figure out how the trading went on, and it 18 was so complicated you really needed those guys to regulate 19 themselves, I would have agreed you really need self- 20 regulation in the sense it was once proposed. 21 We are talking about going back to the New Deal 22 here. What was good policy is still good policy, even since 23 1934. But the way we carry it out is not necessarily even 24 feasible any more. I can see a resolution of a variety of 25 problems, even if you introduce some new challenges in just 1 divorcing the SRO function from the exchange, at which point 2 you get execution centers that may or may not have listing 3 functions because, frankly, someone might decide to get into 4 the listing business. But we would be done with it. 5 MR. COLBY: I would just like to point out that the 6 Congressional budgeting process has not passed New Deal 7 procedures. 8 MR. O'KELLY: May I respond to that? 9 MS. NAZARETH: Yes. 10 MR. O'KELLY: I think when we are talking about the 11 SRO function, we need to divide it into two different 12 buckets. One of them is the SRO's ability to pass trading 13 rules and enforce those trading rules across their members. 14 Another more commonly thought of self-regulatory 15 function is the surveillance function that makes sure that 16 people abide by our rules. If we were to have a single SRO 17 divorced from all the market centers that performed those two 18 functions, that single SRO would be adopting the rules that 19 apply to all the different market centers. Guess what would 20 happened in a very short period of time? With ease of 21 regulation, all the markets would look exactly alike with the 22 exact same trading. 23 I don't think that's what we want to happen. 24 MR. DOMOWITZ: I actually disagree with that. 25 I think that within the same set of standards, 1 going back to standards, the competitive environment will 2 dictate the issues for various execution centers to play in. 3 In other words, the people that support NASDAQ today may, 4 indeed, support NASDAQ tomorrow in that model. The people 5 that support New York today may, indeed support New York in 6 that model tomorrow. 7 They do so because they believe it is better for 8 how they trade and what they trade. The fact that you have a 9 uniform set of standards does not mean that you have the same 10 market structure. 11 MR. LEITNER: See, that is where the devil is in 12 the details. The New York Stock Exchange has rules unique to 13 its floor model that have to do with the handling of customer 14 orders and what our obligations are, and whether we can trade 15 alone or against trading ahead, regulating block trading, so 16 there is no need necessarily for the same type of rules on 17 the NASDAQ marketplace, so I don't know what you mean by 18 standards. 19 It absolutely fits in the New York model. They 20 believe in them. We believe in them. We follow them. But 21 the NASDAQ market is different. We go back to Matt's point, 22 which is, these markets essentially are different models. 23 And they have over time, I think, there has been some 24 convergence and maybe even some of the New York rules will 25 move in the NASDAQ marketplace as they evolve. 1 MR. DOMOWITZ: Yes, but the counter-example is very 2 simple, an ECN, when that started, for example, it was a very 3 different model than NASDAQ. It was extraordinarily 4 different model. It thrived, it grew, and the limit order 5 book market existed and actually was profitable. It operated 6 under NASD rules, it operated under the same rules as the 7 market-makers. 8 Although your point is well taken that the nature 9 of the rules also determines the nature of the surveillance 10 of associated practices. We have clear counter-examples of 11 the same regulatory regime can still breed different markets. 12 MR. LEITNER: But the era when the SEC could make a 13 thousand flowers bloom with a no action letter, I think, are 14 gone. And that's the issue we are dealing with. 15 MR. DOMOWITZ: I will agree with that, too. 16 MS. NAZARETH: Matt? 17 MR. DeSALVO: We would be more of an advocate of 18 separation of member regulation and market regulation. 19 Member regulation involving such things as common 20 sales practices and books and records and registration and 21 finance standards. But the market regulation side, we are 22 comfortable with keeping it at the site of the transaction. 23 So it is a differentiation between member and market. 24 MS. NAZARETH: Does anybody want to have the last 25 word before we wrap up? 1 MR. DeSANO: It would just be interesting to us, 2 every time we call down to ask a question to New York, we 3 never walk away feeling like justice has been served, ever. 4 And it may be that maybe there are just some things we don't 5 understand. But there is no place for us to go, and there is 6 no appeals process. There is nothing that is separate to say 7 this is true, because there is no teeth in it for us because 8 of the standing right now. 9 It would be interesting to us to have a separate 10 independent appeals process to add some justice to it for the 11 little people. 12 MS. NAZARETH: One last final word, or would you 13 prefer to end on that note. 14 MR. QUICK: I would not rather end on that note, 15 but there are answers but there are other orders that 16 interact other institutional orders at some time at the point 17 of sale in the auction process. And there is time priority. 18 There are things that missed down there, but sometimes they 19 just fall on deaf ears because they doesn't like the rules. 20 But the rules are the rules, that's why the NYSE 21 has been so successful because of the standards and rules 22 that our SRO and our market surveillance staff and 23 enforcement division carry out. The exchange is not run by 24 the members. The exchange is run by a board comprised of 25 twelve industry people, twelve outside directors that 1 +represent industries, and we have three management directors 2 on the board. 3 MS. NAZARETH: It has been a long day, but I think 4 it was an extremely productive one. I want to thank everyone 5 for your participation on behalf of the staff and the 6 Commission as well. 7 (Whereupon, at 4:30 p.m., the meeting was 8 adjourned.) 9 * * * * * 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25