1 1 UNITED STATES SECURITIES AND EXCHANGE COMMISSION 2 3 4 5 6 JOINT SEC-CFTC ROUNDTABLE 7 TO DISCUSS SWAP EXECUTION FACILITIES AND 8 SECURITY-BASED SWAP EXECUTION FACILITIES 9 10 11 Wednesday, September 15, 2010 12 9:00 a.m. 13 14 15 16 17 Securities and Exchange Commission 18 100 F. Street, N.E. 19 Washington, D.C. 20 21 22 23 24 Diversified Reporting Services, Inc. 25 (202) 467-9200 2 1 PARTICIPANTS: 2 Scott Bauguess, SEC 3 James Brigagliano, SEC 4 Nancy Burke-Sanow, SEC 5 Robert Cook, SEC 6 Jon DeBord, CFTC 7 Thomas Eady, SEC 8 Mauricio Melara, CFTC 9 Dhaval Patel, CFTC 10 Gregory Price, CFTC 11 Richard Shilts, CFTC 12 Sebastian Pujol Schott, CFTC 13 Heather Seidel, SEC 14 Riva Spear Adriance, CFTC 15 16 PANEL ONE: 17 William De Leon 18 Global Head of Portfolio Risk Management, PIMCO 19 Yves P. Denize 20 Director & Associate General Counsel, TIAA-CREF 21 Andrew Downes 22 Managing Director, UBS Securities, LLC 23 Richard DuFour 24 Executive Vice President, Chicago Board Options Exchange 25 3 1 PANEL ONE (cont.): 2 Julian Harding 3 Executive Director, Tradition Brokerage, Chairman, 4 Wholesale Markets Brokers Association Americas 5 John J. Jeffers 6 Senior Vice President and General Counsel, 7 OTC Global Holdings 8 Ben MacDonald 9 Global Head of Trading, Bloomberg, L.P. 10 Lee Olesky 11 Chief Executive Officer and Co-founder, Tradeweb 12 Stephen Semlitz (via video) 13 Managing Director, HESS 14 Heather Slavkin 15 Senior Legal Policy Advisor for the Office of 16 Investment, AFL-CIO 17 Jeff Sprecher 18 Chairman and Chief Executive Officer 19 IntercontinentalExchange, Inc. 20 S. "Vish" Viswanathan 21 Professor, Fuqua School of Business, Duke University 22 23 24 25 4 1 PANEL TWO: 2 Athanassios Diplas 3 Managing Director and Global Head of Counterparty 4 Management Group, Deutsche Bank 5 William De Leon 6 Global Head of Portfolio Risk Management, PIMCO 7 Yves P. Denize 8 Director and Associate General Counsel, TIAA-CREF 9 Bryan T. Durkin 10 Chief Operating Officer and Managing Director, 11 Products and Services, CME Group 12 Julian Harding 13 Executive Director, Tradition Brokerage, Chairman, 14 Wholesale Markets Brokers Association Americas 15 Edward S. Knight 16 Executive Vice President and General Counsel, the 17 NASDAQ OMX Group, Inc. 18 Michael Masters 19 President, Masters Capital Management, Americans 20 for Financial Reform 21 Richard McVey 22 Chief Executive Officer, MarketAxess 23 Stephen Semlitz (via video) 24 Managing Director, HESS 25 5 1 PANEL TWO (cont.): 2 S. "Vish" Viswanathan 3 Professor, Fuqua School of Business, Duke University 4 Philip Weisberg 5 Chief Executive Officer, Fxall 6 Brian S. Yelvington 7 Director - Strategy, Knight Capital Group 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 6 1 C O N T E N T S 2 PAGE 3 Opening Remarks by SEC and CFTC Staff 4 5 Robert Cook, Director 6 Division of Trading and Markets, SEC 8 7 Richard Shilts, Director, 8 Division of Market Oversight, CFTC 10 9 10 Panel One 11 Swap Execution Facilities ("SEFs") and 12 Security-Based Swap Execution Facilities ("SB SEFs") 13 - Definition and scope of SEFs/SB SEFs 14 14 - Scope of exception from mandatory 15 trading requirement 53 16 17 Panel Two 18 19 Compliance with Core Principles for SEFs and SB SEFs 20 - Block trades 79 21 - Surveillance, investigation, and enforcement 22 of SEF/SB SEF rules 97 23 - Cross-market issues 113 24 - Obligation of SEFs/SB SEFs to provide 25 impartial access 122 7 1 P R O C E E D I N G S 2 MR. COOK: Good morning. I think we would like to 3 get underway here. My name is Robert Cook. I'm the director 4 of Trading and Markets at the SEC and on behalf of the staff 5 of the SEC and the CFTC, it's my pleasure to welcome you this 6 morning to this third in a series of joint staff roundtables 7 on the implementation of the Dodd-Frank Wall Street Reform 8 and Customer -- Consumer Protection Act. 9 To my right is my colleague from the CFTC, Rick 10 Shilts, director of the Division of Market Oversight at the 11 CFTC. It is a great pleasure to be collaborating with Rick 12 and the rest of the CFTC staff on this initiative, and I 13 would like to thank the staff of the CFTC and the SEC for all 14 of their very hard work in putting this roundtable together 15 and their ongoing close cooperation on the implementation of 16 the Dodd-Frank Act. 17 I need to read a mandatory announcement related to 18 the use of this room. So if you'll bear with me, but listen. 19 In the unlikely event we must evacuate the building, please 20 exit the multipurpose room, make a right to reach the main 21 staircase to the main lobby and exit the building on F 22 Street. Once out of the building, please follow SEC 23 employees or SEC guards to the assembly area at 3rd and G 24 Streets. If you're unable to easily use stairs, please 25 notify an SEC employee or guard. In the event we must 8 1 shelter in place because of an incident occurring outside the 2 building, we ask that you remain in the multipurpose room -- 3 that's the room we're in -- and await further instructions. 4 Thank you. Back to our regularly scheduled program here. 5 The Dodd-Frank Act tasked the SEC and the CFTC with 6 bringing greater transparency and oversight to the OTC 7 derivatives markets. At our roundtable yesterday, we 8 discussed the reporting and public dissemination of 9 transaction information regarding swaps. 10 The purpose of today's roundtable is to provide a 11 forum for discussing the implementation of the Dodd-Frank Act 12 with respect to the trading of OTC derivatives. In 13 particular, we'll be focusing on transitioning that trading 14 onto regulated markets and issues regarding the creation and 15 regulation of a new type of market called a swap execution 16 facility or SEF. 17 Under the Act, the agencies are directed to 18 establish a registration framework for SEFs and to set the 19 regulatory standards for swaps to move from an unregulated 20 trading environment to these new trading venues. In 21 addition, the Act requires that with respect to swaps that 22 are required to be cleared, the counterparties must execute 23 the transaction on an exchange,designated contract market or 24 a SEF unless no exchange, DCM or SEF makes a swap available 25 to trade. The Act also requires that in order to maintain 9 1 registration, a SEF must comply with certain core principles 2 established in the Act, as well as any other requirements 3 that the CFTC or the SEC may impose by rule or regulation. 4 Today we'll be focusing on our respective 5 rulemakings regarding SEFs, including the definition of a SEF 6 and its scope and activities, the mandatory trade execution 7 requirement, the creation of a registration framework for 8 SEFs and the SEF's compliance with and enforcement of its core 9 principles. 10 We are very pleased to welcome to this discussion 11 two panels of experts who have kindly agreed to join us today 12 and share their insights, advice and recommendations. These 13 panelists represent investors, end users, dealers, academia, 14 exchanges and potential SEFs. We are grateful for your time 15 and participation and we expect that your comments will be 16 very valuable to the staffs as we develop proposals for a 17 regulatory framework applicable to SEFs. 18 I would like to note that today's roundtable is not 19 the only opportunity for interested parties to share their 20 views on SEFs with us. Each agency has open mailboxes into 21 which you may send comments on all of our OTC derivatives 22 rulemakings. Then, of course, the types of rules that we are 23 discussing here today will be formally published for public 24 comment before they are adopted. 25 I need to make clear that the remarks, questions or 10 1 lines of inquiry that you may hear from me or other SEC staff 2 today represent only our respective views and not necessarily 3 the views of the Commission, the individual commissioners, or 4 our colleagues on the Commission staff. I should also make 5 everyone aware that this meeting is being webcast and your 6 remarks will be recorded. 7 With that, let me give Rick an opportunity to offer 8 any introductory comments. 9 MR. SHILTS: Thank you, Robert. I also want to 10 welcome all -- everyone here today, especially all the 11 panelists for taking times out of their busy schedule to come 12 here and give us their views on these important topics. Also 13 to express my thanks both to Robert and the SEC staff, as 14 well as the CFTC staff, in arranging for this roundtable, as 15 well as all the other roundtables. It's a lot of work. 16 We're looking for a great discussion today and 17 these are some of the most important topics that we're 18 addressing with respect to these rulemakings and with respect 19 to the definitions and responsibilities associated with swap 20 execution facilities. 21 As Robert mentioned, the comments that I make or 22 other CFTC staff, similarly we're not speaking for the 23 Commission at the CFTC or the CFTC in general, and also to 24 note that, again, this is not the only opportunity for 25 interested parties to comment. We have a site on our -- a 11 1 box on our website where traders and others can submit 2 interested comments. And also, we're looking forward to the 3 comments as the rulemakings come out, which we'll take into 4 account over the next couple of months after we do publish 5 them. So with that, we're looking forward to a great 6 discussion today, and I'll turn it back to Robert. Thank 7 you. 8 MR. COOK: Thanks, Rick. 9 So with that, let me welcome our first panel. And 10 again, thank you for your participation today, and we're 11 looking forward to your contribution to this discussion. 12 Just as a way to kick it off, if I could ask that we just go 13 down the line and if each of you could give us your name and 14 state your affiliation please. 15 MR. SPRECHER: Hi. I'm Jeff Sprecher, chairman and 16 CEO of IntercontinentalExchange. 17 MS. SLAVKIN: Heather Slavkin, AFL-CIO. 18 MR. OLESKY: Lee Olesky, Tradeweb. 19 MR. MacDONALD: Ben MacDonald, Bloomberg. 20 MR. JEFFERS: John Jeffers, OTC Global Holdings. 21 MR. HARDING: Julian Harding from the Tradition 22 Group representing the Wholesale Markets Brokers Association. 23 MR. DuFOUR: Richard DuFour, Chicago Board Options 24 Exchange. 25 MR. DOWNES: Andrew Downes, UBS. 12 1 MR. DENIZE: Yves Denize, TIAA-CREF. 2 MR. DE LEON: Bill De Leon, PIMCO. 3 MR. COOK: Thank you. And on the screen, Steve? 4 MR. SEMLITZ: Steve Semlitz from HESS Energy 5 Company. 6 MR. COOK: Great. Thank you. And then we have 7 staff from the CFTC and the SEC bookending the panel here. 8 Maybe we could just quickly go down and you could introduce 9 yourself and your division, please. 10 MR. DeBORD: Jon DeBord, DCIO. 11 MR. PRICE: Greg Price, DMO. 12 MS. PATEL: Dhaval Patel, Office of General 13 Counsel. 14 MR. BRIGAGLIANO: Jamie Brigagliano, SEC Trading 15 and Markets. 16 MR. EADY: Tom Eady, SEC Trading and Markets. 17 MR. BAUGUESS: Scott Bauguess, Division of Risk 18 Strategy and Financial Innovation. 19 MR. SCHOTT: Sebastian Pujol, CFTC, Division of 20 Market Oversight. 21 MR. MELARA: Mauricio Melara, Division of Market 22 Oversight, CFTC. 23 MS. ADRIANCE: Riva Adriance, Division of Market 24 Oversight, CFTC. 25 MS. SEIDEL: Heather Seidel, Division of Trading 13 1 and Markets, SEC. 2 MS. BURKE-SANOW: Nancy Burke-Sanow, Division of 3 Trading and Markets, SEC. 4 MR. COOK: Thank you. So the format today is that 5 the staff from the CFTC and the SEC will ask questions and 6 throw it out there. Anyone will be free to jump in and 7 answer. When you speak, if you could please try to remember 8 to begin with your name so that those who may not have a 9 clear line of sight to you will know who is speaking. 10 Also, in the category of information that would 11 have been useful to have had earlier, just make sure you push 12 the button when you want to speak and it will be helpful if 13 you push it again when you're done so that we don't have 14 unnecessary feedback. 15 We have a -- this panel is going to run from -- 16 until a break at 10:45. So we have a lot to cover, we have a 17 lot of panelists and we have, I think, a lot to say about these 18 issues. And so, we just ask that when you make your remarks, 19 if you could limit your air time as appropriate to give 20 everyone a chance to weigh in. And we may need to move the 21 discussion along from time to time, too, because of a number of 22 topics we would like to make sure we get the benefit of your 23 thinking on. So bear with us if it comes to that. 24 So with that, we'll begin the first panel. There 25 are two main topics that we want to talk about on this panel. 14 1 One is the definition and scope of a SEF, and the second is 2 the scope of the exception from the mandatory trading 3 requirement for SEFs. And with that, let's begin with our 4 first question. 5 MR. BRIGAGLIANO: Yeah. I'm going to ask for your 6 indulgence while I engage in a sometimes tedious exercise of 7 reading a definition, but it's so important. It's the 8 cornerstone of our discussion today. And that's the 9 definition of swap execution facility in the legislation. So 10 it's defined as, "A trading system or a platform in which 11 multiple participants have the ability to execute or trade 12 security-based swaps, or swaps, by accepting bids and offers 13 made by multiple participants in a facility or system through 14 any means of interstate commerce, including any trading 15 facility that (a) facilitates the execution of swaps between 16 persons and (b) is not a national securities exchange." 17 So, I would like to hear what the panelists' views 18 are on the type of trading that would meet this definition. 19 Do the goals of impartial access and pre-trade price 20 transparency dictate a model such as a fully displayed 21 electronic limit order book for a SEF? Are there swaps that 22 currently have enough liquidity to trade on this type of 23 market? And I invite the panelists to jump in and give their 24 thoughts. 25 MR. OLESKY: I'm happy to start. It's Lee Olesky 15 1 from Tradeweb. 2 To start off with, we -- getting to the definition of 3 SEF -- our view is that a SEF should not be interpreted merely 4 as a central order book. And the starting point for that, I 5 guess, is the fact that there is such a definition or a word 6 in the law - SEF - and that's separate from a DCM. So we would 7 have to say that there is an assumption that this SEF is 8 something other than DCM, otherwise it would be redundant to 9 have both. 10 So our starting point is that this is something 11 separate and that there should be an opportunity to have 12 flexible trading models and protocols in order to support 13 that model so long as you, you know, you're fitting within 14 the core principles and you're achieving the policy 15 objectives of what the legislation is about. 16 So that's pre-trade transparency. You know, we 17 would argue things like electronic trading are great 18 indications of that pre-trade transparency, the ability to 19 execute a transaction at a price that you see on a screen, 20 and to have multiple participants. And when we get into the 21 multiple participant point, we think that means there should 22 be multiple participants providing liquidity and multiple 23 participants being able to access the liquidity. 24 MR. HARDING: Julian Harding. There are two 25 components, I guess, in the statute that you've carefully read 16 1 that are significant. The first being the multiple to 2 multiple issue, which I think intended to drive at creating 3 forces in a SEF environment, a genuine marketplace. 4 Multiple to multiple insistence implies a marketplace that, 5 in fact, the inter-dealer brokers for some 50 years or so 6 have generally operated and chestrated a great dynamic system 7 where each participant is a market maker and market taker in 8 the same breath. That is a different thing to the central order 9 book system that Lee is referring to and as you asked about. 10 The second component is the injection of the 11 language by "any means of interstate commerce," which is an 12 important element for us and for, I think, for the marketplace 13 in that it preserves the ability to transact an organized 14 transactions not just in electronic format, which one would 15 contend allows for a better chance of optimizing liquidity in 16 an institutional setting than just having been constrained to 17 an electronic format. 18 MR. DuFOUR: I would add that, you know, although 19 some products traded in the over-the-counter market may have 20 sufficient liquidity to where they -- there might be a 21 populated order book, that I would think in most cases, or in 22 many cases, it would -- a typical SEF would be more like a 23 request for a quote system where you might put up a trade you 24 want to do, but there is not necessarily going to be bids 25 and offers sitting in a book. 17 1 MR. DOWNES: Andrew Downes. I think as you look at 2 the definition or the rulemaking in terms of defining a SEF, 3 I think you need to take account of the liquidity of the 4 market that we're talking about. And I'd say, you know, 5 if you think that currently the markets that trade on 6 exchanges, the order of volume in terms of those exchanges is 7 in the hundreds of thousands, or at least tens of thousands 8 or thousands. I think if we look at data for, say, the 9 interest rate swap market, the most liquid point is the 10 10-year U.S. dollar interest rate swap. I think that trades, 11 if we look at June and say 508 times a day, then that gives 12 you a sense of the liquidity in that market. 13 If you look at the CDS market and take, for 14 example, the most liquid name in the investment grade index, 15 I think that's G.E. And that trades about 15 times a day. 16 So I think, obviously, there's an exercise to define what 17 should be SEF traded. And no matter where you draw that 18 line, there will be a spectrum of liquidity of that which 19 needs to be SEF traded. And therefore, the definition of 20 what is a SEF needs to encompass a plurality of models 21 because at one extreme, say if you look at certain most 22 liquid on-the-run indices, one might think that's susceptible 23 to a central order book approach, but at the other extreme, 24 if you've got something that's only trading 15 or so times 25 a day, that would probably be better traded through an RFQ 18 1 model, providing it meets the requirements in terms of 2 multiple participants, et cetera, et cetera. 3 So I think it's important to, for the definition to 4 encompass multiple models. I think that's the key. And 5 obviously to the extent that you encompass multiple models 6 that allows for innovation, creativity and competition and 7 that will drive liquidity to the best place depending on the 8 spectrum of the product that we're talking about. 9 MR. MacDONALD: Ben MacDonald, Bloomberg. I would 10 agree with a lot of the points that are being made. I think 11 one of the interesting things about the derivatives market is 12 that even looking at the benchmark space, there is ways to 13 trade around the benchmarks, and not everything is -- can 14 really be classified as just a pure benchmark or a 15 benchmark trade. 16 I think that really goes to the point that there 17 are multiple models that exist within the potential SEF 18 space. RFQs are very favorable, for instance, in certain 19 instances, but there are other models that intermediate by 20 buyers and sellers. And I think it's important, especially 21 in the early days of the rulemaking, to allow for all of 22 those models, as long as they meet the principles, to coexist 23 together. And I think that will drive, you know, that will 24 drive a lot of innovation and liquidity in the market. 25 MR. SHILTS: Could I just -- as you're responding 19 1 to the questions here, I'm just interested in you talking about 2 the RFQ models, whether people think that the existing RFQ 3 models would comply with the multiple to multiple requirement 4 in the definition. 5 MR. DuFOUR: Richard DuFour. You know, CBOE has an 6 existing RFQ model, which we call FLEX, which actually the 7 other options exchanges have something similar too, and there 8 are some roughly 200 trading firms that are qualified and use 9 that. It doesn't mean you get 200 quotes every time you, you 10 know, put in a request, but people monitor it on a full-time 11 basis. So I think it would meet your definition of multiple 12 participants. 13 MR. SCHOTT: If I could follow up to that. When 14 you say that up to 200 people could see the request, is that 15 only on one side or is that both the buy side and sell side 16 we would be able to see? 17 MR. DuFOUR: You can -- when you put in a request 18 for a quote, it can be -- you can put in a request to buy or 19 sell or you can just ask for a quote on a particular product 20 without indicating whether you're a buyer or a seller. 21 Anyone that's monitoring the system will see whatever you put 22 in. 23 There's typically then a time established by the 24 person putting in the request, a maximum, I think, of five 25 minutes, but generally much shorter, in which people then 20 1 respond. So and then at the end of the response, you have 2 the choice of doing nothing or going ahead and trading or you 3 could, you know, you could be a firm and you could actually 4 better the quote yourself and facilitate a customer. 5 MS. ADRIANCE: And just to follow up on that, 6 when -- once -- you said like 200 people see this bid offer, 7 just to -- either side with both -- you know, where they're 8 requesting both and 200 people see it, 200 people can 9 respond. What happens when there is a response? Is it 10 limited -- is the ability to transact at that -- with -- on 11 the response limited to the person who initiated the RFQ or is 12 others in the market able to also join in? 13 MR. DuFOUR: I believe it's limited to the person 14 that put in the request. 15 MR. OLESKY: This is Lee Olesky. We have a market 16 in the U.S. treasury market, government bond market, we trade 17 about 25 billion a day on average in the treasury market. 18 And our RFQ system, the way it works in the U.S. treasury 19 market, is you do have a request, prices come back from a 20 group of banks. That's part of the session is a private 21 session between the banks and the customer, but after the 22 trade occurs, we flash the price so that the entire market 23 sees what the price was for a transaction. 24 So in terms of, you know, sort of policy standard, 25 what are you trying to accomplish there, well, you have the 21 1 pre-trade price transparency you're after and you actually have 2 post-trade price transparency for the market to observe, but 3 you don't have a situation, at least in the treasury market, 4 where you have the risk of interference between that 5 privately negotiated transaction between two institutional 6 participants. And that seems to have worked fairly well in a 7 number of fixed income markets. 8 MS. ADRIANCE: If I could just ask one more 9 question, to go back before we move on, because right before 10 we started talking about the question about the RFQ and what 11 does that mean, there was, I have it down that, I think it 12 was -- I'm sorry, Bloomberg -- I'm going to mess up your 13 name -- Ben MacDonald. Yeah. You said that -- you referred 14 to the fact that there is different -- you know, some 15 instances in an RFQ model might be the appropriate model to 16 execute in that particular swap, but then there may be other 17 instances where another model would be appropriate, could you 18 talk more about what is the other models that you're talking 19 about. 20 Are you talking about central limit order book or 21 are you also thinking of third or fourth in time, you know, 22 to other models besides central limit order book and RFQs 23 because so far we've talked about those two. And it has been 24 mentioned like but there is voice brokers -- but I wasn't 25 sure what you were referring to. 22 1 MR. MacDONALD: Sure. At Bloomberg we actually 2 operate two models for trading. One is a traditional RFQ 3 model where participants go out and launch an inquiry to 4 multiple dealers and then execute by inquiry. We also 5 operate a model, which we call single dealer franchise. And 6 essentially what we do is that we allow all of the liquidity 7 makers to use Bloomberg as a distribution mechanism to 8 display over a page on Bloomberg the -- you know, where they 9 might have an ask, for instance, or where, you know, the 10 markets that they're interested in, obviously low liquid 11 stuff, is available in that market. 12 Our customers can look at as many pages as they 13 want over the system, which means that they get, on the one 14 hand, the price transparency and discoverability of looking at 15 multiple quotes, but at the same time when they get to the 16 actual executing part of the transaction, it allows them to 17 execute on a one-to-one basis, which is important in many 18 cases, you know, for both the buyer and the seller of the 19 transaction who have both got concerns about that, you know, 20 trade. So, you kind of get both of those models. 21 If you draw -- you know, in the treasury market and 22 the cash markets where that model is a lot further along 23 because there's a lot more electronic trading in those 24 markets, you know, you can go as far as looking at people to 25 display firm pricing in those markets in a very liquid space, 23 1 which allows people to essentially, you know, achieve the 2 same, you know, very fast level of liquidity. 3 MR. OLESKY: If I could just tag onto what Ben said 4 because we have similar concepts, and the variations in 5 different models, which I think was your question, we equally 6 have a model that's similar to what Ben described, but is yet 7 slightly different in that we call it streaming prices. So 8 we have a different terminology for it, but in the interest 9 rate swap market in Euro denominated swaps, a client can come 10 in and request a stream of live prices from the bid side and 11 offer side from a particular bank and have a live market in 12 front of them to click to trade. Is that really an RFQ? 13 I think you do start to get into these definitional 14 issues of what's an RFQ, what's a request for stream, 15 what's -- but fundamentally what we're talking about here is, 16 you know, access to these live markets, real good pre-trade 17 transparency, post-trade transparency, and fundamentally an 18 ability for the market and the buy side customers and the 19 liquidity providers to be able to transact, which has been 20 evolving, you know, in the last 12 years we've been in 21 business from a more standardized type of business and fixed 22 income such as U.S. treasuries out to other products and now 23 into the derivative space. 24 But we think it's that flexibility that is so 25 critical so that you can have the innovation to accomplish 24 1 what the market needs are so long as it's fitting within the 2 policy objectives that have been laid out by the legislation 3 in the core principles. 4 MR. VISWANATHAN: Vish Viswanathan from Duke. I 5 guess one of the things that concerns me about this RFQ 6 model, if I'm an end user and I'm going to two different 7 platforms, I literally have to request a quote from both 8 platforms. I can't actually compare prices. There is no way 9 to aggregate as there would be in stock markets. 10 So the question then is, from an end user 11 perspective, is this the most efficient way or is there a way 12 to kind of induce more cross-market competition. If there 13 are a multiple of these entities that are created, how do you 14 induce them to compete because the whole idea is to have 15 innovation and competition and so how can you do comparison 16 shopping in this context is the big issue. 17 MR. HARDING: Julian Harding. It strikes me that 18 there is a trenchant change in the air and all different forms 19 of execution that exist now will have to adapt in some form. 20 I think we should be under no illusion about that. And as an 21 interdealer broker, as an interdealer broker myself and as 22 representing the association, we are under no illusion at all 23 that we're going to have to adapt in certain areas to meet 24 the regular fee standards and policy standards that are 25 coming out over the last year. And in the same way, it seems 25 1 to me that probably an RFQ system, given it's a single 2 requestor, will have to make some adaptations as well. 3 MR. DE LEON: Bill De Leon, PIMCO. I think it's 4 important to differentiate because as we've seen through the 5 evolution of the stock market where there have been multiple 6 sources of prices and central order books, that both the 7 order books themselves, as well as the end users, have spent 8 an enormous amount of time, energy and their own money to 9 come up with systems to pull and pull that information 10 together so that they can look at multiple sources of 11 information. 12 It's very analogous to what goes on now in the OTC 13 market, especially in fixed income, where as an executor of a 14 trade, you need to do price discovery talking to multiple 15 dealers to get that information and/or looking at other price 16 sources. Not to show favoritism, but a few of the people 17 here who have talked have some very good SEF-like models already. 18 I've traded on them, as many people in this room probably have, 19 and there are pluses and minuses to them. And they're a 20 combination of RFQ, streaming quotes, central order books. 21 And they provide a lot of pre and post-trade transparency. 22 What SEFs in our view will not provide is 23 continuous price making because there is no central market 24 maker, per se. And that's something we should focus on. In 25 addition, as we move into the derivative market and you move 26 1 away less from the widget market, and not to pick on anything 2 in particular, if you take a stock, it's very nice because it 3 is a defined widget. 4 IBM, for example, is a single stock name. There is 5 one piece of information: it's traded, how many shares trade, 6 or are perceived to have traded at what sets of prices. And as 7 you move away from that and you move to derivative markets, 8 you wind up with a lot more bespoke characteristics. So 9 getting that information and tracking it becomes much more 10 difficult. And that's a plus and a minus. It allows 11 standardization of things to be moved away to customization 12 to meet different product needs for different end users. 13 It also has the stigma of as you move away from the 14 widget effect to a customized event, the footprint left by 15 those using it becomes much more known. For example, certain 16 people trade certain instruments. It will be known that only 17 certain people do that. If everyone sees that information 18 and it's not aggregated, you're going to know who is playing 19 and if they are a large player or there is a small player, 20 that information is given away. 21 So, you lose that concept of anonymity, which I think is 22 an important function of the market, because while people have 23 a right to know what is out there and where the fair level of 24 a trade may be prior to a trade, not everyone has the right 25 or ability to know every single trade in every detail because 27 1 there are things that are not standardized and is a function 2 of market liquidity and need. 3 So I think that any SEF model or SB SEF model will 4 have to take that into account and allow certain 5 flexibilities because otherwise, you'll wind up with people 6 not wanting to use it and as we've just alluded -- we have 7 some very good statistics here -- there are a lot of things 8 that trade -- that don't trade that actively especially in 9 the standard form. 10 Ten-year swaps. A lot of players use 10-year 11 swaps. They look at that quote every day. There aren't that 12 many trades that occur every day. And the reality is when a 13 10-year swap does trade, most end users don't trade the 14 10-year swap. They will trade something that is around a 15 10-year swap. They may trade something that's 10 years and a 16 quarter, they may trade something that's 9 years. The coupon 17 will be different, the dates will be different, there are 18 all these unique functions. 19 So to the extent that you're forcing things to be 20 on an SEF and all that information needs to be given up, 21 you're going to be telling an awful lot of people a lot of 22 customized information about the end user which will hurt 23 them and not help them. In addition, it will make using SEFs 24 harder because to track and display that information will 25 sort of be a player's curse. I'm sorry for taking so much 28 1 time. 2 MS. SLAVKIN: Heather Slavkin. I understand the 3 concerns about accommodating a lot of different business 4 models. I'm not sure I agree with the argument he just 5 made, however, because the legislation envisions that the SEFs 6 and the trading requirement will apply to products that have 7 to clear. And the clearing requirement applies to liquid 8 products. Standardized and customized transactions are not 9 going to be required to clear. So I don't think that we 10 should take customization into account when determining what 11 types of information should be disclosed to SEFs or how SEFs 12 should operate. 13 MR. COOK: Let me sort of ask that -- sort of -- 14 and Bill or Yves, you can pick up on this. Sort of one line of 15 reasoning that I think we're hearing is that it's a diverse 16 market. Lot of different types of products, lot of different 17 type of platforms need to take that into account in order to 18 accommodate the status quo to some extent. One could also 19 argue that the statutes require the agencies to think about a 20 new market, a new paradigm for how swaps are being traded and 21 which would have the effect that some of the existing models 22 might not work with that new paradigm. 23 And I would be interested in getting, you know, 24 also particularly from the end user perspective, in the first 25 instance, but everyone's perspectives, on how we weigh those 29 1 concerns. And if the idea is that we really are meant to be 2 moving to a new market structure, how do we get there? Is it 3 a one time leap or is it a transitional exercise? 4 MR. OLESKY: This is Lee Olesky, Robert. The 5 interesting thing is -- 6 MR. COOK: If -- 7 MR. OLESKY: Oh, I'm sorry. 8 MR. COOK: If you could Yves go after. Go ahead, 9 but -- 10 MR. OLESKY: I just want to make a quick comment, 11 which is the derivative markets right now are not there yet. 12 If you look at the percentage of the trades electronically on 13 vehicles such as Tradeweb, Bloomberg, any of these other 14 entities, the market is not there yet. It's a small percentage 15 of the market relative to the total size of the market. So 16 in terms of change and structural change here, you would have 17 significant structural change if there was a mandated 18 requirement to take most of these standardized transactions 19 and run them through a SEF because that's not happening 20 today. 21 So first and foremost, establishing the mandatory 22 nature of this that's in the law and defining what's standard 23 is going to create some significant change in the 24 marketplace. That's how we look at it relative to other cash 25 markets. This is still a market that's evolving towards 30 1 electronic and it's still only a small percentage of the 2 market. 3 MR. SEMLITZ: Steve Semlitz from HESS Co. In the 4 energy markets, I think we have just the opposite occurring 5 where everything was OTC many years ago and now an inordinate 6 number of products do trade in a derivative form. The energy 7 market probably has the largest amount of customized types of 8 trades. Many of those trade OTC and then get cleared, but 9 overall, there are tremendous numbers of derivatives trading 10 in forums like ICE, that trade and clear. There is no -- 11 you're not requesting quotes. 12 And the real issue for everyone is can you attract 13 enough eyeballs and enough people to look at the screens and 14 figure out where there is an opportunity to trade because 15 it's an infinite universe of trades that can occur. The 16 definition of what a trade that has to be on a SEF will 17 be -- that's where the -- 18 (Interruption to video call.) 19 MR. SEMLITZ: -- because in the energy markets, 20 there are just millions of types of customizable trades that 21 occur. And you get quotes by going to a market maker and 22 saying please spend the time to think about how to quote this 23 product and then quote it. The question I would have is when 24 you go to a request for a quote type of system and people 25 aren't quoting their best customer and their high volume 31 1 customer, whatever it is, whether -- 2 (Interruption to video call.) 3 MR. SEMLITZ: -- for anyone to continuously go 4 through all the requests for quotes that you might need. So 5 where you draw that definition is going to be the key. 6 MR. DENIZE: I just wanted to underline -- this is 7 Yves Denize -- from the end user perspective, wanting to 8 underline the difficulty in standardization. As you look at 9 the end users in the various industries now, for instance, 10 our life insurer, for instance, has very specific needs and 11 comes to the derivatives market for very specific risk 12 mitigation needs. And it can't be overstated that defining 13 that standardization process in a way that can be freely 14 traded where you won't need a phone call and a discussion 15 about those terms is going to be very difficult. 16 And that process that will determine what should be 17 traded on a mandatory basis, should not be -- should be an 18 organic process that really reflects the growth in the market 19 and the growth of these trading facilities. And it should be 20 done in an organized manner, a centralized manner, so that we 21 have - are able to settle our expectations as end users when 22 and how the types of derivatives that we are used to being 23 able to access are going to be forced onto a trading system 24 that may change our model as to how we're mitigating risk. 25 MR. DE LEON: Bill De Leon. Yeah, I think it's 32 1 important to separate a few things here, which we've sort of 2 not brought up and I know it came up in some of the previous 3 panels. So I apologize for revisiting. The important things 4 in our view is to reduce systemic risk in the system and 5 ensure that people have access to decent price information so 6 that they're transacting approximately where the market is. 7 And the things to achieve that are not necessarily driven by 8 standardization of everything down to the widget level. 9 Clearly, the more widget-like something is, the easier it is 10 to turn it into a central order book. 11 If you look at futures, which they're a great 12 example, farm contracts, S&P futures, we can talk about many 13 different things. They're very standardized instruments. 14 You know what all the details are. And they lend themselves 15 very easily to being traded on an exchange. Everyone knows 16 exactly what it is. You don't have to stand behind and 17 figure out customized things. They've lent some issues to 18 it, however, because they don't necessarily meet end user 19 hedging abilities, especially FAS-133 and cash flow hedging. 20 And that's something. But you've achieved a widget thing 21 which allows a lot of people to look at a lot of information 22 and trade and transact. 23 The nice thing about a SEF or the concept of it, in 24 our view, is that it allows multiple participants to access 25 and trade with anonymity. That's the important thing because 33 1 if you take that away, you're just creating sort of a 2 marketplace for people to sort of share information and try 3 to get things done. The point is, and I know this is on a 4 later panel, a SEF needs to ensure that people who 5 participate in it get the information they want or can 6 provide it to trade. And then when they do a trade, it gets 7 given up and cleared so that it goes through and reduces 8 systemic risk and you're not worried about who you traded 9 with. 10 Our clients don't want us, necessarily, trading 11 with the guy who is making a market in derivatives sitting in 12 his, you know, basement. That's not what we want, right? 13 The SEF concept would allow clearing and it would take it 14 away. So if the guy in a basement was doing it, he would be 15 doing it via clearing mechanisms. Some would be providing 16 capital and we would reduce risk. So if he wanted to provide 17 the best price, we would be happy. So I think we need to 18 focus on these aspects, and I think maybe that's getting to 19 what you're talking about. 20 So we need to sort of not lose sight of what we're 21 trying to achieve through the SEF, which is allowing multiple 22 people to come together, provide prices, provide information, 23 allowing more people to transact and not having sort of an 24 oligopolist situation where only dealers provide prices. 25 We're very much in favor of that, but it has to be done in a 34 1 way realizing that when there are lots of unique things, it's 2 more difficult to get that information, as well as by the 3 nature of unique information, it tells you a lot about who is 4 playing so as you move away from that. 5 And one other thing which I think is very important 6 and I think we shouldn't lose sight of, and I'll stop taking 7 so much time. And I apologize. Is that there is a huge 8 legacy set of positions that exist and the ultimate goal to 9 reduce systemic risk is to move those positions over to a CCP 10 like framework. How do you handle those positions so that 11 they can be traded in a liquid manner where the end user -- 12 and the important thing to remember -- end user is mom and 13 pop because institutions frequently represent very small 14 investors that just pool their money with a large investment 15 fund -- that this is sort of the view. And you don't want them 16 paying high bid ask or information fees to make that happen. 17 MR. EADY: So it sounds like the panelists believe 18 that there are some desirable characteristics of the RFQ 19 model that they would like to preserve. One of the things 20 that we're focused on, obviously, is pre-trade price 21 transparency, and we talked about it a little bit here. The 22 question I would have for the panelists is do you believe 23 that there is adequate pre-trade price transparency with the 24 RFQ models that exist today or if not, what would you suggest 25 as improvements to pre-trade price transparency to help us 35 1 better meet that policy goal. 2 MR. DuFOUR: Richard DuFour. I can only speak to, 3 you know, what we do today, which is not really a SEF, but 4 the FLEX system I mentioned before, which is targeted at the 5 over the counter market kind of draw from that impact. The 6 last two or three years has -- the volume has increased 7 substantially, I think, because of what's been going on. And 8 I would argue that there is, you know, adequate transparency 9 in that. And the key thing, you know, in responding to a quote 10 is to know the terms of the trade, which you can vary, and, 11 you know, price and size. And I think those elements are all 12 met. And I suspect in some of these other systems, but I 13 won't speak for them. 14 MR. OLESKY: I guess I would just reflect what I 15 was saying a little bit before, which is the state of the 16 market right now with derivatives trading through what we 17 envision to be SEF like entities, such as some of the 18 entities that are on this panel, is still in its early days. 19 And as a result, you don't have a lot of activity, not nearly 20 the same activity that you have in the other market. 21 So what should happen here naturally, as more 22 activity moves to these vehicles, you're going to have 23 enhanced transparency in terms of more participants joining 24 in, more prices coming in, more competition because for the 25 first time, this is going to be a mandatory process. And if 36 1 that market today, and let's just, you know, lump together 2 credit and rates, is less than 5 percent through these kinds 3 of vehicles, if you envision a world where there is 75 percent 4 of the activity going through those vehicles, you're going to 5 have a lot more transparency. 6 MR. MacDONALD: I would agree with a lot of the points 7 that are being made on the SEF. Ben MacDonald from Bloomberg. 8 I think that the derivatives market is relatively early in its 9 stages. I think we will draw comparisons in the cash market 10 where the RFQ process is very prevalent, as are some of the 11 other things that we're talking about here. And I think 12 that, you know, the important thing from a kind of rulemaking 13 standpoint is really going to be allowing multiple models to 14 exist because I think that's ultimately going to promote 15 liquidity. 16 I think that the buyers -- the takers and makers 17 of liquidity, will end up, you know, in the place which is 18 the most efficient and suits their needs. And I think if you 19 look -- if you draw the analogy today in the cash base, there 20 are multiple models and people use those models for different 21 reasons. But there is a vast majority of trading that 22 happens over, you know, the RFQ model over the streaming 23 model, you know, as well as other end users as well. And I 24 think the really important thing today is to allow that 25 market to continue to grow. And I think the risk we're 37 1 just -- we're then posing, one model is that it will achieve 2 the opposite to that. 3 MR. DOWNES: Andrew Downes, UBS. I would say that 4 if -- even with an RFQ model, if that's encompassed in the 5 SEF definition, should provide more transparency than 6 currently exists because you need a trigger for the 7 discussion and that trigger will be a price around the 8 benchmark tenors or trades. And having that will 9 obviously provide more transparency. 10 MR. DuFOUR: I would also add that this is really 11 on the issue of the customization, that I believe a great 12 deal of the customization that's done, I would even say, you 13 know, the majority of it, can be done in a standardized form 14 and that one of the challenges for you as regulators is to make 15 sure that customization doesn't become a loophole for avoiding 16 the exposure and clearing it. 17 MR. HARDING: Julian Harding. Moving slightly away 18 from the RFQ concept and talking a little more of the existing 19 interdealer, the broker venue, the existing structure is a 20 little bit -- is not talked about very much, which is that there 21 is a sort of an outer area of end user participants and there 22 is an inner area of banking or dealing participants and there 23 is a third inner area in which sit the interdealer, the brokers. 24 In terms of the transparency issue, I think a lot 25 of this will be attended to by the fact that in the new 38 1 imagined environment, there will be introduced a great deal 2 more different sorts of participants by the mandated areas 3 that Lee just alluded to. First off, the fact that clearable 4 trades now will be mandated through SEFs and DCMs will do 5 something. Secondly, a whole new batch of players 6 determined -- yet to be fully decided upon, but certainly in 7 amongst the legislation, will be forced to go through SEFs as 8 well. Brand new entrants to that marketplace to be -- what I 9 would say the previous marketplace, the one we described as a 10 SEF. 11 So, that is quite a dramatic change in and of 12 itself and the transparency might even be contended -- the 13 transparency desires may come from primarily a lot of those 14 participants who, in fact, now will be fully participating in 15 those marketplaces, which is rather a large change. I think 16 it's the same theme, but extending the theme that Lee was 17 mentioning. 18 MR. COOK: I want to make sure we get, you know, 19 all the staff's questions in there. 20 MR. BRIGAGLIANO: Can I ask the panelists whether 21 they believe there should be a firm quote requirement for 22 swaps and if so, should that depend on a liquidity threshold? 23 MR. DE LEON: Bill De Leon. I think you've seen 24 the market adapt to that concept already, that people put out 25 quotes on many things in the OTC market, and that's how a lot 39 1 of people get transparency. And the understanding is that 2 people who put quotes out, understand that there's going to 3 be some minimum size that they will transact on. Sometimes 4 the quote information has that price information as well as a 5 quote size or there is some indication. 6 We've found that dealers who send us information 7 who don't include size and/or when you go and say I want to 8 trade on one side of your market. And they say, oh, no, that 9 was just an indication; we're not willing to trade. You tend 10 not to get a lot of repeats. 11 So, I would agree that having a quote without some 12 concept of size and maybe it displayed or maybe there is an 13 understood rule and obviously depending on the product, the 14 size needs to be different, I think that makes sense because 15 people flashing prices on screens not standing behind them 16 I think gives -- lends to manipulation and/or misleading 17 information. And it doesn't help anyone to have that, to say 18 I'll buy it here at 25. Oh, okay. I'll sell it to you at 19 25. Oh, no-no. That's only not good for any real size or I 20 take it away. I think that hurts the market in liquidity in 21 any process. 22 So, I think the concept of a firm size is good 23 because it enforces that when you see this price there, well, 24 it may not be all the size you want to do. At least you know 25 you can do a transaction and it's not getting you to waste 40 1 your time because otherwise, you're going to give away 2 information to somebody. 3 MR. DOWNES: Andrew Downes. I think in answer to 4 the second part of your question, I think that definitely 5 does have to be a liquidity requirement for firm quotes if 6 there were to be that requirement at all. Obviously, there 7 is a lot of disparate standardized contracts that trade, and 8 some in very small volumes, and I think it's obviously 9 difficult to maintain prices across, you know, a number of 10 tenors and hundreds of names all at once. That's just not 11 feasible from a market making perspective. 12 So, I think you need to have a liquidity requirement 13 and obviously, to the extent you were going to require firm 14 prices, that should really just be around the benchmark and 15 then obviously the benchmark price that is shown can be a 16 trigger for a discussion if someone wants something other 17 than the benchmark. 18 MR. OLESKY: Hi. It's Lee Olesky. One of the 19 beauties of the way the market works right now, and this is 20 following up on the two previous comments, is because it's 21 fully disclosed, you know who is putting out the price. You 22 know who you're asking for the price for. And I think we 23 heard from Bill, if you are putting out prices and you're not 24 standing behind them, you're not going to be asked for prices 25 in the future. 41 1 And so we see that all the time. If you're not 2 going to stand behind your prices, you know who it is, and 3 this is the key to the anonymity of the relationship between 4 the user and the liquidity provider, you're not going to go 5 back to them. So, there's a self-enforcing mechanism 6 that works quite well in terms of standing behind your 7 prices. And, you know, we've experienced that in the last 12 8 years. If you're putting out prices and you're not standing 9 behind them, you're not going to get that inquiry in the 10 future. 11 MR. SEMLITZ: I want to grab a question because it 12 really goes to futures and derivatives and there is an entire 13 industry that grew up around trading futures and those firms 14 are canceling inordinate amounts of their orders, maybe up to 15 97 percent of their orders. So, their orders aren't there to 16 provide liquidity and many of their orders aren't there to 17 really execute they're there to paint the market. We've got 18 rules already in place to regulate this. And that's the 19 current state of the market. So the question I have is why 20 are we -- you know, why are we so concerned with request for 21 quote and whether people will be firm on their quotes, but 22 we're not doing anything about the futures markets. 23 MR. MacDONALD: Ben MacDonald, Bloomberg. Just to 24 kind of follow up on a lot of the comments. I think this 25 goes back to the point of having, you know, different types 42 1 of SEFs and to draw -- I mean, you know, I think one of the 2 issues here is because the derivatives market is so new in 3 the electronic trading space, I think we all draw comparisons 4 with how we saw the cash space evolve. Natural evolution in 5 the cash space was firm pricing and, you know, you've got to a 6 point now where you can access hundreds of quotes on a 7 system. And the way the liquidity providers differentiate 8 themselves, one of the ways, is by firm quotes and, you know, 9 people stand up to larger sizes from depending on what their 10 appetite is. 11 So I think, you know, it's one of the reasons why 12 it's very important to allow different models to exist 13 because I think you'll naturally get this evolution, which 14 we've seen in the cash space, across, you know, both the rates 15 and the credit market. 16 MR. OLESKY: Just to make another point on that is 17 the -- in a sense, anonymity allows for backing away. I 18 mean, I'm not going to be critical of the futures market; 19 that's not my area, but I think when you have anonymity, it's 20 easier to back away. When you don't have anonymity, when it's 21 a privately negotiated structure or deal, you can't back away 22 or you won't do that business in the future. 23 MR. DE LEON: One of the things, though, to sort of 24 go back to that is as you -- you know, you look at your 25 model, which works very well, people put up a quote and you 43 1 know which dealer it is. Conversely, when you look at the 2 futures market, this whole concept is that it's anonymous. 3 Whatever goes on, though, what you do know is if someone puts 4 up a price and a size, they may be painting the screens as 5 sort of the example that was laid out, but you do have the 6 ability to do that trade. And there are times when people 7 paint the screens and the market comes out and they don't 8 pull their level quickly enough and they're held to that 9 trade. 10 And also you tend to know when people paint 11 screens, that is to say, well, if the market is $1 at $2, and 12 they put a level out at $6, but they're willing to sell an 13 awful lot of it, well, they know, and everyone knows, well, 14 they're just painting the screen because it's so far off the 15 market, you're not going to look at it and you know it's 16 someone playing a game. Unfortunately we've not seen that, 17 and that's just gamesmanship and it's -- I'm not a fan of it 18 or in favor of it, but if that person puts out a large block 19 instead of at $2, at $2.02 and it's a large block, well, you 20 might lift them on that offer because you may go, you know 21 what? That's not painting the screen. That's them asking a 22 little bit of bid-ask for a large block. So I think you need 23 to differentiate the two. 24 And the market does self-reinforce because people 25 tend to look at it and go, these guys are constantly showing 44 1 blocks off the market and I'm not going to look at it. So I 2 do think that people self-correct for that even though it's 3 anonymous. What I just want to get back to is the concept 4 that if you're going to have a SEF model, the concept -- and 5 it's going to be cleared, the concept is people having access 6 need to be anonymous and be able to do things. 7 So, I think some of the points that have been made 8 here is clearly there would be less customization over time, 9 which I think is a natural outcome of the market, but we have 10 this huge outstanding book of business that still needs to be 11 traded and managed and eventually possibly cleared to reduce 12 systemic risk to the system, which is the ultimate good. And 13 how do you do that if you force the new model not to 14 incorporate the existing body of positions. 15 MR. VISWANATHAN: I just want to follow up on this. 16 I guess to me one of the presumptions of the Act is clearing 17 itself will induce innovation. The fact you clear will lead 18 to more standardization over time and hopefully it will 19 mold - instead of just a request for a quote more actual 20 quote-setting behavior, but it might be that the regulators 21 need to set thresholds that if you meet certain volume hurdles, 22 you're a big enough market that you want to prod more quote 23 behavior because posting quotes, I think, makes a big 24 difference to end users. 25 It goes from a negotiation relationship, which is 45 1 kind of person to person -- RFQ model each person is getting 2 a separate set of quotes. I'm seeing prices, but I'm not 3 seeing the quotes that somebody else gets. But the posting 4 quote model, it's completely different. I know the quotes 5 that other customers have got for some standardized product. 6 So it changes the nature of the game in essence. 7 MR. COOK: So you think if an RFQ model had a 8 certain volume, then maybe you would say it needs to 9 transition to a full disclosure model. 10 MR. VISWANATHAN: Yeah. I would say that that 11 would be part of the regulatory kind of rule-making or 12 decision-making. 13 MS. ADRIANCE: That actually leads very nicely into 14 my question, which is there have been several people who have 15 talked about RFQ models, the versions that have streaming 16 quotes, you know, whatever you want to call them. From what 17 we've heard before and today, the different versions of these 18 models, you know, have differed. There is many different 19 systems that's been mentioned. In fact, several people are 20 saying we need to have different models. 21 And I'm really curious, if, you know, if we as 22 regulators, in addition, you know, we say everybody knows the 23 control limit order book. We know how that works. We know 24 that it has certain advantages. What I think is being said 25 is that there are certain instances where a central order 46 1 book just isn't going to work, like liquidity or whatever has 2 been mentioned. 3 So, in addition to a central limit order book, 4 several people have mentioned other models. What I'm trying 5 to get more information on is if we could actually -- I 6 realize that a number of you already have models in place. 7 But if we were going to design a new model, if as regulators 8 we were going to say, well, what is the best additional model 9 in addition to a central limit order book that we want to . 10 encourage. Separate from where we are today, where do we want 11 to get to to deal with situations where there is the, you say, 12 the less liquid slots. 13 What would that model be that would provide 14 in a sense, the regulatory goals of the Dodd-Frank Act, 15 you know, pre-trade price transparency, yet there is, as it 16 has been mentioned, anonymity, you know, that there is this 17 marketplace because many of the RFQ models, people have said, 18 well, they're multiple to multiple because you have multiple 19 people sending out requests for quotes and you have multiple 20 people that can respond to that individual person. 21 Is there something between that RFQ model -- I know 22 there's been -- streaming quotes have been mentioned. That's 23 something that can go to a lot of people who can react. But 24 is there something else that is either out there that you're 25 aware of, something between central limit order book and 47 1 RFQs, that is a beefed up version of an 2 RFQ model or some other model that is either out there or 3 that you can envision being a practical step if we're trying 4 to evolve, you know, certainly as we've mentioned, as certain 5 swaps become more liquid, if there is some and we have some 6 kind of standard that we say, okay, if it reaches a certain 7 liquidity it needs to move up the gradation of, in terms of 8 models, is there something between this RFQ model, and the 9 central limit order book. 10 One thing that was mentioned was streaming quotes, 11 but that's not something that you can transact or 12 that you can expect to necessarily transact I think. So I'm 13 curious about what is -- what else might be in the middle 14 between RFQs and central limit order book that's either out 15 there or that you can envision being a useful thing to have 16 out there. 17 MR. SPRECHER: This is Jeff Sprecher from ICE. 18 What exists in the futures market for exactly that need, 19 particularly in the options on futures market, is the 20 requirement that there can be pre-trade conversation, 21 arranging of somewhat customized deals. But when the moment 22 comes to actually cross the trade, it is advertised, for some 23 period of time, to a broad market before it's crossed. 24 So in other words, if the buyer and the seller have 25 already found each other, one person puts up their bid or 48 1 their offer, counts to three, and then the other one can 2 cross it. That allows for a price improvement capability 3 inside that bid and offer. 4 Let me make one other point while we're on that. 5 And one of the issues that we have at ICE, as an operator of 6 futures exchanges, many to many OTC markets, dealer to 7 client OTC markets and inter-dealer OTC markets. So really 8 covering across all trading types. One of the concerns we 9 have is that in both Commissions requirements, to institute 10 the core principles, as well as the aspirations of pre-trade 11 price transparency and some of the other aspirational aspects 12 of the bill, that we not try to go through market type by 13 market type or market by market and somehow give a broad 14 set of exemptions. 15 It seems like that, however, those core principles 16 are implemented, they ought to be consistent across all 17 models, and then if the Commissions want to allow various 18 trading models to exist, they shouldn't exist because of some 19 kind of regulatory arbitrage differences between the way the 20 core principles are implemented. 21 MR. HARDING: Julian Harding. We seem to have 22 stuck with, a little bit, the RFQ model, which as I 23 cautioned before, could -- I think needs to adapt a little to 24 adhere to the strict definition that was read out at the 25 beginning. The -- I would offer the inter-dealer broker 49 1 over-the-counter market places do offer what you've been 2 suggesting might be a middle road between a central order 3 book and an RFQ. 4 The over-the-counter markets to date generate 5 liquidity when compared pari passu to any other venue for a 6 similar product area in an overwhelmingly greater manner. And 7 if we agree that liquidity generation or liquidity 8 preservation and improvement is a central tenet of safety 9 and certainly is of the ability to clear trades and therefore 10 create further levels of safety, then the inter-dealer broker 11 model, the over-the-counter existing models can offer a lot 12 of what you want in that we are looking at multiple parties 13 transacting with multiple parties, accepting bids and offers 14 from multiple parties, in a very dynamic environment where 15 each player in that marketplace can be, at any one time, and 16 even simultaneously, a market taker and a market maker. 17 MR. EADY: So let's expand on that a little bit 18 because my understanding is that the inter-dealer market is a 19 market among dealers and not others. So when we get to the 20 other policy goal of impartial access, are you suggesting 21 that we open the inter-dealer market up to other participants 22 who want to be involved for the benefits that you just 23 described? 24 MR. HARDING: Yes, and I said in my previous 25 comment, that the new environment clearly envisages, by 50 1 virtue of mandating certain existing parties to go through 2 the SEF or the DCM and furthermore, mandates extra new 3 parties that previously were not, possibly due to the fact of 4 counterparty credit issues, were not having easy access or 5 any access to those marketplaces. In the new environment, in 6 newly cleared products, the counterparty credit issue is 7 removed and those same newly mandated participants are going 8 to have free access to those same marketplaces. So yes, 9 indeed. The answer is yes. 10 MR. MacDONALD: I think there is -- this is Ben 11 MacDonald from Bloomberg. I think one of the things that's 12 getting a little bit lost in the debate, perhaps, is that, 13 you know, a large part of the derivatives market is what I'll 14 call semi-standardized and by that, you know, the benchmark 15 is where everybody is going to be basically looking as a 16 reference but the reality is that a lot of the way trading is 17 done on these standardized products is the cash flows and 18 models slightly outside of, you know, the point which is made, 19 you might want to do a 10-year swap with, you know, three months 20 forward, or something like that. 21 And so, the risk is, it just becomes very, very 22 cumbersome, from a technology perspective, to be able to, you 23 know, if you change the model from an RFQ model, or don't have 24 that in your RFQ model, the RFQ model is actually the most 25 efficient model that we see today because it allows, you know, 51 1 people to go out with this kind of semi-customized trade and 2 dealers can quote it or the price makers can quote that 3 transaction and come back, come back with a price. 4 It's just -- I think it becomes very, very 5 complicated, you know, the more kind of -- the more you 6 change that model and the more you get into the kind of 7 semi-standardized space, I think those models become very 8 complicated to maintain. So I'm just trying to kind of 9 understand the thoughts around that. 10 MR. OLESKY: Yeah. One comment I would like to 11 make is I think what happens here, again, as we're trying to 12 anticipate the effect of these rules on the marketplace, and 13 I'll go back to that point. With the establishment of SEF, 14 as more and more activity goes through SEF, I think you're 15 going to start to see potentially more volume occurring, 16 which will drive business into different types of trading 17 models. And it will be driven, to a large extent, I think by 18 market participants. 19 If there were thousands of participants, as there 20 are in exchange models, who wanted to participate in a 21 10-year swap, then I think you would have an order book real 22 quick because I'm pretty sure, you know, all of us would want 23 to open up our markets to an exchange type environment if 24 there are enough participants and enough liquidity to support 25 that kind of model. 52 1 And I do think we will see an evolution, and we've 2 seen this evolution in other products where, for example, you 3 have the treasury market, you have other markets. As they've 4 gone electronic, they become more liquid, they pull in more 5 participants, and you have more of an order book type of 6 model occurring, whether it's in IDBs or, you know, through 7 Tradeweb or Bloomberg or wherever. I think you're going to 8 see an evolution in certain products where they will move to, 9 naturally move to different types of trading protocols. 10 MS. SLAVKIN: I have a concern that the 11 conversation about maintaining the RFQ model is the status 12 quo that you mentioned earlier. And it seems to me that the 13 legislation envisioned, and what we should be trying to 14 aspire to, is something as close to an order book as 15 possible. As soon as we start adding the human element in, 16 you invite the possibility for manipulation. 17 So, I also think that this can be, you know, made 18 electronic and have as many participants as possible who are 19 interested in participating in the process to have access to 20 it. That should be the goal of this process and to move as 21 much of the market onto that type of model as possible. 22 MR. COOK: Okay. Thanks. I want to make sure we 23 have time for our other topic we want to get to on this 24 panel, then we can circle back in the end, which is the 25 exception to the mandatory trading requirement. Why 53 1 don't we start with a line of questioning on that. 2 MS. ADRIANCE: We've just been talking about what 3 is this -- the model that have -- the different versions of 4 the models that have the pre-trade price transparency, et 5 cetera. We know that there is language in the Dodd-Frank Act 6 that refers to, for instance, block trades. There is a 7 possibility of some other language that might allow other 8 situations, which might not be -- which might be trades done 9 without a pre-trade price transparency. And we're trying to 10 understand what that means. 11 What -- your views of under the Dodd-Frank, is 12 there something in addition to block trades that should 13 have -- that can be a method that trades can be executed 14 without requiring pre-trade price transparency and, you know, 15 what might those methods be and as well as under what 16 circumstances would it be appropriate to allow, you know, as 17 regulators we're supposed to be sorting out well, when and 18 under which circumstances are block trades allowed? Can they 19 be accepted as just block trades? Is it an adjunct to these 20 other markets that have the, as it was discussed, central 21 limit order books, RFQs or these other models in between. 22 Where does this fit in, these exceptions to this rule of 23 pre-trade price transparency. 24 MR. DE LEON: I would like to -- sorry. You know, 25 block trading and post-trade transparency is something I 54 1 think that offers very different information than pre-trade. 2 And there are different social aspects to it, in terms of 3 information, information flow and who benefits from that 4 information. And I think it's important to separate those 5 out. 6 What I think, and I think that most people agree 7 on, pre-trade transparency is incredibly important to make 8 sure that as many people have access to (a) get the best 9 price information out there, (b) be allowed to trade and 10 offer liquidity, and (c) when they transact, get that best 11 price. That's a common good, I think, because you want 12 people to trade and get the best levels and not be forced to 13 pay a substantially higher bid ask price. 14 What happens post-trade, however, is what 15 information gets given out and how it is given out and who 16 gets it. Post-trade is now a transaction barring a central 17 limit order book where there is a trade done where that size 18 then gets done. But what information gets given out 19 afterwards and at what rate. 20 You very quickly run into a situation where if you 21 don't have a central limit order book where it's clear that 22 there were willing buyers and sellers to offset each other, 23 and that was the clearing price, where the market starts to 24 move. If I want to buy a lot of something and the central 25 order limit book isn't deep enough, I'll move the price up. 55 1 If I want to buy something, the central limit order book is 2 big enough, actually it can trade through me. 3 When you do something, a block trade, which is now 4 off central order and it's occurred, what information do you 5 give out and when and why. I think it's really important 6 because you create the situation of the buyer's curse and 7 the, you know, or the winner's curse. And you need to 8 prevent that because that's not a social good. Because that 9 means that the two people transacting are taking extra risk 10 on both sides and everyone else who doesn't take part in that 11 transaction gets a lot of free information that they can then 12 use against both people transacting. 13 For example, if I want to buy a lot of something 14 and I call someone and I ask, well, where can I buy it. 15 Well, if the typical order size is one car and I want to buy 16 five thousand cars, well, the price for that is going to be 17 very different. You have to produce 5,000 cars, where are 18 they available. There is a uniqueness factor for it. And 19 if the person who sells me them doesn't have the inventory 20 and needs to work out of it, everyone is going to know he is 21 short them so they can run the price up on that person. 22 So there is a real disincentive for that 23 information to get out. What is important is that that trade 24 did occur, people should know a large trade occurred, but how 25 large and exactly what details should be protected because 56 1 otherwise, everyone else in the room, or everyone else in the 2 market, will have an unfair advantage. So the person who 3 sold the car is going to want to protect themselves and 4 charge a higher price. The person who buys them will pay a 5 higher price. And conversely if I wanted to buy more cars 6 after that first block, it's going to be more difficult. So 7 I think when we talk about block trading, we have to 8 understand the social implications and who is benefitting, 9 versus who is getting hurt and what is the information. 10 MR. COOK: I would like to steer us away, maybe -- 11 thanks for those comments -- from yesterday's topic of 12 transparency and reporting and maybe back a little bit to the 13 mandatory clearing requirement. Because what the statute 14 says is that if a swap is subject to the clearing 15 requirement, then it must be traded on an exchange or SEF 16 unless no SEF or exchange makes the swap available for 17 trading. 18 So, I think what would be helpful is to get your 19 thoughts on how we should be interpreting the words 20 "available for trading" when we're defining the scope of this 21 exception from the mandatory trading requirement. 22 MR. HARDING: Julian Harding. Robert, this might 23 point to the big differences between an exchange or DCM 24 situation and a SEF situation. It is -- whilst it is 25 possible to conceive that a newly clearable swap may be 57 1 turned down because of structural concerns of some kind or 2 something else within an exchange environment, as being able 3 to be listed for whatever reason, it is similarly almost 4 inconceivable that a SEF that one might imagine with a 5 different sort of structure, would ever turn down the chance 6 to operate a market to trade a newly cleared swap. 7 So it doesn't point to the differences. And it's a 8 common thread in the discussions over many, many months now 9 that the word "listing" often comes up in terms of SEFs as 10 well and if SEFs reflect the marketplaces that I believe they 11 do, the listing concept is not correct and, in fact, the 12 SEF will quickly and effectively create a marketplace or 13 operate a marketplace for that newly created swap. So the 14 exception, to me, never made that much sense. 15 MR. COOK: So you would interpret it very narrowly 16 and so that if it's available to be traded, then the 17 exception would not be -- would not apply. 18 MR. HARDING: Yes. For the SEF? Yes. 19 MR. COOK: Yeah. Okay. 20 MR. DuFOUR: Richard DuFour. I think the real key 21 isn't going to be at the SEF level, but at the clearing 22 level, that if a clearing -- there is a clearing entity out 23 there that believes they can clear the contract with whatever 24 the specifications are, then a SEF will be happy to act as 25 the, you know, the place to match buyer and seller. 58 1 MR. SPRECHER: This is Jeff Sprecher from ICE. 2 Just to follow up on that comment. As probably one of the 3 largest, if not the largest, operator of the OTC 4 clearinghouses, the reason that we've become successful in 5 doing that is that we've made arrangements with the industry 6 to give us price transparency because the remedy in a 7 clearinghouse on a default is to liquidate the position. And 8 so by default, we need to know at all times where the market 9 price is, which means we need price transparency. So 10 clearing will follow markets where there is price 11 transparency. Not the opposite. 12 And also, let me just expand my comments to say, 13 again, as an operator of futures exchanges, we recognize 14 there are contracts that are large size that are parts of a 15 tailored risk profiles and the like that are arranged off 16 exchange and are given to us as blocks. We try to maintain 17 that at a reasonable amount of quantity, usually in our 18 futures exchanges, it's under 10 percent of our volume, and 19 we try to influence that by the pricing that we charge to 20 accept blocks, the rules that we put in place to force 21 advertising of trades, and the like. 22 In ICE's OTC energy markets, which are probably the 23 most liquid two-way bid offer, OTC markets that exist today, 24 97 percent of all the trades that go across our platform are 25 cleared and about 15 percent of the volume that comes into 59 1 our OTC market or 15 percent of the number of trades that 2 come in are done away. Those are typically, again, large 3 size, customized deals, and so on and so forth. So I think 4 that the market can develop around a price transparent market 5 and still accommodate some customization for both clearing 6 and trading. 7 MR. DOWNES: Andrew Downes. With respect to the 8 available to trade, meaning I think that has to have, 9 implicit in it the level of liquidity that's necessary for 10 trading actually to occur; i.e., that someone can come in and 11 there is trading available. I don't think it's a matter of 12 build it and it's mandatory to come. I think it's more build 13 it. They may come. Then if they come, it should be 14 mandatory to come. That's the way I would understand the 15 legislation. 16 And I think, you know, if you're looking at 17 liquidity, which is back to the sort of plurality of models 18 to be encompassed in the SEF definition, as I said and people 19 have said earlier, there will be a large range of liquidity. 20 So I think you need to think about, in each case, what is 21 available to trade based on, you know, the frequency of 22 trading of the instrument. 23 You know, in some cases, if there's only X amount 24 of, say, 10 trades a day, that's not really going to lend 25 itself to a central order book because there just won't be 60 1 bids and offers seeking each other in order to clear. So I 2 would say liquidity is absolutely key. 3 MR. MacDONALD: I think -- Ben MacDonald from 4 Bloomberg. One of the -- just thinking about the debate a 5 little bit, one of the models which does exist today is the 6 single dealer franchises that we spoke about which allow people 7 to look in multiple offerings simultaneously. The beauty of 8 that model is it actually allows the market to start becoming 9 more electronic and people to post new products out there. 10 And I think, you know, just kind of I guess thinking out 11 loud, the question is whether, you know, that -- if you have 12 that kind of forum, what happens, you eventually reach a 13 critical mass in terms of liquidity which then allows you to 14 kind of, you know, hit that point. 15 So, I think what you're encouraging, by having that 16 kind of model, is people to post liquidity and it gives you a 17 mechanism to understand at what point those products actually 18 do become liquid and then should, you know, I make a decision 19 as to whether I should -- whether you belong, you know, in a 20 clearing house. 21 MR. DENIZE: Again, just a note. This is Yves 22 Denize. A note to encourage a process in place that provides 23 us with some subtle expectations as to how that would 24 transition into a mandatory trading environment. You know, 25 simply having a SEF raise its hand and say I'm prepared to 61 1 list or prepared to provide a trading venue may not be the 2 right process to have a trade go into mandatory trading. And 3 I think there was some discussion about that over the course 4 of the legislation when we talked about mandatory clearing 5 and how the agencies would be involved and having something 6 be designated for mandatory clearing. A similar process 7 would be at least logical here when you talk about mandatory 8 trading and whether you're looking at liquidity volumes, et 9 cetera, would be welcome. 10 MR. MacDONALD: I think -- actually which raises a 11 very, very good point. I think one of the things which, you 12 know, which we have to think about, and I know this is part 13 of the second discussion is, you know, the mechanisms by 14 which, you know, you can ensure that all of the SEFs kind 15 of, you know, have the same rules and there is clear 16 understanding as to, you know, what belongs in which forum, 17 which, you know, does raise the question as to, you know, to 18 what extent should there be some form of central utility or 19 something like that which is responsible for some of these 20 activities around defining what's available and, you know, 21 where the products should be going. 22 MS. SLAVKIN: I think this language should be read 23 relatively narrowly and that the requirement should flow from 24 the clearing requirement. I think what we're seeing here is 25 a preference that was reflected throughout the Dodd-Frank Act 62 1 on the part of Congress to avoid mandating private market 2 actors to do pretty much anything, but allow the regulators 3 to make the determinations as necessary. And so I think it 4 would be a mistake to put too much emphasis on this language. 5 MR. VISWANATHAN: I would also kind of say you 6 probably want a narrow interpretation simply because it could 7 be that, you know, the 10 year T-bond is traded and I'm 8 customizing the 9 1/2 year, the 9 1/2 year may not be traded, 9 but for all practical purposes, you know, I'm pricing it off 10 something that's out there. 11 So if you're going to make exemptions, you're kind 12 of worried that if you make too many exemptions, people will 13 use the exemptions to -- the 9 1/2 year's exempt and maybe I 14 don't want to disclose. I simply trade the 9 1/2 year 15 instead of the 10 year. So you have to worry a little about 16 in the rule-making process, you start inducing regulatory 17 arbitrage. So my view is that exceptions should be narrow 18 and the rules should be very clear on when transitions occur. 19 MR. DOWNES: Andrew Downes. I would just say in 20 terms of the requirement to clear compared to the, you know, 21 the requirement to SEF trade, I think when you look at 22 available to trade and take into account liquidity, the 23 liquidity measure or standard is different from that which I 24 would say is relevant for clearing. If you're looking at 25 clearing, what you want is enough price transparency in 63 1 order to have safety for the clearer to be able to calculate 2 the exposures and get the margins. 3 And you can get that with respect to trades, which 4 actually aren't trading on a daily basis in the market 5 because people clearing members will have those trades on 6 their books and they will be marking them on a daily basis. 7 And the risk that you take where people aren't seeing those 8 markets is mitigated to the extent that you can require 9 additional margin and additional safety factors. 10 If you distinguish that from liquidity in the 11 context of trading, it may be that people have trades on 12 their books, as I mentioned, for clearing, but there's no 13 trading going on and no bids and offers to find each other on 14 a daily basis. So I think you need to look at liquidity in 15 different ways for each of the two exercises, clearing and 16 execution. 17 MR. MELARA: If I could follow up on liquidity 18 measures, we heard frequency of trading referenced. Could the 19 panel expand, to the extent they can, as to what other 20 factors we should be looking at when considering this 21 particular issue. 22 MR. DOWNES: I would say other factors that would 23 be relevant, aside from frequency of trading, is the amount 24 of market makers or people that are involved in the market. 25 I think if you've got, you know, only a couple of people that 64 1 are making markets, that's clearly less of the market than if 2 you've got, you know, 16 market makers all participating on a 3 daily basis. So I think that's another key aspect. 4 I think you also need in different, depending on 5 the product, you need to look across the curves, sort of 6 tenor points and find out, you know, what proportion of the 7 contract is trading at which points because there will be 8 some names that we would regard as relatively liquid, but 9 they'll only be liquid, say, at one point, as opposed to the 10 other points. So I think those are some of the key things 11 that people need to consider. 12 MR. OLESKY: I would agree, Andrew. I think in 13 addition to the velocity of trading and the frequency of 14 trading, it would really be the breadth of the market, is 15 critical component here, how many participants are there 16 willing to take on that risk. If you're talking about a 17 principal market, it's almost as important as the velocity 18 and the frequency with which things trade because that really 19 does establish. If you only have two folks making those markets, 20 I would say that's a little bit less liquid instrument. 21 MR. DuFOUR: I think there is -- Richard DuFour. I 22 think there is a couple of other measures you can also look 23 at. Some of it would just be in a -- well, if you had 24 something that was traded in an order book, you know, it's 25 kind of the depth of the book; if it's a request for quotes, 65 1 you would look at, you know, how many people had responded 2 and the sizes for the responses. And you also can look at 3 the open interest in the particular series or contract that's 4 been created. Typically greater open interest will indicate, 5 you know, greater liquidity. 6 MR. HARDING: Julian Harding. I think also 7 following from Andrew's point that certainly the number of 8 participants is a crucial measure. But broadly, there are -- 9 there is a -- there are two forms of liquidity, retail 10 liquidity which could be characterized by an equity 11 marketplace where there are 100 pieces of 100 lots on each 12 side of the bid offer spread. And then there is 13 institutional liquidity where that same bid offer spread may 14 be populated by one market maker in an average size. But 15 right behind that, maybe one tick away on both sides, there 16 is a vast amount of size that can be transacted. And that 17 sort of liquidity, I think, is a very important element to 18 preserve in any debates we have about liquidity. 19 MR. OLESKY: Arguably, a definition of liquidity is 20 can you do -- how much size can you do relative to moving the 21 market is one definition. If you can do a lot of size, 22 there's a lot of liquidity. If you can't do a lot of size, 23 there is not so much liquidity. 24 MR. VISWANATHAN: I would agree with that. I think 25 in the end, it has to come with measuring your price impact. 66 1 If every trade moves prices a lot, then you have to say the 2 market -- you would have to measure, somehow, the size and 3 the depth of the market of something. 4 MR. COOK: If a product goes through the process of 5 being qualified to clear, so you have a clearing agency who 6 thinks there's enough price transparency or modeling 7 capability around that product field to accept it, how likely 8 is it that product wouldn't be available in -- for trading in 9 some facility that would qualify as a SEF? I mean, how big 10 an issue do you think this will be as a real world manner once 11 it makes it through the -- over the threshold of being clearable 12 then why wouldn't it normally be tradeable in a liquid market? 13 MR. MacDONALD: I think it's unlikely that anything 14 which is accepted by the clearinghouse wouldn't be 15 immediately available on the SEF. I think that, you know, 16 especially if you've got multiple SEFs, there will be a, you 17 know, competitive aspect to being first to market with anything 18 which is made available on the clearinghouse. 19 MR. DE LEON: I'm not sure that you'll see that 20 because as we've seen, there are multiple security -- 21 multiple things that trade or are cleared or are clearing 22 eligible and especially if you look at the CDS market where 23 there is the big bang and small bang, so there's been a lot of 24 standardization already, but a lot of smaller names don't 25 trade that actively. 67 1 So, I would say that it's in the public good to have 2 these CDS cleared and margined and marked on a daily basis, 3 but it's not necessarily clear that they will be trading very 4 frequently. They trade once or twice a week as is. So -- 5 and that's only by a limited number of players and dealers. 6 So given that, I don't foresee a pickup in that dramatically 7 as it is SEF eligible. It will obviously help over time, but 8 there are things that are not that liquidly traded as is. So 9 just by definition, it doesn't mean there will now be a full 10 deep market in it. 11 MS. ADRIANCE: In terms of that, you mentioned 12 that, you know, just because it's determined to be clearable, 13 it does not necessarily mean that it will trade frequently. 14 The question that we began on was what does "makes available 15 for trading" mean. If it's determined clearable, but it's not 16 trading frequently, is that available for trading because 17 it's being offered by some SEFs and so therefore there is 18 mandatory trading or are you suggesting that there is -- it's 19 not being really made available. I'm not sure really what 20 you're saying. 21 MR. DE LEON: I just -- 22 MR. DOWNES: I would say if you look at what's 23 clearing, you know, there's obviously -- say you look at the 24 less frequently traded single names, which should be the 25 constituents of the investment grade index, I think on 68 1 average across all of the constituent names, they trade four 2 times a day. That is not very frequent. So, what I would say 3 is whilst all those names or most of those names will clear and 4 many of them already clear, it would be unlikely, at least 5 for some subset, that they may trade in a SEF depending on 6 the model for the SEF. 7 So if one were to prescribe that a SEF should be a 8 central order book, I don't think those names would trade on a 9 central order book because there's not enough bids and offers 10 seeking each other. But to the extent that the SEF definition 11 is wider and can encompass more models, then there's more 12 likelihood that some names can be picked up and traded in a 13 different model. 14 MR. OLESKY: I wanted to go back to your question, 15 Robert and I think yours as well Riva. One of the things, 16 and Ben made this point, I think if it's clearable, a SEF is 17 going to want to have it on their system. What might stop 18 that from happening is if there is not equal access to that 19 central counterparty and clearing so that it's not economic for 20 that SEF to actually be able to do that business. So this 21 gets back to what might stop an organization from actually 22 making something available. Well, if we can't be competitive 23 in the marketplace, vis-a-vis the customer base, that would 24 be a hindrance. 25 MR. MacDONALD: I think in a funny way -- 69 1 MR. SEMLITZ: I'm sorry. Go ahead. 2 MR. MacDONALD: I was going to say I think it also 3 goes back -- I think there is two issues, there is 4 availability and liquidity, which is ultimately going to 5 drive what's on the screen. I think it does kind of go a 6 little bit back to the RFQ model as well because the relevance 7 of that model is that it allows you to trade illiquid products 8 by sending out a quote rather than actually only being able to 9 trade if that product is available on the screen. So, I think 10 that remains a very important part especially as this market 11 continues to grow. 12 MR. SEMLITZ: You know, it's one thing to say 13 something is available to be traded on a screen, but it's not 14 trading, and if the purpose -- unless you want to exclude end 15 users or people who need semi-customizable products from 16 trading and executing what they need, if you don't 17 differentiate between clearing and available to be traded, 18 you're going to end up constraining trade by not permitting 19 people to execute trades because there are no quotes on a 20 particular SEF. So you're going to have to separate the two. 21 MR. DuFOUR: What I would visualize happening is 22 you would have the relationship between a SEF and a clearing 23 entity and this concept of a product being listed I think 24 isn't quite the right way to think about it, but rather, and 25 I'm thinking of a request for quote model, I would come in 70 1 and request for a quote for a product, you know, 2 specifications, and the listing, if you will, would be 3 created as a function of the trade as long as it was within 4 some parameters that had previously been agreed upon with the 5 clearing corporation and we can clear the following things. 6 You can vary the following, you know, aspects of 7 the contract, then the product would become listed or 8 available because a trade took place and now there would be 9 open interest carried in that particular product and, you 10 know, someone else could come in and request quotes for 11 order. 12 MR. HARDING: Julian Harding. Just picking up on 13 what Lee was just saying, which needs emphasis, I think, 14 again. It is in the statute that there is full and open and 15 non-discriminatory access from a competing SEF to a DCO and 16 it's important, and it's an important proviso. I think it's 17 important to preserve the ability of competing SEFs to have 18 this non-discriminatory access, which has to be complete. 19 There can't be subtle or nuanced or discreet ways to 20 discriminate against the access that SEF has. 21 This point is well taken that there is -- one could 22 possibly imagine a situation where a SEF's desire and ability 23 to organize trading markets in that cleared product, 24 may be in some ways stifled or stultified by a 25 disadvantageous access to that DCO. 71 1 MR. SPRECHER: This is Jeff Sprecher from ICE. I 2 just want to make the point again, that again as the operator 3 of a leading CDS clearinghouse, one of the most complicated 4 derivatives, we can't clear a product unless there is a 5 liquid market. And not only does there have to be a liquid 6 market, we need pre-trade price transparency, which means it 7 will be very difficult for market operators like us to clear 8 contracts that simply exist through a request for quote, for 9 example. 10 The reason I say that is that if a market is an 11 illiquid market and let's say it's 20 bid at 50, which 12 means there is somebody willing to buy at 20 and sell at 50, 13 if somebody trades and does a trade at 50 and we receive it 14 at the clearinghouse, we can't mark our positions to market 15 at 50. If we have to sell, we're going to have to sell at 16 20. 17 So, we need to know the bid-offer spread and we 18 need -- and then in other words, we need the pre-trade price 19 transparency in order to properly mark to market unless we 20 want to margin somebody at a hundred percent, which is the 21 only other -- in other words, everybody prepays for all their 22 business, which I think probably is self-defeating in the 23 marketplace. 24 As Julian mentioned, the single name CDS market is 25 relatively illiquid as it has existed; however, the reason 72 1 that we're able to clear that, is we run a separate market 2 with market participants where we run a daily auction to give 3 us pre-trade and post-trade price transparency. And it's 4 only because we have found market participants willing to 5 accept that daily risk and with products that can trade on a 6 daily basis that we're able to clear the portfolio that we 7 clear right now. 8 MR. SCHOTT: I would like to follow up on that and 9 also bring it back to a question that I think Tom has asked 10 some time ago emphasizing that pre-trade price transparency 11 was one of the goals we were trying to achieve here. When 12 the different answers that came across this question talked 13 about different potential SEF models, but I'm not sure we 14 squarely addressed pre-trade price transparency. We talked 15 about central limit order books and RFQs and so forth. 16 Is there any reason why, regardless of the model or 17 models that the Commissions may adopt, why the models could 18 not include pre-trade transparency as part of how the 19 SEF operates, so that even if it's an RFQ, you can see all 20 responses, all participants can see the responses to the 21 quotes. 22 MR. DE LEON: I think, and I'll let some of my 23 other colleagues who actually run sort of flex type models is 24 that when you do put an RFQ out to these things, when it 25 comes back, it's sort of public information and everyone has 73 1 access to see that information. Whether or not you can trade 2 on it is a function of whether or not you're a member or 3 you're clearing through a member to that exchange. And 4 that's a different discussion. But assuming one of those two 5 things is the case, you would see that RFQ and you could 6 then go back and either transact on it or counter back with 7 another level. Please correct me if I missed something. 8 MR. DuFOUR: That's correct. 9 MR. OLESKY: I would just say a slight variation. 10 I think one of the RFQ models that we run has actually -- 11 does not have a feature where the whole market sees the 12 inquiry that's coming in from a buy side customer. And the 13 reason for that is to protect that buy side customer from -- 14 and give the confidence to those providing liquidity that 15 they can actually do the transaction before it becomes 16 public, and potentially, that information could be used 17 against each of those market participants. That's kind of 18 how the voice market has functioned for years. 19 And the risks there, in terms of the tradeoffs, are 20 transparency, immediate transparency for the whole world 21 versus the liquidity and price formation. And if you make it 22 immediately obvious to the whole world that someone wants to 23 do a trade for a billion five-year U.S. treasuries, it's 24 going to be very hard for any liquidity provider to step up 25 and say I'm going to provide you with that liquidity because 74 1 they're going to be concerned about how they're going to 2 hedge that risk and get out of that risk. 3 And in a principal market like for example just to 4 use the U.S. treasury market, in the U.S. treasury market, it 5 would be very difficult for primary dealers to extend the 6 kind of liquidity they do instantaneously over TradeWeb and 7 other platforms if they didn't have that protected section 8 when the RFQ is going on. And I think what would be the 9 result is if you open that up so that everyone can see it, 10 you're going to have less willing participants in terms of 11 opening up there and taking on that risk that they're 12 providing to a customer when they're making a market, for 13 example, in the U.S. treasury market. 14 MS. ADRIANCE: And just to follow up on that, what 15 you've just mentioned was, in a sense, a large block trade 16 that another party may not want to take on the risk of a 17 large block if it's a fully transparent marketplace. But 18 what you've just described is a large block, which is clearly 19 mentioned in the Act, under Dodd-Frank. And in that sense, 20 there is, like for instance, to use the model that we have in 21 the futures world, you can have a centralized marketplace and 22 you can still have exceptions to the rule; for instance for 23 blocks or U fees or whatever. 24 I guess part of our question here is, is there 25 something -- while there may be situations that are 75 1 appropriate that there is a block trade with a large notional 2 amount, or whatever, what should be the rule, the basic 3 marketplace, what variations of models. I think Sebastian 4 was asking if the variations of models that are appropriate 5 for the overall marketplace, can all those different flavors 6 of models somehow offer pre-trade price transparency as 7 separate from those situations, those -- the exceptions to 8 the rule, the block trades, that wouldn't offer that. 9 So if we can get any thoughts as to back to, 10 in a sense, the basic rule, what is this -- the usual model, 11 whether it's central limit order book, RFQ's, streaming 12 quotes, these other versions that we've been talking about, 13 can all of those be -- either have now or be adapted to 14 provide pre-trade price transparency for that basic 15 marketplace. And we'll view that as the -- as not the 16 exceptions, but the basic model. 17 MR. DE LEON: I think you're sort of alluding to 18 something we already see in the market now, which is the 19 concept of people do pre-trade transparency either because 20 things are electronically available or they pick up the phone 21 and call and ask five people where do you see the following 22 or show me a two-way on the following to get information 23 back. And then once they've done that, when they want to do 24 a block trade, they'll pick one or maybe two people and then 25 show it to them on the block size because as -- not to delve 76 1 into the other thing, I apologize for that earlier, but when 2 you do do a block trade, obviously you want to be careful 3 about what information you show and how. 4 So, to the extent that you can do an upstairs block 5 trade, but you've gone through some of the other methods to 6 get pre-trade information, either through RFQs, looking at 7 central order book, et cetera, streaming levels, you have an 8 idea of where to trade and what to trade and then you can do 9 your block trade. 10 And another thing to sort of incorporate, depending 11 on how these SEFs work, there are some issues, in terms of 12 going to the treasury model, just to dwell on it -- there are 13 compliance and other legal issues because some accounts can't 14 trade with certain people. And it's not a credit quality thing, 15 it's sort of a structural thing. So, there are certain rules, 16 and especially in Tradeweb, where you may not choose to show 17 prices to a specific set of dealers because you're 18 legally not allowed to trade with them due to compliance or 19 legal or client guidelines. So that's sort of just an 20 adjunct. 21 But I do think that there are plenty of ways to do 22 pre-trade transparency that would work in the sort of 23 SEF model and then post that, you would then do either an RFQ 24 or you would do an upstairs block trade where you don't 25 necessarily show out to the universe what you want to do and 77 1 advertise it up-front. 2 MR. COOK: Thanks. So we're coming to the end of 3 our time. I just want to give anyone that wants to get in on 4 this question a final opportunity to do so, keeping comments 5 very brief. And let me ask Steve, do you have anything you 6 want to add before we wrap up? 7 MR. SEMLITZ: No, nothing here. 8 MR. COOK: Okay. Thanks. Anyone else? 9 MR. OLESKY: I just wanted to reiterate a point I 10 made earlier, which is I think as we -- this is such a big 11 change. If you move derivatives markets into a 12 SEF environment, which is largely electronic, I think one of 13 the outcomes is you're going to have more participation, 14 you're going to have more prices coming in, and you're going 15 to have more pre-trade price transparency. And it's hard to 16 talk about each of these bits of the legislation in piecemeal 17 because I think they are all interrelated. The block rules 18 are tied to the RFQ, which is tied to liquidity issues, which 19 is tied to pre-trade price transparency. They all kind of 20 need to be viewed, I think, holistically. 21 MR. HARDING: Julian Harding. I would just like to 22 almost endorse what Lee has said before, that the new SEF 23 environment should be required of it a level of transparency 24 that does not hamper liquidity. That for me should be the 25 statement we should make, that liquidity maintenance or 78 1 preservation is of paramount importance and we should 2 offer -- we should construct marketplaces that insist upon 3 the level of transparency -- the highest level of 4 transparency that does not hamper liquidity. 5 MR. DuFOUR: Richard DuFour. I would add to -- I 6 would agree with both of them, and I would add to that the 7 importance of -- it came up earlier, the issue of a firm quote, 8 that if I'm going to be able to trade anonymously, in these 9 systems and I put in a quote, I have to be held to it, you 10 know, for a period of time. 11 MR. COOK: Very good. Well, thank you very much 12 for your participation on this panel. It's been very helpful 13 and we appreciate your time and your contributions. We will 14 be taking a 15 minute break and we'll reconvene at 11:00 for 15 our second panel. Thank you. 16 (A brief recess was taken.) 17 MR. COOK: Well, welcome back to our second panel 18 of the day on SEFs. This panel will focus on the compliance 19 with core principles for SEFs. There are four key subtopics 20 that we would hope to touch on, one is block trades, the second 21 is surveillance, investigation and enforcement of SEF rules, 22 the third is cross-market issues, and the fourth is the 23 obligation of SEFs to provide impartial access. 24 So the format for the panel will be the same as the 25 last, but why don't we again start by asking if folks could 79 1 please kind of go down the line and say your name and your 2 affiliation please. 3 MR. VISWANATHAN: Vish Viswanathan, Duke 4 University. 5 MR. YELVINGTON: Brian Yelvington, Knight Capital 6 Group. 7 MR. WEISBERG: Philip Weisberg, FXall. 8 MR. McVEY: Rick McVey, Market Axess. 9 MR. KNIGHT: Ed Knight, NASDAQ. 10 MR. DURKIN: Bryan Durkin, COO, CME Group. 11 MR. DIPLAS: Athanassios Diplas, Deutsche Bank. 12 MR. DENIZE: Yves Denize, TIAA-CREF. 13 MR. DE LEON: Bill De Leon, PIMCO. 14 MR. HARDING: Julian Harding, Tradition, 15 representing the Wholesale Market Brokers' Association. 16 MR. COOK: And we may be joined by Michael Masters 17 momentarily. 18 So again, the format will be the same. We'll start 19 with the staff asking questions and anyone is free to jump 20 in, and we'll ask again that you'll just bear in mind the 21 number of panelists, the interest in the topics and the short 22 time that we have to get through this material. So with 23 that, let's start with the first question. 24 MR. MELARA: Thank you. The first topic is block 25 trades. And as Director Cook indicated, this regards compliance 80 1 with core principles. So, with respect to the core principles, 2 Section 733 of Dodd-Frank, core principle 2 reads that, it 3 says that, "A swap execution facility shall establish rules 4 governing the operation of the facility, including rules 5 specifying trading procedures to be used in entering and 6 executing orders traded or posted on the facility, including 7 block trades." 8 Now the first panel spoke at length and on 9 various -- in answers to various questions about block 10 trades. So I would like to get a sense from this new panel 11 as to your views regarding block trades and how they impact 12 your various businesses or your perspectives, the perspectives 13 that you represent here today. 14 MR. McVEY: Rick McVey with Market Axess. Happy 15 to jump in and start there. We run an electronic trading 16 network primarily active in institutional credit markets 17 today. We primarily utilize an RFQ protocol, although there 18 are a variety of different trading protocols available on the 19 system, and we compete with many entities that have 20 alternative models. 21 With respect to block trades, what we have found in 22 our 10 year history is that the larger the trade size 23 becomes, the more likely it is that both the buyer and seller 24 have interest in bilateral transactions or in narrowing the 25 audience for an RFQ or an auction process. In our opinion, 81 1 those trades can take place electronically in an RFQ model 2 through anything from a bilateral transaction to an inquiry 3 they might want to share only with two or three key 4 counterparties at a time. 5 So in our opinion, block trades can be accommodated 6 electronically and meet the market objectives of sourcing 7 liquidity without exposing an excess amount of market risk 8 that may lead to front running in the marketplace. We also 9 are fans of the compromises that have been made around price 10 reporting for block trades with TRACE, wherein corporate 11 bonds, there are size thresholds above which a trade is 12 reported as only having taken place above that size 13 threshold. 14 So for instance, for high grade corporate bonds, 15 the transaction is reported as 5 million plus irrespective of 16 the block trading size. In high yield, which is an even less 17 liquid market, TRACE reports at 1 million plus. I think 18 those are sensible compromises that have been made between 19 the regulators and the industry that could be applied 20 successfully in the OTC derivative space. 21 MR. MELARA: Thank you. Anyone else? 22 MR. DE LEON: Yeah, hi. Bill De Leon. We tend to 23 agree with that view, that it is incredibly important that 24 while the public has information to be achieved that you want 25 the trade to occur on a SEF, it is important that you 82 1 bifurcate sort of what information goes out there. Because 2 as the trade size increases or it becomes customized, it 3 tends to be more bilateral in nature and you want to limit 4 what information is given out. 5 And we think the TRACE-style compromise of treating 6 it as above a certain amount accomplishes that. It lets 7 people know a block trade occurred. It doesn't tell them 8 exactly how much, which insulates the players, and it allows 9 the bilateral conversation to occur in a way such that both 10 the buyer and the seller are not giving away too much 11 information, which leads to the free rider problem and sort 12 of the loser's curse. So, I do echo those points and I think 13 they were said more eloquently than I just did there. 14 MR. DURKIN: Bryan Durkin from the CME Group. 15 Clearly, as part of our model, block trades has a relevance 16 and has a place in the markets and the determination of 17 what's appropriate, in the context of thresholds, one needs 18 to take into consideration liquidity of the product and the 19 platform in which the product is available. 20 And then, you know, lastly, you know, I think in 21 considering thresholds, one has to consider equivalent 22 products or equivalent markets. And so if there is an 23 equivalent product or market to benchmark off of, those 24 thresholds need to be taken into consideration so as not to 25 take away from the transparency or the liquidity of a 83 1 centralized market in which block thresholds have been 2 established for some period of time. 3 MR. DIPLAS: Hi. Athanassios Diplas. I would 4 agree with some of the comments that Bill De Leon over here 5 made. One of the primary objectives or policy objectives 6 obviously has been to ensure that the clients get to transact 7 their desired size in the best possible price. And 8 transparency obviously is one of the means in achieving that 9 goal, but obviously there is a limit to how far we can take 10 that when the transaction size gets large. And that is why I 11 think it is these combination of places you have a lot of 12 transparency for the smaller type transactions combined with 13 the ability to share the protected transacting parties for 14 larger size transactions is a very smart way to go about it. 15 As people have said before, we have seen examples 16 of how TRACE can set thresholds. Also we have seen now in 17 the current legislation and proposals how they have also set 18 thresholds that take into account the liquidity aspects of 19 the underlying product. In doing that analysis, obviously, 20 we need to -- I think we need to base it on facts, rather 21 than obviously our own beliefs, and all of us come from 22 different angles, so the beliefs are going to be multiple. 23 And in that respect, you know, I urge you to look, obviously, 24 at the data in the underlying markets. 25 What we have done, as part of market participants, 84 1 the different dealers along with the large buy side clients, 2 as part of the commitments to the international regulators, 3 we have done a study on transparency and we have delivered 4 now a three month slice of data for credit rates and equities, 5 along with credit rates quotes that are associated with those 6 instruments, and that is, I think, an extremely valuable 7 piece of information that's going to allow the supervisors to 8 make those determinations in terms of what constitutes a 9 large market size moving transaction, based on actual data. 10 What you will see there, obviously, is that the market, the 11 OTC market, is much more diverse than something like a futures 12 market, which tends to focus on very high concentrated 13 widgets. 14 The breadth of the market is managed here and I 15 think it is so for a reason in that it tries to accommodate 16 client needs. And you will see that basically a lot of 17 transactions happen in the benchmarks and those would be then 18 accommodated very easily in the SEF, in the standard 19 SEF example, but some of the others that are either smaller, 20 still clearable, because they have been admitted to a 21 clearinghouse because either they were older or they have 22 naturally aged, but they don't naturally trade, those need to 23 be subject to the block trading requirement. 24 MR. WEISBERG: Phil Weisberg from FXall. When we 25 were originally starting our electronic market, we had a lot 85 1 of discussions with market participants about how we would 2 have to set it up to enable them to do most, if not all, of 3 their trading on our platform and what was important to the 4 end users, that we have transacting, the market participants 5 that are very diverse, was choices. So the ability to 6 transfer the whole risk in a block in one immediate activity 7 or the ability, if it was available, to take more execution 8 risk and break that, you know, trade up over time. 9 So, they initially asked us, 'can you make sure that 10 your trade protocol has the ability to let a market maker 11 know if I'm asking just them or I'm asking everybody,' because 12 they felt it would impact the price that they would, you 13 know, receive. So, we just would encourage the regulators to 14 draft the regulations in a way that would maximize the amount 15 of trades that could occur on a SEF, allow blocks to happen 16 on a SEF if possible and with respect to the reporting of 17 block trade information, we would agree with Mr. McVey's 18 comments that TRACE-like distribution mechanisms, where 19 people were informed of trades and slightly larger trades 20 have a little bit of a delay and are reported as large trades 21 as opposed to the exact size would be the best compromise to 22 achieve the objectives. 23 MR. DENIZE: Just a quick note -- Yves Denize -- 24 because of our discussion in the last panel about how 25 different types of transactions will have vastly different 86 1 liquidities, one size limit probably won't fit all. So, I 2 think the example used was 5 million plus, for instance, but 3 clearly as you look at different types of trades, there will 4 be different types of limits applicable for the block trade. 5 MR. MELARA: If I may follow up along the same line 6 of questioning as the liquidity measures question in the last 7 panel, if there are these factors that you would like to share 8 with respect to looking at block trades, depending on the 9 asset class and/or the market that trades them, I think that 10 would be useful. 11 MR. DE LEON: I think it's very important to, as 12 Yves pointed out and was initially brought up, different 13 asset classes and products will have different thresholds for 14 what's considered a block trade. So, if you take the generic 15 interest rate market or treasury market, clearly trading a 16 hundred million dollars of long bonds is a very different 17 trade than trading a hundred million dollars of two year 18 treasury notes. The duration and market impact there is very 19 different between those two, even though they're both about 20 the same notional and the same market value. And I'm just 21 using that as an example. I'm not giving guidance on where I 22 think the threshold should be. 23 So, I think as you go through the market, you need 24 to take into account not only the size of the trade, but the 25 type of market and the market risk that's associated with 87 1 it, so as well as what are the market conditions at the time. 2 So, for example, doing a hundred million dollar trade in U.S. 3 long government bonds at 10 o'clock eastern time is a very 4 different liquidity stress, and I would say it's relatively 5 low, than if I were to try to do that in Asia time or if I 6 were to try to do that after an economic event occurred, you 7 know, non-farm payroll comes out at 8:30 New York time, but 8 try to do a large trade at 8:31, what's considered a block 9 trade would be different. 10 So, I think you need to scale market conditions, as 11 well as products, into that factor, so to get a better feeling 12 because it's definitively not a one-stop fits all type 13 approach in terms of what a block trade is. 14 MS. ADRIANCE: It sounds like what you're 15 suggesting is that rather than a number be determined to be an 16 appropriate threshold, that you're talking about some process 17 or some algorithm or some calculation through which we, or 18 somebody, should be going through to determine what's the 19 appropriate size. Is that correct? 20 MR. DE LEON: Yes, that's what I was implying. 21 Some concept of risk is best -- and I apologize for 22 being technical, but the risk associated with something in 23 terms of -- I think this came out in one of the earlier 24 panels. What is the liquidity cost of doing a large trade 25 might be your standard. So, for something that, you know, a 88 1 $5 million trade has the ability to move the market X amount 2 where in another market, you can do a $500 million trade 3 without moving the market -- with moving the market the same 4 amount. That would get you a similar sort of concept there 5 where liquidity and price discovery equate what a block is. 6 MR. DIPLAS: I think to get it started along those 7 lines, I mean, you can see right now in the market if you 8 look at the pre-trade quotes that are available -- and there 9 is a big database of those that we have accumulated -- you 10 can see, for example, what the market participants and dealers 11 in particular are willing to quote as a standard transaction. 12 So, you're looking at an on-the-run index someone is willing 13 to quote, you know, 200 million two ways and are sending -- 14 blasting that to everyone. Clearly that someone doesn't 15 think that as a market moving transaction are willing to send 16 that out to everyone. And that would not constitute the 17 block. 18 I think if you go three times that amount, it 19 definitely -- it will have a market moving impact. So, to 20 the extent that you have that space to get all those numbers 21 out, the same thing for single names, a low beta name, 22 meaning a name that doesn't move that much, is often quoted, 23 you know, at 20 million two ways, but the high beta name is 24 quoted at 10 million this way. You go to high yield, the 25 number drops a lot of times to two by two. So, that is - at 89 1 least gives you the basis from which to start because I think 2 obviously you turn to an algorithm, such a huge number of 3 names is going to be a very impossible task for you. So, at 4 least you need to start with something simpler. 5 MR. EADY: Athanassios, when you said this in a 6 database somewhere, I mean, what are you talking about? 7 MR. DIPLAS: Yeah. The numbers -- the way the 8 market is communicating this level right now is actually by 9 blasting out Bloomberg messages to everybody. So, clients 10 actually have the problem of getting way too much 11 information. Vendors have stepped in and actually have taken 12 all the information and started sharing it in a way that 13 actually becomes useable for market participants and they -- 14 given the fact that they have stock in a best bid offer. 15 Now these vendors have kept all that information 16 and as part of the study I mentioned earlier, we gave a 17 three-month slice of that information to regulators so they can 18 actually see how -- what kind of quotes were actually given 19 out and how the market was trading. At the same time, you 20 can see where actual trades occurred. So, you see if the -- 21 how well correlated basically the trade was to the actual 22 quote. So, if the trade was done at five times the size, you 23 would see a difference. 24 MR. YELVINGTON: Additionally, just to kind of echo 25 the comments that Bill made earlier, you know, you have to 90 1 take into account not only the time at which the transactions 2 occurred, acknowledged that for various instruments that may 3 be a little bit more granular, I was thinking here possibly 4 in the area of credit derivatives, it changes through time. 5 A particular event such as, you know, a corporate disaster of 6 some sort may increase the trading activity of particular 7 name in such a rapid fashion that what would have been a 8 block trade before is no longer a block trade. And that can 9 happen in a matter of a day. 10 MR. McVEY: Go ahead. 11 MR. VISWANATHAN: The only thing I want to caution 12 is it's hard for me to imagine regulators running through a 13 complicated process every day trying to figure out what the 14 depth of every market is to determine what the threshold is. 15 It doesn't seem to be -- I mean, clearly, there has to be 16 some mechanism to take the volume, the price impact, perhaps, 17 over the last six months. But beyond that, perhaps another 18 approach is simply to say all trades are disclosed. But if 19 you're over this threshold, because it's impulse like, you get 20 a delay of one day or something and leave it at that. 21 So, all trades are eventually disclosed. Some are 22 disclosed with a threshold. You know, for this market, it's 23 1 percent of volume. So, a hundred, you know, a hundred 24 million, or whatever it is. But if a day later you're told, 25 without being told exactly that it was that trade, they had a 91 1 trade of, you know, maybe 300 million occurred at that price. 2 And that might be a compromise that might work rather than 3 requiring the regulators to collect, you know, information 4 that it may be difficult to actually do. 5 MR. McVEY: Yeah, I would agree with that. I think 6 simple is better with respect to block trading rules. 7 Fortunately for all of us there is more and more data 8 available on OTC trading activity through the growth in 9 central clearing and also the DTCC warehouse. And I think 10 it's instructive, in terms of which instruments have 11 different levels of liquidity and what block trading rules 12 might apply to, say, a CDS index versus an inactively traded 13 single name. 14 But I think it's important to keep the rules 15 simple. I think TRACE has worked and it's generally simple. 16 The only caveat to that is what's liquid today may not be 17 liquid in six months. So, I don't think you could do this one 18 time and forget about it. I think it requires some regular 19 monitoring, but I think simple rules would be better. 20 MR. DURKIN: Just to echo Bill's comments, though, 21 earlier, and also to compliment your comments, yes, simple is 22 better in the context of the users' need to understand what 23 the rules and the requirements are to be able to accomplish 24 these block provisions; however, there definitely is 25 something to be said for the differences in liquidity based 92 1 on time zones. And there are levers that you can put in 2 place, as Bill has suggested, to accomplish that to deal with 3 those market idiosyncrasies based upon time zone. 4 MR. MASTERS: Just to make one point that I made 5 yesterday in the block panel, I mean, there is a key 6 difference here between pre-trade transparency and reporting 7 and post-trade. I mean, clearly, the post-trade regime is 8 much narrower than the pre-trade period where people are 9 going to, you know, Alltax or -- I'm dating myself here -- 10 but, you know, other vendors to try to discover what the 11 price should be. 12 The post-trade regime, as a general rule, should be 13 much tighter. You know, the public needs to see the trade 14 ASAP and I understand the dealer has to hedge, but in many of 15 these markets, as we all know, these over the counter trades 16 can be broken out into, you know, the least common denominators. 17 And it doesn't take long to -- not as long as maybe some would 18 like to postulate, to actually get your hedge done. 19 So, there is an offsetting public interest here in 20 the sense of we would like to see the data as soon as 21 possible, including market participants, regulators, but we 22 would also like to see it in a standardized format. We would 23 like to see the data in a universal way, so that we can all 24 comprehend the data. 25 So, if we're doing, you know, a billion dollar 93 1 interest rates swap-- the trader on the other side of that 2 swap knows how to break that down into its component parts 3 because he has got to do his own hedge. And so, I would like 4 to see it, in terms of a delta equivalent, to its nearest 5 listed equivalent, if possible, so that we can compare apples 6 to apples. And that's critical not only for regulators to be 7 able to do things like position limits and so forth, but it's 8 also critical for market participants to be able to be 9 involved in these markets because if you want more liquidity, 10 we have got to see prints. 11 We've got to see those prints and see where things 12 happen and that stimulates activity. If we don't see the 13 prints, the post-trade prints, on a quick basis, then the 14 market -- some of these markets are going to be what they 15 are now, in many cases, which is sort of back waters. 16 MR. KNIGHT: I would like to echo that point. Ed 17 Knight, NASDAQ. We certainly believe that the statute seeks 18 to accommodate block trades. We think our own model, public 19 exchanges, are not as efficient in handling block trades. 20 There needs to be an alternative. But what you're talking 21 about is creating private markets and tolerating private 22 markets. And the question is, how much of one do you want. 23 And so, it comes down to the definitions. And to 24 some degree, if the private market becomes so large, you're 25 going to destroy the ability to have a public market with 94 1 transparent price discovery. So, you've got to balance a 2 number of factors. 3 MR. McVEY: I would just, you know, add to that 4 with respect to our electronic trading experience and the 5 speed of price reporting. And I think it does go back to the 6 benefits of electronic trading, even if the transaction is 7 done bilaterally, which it can be done in an RFQ model. On 8 average, for a trade that's completed on MarketAxess through 9 the APIs that we have that facilitate straight through 10 processing, the trade leaves our system, goes to a dealers 11 trade capture system, is immediately sent onto FINRA for 12 trade reporting, and is back and available publicly within 13 one minute. 14 So, I think in a market that we're talking about 15 where most instruments trade relatively infrequently, that's 16 a great example of how e-trading reporting facilitates 17 immediate and real time transaction price reporting for the 18 overall market. 19 MR. DE LEON: I would like to separate sort of a 20 couple of concepts because I'm not sure I agree with some of 21 this line of thought. I think that the public good is to 22 reduce counterparty exposure, to reduce systemic risk, and to 23 have pre-trade price transparency for things. I do not think 24 that everyone has a right to know what everyone else 25 does. 95 1 And so, I sort of disagree with that. And I think 2 TRACE does an incredibly good job of providing good real time 3 information about what has traded in terms of a block, in 4 terms of sizing, and coming up with a meaningful mechanism 5 for it, but every single transaction that gets done creates 6 several issues. It creates the free rider problem, it 7 creates the winner's curse, and it also takes the incentive 8 to do people's own research away, because when someone does 9 a trade, it means that they have done economic work, they've 10 done financial work, they've got an investment guideline 11 they're trying to achieve on behalf of their investors, many 12 of which who are small investors who have pooled their 13 assets, possibly. 14 And that information and research, by doing a 15 transaction, is signaling information. And everyone doesn't 16 have to have that information for free. And that sort of 17 defeats the purpose. The purpose here is to make sure that 18 when people transact, that they get the best price possible, 19 not that they know what everyone else in the market is doing. 20 And I think we need to separate that. And TRACE has done an 21 incredibly good job of sort of mitigating that issue. And 22 it's a very good compromise. 23 And if you look at the equity markets, to get around 24 size and block trading, there are people who spend millions 25 upon millions of dollars to come up with algorithmic models to 96 1 hide what they're doing electronically. So, they break up 2 block trades into small little pieces. So, I think that the 3 market will go out of its way to avoid this. And it will 4 cost money, it will cost investors money, it will hide 5 information. 6 MR. COOK: If we could just finish -- wrap this up 7 quickly. So please go ahead, but I want to move onto some -- 8 9 MR. DIPLAS: Yeah, I'll do that quickly. 10 Athanassios Diplas. 11 I think TRACE is a good example. And we have seen 12 TRACE has affected behavior. That -- it doesn't have to be 13 good or bad, but it changes behavior. And where we set those 14 limits is what dictates the behavior of the counterparty. So, 15 in the example that Bill mentioned that the large trade is 16 broken into smaller pieces in order to be actually liquidated 17 in the market, that means that the risk has moved from the 18 dealer that normally would take it in a large chunk to his 19 counterparty that actually has to take that risk. So, that is 20 something we need to be cognizant of. And where we set that 21 limit is going to dictate that behavior. 22 And the last part is that I agree with Bill, there is no 23 inalienable right for everyone in the market to know where 24 every single trade has occurred. The point of transparency 25 is to enable the participants to do their transactions in the 97 1 best possible price. But if that comes at the detriment of 2 the two transacting parties, clearly we moved too fast. 3 So, it's an issue about achieving that fine balance. 4 MR. COOK: Okay. Thank you. Why don't we move on 5 to the -- some of the other core principles we want to 6 explore. Heather, do you want to ask the next question? 7 MS. SEIDEL: The Dodd-Frank Act, under 8 the Dodd-Frank Act, SEFs, swap execution facilities, have an 9 obligation to monitor trading on their markets and also to 10 enforce the rules that they put in place with respect to 11 trading on their markets. 12 I guess if we could get the panelists' views on sort 13 of in this new world where you have markets that have 14 obligations that they might not have now, how would that 15 work? You know, what do the markets think about how they 16 might go about carrying out their obligations, market 17 participants that would be subject to those obligations, 18 views on, you know, sort of how this would work in the new 19 structure where there are sort of regulatory obligations on 20 the SEFs. 21 MR. HARDING: Julian Harding. Just as a slight 22 preamble, the Wholesale Market Brokers' Association gamely 23 attempted to send off a discussion draft on the core 24 principles for SEFs, as soon as it was able, to both the 25 agencies here. And I hope everyone knows that. I have a 98 1 copy here if anyone needs one. As we know, the core 2 principles for SEF were born of something very similar for a 3 DCM and we've attempted to be flexible and adjust some of the 4 core principles to reflect a more competitive SEF market that 5 we might imagine for the future. 6 But on the specific point, you're asking, Heather, 7 the -- our view on the SRO type of issue. We endorse the 8 idea of an independent SRO populated, hopefully, by 9 practitioners who understand, at least a large body of which 10 will understand, the swap marketplaces, which is not the case 11 right now, probably in existing possible entities that could 12 fulfill the function. And we would also like to imagine a 13 situation where there is one unifying SRO for all SEFs, so as 14 to avoid any sort of imagined regulatory arbitrage where one 15 SRO allied to a single SEF might interpret its obligations 16 differently. 17 MR. KNIGHT: I would like to make a comment just on 18 the statute and how we view it. I think, from the 19 perspective of a stock exchange and subject to exchange-like 20 regulation, our first question was what did Congress intend 21 here. And to me, the most revealing language is in the 22 Senate report where the report states trading more 23 derivatives on regulated exchanges should be encouraged, 24 because it will result in more price transparency, efficiency 25 and liquidity. In order to allow the OTC market to adapt to 99 1 more exchange trading, the legislation provides for what was 2 then called alternative swap execution facility. 3 To me, they're equating the regulation in this area 4 to something very much like an exchange. Then when you look 5 at the language of the statute, the key responsibilities of 6 an exchange is the ability to enforce its rules against 7 members. Its responsibility to do that, its obligation to 8 take disciplinary actions and to investigate. Those verbs 9 are all found in the core principles here. 10 So, and I think there is some logic to this because 11 given the history of the financial crisis, given how the 12 markets operated, I think the system of regulation of 13 futures markets, the cash equities markets, seem to work. And 14 Congress noticed that and wanted something like that in this 15 space. I mean, we complain, at times, about the level of 16 regulation, but it works. And I think the record is there. 17 Those markets did not shut down. They continued to operate. 18 The taxpayer did not have to step in and fund them to keep 19 them going. People did not necessarily like the prices they 20 were getting, but they were continuously operating in a 21 well-regulated manner. 22 So, I think that system of regulation, which 23 Congress did not make any major changes to in this latest 24 round of reform, is what they were looking to hear from this 25 language as a legal matter. Now, the self-regulatory 100 1 organization model, we use it. We outsource to a fully 2 independent, what we believe, the gold standard of regulation 3 to FINRA. We think that opportunity is helpful to the market 4 overall. It provides us with a degree of independence that I 5 think is irrefutable in how we, in particularly, enforce our 6 rules against our members, so that there is no question about 7 impartiality. But I think strictly from looking at the 8 statute and the history of the statute, what the Congress 9 appears to be asking for is something much like the way an 10 exchange is regulated. 11 MR. SCHOTT: Julian and Ed, you guys both in your 12 answers sort of raised two distant concepts. One is who is 13 doing the regulation, and you mentioned the SRO model. And I 14 think we definitely need to talk about that, but I want to 15 just sort of go a step back and go to the first part of your 16 answers. And that is what ought they be doing. 17 So Julian, you mentioned the SEF core principles 18 sort of born of the DCM core principles. And we have certain 19 expectations around what appropriate trading practices are in 20 a DCM and what practices are of concern and so forth. When 21 investigations are conducted, violations are found, there are 22 disciplinary actions. There is a structure in place. And 23 it's fairly standard across the DCM. 24 So, I'm sort of wondering what the opinions around 25 the panel are in terms of what are the sorts of conduct 101 1 that -- whether it's the SEF itself or third party SRO, ought 2 to be looking for. What is a prohibited trading practice on 3 a SEF? How does it differ from a DCM? And once those 4 practices are found, what sort of standards ought to be put 5 in place around sanctioning the conduct? 6 And as you answer that, think about, you know, are 7 some of the concepts, at a big picture level, relevant in the 8 SEF context as they are in the DCM context. We have a lot of 9 rules in the DCM context that revolve around the 10 intermediation of customer trade. And so you are protecting 11 the customer. Are those sorts of concepts relevant in the 12 SEF world? 13 MR. KNIGHT: One comment, which is, be careful if 14 you're thinking about setting standards that are different 15 based upon the instrument standards and behavior. I think we 16 are seeing more and more, and I guess that's another topic 17 here, how these markets are interconnected. We do not want 18 regulatory arbitrage between them. And again, you have a 19 long record of certain principles that I would have to think 20 you believe work and you've got a statute where, you know, 21 I've been asked does this create a national market. 22 I -- and looking at the SEC language in the 1970's, 23 I think this goes well beyond that where the SEC was given 24 the authority to do it without a lot of, frankly, guidance 25 from Congress. Congress has gone a step further in giving 102 1 you explicit guidance about a market that it wants structured 2 for the whole nation. It doesn't say this only applies in 3 the lower 48. This is a national market with national 4 standards and you've got a broad ground of authority to write 5 rules here. 6 MR. MASTERS: I would just say that there is -- you 7 know, in terms of what you were saying with regard to 8 conflicts of interest, there is a lot of issues here in terms 9 of, you know, what, I think, Congress intended, you know, in 10 terms of, you know, when you look at SEFs versus DCOs, for 11 instance, and you talk about, you know, compensation for 12 referring business. 13 I mean, there is a lot of real tricky conflicts of 14 interest here that I don't think the public wants to see. 15 You know, clearing and matching, you know, really need to be 16 separate businesses. I mean, someone that clears ought to 17 have different alternatives. There is an issue, you know, 18 the old, you know, payment for order flow kinds of issues 19 that we've seen in the exchanges. I mean, there are some 20 things that we've dealt with that didn't work so well at 21 equities that we would like not to see in this sort of pro 22 forma regime that you're setting up. 23 But there's -- I think there is, as far as I'm 24 concerned, there is going to be a very severe look at the 25 whole notion of conflicts of interest in many different forms 103 1 and how entities relate to each other and whatnot. And 2 moreover, if those situations are anti-competitive. We want 3 a broad, diverse group of participants involved, but we don't 4 want levels set that prohibit one group at the expense of 5 another group. 6 MR. HARDING: Julian Harding. I guess the most 7 important difference between the SEF and the DCM situation is 8 that the SEFs in the future, as the IDBs are now, are 9 competitive entities. And I think that there is a lot of 10 questions of quality in the future of the surveillance or 11 monitoring that will be undertaken by SEFs. It's just that 12 they can only really do it in the -- within the confines of 13 their own execution facility. So, they can -- as has been 14 coined before, they can see what's going in their own 15 classroom, but they can't see what's going on in the overall 16 school or the playground. So, I think that's a primary 17 difference as to what they're quoted as seeing. 18 MR. WEISBERG: I think the regulators need to be 19 mindful that often these markets that they're writing rules 20 for are global markets. So, I would say even though the rules 21 may apply nationally, the marketplaces themselves are global. 22 Our clients are global. We are in a currency market, which 23 is global. And that really transcends all of the rules that 24 people are writing. 25 So, there is no open and close for a foreign 104 1 exchange market. It's a continuous market. But there are 2 liquidity differences during times of the day and as an 3 unregulated market, it performed quite well through the 4 prices customers were always able to transact. So, I think 5 the regulators have to make sure that continues to be the 6 case after the introduction of rules that I think we all feel 7 could be beneficial for the entire industry. 8 With respect to, you know, compliance, we -- in 9 order to create a competitive SEF market, we think it's 10 important that SEFs are able to outsource or delegate the 11 surveillance function to third parties, appropriate 12 third-party providers. We think the enforcement needs to be 13 simple and not necessarily cumbersome and involve 14 extraordinary expenses for the SEFs. And we think a 15 suspension or revocation of trading privileges on a SEF, 16 is oftentimes a very effective mechanism to ensure that 17 people comply with the rules. 18 MR. DIPLAS: Okay. Athanassios Diplas. In terms 19 of the exchange model versus not, I think it is clear, and 20 the drafters were very clear, that they did not want just the 21 exchange model. That's why they wanted to be able to define 22 the SEF. And they did that in recognition of the complexity 23 of the market, of the derivatives market, that meets 24 something more expansive. 25 So, that creates certain problems and also, you 105 1 know, certain challenges, basically, in that for example, we 2 would have multiple facilities for the instruments traded. 3 We also have multiple venues in which instruments clear. 4 And therefore, as we discussed in the swap data report, 5 sometimes you will be able to look for enforcement of the 6 principles in the actual SEFs and then you might have to 7 go wider as in the data report so we can get a better slice 8 of the whole market. So, I think we need to take a step back 9 and go for a little bit more open model. 10 I don't have strong views of the outsourcing of 11 the -- some of these functions or not, but the point is the 12 market is much more complex than a single silo market that is 13 containing one exchange. 14 MR. SCHOTT: Just one last question on this. Oh, 15 no, please, go ahead. 16 MR. DE LEON: No, I just -- to reiterate some 17 points. I think it's very important that -- I'm agreeing 18 with Athanassios in terms of not being overly burdensome in 19 terms of the regulatory things, so it's not meant to replace a 20 DCM. However, it is very important to keep in mind that as 21 regulations are written and rules are determined, that this 22 is a global marketplace and we want to be careful to avoid 23 regulatory arbitrage or a preferential market treatment. 24 And to the extent that we've seen this now already, 25 a few people on this panel constantly are competing with 106 1 overseas based markets in terms of clearing futures and their 2 offerings, which I think is a good thing in terms of 3 competition is good in this market as long as the right 4 safeguards are in place for futures. We're going to see 5 something similar for clearing of derivatives as well as for 6 SEFs. 7 We want to make sure that whatever rules are 8 created don't give an outright advantage to an overseas or 9 one exchange versus another due to location, because you'll 10 wind up losing control or having less impact where things go 11 and you'll have both the ability to arbitrage by doing trades 12 similar in nature on one exchange versus -- one SEF versus 13 another. And more importantly, there will be a preferential 14 treatment to where you want to clear your business, which may 15 not be here if it's too onerous. 16 MR. COOK: Thank you. I just want to follow up on 17 this, pick up certain aspects of the conversation here, where, 18 Ed, you mentioned the national market system, and other people 19 have drawn analogies to the equity markets and this question 20 of arbitrage and different rules resonates a little bit with 21 some of the debate that is happening in the equity market now 22 where there has been a mandate to develop a national market 23 system. And there is, has been an effort to balance 24 competition of orders with competition of trading venues. 25 And, you know, that's constantly the issue is to get the 107 1 right balance there. 2 So, the question I would ask is when we're talking 3 about the rules that SEFs will impose on their members, how 4 uniform should we expect those to be. How much -- what's the 5 cost of allowing SEFs, each SEF to develop its own rules. 6 What's the benefit of allowing each SEF to develop its own 7 rules. If you're an investor trying to trade in these 8 markets, how important is it to you to have a uniform set of 9 rules that you can follow, so you have certainty, regardless 10 of which platform you're trading on, as to what the rules of 11 the game are, how much will imposing a requirement like that 12 restrict the innovation of models that some people have 13 argued on this panel and the last panel is important to the 14 development of this market. 15 MR. McVEY: I would tend to favor a uniform set of 16 rules for all SEFs. And I agree with the comments earlier 17 that being able to discharge someone -- not discharge, but 18 share some of those responsibilities with a third party that 19 may run a business in supervisory oversight of SEFs would be 20 important from a cost standpoint. 21 I think historically, electronic trading venues 22 have successfully competed along the lines of technology, 23 liquidity and price, and I do think that you can enhance 24 competition in this space while at the same time, having 25 consistency and efficiency in the rule process. 108 1 MR. VISWANATHAN: I would tend to agree. You 2 probably want to standardize clearing rules. The last thing 3 you want to do is to find that you could clear across 4 different SEFs in different ways. And you probably want to 5 standardize some transparency rules and reporting rules. But 6 within that, you want to allow -- from the equity markets, 7 we've learned that allowing some degree of competition, kind 8 of shakes the status quo in positive ways because it can lead 9 to other outcomes as we've seen with stock pools of 10 liquidity. But if you can get the right innovations and not 11 restricted, I think that would actually be socially 12 beneficial. 13 MR. MASTERS: I would just -- this is Mike Masters. 14 I would just say that, you know, I think the key focus here 15 for the regulator is to prevent a race to the bottom in the 16 sense of, you know, establish a uniform set of rules. 17 Innovation is great except when it concerns regulatory 18 arbitrage. And if you don't have a standard of rules that 19 people across the board are -- have to comply with, then 20 you're going to get a race to the bottom because, you 21 know, one person has a -- you know, you have an easier time 22 doing business here than there and people won't actually go 23 there and then you're defeating your purpose as a regulator. 24 So, I think it absolutely has to have -- there needs 25 to be standard rules. It has to come from the regulator. 109 1 And I think the competition and stuff will take care of 2 itself once people know what the playing field is. 3 MR. SEMLITZ: Don't you think that there's going to 4 be issue with international versus domestic. That, even if 5 you have standardized rules in the United States, that 6 overseas you're going to have a different set of rules that 7 already exist today. And it will be just compounded as we 8 go forward. So, then how are you going to deal with the fact 9 that U.S. rules, even if they are standard, are then offshore 10 rules? 11 MR. MASTERS: I think that in 2010, fiduciaries in 12 general want more regulation, not less, as a matter of 13 precept, in terms of where they're trading and where they're 14 clearing. You know, there is -- you know, quite frankly, 15 there's -- I'm not sure there is that much you can do about 16 overseas other than linking -- trying to link regulatory 17 groups together around the world. And I think that's 18 something that needs to be done proactively by regulators. 19 But clearly, I mean, someone has to come up with a 20 standard. And I actually think if a market is known for 21 being the one with the most integrity, the one with the best 22 chance of having the most fiduciaries and institutional 23 investors around the world where there is the most trust in 24 that market, I think that's going to naturally attract 25 business. And again, maybe you can get, instead of a race to 110 1 the bottom, you can get a race to the top. 2 MR. KNIGHT: I mean, what we've observed in Europe 3 is that the regulators there are watching very closely what 4 the U.S. is doing and often waiting to see what they do and 5 follow what they do. And I think some of them are pretty 6 open about that and other countries the same way. So, I think 7 to say don't do that because people won't follow I'm not sure 8 is correct. 9 In terms of uniform rules across market centers, I 10 think there is the factor of, of course, different market 11 structures and we would need -- if a SEF had a floor, a 12 different set of rules and additional principles and all the 13 electronics, but I think it should be core principles that 14 apply across the various venues to promote competition. 15 MR. DURKIN: I agree with that last comment. I was 16 a little concerned that, you know, not every organization or 17 market is exactly the same, that there should be core 18 principles that need to be followed across a sector, but you 19 must allow for some level of flexibility on the innovation 20 side of things. So, you know, the rules of the game, you 21 know, should be very consistent across the industry, but 22 there should be capabilities there for a market to be able to 23 innovate. 24 MR. DENIZE: Again, as an end user, I think we 25 certainly are supportive of measures and processes that would 111 1 lead to supporting competition, but our traders don't want to 2 have to figure out the arbitrage among, you know, 20 SEFs and 3 what rules they might be applying to themselves and so forth. 4 So, the uniform set of rules makes our decision-making a lot 5 more streamlined and we can focus on the true issues that 6 we're focused on for our products and for our participants, 7 which would be the elements of the economic products and the 8 economic risk mitigants that we're seeking. 9 One of the issues, as you do set up the uniform set 10 of rules, that we are concerned about, is to be -- to ensure 11 that the governance process certainly includes the voice of 12 the end users from the outset and throughout the process. 13 One of the clear observations of the OTC derivatives market, 14 prior to the crisis, was its lack of transparency and lack of 15 openness. 16 As we move to a regulated environment, the end 17 users have a strong voice that need to be heard throughout 18 that process. Even though we may not own the clearinghouses, 19 we may not own the trading systems and we don't necessarily 20 need to do that, although we should be, perhaps, members or we 21 should be participating in the advisory committees and the 22 rules and governance committees that are applying these 23 uniform sets of rules. 24 MR. DIPLAS: I would agree with some of these 25 comments with respect to the end users. I think if you have 112 1 seen something encouraging over the last couple of years is 2 that the governance structure in the market has changed to 3 include the end users, and I think this is going to continue. 4 The same thing goes for the CCPs and, as you've mentioned, 5 the advisory committees, including the end users, and this is 6 going to continue. 7 The second thing that's also encouraging is that I 8 think we have good evidence right now of international 9 regulatory cooperation. We have the global supervisors 10 forum, which now includes, what, 30 or 40 participants, more 11 or less. And I think that that -- they're taking that job 12 pretty seriously. So, in order to kind of -- it's a balance 13 of trying to ensure that we can have innovation and at the 14 same time have actually a well regulating environment we need 15 to, I think perhaps, look at the more hybrid system. 16 We start at the top with the core principles. 17 Perhaps we have some kind of set of super rules that actually 18 apply to everybody and then you have a certain lower set of 19 rules which actually are more flexible. But there are the 20 different steps to operate, because they deal with 21 different products and they need that flexibility. It allows 22 the new participants to come in and offer something better. 23 So, that would be a bit more flexible and balanced approach, 24 I think. 25 MR. YELVINGTON: I think also having, at the very 113 1 least, a core set of rules around the regulation there really 2 might engender quite a bit of innovation, because it takes 3 away a lot of the business risk for somebody looking up to 4 set up a SEF. It actually helps them know their boundaries 5 at least on a minimum basis. The same could be said with the 6 conversation earlier this morning when people were 7 discussing, you know, a standard for RFQ versus central limit 8 order book. 9 If you define the minimum and provide a, you know, 10 a metric by which the market can evolve itself, then we'll 11 get a better answer as to what's preferred by the market for 12 different instruments, and I think that you'll probably end 13 up seeing, not only on the regulatory side, but on the market 14 structure side, different instruments will trade in different 15 markets just because they trade better there. 16 MR. COOK: So let's move onto the cross-market 17 portion. 18 MR. SCHOTT: I think we've already touched on a lot 19 of these issues, but, you know, one of the things that we're 20 struggling with is how do you appropriately regulate a market 21 where a product can be trading across so many different 22 platforms. Someone mentioned the schoolyard model. What 23 happens in a classroom model? Who regulates the schoolyard? 24 We can say the SEFs can regulate what's happening in their 25 own market. That leaves a gap. 114 1 People have suggested a super SRO. I was wondering 2 if just the panels could give their opinions a little bit as 3 to who that entity might be, what sort of a relationship 4 between the SEF and the SRO, what role the SEFs should still 5 have, you know, either in terms of overseeing the services 6 being provided for them. If there are super rules versus 7 some local rules for SEF, does each SEF oversee its 8 particular rules and leave the sort of the master rule book 9 to the overall SRO. This is an area where we're just giving 10 a lot of thought to. So your opinions would be welcome. 11 MR. DE LEON: Hi. This is Bill De Leon. I think 12 it's important to think about the roles that SEFs play in 13 terms of the overall connectivity and the whole process. And 14 that will sort of give you some guiding principles. I agree 15 with the comments here that there should be some set of 16 minimum standards for SEFs in terms of what they report, how 17 they regulate requirements in terms of trading information, 18 but there should be innovation. 19 The two important things regarding that that I see 20 is once a transaction is done on a SEF, it has to be given up 21 into a CCP. So, you need to know that there is a process and 22 capital behind the trade that's done on a SEF, so that it 23 winds up in a CCP and is matched. And that's an incredibly 24 important thing. To the extent you're trading directly on 25 the CCP -- so, if I were to use ICE and then give up into ICE 115 1 or I trade on the CME and give up into the CME, it's one 2 unit. I have a lot less risk. But if I trade on some other 3 third party vendor and then it gets given up into ICE or CME, 4 or whoever it may be, I need to know that the process between 5 the transaction and it getting there is going to be highly 6 monitored and there is going to be capital there. 7 So, just like all of us do detailed counterparty 8 exposure management and credit review of any CCP we're going 9 to use, because we want to make sure that the CCP stands 10 behind it and whoever we are using to clear our trades are 11 sufficient, when I do a transaction with a SEF, I need to 12 know that they need some minimum standards that I have 13 incredibly high degree of confidence that that trade winds up 14 in the CCP. 15 It would be analogous to what we see when we trade 16 equities. I can trade equities on multiple venues and then 17 it gets given up into my actual clearing account. The beauty 18 of most of the equity markets in the U.S., and probably all, 19 is that I do a trade and very few people worry about where I 20 did that trade and it getting cleared and me taking 21 counterparty risks. Obviously I take credit risk or equity 22 risk by doing that trade, but that component is taken out. 23 Whatever regulatory environment, whatever rules are 24 set, I would argue you need to make sure that the three legs 25 here are similar, that the -- when I do the trade, who I use 116 1 to do my SEF execution is going to stand behind the 2 transaction, the SEF itself will stand behind it, and then it 3 will wind up in my CCP. And that connectivity and the credit 4 risk there is limited or removed as much as possible and 5 there is an overriding set of principles and rules, so that I 6 can sleep at night, about the counterparty exposure, the 7 clearing, the connectivity. 8 MR. SCHOTT: Do other panelists have thoughts just 9 around some of the more practical elements of establishing 10 this SRO and sort of the legal relationships between the SEFs 11 and the SROs, the financing of the SRO? 12 MR. SEMLITZ: To have a consistency in 13 regulatory. It would seem to me that the CFTC ought to be 14 regulating all SEFs that relate to commodities, and the SEC 15 ought to be regulating all SEFs that relate to securities and 16 apply the same standards you're using today across all 17 markets. Therefore, it takes out the regulatory arbitrage. 18 You've got counsel at every firm who is used to dealing with 19 the regulators who are regulating them already and that 20 creates a tremendous amount of consistency across existing 21 markets today and the markets that are going to exist 22 tomorrow. 23 MR. SCHOTT: Speaking of markets existing tomorrow, 24 we have, I forget what count we're at now, 330 some days 25 before sort of this goes live. If that third entity isn't 117 1 set up yet, what's the interim solution? Any ideas will be 2 accepted. 3 MR. KNIGHT: I think if the demand is there, and I 4 believe the demand will be there, there will be offerings 5 here and I -- you know, you look again at the FINRA example, 6 both the New York Stock Exchange and NASDAQ use them, and 7 they facilitated the entry of many other competitors in the 8 space. So, I don't think you can no longer say that 9 regulation is a hurdle to competition in our space. I can't 10 see why a similar model wouldn't work. 11 MR. DURKIN: And I think there is something to be 12 said for the template that, you know, you've already 13 established in terms of the respective agency's oversight 14 into the centralized markets as they exist today and, you 15 know, the establishment of, you know, rules, criteria, 16 capital requirements, you know, a lot of the things that 17 Bill, you know, had expressed, you know, the need and desire 18 to have in place. So, there is that confidence for this 19 central counterparty risk management capabilities. I mean 20 there is a very solid template out there that can be built 21 upon to adapt to this new evolution. 22 MS. ADRIANCE: Thanks. In terms -- I mean, 23 we've -- there was a little silence there when it came to 24 suggestions for a possible third party, except there has been 25 one or two suggestions. But in terms of this -- to get back 118 1 to what was mentioned earlier, the interaction between the 2 individual SEF, well, I should say derivative SEFs, and if 3 there is this third party that is out there in the shadows 4 that we do not yet know who exactly it's going to be and it 5 may be more than one, or maybe it's one, it would -- so there 6 are several questions there. 7 One is what, you know, what is the -- what 8 responsibility -- it is one that Sebastian asked earlier. 9 What is responsibility that you see being given up as 10 compared to what is being retained. It was mentioned 11 earlier, for instance, that an appropriate step, if there has 12 been a some kind of you've gone to the design process, there 13 could be some kind of -- one appropriate step would be that a 14 trader has lost access to that SEF. Well, is it just to that 15 SEF or is it across all SEFs. And who does that? 16 Is it -- so if it's just one SEF, that SEF can just 17 pull the plug. But should it be -- you know, if something 18 has happened on one SEF or has happened on multiple SEFs, 19 should there be some coordination between what happens? Is 20 this third party, this mystical third party that is going 21 to do this? Should it be that each SEF has a responsibility 22 to do its own after being notified by the CFTC, the SEC, some 23 third party? 24 How is this interaction going to work when you've 25 got -- you may have issues that are not going to be kept 119 1 to necessarily one SEF or in each SEF you -- if there 2 are several parties who end up being SEFs, you may not 3 know, on one hand, when you're seeing a problem, if it's 4 happening on other SEFs. What should happen? 5 MR. KNIGHT: Well, I mean, in the exchange space, 6 there is some robust principles in this area. There is the 7 principle referring things to the appropriate regulator when 8 you see a problem even if you have a suspicion of a problem. 9 So, it gets widely disseminated. You have access rules that 10 make it clear when you can deny access to someone for a 11 regulatory problem. And again, you have to share that 12 information. 13 And so, I think there are processes in place that 14 you can borrow that are pretty well established. You know, 15 one is all stock exchanges -- and the SEFs I wouldn't see why 16 you wouldn't do the same -- you are regulatory officers and 17 you have an obligation to enforce the rules. Even though you 18 may be a salesman, if you see something, you have to report 19 it. You know, these basic principles, I think, serve us 20 well. 21 The question of what you would outsource or not, I 22 think it really depends on who the vendors and what the 23 capabilities of the SEF are, but generally, the most awkward 24 aspects of these, if you will, self-regulatory 25 responsibilities are enforcing rules and disciplining and 120 1 investigating your own members who are also customers. 2 So, that is what we outsource to FINRA. And, but 3 the real time surveillance of the market and issues related 4 to that we conduct internally. The process of determining 5 what fits on our market or not we do internally through an 6 independent group that's again subject to oversight by the 7 SEC. But, you know, the awkward piece is enforcing those 8 rules against people who your salesmen are visiting the day 9 before. And that's where outsourcing works. 10 MR. DURKIN: Well, I would have to take a little 11 issue just from our existing model, and I'm using the 12 centralized market at the CME group. I mean, we are, you 13 know, a self-regulated organization. We have very stringent 14 rules and regulations. We have a very aggressive market 15 surveillance and trade practice surveillance program. 16 We do take actions and we do review all of our 17 activity that occurs across all of our products every day. 18 And if we find trade practice abuses, we take appropriate 19 action and it gets publicized to the CFTC as well who 20 may also take action and decide to pick up the case. So, I 21 would just be, you know, very careful in saying the 22 awkwardness of being able to monitor and regulate your own 23 markets because that model has worked and it has been in 24 place for many, many years. 25 MR. KNIGHT: Yeah. I premised by saying if you 121 1 have the capability. You all do it well and I wasn't in any 2 way criticizing your model, but I think the option of being 3 able to do that is healthy out there. I think. 4 MR. McVEY: It was a multi-part question. So I'll 5 take a crack at a few pieces of that. But FINRA has been 6 mentioned several times on the panel today as a logical 7 entity to provide some of the oversight services. I think 8 NFA is another logical entity that would be willing to 9 compete for that business if it's made available. 10 I personally think that any SEF has to be 11 responsible for the fairness and safety and reliability of 12 their own marketplace, but I would agree that when it comes 13 to arbitration and enforcement of the rules and fines, that's 14 an area where we think it would be logical to lean on a third 15 party like FINRA and the NFA. 16 And I also think that since I believe you do want 17 to encourage competition in the SEF space, a large trader 18 reporting is another piece that could be handled better on an 19 aggregate basis by one of those entities, as opposed to a 20 variety of different SEFs. 21 MR. DIPLAS: I would agree with that part. I 22 think, to go back to Riva's initial question, you need that 23 third party because you acknowledged that the market is a 24 little wider than that one silo containing one SEF. And so, 25 the SEF can actually do on its own what it can see, but you 122 1 need that third party to deal with the stuff that that one 2 SEF might not be able to see. 3 I would be agnostic in terms of which entity, in 4 particular, could actually carry out that task, but at the 5 same time what I would want that entity to have as a 6 qualification is the knowledge and experience in the 7 underlying asset class. The way SEFs are most likely going 8 to be organized is along asset classes. And so that 9 expertise is, I think is going to be very important. 10 Now some SEFs might actually do multiple asset 11 class in which case, in some ways, that third party could 12 cover those or it could be a multiple of those. But I think 13 the expertise in that asset class has to be demonstrated and 14 something that we need to focus on to make sure they 15 understand that because what works in equity, doesn't work 16 in trading, et cetera. 17 MR. COOK: I want to make sure -- thank you. I 18 want to make sure we have time for our last topic, which is 19 an important one, and it's standards for SEFs to fulfill 20 their obligation to maintain impartial, open access. So, 21 why don't we begin the questions on that topic. 22 MS. SEIDEL: Well, I think the -- you know, as 23 Robert noted, there are obligations on SEFs in the Act to 24 ensure impartial access to their markets. And so, we would -- 25 you know, sort of an open question as to how would SEFs go 123 1 about doing that and what types of entities, you know, would 2 or should be allowed to have direct access to the markets. 3 MR. McVEY: I will take a crack at that since no 4 one seems to be diving in in any hurry. But I think the 5 criteria for access have to be fair and impartial and, you 6 know, ideally, publicly disclosed. We have an institutional 7 credit market and as a result, qualified broker-dealers are 8 welcomed to make markets on the market access system and 9 qualified institutional buyers are welcome as buy side 10 participants. So, I think if there are fair and impartial 11 criteria and they're publicly disclosed, then I think the 12 open access issue would be addressed. 13 MR. DURKIN: I would just say that every registered 14 entity should have the ability to set the terms and 15 conditions for their participation on their direct system. 16 And within that, there should be, then, impartial access. 17 But I do believe that, you know, every registered entity 18 should have that flexibility to determine what those 19 requirements are. 20 MR. DE LEON: Hi. This is Bill De Leon. Yeah, I 21 think it's important that things be impartial, but there have 22 to be minimum capital and/or regulatory standards for anyone 23 to have access or to provide access to clients. This is 24 analogous to sort of the know your client and as well as sort 25 of to become a member of an exchange, you would have to pass 124 1 certain minimums. 2 What those minimums are, I'm not referring, you 3 know, someone should set, but there needs to be a uniform set 4 that are not, you know, capriciously set or sort of too high 5 to hit the bar for anyone joining. But there obviously needs 6 to be a bar set, because there is risk associated with the 7 transaction being done that the person who does the 8 transaction can't stand behind it or that even if they can, 9 that the SEF needs the capital to make sure it gets all the 10 way through. 11 So, you need to ensure that the SEFs themselves and 12 the clients who use the SEFs on the way to having things 13 cleared in a CCP are not creating systemic risk in the 14 system. So, there needs to be these minimums. 15 MR. SHILTS: Bryan, is that what you were pretty 16 much referring to? 17 MR. DURKIN: That is, you know, that they need to 18 equally enforced across the participants and that, you 19 know, you have these types of minimums established, the main 20 predicate of maintaining market integrity. 21 MR. DIPLAS: Athanassios Diplas. I would agree 22 with both Bryan and Bill in that respect that standards are 23 needed, but those standards have to be objective and they 24 have to be transparent. But standards are needed, especially 25 since we are actually linking the SEFs to clearing venues and 125 1 we need to ensure that actually the trades are -- actually 2 are going to get where they're supposed to get and the 3 participants have claimed, for example, that they can offer 4 clearing services are indeed able to do so. 5 MR. HARDING: Julian Harding. Yes. It can't be 6 emphasized enough the need for impartial non-discriminatory 7 access to the DCOs from the SEFs. The existing IDB, entity, 8 the broker marketplace, as I said a little earlier today, 9 will necessarily be expanded to include those entities that 10 are mandated to now trade through a SEF or a DCM for a 11 clearable swap. And further to that, there will be an 12 evolution, I think, the way the public policy is constructed 13 towards further entities who are not mandated to be -- to 14 trade through a SEF to be -- to trade through a SEF in the 15 future. 16 The constraint that has existed to date has been, 17 as I mentioned before, difficult creditworthiness issues 18 around the time of the trade, which now with clearable 19 solutions being offered and mandated more broadly will be 20 removed. And that will open up the marketplace, the SEF 21 marketplace to a greater -- a broader population for sure. 22 MR. MASTERS: I would just echo those comments. I 23 mean, there is a clear anti-competitive provision in the 24 statute here that we're talking about. I mean, there is -- 25 you know, especially with regard to SEFs. And one of the 126 1 points I was making earlier with regard to conflicts of 2 interest with SEFs and DCOs, is you don't want DCOs, you 3 don't want a linkage there where there is anti-competitive 4 kind of behavior. You want access from participants. 5 It's one thing to clear a trade, it's another 6 thing, you know, with regard to a SEF. I mean, trading and 7 clearing, you know, we've got to make sure that there is -- 8 that the market is anti-competitive, that we allow a 9 significant amount of participants to get involved in these 10 markets, especially in the fact, because they are going to 11 evolve. They're going to move forward. And I think that's 12 the healthiest alternative. 13 MR. KNIGHT: Appropriately, there has been a lot of 14 focus on the initial access to the market. But I think, just 15 from a regulatory perspective, there is this question of once 16 you have access, you have a group of members, if you will, 17 there are times when you have to, for regulatory purposes, 18 remove someone. And that has to be done also in a uniform 19 manner. You can't have ambiguity around that. 20 MR. YELVINGTON: I would also like -- you know, in 21 the spirit of the legislation here, what you want to do is 22 get some -- you know, the way I read it, you want to get more 23 participants to help pricing, to help markets in troubled 24 times. And when you're thinking about crafting the actual 25 regulatory framework around this, probably the best thing 127 1 that could be done would be to craft a framework that is 2 extremely objective and not have a lot of opacity or 3 subjective type of requirements. And also make sure that 4 there is an ability for a firm to transfer in, a firm that 5 could grow and mature and become a SEF member or a CCP 6 member. 7 MR. DIPLAS: I think this is a final 8 clarification though. I think that we're mixing up a little 9 bit the access issues between access from SEF to 10 clearing venues and access within the SEFs. And these two 11 are two distinct issues. I agree that, first of all, all 12 clearing venues should allow access to the various SEF as 13 long as they actually fulfill their own criteria. The topic 14 here is, of course, access into the SEF itself. 15 In that as well, we have to be clear that the advent 16 of clearing does not eliminate counterparty risk. It greatly 17 reduces it, but there are still aspects of the counterparty 18 risk that are varying types of the performance of the 19 clearing broker to actually get the trade there. There is 20 also sometimes exposure that someone gets within the same 21 account with other clients, et cetera. 22 So, it's noticing that basically counterparty risk 23 isn't basically being completely eliminated through the new 24 system. It has been reduced, but clients, again, will have 25 to do their own homework to ensure that the counterparty 128 1 they're trading with is actually within their trading 2 parameters. 3 MR. YELVINGTON: I agree on that point. I would 4 also point out that, you know, far from eliminating 5 counterparty risk, what you've done here is neutralized it. 6 And, you know, in an act of neutralization, you should seek 7 to, again, on an objective basis, kind of diversify those 8 counterparty risks out in such a way that, you know, you've 9 reduced the dependence on one individual or a small number 10 of individual participants. And whether we're talking about 11 access to the SEF or access to the CCP, still maintain that 12 the objective nature of things should be maintained. 13 MR. HARDING: Julian Harding. Just a -- you're 14 right to qualify what I said in terms of removing risk. 15 Absolutely. 16 I would like to bring up the historical precedent 17 of the energy markets going back seven or eight years. In a 18 post Enron crisis environment, there was a natural, dare one 19 say, organic shift towards a clear solution that was offered 20 by certain people around this table and the -- what might be 21 SEFs in the future, but the IDBs at the time fully embraced 22 that and, in fact, were definitely part and parcel of a 23 broadening of the population that could avail themselves of 24 the IDB markets and other markets by virtue of that move 25 towards a cleared solution. So it did work. It has worked 129 1 very well, naturally, seven or eight years ago and to this 2 day in the energy spectrum. 3 MR. DE LEON: Bill De Leon. I just want to get 4 back to what Athanassios was talking about on environment. 5 Neutralization in and of itself is a good concept, but if the 6 entries into that neutralization pool bring down the overall 7 quality of it, it actually is worse than having few 8 participants. So, you need to have a minimum standard. It is 9 not just having additional participants being good, you have 10 to have the right people. And there just has to be a minimum 11 standard and they have to be enforced. 12 And that is why when you look at the futures market 13 it has worked so incredibly well. When you look at the 14 equity markets, it has worked so incredibly well. There are 15 minimum standards in terms of becoming a clearing member, 16 there is a minimum standard in terms of being able to execute 17 trades and give things up and the rules in terms of 18 segregation of assets and cash are very well defined for 19 certain aspects of the market. 20 So, you have to have the same concept apply. Just 21 because you're executing a trade the person who is doing it 22 or the exchange you're using and who you're using to 23 facilitate that needs to be able to stand behind it. If they 24 can't, they're going to dramatically weaken the confidence in 25 the system and could possibly lead to people taking risks 130 1 that they're not aware of and/or making the system less 2 stable, which we've seen during the crisis. 3 So, I would just want to reiterate that I agree with 4 the concept of more participants diversifying things as long 5 as they are the right participants. And by that, I define 6 the correct capital and the correct regulatory oversight 7 meeting certain regime minimums. 8 MS. ADRIANCE: Just to, in terms of the, what has 9 been talked about, okay, there was a -- mentioned impartial 10 access to the SEFs and impartial access to the clearing, 11 central clearing parties. One thing that has not been 12 mentioned -- it was mentioned that there should not be some 13 tie between the SEFs and the clearinghouse that would not be 14 partial access to the clearinghouse. 15 In terms of what the -- we need to just go a little 16 bit further from just in addition to what standards of who 17 can access the SEF, can the SEF -- is it impartial access if 18 the SEF is saying, okay, here are my standards. They're 19 based on, you know, as mentioned, in terms of the financial 20 standards, or whatever, that are real basic and they look 21 really good, but get sent into us as the regulators and it 22 looks good, can the SEF differentiate between what does 23 access mean. 24 Is there a different bandwith that's provided? Is 25 there different pricing that's charged, depending on the 131 1 amount of trades? What are -- what is impartial and what's 2 not impartial when it comes to what the SEF can -- how the 3 SEF can treat the traders separately from these overall 4 basics? In a sense, if it's risk management kind of standards, 5 what other kind of links can they put in? 6 Can it be that the SEF can offer certain extra, you 7 know, they can offer more information if that trader also 8 uses other services that particular SEF, you know, and 9 affiliated businesses or if they use the particular -- you 10 know, it's not that they don't offer the ability to link -- 11 to do trades and link to non-affiliated clearinghouses, but 12 if you do do the trade on the system and you link to the 13 affiliated clearinghouse, can they charge differently? Can 14 they treat you differently? How does that play? You know, 15 does impartial access have some kind of impact on that? 16 MR. VISWANATHAN: Vish Viswanathan from Duke. I 17 guess you're kind of opening the can of worms that payment 18 for order flow opened in equity markets if you start allowing 19 me to bundle clearing and, you know, trading in some way. You 20 say oh, if a clearinghouse access, if you trade through, you 21 know, this particular SEF, you know, I'm going to give you a 22 discount, I think that defeats the purpose of the -- you can't 23 allow that kind of bundling. Neither can I think you can 24 allow differentiation of information in some way. 25 Now with any auxiliary services, I don't know. It 132 1 seems to be very clear that all the participants in the SEF 2 should receive from the SEF itself the same information. 3 There can't be these kind of inducements to kind of bundle 4 clearing and trading. 5 MR. HARDING: Julian Harding. I would echo that 6 entirely. The existing over the counter marketplaces and 7 especially those operated by the IDBs do not have anything 8 like you're mentioning to date. The example I mentioned of 9 the energy markets in the last seven or eight years having 10 transitioned into a different sort of framework, there is no 11 sign of that either. So, there is new entrance that came in. 12 We're not -- I'm not in any way disadvantaged. It's not 13 something that would ever come up. 14 And frankly, I think, in a proper genuine 15 marketplace, the totality of market participants all expect 16 certain standards to be met. And they are uniform standards. 17 And I don't think, in a genuine multiple to multiple 18 marketplace, you could have anything less than that. 19 MR. DIPLAS: I think that -- I would agree that in 20 terms of the access to information, it's a business that's 21 going to drive the fact that you cannot differentiate 22 customers with different information. Everybody would want 23 to have that. So, I think that takes care of itself. 24 In terms of pricing, I think we need to be flexible 25 in terms of what people do with their pricing. In the current 133 1 exchanges right now, I think most participants have different 2 prices that they can negotiate depending on their volume. 3 And I think that the inter-dealer broker, it is definitely 4 the case. I mean, it will be -- I cannot think that there is 5 anybody that has the same price with their broker than 6 anybody else. I don't even know what the other banks pay. 7 People negotiate these things individually. So, I think to 8 the extent that this is a private business, you should be 9 able to negotiate these things. 10 MR. HARDING: Sorry. I thought you were referring 11 to within one SEF the notion that certain people have 12 different access and certain have different information. 13 That's what I was referring to. But between the SEFs, there 14 is going to be -- it's a competitive marketplace, and there's 15 going to be differences in approach and in pricing and 16 various things, but -- 17 MR. DIPLAS: Sorry. Athanassios Diplas. Just to 18 clarify, perhaps I didn't understand you. What I'm saying, 19 if we go currently to, to a broker, let's say that 20 that broker, just for argument sake, is a SEF itself. 21 Currently, I don't know, we pay totally different price than 22 someone else does. And I don't know how you determine that 23 price. Part of it is negotiation, part of it is the volume 24 that you bring to the business, et cetera. 25 So, you don't have an environment right now that 134 1 we all pay the same price. So, I don't see how in the SEF it 2 might necessarily be the case. It might be. That could be 3 the model that you choose. I would prefer not to prescribe it 4 to you that that is your only choice. 5 MR. HARDING: What I'm getting at, I guess is if 6 you, as you say, for instance, go to two or three brokers, 7 you're going to choose the best price to operate that. I 8 don't understand when you say I'm going to get the same price 9 as someone else, you'll -- 10 MR. DIPLAS: Perhaps I wasn't clear. Within your 11 own brokerage -- let's take an example. I'm not picking on 12 you. Within your brokerage, I don't think all your clients 13 get the exact same price for brokerage. Most people -- from 14 what I understand, most people have different prices. 15 MR. HARDING: Are you talking about commissions? 16 MR. DIPLAS: Yes. 17 MR. HARDING: Oh. 18 MR. DIPLAS: That's what I mean. I mean, that's 19 the price that -- the SEF charge. 20 MR. HARDING: I'm sorry, Riva. Were you referring 21 to commission schedules? I didn't realize that at all. 22 MS. ADRIANCE: No. Actually, I was referring to 23 looking at one SEF, since we were talking about regulation of 24 SEFs, can a SEF if, you know, since they have this 25 responsibility to provide open, impartial access, can they 135 1 charge differently for different participants, can they 2 provide different bandwidths for different participants, can 3 they, in a sense, provide some different information. 4 If it can -- what does impartial access mean. I'm 5 trying to get down to besides its overall term, okay, we've 6 got to have clear standards and let people in under clear 7 standards, can they -- is there any kind of differentiation 8 of how a SEF can treat the different traders on that SEF. 9 MR. COOK: Yes. I think we're talking about 10 commissions but by another name. So differential fees 11 for differential services even though once they meet some 12 across the board objective access requirement. 13 MR. WEISBERG: To be economically viable, and I 14 think everybody wants economically viable SEFs, I think they 15 have a responsibility to set objective access criteria and 16 disclose what those are. So, that may mean they can't decide 17 a particular person couldn't fall in that class, but once 18 they define that objective criteria and say okay, if we want 19 to have a class of participants that are market makers and 20 they're going to uphold a responsibility to be on a price all 21 the time, those people obviously have different bandwidth 22 requirements than people who trade intermittently, 23 infrequently. 24 And your economics are different. People who buy a 25 lot of things from you oftentimes get different prices than 136 1 people who buy occasionally just because the economics of 2 serving them are different. So, I think it should be okay 3 for a SEF to define those classes of participants so long as 4 they're disclosed, and that they can objectively apply 5 those criteria uniformly across all of their clients. 6 And I think those criteria may be different from 7 SEF to SEF because they may have slightly different economic 8 models, slightly different operating models, slightly 9 different market mechanisms. In some cases, the information 10 requirements could be different, but we shouldn't prohibit 11 somebody from going from one class to another so long as they 12 meet, you know, those criteria. 13 And I think it is super important for people to 14 recognize there are likely to be many more SEFs than DCOs, 15 that there is a lot more risk sitting in DCOs than are in 16 SEFs. The capital requirements could very well be different 17 between a DCO and a SEF. And what that's going to mean if 18 you want to create a competitive SEF market with fewer DCOs, 19 that, you know, you have to make sure that it's not just 20 bundling of price access, but access to the APIs, access to 21 technical environments, access to the quality assurance 22 departments, those types of things, that all need to be 23 opened up, so that any SEF could successfully send a trade to 24 a DCO. 25 MR. SCHOTT: You mentioned different classes of 137 1 participants. But within a class, assuming they are 2 providing or receiving the same service, would you still 3 allow different pricing? 4 MR. WEISBERG: I don't think that's an absolute 5 requirement. It is -- we try to make ours as uniform as 6 possible and I think it would be possible for SEFs, to make 7 them uniform. 8 MR. SCHOTT: I should have turned that the other 9 way around. Would you allow for diverting pricing. 10 MR. WEISBERG: I don't think it's a requirement 11 that there is. 12 MR. DE LEON: You know, what Athanassios was 13 alluding to and we've seen in the markets, different dealers 14 for even something as generic as equities will charge 15 different prices depending on your access point, what the 16 type of trade is, whether it's voice, whether it's 17 algorithmic driven, whether or not you're getting research. 18 Some people do soft dollars. 19 There are a whole host of things that are involved 20 with the pricing of getting access to trading. And the 21 market bears what it bears and people negotiate and do those 22 things. But I think what the important thing is, that the 23 market will drive prices to where they go for the thing, but 24 everyone should have access to it as long as they meet the 25 certain minimum standards required to show that they're 138 1 responsible traders. 2 Just like you have KYWC, you have, you know, your 3 client issues in terms of equity and opening an account or 4 fixed income or anything, you'll have similar issues 5 associated with SEFs and people have access to SEFs and to 6 DCOs, the pricing or commissions, to be blunt, will be driven 7 by what people think is fair and it will be competitive. And 8 I don't see a problem with that as long as it's not gouging. 9 MR. YELVINGTON: I agree with that. I mean, having 10 the ability for different SEFs to, you know, tailor their 11 businesses to their clients is a good thing and, you 12 know, although it's maddening sometimes, I think there are a 13 lot of things that we all buy that we all pay different 14 prices for, depending on who we are. It helps the businesses 15 to grow to have that ability. 16 You have to, when you're setting this up, from a 17 purely regulatory perspective, however, balance that with 18 what does that do with your reporting lines and what are you 19 not seeing. What is being, you know, traded opaquely that 20 you're not seeing the data on. If certain benefits are given 21 at two different prices, there is going to be a reason why. 22 And I think the businesses have the rights, SEFs have the 23 right to charge two different prices. From a regulatory 24 perspective, however, you have to ask am I giving up some 25 information here and what am I not seeing. 139 1 MR. COOK: Ok. Thank you. So we've reached the end 2 of our time. I think we should bring this to a close. Rick any 3 final comments you want to make? 4 MR. SHILTS: I wish to thank everyone for the great 5 discussion. 6 MR. COOK: I want to thank everyone for your 7 participation today. I just remind everyone that we do have 8 open mailboxes on our website so anyone who has an interest 9 in these topics please submit your thoughts in writing. We 10 would be -- look forward to reading them. Thank you. 11 (Whereupon, at 12:30 p.m., the roundtable was concluded.) 12 * * * * * 13 14 15 16 17 18 19 20 21 22 23 24 25