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