0001
 1           U.S. SECURITIES AND EXCHANGE COMMISSION 
 2   
 3   
 4   
 5                 MEETING OF THE FIXED INCOME 
 6             MARKET STRUCTURE ADVISORY COMMITTEE 
 7                                
 8   
 9   
10   
11   
12                    Monday, July 16, 2018 
13                          9:29 a.m. 
14   
15   
16   
17   
18   
19   
20   
21              Securities and Exchange Commission 
22                   Multipurpose Room LL-006 
23                      100 F Street, N.E. 
24                   Washington, D.C.  20549 
25   
0002
 1   PARTICIPANTS:
 2   
 3   COMMISSIONERS:     
 4   Commissioner Kara M. Stein  
 5   Commissioner Hester M. Peirce 
 6   COMMITTEE MEMBERS: 
 7   Michael Heaney, Chairman 
 8   Brett Redfearn, DFO 
 9   Matthew Andresen 
10   Brian Archer 
11   John Bagley 
12   Gilbert Garcia 
13   Tom Gira 
14   Larry Harris 
15   Amar Kuchinad 
16   Lynn Martin 
17   Richard McVey 
18   Suzanne Shank 
19   Larry Tabb 
20   Kumar Venkataraman 
21   Elisse Walter 
22   Rachel Wilson 
23   Mihir Worah 
24    
25   
0003
 1   GUESTS: 
 2   Horace Carter, Raymond James  
 3   Bernard Costello, Morgan Stanley  
 4   Ric Edelman, Edelman Financial Services  
 5   Chris Ferreri, Hartfield, Titus and Donnelly 
 6   Doug Friedman, Tradeweb  
 7   Neil Hamburger, JPMorgan Chase  
 8   Jon Klein, Bank of America Merrill Lynch  
 9   Ben Macdonald, Bloomberg  
10   Amy McGarrity, Investment Advisory Committee  
11   Rick McVey, MarketAxess  
12   Miranda Morad, MarketAxess 
13   Tim Morbelli, Alliance Bernstein 
14   Alex Sedgwick, T. Rowe Price 
15   Steve Shaw, BondSavvy 
16   Ben Smelser, Breckinridge Capital Advisors  
17   Michael Surowiecki, PIMCO 
18   David Umeda, Charles Schwab  
19   Thomas Urano, Sage Advisory Services  
20   Ron Valinoti, Triangle Park Capital Markets Data  
21   Bill Vulpis, ICE BondPoint  
22   Chris White, Viable Markets and BondCliQ 
23   Mark Yallop, FICC Markets Standards Board  
24    
25   
0004
 1   STAFF: 
 2   Chyhe Becker 
 3   Ben Bernstein 
 4   David Dimitrious 
 5   Tom Eady  
 6   Amy Edwards 
 7   Abby Kim  
 8   Rebecca Olson 
 9   John Roeser 
10   David Shillman 
11    
12    
13    
14    
15    
16    
17    
18    
19    
20    
21    
22    
23    
24    
25    
0005
 1                        C O N T E N T S 
 2                                                       PAGE 
 3   OPENING REMARKS                                        7 
 4             Commissioner Stein  
 5             Commissioner Peirce  
 6             Brett Redfearn, Director, Division of  
 7             Trading and Markets  
 8             Michael Heaney, Committee Chairman  
 9   PANEL I:  PRE-TRADE TRANSPARENCY UNDER MiFID II       18 
10             Mihir Worah, Moderator 
11             Neil Hamburger, JPMorgan Chase  
12             Miranda Morad, MarketAxess 
13             Michael Surowiecki, PIMCO 
14             Mark Yallop, FICC Markets Standards Board  
15   BREAK                                                 81 
16   PANEL II: CURRENT STATE OF PRE-TRADE TRANSPARENCY  
17   IN THE U.S. CORPORATE BOND MARKET                     81 
18             Mihir Worah, Moderator 
19             Jon Klein, Bank of America Merrill Lynch  
20             Ben Macdonald, Bloomberg  
21             Tim Morbelli, AllianceBernstein 
22             Steve Shaw, BondSavvy  
23             Thomas Urano, Sage Advisory Services  
24             Chris White, Viable Markets and BondCliq  
25   LUNCH BREAK                                         153
0006
 1                 C O N T E N T S (CONT.) 
 2   PANEL III: CURRENT STATE OF PRE-TRADE TRANSPARENCY  
 3   IN THE U.S. MUNICIPAL SECURITIES MARKET            158 
 4             Lynn Martin, Moderator  
 5             Bernard Costello, Morgan Stanley  
 6             Ric Edelman, Edelman Financial Services  
 7             Chris Ferreri, Hartfield, Titus and Donnelly 
 8             Ben Smelser, Breckinridge Capital Advisors  
 9             David Umeda, Charles Schwab 
10             Ron Valinoti, Triangle Park Capital Markets 
11   Data  
12   BREAK                                              210 
13   PANEL IV: ELECTRONIC TRADING VENUE REGULATION --  
14   DRAFT RECOMMENDATIONS                              210 
15             Brett Redfearn, Moderator  
16             Horace Carter, Raymond James  
17             Doug Friedman, Tradeweb 
18             Ben Macdonald, Bloomberg  
19             Rick McVey, MarketAxess 
20             Alex Sedgwick, T. Rowe Price  
21             Bill Vulpis, ICE BondPoint  
22   ETFs and Bond Funds Subcommittee Update           254 
23   Adjournment                                       262    
24          
25    
0007
 1                     P R O C E E D I N G S 
 2             MR. HEANEY:  Good morning.  I believe we have a 
 3   quorum, so I would like to call the meeting to order.  
 4   Thank you for joining us for the now third meeting of the 
 5   SEC Fixed Income Market Structure Advisory Committee.  
 6   And in addition to the members that we have in the room, 
 7   I believe we have Elisse Walter and Matt Andresen on the 
 8   phone.  Good morning, Matt.  Good morning, Elise.  
 9             MR. ANDRESEN:  Good morning.  
10             MS. WALTER:  Good morning.  
11             MR. HEANEY:  I'd also like to note that one of 
12   our esteemed colleagues, Tom Thees, is no longer a member 
13   of the committee as he recently left the firm that he was 
14   representing to move on to another venture outside the 
15   industry.  
16             I'm sure I'm not alone when I'd like to express 
17   our gratitude for the hard work that he put into this 
18   committee, providing important insights into the 
19   operation and structure of the fixed income markets that 
20   we continue to analyze and be considered by the committee 
21   today.  
22             I would like to now move it over to our 
23   introductory remarks. Unfortunately, Chairman Clayton and 
24   Commissioner Jackson will not be here this morning.  
25   Chairman Clayton will try to attend this afternoon's 
0008
 1   sessions.  But I'll begin by welcoming Commissioner Stein 
 2   in person, and asking Commissioner Stein to make her 
 3   opening remarks.  
 4             COMMISSIONER STEIN:  Welcome to everyone.  I 
 5   have very brief remarks.  It's mainly to thank you again 
 6   for this pro bono activity, giving us your best thoughts 
 7   on some of these important topics.  We have, I think, a 
 8   really interesting day ahead of us, three panels on pre-
 9   trade transparency and a draft recommendation related to 
10   electronic trading venue regulation.   
11             But all of the panels are addressing a very 
12   important topic, transparency.  And as I think everyone 
13   in this room is aware, I think transparency is the grease 
14   that keeps the machinery of trading moving.  So I'm 
15   extremely interested in people's best thoughts on how we 
16   can best facilitate transparency in the fixed income 
17   markets.  
18             I'm also interested in hearing everyone's 
19   thoughts on the Technology and Electronic Trading 
20   Subcommittee's draft recommendation on oversight of 
21   electronic trading platforms for corporate and municipal 
22   bonds.  In particular, I'm interested in your views on 
23   the appropriate oversight of this important marketplace. 
24    How should it be regulated?  And is there a need for 
25   differentiated regulation?  
0009
 1             So I look forward to today's discussion.  And 
 2   once again, I want to thank everyone for coming to 
 3   today's meeting.  
 4             MR. HEANEY:  Thank you, Commissioner Stein.  
 5             Commissioner Peirce?  
 6             COMMISSIONER PEIRCE:  Thank you, and thank you 
 7   all for being here today.  Again, I very much appreciate 
 8   your willingness to serve on this committee.  It's very 
 9   valuable for us to hear your perspectives, and I'm 
10   looking forward to all of the panels today.  I may not be 
11   able to be here for all of them, but I will watch after 
12   the fact.  So I look forward to that.  
13             And I'm especially looking forward to having a 
14   discussion on municipal securities, which I think will be 
15   very valuable for us.  Thanks very much.  
16             MR. HEANEY:  Thank you, Commissioner Peirce.  
17             Next I'd like to turn it over to Brett 
18   Redfearn, Director of Division of Trading and Markets and 
19   the committee's Designated Federal Official, for his 
20   opening remarks.  
21             MR. REDFEARN:  Thank you, Michael.  I'd like to 
22   welcome everyone to our third FIMSAC meeting.  Let me 
23   briefly introduce my colleagues sitting here with me 
24   today.  
25             From the Division of Trading and Markets, to my 
0010
 1   right we have Dave Shillman and John Roeser, both 
 2   associate directors in the Office of Market Supervision; 
 3   Tom Eady, senior policy advisor in our Office of 
 4   Analytics and Research; David Dimitrious, senior special 
 5   counsel; and Ben Bernstein, attorney advisor in the 
 6   Office of Market Supervision.  
 7             Also joining us on this side are Rebecca Olsen, 
 8   acting director of the Office of Municipal Securities; 
 9   and Amy is here somewhere -- is Amy down there as well?  
10   Amy Edwards, assistant director of the Division of 
11   Economic and Risk Analysis.  
12             Before we get started, as we always say, any 
13   views expressed by the Staff in this forum are ours alone 
14   and cannot be attributed to the Commission or the 
15   Commissioners.  
16             So transparency topics have been a key focus of 
17   FIMSAC so far.  At the last meeting, this group approved 
18   a recommendation related to post-trade transparency in 
19   the block market, which we are actively considering in 
20   cooperation with FINRA.  Today we will be spending a 
21   large portion of the day discussing another important 
22   transparency subject, the availability of pre-trade price 
23   information.  
24             In the context of post-trade transparency block 
25   pilot discussion at the last FIMSAC meeting, I noted that 
0011
 1   striking the right balance is important.  Transparency 
 2   can promote efficient markets through lower search costs 
 3   and greater price competition.  
 4             Arguably, it can also impair liquidity and 
 5   market quality in certain segments if it increases market 
 6   impact and risk in the provision of capital.  Striking 
 7   this balance is equally important when it comes to 
 8   considering the appropriate availability of pre-trade 
 9   price information in our corporate bond and municipal 
10   securities markets.   
11             Pre-trade transparency was a prominent topic of 
12   discussion at the FIMSAC's inaugural meeting.  It was 
13   discussed by one participant, who urged the Commission to 
14   give it due consideration.  A large buy side firm was 
15   described as "the holy grail for the marketplace that 
16   could improve the market's efficiency and incent new 
17   capital into the market."  
18             It was also suggested that it could also 
19   improve market participant confidence to trade bonds that 
20   they might otherwise be hesitant to trade, and increase 
21   the velocity of trading and improve liquidity on 
22   electronic platforms.  
23             This market participant described the levels of 
24   pre-trade information available to it, and recognized 
25   both the technological challenge in collecting and making 
0012
 1   sense of that significant amount of information and the 
 2   critical importance of doing so to succeed in today's 
 3   fixed income markets.  
 4             The same market participant recognized the 
 5   challenge for the buy side to change from the traditional 
 6   role as price takers to price makers, and noted that 
 7   enhanced pre-trade transparency can assist in that 
 8   transition by providing more confidence about the levels 
 9   bonds should trade.  
10             Another market participant, a large sell side 
11   firm, also recognized the importance of transparency, 
12   including pre-trade transparency, and the goal of getting 
13   more information to the marketplace.  At the same time, 
14   that sell side firm cautioned that not all bonds are 
15   created equal and that what might be good for some bonds 
16   might not be good for other bonds.  
17             One large buy side firm also supported pre-
18   transparency, but advised that such transparency be 
19   anchored in transactable or firm prices only.  On the 
20   other hand, another large buy side firm suggested that 
21   pre-trade information should not be limited to 
22   transactable quotes, only drawing a parallel to NASDAQ in 
23   the early days when it streamed quotes that were not 
24   executable.  
25             We've already heard a diverse set of views, and 
0013
 1   today's panels will provide an opportunity to begin a 
 2   deeper dive into this important area of focus by 
 3   exploring the current state of pre-transparency in the 
 4   United States and the European Union.   
 5             We will begin today considering the approach to 
 6   pre-trade transparency pursued in the European Union 
 7   through MiFID II.  Recognizing that the European Union 
 8   bond market may be different than the U.S. corporate bond 
 9   and municipal securities markets in significant aspects 
10   of its operation and structure, the MiFID experience 
11   could provide some insights into how one jurisdiction has 
12   approached pre-trade transparency in bond markets.  
13             After a discussion of MiFID, the committee will 
14   hear from panels of experts on the current state of pre-
15   trade transparency in both the corporate bond and 
16   municipal securities markets.  
17             I look forward to benefitting from this 
18   committee's thoughts on how we can strike the right 
19   balance on pre-trade transparency in an effort to improve 
20   the fairness and efficiency of our markets for all 
21   investors, including, very importantly, retail investors.  
22             The corporate bond and municipal securities 
23   markets continue to benefit from increased technological 
24   advancement.  It is worth this committee's effort to 
25   consider whether these technologies are being 
0014
 1   sufficiently leveraged to provide pre-trade pricing 
 2   information in a manner that optimally serves our markets 
 3   and all investors.  
 4             As a further reflection of the increasing 
 5   impact of technology on our fixed income markets, later 
 6   this afternoon we will consider the Technology and 
 7   Electronic Trading Subcommittee's preliminary 
 8   recommendation on the oversight of electronic trading 
 9   platforms.   
10             The subcommittee's preliminary recommendation 
11   today proposes to create a joint working group of the 
12   SEC, FINRA, and the MSRB to review the regulatory 
13   framework for fixed income electronic trading platforms. 
14    I look forward to the full committee's consideration of 
15   the preliminary recommendation, and any efforts to 
16   improve the regulation of fixed income trading will 
17   benefit from this committee's guidance.  
18             Thank you all again for devoting so much of 
19   your time to this committee.  I look forward to today's 
20   discussion.  So now I'm turning it back over to you, 
21   Michael.  
22             MR. HEANEY:  Thank you.  Thanks, Brett.  And 
23   thank you again for joining us for the third meeting of 
24   FIMSAC.  I would like to take this opportunity to thank 
25   all the panelists who have made their way down to 
0015
 1   Washington, many from outside the city itself, but a 
 2   large portion who came in from London.  So thank you very 
 3   much.  
 4             We do have a full agenda, so I'll try to be 
 5   brief as well and recap what the day will look like.  
 6   You'll hear updates of the working priorities of now the 
 7   four subcommittees.  As you know, following our January 
 8   meeting, we formed three subcommittees, Transparency, ETF 
 9   and Bond Funds, Technology and Electronic Trading.   
10             But based on several months of experience with 
11   FIMSAC and the work of the Transparency Subcommittee, 
12   Brett and I decided to replace the Transparency 
13   Subcommittee with two subcommittees, Corporate Bond 
14   Transparency Subcommittee and the Municipal Securities 
15   Transparency Subcommittee.  
16             At the last meeting, the committee approved the 
17   recommendation from the Transparency Subcommittee to 
18   conduct a pilot program for block-sized trades in 
19   corporate bonds.  As you recall, the Transparency 
20   Subcommittee did not propose including municipal 
21   securities in its recommendation because we considered 
22   the considerations of the uniqueness of that market.  
23             These considerations suggested that perhaps the 
24   corporate bond and municipal securities markets may 
25   warrant different approaches to the transparency 
0016
 1   policies.  We therefore felt that FIMSAC would benefit 
 2   from having two separate subcommittees, each devoted to a 
 3   particular asset class.  
 4             Other than the focus on corporate bonds and 
 5   municipal securities exclusively, the two subcommittees' 
 6   mandates are the same as that prior of the original 
 7   Transparency Subcommittee.  
 8             I want to thank Mihir again for his continued 
 9   willingness to serve as the chair of the Corporate Bond 
10   Transparency Subcommittee, and I'd to thank Lynn Martin 
11   for serving as the chair of the Municipal Securities 
12   Transparency Subcommittee.  
13             To continue our focus on transparency issues, 
14   we will be devoting much of our day to the consideration 
15   of the current state of play of pre-trade transparency, 
16   as Brett alluded to, not only in the U.S. but in Europe 
17   as well.  These level-setting panels should inform us all 
18   and help the subcommittees consider their agendas as we 
19   go forward over the next two to three months.  
20             For our first panel, we'll hear from the market 
21   participants about the approach to pre-trade transparency 
22   for fixed income securities in the European Union, as 
23   required by MiFID II.  The panelists will explain MiFID's 
24   approach to pre-trade transparency, discuss the 
25   implementation process, and offer thoughts on what the 
0017
 1   impact has been to the market after the first six months.  
 2             Our next two panels will discuss the current 
 3   state of pre-trade transparency in the U.S. corporate 
 4   bond and municipal securities markets.  During these 
 5   panels we will hear from a variety of market participants 
 6   about what pre-trade price information is available 
 7   today, the access to pre-trade information by both retail 
 8   and institutional investors, and the methods market 
 9   participants are using to transmit, receive, aggregate, 
10   and basically use this essential market information.   
11             Our first panel will consider the U.S. 
12   corporate bond market.  After lunch we'll return for the 
13   municipal securities market discussion.   
14             Following a short afternoon break, our 
15   discussion will turn towards consideration of the 
16   Technology and Electronic Trading Subcommittee's 
17   preliminary recommendation to review the oversight 
18   framework of electronic trading platforms for corporate 
19   bonds and municipal securities.  We'll then conclude the 
20   day by hearing from the ETF and Bond Funds Subcommittee 
21   on the status of the work and the topics they've been 
22   considering going forward.   
23             I would just like again to take the opportunity 
24   off, script, and thank the members of FIMSAC.  This has 
25   been an intense last two months, with most committees 
0018
 1   meeting or having conference calls every week for the 
 2   last two months.  
 3             So there's been a lot of hard work, a lot of 
 4   energy, a lot of enthusiasm.  And I will say to 
 5   collectively work together for this has been fairly 
 6   remarkable to watch.  So again, thank you.  
 7             I just would remind everybody as we go forward 
 8   during the discussions, if you have a question again, 
 9   just to raise your name plate so we can keep this 
10   somewhat orderly.  And I'm sure there'll be quite a 
11   robust and interesting dialogue in Q&A period.  
12             So now let's turn to our on pre-trade 
13   transparency, and I'll turn it over to Mihir, the chair 
14   of the Corporate Bond Transparency Subcommittee, to 
15   introduce our panelists and moderate the discussion of 
16   the MiFID II approach.  
17   PANEL I:  PRE-TRADE TRANSPARENCY UNDER MIFID II 
18             MR. WORAH:  Thanks, Michael.  So again, I don't 
19   want to repeat a lot of what Brett and Michael said.  But 
20   the Corporate Bond Transparency or liquidity Subcommittee 
21   has started focusing on pre-trade transparency.  
22             You look at markets with executable prices 
23   available to all -- the equity markets, the Treasury 
24   futures market, depth of markets available to most 
25   participants.  And the goal is:  Can the corporate bond 
0019
 1   market get there?  What are the costs?  What are the 
 2   benefits?   
 3             Should we allow market solutions?  We're 
 4   already getting there with a lot of the electronic 
 5   trading platforms.  Or should there be some regulatory 
 6   nudge to getting corporate bond markets there?  So that's 
 7   how I see the goal of our subcommittee -- come up with 
 8   the recommendation on what, if anything, can we do to 
 9   improve pre-trade transparency in the corporate bond 
10   markets.   
11             To get there we decided to level set, A, with 
12   looking at the European experience, MiFID II attempts to 
13   open up and enforce a certain pre-trade transparency 
14   regime in European markets.  So we're fortunate to have a 
15   number of esteemed industry participants come visit us 
16   from over the pond.  
17             The good news is many of them are from London 
18   and they didn't have to miss an England World Cup final, 
19   so flying here wasn't too bad.  And one of them is from 
20   Germany, and he was ready to come anyway pretty early on.  
21             (Laughter.) 
22             MR. WORAH:  But thanks for coming.  So we have 
23   a panel talking about MiFID.  And then we've got a second 
24   panel following this.  There'll be a short break, and 
25   following that we'll have a panel talking about the 
0020
 1   experience in the U.S. -- where we are the state of the 
 2   art, where we are, where we hope to be.  What are some of 
 3   the industry solutions?  What's the experience from sell 
 4   side, buy side, retail market participants in the U.S.?  
 5             So we hope this panel sheds some light in terms 
 6   of what's happening in Europe.  What were the thoughts 
 7   going into the design of the pre-trade transparency 
 8   regime?  And what are some of the early outcomes?  And 
 9   what can we learn?  
10             So participating in this panel, we have, 
11   starting from the top of the room, Neil Hamburger from 
12   JPMorgan; Miranda Morad from MarketAxess; Michael 
13   Surowiecki from PIMCO; and Mark Yallop, who's on the FICC 
14   Market Standards Board.  So we hope a wide array of 
15   views, and we hope to have a pretty active discussion, 
16   and hope that you folks can educate us on the European 
17   experience.  
18             So I think we start with Mark.  So Mark -- all 
19   the participants -- why don't you introduce yourself 
20   briefly, what you do and how that colors your outlook?  
21             MR. YALLOP:  Certainly.  Thank you.  So I've 
22   spent 35 years in the private sector, working in and 
23   running trading and sales businesses for large sell side 
24   firms and for, at the time, the largest inter-dollar 
25   broker in the world.   
0021
 1             I now sit on the Prudential Regulatory 
 2   Committee at the Bank of England, and I chair the entity, 
 3   the FICC Markets Standards Board.  So I should say just 
 4   up front that I'm not speaking here on behalf of the Bank 
 5   of England.  What I say is my personal views, not those 
 6   of the bank.  
 7             What I thought I would do, just to kick the 
 8   discussion off, is to talk about two bigger, more 
 9   strategic questions.  First of all, what was it that the 
10   Europeans thought they were trying to do when they 
11   conceived the idea of MiFID?  And secondly, conceptually, 
12   was MiFID II the right way to approach the question of 
13   how to improve liquidity in the wholesale bond markets, 
14   or could other things have been done?  
15             So starting with the first of those questions, 
16   you may or may not be aware that MiFID II, which along 
17   with another acronym, EMEA, is the European realization 
18   of the 2009 G20 declaration back in Philadelphia that I'm 
19   sure we're all familiar with.  So it's taken 7 years from 
20   that time, when the Europeans took up the matter 
21   seriously in 2011, up to the beginning of 2018, this 
22   year, to develop and introduce MiFID II.  
23             At the outset, I think it's also worth 
24   remembering that European securities markets are 
25   significantly less well-developed than the markets that 
0022
 1   you're familiar with here in the U.S.  Just three or four 
 2   statistics:   
 3             European corporates take, in relative terms, 
 4   five times more bank debt than they do securities for 
 5   financing their businesses than U.S. corporations do.  
 6   Household investment portfolios in Europe have, in 
 7   relative terms, half the volume of securities in their 
 8   assets than the U.S. counter parties have.  And European 
 9   investors pay, in absolute terms, about eight times what 
10   their U.S. peers do to access real time securities price 
11   data.  So there are really big differences in the 
12   starting points across the Atlantic.  
13             MiFID II was intended -- is intended -- to have 
14   a very wide-ranging impact.  Some would argue perhaps too 
15   ambitious.  It is the biggest change in a generation in 
16   market structure and in trading venues.  It extends the 
17   previous MiFID transparency regime, which was just for 
18   equities.  It creates for the first time transparency 
19   regimes for non-equities, particularly fixed income and 
20   commodities.   
21             It extends the regulatory regime for platforms. 
22    It creates new platforms, and brings high-frequency 
23   traders into regulation-capped dark pool trading, and 
24   does a number of other things that I'm sure we'll talk 
25   about in the course of this morning.  
0023
 1             It unbundles research to address conflicts of 
 2   interest.  In that field, it improves investor protection 
 3   by targeting price transparency, best execution, dealing 
 4   commissions, and a host of other aspects of trading in 
 5   markets.  
 6             And it's a significant step towards trying to 
 7   harmonize the rules governing European securities 
 8   markets, which is an important precursor to developing a 
 9   single European capital market.  
10             Right at the outset, if you look back in 
11   history, the European Commission, when they first 
12   conceived this project, set out four objectives for MiFID 
13   II, which were, first of all, strengthening investor 
14   protection; secondly, reducing the risks of a disorderly 
15   market; thirdly, reducing systemic risk; and fourthly, 
16   increasing the efficiency of financial markets and 
17   reducing unnecessary costs for participants.  
18             So those were, in my view, entirely rational 
19   and appropriate objectives for the European Commission to 
20   set.  And I doubt there would be any equivalent bunch of 
21   people in a similar position in any other part of the 
22   world who would have perhaps expressed themselves 
23   differently.  
24             Again, if you go right back to the outset of 
25   the MiFID II project, European regulators talked about 
0024
 1   their mission in the following kinds of terms, and I'll 
 2   give you four direct quotes from European regulators at 
 3   the time back in 2011 and '12.  
 4             First of all, for example, to make European 
 5   markets "more transparent, efficient, attractive, and 
 6   safer to invest in."  Secondly, "improving market 
 7   integrity, enhancing competition, and providing further 
 8   necessary protection for consumers and market users."  
 9             Third, "The desirability of greater 
10   transparency is obvious:  Greater information, improving 
11   price discovery, and the liquidity in the markets.  MiFID 
12   II weighs in favor of more lit, on-venue, transparent 
13   trading."  
14             And they also said, "We support greater 
15   transparency in non-equity markets.  But transparency in 
16   those markets, which are often characterized by lower and 
17   episodic liquidity requires careful calibration."  So I 
18   think not only amongst the Brussels bureaucrats, but also 
19   among the practical national regulators, there were a 
20   sensible set of objectives right at the outset.  
21             Pre-trade transparency, the topic of this 
22   session, was of course one of the most debated topics at 
23   ESMA and among the national market regulators.  And they 
24   all recognized the difficulty of finding a balance 
25   between improving transparency to aid price discovery 
0025
 1   whilst ensuring at the same time that whatever rules they 
 2   came up with didn't kill off market liquidity because it 
 3   gave too much transparency.  
 4             Industry and national regulators were heavily 
 5   divided on how to do this, but eventually and instrument-
 6   by-instrument approach that I'm sure my colleagues here 
 7   will talk about in a minute was developed.  And a system 
 8   of waivers and phasing-in of reporting for larger-sized, 
 9   less frequently traded, less liquid securities.   
10             And the deferral of the publication of post-
11   trade details on transactions was decided upon as the 
12   means of implementation.  Those limitations on 
13   publication are popular with the sell side firms in the 
14   market, but they're less so with the buy side.   
15             Perhaps finally on this question, I might just 
16   point out that when the pre-trade price transparency 
17   system was devised, it was expected, based on some 
18   analysis that ESMA did, that about 4 percent of all the 
19   bonds in issue in Europe, about 2,000 different issues, 
20   2- to 2-1/2 thousand different issues -- most of which 
21   would, by the way, be sovereigns -- would eventually be 
22   captured.  
23             But the phasing-in of reporting obligations 
24   means that a much smaller proportion of being captured at 
25   the moment, perhaps only a couple of hundred or about 
0026
 1   half a percent of all of the bonds that are being traded 
 2   in Europe.   
 3             As you can imagine, the fact that 95 percent of 
 4   trades are unlikely ever to be captured has caused quite 
 5   a few people to question the value of the transparency 
 6   parts of MiFID II, a question, I'm sure, we'll come back 
 7   to later.  
 8             Let me just move on quickly then to my second 
 9   observation, which is about whether MiFID II regulations 
10   were conceptually the right approach to improving 
11   liquidity.  And to start off with, a philosophical point 
12   about the regulation of wholesale markets.  
13             MiFID II, if you look at it in its entirety, is 
14   about 1.7 million paragraphs of new regulation.  Some 
15   firms have spent over $100 million in preparing for the 
16   implementation of MiFID II.  This detailed and perceptive 
17   approach of MiFID II is one valid approach to take, and 
18   it is typical of what has happened in Europe.  But it 
19   isn't the only way to go about regulating securities 
20   markets.  
21             I would just point out that experience shows 
22   that perceptive regulation invites regulatory arbitrage 
23   and often some unintended consequences.  It is also hard 
24   for securities regulators, even those who have extensive 
25   resources, to keep pace with new product and market 
0027
 1   investigation by a highly motivated and well-resourced 
 2   private sector.  
 3             Third, well-functioning markets need 
 4   discipline.  But market discipline needs to be actively 
 5   supported by market participants and not just imposed on 
 6   them by regulators.  
 7             And fourth, trust in financial markets, which 
 8   has taken a significant hit, is unlikely to be restored, 
 9   in my view, simply by tougher regulation.  Market 
10   participant behavior has to change, and to be convincing, 
11   that change has to be driven by the participants 
12   themselves, not just imposed by regulators.  
13             There is another possible approach to improving 
14   comfort in markets if that is the underlying aim here, 
15   which the FMSB, FICC Market Standards Board, is 
16   championing following the U.K.'s fair and effective 
17   markets review.   
18             And that is to get market participants, all of 
19   them who have an interest in wholesale markets, the 
20   issuers and the investors, market makers and price 
21   takers, data providers and infrastructure firms, to 
22   develop practical standards that lay out how business 
23   should be conducted and for those standards to be adhered 
24   to and tested in the public domain.   
25             It's harder to do this in retail markets.  But 
0028
 1   the concentrated nature of wholesale fixed income markets 
 2   makes it possible, I believe, to get all interested 
 3   parties around the table to debate appropriate business 
 4   practice.  And the private sector standards-driven 
 5   approach could, I believe, have been used to derive many 
 6   of the benefits of MiFID II at somewhat lower cost and 
 7   rather more quickly.  But -- I emphasize -- this would 
 8   have entailed the EU approaching post-crisis reforms in a 
 9   way that would have been untypical for them.  
10             I know the U.S. is traditionally in favored, 
11   detailed rulemaking as a means of regulation.  But if I 
12   have one request for you at this meeting this morning, it 
13   is that the SEC gives careful consideration to the 
14   lessons learned from the FMSB approach as well as 
15   considering more perceptive details, rule-based 
16   approaches that you've seen in the European Union and in 
17   ESMA.  
18             Finally, whilst I believe there will be 
19   significant value in pre-trade transparency provided that 
20   some of the obstacles my colleagues will talk about in a 
21   minute can be addressed.  There are, I think, a few 
22   important, practical points to bear in mind when 
23   considering the impact of MiFID II and pre-trade price 
24   transparency.  
25             First of all, as I've mentioned already, 
0029
 1   European and U.S. markets start in very different places 
 2   as regards trade transparency because of TRACE and the 
 3   work done by CFTC on derivatives reporting.  
 4             Secondly, it's obvious that bonds are a much 
 5   more heterogeneous asset class than equities, trading 
 6   perhaps 100,000th or even a 10,000th as frequently as 
 7   equity equivalents.  And this liquidity in the bond 
 8   market declines as bonds move off the run and age.  
 9   Electronic markets, on the other hand, flourish when 
10   they're used for homogeneous liquid instruments like cash 
11   equities, book foreign exchange, and on-the-run 
12   government bonds.   
13             Third, experience shows that electronic markets 
14   tend to increase the number of trades but not necessarily 
15   the true liquidity in markets because of the way in which 
16   they facilitate algorithmic trading and smaller order 
17   sizes.  Apparent liquidity in peacetime, favorable market 
18   conditions, is not always available in stressed, wartime 
19   conditions.  
20             Fourth, electronic trading can fragment 
21   liquidity.  And while fragmented markets may show tight 
22   bid-ask spreads for execution, they may not offer the 
23   best overall total cost of execution when post-trade 
24   processing and infrastructure costs are factored in as 
25   well.  
0030
 1             Fifth, electronic markets not only permit 
 2   traditional forms of misconduct to flourish, but they 
 3   also create options for new forms of misconduct to be 
 4   practices.  And as we push more markets electronic, I 
 5   think this needs to be borne in mind.  
 6             And lastly, in bond markets, I think much more 
 7   and in equities, liquidity is also a function of other 
 8   parts of the fixed income value chain -- for example, 
 9   what's happening in repo markets, which needs to be 
10   factored into this debate as well.  
11             I should emphasize finally that I speak as an 
12   advocate of electronic trading for many, many years, as 
13   those in the room who know me will attest.  It can bring 
14   huge benefits.  My point in making these slightly gloomy 
15   observations is only the vulnerabilities of electronic 
16   markets don't always receive enough attention in the rush 
17   to promote them.  Thank you very much.  
18             MR. HEANEY:  Thanks, Mark.  And I think we save 
19   questions, as far as possible, till all the panelists are 
20   done.  
21             So Michael, why don't you introduce yourself 
22   and walk through your presentation.  
23             MR. SUROWIECKI:  Thank you, Mihir.  And good 
24   morning, everyone.  My name is Michael Surowiecki, and I 
25   work as a portfolio manager for European fixed income at 
0031
 1   PIMCO in Munich, Germany.  
 2             Let me give you a description of what MiFID II 
 3   looks like.  I haven't done it myself, but it is said 
 4   that depending on how much of the supporting legislation 
 5   you pull in, you're going to be printing anywhere between 
 6   15- to 30,000 pages of documents for MiFID.  That's a 
 7   stack that reaches almost 10 foot high.  It reaches the 
 8   ceiling of this room.   
 9             By comparison, the Dodd-Frank Act, I believe, 
10   has only 3,000 pages; the Bible, 1200; and so in terms of 
11   relative size, MiFID is an order of magnitude bigger.  
12   And also, in true European spirit, the entire law has 
13   been translated into 23 European languages.   
14             Now, with all this complexity, I hope you may 
15   appreciate two things:  First of all, how reasonably 
16   challenging it will be for me to give you a summary in 
17   the next five minutes.  But also, more importantly, the 
18   time, cost, and effort spent by the financial industry 
19   when it came to implementing MiFID II.  
20             There are many different numbers floating 
21   around, but the one I hear about most frequently is that 
22   collectively, the industry spent around 2.5 billion U.S. 
23   dollars on implementing MiFID II.  In fact, one large 
24   European asset manager -- not us -- has claimed to have 
25   assigned almost 300 employees towards the project of 
0032
 1   MiFID implementation.  And so perhaps one thing to keep 
 2   in mind is that the burden of such extensive legislation 
 3   is perhaps most heavily felt by the smaller firms and not 
 4   by the larger ones.  
 5             I'll be focusing on the transparency aspects.  
 6   I prepared a deck of slides, which I think everyone has 
 7   in front of himself.  Let's jump to page 3.  Under MiFID, 
 8   the way I look at it, there are three important concepts 
 9   that form the basis of the transparency framework.  
10             The first one is a distinction between pre- and 
11   post-trade transparency.  Post-trade transparency is 
12   nothing new.  It exists already in the U.S. in the form 
13   of TRACE and the derivatives reporting obligation under 
14   Dodd-Frank.  And under MiFID, it's really the same idea.  
15             The real innovation is in the pre-trade 
16   transparency.  And in taken to its full orthodoxy, it 
17   literally means that as quotations are made to clients 
18   for securities, that these quotes need to be disclosed to 
19   the market in realtime.  
20             Now, before I continue, I should mention that 
21   the actual implementation may differ between different 
22   asset classes.  So if you revisit this topic, you might 
23   spot little differences, depending on whether you're 
24   looking at the equities, non-equities, derivatives, non-
25   derivatives, corporate, or government bonds.  I'm 
0033
 1   skipping over all these details because I think they 
 2   distract from the general overview of what transparency 
 3   is all about.  But I wanted you to be aware of it.  
 4             Let's move on to page 4.  The second important 
 5   concept is one of liquid versus illiquid.  This makes 
 6   sense because in a way, the trading of illiquid 
 7   securities might be impacted negatively by too much 
 8   transparency.  And so the European regulator has defined 
 9   what's liquid and what's not liquid.  
10             And this is done on the security level.  So as 
11   mentioned before, there are literally thousands of 
12   securities out there, and each one of them gets a 
13   determination as to whether it's liquid or illiquid.  And 
14   so one important take-away from this is that it's 
15   probably very difficult for a human being to be keeping 
16   track of all this at the icing level, and the help of 
17   modern computer systems is greatly needed.  
18             How do those liquidity definitions work?  Well, 
19   the regulator has determined what's liquid and illiquid, 
20   and these definitions work reasonably well.  There are, 
21   however, some cases where they don't work quite so will, 
22   and I've highlighted two of them here.  
23             The first one is the screenshot in the upper 
24   right.  We're looking at a 10-year German benchmark bond, 
25   which is deliverable into the current 10-year Bund 
0034
 1   futures contract.  This is a highly liquid contract 
 2   that's equivalent to your T-note futures contract here in 
 3   the U.S.  And yet the regulator has determined this bond 
 4   to be illiquid.  
 5             The other example is the one below.  We're 
 6   looking at a government bond issue by the Republic of 
 7   Malta.  Now, Malta is an island in the Mediterranean Sea. 
 8    It's a lovely destination to go on holidays.  However, 
 9   it is not particularly well-known for the liquidity of 
10   its bond market.  And so this is surprising.  
11             Now, all cynicism aside, what's the take-away? 
12    I believe, for the purpose of what you may have in mind, 
13   their recommendation could be that perhaps if you are 
14   tasked with designing similar rules, be mindful that they 
15   might be at odds in certain cases with how the market 
16   perceives actual liquidity.  
17             Let's move to page 5.  The third important 
18   concept under MiFID relates to market infrastructure.  
19   This is important because -- think about it -- if you 
20   want to regulate trading, you first need to define the 
21   types of trading that are taking place.  
22             So the European regulator has defined two broad 
23   categories of trading.  The first one is trading that 
24   takes place on a -- call it a venue, a public place, a 
25   marketplace, where people meet to trade.  This could be 
0035
 1   an exchange, like the New York Stock Exchange, or a 
 2   trading platform, like MarketAxess or Bloomberg.  
 3             The other type of trading is the more informal, 
 4   over-the-counter trading of securities.  The reason this 
 5   is important is because MiFID makes frequent reference to 
 6   the type of trading, and is prescriptive in terms of how 
 7   transparency and other aspects should be handled, 
 8   depending on the type of trading that's taking place.  
 9             So let me walk you through an actual example of 
10   how all of this comes together.  On page 6, I'm showing 
11   you a decision tree to determine pre-trade transparency 
12   under MiFID.  The first step relates to the trading 
13   venue.  This is what I just said.  Where do you trade?  
14   Let's suppose we're trading a bond on a platform; that's 
15   in any case what we do most frequently.  So the first 
16   decision is, do you trade on a platform or do you trade 
17   outside a platform?  That determines whether the trade 
18   might potentially be in scope.  
19             If it is, then we move on to the next step, 
20   which is on page 7, which leads you to what I would call 
21   a liquidity test.  A liquidity test relates to what I 
22   mentioned earlier, whether a bond is deemed liquid or 
23   illiquid.  And it's really very simple.  If it's liquid, 
24   then the bond might be in scope for pre-trade 
25   transparency.  If it's illiquid, it will not be in scope.  
0036
 1             I've put another great example on this page.  
 2   It's a U.S. Treasury bond.  Treasuries are in scope as 
 3   well as long as they are traded in the European Union.  
 4   This bond is deliverable into your ultra-long futures 
 5   contract, and is also illiquid.  
 6             The final step on page 8 relates to the size of 
 7   your trade.  And here is where it gets a bit more 
 8   detailed.  Now, the common-sensical way to think about it 
 9   is that, basically, a large trade shouldn't be in scope 
10   for pre-trade transparency.  Only small trades should be 
11   in scope.  And so the regulator has created threshold 
12   levels at which trades are either large or small.  
13             The one threshold -- and it's one of the 
14   acronyms that I use very frequently so that's why I'm 
15   bringing it up -- the one threshold is called LIS.  The L 
16   stands for "large."  The LIS threshold is a number, let's 
17   say 5 million.  Any trade above 5 million notional would 
18   hence be deemed large, large in scale ,and it wouldn't be 
19   in scope.  
20             The other threshold is one for small trades.  
21   It's called SSTI.  This is a number, let's say one 
22   million, for example.  So any trade falling below that 
23   threshold is deemed to be a small trade.  And in that 
24   case, it is in scope for pre-trade transparency.  
25   Anything in between is medium and receives a hybrid kind 
0037
 1   of treatment, which I will skip.  
 2             With that said, there are three important 
 3   things I would like to highlight.  First of all, these 
 4   thresholds are set to increase over the coming years.  
 5   Right now they're reasonably small, and only a very small 
 6   percentage of trades are actually captured by pre-trade 
 7   transparency.  However, the thresholds are set to go up 
 8   by a factor of 3 to 5 over the coming years, thereby 
 9   greatly increasing the scope for pre-trade transparency.  
10             The second thing to mention is that there is no 
11   consideration as to the duration risk of the bonds.  So 
12   one million in a one-year bond, that's the same treatment 
13   as one million in a hundred-year bond.  So perhaps, as 
14   another suggestion, if you were to design similar rules, 
15   perhaps there is consideration that could be given as to 
16   the difference in duration between different maturity 
17   bonds.  
18             I have more slides on post-trade transparency, 
19   but it's not in scope for this panel discussion, so I 
20   will stop here.  Thank you very much.  
21             MR. HEANEY:  Thanks, Michael.  
22             Let's move to Miranda for an industry 
23   technology perspective.  Or were we going to do Neil 
24   first?  Neil, you can go.  Why don't you go?  Sorry.  I 
25   mixed up the order.  Neil for the sell side perspective.  
0038
 1             MR. HAMBURGER:  Okay.  First of all, thank you 
 2   very much to the Commission and the committee for 
 3   inviting me here to speak today.  My name is Neil 
 4   Hamburger.  I work at JPMorgan.  I've been working for 
 5   JPMorgan since I did a summer internship back in 2000.  
 6   I've been a market-maker for 11 of those years, both in 
 7   London and in New York, trading high-grade, high-yield 
 8   bonds, CDS and index.  
 9             I was also on the buy side, essentially, at 
10   JPMorgan, running the portfolio of risk generated by our 
11   correlation business.  And then for the last 5 years, 
12   I've been working in our -- it's just me -- global credit 
13   trading strategy; so essentially, an internal consultant 
14   for the credit trading management.   
15             Also, as I read the CVs of you esteemed people 
16   around the room, I feel I should mention that I'm also 
17   the chair of the board of CDSClear, which is essentially 
18   the LCH for CDS.  And I'm also on the board of the Jewish 
19   Orthodox Feminist Alliance, which essentially -- its main 
20   aim is to not be an oxymoron.  
21             (Laughter.) 
22             MR. HAMBURGER:  The precise question that I was 
23   asked to address today is:  On the evidence of what we've 
24   experienced, what can we learn in Europe that we can tell 
25   to you in the U.S. about pre-trade transparency?   
0039
 1             And I think first of all, I agree with 
 2   everybody here.  But I'd just like to say it again.  Pre-
 3   trade transparency -- I think everyone's in favor of, 
 4   writ large, meaning investors should have a good idea 
 5   ahead of trading of what's a fair value and a sensible 
 6   value for the transaction they're about to try and 
 7   undertake?  
 8             Pre-trade transparency in MiFID means Investor 
 9   One comes in to attempt a trade.  And information about 
10   that trade and market makers or other market 
11   participants' responses to that trade while that trade is 
12   in progress should be made transparent to the rest of the 
13   market.   
14             MiFID is quite clear.  It's for two purposes.  
15   One -- sorry, not for two purposes.  You have to do two 
16   things.  One, you have to make the information open for 
17   the world so that the world knows what's happening.  And 
18   two, under certain extra circumstances, you need to make 
19   the quotes available for other market participants to 
20   interact with at the same time.  Okay?  So we're talking 
21   about that specific type of pre-trade transparency.  
22             The answer to the specific question I've been 
23   asked, what's been the impact on the European markets, is 
24   honestly, I could answer very quickly:  We don't know.  
25   All right?  It's been going for six months.  You've heard 
0040
 1   of the scale of the project.  So it's not the only thing 
 2   that's happened.  So in terms of a controlled experiment, 
 3   it's obviously appalling.  There are a million things 
 4   going on at the same time in the wider marketplace and 
 5   the world, but also just the MiFID rules.  
 6             If we were to look for symptoms or for 
 7   experiences, I think one way to control would be to 
 8   compare what's happened in the European versus the U.S. 
 9   markets.  In general, the trends have been pretty similar 
10   in terms of volume, bid-offer, profitability.  One thing 
11   that's been quite different is electronification.   
12             So in the European market -- which started, I 
13   think, more electronified than the U.S., something like 
14   40 percent going through electronic venues in Europe last 
15   year versus 30-ish in the U.S. -- Europe has flipped to 
16   be more 60/40, 60 now electronic, whereas the U.S. has 
17   stayed pretty constant.  So I think for those in favor of 
18   more electronification, that's a positive.  
19             However, I would caution that I think the vast 
20   majority of that extra electronic trading is really what 
21   we would call process trades.  It's RFQs to one.  So in 
22   terms of, I think, what Brett was maybe referring to, the 
23   holy grail of electronic trading, I don't think RFQs to 
24   one is what people are considering.  
25             If I then move slightly away, then, from the 
0041
 1   specific question and into what we can see, what are 
 2   positives, what are NGVs, from the pre-trade transparency 
 3   regime, I would say certainly the positives are the aims, 
 4   as Mark mentioned earlier.  I would say one phrase you've 
 5   heard repeatedly, I think, shows one of the initial 
 6   negatives, but we can turn it into kind of a backhanded 
 7   positive, which is MiFID is split in half.   
 8             It's split into equities and non-equities.  I 
 9   think any rule which starts, we have a rule for blondes 
10   and a rule for non-blondes, I think on the one hand shows 
11   that you're aware that non-blondes exist, but I think 
12   also shows that you have some way to travel in your 
13   journey for understanding and appreciating the diversity 
14   of all the non-blondes who are out there in the world.  
15             So I would caution that MiFID seems to come 
16   very much with an idea that equities is the right model. 
17    And everything should eventually conform to the equity 
18   model.  I think the reality, as we've spoken about, is 
19   credit is very different, and even within credit there's 
20   a vast array of differences.  I think only 1 percent of 
21   the corporate bond market is as liquid as the least 
22   liquid equities.   
23             So it's really very hard, and probably 
24   inadvisable, to take too many lessons from the equity 
25   market over to the credit market.  I'd like to credit 
0042
 1   Professor Harris with that statement.  That's not in my 
 2   master's.  
 3             Through the process of engaging with the 
 4   regulators about MiFID and pre- and post-trade 
 5   transparency, a couple of things changed.  So I don't 
 6   know if you've read in detail some of the rest of 
 7   Michael's presentation.  He talks about the different 
 8   nomenclature that was introduced in MiFID about the 
 9   different venues -- multi-trading vicinities, regulated 
10   markets.  
11             And then they introduce a new concept called an 
12   SI, a systemic internalizer, which means, practically 
13   speaking, a bank or a market maker.  But the sound of it 
14   sounds very much like the regulator thinks of it as a 
15   venue where market makers essentially match buyers and 
16   sellers.  
17             I think the reality is that market makers take 
18   a lot more risk than that.  And it's the taking of risk 
19   which makes a big difference, differentiates both the 
20   credit markets from other markets, and also certain bonds 
21   from certain other bonds.  
22             The regulators did take account of that later 
23   on -- for example, in the post-trade transparency, they 
24   allowed the SIs to remain anonymous, and now it just says 
25   SI, whereas if it's traded on a venue, on an exchange, 
0043
 1   then you give the name of the venue.  So that did 
 2   improve.  
 3             The other thing I think they did very well in 
 4   Europe is the calibration.  I think while I can't tell 
 5   you today what's the impact of the pre-trade 
 6   transparency, and a lot of that is because you've heard 
 7   it's on a small number of bonds and a small number of 
 8   tickets, it is intended to increase over time.  There's a 
 9   four-year schedule.   
10             But that schedule only materializes on an 
11   evidence-based approach.  The regulators need to examine 
12   what's happened.  They need to make sure that the market 
13   hasn't been adversely affected.  And only if that's the 
14   case will they move to the next level of transparency.  
15             There's also an automatic handbrake -- or, 
16   sorry, not an automatic handbrake.  There is a regulatory 
17   handbrake built into the rules.  So if the regulator or 
18   if the competent national authorities determine that the 
19   rules are not working, they can temporarily suspend the 
20   rules.  
21             I would say, in terms of where it can be good 
22   and bad, I would say there's an interesting problem in 
23   the credit markets, which is -- I think we're talking 
24   about -- as Brett mentioned, we're talking about the 
25   retail investor.  I think the institutional investor 
0044
 1   feels fairly well-informed pre-trade in the credit 
 2   markets.   
 3             And I think again, if you go to page 3 of 
 4   Michael's presentation, on the top right he has a lovely 
 5   table showing a Bloomberg screen, which has on it three 
 6   squiggly lines, which are pre-MiFID.  So these are -- 
 7   this is budget's BGN.   
 8             So this is the constantly evaluated price 
 9   stream.  And there are other vendors who have the same 
10   market access, have EVB.  And these streams are available 
11   to clients.  So pre-MiFID, clients would have thought, 
12   and the market was somewhere between the red line and the 
13   green line.  
14             In the U.S., you already have the white dots.  
15   These are the trades -- this is our European TRACE 
16   equivalent -- and the pre-trade of the new diamonds, the 
17   new brown diamonds.   
18             It seems to me that the white dots are already 
19   taking place around the white line, around the mid, and 
20   within the bid-offer.  And therefore, even pre-MiFID, 
21   there was a very good amount of pre-trade transparency.  
22   Giving trade information I think is useful.  And 
23   certainly in the States I think it's been useful to 
24   investors in general.  There's obviously lots of 
25   questions that I know you guys have been struggling -- 
0045
 1   have been thinking about for a long time.  
 2             But my question is:  If you already have the 
 3   squiggly lines and the white dots, how much extra 
 4   information do you get from the brown diamonds?  And I 
 5   mean that very seriously because the white dots, the 
 6   trades, have their own quality control inbuilt.   
 7             If a trader gives a poor price or an incorrect 
 8   price or a too-good price, it's very difficult to look at 
 9   the tape and know what was going on.  Was the primary 
10   trader in the bathroom and the junior made a price?  Was 
11   the primary trader on holiday?  Did they really want to 
12   do that trade for some reason that's unrelated to 
13   anything related to that instrument?   
14             Obviously, within the pre-trade transparency -- 
15   as we've spoken about it, there's lots of issues about 
16   speed.  You move towards whoever can act on the pre-trade 
17   fastest, getting the quote.  So what happens to the 
18   client who initiated the interaction?  How should a 
19   dollar price the trade?   
20             Should they think about the possibility of more 
21   than one client trading on the price?  And if not, then 
22   there would be a disincentive to initiating a trade, 
23   which is another problem for me.  Also, can you wait 15 
24   minutes from the trades to see the post-trade, to see the 
25   white dot?   
0046
 1             I think, in general, I'm a big fan of the way 
 2   the market has developed in the last few years.  I think 
 3   around the room we have good representatives from the 
 4   plethora of new electronic venues, electronic ways to 
 5   execute.  We have Trumid.  We have MarketAxess.  We have 
 6   Bond Sessions, BondPoint.  We have lots of places where 
 7   market participants can interact, where central limit 
 8   order books where they're appropriate, where investors 
 9   who want to use them can develop.  
10             I'm very cautious of markets being forced down 
11   any one avenue, which may or may not be appropriate for 
12   different instruments at different times, as I think has 
13   been mentioned before.  
14             So just to finish, I would say myself, but also 
15   JPMorgan, we're in favor of encouraging an organic, 
16   diverse ecosystem within the markets rather than 
17   mandating one specific direction.  
18             MR. HEANEY:  Thanks, Neil.  
19             Miranda?  
20             MS. MORAD:  Thank you.  I'd like to thank the 
21   Commission and the committee for inviting me today.  My 
22   name is Miranda Morad.  I'm the European general counsel 
23   for MarketAxess.  MarketAxess is a global electronic 
24   trading platform which specializes in RFQ in credit.   
25             And also it has a subsidiary in the U.K. called 
0047
 1   Trax, which has access to one of the best data sets, 
 2   certainly pre-MiFID II for trading in fixed income in 
 3   Europe, and post-MiFID II, after ESMA still has one of 
 4   the best data sets and we were heavily involved in 
 5   advising and helping ESMA to calculate the liquidity 
 6   regime.  
 7             So with that in mind, thank you to my fellow 
 8   panelists for setting the scene.  Some of what I say will 
 9   expand on some of those observations, which will 
10   hopefully be helpful to you, and perhaps some new things 
11   from the perspective of an electronic trading platform in 
12   Europe.  
13             I think a good place to start is a recent 
14   speech by Steven Maijoor, the chair of ESMA.  In June, he 
15   reflected to FESE, the Federation of European Exchanges, 
16   on whether or not MiFID had achieved its aims.  And I was 
17   particularly interested when preparing for this session 
18   as to his thoughts as to whether or not MiFID had 
19   achieved its aims in the context of transparency, and 
20   pre-trade transparency in particular.  
21             He made a couple of comments.  He reiterated 
22   that the particular aims of MiFID in that context were to 
23   improve transparency and price formation, and also to 
24   move transactions onto regulated platforms.  He 
25   concluded, when evaluating whether or not MiFID has 
0048
 1   achieved its aims, that there have been significant 
 2   concerns, some of which have been voiced already this 
 3   morning, that the introduction of transparency would 
 4   cause damage to the markets and impede liquidity.   
 5             And his conclusion was that that did not 
 6   materialize, and we would agree.  He also concluded that 
 7   ESMA had been extremely cautious in its calibration of 
 8   liquidity, and we would agree.  He then went on to 
 9   conclude that perhaps ESMA has been too cautious because 
10   of the paucity of data and the relative uselessness of 
11   that data.   
12             And at this point, I would stop and pay 
13   attention before jumping to agree because I think it's 
14   quite context and there is a lot of focus on simply the 
15   number of bonds that have been made transparent, and not 
16   so much focus on why there are so few data points 
17   available, and why is it so difficult to understand or 
18   make sense of that data that is available.   
19             So that's what I'm going to talk about.  Why is 
20   there so little data, and why at the moment is it of 
21   little use?  
22             I think, as a starting point, it's good to 
23   remember that Europe started from a different starting 
24   point from the USA.  When we were looking at pre-trade 
25   transparency and how to calibrate it, there was no 
0049
 1   transparency at all in Europe.  No TRACE-like equivalent, 
 2   no pre-trade transparency, no post-trade transparency.  
 3   And it was all a new thing in Europe in the non-equities 
 4   market.  
 5             So there was minimal information available to 
 6   ESMA at the design phase.  And when introducing realtime 
 7   post-trade and pre-trade information at the same time, it 
 8   was incredibly complex.  And I think there were three 
 9   reasons for that complexity.  
10             Firstly, the calibration of liquidity itself, 
11   which some of my panelists have also touched upon, and I 
12   will also give you a little bit more color.  
13             Secondly, the unique challenge of pre-trade 
14   transparency as MiFID views it, rather than as the market 
15   views it, a pre-one trade at a time.  The unique 
16   challenge of that is that you have to take into account 
17   the protocols.  And as Mark has already mentioned, if 
18   regulation is to keep pace with innovation, that's an 
19   ongoing challenge.  
20             And thirdly, the very complex system of waivers 
21   and deferrals, which Michael has also commented on, and I 
22   will give you some details.  
23             So going back to the first one, which is the 
24   liquidity calibration -- and we keep talking about the 
25   liquidity calibration here, but it is key because 
0050
 1   enshrined in the method on the concept of pre-trade 
 2   transparency was the understanding -- and I don't think I 
 3   met anybody who disagreed with this -- that pre-trade, 
 4   pre-individual trade, transparency should not be 
 5   introduced into any instrument that wasn't sufficiently 
 6   liquid to support it.  
 7             And herein lies the compare challenge because 
 8   that is the inherent brakes, the airbag, if you like, of 
 9   pre-trade transparency that is in MiFID.  And so we 
10   embarked in 2015 and 2016 of how to identify which bonds 
11   are sufficiently liquid to support transparency, and in 
12   particular, pre-trade transparency.  
13             It has been mentioned few times by my panelists 
14   that what was concluded was an instrument-by-instrument 
15   approach.  In fact, it's a combination of two approaches. 
16    It's impossible, of course, to take an instrument-by-
17   instrument approach, which is dependent upon historical 
18   activity for new issuances.  
19             So there are two approaches, a blend of two 
20   approaches, for new issuances.  The liquidity is 
21   calibrated by issue size.  And specifically for credit, 
22   any issuances that are over one billion in issue size are 
23   deemed to be liquid.  And then following about five 
24   months after issuance, you move to an instrument-by-
25   instrument approach, which is based upon the historical 
0051
 1   activity of the bond.   
 2             Now, during 2015 and '16, on behalf of -- 
 3   representing MarketAxess and Trax, we analyzed the data 
 4   because we had access to the majority of information 
 5   about what has traded in Europe in the fixed income 
 6   market.  And we told Europe, and we explained our data, 
 7   and what we explained was, it's impossible to predict 
 8   liquidity.   
 9             What instruments are going to be frequently 
10   traded in credit by reference to either issue size or 
11   historical activity.  If anything, issue size has the 
12   highest positive correlation, but it's still extremely 
13   poor.  And really, what you end up with is how tolerant 
14   are you for misclassifications?  That's really the 
15   question because nothing is a good indicator.   
16             And anybody who's a practitioner in the market 
17   will tell you, that's because credit instruments trade 
18   initially upon issuance a few days.  We're not talking 
19   three to five months.  And then they're held, and they 
20   don't trade.  And then they will trade upon sporadic 
21   events, news-driven events, that are unpredictable.  
22             So therefore, just because something traded 
23   yesterday or last week is no indication that it will 
24   trade tomorrow or next week.  And this is the difficulty. 
25    Municipal bonds aside, we have on MarketAxess almost 
0052
 1   600,000 credit instruments available to trade; only about 
 2   25,000 instruments, unique instruments, to trade in any 
 3   quarter.  And they don't necessarily trade again the 
 4   following quarter.  So the MiFID liquidity calibration 
 5   exercise is always going to be an exercise in tolerance 
 6   for misclassifications.  
 7             Secondly, the question of what is liquid, is 
 8   something liquid, depending on how often it trades a day. 
 9    MiFID has in itself a definition of what is liquid.  It 
10   says it's -- where there is a ready and willing buyer and 
11   seller on a continuous basis.  And this translated itself 
12   into, when a bond trades 15 times a day, continuously 
13   over 80 percent of the available trading days, and over a 
14   nominal amount of 100,000 per trade.  
15             Now, Steven Maijoor said that ESMA had been 
16   cautious, and they have been cautious.  And they have 
17   been cautious.  It started at 15 trades a day.  So right 
18   now, the calibration is based on something that has 
19   traded 15 times a day.  Let me tell you what that looks 
20   like.  
21             This is a sneak peek into the future because 
22   ESMA has not yet published its calibrations for Q2.  But 
23   we've already done our calculations.  The current regime, 
24   based on Q1, everybody's very upset about.  There's only 
25   220 bonds that are liquid.  And we estimate that on the 
0053
 1   basis of the Q2 trades, it should be about 290 bonds.  
 2   That's 290 bonds, which represents 64 credit instruments 
 3   out of 25,000 traded per quarter, and the rest are 
 4   sovereign instruments.   
 5             So ESMA is considering peeling back the regime. 
 6    We've talked about the fact that it will become more and 
 7   more transparent over 4 years.  Let's have a look and see 
 8   what that looks like.  And if you want to know the 
 9   equivalent volumes, at 290 bonds, 64 of which are credit, 
10   that's 1.21 percent of volumes.  If it's peeled back, the 
11   next stage after 15 trades a day goes down to 10 trades a 
12   day, and then seven, and then two on the regime that Neil 
13   said, as long as it's been proven that it won't cause any 
14   damage to the markets.  
15             If we have a look at what happens if you take 
16   10 trades a day, which will bring about a huge number, 
17   450 instruments -- of which 88 are credit.  Let's go to 
18   twice a day.  You don't even get the initial aim of 4 
19   percent of the market.  It's a total of 1,804 
20   instruments, of which 850 are credit, representing at its 
21   peak 3 percent of instruments and 21 percent of the 
22   volume is in credit.  It's an exercise in tolerance:  Can 
23   the market tolerate it?  
24             So that's one of the first reasons, is the 
25   calibration itself is incredibly complex and largely 
0054
 1   inaccurate because it's based upon looking backwards.  
 2   And that's not the way in which the credit markets work.  
 3             Secondly, let's look at what are the unique 
 4   challenges for pre-trade transparency as opposed to post-
 5   trade transparency.  The unique challenges of pre-trade 
 6   transparency is you have to make it work for the 
 7   protocol.   
 8             So if you look at the MiFID structure, it 
 9   provides for several different trading protocols in non-
10   equities.  There is a central limit order book; quote-
11   driven, which is firm quotes; periodic auction books; 
12   RFQ; voice, and other.  Now, other is supposed to be the 
13   flexibility that will be able to keep pace with 
14   innovation.  But it is inherently limited.   
15             So what that means is that it makes it -- that 
16   the barriers to entry for constructing a trading platform 
17   in the European markets are quite high because if you 
18   have two different trading protocols, you have two 
19   different mechanisms for making public the pre-trade 
20   transparency, whereas post-trade transparency is protocol 
21   agnostic.  This just increases the complexity.  
22             Lastly, and my panelists have already commented 
23   on this, there is a complex system of waivers and 
24   deferrals.  Waivers are for pre-trade, and ultimately, 
25   the waivers work to make sure that block sizes are not 
0055
 1   made pre-transparent.   
 2             So in the credit markets at the moment, any 
 3   trades under 300,000 Euros have full transparency.  Any 
 4   trades between 300,000 Euros and one million have a 
 5   weighed indicative average.  Sometimes on an RFQ market, 
 6   that will be a weighted average of one quote.  
 7             The deferrals regime -- and I do think this is 
 8   important, even though we are speaking about pre-trade 
 9   transparency, for the reasons that Neil has mentioned -- 
10   in the markets, the concept of pre-trade transparency is, 
11   where can I find the best price?   
12             The deferral regime in Europe is far more 
13   conservative than that that exists in the United States. 
14    For the large majority of trades, all non-liquid 
15   instruments are ultimately made transparent.  But the 
16   large majority are generally made transparent after four 
17   weeks.  
18             SO what that means is that MiFID has been 
19   successful in providing non-realtime transparency for the 
20   purposes of TCA, best execution, but very little at the 
21   moment in the way of realtime transparency, whether 
22   you're talking about immediately pre-trade or immediately 
23   post-trade in the credit markets.  
24             So altogether, this amounts to Steven Maijoor's 
25   conclusion that many would disagree that there's very 
0056
 1   little realtime information available, and even less of 
 2   which is available pre-individual trade.  And this is 
 3   because the basic premise is that an instrument must be 
 4   liquid to support it, and the credit markets are not 
 5   liquid.  
 6             So there has been an enormous amount of effort. 
 7    And I would agree with Neil that it's very difficult to 
 8   come to a conclusion that there has been any benefit to 
 9   it yet.  It's too early to tell.  However, I think it is 
10   unlikely to change soon, for some of the reasons I've 
11   just explained.  The next stage is only increasing to a 
12   few hundred bonds.   
13             And also, unfortunately, I think it's about to 
14   get more complex in Europe because of the impact of 
15   Brexit, which I won't go into any more detail.  But it 
16   will make things more complicated and difficult to 
17   understand.  
18             So why -- moving on to the next section, which 
19   is, why is it that of the information that is out there, 
20   it's of little use and difficult to understand?  Well, 
21   the first answer is because there's not much of it out 
22   there.  We've gone over that ad nauseam.   
23             The second is something that has not been 
24   mentioned yet.  Earlier on, somebody made the comment 
25   that transparency is the grease that makes trading go 
0057
 1   round.  I would say that reference data is the grease 
 2   that makes transparency work.  This is the one challenge 
 3   that was not anticipated all the regulators did not heed 
 4   post-MiFID II, and that is of reference data.  It is 
 5   notoriously unreliable, difficult to create accuracy, and 
 6   those that bring it together are not regulated.  
 7             In MiFID II, we talked about the explosion of 
 8   regulated venues.  These are not primary markets; these 
 9   are secondary markets with no obligation to have a 
10   primary listing on their markets, no nexus to the issuer. 
11    And it has been an enormous challenge of the regulators. 
12    And the obligation is on the trading venues to create 
13   and report accurate reference data.  Let me give you an 
14   example of what we, as a trading venue, are required to 
15   report daily to ESMA.   
16             We are required to report daily a reference 
17   data report with 48 fields of every single bond that is 
18   available to trade -- I mentioned that is 600,000 -- and 
19   also, daily, 43 fields on a different report of 
20   quantitative information about the bonds that did trade. 
21    It's not one report; it's two.  And getting this right 
22   is critical to making the transparency machine work.  And 
23   it has been very, very, very challenging.  
24             So there is a major focus out there of the 
25   trading venues and ESMA upon the reference data 
0058
 1   challenge.  But there is an enormous depending upon the 
 2   data vendors.  And they are unregulated, and they are 
 3   late to the party.  So it will take time for this to 
 4   shake out.  But I find it hard to underestimate the 
 5   impact of the reference data challenge to make this work.  
 6             And the last reason why the information is of 
 7   little use yet is because it's dispersed amongst numerous 
 8   venues.  One of the impacts of MiFID was to increase the 
 9   number of regulated organized trading venues.  We've gone 
10   from 76 regulated markets pre-MiFID II to 112.   
11             The number of SIs have gone from 11 to 109, and 
12   there will be more to come in September when it becomes 
13   mandatory.  The number of MTFs have exploded from 61 to 
14   192.  And there is a new kind of regulated market 
15   designed to capture voice trading, which is an OTF, an 
16   organized trading facility, and there are 72 of those.  
17   Those are all responsible for disseminating data, 
18   hundreds of different points of disbursement.  
19             It has been a success in that hitherto 
20   unregulated platforms now are regulated.  Bloomberg, 
21   which is the largest fixed income trading platform, 
22   hitherto unregulated, is now regulated.  And I agree with 
23   Neil that there has been a move towards electronic 
24   trading.   
25             But this is not necessarily for price 
0059
 1   formation.  This is actually more RFQ, so one process 
 2   trades where the price formation happens off-venue.  But 
 3   the obligation to make the report has created a 
 4   condition, an environment, which pushes the reporting 
 5   onto the venue because it's such a complex job to do.  
 6             So we have seen, anecdotally, price formation 
 7   occurring off-venue and process trades becoming on-venue. 
 8    So it's a very -- that's an indication of how 
 9   challenging the regime has been to implement.  And 
10   ultimately, no one can trade in Europe without a solution 
11   for transparency unless -- unless -- that instrument is 
12   not traded on a trading venue.   
13             And since the reference data reporting is so 
14   challenging, the systems are only -- the regulated 
15   systems -- and this is another lessons learned -- the 
16   regulated systems are only built to accept 100 percent 
17   accurate and complete information.   
18             So a file with thousands of fields in it for a 
19   reference data report will be rejected if one field is 
20   missing.  Then you don't have your instruments on the 
21   venue.  Then the regulator doesn't see it.  The regulator 
22   doesn't know what they don't see.  And there becomes an 
23   opacity or a move towards OTC trading because there is no 
24   mandatory trading obligation for credit.  
25             So to conclude, there is little out there, it 
0060
 1   is true.  There is a reference data challenge.  It is 
 2   dispersed among many.  There is a complexity which raises 
 3   the barriers to entry into the European markets.  And 
 4   there were ex ante expectations of which instruments 
 5   would be liquid, which in my view is not the appropriate 
 6   way to manage -- to predict liquidity.  
 7             So so far for pre-trade, there has been very 
 8   little benefit.  But there has been an exploration of 
 9   data.  It has been more useful, less so in realtime.  It 
10   may become more useful in realtime as the reference data 
11   challenges shake out.  I think it will take longer than 
12   six months to come to any conclusions.  And the 
13   conditions for on-venue trading are now there, which 
14   hopefully is a good thing.  
15             Thank you.  
16             MR. HEANEY:  Thanks.  Very informative.  Let's 
17   turn it over for questions either from FIMSAC, SEC staff, 
18   others.  
19             MR. HEANEY:  Mihir, before we maybe go to the 
20   SEC staff for the first round of questions, do Mark or 
21   Michael, any of the panelists, have questions among 
22   yourselves?  Obviously there were four very deep, very 
23   interesting presentations.  Whether there's any questions 
24   among the panelists for other panelists?  Otherwise we'll 
25   move to the SEC staff.  
0061
 1             MR. YALLOP:  I guess the only point I would add 
 2   to the many excellent points made by my colleagues is 
 3   that in this natural experiment that is going on, with 
 4   all the defects that have just been pointed out, there 
 5   are multiple other macroeconomic disturbances to the way 
 6   markets trade and the way actors in markets are acting, 
 7   including factors such as the ECB 2E program, which is 
 8   still running the effective credential reform on the 
 9   capital banks issue and liquidity and banking activities 
10   of banks, major market makers in the world as well as the 
11   relationship between bank debt, which is half, I think, 
12   of all fixed income actions in Europe and sovereign risk.  
13             So if you're looking for reasons why the 
14   current period is tough to observe to draw solid 
15   conclusions from, we shouldn't forget those other factors 
16   as well.  
17             MR. HEANEY:  Any questions from the SEC staff? 
18    Commissioners?  
19             MR. REDFEARN:  Yes.  I have a question, and I'd 
20   be interested in any of your responses on this.  
21             I think the question is that understanding the 
22   complexity in the early stage of this evolution, for 
23   those instruments that are experiencing this -- so let's 
24   take a very narrow universe.  Neil, you talked about how 
25   this comes out when the trade is in progress.   
0062
 1             I guess the question I have is:  Can you 
 2   elaborate a little bit?  I know you said that it's too 
 3   early to have any sort of final impacts on this.  But I 
 4   am interested in what observations you've made so far, 
 5   whether they be anecdotal or moreso than that, with 
 6   respect to how often are then those -- if those 
 7   effectively -- that pre-trade transparency goes out 
 8   there.  
 9             Are they broken up?  Are there participants who 
10   are faster and getting to the trade quicker?  Is this 
11   causing any competition in terms of the narrowing of 
12   spreads?  And is there any observation in that limited 
13   universe vis-a-vis post-trade transaction quality, i.e. 
14   how is it affecting sort of the way you look at impact or 
15   execution quality relative to names in which you're not 
16   seeing that?  Or any one of those questions.  
17             MR. HAMBURGER:  I only look to Miranda because 
18   the honest answer to your question is essentially 
19   everything Miranda said about how difficult it is to use 
20   the data at the moment.   
21             From a JPMorgan perspective, so the pre-trade 
22   transparency is owed by the venue.  So if we're trading 
23   on Bloomberg, Bloomberg knows it.  If we're trading on 
24   MarketAxess, MarketAxess owes it.  So JPMorgan only owe 
25   it when we're trading on voice.  So we only owe it when 
0063
 1   we're trading on voice on a liquid instrument under SSTI.  
 2             Clients don't normally phone us when they want 
 3   to trade on a small size.  So I think last week we had 
 4   15, one five, in total pre-trade transparency events in 
 5   corporate credit trading.  So we're really talking an 
 6   extremely limited number.  
 7             Then we put that up on the JPMorgan public pre-
 8   trade transparency page during that inquiry.  So you 
 9   would -- in order to transact on it, you would need to be 
10   -- in order to transact on it, you would need to be a 
11   client in the same tier as the client who initiated the 
12   inquiry.  You would need to be looking at the JPMorgan 
13   page at the time.  And then you would have to phone up to 
14   trade.  It's not a click-to-trade.   
15             And that's largely because you don't want the 
16   initiating client to be front-run.  So you don't want it 
17   to be click-to-trade.  One of the, I think, big 
18   discussion points around pre-trade transparency is in 
19   order to -- if you were a client and you wanted to -- 
20   when would you want to trade on that?  When would you 
21   want to trade on the information?  
22             And I think in the credit markets, true end-
23   user clients, unlike market makers, think very carefully 
24   before they trade about what instrument they want to 
25   trade in how many; whereas a market maker tends to be 
0064
 1   more responsive.  
 2             You would need to have already on your sheet of 
 3   paper this trade that's now popped up in front of you.  
 4   And you need to decide very quickly want you want to do. 
 5    Because of the number of instruments that we talk about 
 6   so frequently, that meeting of -- that coincidental 
 7   meeting of desire and inquiry is very rare.   
 8             And I think that's also one of the reasons why 
 9   central limit order books and other of these venues 
10   haven't taken off quite as much as some of the advocates 
11   would want them to be.  I think at the moment there's 
12   around 4 percent of the market that goes through those 
13   types of venues.  
14             And I think it's because it's not that frequent 
15   that there is this coincident desire.  What market makers 
16   offer is the -- is substitutability.  Market makers are 
17   willing to buy one thing to sell something else soon, 
18   whereas end users tend to have a specific request and a 
19   specific requirement which is not very substitutable.  
20             MR. REDFEARN:  So just following on to that, 
21   then imagining how this is going to evolve, I mean, again 
22   it's early and I know this will be expanded.  But do you 
23   envision that in the -- with the way that this is 
24   currently set up -- give me some thoughts on how this is 
25   going to evolve.  
0065
 1             What are we going to know -- we come back in a 
 2   year from now.  Are we still not going to see anything?  
 3   Has this just been designed -- it sounds like Miranda's 
 4   suggesting the design and the way that we determine 
 5   liquidity might not be exactly right.  So where are we 
 6   headed, and what can we expect to see from all of this 
 7   hundreds of millions of dollars in effort that's gone 
 8   into this?  
 9             MR. HAMBURGER:  Yes.  So I think there's two 
10   more tensions.  One is the tension between what type of 
11   investor you're trying to serve, the retail investor or 
12   the institutional investor being the two ends.  And then 
13   the other tension is the tension between the costs and 
14   the benefits.   
15             So I think, from a retail investor perspective, 
16   any extra transparency is good new.  The table that we 
17   have in Michael's presentation is for institutional 
18   investors.  I don't know how many retail investors have 
19   Bloomberg.  
20             Risk is:  How do you make information available 
21   to the retail investor that isn't also available to 
22   institutional investors and other dollars?  So Miranda 
23   mentioned something to me on the way in which is very 
24   revealing which I hadn't thought about, the angle of 
25   which is if you make all quotes public and all in one 
0066
 1   place and from the beginning of the trade, as a dealer 
 2   that's just made the price, if I notice that my price is 
 3   the best price, I'm immediately going to change it.   
 4             I'm immediately only going to want to be the 
 5   best price by the minimum possible.  I might still want 
 6   to win it, but I'm going to be the best price by a 
 7   smaller margin.  So I'm also only going to do that if it 
 8   matters.  So I'm only going to be concentrating on that 
 9   when it's for an institutional ticket.  
10             So you can -- there's a tension between 
11   benefitting the retail customer and a potential cost to 
12   the institutional customer.   
13             MR. HAMBURGER:  I hear the desire to increase 
14   the participation of retail in credit, which is fairly 
15   low at the moment.  But I think there's a tension there 
16   also between whether you try to improve the market for 
17   the current participants or whether you risk negatively 
18   affecting the market for current participants for the 
19   potential future involvement of other participants.  
20             And then the second tension was about the cost 
21   and the benefits.  I think when we're looking at only 
22   using pre-trade transparency for liquid markets, I think 
23   we are implicitly agreeing that illiquid markets will 
24   have a greater price impact negatively from pre-trade 
25   transparency.  
0067
 1             But I think that also misses the benefits.  
 2   Liquid bonds tend to have very good pre-trade 
 3   transparency, meaning everybody knows where liquid 
 4   instruments are.  It's really very easy to find out, to 
 5   Google it or see a run or see the last trade.  It 
 6   probably traded fairly recently.   
 7             From a market maker perspective, I don't mind 
 8   too much if other clients see prices that I'm making 
 9   right now because there's not that much information in 
10   the prices that I'm making right now.  So I think at the 
11   liquid end of the spectrum, you have low cost to the 
12   market in price impact, but also low benefit from the 
13   current state; whereas at the other end of the spectrum, 
14   on the illiquid end of the spectrum, there's a lot of 
15   benefit to be had from the investor having more 
16   information about where the market is.  But that comes 
17   with increased cost.  
18             So I think much of the argument focuses on 
19   finding the sweet spot in the middle.  But it's not clear 
20   to me that there is a sweet spot because the costs and 
21   the benefits are positively correlated.  They don't cross 
22   over anywhere.  They start -- they're both high at the 
23   same time and they're both low at the same time.  
24             MR. HEANEY:  Kumar? 
25             MR. VENKATARAMAN:  Thank you, Michael.  I have 
0068
 1   a response to some observations that Neil made.  First of 
 2   all, I think perhaps the U.S. markets are a bit different 
 3   from what you're seeing in Europe.  In the U.S. markets, 
 4   there's quite a bit of participation by retail investors. 
 5    If you look at the TRACE data, for example, 60 to 70 
 6   percent of the reported transactions are of transaction 
 7   sizes less than 100,000.  
 8             I don't know what the statistics look like in 
 9   Europe, but that's exactly what was told about U.S. 
10   markets before TRACE data was collected.  So perhaps, 
11   once you look at the distribution of trade size in 
12   Europe, we may have a little bit more information about 
13   what may be going on.  
14             Second, I think you raised a very good point 
15   about whether institutions are fairly well-informed about 
16   what's going on in the market, and how much extra 
17   information does pre-trade transparency provide.  Is that 
18   information asymmetry problem that pre-trade transparency 
19   will solve?  
20             And I would say, based on some of the studies 
21   that I've conducted as well as other studies conducted 
22   using data in the U.S., there certainly seems to be an 
23   information asymmetry problem.  So let me give you some 
24   statistics.  
25             If you look at the average round trip trade 
0069
 1   executions' cost for customers in both corporate bonds as 
 2   well as municipal bonds, they decline with trade size.  
 3   So the retail-sized transactions, for example, are about 
 4   five times larger than those for block transactions.   
 5             And further, if you look at insurance 
 6   transactions, and you know more about insurance companies 
 7   because they are required to report the transactions to 
 8   their regulator, and identity of the participant is known 
 9   -- so if you look at that data, what you find is that 
10   insurance companies that are large obtain better prices 
11   or have lower transactions' cost than insurance companies 
12   that are small.  And this is for same type of trades 
13   executed on the same day in the same bond, et cetera.  
14             So if you go back to Michael's structure 
15   theory, transactions' cost should represent compensation 
16   to dealers for supplying liquidity services.  And this 
17   should be a function of the risk of market-making, which 
18   depends upon inventory costs and adverse selection costs. 
19    And it's very difficult to argue that retail-sized 
20   transactions have higher inventory costs or that retail 
21   investors are more informed that institutional investors.  
22             So what it points to is a problem with sort of 
23   the search frictions in the market.  The fact is that 
24   retail investors just simply do not have access to the 
25   best prices.  They don't know how to compare their 
0070
 1   executions with what is being observed in the market.  
 2             And further, once trades reporting was 
 3   introduced in the U.S., if you compare the same bond 
 4   before versus after, you find that the difference has 
 5   declined, but the difference still exists between the 
 6   execution quality of large versus small institutions.  
 7             So Even within the institutional world, at 
 8   least based on some of the evidence that we have, it does 
 9   appear to be the case that there is that information 
10   asymmetry problem that perhaps pre-trade transparency 
11   will help some.  Thank you.  
12             MR. HEANEY:  Larry Tabb?  
13             MR. TABB:  Thanks.  I guess the question I have 
14   is:  What would be your recommendations if you started 
15   kind of over -- I wouldn't say over again, but where 
16   would you go from here?  So the recap, it seems like 
17   post-trade is working pretty well.  People are getting 
18   some pretty good information out of post-trade.  
19             Pre-trade is active only on a very limited set 
20   of bonds, and to a certain extent people are quoting to 
21   places where nobody's going to find that information 
22   anyway.  And so they're, in effect, quoting to basically 
23   get out of the obligation of quoting, or really they 
24   don't want those -- you know, inventory or those quotes 
25   accessed.  
0071
 1             There's a tremendous problem with reference 
 2   data.  If we're looking to try to do something similar 
 3   but different in the U.S., where would you say the 
 4   pitfalls are and where would you say we should kind of 
 5   focus?  Or should we just focus on post-trade and let 
 6   pre-trade develop when electronic platforms kind of 
 7   mature?  
 8             MR. HAMBURGER:  So I will leave the retail 
 9   comments, I think, acknowledging the truth of everything 
10   that you say.  Blank sheet of paper, I think one of the 
11   things that the markets have in terms of pre-trade is 
12   generic pre-trade.  And I think generic pre-trade 
13   information has many of the benefits for the investor 
14   looking to undertake a transaction with not as much of 
15   the costs as specific pre-trade transparency.  
16             Generic pre-trade transparency I would -- I'd 
17   be talking about the squiggly lines on the page.  Do I 
18   think generic pre-trade transparency should be mandated? 
19    No, because I'm not in favor of mandating anything.  Do 
20   I think investors and investment platforms should reward 
21   providers of generic, good generic pre-trade 
22   transparency?  Yes.   
23             In Europe, we have -- Bloomberg is the dominant 
24   player.  And Bloomberg has a function -- I mean, for a 
25   long time you've had an all-quote page where the dealer 
0072
 1   puts up there the bid and ask price on the screen.  But 
 2   there's a function where the price changes color.  So it 
 3   changes to magenta if you have shown that you've stood by 
 4   your price 95 percent of the time.  
 5             That for me is a very good pre-trade 
 6   transparency set of information.  And if clients rewarded 
 7   dealers for that quality of information, I think that 
 8   leads to some good behavior.  
 9             MR. HEANEY:  Mark, would you like to respond to 
10   Larry's question?  
11             MR. YALLOP:  I think, actually, I should just 
12   respond to an earlier question first.  I think what is 
13   likely to -- well, anybody who's worked in Europe knows 
14   that it is foolish to bet against the resilience of 
15   European authorities or their determination to see a 
16   project through to its completion.  
17             So in response to the question, where do you 
18   think we'll be in a year's time, I would suggest the 
19   regulators will have cleaned up the technology problems 
20   in the receiving end of the reporting framework just 
21   elegantly described.  That will make a material 
22   difference.   
23             Changes will be trained to try to bring more 
24   data vendors into the regulatory act.  And if ESMA is 
25   smart, they will choose to focus on a much more 
0073
 1   restricted number of securities that they're trying to 
 2   operate this pre-trade price transparency framework for.  
 3             With a blank sheet of paper, as I say, I think 
 4   I personally would not have embarked on the highly 
 5   prescriptive approach to defining transparency in the way 
 6   that MiFID II has done because all the problems that 
 7   we've just been talking about in the last hour could have 
 8   been forecast.   
 9             And I would put a bunch of market participants 
10   in the room a figure out what the buy side, wholesale buy 
11   side, really wants to see in terms of price transparency, 
12   which may be only modest increments beyond or on top of 
13   what is already available in the market, and figure out 
14   the best way to deliver them.  
15             So I hope that answers your question, Larry.  
16   But that's the way I would have gone.  
17             MR. HEANEY:  Larry Harris?  
18             MR. HARRIS:  Thank you.  In trying to 
19   understand the European experience and its applicability 
20   to the United States, it's important for us to know the 
21   differences in the issues and other behaviors that might 
22   affect liquidity.  So I have four points that I'd like to 
23   make and I hope that you can respond to because I don't 
24   fully understand them.  
25             The first is I suspect that the issue sizes in 
0074
 1   Europe are smaller than the United States, in large part 
 2   because I know that the banking system is more important 
 3   in finance in Europe than it is in the United States, and 
 4   the firms are also smaller.  So if indeed the issue sizes 
 5   are smaller in Europe than the United States, then we 
 6   would expect that liquidity conditions would be 
 7   different.  
 8             The second issue has to do with what some 
 9   people call the home country bias.  There's a lot of 
10   evidence that suggests that investors prefer to invest in 
11   the securities that they're most familiar with.  So in 
12   this context, the Spanish would be investing in Spanish 
13   securities and the English in English securities and so 
14   forth.  And as a consequence, the market becomes quite 
15   segmented.  And that segmentation might also reduce the 
16   liquidity in Europe for a particular issue vis-a-vis 
17   compatible issues in the United States.  
18             The third issue is perhaps the wildest one, but 
19   still worth commenting.  With the European Central Bank 
20   purchasing corporate bonds in very large size, there's 
21   been substantial absorption of those bonds.  And as a 
22   consequence, perhaps there's less liquidity remaining.  
23             And then the final issue is the one that I have 
24   the absolute least certainty about, and perhaps you can 
25   comment on, is whether the European bonds tend to be more 
0075
 1   complex than the U.S. bonds.  I have some suspicion that 
 2   they might be because while I know that reference data is 
 3   extremely important everywhere, Miranda made a point of 
 4   talking about the importance of reference data to 
 5   liquidity.  And if it's the case that the European bonds 
 6   are indeed more complex than U.S. ones, we have empirical 
 7   results that indicate that complexity reduces liquidity 
 8   as well.  
 9             So those are four questions that I don't really 
10   fully understand.  But I'll hope that your expertise 
11   could help us out.  
12             MR. SUROWIECKI:  So perhaps addressing the 
13   first three questions -- I don't know the answer to the 
14   forth -- but intuitively, without having reference data, 
15   I would say yes.  Issue sizes are probably smaller in 
16   Europe.  Domestics, yes, they are a swing factor behind 
17   market moves.  That's correct as well.  
18             And as for the ECB, indeed that is a big 
19   factor, though perhaps -- I don't have specific data for 
20   corporate bonds.  But in liquid securities such as German 
21   bunds, for example, they've absorbed so much that these 
22   things trade very specially in repo.  Repo was mentioned 
23   before for the functioning of a liquid market.  German 
24   bunds can trade 20, 30 basis point special due to the ESB 
25   absorption.  
0076
 1             MR. HEANEY:  Rick?  
 2             MR. MCVEY:  Just thank you all for your 
 3   comments.  And following on Larry's point and Larry 
 4   Tabb's as well, just trying to draw comparisons between 
 5   the transparency regime of MiFID II and the transparency 
 6   regime here in the U.S., it seems to me through the 
 7   discussion that the vast majority of complexity in the 
 8   transparency rules of MiFID II are due to the pre-trade 
 9   rules, and the caution about the potential harm to 
10   liquidity.  
11             The bigger surprise to me is the easy part, 
12   which is post-trade, and the fact that the vast majority 
13   of credit securities have a four-week delay before post-
14   trade transparency is available to investors.   
15             And my question is:  Do you believe that 
16   investors would be better off in Europe if the initial 
17   focus was on post-trade transparency and narrowing that 
18   gap in the reporting delay for the majority or all 
19   securities to look more like the U.S. transparency regime 
20   in terms of promoting market transparency for both retail 
21   and institutional investors in Europe?  
22             MR. HAMBURGER:  So I would say certainly I 
23   think it would be a good idea to do one thing at a time. 
24    I think post-trade can be very powerful.  I think the 
25   experience of post-trade in the States has not been 
0077
 1   without its constructive discussions.   
 2             I think the buy side in general now, not in 
 3   totum but the buy side in general, is now feeling the 
 4   pain from the difficulty of transacting in block size on 
 5   the wire.  And we've seen the continued reduction in 
 6   average ticket sizes and the significant reduction in 
 7   large ticket size volumes going through.  
 8             To your four-week point, I think it's 
 9   considerably underestimated how long it takes a market 
10   maker to turn that position.  I think we see, on average, 
11   between 5 to 10 days, on average, on a five-fold basis 
12   for a market maker to turn a North America high grade 
13   position or a European high grade position.  
14             I would say to Professor Harris's point before, 
15   I think all of his observations are probably true.  But 
16   I'm not sure that it has an enormous impact on liquidity 
17   in the market on this one measure of average time to turn 
18   over a position.  
19             The reason I stress that is because when Client 
20   A is coming in to ask for a risk price from the market 
21   maker, the market maker wants to know that he has a 
22   certain amount of time before he can -- in order to turn 
23   that position around before the whole market moves to 
24   that new price point or moves to the price point of that 
25   trade.  
0078
 1             And I think it's important to calibrate -- if 
 2   we call it a deferral -- the deferral time, even on a 
 3   per-instrument basis, if possible, although given the 
 4   difficulty that Miranda said of trying to estimate future 
 5   liquidity from looking historically, but we don't have 
 6   any other better ways to do it, really.  
 7             And I think we should be calibrating to how 
 8   long do we think it's going to take to turn that position 
 9   around.  So I totally understand why, when we talk about 
10   post-trade transparency as being the most important, 
11   particularly in the U.S., the big leader of transparency, 
12   that's why I'm kind of much more in favor of generic pre-
13   trade transparency, certainly, rather than specific pre-
14   trade transparency, but even potentially over specific 
15   post-trade transparency because every time you have 
16   specific trade information, everybody involved in that 
17   trade is affected one way or another.  
18             And the knowledge that that's going to happen 
19   affects the behavior at the inception and even pre the 
20   inception of that trade.  
21             MR. HEANEY:  We'll just go one last question.  
22   We'll go to Amar, and then I'll turn it back to Mihir.  
23             MR. KUCHINAD:  I'll try to make this quick; I 
24   know we're on a time frame.  
25             WE spend a lot of time on this committee, and I 
0079
 1   think just in general, talking about market solutions 
 2   versus prescriptive or less prescriptive regulatory 
 3   solutions.  It sounds like, from what we've heard, that 
 4   post six months of MiFID II being affected, the market in 
 5   general has realized a few issues.  
 6             Are you seeing private market solutions come in 
 7   to try to address those issues?  Or is this something 
 8   that you think, as market participants, that you're going 
 9   to need regulatory intervention to solve -- ref data, the 
10   fragmentation of post-trade data and pre-trade data?  
11             MS. MORAD:  Yes.  We are seeing some solutions. 
12    Also, to add to what Neil was just saying now, market 
13   access itself has a post-trade tape called Access All 
14   which provides post-trade transparency in a way that is 
15   not specific to the trade but is carefully masking enough 
16   of the details of the trade to provide a safety net for 
17   the market makers, but still to provide price 
18   transparency.  
19             And one aspect that it takes into account that 
20   the regulatory regime does not take into account is the 
21   number of principals to the trade.  So if there is more 
22   than one principal to the trade and a minimum of three 
23   trades, the product makes the trade transparent, or 
24   enough details of the trade transparent, in a safe way.  
25             So that's an example of a market solution, 
0080
 1   which was already developed pre-MiFID II.  But it's not 
 2   what the regulators decided to emulate.  But it is one 
 3   that is working in Europe.  
 4             MR. WORAH:  Great.  So I think this has been 
 5   really informative, and it sets the stage for our next 
 6   discussion on where we are in the U.S.  And I think big 
 7   picture for all of us, it's really that this committee, 
 8   the SEC, the FIMSAC, is grappling to solve is this 
 9   chicken and egg problem.  
10             There is -- in homogeneous markets, you have 
11   pre-trade transparency and a lot of liquidity.  And what 
12   comes first?  Does this the liquidity and the homogeneity 
13   beget the pre-trade transparency?  Or is it the other way 
14   around?  If you force the pre-trade transparency, are you 
15   going to get liquidity?  So we'll be struggling to find 
16   the sweet spot. And thanks a lot for your insights.  
17             MR. HEANEY:  Real quickly, we're back 15 
18   minutes later.  So let's just have a 15-minute break.  
19   We'll start the second session.  And as Mark Yallop 
20   suggested, it should be market participants getting 
21   around the panels, which is exactly what we have, sell 
22   side, buy side, electronic venues, to thrash it out here 
23   in the U.S.  We'll start at 11:30, and that'll end at 
24   12:45.  
25             (A 15-minute recess was taken.) 
0081
 1             MR. HEANEY:  All right.  Why don't we get 
 2   started.  We'll now turn to the next panel, which Mihir 
 3   will also moderate.  We appreciate his double duty on 
 4   moderation today.  
 5             But we're going to talk about pre-trade 
 6   transparency conditions in the U.S. corporate bond 
 7   market.  And we do have a large group of panelists, and 
 8   again, I thank them for coming down to D.C.  And at this 
 9   point, I'll turn it over to Mihir.  
10   PANEL II: CURRENT STATE OF PRE-TRADE TRANSPARENCY  
11   IN THE U.S. CORPORATE BOND MARKET 
12             MR. WORAH:  Thanks.  So as far as this panel is 
13   concerned, we've tried to organize a collection of 
14   panelists that really represents the state of the art in 
15   terms of pre-trade transparency in their U.S.; different 
16   perspectives -- buy side, sell side, folks that interact 
17   primarily with retail, institutional as well as 
18   technology providers.  
19             So hopefully, this is the kind of collection 
20   that Mark was talking about, get industry participants 
21   together and try and come up with the right solution that 
22   addresses enough disclosure, pre-trade transparency.  And 
23   again, the goal is always to improve liquidity and have a 
24   better-functioning market.  
25             So with that in mind, let me introduce the 
0082
 1   panelists.  And again, when you folks speak, just go in 
 2   some depth in terms of your organization and what you do. 
 3    But just to introduce, we've got Jon Klein from the Bank 
 4   of America; Ben Macdonald from Bloomberg; Tim Morbelli 
 5   from AllianceBernstein; Steve Shaw, BondSavvy; Thomas 
 6   Urano from Sage Advisory Services; and Chris White from 
 7   BondCliQ. 
 8             So with that, let's welcome our panelists.  And 
 9   Chris, if you can lead us off.  
10             MR. WHITE:  Thank you very much.  Good morning, 
11   everyone.  I'm Chris White, the founder and CEO of 
12   BondCliQ.  It's an honor to be invited to the SEC to 
13   participate in today's Fixed Income Market Structure 
14   Advisory Committee meeting.   
15             In a moment we'll be discussing one of the most 
16   important topics in the U.S. financial system, which is 
17   pre-trade transparency in the U.S. corporate bond market. 
18    Just a fair warning to the committee members:  This is 
19   such an important event for me that my father ended up 
20   showing up.  So careful with your questions or you will 
21   be accosted.  
22             (Laughter and applause.) 
23             MR. WHITE:  Anyway, I just wanted to start by 
24   saying, I've dedicated a lot of my life and my current 
25   career to the study of markets.  And what I'm doing with 
0083
 1   BondCliQ right now is delivering, or attempting to 
 2   deliver, a solution to the marketplace that addresses 
 3   exactly what we're talking about in terms of pre-trade 
 4   market transparency.  Given the impressive backgrounds of 
 5   my fellow panelists, I'll try my best to be brief so that 
 6   all perspectives can be given ample time to present and 
 7   be heard.  
 8             We will collectively, I know, do our best to 
 9   provide the best, FIMSAC, and the investing public with 
10   our full knowledge on the current state of the U.S. 
11   corporate bond market pre-trade transparency, and best 
12   guidance on how to improve the quality and the 
13   availability of the substantial market data set.  
14             My personal perspective on U.S. corporate bond 
15   market pre-trade data is derived from over 20 years 
16   working in the U.S. market financial system, with the 
17   majority of my career focused on market innovation 
18   particularly for fixed income and the corporate bond 
19   market.   
20             I was fortunate enough in the early days to 
21   work for Rick McVey of MarketAxess.  These were the early 
22   days of RQ trading.  And then I eventually transitioned 
23   to work on electronic business development with roles at 
24   Barclay's and Goldman Sachs.  With working on the sell 
25   side, my responsibilities could be described as helping 
0084
 1   global credit desks leverage technology to optimize 
 2   corporate bond market making.  
 3             This requires constant monitoring of market 
 4   trends, reviewing of new technology solutions, and a deep 
 5   understanding overall market structure, not just in the 
 6   EUS corporate bond market but across all financial 
 7   markets and all financial systems.  
 8             Since 2016, I've launched two separate 
 9   companies, Viable Markets, a capital markets fin tech 
10   strategic consulting firm that works with large financial 
11   institutions on complex innovation projects -- many of 
12   those projects involve market data, electronic trading, 
13   infrastructure, settlement solutions for corporate bonds; 
14   and also BondCliQ, which I'm going to provide you more 
15   details on today.  But the entire goal of BondCliQ is to 
16   deliver higher quality market data information for the 
17   U.S. corporate bond market.  
18             To help frame today's discussion, I've provided 
19   a presentation for the group.  It's called, "Developing 
20   Pre-Trade Corporate Bond Market Transparency."  The 
21   materials have been provided in advance of this session 
22   and are currently available on the SEC website.  
23             I'll just say, to start, as with almost any 
24   fixed income-related topic, I think the catalyst for the 
25   reason why we're discussing pre-trade transparency today 
0085
 1   starts with the 2008 credit crisis.   
 2             There are two aspects of the crisis or results 
 3   of the crisis:  a rapid increase in the outstanding size 
 4   of the corporate bond market, and a decline in the 
 5   balance sheets being used by dealers who support the 
 6   market that have exposed, I believe, a fundamental flaw 
 7   in U.S. corporate bond market structure, which is the 
 8   lack of high-quality data.  
 9             I didn't say the lack of data because there is 
10   a ton of information out there in the corporate bond 
11   market.  However, as my presentation will walk you 
12   through, this data is very difficult to come by and is 
13   non-standard.  
14             So while modern markets rely heavily on 
15   solutions driven by high-quality centralized market data, 
16   the U.S. corporate bond market is characterized by an 
17   institutional market that has fragmented and non-standard 
18   pricing investigation.   
19             And I say an institutional market because I 
20   believe that a key focus -- at least to my perspective a 
21   key focus -- is the institutional market, the one million 
22   and up market, which Constitutions about 85 percent of 
23   the average daily volumes in the U.S. corporate bond 
24   market.  This is the area of the market in which 
25   transparency is really elusive, and pricing information 
0086
 1   is difficult to standardize.  
 2             The other topic that I think is really key, and 
 3   it was touched on slightly in the MiFID discussion, is 
 4   the asymmetric nature of information in the marketplace. 
 5    We literally have a market system in which different 
 6   market participants can be looking at the same bond but 
 7   seeing totally different pricing information related to 
 8   that bond.  That's across customers, and that's 
 9   especially the case when we're comparing the level of 
10   access between the buy side and the sell side.  
11             I think one of the things that has not gotten 
12   enough attention or discussion is the lack of access to 
13   information that the actual sell side has in the 
14   marketplace.  If we are expecting the sell side to be 
15   able to support the market with less balance sheet 
16   available, then we must have the sell side receiving 
17   equal access to pre-trade information in line with what 
18   the buy side is currently seeing.  
19             These conditions, I believe, are what are 
20   reducing reliable liquidity in the institutional market. 
21    But I am very optimistic because I think that the right 
22   pre-trade transparency methodology will improve the 
23   quality of secondary market liquidity for block trading.  
24             I am so optimistic because there is a 50-year 
25   case study on this topic.  This isn't the first time 
0087
 1   there's been a discussion about what to do about 
 2   institutional liquidity.  In fact, if we could go back in 
 3   time about 50 years ago, there's probably a similar 
 4   committee meeting.  Maybe the suits were a little bit 
 5   different.  And there were as many people in the room.  
 6             But there was certainly discussion about the 
 7   lack of institutional liquidity in the U.S. equity 
 8   market, and what were we going to do about it?  Because 
 9   we had larger customers never before in the U.S. equity 
10   market.  The market was bigger than ever before.  Yet 
11   block trading had stagnated, and overall turnover in the 
12   market had stagnated.  
13             The solution ultimately became the construction 
14   of pre-trade transparency.  What the piece of 
15   architecture was that was introduced into the market was 
16   something called the consolidated quote system, which is 
17   effectively a non-executable electronic bulletin board of 
18   market maker quotes.  This provided an up-to-date view of 
19   bid and ask prices organized into a montage.  
20             Prior to the organization of this information, 
21   if you wanted to look up a price -- for example, on a 
22   NASDAQ stock or a non-listed stock -- you would have to 
23   look in a book called the pink sheets.  Now, the pink 
24   sheets still exist today; however, most trading and most 
25   market information is referenced electronically, and that 
0088
 1   started way back in 1971 with the launch of NASDAQ, the 
 2   AQ in NASDAQ standing for "automated quotation system."  
 3             For the system that we've created, we've 
 4   approached the protocols for consolidated quote system a 
 5   little bit differently.  But on our system, buy side 
 6   participants will see consolidated quote inform; they'll 
 7   see attributed quotes with size details.  However, the 
 8   sell side will also see the quote information, but it'll 
 9   just appear as anonymized prices with no size details.  
10             I just want to point out that this 
11   architecture, this central architecture, has been proven 
12   to be a critical component to the modernization of 
13   financial markets, if you look at the examples I've 
14   provided below.  Given some of the questions of the 
15   committee members to the earlier panel, I can see that 
16   there's a lot of interest in moving the market forward.  
17   This is why we're all here today.  
18             Well, I'm here to tell you that there is a 
19   clear, sequential blueprint that indicates that unless we 
20   get pre-trade and post-trade information organized in any 
21   market, the market will have a lot of difficulty 
22   modernizing and becoming more efficient and maintaining 
23   reliable liquidity.  
24             So the intended impact of our pre-trade effort, 
25   which is a consolidated quote methodology for enhancing 
0089
 1   pre-trade transparency, is primarily to help market 
 2   makers optimize balance sheet usage.  Let me describe 
 3   what the experience is like, given that most of my time 
 4   has been spent working on strategy initiatives on the 
 5   sell side.  
 6             You're dealing with a market right now as a 
 7   market maker which is twice as big as it was 10 years 
 8   ago, where the customers are now larger than you in terms 
 9   of their overall size, and where the information that 
10   you're providing to the market every day is often being 
11   used against you and you're being negatively selected.  
12             The answer to improving the market is not to 
13   just allow more balance sheets to enter into the 
14   environment.  The answer is to actually give the dealers 
15   the amount of access to pre-trade pricing information so 
16   that they can accurately calculate risk and manage their 
17   balance sheets more efficiently.  
18             The quote in the previous panel said that it 
19   takes 5 to 10 days to turn over an investment-grade 
20   inventory position.  Given that balance sheet capacity 
21   has pretty much stagnated while the market continues to 
22   grow, we should all be very, very concerned at the lack 
23   of optimization of dealer balance sheet and the time it 
24   takes for a dealer to turn over a position.  Optimizing 
25   balance sheet and burning over your positions is a direct 
0090
 1   result of actually have access to the pre-trade market 
 2   information you need to make informed decisions.  
 3             The next thing that we hope to achieve with our 
 4   version of a consolidated quote system is to improve the 
 5   overall integrity of pre-trade institutional pricing data 
 6   in the marketplace.  I say "integrity" because I think 
 7   that if you speak to many buy side institutions, they'll 
 8   say that they have ample data; but figuring out just 
 9   what's real in the marketplace is a real challenge.  
10             I know Tim Morbelli from many previous 
11   conversations, and he's done a fantastic job creating and 
12   driving innovation at AllianceBernstein to create a 
13   system that actually can figure out which levels are 
14   real.  I won't steal any more of Tim's thunder.  
15             But not everybody has a Tim Morbelli, and not 
16   everybody works with Alliance.  So for other buy side 
17   customers, this is a real challenge.  With our system and 
18   with other consolidated quote systems, the organization 
19   of data and the competition around showing the 
20   marketplace that you can provide a quote ultimately 
21   improves the overall quality of the data in the 
22   marketplace.  
23             And then finally, our consolidated quote 
24   system, and all consolidated quote systems, are really 
25   designed to consolidate information, which ultimately is 
0091
 1   to reduce the adverse information asymmetry that's in the 
 2   marketplace.  I think we would all agree that prior to 
 3   the Kelly blue book being invented, going onto a used car 
 4   lot and buying a car was an act of bravery.   
 5             But now that there's more standardized 
 6   information in the marketplace, I feel comfortable, and I 
 7   don't know much about cars, going onto a car lot and 
 8   being able to negotiate a price and get a good deal 
 9   because there's enough ample, high-quality reference 
10   information so that I'm not adversely selected.  
11             Now, oftentimes when I give this example, 
12   people think of the buy side institutions as the person 
13   walking onto the car lot, and obviously the dealers as 
14   the car dealership.  But I would just say, from my 
15   observation working with market makers, it's actually 
16   quite the opposite in a lot of areas of the corporate 
17   bond market, where the market makers themselves are 
18   relying on the goodness and kindness of the buy side to 
19   tell them where the actual market is before they make a 
20   decision around providing on-demand block liquidity.  
21             I think the combination of results that are 
22   outlined here for what we're trying to get to with the 
23   consolidated quote methodology will improve liquidity 
24   conditions for the institutional corporate bond market.  
25   I'm confident that it will improve those conditions 
0092
 1   because in all of the previous examples that I pointed to 
 2   in my other slide, turnover waits and institutional 
 3   liquidity instantly improved, just -- for example, the 
 4   turnover rates on stocks listed on the London Stock 
 5   Exchange prior to 1986, they did not over a consolidated 
 6   quote system.  They introduced one at the end of 1986.  
 7   1987, turnover rates on the New York Stock Exchange went 
 8   from -- the monthly turnover or the monthly volume traded 
 9   equaled the annual volume in 1986.  
10             So the impact can be quite profound.  And it's 
11   actually quite simple.  When there's efficiency around 
12   the distribution of pricing information, more people are 
13   willing to trade and they're willing to take more risk.  
14   I think we are comfortable with taking risks as human 
15   beings as long as we have the information to be able to 
16   calculate those risks.  More information will give us the 
17   ability to calculate the risks for block trading and 
18   increase the overall efficiency of the block trading 
19   process.  
20             Finally, I know there may be some people in the 
21   audience or listening to the webcast who are saying, 
22   bonds are not equities, which I'm very familiar they are 
23   not.  But they are both securities, and there are 
24   overwhelming keys to establishing better pricing data, or 
25   at least themes to establishing better pricing data, in 
0093
 1   securities markets.  
 2             I know that the changes that will be needed to 
 3   make this appropriate for the U.S. corporate bond market 
 4   are important and idiosyncratic and something that we're 
 5   working on every day at BondCliQ to make sure that we 
 6   have the appropriate solution around a consolidated quote 
 7   system.  But I'd like to repeat these keys just to the 
 8   SEC and the FIMSAC committee because I think this is 
 9   something we should keep in mind when we're evaluating 
10   almost any idea around pre-trade transparency.  
11             The first key is that customer order flow must 
12   be linked to market maker price proficiency.  Without 
13   this linkage, market makers, who are the provider of 
14   prices in the marketplace and the provider of pricing 
15   information, do not have the proper incentive to 
16   consistently, accurately, and competitively price.  
17             If we give them that incentive, then market 
18   makers will also fall in line.  Again, in the previous 
19   panel, there was a mention of the Magenta Line for 
20   Bloomberg's ALLQ platform, how impactful that was in 
21   improving behaviors around standing up to prices for the 
22   ALLQ system.  
23             Having witnessed that firsthand, I can attest 
24   to the previous panelist's comments.  A lot of my job 
25   during the time the Magenta Line became popular was 
0094
 1   monitoring our response rates and making sure that we 
 2   were able to maintain Magenta status because that was 
 3   something that was important to our customers.  
 4             If there's a venue that can obviously 
 5   articulate who is able to provide quotes in key bonds the 
 6   customers are looking to trade, customers can then use 
 7   those standards to require dealers to provide pricing 
 8   information in order to see their order flow.  
 9             Also, pricing inauguration must be standardized 
10   to ensure integrity and accuracy.  I think many of us in 
11   the business do have a little inside joke about just what 
12   the different perspectives are and what really 
13   constitutes a quote.  I think adding those standards is 
14   really important to moving forward in the market.  
15             And then finally, all market participants for 
16   at least a subsection of the market must see the same 
17   price at the same time.  Price information must be 
18   centralized and made available to all.  This is not for 
19   the entire 39,000 CUSIPs that are outstanding.  This is 
20   actually for just a subsection of the market.  
21             I know I'm running short on time, so I'll just 
22   very, very quickly go over what ends up being frequently 
23   asked questions when I talk to people about our 
24   initiative.   
25             The first one:  Is the corporate bond market 
0095
 1   too fragmented for this to work?  
 2             MR. WORAH:  Chris, can it wait for the Q&A?  
 3             MR. WHITE:  No problem.  Well, then, I will 
 4   pass it off.  As you can see, I'm very enthusiastic about 
 5   this topic, but I'll pass it off to the other panelists. 
 6    And I thank you for your time.  
 7             MR. WORAH:  Thanks, Chris.  
 8             So let's go to Tom Urano from Sage.  
 9             MR. URANO:  Great.  Thank you.  Thank you, 
10   everybody.  My name's Thomas Urano.  I'm a portfolio 
11   manager at Sage Advisory.  It's an R&B here, and I can 
12   say my partners at Sage are honored that we were asked to 
13   participate in this panel.  
14             My personal history:  I've been involved in the 
15   bond market as an analyst, as a trader, as a portfolio 
16   manager, for 20 years now.  So I think I'm going to bring 
17   an interesting practitioner's perspective to how the bond 
18   market operates.   
19             Sage, just as a quick introduction, Sage is a 
20   fixed income -- is a boutique manager.  Right?  We 
21   specialize in fixed income portfolios serving primarily 
22   institutional clients.  We manage about 13 billion in 
23   assets.  Those assets are invested in individual bonds.  
24   So Sage transacts 150 to 250 trades per day off of our 
25   desk.  Our trade sizes range from block, institutional 
0096
 1   size, down to retail level.   
 2             So again, we bring an interesting perspective 
 3   to this discussion about pre-trade transparency.  So I 
 4   hope, as a practitioner, I'll be able to shed some light, 
 5   at least from a buy side manager's view, on how we see 
 6   the market and how we handle pre-trade transparency.   
 7             For us, pre-trade transparency is asking a 
 8   simple question.  The question is:  How do I best execute 
 9   a trade?  And in order to determine that, there's a 
10   variety of different components that go into it.  But one 
11   of the things that's been mentioned here today several 
12   times is the data and information.  
13             High-quality data, real data -- information is 
14   key.  I can tell you, over the last 20 years, the 
15   evolution of the data has become available to buy side 
16   managers has evolved and grown substantially.  From my 
17   perspective, we have what I would call unofficial pre-
18   trade transparency.  All right?   
19             We have the ability to source information from 
20   a variety of different areas.  Dealers, right, we see 
21   their quotes, their runs, their inventories, where 
22   they're willing to transact.  So that provides us some 
23   color in the market.  TRACE has been a godsend in terms 
24   of providing the data, not only post-trade but also 
25   giving us an understanding of where the market is near 
0097
 1   realtime -- not quite realtime but near realtime.  Right?  
 2             And then lastly, electronic trading platforms 
 3   have grown substantially.  As a portion of our business, 
 4   electronic trading platforms sit in the top five of our 
 5   counter party list.  And the amount of information and 
 6   data that's available on electronic platforms is 
 7   significant.  Market access has been a huge benefit for 
 8   us in terms of sourcing information and data, finding 
 9   markets, finding liquidity.  So electronic platforms have 
10   also provided a substantial benefit for us.  
11             So for us, when I think about pre-trade 
12   transparency or how do we best execute a trade, we think 
13   about three things.  We think about price.  We think 
14   about trade size.  And we think about market impact.  All 
15   right?  So price for us is, where is the market?  Where's 
16   the market?  And price, we could have a lengthy 
17   discussion about the right price for a bond.   
18             I've heard a couple of quotes or comments 
19   earlier this morning that what trades in the bond market 
20   is just a fraction of the market itself, and that's very 
21   true.  One thing that was mentioned earlier with respect 
22   to MiFID II was applying some of these pre-trade 
23   transparency rules to the most liquid bonds.   
24             Ironically, the most liquid bonds are not where 
25   you need pre-trade transparency.  You need pre-trade 
0098
 1   transparency in the illiquid bonds.  And so it becomes a 
 2   challenge because you're trying to source data, 
 3   information, determining where the market is using what? 
 4    Using liquid bonds because that's where the data is 
 5   available.  Right?  
 6             But then the real challenge is how you 
 7   transpose that information to illiquid securities.  So at 
 8   Sage, if we're trying to transact in a bond that hasn't 
 9   traded in the last two months, yes, it becomes a 
10   challenge.  And the challenge is, how do you determine 
11   the right price?  
12             The bond market in my mind is not a  series of 
13   individual securities all trading uniquely different from 
14   each other.  The bond market is a matrix, if you will.  
15   And the way traders and portfolio managers, the way 
16   dealers, even, price individual securities in illiquid 
17   securities, is off of the liquid securities.  So on the 
18   illiquid front, yes, there's no observable prints.  
19   There's no way to determine what the right price is.  
20   What there are is comparables.  
21             So you look to sectors.  You look to where the 
22   most liquid comps are in the market.  You apply a 
23   liquidity discount.  You apply a credit discount if you 
24   need to, a structural discount if you need to, and then 
25   you make a reference.  You end up coming up with a 
0099
 1   reference price for these illiquid securities.  
 2             So for us, pre-trade transparency on a pricing 
 3   function is entirely dependent on data, information.  We 
 4   look to liquid markets first to see where certain liquid 
 5   credits trade.  And then we'll infer discounts or 
 6   premiums to certain illiquid securities based on liquid 
 7   markets.  
 8             Size is another issue.  As you are developing a 
 9   trade as a portfolio manager, thinking about how we're 
10   going to maneuver a portfolio around, thinking about the 
11   size in the market.  And again, the discussion this 
12   morning, it has come up several times during the 
13   discussion that trade sizes aren't substantial.   
14             Five million plus is the kind of blacked-out 
15   aggregate trade size you see in printing.  Quite 
16   honestly, in our experience, trading with the dealer 
17   community, you can trade 5 to 10 million.  But once you 
18   start trading north of 10 million, asking dealers to 
19   position it, they're reluctant.  They're reluctant.  
20             So liquidity in the market becomes an issue in 
21   terms of size.  So for us, pre-trade transparency is also 
22   about size.  How much size am I trying to trade?  Will I 
23   be able to transact large size?  If I want to move a $20 
24   million piece of a position, do I need to do it in two 
25   pieces of 10?  Four pieces of five?  Do I need to do it a 
0100
 1   million at a time?   
 2             And there's no single answer to that.  But that 
 3   is one of the issues that we grapple with, is size.  If 
 4   we're going to go out into the market and trade block 
 5   institutional positions, what's the best size to do?  
 6             And then finally, market impact.  Market impact 
 7   is a big issue for us, too, particularly as you're 
 8   trading in the illiquid securities.  Market impact on 
 9   liquid securities, less important on the illiquid side.  
10   That becomes an issue.  
11             Case in point, a recent transaction we were 
12   working on, it took us two weeks to move out of $15 
13   million of an exceptionally illiquid security.  We had a 
14   market maker quote us a price on 5, and we asked to move 
15   the whole 15, and he quoted it substantially back from 
16   there, just based on size.  
17             So for us that's a significant market impact.  
18   And there is a case where, again, these situations are 
19   not unique.  They are all interrelated -- size, price, 
20   market impact.  In this particular case, it was in our 
21   best interests to piece it out, bit by bit, so we did.  
22   So two weeks to move $15 million worth of bonds is 
23   amazing, but it tells you something about the state of 
24   liquidity in the bond market as well.  
25             So pre-trade transparency for us, again, 
0101
 1   becomes a function of price, size, market impact.  So at 
 2   Sage, how have we developed this or addressed this 
 3   process?  And again, hearing the rules that were put in 
 4   place for MiFID to me sounds like near catch-up to where 
 5   we are in the U.S.   
 6             It would be my opinion that we have this 
 7   unofficial pre-trade transparency process in our ability 
 8   to aggregate data of what's available, right, from 
 9   dealers, from TRACE, from electronic platforms.  But at 
10   the end of the day, I think data become the key.  That is 
11   the lifeblood of all of our analysis in order to 
12   determine where we execute trades.  
13             And data comes in a variety of forms, whether 
14   it's TRACE.  Data from electronic platforms on the market 
15   access side is rather substantial.  We see markets.  We 
16   see public markets from dealers.  We see public markets 
17   from other buy side clients on the market access 
18   platform.  So there's almost this inherent pre-trade 
19   display of depth of market at various prices through the 
20   market access venue.  
21             So again, trying to source as much data, having 
22   it available, and having it immediate, from our 
23   perspective, is key in handling pre-trade transparency 
24   for Sage.  Thank you.  
25             MR. WORAH:  Thanks, Tom.  
0102
 1             Let's move on to Tim Morbelli from 
 2   AllianceBernstein.  
 3             MR. MORBELLI:  Thank you.  Good morning, 
 4   everyone.  My name's Tim Morbelli.  I'm a trader at 
 5   AllianceBernstein, a $535 billion asset manager.  I also 
 6   work on our trading technology development and the 
 7   architect of the ALFA product, which I'll touch upon 
 8   today.  Prior to joining AB 7 years ago, I was a market 
 9   maker on the U.S. credit desk at Credit Suisse.  
10             As a large institutional asset manager, pre-
11   trade transparency has been an important aspect of our 
12   business.  Trying to manage a large position, clear a 
13   block of risk, or determine proper valuation are daily 
14   tasks.   
15             Our view of pre-trade transparency may be 
16   different than other market participants.  In 2015, a key 
17   topic of discussion was around the liquidity in the 
18   credit markets.  Rather than focusing on liquidity 
19   directly, we decided to solve for efficiency to address 
20   the market liquidity.  That's where the idea of ALFA was 
21   born.  With a greater efficiency, we believe that we 
22   could achieve better liquidity for our clients.  
23             To give a quick overview of ALFA, it was during 
24   this time in 2015 that we started seeing a bifurcation of 
25   market liquidity.  Different venues were popping up, 
0103
 1   claiming that they had the solution for market liquidity. 
 2    We asked ourselves how could we, as a trader, 
 3   efficiently monitor these different pockets of liquidity?  
 4             That is what ALFA does.  It provides the 
 5   investor a holistic, aggregate view of the market from 
 6   providing all the relevant information available to make 
 7   that investment decision.  This can be done on an 
 8   individual security or, what's been more powerful, bond 
 9   attribute-based aggregation.  As Tom noted, these are 
10   comparables.  These help us make that investment 
11   decision.  
12             We officially launched ALFA at AB in October of 
13   2015, and shortly after we found there was a real 
14   interest from other asset managers and that this 
15   technology could be a market solution.  So we partnered 
16   with a technology company to commercialize the product, 
17   and there are several firms who are actually currently 
18   operating the product.  We found that this was not only a 
19   solution for the credit desk, but actually across all 
20   fixed income asset classes.  
21             Our goal, aside from ensuring that we were 
22   capturing the liquidity when it was available, was to 
23   increase pre-trade transparency at AB.  We, on average, 
24   get 3 million messages a day.  That's a lot of data we 
25   can use and leverage within our process.  
0104
 1             With the better picture of the market, we 
 2   increase the investor confidence in a bond's valuation.  
 3   With that increased confidence, you in essence increase 
 4   the likelihood of trading.  Within ALFA, we have also 
 5   created our own ALFA Price.  We use this to display bond 
 6   prices, but also we use vendor prices as well, some of 
 7   which is sitting on the FIMSAC committee today.   
 8             In our opinion, the credit markets are going 
 9   through an evolution.  We've seen more willingness to 
10   change over the past 2 years than the previous 20.  There 
11   are three things that have changed the most at our firm 
12   over the past several years:  our investment process, 
13   trader skill sets, and new styles of trading.  
14             An example of how our investment process has 
15   changed:  Previously an analyst would come up with a 
16   trade idea, discuss that idea with a portfolio manager, 
17   check the quant analytics, then send the order for 
18   execution to the trading desk, not understanding the 
19   liquidity dynamics.  
20             We've changed that process completely around, 
21   and we actually bring the liquidity discussion to the 
22   forefront of the investment process.  We want orders to 
23   come to the desk that are actually achievable.  We also 
24   maintain a larger percentage of our portfolio in cash-
25   like instruments, especially for our more retail-oriented 
0105
 1   products.  
 2             The skill sets we're looking for in traders has 
 3   also changed.  We target individuals with diverse 
 4   backgrounds, whether it be from derivatives, operations, 
 5   technology, or even equities.  As we're beginning to see 
 6   an equitization of the fixed income market, and we want 
 7   to make sure we have a dynamic, inquisitive team going 
 8   forward.  
 9             Even how we trade is changing.  For larger-
10   scale client flow, we're actually able to leverage what 
11   we call a portfolio trade.  We're able to trade a 
12   diversified portfolio at or near mid-market versus taking 
13   individual bond or idiosyncratic risk as you deploy that 
14   capital.  This has shown real promise for our firm for 
15   both efficiency of investment, but more importantly, 
16   lowering transaction costs for our clients.  
17             Without all the data feeding into our ALFA 
18   system, we wouldn't be able to efficiently bring the 
19   liquidity discussion to the beginning of the investment 
20   process.  We wouldn't be able to quickly engage the 
21   market liquidity when we do see those bonds trade.  We 
22   wouldn't be able to actually trade with a high confidence 
23   without that valuation, or mitigate turnaround time for 
24   portfolio trading.  
25             The speed in how we actually go and trade is at 
0106
 1   the core of what we do today at AllianceBernstein.  Thank 
 2   you.  I look forward to hearing any questions around 
 3   this.  
 4             MR. WORAH:  Thanks, Tim.  
 5             Let's get a sell side view of pre-trade 
 6   transparency in the U.S., Jon Klein, Bank of America.  
 7             MR. KLEIN:  Thanks.  My name is Jon Klein from 
 8   Bank of America.  Thank you for including me on the 
 9   panel.   
10             I've been with Bank of America for about 4 
11   years now.  My official title is head of U.S. investment-
12   grade trading, mostly responsible for our portfolio 
13   trading as well as ETF.  Create redeem efforts in 
14   investment-grade.  I spent years at Credit Suisse as well 
15   as Bear Stearns, and started my career on buy side at 
16   Merrill Asset Management quite a while ago.  
17             From our perspective, what does pre-trade 
18   transparency mean?  And when I say that, I think there's 
19   a difference between pre-trade transparency and 
20   liquidity.  I think they are interrelated, but I think 
21   that they have different places in the trading process.  
22             If you were back in the not too distant future, 
23   a lot of our information was delivered to clients in a 
24   very archaic way.  It was either messaging, voice, and go 
25   back even further, it was actually fax.  From our 
0107
 1   perspective now, our biggest focus on pre-trade 
 2   transparency is that it's important to clients in how are 
 3   we delivering our information in an effective and 
 4   efficient manner.  
 5             So first and foremost is an investment in 
 6   technology.  We spend a large majority of our time 
 7   figuring out how our clients want to see our information, 
 8   how best to deliver that information, and then ultimately 
 9   what results from the delivery of that information.  
10             And what I mean by that is everything from 
11   pricing data, where we're making generic markets on all 
12   the securities that we're responsible for, to our 
13   inventory and risk positions when we want to disseminate 
14   what we are trying to sell or buy; and even to go so far 
15   as algorithmic trading, which is developed on our side as 
16   a response to the growth in e-trading, which has been 
17   discussed kind of throughout the last two panels.  
18             But the goal for us is to create transparency. 
19    Ease of execution, especially on smaller trades for 
20   customers, to allow us more time to focus on block 
21   trades.  I know the discussion around block liquidity 
22   came up earlier, and I suspect it'll come up throughout 
23   the course of these conversations.  
24             But really, if you look at the order flow that 
25   we're receiving and that we believe our customers are 
0108
 1   receiving, a lot of it is very small.  It's timely.  It 
 2   takes up a lot of our customers' time.  It takes up the 
 3   trading desk's time.  And I think if we can find 
 4   efficient manners to deal with that, it allows the market 
 5   to focus on bigger block trades.  
 6             Just like everybody, we are looking at new 
 7   entrants, whether it's dark pools, all-to-alls, 
 8   aggregators.  Again, our job is to deliver what our 
 9   customers want and move in the direction that they are.  
10             From our perspective, how are we looking at 
11   pre-trade price transparency?  So obviously, TRACE, 
12   there's a tremendous amount of information that we 
13   receive via TRACE.  We went to use it to identify large-
14   scale trends, themes, positioning, over the -- whether 
15   it's a week, a month, a day.  But we look at the TRACE 
16   data that we're receiving and use it in a much bigger 
17   fashion than just on trade-by-trade basis.  
18             As I mentioned earlier, obviously the focus is 
19   on tech, both in responding to what our clients want to 
20   see but as well as trying to capture as much flow as 
21   possible.  Clearly the more that we can turn over our 
22   inventory, free up capital for clients, and deliver 
23   information allows us to be more active in the 
24   marketplace.  
25             Our market has expanded dramatically globally 
0109
 1   over the last 5 years.  Many of our foreign clients need 
 2   to receive information electronically, and that obviously 
 3   has been a big focus, to deliver that.  
 4             Lastly, which I'd say over the last 2 years has 
 5   become a much bigger part of our business, is the ETF and 
 6   portfolio-based trading.  And why that's relevant for a 
 7   discussion around pre-trade transparency is we utilize 
 8   different pricing frameworks, some of which rely on third 
 9   party pricing that is available to the market for pricing 
10   and acting on those sort of inquiries.  
11             I think, going forward, where do we go as a 
12   dealer community?  I think ETF and portfolio-based 
13   trading will be a large part of the market going forward. 
14    I do think we will solve the information delivery issue; 
15   as I said, if you go back 5 years, a fair amount of our 
16   information was being delivered via voice or in a very 
17   inefficient manner relative to the size, scale, and 
18   amount of inquiry that we were receiving.  
19             Based on where we are today and where the 
20   technology is, I think that aspect of our market will be 
21   solved as far as everybody knowing exactly what dealers 
22   want to do, at what price they want to do it, and 
23   therefore making the buy side's job a bit easier, at 
24   least in terms of accessing liquidity where it exists.  
25             I think, big picture, what the market decides 
0110
 1   to do with that information obviously will dictate the 
 2   liquidity conversation.  But I do think on the pre-trade 
 3   transparency side, the market is moving relatively 
 4   quickly to getting a fair amount of information to our 
 5   customers.  
 6             MR. WORAH:  Thanks, Jon.  
 7             Ben Macdonald, Bloomberg.  
 8             MR. MACDONALD:  Hi.  My name is Ben Macdonald. 
 9    Thank you for having us today, here today.  I'm 
10   responsible for Bloomberg's enterprise products.  And so 
11   I guess we play quite a unique role in the sense that 
12   we're relatively central to the price discovery process. 
13    A lot of our clients use us both for data, but also for 
14   the analytics to infer value.  
15             So we've been through this journey in fixed 
16   income, and I thought it would be useful just to explain 
17   a little bit what we do and how it's evolved over time 
18   because I think that it will continue to evolve.  And a 
19   lot of the panelists here have talked about that.  
20             Originally Bloomberg's core mission, which it 
21   still is today, was to bring transparency to the market. 
22    And really, one of the first functions, as we them, was 
23   building the terminal.  It was one that allowed a user to 
24   input a price and figure out whether the spread was the 
25   correct one or not.  
0111
 1             And that rapidly evolved into something that we 
 2   refer to single data pages, where the next extension with 
 3   that was the buy side would say to the dealer they spoke 
 4   to, "Hey, why don't you make your prices available to me 
 5   over Bloomberg or any other mechanism so that I can just 
 6   be more efficient from a workflow perspective?"   
 7             And those prices are today and have always been 
 8   made available on a permission basis, and are the -- 
 9   between, really, the function, the relationship between 
10   the two counter parties.  Over time, that evolved to more 
11   and more market makers making their prices available to 
12   their counterpart who, then again from a workflow 
13   efficiency perspective, really rather than looking at 15 
14   different pages to figure out where the price was at, 
15   wanted to look at all of the prices on one page, which 
16   was really the genesis of something that we call ALLQ 
17   system, ALLQ, which means "all quotes."  Everything on 
18   Bloomberg is for that acronym.  
19             After that, what happened was the market 
20   eventually -- the market really became more transparent 
21   because more and more people had to make prices in order 
22   to compete.  And that really became -- the next step of 
23   that was the creation of something that we call Magenta 
24   Pricing.   
25             And what Magenta Pricing is, really, a pre-
0112
 1   trade lookback on history.  So Magenta Pricing is really 
 2   just saying, the price that this dealer is making is a 
 3   reflection of how good they've been actually performing 
 4   at that price when somebody's actually executed a trade 
 5   with them.  
 6             And so really, that's where our platform is 
 7   today, which allows people to get a comprehensive view of 
 8   prices, first for a given security.  Now, as a lot of 
 9   people have noted, one of the challenges in fixed income 
10   is there are anywhere between 400 and 2,000 securities 
11   that are liquid and where you can get deep pricing.  And 
12   really, for the rest of the market, that pricing isn't 
13   available.  
14             And so what a lot of people do is really try 
15   and figure out -- look at analytics and relative value 
16   and various mechanisms to really figure out, for the 
17   things where there aren't prices readily available, what 
18   is the evaluated price?  What's the approximation of 
19   value that I can get at?  And so a lot of the analytics, 
20   some of which have been used in some of these 
21   presentations, are really around doing that.  
22             I don't personally believe the answer to 
23   liquidity in fixed income is trying to re-factor the 400 
24   bonds that trade liquid.  I think, really, there are more 
25   things that should happen and are starting to happen that 
0113
 1   are much more exciting for us in industry.   
 2             We're seeing more and more technology 
 3   innovation.  The conversation is becoming much more 
 4   around, how do I automate aggregation of the lot of the 
 5   prices that are available to me, and look at value 
 6   between those different price sets to be able to start 
 7   driving a decision?  
 8             I think the changes that are happening in the 
 9   index space are incredibly important.  The thing that 
10   fixed income -- one of the things that fixed income seems 
11   to lack versus the equity market is -- or was lacking -- 
12   was really a very vibrant index space.  And if you look 
13   at a lot of the price action that happens in equities, a 
14   lot of it is along where the industries are.  That, until 
15   very recently, wasn't the case in fixed income.  
16             And so I think what we're going to start seeing 
17   over the course of the next few years is, as a lot of the 
18   index businesses have moved out of being bank-owned into 
19   being independently owned, we'll start to see a lot of 
20   innovation and standardization of products, which I think 
21   will be one of the elements.   
22             And I don't think there's any silver bullet in 
23   this market.  It will be one of the elements that brings 
24   more standardization to the market, which will start 
25   fueling the ability for people to start trading relative 
0114
 1   value between things like ETFs, indices, et cetera, et 
 2   cetera.   
 3             So we think the market is starting to change.  
 4   I'm not convinced that the answer to a lot of woes is in 
 5   the liquid space.  We don't hear from our clients that 
 6   they have a problem with liquid bonds.  We hear from our 
 7   clients that actually, the problem is with the things 
 8   that don't trade.  And it's a very hard problem to crack.  
 9             I do think the market is still relationship-
10   driven whether it is over our platform or other 
11   platforms.  A lot of the fixed income market still trades 
12   on a bilateral basis.  And even things like RFQ or a 
13   request for quote systems in the end are nothing more 
14   than a series of multilateral conversations.  And I think 
15   that we'll continue to see that.   
16             I'm personally not convinced that -- and we've 
17   seen a lot of new businesses start up in this space.  And 
18   I think it's very hard as a practical matter for people 
19   to build enough critical mass because of this problem 
20   around liquidity.  
21             So I'll stop there.  I'm sure there'll be lots 
22   of questions.  
23             MR. WORAH:  Sure.  And while many institutional 
24   investors at the end of the day represent individual 
25   retail investors, let's hear from Steve Shaw and 
0115
 1   BondSavvy, who directly works for the individual 
 2   investor.  
 3             MR. SHAW:  Great.  Thank you, Mihir.  And thank 
 4   you to Michael and the rest of the FIMSAC for having me 
 5   speak today.  It's a privilege to be here.  
 6             I founded BondSavvy because I believe that 
 7   individual corporate bonds are underrepresented in 
 8   investor portfolios across the United States.  There are 
 9   about -- it's only about less than 1 percent of investor 
10   portfolios are in individual corporate bonds, and I 
11   wanted to change this.  
12             And what BondSavvy does is we make 25 CUSIP-
13   level recommendations per year, both the buys and the 
14   sells.  And we also provide investors with educational 
15   videos and blog posts so that they can become familiar 
16   with investing in individual corporate bonds.  
17             Prior to BondSavvy, I led Tradeweb Direct.  I 
18   was a senior executive at BondDesk, two leading fixed 
19   income ATSs, and I was also an investment banker and a 
20   senior corporate executive focused on mergers and 
21   acquisitions.  
22             I'm going to focus on two different things 
23   today.  And just so folks know, previously I provided to 
24   the SEC the BondSavvy comment letter that's available on 
25   the SEC website.  And I'm going to touch upon different 
0116
 1   parts of that as I go through my presentation.  I'm going 
 2   to talk about what pre-trade transparency looks like for 
 3   retail investors today.  And then I'm going to go through 
 4   some of the recommendations that I personally have for 
 5   that.  
 6             When we look at pre-trade transparency, we 
 7   believe it's not just about the trade because a lot of it 
 8   is about the financial disclosures that are available to 
 9   investors on the specific investment.  And I believe that 
10   when you compare the financial disclosures that are 
11   available for a corporate bond investor, you've got 
12   everything that an equity investor has.  
13             You have all the SEC filings.  You have 
14   interviews with the CEO.  You've got proxy statements.  
15   You've got everything.  You then have all the documents 
16   that pertain to that specific issuance.  So I believe 
17   that if you compare an investor investing into an 
18   individual bond, that that investor has a greater level 
19   of transparency than one investing in, say, a bond fund 
20   or an ETF, which is always changing and turning over.  
21             The ability to assess the pricing of the bond 
22   is obviously another big piece.  And I'm going to focus 
23   folks' attention to graphic 4 of the paper that I put 
24   together, on page 5.  And many folks will be surprised as 
25   to the quality of pricing in retail fixed income.  
0117
 1             And what's interesting is that I hear a lot of 
 2   institutional investors talk about, well, it's really 
 3   hard to get liquidity for 10 million, 25 million.  And 
 4   that's right because there aren't that many counter 
 5   parties, compared to in retail fixed income, it's pretty 
 6   easy to find someone to trade 5 bonds or 10 bonds or 25 
 7   bonds.  There are over a hundred dealers who can make 
 8   that type of market.  
 9             When you look at graphic 4, this is the depth 
10   of book for an Apple bond.  So Apple 345 is a '45.  You 
11   see that there were 10 dealers quoting on the bid side.  
12   There were eight dealers quoting on the offer side.  A 
13   couple things worth noting.   
14             You look at the spread.  So you look at the 
15   spread; the yield to maturity on the bid side was 3.99 
16   percent, on the offer side was 3.97 percent.  So a 2-
17   basis-point spread, which I believe compares very 
18   favorably to the institutional market.  You look at the 
19   market access, bid-ask spread index.  In this particular 
20   case, it's actually inside of where it is for block 
21   trades.  
22             Another point worth mentioning is that for 
23   retail fixed income, my order referred to it as a very 
24   egalitarian type of market.  Now, when you look at the 
25   quantities associated with these different quotes -- 
0118
 1   let's look at the right side.   
 2             You look at the offer side, and you'll see the 
 3   quantities associated with this.  This quote for 91 spot 
 4   59.  The minimum quantity is only 10 bonds.  So someone 
 5   who can invest 10 bonds, about $10,000, can get the same 
 6   price as someone who's going to buy 100 bonds or 250 
 7   bonds.  So the thought that you need to be a large 
 8   investor to be treated fairly when you buy corporate 
 9   bonds -- just not true, as shown in this case.  
10             Now, people think, well, that's an Apple bond. 
11    So what does it look like when you deal with a non-
12   investment-grade bond?  So when you look at page 6, this 
13   is a DSH DBS Corp, 7-3/4 of '26.  And people think, well, 
14   a non-investment-grade bond.  What type of liquidity?   
15             Maybe it'll get three dealers.  Maybe it'll get 
16   four dealers.  Well, you have 10 dealers on the bid side 
17   and you have 12 dealers on the offer side.  And you look 
18   at the bid-ask spread.  Again, it's 10 basis points on a 
19   yield basis.  It is about half a point from a dollar 
20   price standpoint.   
21             So what's important for folks to realize is 
22   that the retail corporate bond market is strong.  There 
23   are lots of dealers.  There are lots of trades that are 
24   taking place, which I'm going to show you in a moment.  
25   It's a strong market where more investors can transact.  
0119
 1             We then move on to page 19 and graphic 9.  This 
 2   is a bond, Bed, Bath & Beyond, of '24.  This is a bond 
 3   that BondSavvy recently recommended.  And so when we were 
 4   looking at that particular bond, we obviously did our 
 5   financial analysis and looked to see how the company was 
 6   doing and all that kind of stuff.  We looked at the depth 
 7   of book, which was strong.   
 8             Then of course we looked at trades.  And this 
 9   bond -- so BED BATH has three bonds.  It has one in '24 
10   that's 300 million size, one in '34 that's 300 million, 
11   then a larger one in '44.  This is a $300 million 
12   issuance.  And you'll see that there were over -- there 
13   were about 20 trades that took place in an hour, right 
14   around when we were looking to invest in this bond.  
15             So if you go down to the bottom of the table, 
16   that was 11:12 a.m., right around lunchtime, of above.  
17   The trades that I boxed are the inter-dealer trades.  So 
18   you had about nine dealer-to-dealer trades that took 
19   place in an hour.  
20             And in this case, it's a fairly simple process 
21   in that you see where all of these trades are.  They're 
22   all right around 85 and a half, some closer to 86.  And 
23   this is happening in a small issuance size.  So it shows 
24   that there's sufficient trading activity that's happening 
25   in individual corporate bonds.  
0120
 1             So what are some recommendations that we have? 
 2    Because while the market is strong, investor 
 3   participation is still relatively weak.  So what are some 
 4   things that we can do to change this?  Turn to page 12 on 
 5   graphic 11.  And what I'm going to focus on first is the 
 6   way to make it simple for investors to understand the 
 7   pricing of a security.   
 8             SO right now if you go onto your broker's 
 9   account, and you go and you'll search through 9,000 
10   bonds, and you'll see the TRACE activity.  It's very 
11   tedious to get through in that you'll see 20 trades, and 
12   you got to click through, and then you'll get carpal 
13   tunnel if you keep clicking through to see the full -- 
14   all the trades that happen.  
15             What's interesting, and I'm going to refer back 
16   to this BED BATH bond of 24 -- if you look at the trading 
17   activity of this bond in graph 11, you'll see that the 
18   trade that I've boxed, 94 spot 63, that was one of the 
19   last trades of the day.   
20             And you'll see how that price was substantially 
21   higher than the trades that were right before it.  All 
22   the trades that were right before it were kind of 91 and 
23   a half, right around 92.  And the reason for that is 
24   because there was a big markup in it that bond, about a 
25   2-point markup.  
0121
 1             So my first recommendation relates to the 
 2   graphs that are currently provided in FINRA.  So right 
 3   now most retail brokerages -- they show it on table 4 -- 
 4   my recommendation is that the retail brokerages show it 
 5   in graph format to make it easier for folks.  
 6             But the change that I would recommend here is 
 7   that if you look at the -- if you look at this graph, 
 8   okay, and I'm on graphic 12 on page 13, you'll see that 
 9   this BED BATH bond had a supposed spike in its price.  
10   And if you look at this where I circled in green, the 
11   last trade price, it's that 94 spot 63 price that I 
12   showed in the previous slide.  
13             Now, when you look at this -- when you look at 
14   the chart, you're like, oh, well, what was happening in 
15   this bond?  And the truth of the matter is nothing 
16   happened; it's just that there was a markup.  The price 
17   that should be shown on this price should be the dealer-
18   to-dealer trade.  And I believe that if we show that 
19   price, it'll present a more fair and a more indicative 
20   price in terms of what actually can be transacted in the 
21   marketplace.  
22             The next recommendation I would make with 
23   regard to these charts is to add credit spreads, so 
24   spread to the comparable Treasury.  My goal is to put the 
25   retail investor on the same playing field as an 
0122
 1   institutional investor.  So the institutional investor, 
 2   obviously they have got lots of investment at their 
 3   disposal.  They have all the fancy analytics.  
 4             I'd like to see an individual investor see 
 5   credit spreads because right now when I go out and talk 
 6   to investors, they believe that the only thing that 
 7   drives bond price, corporate bond prices, is interest 
 8   rates.   
 9             They don't even necessarily recognize that 
10   there is a credit component to it.  And I believe that if 
11   we add credit spreads to these charts, that'll help 
12   significantly change the discussion around corporate 
13   bonds, have investors really understand how corporate 
14   bonds trade and why the prices move.  
15             The next recommendation relates to the behavior 
16   around investing in bonds.  And I'm on page 14, graphic 
17   13.  Now, as I've said before, there aren't that many 
18   people who invest in individual bonds.  But those who do 
19   generally buy them and then they put them away for 
20   safekeeping and they hold them maturity.  They create 
21   what's called a laddered bond portfolio.  
22             And we believe that creating a laddered bond 
23   portfolio is not the best outcome for investors because 
24   it restricts their choice.  It lowers the return, and it 
25   also reduces the amount of secondary trading activity 
0123
 1   that happens in the market because those bonds just 
 2   aren't traded.   
 3             I've created a chart here in graphic 13 that 
 4   shows the pricing distribution of investment-grade bonds. 
 5    Now, these were bonds that were available on 
 6   fidelity.com.  There was about 2,000 bonds that had 
 7   yields to maturity of over 4 -- at least 4 percent.  
 8             And you'll see the distribution is that between 
 9   par and 110, about 35 percent of the bonds were priced 
10   between there.  And you'll see it's kind of an even split 
11   between what's below and what's above.  But what's 
12   important for investors to understand is that bond prices 
13   don't go up to 200.  Okay?  Bond prices are not stock 
14   prices.  Bond prices have effectively a ceiling.  
15             And you'll see here the ceiling is effectively 
16   150.  Only nine bonds out of the 2,000 got to 250.  And 
17   so what I want investors to realize is that if you buy a 
18   bond at par and that bond goes up to 125 or 130, and 
19   let's say that it does that in two years, and the bond 
20   matures in 2024, it doesn't make a whole lot of sense to 
21   hold that bond to 2024 to get par back.  You could have 
22   gotten 130.  
23             And when retail brokerages have conversations 
24   with their clients, I think it's really important for 
25   them to help their clients understand this dynamic, and 
0124
 1   that you don't necessarily always want to hold the bond 
 2   to maturity.   
 3             There can be a solid rationale for selling it 
 4   prior to maturity.  And if you have more investors 
 5   selling it prior to maturity, you obviously have more 
 6   bonds that are trading in the marketplace.  You have a 
 7   more active secondary market.  
 8             Next is recommendation 3.  And this relates to 
 9   the statement prices that are available on retail 
10   brokerages.  And I'm on page 15.  If you think about it, 
11   so we have baby boomers retiring each and every day.  
12   They're going to be transitioning from heavy stock 
13   concentration to more fixed income concentration.  
14             The question is, what investment vehicles are 
15   they going to use?  They're currently only comfortable -- 
16   you see bond funds.  That's fine.  But to get them 
17   comfortable investing in individual bonds, we need to 
18   address the point of what people perceive to be the bid-
19   ask spread or a wide bid-ask spread.  
20             And the challenge in the market is that you buy 
21   a bond at 98.  You might be able to sell that bond at 97 
22   and a half that day.  But the price that you see on your 
23   statement or the price that you see online is what's 
24   called an evaluated price.  So that has all different 
25   sorts of factors that are going into it.  
0125
 1             And more often than not, the market price that 
 2   you get is better than the evaluated price.  The 
 3   evaluated price makes a lot of sense for a security 
 4   that's not trading very often; maybe it's a municipal 
 5   bond.   
 6             But what I'd like to see is that if there's an 
 7   active quote in the size of the position that the client 
 8   holds, that should be the price that's shown on the 
 9   statement.  So if the client can sell it at 97 and a half 
10   and there's a quote there, that's the price that should 
11   be reflected on the statement, not the evaluated price.  
12             The last point I want to make is with regard to 
13   how financial services firms approach the discussion of 
14   individual bonds versus bond funds.  And it seems that, 
15   based on a lot the discussions that I have, the default 
16   for a financial advisor is to put it into a bond fund.  
17   It's easier.  The advisor doesn't have the risk of making 
18   a mistake.  He'd rather put the risk off on a large buy 
19   side firm.  
20             But what's important for the industry to 
21   understand and to push forward is that we want investors 
22   to be as familiar with as many different types of asset 
23   classes as possible.  As I said before, individual bonds 
24   are far under-represented in investor portfolios.   
25             And so when we have a discussion with clients, 
0126
 1   it needs to be even.  We need to talk about how 
 2   individual corporate bonds can be an alternative; the 
 3   benefits of individual bonds in terms of fixed coupon, 
 4   repayment of principal at maturity, the ability to be 
 5   selective.  I believe if we have those conversations, 
 6   it'll create another asset class that investors can use 
 7   very well to their advantage.  Thank you.  
 8             MR. WORAH:  Thanks, Steve.  
 9             MR. HEANEY:  So why don't we open it up now to 
10   our panelists for questions -- excuse me, to the 
11   committee members for the panelists on questions and/or 
12   comments.  But we'll start with Larry.  We'll start with 
13   you on some of the comments and topics you may want to 
14   raise.  Larry is a member of the Corporate Bond 
15   Transparency Committee -- Subcommittee.  
16             MR. HARRIS:  Thank you, Michael.  Mihir asked 
17   me to provide a quick survey of the different types of 
18   regulatory changes that we might propose from the 
19   Corporate Bond Transparency -- Pre-Trade Transparency 
20   Committee.  
21             Before I do that, let me remind you I'm going 
22   to mention Interactive Brokers.  So I want to remind you 
23   that I'm a director of Interactive Brokers, and that I 
24   only speak for myself, and that Interactive Brokers has 
25   not censored me in any way, either.  
0127
 1             So I want to talk about three types of 
 2   initiatives that we might engage in.  The first is a pre-
 3   trade disclosure of markups for riskless principal 
 4   trades.  Second is pre-trade dissemination of quoted and 
 5   indicated prices, which has primarily been the topic of 
 6   this session.  And the third is a question of 
 7   representation of customer limit orders.  
 8             This is not a complete LIS of all the things 
 9   that we could be talking about with respect to pre-trade 
10   transparency.  Steve just proposed some ways that FINRA 
11   might better organize their data, and he also talked 
12   about the presentation of fundamental information or 
13   representative information also, which are other things 
14   that we might talk about.  
15             Let's first talk about markups and commissions. 
16    So riskless principal trades are essentially broker 
17   trades for which the markup is the commission.  Pre-trade 
18   transparency, in particular transparency as to what the 
19   markup's going to be, like if there were a schedule, 
20   would allow customers to shop on price.  So it might be a 
21   reasonable thing to ask for.  
22             We'll note that most small riskless principal 
23   trades are arranged electronically so that the search 
24   costs are minimal, which means that it's not unreasonable 
25   to expect that there might actually be a fixed commission 
0128
 1   schedule.  
 2             And more generally, the notion that we're 
 3   regulating broker dealers with one framework probably 
 4   needs to be challenged.  We see that there are entities 
 5   that are doing brokerage and entities that are actually 
 6   positioning their capital.  And they probably need to be 
 7   regulated separately.  
 8             So risks of regulation is making the riskless 
 9   principal trade markups pre-trade transparent, and I want 
10   to talk a little bit about that.  But to shop on price, 
11   if you only look at the post-trade markups, you would 
12   need to have accounts with multiple brokers so that you 
13   could compare what their markups are.  And you'd have to 
14   do trades with multiple brokers.  That's quite a burden, 
15   especially for retail traders.  
16             Now, a couple arguments for and against this 
17   type of regulation.  The argument for is pretty obvious. 
18    The publication of markup schedules, which affect the 
19   commissions, would promote competition against brokers.  
20   The argument against doing this regulation is that it may 
21   not be necessary.  
22             With pre-trade markup transparency, it creates 
23   the incentives for some brokers to disclose their markup 
24   policies because now a broker can say, "Take a look at 
25   what your broker's giving you and compare it to what we 
0129
 1   promise to give you."  So it may be the case that the 
 2   post-trade markup will produce what we want.  
 3             Now let's talk a little bit about the pre-trade 
 4   price transparency, which has been the primary topic of 
 5   this panel.  So with respect to quoted and indicated 
 6   prices, most corporate bond volume trades in markets with 
 7   two-sided quotes were indications.  Of course, most bonds 
 8   trade by appointment only.  But most volume is trading in 
 9   pretty liquid markets, at least relative to what we 
10   normally think about when we think about the typical 
11   bond.  
12             Now, the quoted sizes are generally small.  But 
13   dealers often replenish size automatically.  There are a 
14   lot of proprietary traders now who are making market in 
15   bonds, and it looks an awful lot like equities.  It's not 
16   the equitization of the markets; it's the securitization, 
17   and it's the modernization of markets.  They've both gone 
18   in the same direction; just equities went first.  
19             Indications -- well, indication's not strictly 
20   firm, but dealers usually stand behind them.  And they 
21   fade when information's present or when they suspect a 
22   large order, and that's completely appropriate.  We 
23   shouldn't be requiring dealers to offer liquidity under 
24   circumstances that they don't expect.  
25             So what are the benefits of greater pre-trade 
0130
 1   transparency?  By this I mean collecting information 
 2   about quotes and indications ahead of time.  Obviously, 
 3   it improves the competition for best price because 
 4   traders can more easily shop for price, and you've just 
 5   given them the tools to look around, and you don't have 
 6   to make a whole bunch of inquiries.  It's all organized 
 7   in one place.  
 8             It increases the dealer incentives to compete 
 9   for price because that's now you get order flow.  It 
10   reduces the cost to clients of monitoring their best 
11   execution, which is really important.  And then finally, 
12   tangentially but still very important, it improves the 
13   pricing of bond funds.  So that's an investment 
14   management issue, not a trading issue, but it's a 
15   derivative issue that's very important to us.  
16             Now, there are a bunch of market-based 
17   innovations.  We've heard from several people on the 
18   panel about these already.  Many vendors are now 
19   assembling pre-trade price information, as Chris White 
20   has described in his firm and others as well.  This 
21   information largely benefits broker dealers, proprietary 
22   traders, and the large buy side instructions because 
23   these are the ones who are getting the information.  
24   Retail doesn't see much of it.  
25             Now, the retail traders, though they don't 
0131
 1   generally see the price information as they do in 
 2   equities and probably would benefit if they did, will 
 3   note that they do get benefit from this information 
 4   because they benefit indirectly when their brokers, using 
 5   this information, can obtain better prices for them.  So 
 6   it's not -- the fact that they're not necessarily seeing 
 7   it -- some do, but that they're not necessarily seeing it 
 8   doesn't mean it's not valuable to them.  
 9             Let's talk just a little bit about best 
10   execution because best execution interacts tremendously 
11   with the notion that there are interesting pre-trade 
12   prices.  So we know best execution is a broker dealer 
13   responsibility.  Unfortunately, the history of best 
14   execution is that in all markets, regrettably, some 
15   broker dealers -- most not, but some -- don't take this 
16   responsibility seriously, especially when dealing to 
17   their clients.  
18             So just as an aside, let me tell you why Reagan 
19   MS imposed the trade-through requirement in equities, in 
20   particular in NASDAQ.  It had nothing to do with market 
21   structures issues.  It had to do with making it easier 
22   for the SEC to enforce best execution requirements 
23   against NASDAQ dealers who weren't honoring it.  So there 
24   were 3 percent of trades that were trading through, which 
25   at the time in equis was not appropriate, and this just 
0132
 1   created a very bright line standard.  
 2             So the best execution problem is that there's 
 3   slippage in best execution.  It's greatest when the 
 4   customers can't measure the quality of their executions. 
 5    To measure the quality of executions requires pre-trade 
 6   information and some analytic sophistication.  The 
 7   problem affects retail traders most, but also many 
 8   institutional traders.  So if we give them the pre-trade 
 9   information, they'll be able to do better off.  
10             So what are the regulatory issues here?  Well, 
11   the pre-trade transparency facilitates the regulation of 
12   best execution and cheapens the surveillance and 
13   enforcement costs, whether the surveillance is by 
14   regulators or by the client themselves.  And of course, 
15   the threat of enforcement deters bad behavior, which is 
16   exactly what we want.  We don't want to enforce; we just 
17   don't want to have the -- we don't want to need to 
18   enforce.  
19             So fair enforcement -- now here I want to talk 
20   about private MBBO versus a public MBBO.  So fair and 
21   uniform enforcement requires some standard national best 
22   bidder offer, something like that, something that we can 
23   all say is accessible and measurable so that if somebody 
24   does something wrong, you can say, "Well, you should have 
25   done otherwise."  
0133
 1             So if we delegate the production of a standard 
 2   MBBO to multiple private entities, that would be 
 3   problematic.  And the reason it's problematic is because 
 4   we have a whole bunch of different standards, and anybody 
 5   who doesn't want to face a serious standard will 
 6   subscribe to the least capable producer of the MBBO, 
 7   which is certainly not something we'd want.  
 8             So what are our policy alternatives in this 
 9   area?  Well, one is to simply not do anything and allow 
10   the private entities to continue to produce data.  And we 
11   may have some issues with enforcement of best execution 
12   or something like that, but we'll get somewhere for sure.  
13             The second is to require broker dealers to 
14   present pre-trade prices from some reputable source to 
15   their clients so the clients actually start seeing the 
16   data.  Clients could demand it themselves, but they seem 
17   not to, or they may not even be aware that they could, or 
18   I'm not exactly sure why it's not getting to them.  That 
19   would -- we'd have to argue about what "reputable source" 
20   is, but it would be a step in the right direction.  
21             We could have some entity like FINRA, perhaps a 
22   consortium of exchanges or ATSs or something, produce and 
23   disseminate an MBBO, as occurred in equities many years 
24   ago.  And the existence of the MBBO undoubtedly would 
25   cause dealers to convert most of their indications to 
0134
 1   firm quotes.  
 2             So the problem of the indications versus firm 
 3   quotes is much greater in the equity space because things 
 4   happen faster, and bigger things happen because there's 
 5   more risk.  So if they're able to do firm quotes in 
 6   equities, for sure it can be done in fixed income.  
 7             And the last public policy alternative in this 
 8   space would be to prohibit trade-throughs for 
 9   electronically accessible quotes.  And that's a big step, 
10   but it's a step that at some point people might want to 
11   think about.  
12             The last area that I wanted to talk about is a 
13   potential area for our recommendations, is to talk about 
14   the representation of limit orders.  The current practice 
15   is that most brokers won't represent client bond limit 
16   orders in electronic trading systems, especially retail 
17   and smaller institutional.  
18             Those brokers that will represent them will 
19   display a client order -- the ones that will display a 
20   client order usually restrict display to a short period, 
21   or require all-or-nothing executions.  And we heard about 
22   this a couple meetings ago.  
23             Interactive Brokers is the major exception 
24   here.  Interactive Brokers collects their own version of 
25   a private MBBO from lots of different sources, and they 
0135
 1   allow their clients to trade as though they were trading 
 2   equities -- it looks just the same, same software, 
 3   everything.  
 4             So what are the implications?  These policies 
 5   help ensure that clients trade with their broker dealers, 
 6   which is one of the reasons why the broker dealers are 
 7   not leaders in innovation here, because they like trading 
 8   with their clients.  They make a lot of dealing profits 
 9   and what would appear to be markups that are hard to 
10   understand with reference to the -- especially with 
11   riskless principal trades with reference to the costs of 
12   arranging those trades.  
13             These policies also prevent clients from 
14   competing with broker dealers and proprietary traders to 
15   provide liquidity, which is a really serious problem 
16   because the reason the equity markets are so incredibly 
17   liquid is because the buy side is easily able to meet the 
18   buy side.  There are other reasons, too -- the 
19   concentration of order flow, and assured names, and stuff 
20   like that.  But it's the buy side meeting the buy side 
21   that's so important.  
22             Some industry participants will argue that 
23   these policies protect clients from poor executions, and 
24   we heard that a couple meetings ago.  I'll let you 
25   evaluate that yourself.  So in particular, if a client 
0136
 1   represents a limit order and the limit order gets 
 2   executed as a standing limit order, there is some 
 3   possibility that that would be a bad execution at the 
 4   time.  But in most circumstances, we'd never suspect 
 5   that.  That's at least my opinion.  
 6             So what are the implications for liquidity?  
 7   Well, transaction costs are lowest when the buy side can 
 8   meet the buy side, as I mentioned.  But let me explain 
 9   the reason why.  So dealers widen spreads when they fear 
10   that they trade with informed traders.   
11             If you're afraid that you're going to trade 
12   with an informed trader, you don't want to be tied to the 
13   market because that's how you get hurt.  You widen your 
14   spread so you, on average, can recover from other traders 
15   what you're losing to the informed traders.  
16             In contrast, buy side traders do just the 
17   opposite.  So buy side traders price more aggressively 
18   when they fear that they won't trade because there are 
19   informed traders in the market.   
20             And this is the reason why buy side traders 
21   tend to tighten spreads and make the markets really 
22   liquid, because when there are informed traders in the 
23   market they're afraid they're going to get hit if they're 
24   on the wrong side, no matter what.  But what they're 
25   afraid of is if they're on the right side of the market, 
0137
 1   they won't trade when the buy side gets in front of them.  
 2             So I'm going to finish up very quickly.  
 3   There's a chicken-and-egg problem here.  The industry 
 4   reports that most clients do not want to trade limit 
 5   orders.  But do the clients have little intrinsic 
 6   interest?  Or do they not express interest because the 
 7   facilities are simply not available to them?  I would 
 8   note that a trade-through rule, of course, would make 
 9   exposure far more attractive to clients because it would 
10   allow them to actually trade.   
11             So the regulatory alternatives in this space 
12   is, first of all, do nothing; if the clients want exposed 
13   bond orders, they'll migrate to brokers like Interactive 
14   Brokers and other competitors that support these 
15   services.  
16             We could also require brokers to route standing 
17   client orders to an existing or purpose-created bond 
18   order display facility, so either existing exchanges or 
19   ATSs, or perhaps a special facility that will represent 
20   these orders.  And if we wanted to really provide a lot 
21   of teeth, we'd add a trade-through rule to protect 
22   standing orders, as we've done in the equity markets.  
23             So final comments:  I want to talk very briefly 
24   about alternative regulatory approaches.  
25             MR. HEANEY:  Larry, I'm going to stop you there 
0138
 1   -- I'm sorry -- just because we are running out of time 
 2   and we've got to open this up to Q&A.  
 3             MR. HARRIS:  Okay.  It's really just a minute, 
 4   but if you --  
 5             MR. HEANEY:  SO let's take a pause.  Let's open 
 6   this thing wide open.  Panelists to respond.  Questions 
 7   from the committee members.  Please, just raising the 
 8   card, and we'll go from there.  
 9             Brian? 
10             MR. ARCHER:  Thank you, Mike.  I need just to 
11   respond to some of this from my perspective and working 
12   at Citigroup.  I think there's two issues here.  It's an 
13   institutional issue versus a retail issue.  And if you 
14   look at the institutional market, I would go back to what 
15   Rick said in the January meeting, and I'm going to 
16   paraphrase Rick, if that's okay.  I don't remember the 
17   exact quote.  
18             He said something like, "If it's not fixed, 
19   don't" -- "If it's not broke, don't fix it."  And if you 
20   look at the institutional market from my perspective, 
21   there's a hell of a lot of pre-trade transparency.  And a 
22   pre-trade transparency institutional market, because it's 
23   not a -- it's a best -- it's a last look market, meaning 
24   there's not firm quotes.  But the information available 
25   to institutional clients I think is immense and getting 
0139
 1   better.  
 2             And I think Jon mentioned that, that there's 
 3   more and more innovation on the institutional side of 
 4   what dealers are willing to give their clients pre-trade 
 5   as far as going out and giving them their inventory.  So 
 6   I personally don't think it's an institutional problem, 
 7   and Tom, it's the example that you cited, where it took 
 8   you two weeks to move 15 million bonds.  I don't think 
 9   pre-trade information would have helped you move that any 
10   quicker.  Right?  That's about illiquid versus liquid 
11   bonds.   
12             So from my perspective, it's really a retail 
13   problem.  And I'm not familiar with the retail markets, 
14   but I would differentiate the issue.  So for me to sit 
15   here and advocate that the SEC look to enforce pre-trade 
16   transparency on the dealer community with respect to 
17   institutional clients I think would be a mistake.  
18             Given that, I think it's already there and it's 
19   developing and there's plenty of competition.  When I 
20   look at the competition in the market that we transact 
21   in, I think there's plenty of competition, and 
22   surprisingly, it's getting more competitive, not less 
23   competitive, given a narrowing of bid-offer spreads.  
24             MR. HEANEY:  Larry?  
25             MR. TABB:  Yes.  I think I want to follow up 
0140
 1   from what Brian said, and to a certain extent Larry and 
 2   the team over here.  Are we looking at possibly two 
 3   separate markets and two separate market structures?  It 
 4   looks like, from the first panel, we've had some pretty 
 5   serious challenges with liquidity, given MiFID and at the 
 6   enforcing of pre-trade transparency in the European 
 7   market.  
 8             It seems like it's somewhat developing on the 
 9   institutional side.  Clearly, the retail side there may 
10   be some challenges.  Are we looking -- this is not a 
11   statement.  It is more of a question.  Maybe we shouldn't 
12   be looking at just a one-size-fits-all issue.  Maybe 
13   there should be some way of looking at separating retail 
14   from institutional, and then migrate to some order 
15   handling or possibly limit order rules in terms of 
16   retail.  
17             I don't know if a refocus on -- in the market 
18   structure evolution, the institutional side is the -- 
19   institutions demand better services from the 
20   institutional brokers.  I don't know.  
21             MR. HEANEY:  Any of the panelists like to 
22   respond to any of that?  
23             MR. SHAW:  With regard to that point, I'd say 
24   that for retail investors, you have a very competitive 
25   market.  Okay?  Compared to the institutional market, 
0141
 1   where it seems that there has been discussion around 
 2   changing the trading protocols and those sorts of things. 
 3    The trading protocols for retail already work very well, 
 4   in my -- from my point of view.  
 5             The challenge with retail is getting more 
 6   information out to retail investors that they can use to 
 7   make a successful investment.  That's I would distinguish 
 8   it.  But you have live and executable codes.  You have a 
 9   very liquid market.  You have many dealers providing 
10   these quotes across 9,000 bonds.  
11             MR. HEANEY:  Tom?  
12             MR. TABB:  That's not -- oops, sorry.  Go on, 
13   Larry.  
14             MR. SHAW:  That didn't seem to see what Larry 
15   was getting at.  
16             MR. HEANEY:  Tom, please?  
17             MR. GIRA:  Yes.  I was just going to add that I 
18   think Chris made a good point about the consolidation of 
19   equity quotes back in 1975.  I think another important 
20   thing that happened then was also a firm quote rule.  And 
21   I think if you're going to have, for retail, some type of 
22   pre-trade transparency, I think it's very important that 
23   those indications have meat to them and that there's some 
24   sort of incentive to honor your quote.  
25             And we've talked about sort of Magenta and 
0142
 1   other things.  I'd be curious to see if there's any other 
 2   -- any other panelists have ideas about how we could put 
 3   some oomph behind those indications so that there's not 
 4   inappropriate backing away.  
 5             MR. HEANEY:  Please, Tim.  
 6             MR. MORBELLI:  So part of the information we 
 7   aggregate is also some of the smaller, more esoteric or -
 8   - ATS is a smaller flow.  And there actually are dealer 
 9   platforms to do it, very large retail presence, who 
10   actually, from their dealer balance sheet, back up those 
11   quotes.  
12             So whether the liquidity provider is on their 
13   ATS, if that individual does not honor that quote, that 
14   broker dealer whose platform it is actually does honor 
15   that quote.  And they have a separate balance sheet for 
16   that as well.  So in essence, it is providing a firm 
17   liquidity for one of the largest retail platforms that 
18   are out there.  
19             MR. WHITE:  So Tom, I'll just say, in your 
20   question, I think that quotes are really the father of 
21   firm pricing.  And so until there can be better 
22   coordination and consolidation of the quoted market, I 
23   think expansion of firm pricing and the reliability of 
24   people standing up to their price is going to be 
25   challenged.  
0143
 1             It's just really simple.  If I don't know as a 
 2   market maker what the quoted market is around me, it's 
 3   very difficult for me to have a high degree of confidence 
 4   to stand up to that market.  Larry had mentioned it; it's 
 5   difficult to force a dealer to stand up to a market when 
 6   they're now under circumstances that they did not 
 7   anticipate.  Well, if I don't know where the other bids 
 8   and offers are around me, when I get hit it's a 
 9   circumstance I didn't anticipate if I didn't want to be 
10   the best bid.  
11             And then I'll just respond to Mr. Archer, who I 
12   agree with a lot of your points.  And I think it has been 
13   said on this panel that there is ample pre-trade 
14   information in the institutional market.  But I think 
15   that that perspective is if you were sitting in a buy 
16   side seat.   
17             I think if you were sitting in a sell side 
18   seat, we have not seen the same elevation of the 
19   availability of pre-trade market information, especially 
20   when it comes to institutional markets.   
21             So we see this in the numbers in terms of what 
22   the sell side feels confident printing on a block basis 
23   the universe of CUSIPs that are trading in large sizes, 
24   and getting smaller and smaller, despite the fact that 
25   the market has been getting larger and larger.   
0144
 1             And I think that has to do with dealers relying 
 2   on TRACE as the objective reference point for making 
 3   decisions on large trades as opposed to being able to 
 4   know basically where the quoted market is around them.  I 
 5   think if they have that information, they may be willing 
 6   to take more risk.  
 7             MR. HEANEY:  Kumar?  
 8             MR. VENKATARAMAN:  Thank you, Michael.  So as 
 9   Chris pointed out, if you look at what's going on in the 
10   corporate bond market, there's been an increase in the 
11   issue and its activity but a reduction in the size of the 
12   dealer capital.  And to the extent that we rely on the 
13   traditional set of dealers to supply liquidity, this 
14   leads to a liquidity problem.  
15             And our real solution is to see if we can 
16   expand the pool of liquidity suppliers, and one of this 
17   pool of liquidity supplies the buy side.  As Tim pointed 
18   out, in their investment process now, they're thinking 
19   more carefully about the liquidity mix, implying this 
20   could be a trading signal.  They could jump in and supply 
21   liquidity in case of situations where it might be 
22   profitable to do so.  
23             So my question is:  At this point, how are the 
24   trading protocols set up such that a buy side 
25   institutional can actually jump in and supply liquidity? 
0145
 1    How easy is it for buy side to find the buy side?  Or is 
 2   it still that a lot of this flow is still being 
 3   intermediated by the traditional dealers?  
 4             MR. MORBELLI:  So, I mean, part of that 
 5   evolution that happened in 2015 was where there were 
 6   other electronic venues that came online to allow the buy 
 7   side to meet to actually transact directly buy side to 
 8   buy side.  There are several out there that we can 
 9   actually choose from.  So we do have a suite to pick 
10   from.  
11             But what we're noticing, too, in these 
12   particular venues, well, I should tell you that when you 
13   have a willing buyer and a willing seller, even once you 
14   have that, which is a difficult thing to get in the first 
15   place, they're only agreeing to trade 20 to 25 percent of 
16   the time.  
17             Now, why is that happening when you have a 
18   willing buyer and a willing seller of the same bond?  
19   It's because they don't have the confidence on where that 
20   valuation should be.  So if they're framing the making 
21   and they want to get ion the bid side, or the other one 
22   wants to get lifted on the offer side, they're not 
23   meeting in the middle because they both think they're 
24   right.  
25             That's what the pre-trade transparency is 
0146
 1   about.  And that's what the dealer community helps us 
 2   with as well.  For lack of a better term, but they 
 3   shepherd us through that.  They are there showing, 
 4   "Here's what we're seeing.  We should be skewed more to 
 5   the buy side," or "We should be skewed more to the offer 
 6   side" because of X, Y, and Z.  
 7             So the dealer there is to help facilitate that 
 8   as well for when you don't have that high confidence to 
 9   trade.  
10             MR. HEANEY:  Jon?  
11             MR. URANO:  Yes.  So let me just Yes.  So let 
12   me just add on to that, too, and maybe in two parts.  
13   Number one, to the discussion about quotes, indicative 
14   pricing, and firm pricing, and there's been various 
15   discussions about making it available.  The problem is, 
16   making it available where?  
17             There's not one unique spot where we as 
18   portfolio managers or traders turn to say, "Oh, this is 
19   where all the information sits."  There's not one central 
20   spot to go find all that information.  And that's the 
21   key.  
22             I mean, the discussion is saying, hey, let's -- 
23   for the retail investor, let's put up firm quotes with 
24   size into a market.  Great.  But which market, and who?  
25   And there are a variety of different private vendors that 
0147
 1   can supply those -- Interactive Brokers, ALLQ, 
 2   MarketAxess -- a variety of different platforms can 
 3   supply those; but again, bringing all market participants 
 4   together and putting information in one spot where 
 5   everybody can then know, what is the price?  
 6             In the example just discussed, that buy side to 
 7   buy side trading has certainly grown as a portion of our 
 8   business.  But Tim's correct.  There's a challenge 
 9   because I may know where our price is.  The buy side 
10   counter party may know where his price is.  We're not 
11   agreeing.  No trade happens.  Right?   
12             There's not a central spot for me to go, "Hey, 
13   this is where I'm going to be on the market."  So there's 
14   a difference, in that scenario, we're price-making, not 
15   price-taking.  All right?   
16             So the way to approach it, I guess, is two 
17   different avenues.  On the retail front, providing a 
18   centralized location where dealers can put up liquidity 
19   in quotes, in firm quotes, and executable quotes is fine. 
20    Right?  And that would be a nice spot, if there's a 
21   centralized place where everybody can go to find that 
22   information.  
23             But I'm not sure that works so will on the 
24   institutional front.  The reason being is size, again.  
25   You can put up 10,000, 100,000, $250,000 worth of market 
0148
 1   value with a firm quote; for a retail investor, no 
 2   problem.  But what happens when I show up and I have to 
 3   trade 20 million of that?  All of a sudden I move the 
 4   market tremendously, and that 250,000 piece doesn't do 
 5   much good.  
 6             So again, a different approach for two 
 7   different avenues, the institutional front and the retail 
 8   front.  
 9             MR. HEANEY:  Jon Klein?  
10             MR. KLEIN:  Sure.  Back to the point made a few 
11   minutes ago.  I think it's important, when thinking about 
12   any of this, that any one price -- or any trade is not 
13   dictated alone by where every other dealer is bidding or 
14   fundraising or there seeing it.  Every trade is thought 
15   up as a collection of trades.  What's happening in the 
16   rest of the market?  Where is our inventory?  What do we 
17   expect out of customer demand going forward?  
18             So yes, in terms of seeing an individual trade 
19   and looking at it in isolation and just thinking about 
20   what are all the other prices at the time, it's probably 
21   a small piece of the decision by any one dealer.  But I 
22   would think that the totality of the state of the market 
23   at that time is probably more relevant.  
24             And I think that is sort of what we're aiming 
25   towards, on the pre-trade information delivery.  If we're 
0149
 1   delivering the fact that we have a lot of inventory and 
 2   that we're willing sellers to customers, that's reflected 
 3   in our pricing.  But what is reflected by the rest of the 
 4   market is going to be different.  
 5             So I think it's important, when you think about 
 6   any of this, to not only think about each trade in the 
 7   totality of a bunch of trades that are happening all at 
 8   the same time, as well as the disposition of the market. 
 9    I think it's very hard to isolate any one decision by 
10   just having a bunch of data points solely around that 
11   decision, whether it's prior TRACE prints on that 
12   particular bond, or where other dealers are quoting that 
13   bond at that time, or where ECNs are at that time.  I 
14   really think it's about the totality of the business and 
15   the positioning of the market.  
16             MR. HEANEY:  Tim, did you want to respond?  
17             MR. MORBELLI:  I mean, just one other point to 
18   make.  Part of the portfolio and how we are structured 
19   and carrying that additional liquidity is when we do see 
20   those opportunities out in the marketplace where we can 
21   be a liquidity provider, we want to be able to rotate 
22   that portfolio quickly, where we can actually cash in 
23   that liquidity, that cash-like instrument, and actually 
24   be the liquidity provider for the marketplace.  
25             And that's exactly why we're structured that 
0150
 1   way, and that's why our investment process is built 
 2   around that.  So when we do spot that opportunity and we 
 3   see that we can be the -- call it the balance sheet 
 4   buffer, we take advantage of that.  
 5             MR. HEANEY:  John Bagley?  
 6             MR. BAGLEY:  I just wanted to mention ATSs that 
 7   cater to retail have a similar Magenta Line.  But it's 
 8   really around 98 or 99 percent.  So I just want everybody 
 9   to know that it's not just available in one spot 
10   institutionally.  The execution for retail investors is 
11   usually important, so the bar is left pretty high for 
12   them to execute on.  
13             MR. HEANEY:  Thank you.  
14             Amar?  
15             MR. KUCHINAD:  I'm going to ask a similar 
16   question that I asked the previous panel, which is, from 
17   your guys' perspective as market practitioners, do you 
18   think that the incentives are in place for private market 
19   solutions to this?  Or does there need to be regulatory 
20   intervention to incentivize dealers to share market 
21   intelligence that has been pretty hard to construct and 
22   costly to construct?  
23             MR. WHITE:  I think there are plenty of 
24   incentives for the dealers to start thinking differently 
25   about pre-trade transparency.  I mean, we currently have 
0151
 1   18 dealers providing us with their pricing data on a 
 2   daily basis, and that's because we've worked really hard 
 3   to create an environment in which the dealers feel that 
 4   by giving pricing, they're able to grow their business, 
 5   manage their business more effectively, and ultimately 
 6   return to profitability from a market-making standpoint 
 7   in portions of the market that have been difficult and 
 8   challenging.  
 9             It's interesting to hear.  I mean, the 
10   colleague to my left, Thomas, is talking about there 
11   needs to be something that centralizes the information in 
12   the institutional market.  We're hearing Tim Morbelli say 
13   that they feel confident when they have the information 
14   to step and then provide liquidity according to the 
15   signals they get on ALFA.  
16             I'm just highly confident that if the dealers 
17   have greater access to the priced market, that they can 
18   start making informed decisions about where they want to 
19   step in.  I don't think that that needs to come from 
20   regulatory mandate, too.  I think there are plenty of 
21   active private solutions that are trying to approach 
22   this.  
23             And as we've seen with electronic trading, 
24   nobody mandated the RFQ in the corporate bond market.  
25   Yet we see how important it's become to the everyday 
0152
 1   activity in the marketplace.  I have a similar optimistic 
 2   outlook for how improved market data solutions can come 
 3   into this environment.  
 4             MR. HEANEY:  Ben?  
 5             MR. BAGLEY:  Yes.  I mean, I agree, I think.  I 
 6   think one of the things that has actually -- is helping 
 7   the market right now is the farther there is, an 
 8   homogeneous kind of regulatory framework is aligned, 
 9   different types of competition, and different types of 
10   innovations happening.  
11             You can see it across the spectrum where there 
12   is sell side, buy side.  In the vendor space, people have 
13   traditionally been there.  There's a lot of new entrants 
14   in the market.  And I think it's hugely important because 
15   I think the challenge with fixing in particular is that 
16   there is no one size fits all.   
17             And I don't think there's a silver bullet in 
18   terms of a third platform with the price that everybody 
19   can coalesce around.  I think that you need to have 
20   color.  You need to have context.  It's relationship-
21   based.  And you want to have an ecosystem that has the 
22   ability to evolve as liquidity evolves.  
23             MR. HEANEY:  Brian?  
24             MR. ARCHER:  I think it's important just to 
25   reiterate that there are opportunities for clients to 
0153
 1   trade with clients in plenty of places, including one of 
 2   Rick's protocols.  And just because that's available, to 
 3   Tim's point, buyers don't meet sellers all the time when 
 4   they think they know where the market is.   
 5             And that happens not just on those platforms; 
 6   that happens on dealer desk as well, where we may get an 
 7   order from a client, but we can't get anybody to take the 
 8   other side of that because they don't believe in that 
 9   price.  So just because there's a willing buyer and 
10   seller doesn't mean it meets all the time, either in the 
11   platform space or on a dealing desk.  
12             MR. HEANEY:  I'm going to pass it back to Mihir 
13   to conclude, as I'm sensing there could be a mutiny, 
14   given that our lunchtime for the committee members has 
15   now gone from one hour to 28 minutes.  So Mihir?  
16             MR. WORAH:  Adjourned.  
17             MR. HEANEY:  Thank you very much to the 
18   panelists.  You guys were terrific.  Great second act to 
19   what was a very good first.  Set the stage.  So thank you 
20   very much.  We will be back here at 1:30.  
21             (Whereupon, at 1:00 p.m., a luncheon recess was 
22   taken.)
23   A F T E R N O O N  S E S S I O N 
24             (1:35 p.m.) 
25             MR. HEANEY:  Okay.  Why don't we get started.  
0154
 1   This will be the closest we are to on time all day, maybe 
 2   with the exception of 9:30.   
 3             So we'll turn it over to the U.S. municipal 
 4   securities market conversation around pre-trade 
 5   transparency.  But before I turn it over to Lynn Martin 
 6   to host that and moderate that, first we're going to hear 
 7   from Abby Kim -- excuse me -- the economist in the 
 8   Division of Economic and Risk Analysis at the SEC, who 
 9   recently published the white paper that we sent around to 
10   everybody.  And I believe it is also in the pack here 
11   today.   
12             But that was a paper that was written on pre-
13   trade information in the municipal bond markets.  So 
14   certainly timely.  Abby, thanks for joining us.  And the 
15   floor is yours.  
16             MS. KIM:  Thank you for the Commission and the 
17   FIMSAC for the opportunity to discuss DERA white paper, 
18   "Free Trade Information in the Municipal Bond Market."  
19   This is a joint work with my other two DERA colleagues, 
20   Lewis Craig and Seung Won Woo, who are sitting on with 
21   the audiences.  
22             The purpose of the white paper is to describe 
23   the state of free trade information in the municipal bond 
24   markets.  Using dealer quote data obtained from several 
25   ATSs, the paper provides descriptive statistics of dealer 
0155
 1   codes for municipal bonds on ATSS, a need provides some 
 2   description of the types of municipal bonds for which 
 3   dealers provide quotes, and some of the characteristics 
 4   of dealers providing quotes on ATSs.   
 5             We also use ATS dealer quotes and TRACE data 
 6   from the MSRB to estimate municipal bond transaction 
 7   cost.  We believe that these descriptive statistics and 
 8   accompanying analysis improve our understanding of trade 
 9   information in the municipal bond market.  
10             Here are some of the main findings from the 
11   white paper. You find that only a small fraction of 
12   municipal bonds trades on a given day.  Among the bonds 
13   that traded on a given day, more than 50 percent of bonds 
14   have free trade information in the form of dealer codes. 
15     
16             We also find that codes are available for bonds 
17   on days without trades, suggesting that some dealer codes 
18   are less stale than the most recent bond prices.  And 
19   conditioner on a bond having dealer codes:  More than 90 
20   percent of bonds had one-sided quotes on an average day. 
21    And that's typically on the offer side.  
22             We also find that few municipal bonds have 
23   dealer codes on consecutive days.  To be specific, the 
24   number of days quoted per bond is 16 days, and that's 
25   pretty small compared to 67 business days we used in our 
0156
 1   analysis.  
 2             We also find that bonds with the dealer codes 
 3   have an average issue size more than 1.5 times larger 
 4   than bonds without dealer codes.  But from other bond 
 5   characteristics such as maturity, bond age, credit 
 6   ratings, or whether the bonds being geo bonds or revenue 
 7   bonds, show little relation with the quoting activity.  
 8             We also looked at some of the dealer 
 9   characteristics to providing quotes on ATSs.  We find 
10   that large dealers who trade a large number of bonds, and 
11   who trade higher per volume, are four times more likely 
12   to provide quotes on the ATSS compared to the smaller 
13   dealers, who trade a smaller number of bonds and also 
14   smaller purple.  
15             And next, some of our findings regarding the 
16   transaction cost.  We used the well-known measures of 
17   transaction costs such a quoted spread, effective spread, 
18   price improvement, and also the measures of markups to 
19   the available quotes.  
20             We first found that transaction costs vary 
21   across different types of bonds, but are consistently 
22   larger for customer trades than for inter-dealer trades. 
23    This finding is consistent with existing academic 
24   literature that have documented that customer trade 
25   transaction costs are higher than inter-dealer trade in 
0157
 1   in other fixed income markets.  
 2             When two-sided quotes exist, the majority of 
 3   customer trades, that's about 75 percent of customer 
 4   trades, are executed outside of the best-quoted prices 
 5   even though trade site does not exit quote price much 
 6   often.  As a contrast, the majority of inter-dealer 
 7   trades are executed very close to the best available 
 8   dealer quotes.  
 9             Next, when one-sided quotes exist, the majority 
10   of customer trades execute at a price worse than the best 
11   available dealer quotes.  That's also about the 75 
12   percent of customer trade.  We think that this is 
13   evidence for positive dealer markups in the customer 
14   trades in the municipal bond market in our sample period.  
15             We also find that smaller customer trades, that 
16   is, less than $100,000, have larger markups than larger 
17   customer trades.  And lastly, we found that dealer 
18   quoting activities, such as a code competition among the 
19   dealers, to be specific conditioner on a bond having 
20   quotes.  On average less than five dealers, quotes are 
21   given bond per day.  
22             We think that this findings highlights the 
23   relative lack of pre-trade information accessible to 
24   investors in the municipal bond market.  Thank you.  
25             MR. HEANEY:  Thank you, Abby.  This research 
0158
 1   will be not only good for setting the stage for the 
 2   panel, but as we consider initiatives in the future.  So 
 3   thank you very much.  
 4             As I mentioned, I'll turn it over to Lynn now, 
 5   who'll moderate our panel.  
 6   PANEL III:  CURRENT STATE OF PRE-TRADE TRANSPARENCY  
 7   IN THE U.S. MUNICIPAL SECURITIES MARKET 
 8             MS. MARTIN:  Thank you, Michael.  Thank you, 
 9   everyone for being here today.  We have an interesting 
10   panel to talk about this important topic.  
11             I think as the panel goes on, you will notice 
12   that the complexion of the municipal securities market is 
13   very different from the complexion of the corporate bond 
14   markets, which we've been talking about and where we've 
15   focused most of our attention on today, which is why I  
16   applaud the SEC for bifurcating the transparency topic 
17   into two different topics to look at the different issues 
18   focusing on the two different types of markets.  
19             The SEC staff study on pre-trade information 
20   that was included in our packs gave great, informative 
21   table settings for this afternoon's panel.   
22             Two things before I introduce our panelists.  
23   In the muni markets, near-realtime post-trade information 
24   is generally available through MSRB's EMMA tool.  And 
25   what we will find is that the pre-trade transparency is 
0159
 1   often made unavailable in a generic form, but is moreso 
 2   relied upon when it comes to the brokers making that 
 3   information available to their individual investors.  
 4             The one area where I would say the muni markets 
 5   share in common with the corporate bond markets is that 
 6   the rise of ATSs and technology enablement has increased 
 7   the level of pre-trade transparency in given years.  
 8             So let me first quickly introduce our 
 9   panelists, and then I'll turn it over to some Q&A that we 
10   had prepared to really comment on the state of pre-trade 
11   transparency in the market.  
12             Today we're honored to have joining us Bernie 
13   Costello, who is the head of muni trading at Morgan 
14   Stanley and Co. to provide the sell side perspective on 
15   pre-trade transparency; Ric Edelman, who is chairman and 
16   CEO of Edelman Financial Services, a leading independent 
17   financial advisor; Chris Ferreri, COO of Hartfield, Titus 
18   and Donnelly, a broker's broker; Ben Smelser, head of 
19   muni trading for Breckinridge Capital Advisors, which 
20   represents the institutional investor side of the 
21   equation; Dave Umeda from Charles Schwab, who is going to 
22   provide the online retail brokerage perspective; and Ron 
23   Valinotti, founder of Triangle Park Capital Markets, 
24   who's going to provide the market data perspective.  
25             As I call on each of you, if you could provide 
0160
 1   an overview of your role in the municipal markets 
 2   ecosystem, that would be helpful when framing your 
 3   answers.  
 4             So for the first question, I'm going to turn it 
 5   to Ben, Bernie, and Ric.  I have two general questions 
 6   for you to answer.  Number one:  Given the proliferation 
 7   of municipal securities in the market, with more than a 
 8   million securities available to be traded.  How do you 
 9   find information from a reference data perspective?   
10             We heard in the earlier panels that reference 
11   data is incredibly important.  Reference data, we mean 
12   what's available for trading -- coupons, duration, cost 
13   schedules, fundamental information such as that.   
14             That's incredibly important to the liquidity of 
15   markets.  And how do you also provide -- find details 
16   that enable you to provide what you deem to be fair value 
17   in the market?  When you're answering this question, 
18   please cover two different perspectives.   
19             Number one:  Are there different tools 
20   available to you, depending on the role you're taking in 
21   the markets, be it a price marker or a price taker?   
22             And number two:  Is this information 
23   technologically enabled, or do you have to rely on more 
24   traditional methods such as phone, information, even 
25   reading documents, things of that nature?  
0161
 1             So I'm going to turn it over to Ben.  Would you 
 2   mind starting off with your perspective?  
 3             MR. SMELSER:  Sure.  So Ben Smelser, 
 4   Breckinridge Capital Advisors.  As mentioned, I'm head of 
 5   municipal trading.  I manage 36 billion in separately 
 6   managed accounts, most of that through high net worth and 
 7   also institutional investors.  
 8             So in regarding what information is presented 
 9   to us, what bonds are available for trading, one of the 
10   biggest things that we've done is try to access and view 
11   as much of the market as possible.  And that's through 
12   talking to -- where there are 150 broker dealers, and 
13   seeing the offerings that they have available, the 
14   investigations that they're seeing from their clients and 
15   big ones being -- bonds that are being looked to be sold 
16   by the clients, and the parameters around the big 
17   "Wanted."  
18             So it's less the volume, more of the offerings, 
19   more of the market that's available for us to be seen, is 
20   the way that we've gone about it, trying to take a full 
21   view of what's available to be traded in the market, and 
22   aggregating that internally within our own systems.  And 
23   layering on top of that, our credit and our analysis of 
24   that particular bond and that particular credit.   
25             In terms of market price-taking, price-making, 
0162
 1   the primarily price-making function is through this "Bid 
 2   Wanted" process.  For bonds that are available to be 
 3   sold.  There is no price system input.  There's no 
 4   offering that's out there from the dealer.  These are 
 5   bonds that are offered by clients, or in some instance 
 6   dealers, that are offered to be sold, and the information 
 7   that's available is the CUSIP, the amount, and when 
 8   they're looking for the bid-buy.  
 9             So the way that we do this in a market where, 
10   as Lynn mentioned, there's close to a million CUSIPs 
11   available; there's 36,000 different issuers in that 
12   CUSIP.  They may not have traded in days, weeks, months, 
13   even years.  It is looking at a lot of comparable 
14   securities.  
15             So what we do is we'll look at comparable 
16   securities, being state of the bond that's being offered, 
17   the maturity, the different terms of the bond that's 
18   available.  And certainly in a market that is primarily 
19   tax exempt, the tax status of that bond in that state, 
20   and depending on what the individual investor or the 
21   institution that we're putting it in, it's an institution 
22   and their tax status, which is a big component of the 
23   price that we're going to evaluate and put our terms of 
24   what we think the price should be on that bond.  
25             MS. MARTIN:  Okay.  Bernie, would you mind 
0163
 1   offering the sell side perspective?  
 2             MR. COSTELLO:  Good afternoon.  Thank you for 
 3   having me.  Just a brief introduction.  My name is Bernie 
 4   Costello.  I've been working in the financial services 
 5   business since 1996, and specifically in the municipal 
 6   industry for the past 18 years, predominately as a 
 7   trader.  
 8             I started my career trading municipal bonds at 
 9   Citi.  I traded a variety of different sectors and 
10   products until I left in 2010.  I joined Barclay's.  I 
11   was asked to head their municipal securities trading 
12   desk.  I was head of trading there until 2015.  Presently 
13   I'm a managing director and head of municipal securities 
14   trading for Morgan Stanley.  
15             Just one caveat:  Because of the differences 
16   listening to the conversations that have happened before 
17   and the distinction that's made often between 
18   institutional and retail investors.  It's important to 
19   note that I'm an employee of MSCO, so most of my 
20   commentary and feedback is going to be to be based on the 
21   install side of the business.  That said, the municipal 
22   market is a retail-focused business, and I do have 
23   opinions, and am happy to offer insight.  I just don't 
24   want to misrepresent these opinions of Morgan Stanley 
25   Smith Barney.  
0164
 1             With respect to off-footprint in the municipal 
 2   space, I think it's -- can simply, without giving a big 
 3   commercial, can tell you Morgan Stanley's footprint is 
 4   significant.  We are a full-service broker dealer that 
 5   provides a whole suite of products, services, and client 
 6   solutions to our issuer accounts as well as our 
 7   institutional and retail counter parties.  
 8             With respect to the question that you just 
 9   answered from the sell side perspective, to a point 
10   there, it is a multitude of factors any trader needs to 
11   consider when pricing a security.  I'm going to think 
12   about the comments that I heard in some of the meetings 
13   prior.  
14             There are some differences; there are some 
15   similarities in terms of what we're looking at with 
16   respect to issues around transparency, issues around 
17   liquidity.  But there the similarities end, to the points 
18   that we made.  We have over a million CUSIPs in our asset 
19   class.  
20             So if you're talking about pricing a segment of 
21   our market, which is common, which are large, liquid 
22   CUSIPs and highly-rated credits in stable interest rate 
23   environments, where supply and demand technicals match a 
24   trader.  And that is a good component of our asset class. 
25    A trader can evaluate, fairly easily and consistently, 
0165
 1   just by looking at a handful of data points.  
 2             The most important data point I look at, and I 
 3   think most of the traders look at, certainly, when you 
 4   start thinking about pre-trade price transparency, when 
 5   you speak to someone in munis, they think you're talking 
 6   about post-trade data, how available that is.  It took me 
 7   a couple days to figure out that that's exactly what this 
 8   topic was going to be about -- excuse me -- with respect 
 9   to markets as opposed to historical trade data.  
10             The other things we look at would be inter-day 
11   moves in more liquid interest rate products like 
12   Treasuries, or reference pricing scales like MMD, in 
13   addition to some of the eval prices that are provided by 
14   a variety of data services coupled with some of the 
15   markets.  
16             In those environments, you could readily price, 
17   fairly consistently and accurately, municipal securities. 
18    That said, as we know, there's a huge component of our 
19   market that doesn't trade every single day, and they're 
20   definitely more in-depth analysis when you start talking 
21   about credit, when you start talking about sizes that may 
22   not be wrapped around a million, where you start talking 
23   about environments, and it seems consistently supply and 
24   demand balances differ.  And you got the whole point of 
25   the market, which is shifting one way or another.  
0166
 1             There's those measurements of a municipal's 
 2   price, or how to evaluate those risks in a market, are 
 3   more subjective in nature and not as objective, as 
 4   observed in the data points that we mentioned earlier.  
 5             Those data points, moreso now than ever, are 
 6   readily available to all market participants.  The issue 
 7   that we have is that there is no consistent format.  
 8   There is no consistent format to view those data points. 
 9     
10             I can tell you, having traded municipal 
11   securities at three different firms, one of the first 
12   challenges you face is once you set up your workstation, 
13   is trying to get familiar with all the different systems 
14   that you have -- where you source your MSRB data, what 
15   you're looking at to observe quotes in Treasuries and MMD 
16   and some of the other things that you use the price of 
17   securities, the logistics around that; the systems that 
18   you have to use the technology is you need to each -- 
19   users, you need to each be firm, and market participant.  
20             So while I'm confident an decision state that 
21   the information has never been better and readily 
22   available, the manner in which you mine that data, 
23   unfortunately, for better or for worse, it just is not 
24   consistent and is usually specified to the end users.   
25             The systems that a salesperson on the 
0167
 1   institutional side may use to mine this data is going to 
 2   be specific to their role.  But I'll use it.  Someone 
 3   that's managing a desk of risk is specific to my role, to 
 4   the point I'm an FA who's selling not just munis but a 
 5   variety of different products.  
 6             Their front-event system is going to be 
 7   distinctly different.  While they'll be able to source 
 8   the same data, the systems that they are using are 
 9   different.   
10             I'm happy to go into more detail; I could 
11   certainly pass.  
12             MS. MARTIN:  Okay.  Thanks.  
13             Ric, from your perspective as an independent 
14   financial advisor, how do you source this data?  
15             MR. EDELMAN:  Thank you very much for the 
16   opportunity to be here with you today.  I've been in this 
17   industry for 32 years, and have built what is now one of 
18   the largest independent investment advisory and financial 
19   planning services in the nation.  We serve 36,000 clients 
20   across 43 offices nationally.  
21             And what makes us most unique in the financial 
22   advisory field is that our clients tend to be mass 
23   affluent.  So unlike most folks here today and in our 
24   industry who are dealing with the high net worth 
25   marketplace, the family office market -- we tend to be 
0168
 1   dealing with everybody else.  
 2             Our average household account is half a million 
 3   dollars, and a great many, more than two-thirds, have 
 4   substantially below that.  Our household minimum is only 
 5   $5,000.  So we're dealing with, truly, the retail client.  
 6             And because of that, and the fact that's the 
 7   way it's always been in our practice, long ago we punted 
 8   on the notion of buying individual munis.  It simply 
 9   isn't a level playing field, from our perspective.  And 
10   that's why we're using ETFs and low-cost mutual funds 
11   instead.   
12             The pricing points that Abby described for us 
13   in this new white paper, I think, really articulate 
14   perfectly the concerns that we have discovered in an 
15   active practice of our inability to get effective pricing 
16   information prior; the ability to understand exactly what 
17   the spreads ought to be, let alone what they are, being 
18   presented to us.  
19             And because of that fact, because we have 
20   chosen to go the fund route, we're not trading bonds 
21   directly ourselves.  Instead, we clear predominately 
22   through TD Ameritrade, and when we are faced with a muni 
23   bond trade, we call their trading desk and we let them do 
24   it.  
25             What we have relied on, in addition to the data 
0169
 1   that we've just heard earlier from Abby, market group did 
 2   a similar study a couple of years ago and found that 95 
 3   percent of geo bonds traded less than 25 days out of the 
 4   year.  Ninety percent of revenue bonds traded less than 
 5   50 days out of the year.  
 6             And because there's a million-plus issues out 
 7   there, and because the overwhelming majority of these 
 8   bonds are being purchased by retail consumers who never 
 9   trade them, the ability to get effective current pricing 
10   information is extraordinarily difficult.  
11             And we discover that most retail consumers, in 
12   our experience, aren't even aware that muni bonds can be 
13   negotiated, that just because a broker quotes a given 
14   price, that's not in stone.  That price can be different 
15   at another broker for the exact same bond, the exact same 
16   time.  The pricing can be different.  
17             And so what we do before we go to the trading 
18   desk at TD, we'll look at EMMA.  The problem is, as has 
19   already been cited, is that the data can often be old 
20   because EMMA's really good for trades that have happened. 
21    But since most munis aren't trading frequently, when you 
22   do look at EMMA, the data can be rather old because there 
23   hasn't been a recent trade.  
24             Another source that is pretty nifty is 
25   bondview.com, a commercial space that pulls data from 
0170
 1   MSRB.  It also pulls data from other pricing services, 
 2   such as Thomson Reuters.  And for thinly traded bonds, it 
 3   does have a fair market price.  
 4             But they categorize them in three different 
 5   categories -- one where there is recent market data, one 
 6   where there really isn't, and one where there's no data 
 7   of any kind and they're kind of just making a guess.  
 8   It's rather shocking to us that a marketplace that is so 
 9   large that it's predominately retail-focused has such a 
10   poor level of pre-trade pricing compared to virtually any 
11   other security that might be contemplated by a consumer -
12   - a stock or even corporates.  Certainly in the municipal 
13   markets or what have you.  
14             And so that's why we have tended to emphasize 
15   the fund route, which isn't necessarily a perfect 
16   solution but we consider it, in this environment, for our 
17   kind of client, a much better choice.  
18             MR. SMELSER:  Could I just take two seconds to 
19   respond? 
20             MS. MARTIN:  Sure.  
21             MR. SMELSER:  I don't mean to put you on the 
22   spot and be thought guy, but I'm going to be that guy.  
23             (Laughter.) 
24             MR. SMELSER:  I do think there's a lot of 
25   misconceptions.  I think the telling -- a statement that 
0171
 1   you just made in your commentary -- was a long time ago 
 2   you made a decision to move because of things that I 
 3   think a long time ago were very valid criticisms of the 
 4   investment market.  
 5             But the modern municipal market is nothing like 
 6   you just described, from my perspective.  I definitely 
 7   have some questions about it.  I've read this white 
 8   paper.  I would argue that some of the data points -- I 
 9   mean, you referenced specifically some data sets.  
10             When were they dated?  I don't mean to put you 
11   on the spot, but where was that in 2014?  
12             MR. COSTELLO:  2014.  All right.  The biggest 
13   glaring issue I have with this report and those comments 
14   is that we've had groundbreaking, sweeping changes in 
15   regulation with respect to municipal best acts.  
16             I will argue my 20 years of trading municipal 
17   bonds, there's not one single regulatory change that has 
18   fundamentally changed how retail clients access bonds and 
19   liquidity in the market.  And our data points in this 
20   report are scathing.  Scathing.  Some of the spreads you 
21   talk about, 80 basis points, 7/8ths of a point, nothing 
22   like I experienced in the markets today.  
23             Some of your comments around -- I take it you 
24   had multiple examples of where dealers are executing with 
25   traders at levels that are well below the advertised 
0172
 1   price on an ATS.  This is explicitly illegal in the 
 2   current environment.  I mean, maybe there are smaller 
 3   deletes and other things that don't apply the rules to 
 4   the standards that Morgan Stanley does, but I could not 
 5   do that in today's world.  
 6             Would there be multiple catches and failsafes 
 7   that would prevent such activity.  So the municipal 
 8   market has the exact same concerns that o other markets 
 9   have had with respect to liquidity, with respect to 
10   transparency, with respect to efficiency.  
11             But the way the municipal market has chose to 
12   go about fixing that is to put the onus on the dealer 
13   community because it's such a widely held retail market. 
14    We're not going to force market disclosures or 
15   participants who I know at times get the blame of being 
16   unsophisticated.  
17             They're not unsophisticated individuals.  
18   They're professionals.  They're just not profile traders. 
19    All right?  They run businesses.  They're retirees.  
20   They're perfectly capable of understanding.  But instead 
21   of putting the onus on them to hammer them with all this 
22   information and have them decide what's the best price.  
23              We're at the industry professional zoo, and 
24   we're forcing competition into the market.  So I'm just 
25   hesitant.  I respectfully disagree, and I would argue 
0173
 1   that I'd be very interested to see what the data shows 
 2   post-best execution standards because I can tell you from 
 3   my experience, having marshalled our firm through the 
 4   implementation of that, it is decidedly different.  
 5   Decidedly different.  
 6             MR. EDELMAN:  I won't be able to speak 
 7   specifically to the spread conversation that Bernie has 
 8   raised because we're not trading bonds ourselves.  So I 
 9   can't speak to that point.  
10             But what I can speak to is to why we're using 
11   funds as opposed to individual immunities.  Again focus 
12   on the fact of the nature of the retail client we're 
13   dealing with.  If someone's only got a couple of hundred 
14   thousand dollars in their total investment portfolio, 
15   having them purchase an individual issue with 50- or 
16   $100,000 is an incredible lack of diversification.  
17             And that means we're going to need to buy 10 or 
18   20 issues.  A need when you're doing that, you're only 
19   buying $10,000 or $20,000 per issue.  You're at the 
20   basest of the retail level, and the spreads are going to 
21   be at their maximum.  And this is in addition to the fee 
22   that the advisory practice is going to tack onto this.   
23             It just makes it economically not at all in the 
24   client's best interest, and especially when you're adding 
25   onto this the fact that there isn't any absolute 
0174
 1   confidence in the pricing that we're getting.  If in fact 
 2   we can get an effective price prior to the trade, all of 
 3   that adds up to, from a financial planning perspective, 
 4   why we're focusing on funds as opposed to individual 
 5   securities.   
 6             We think that the individual retail investor, 
 7   who is generally buying munis, is not nearly as 
 8   sophisticated or as experienced as Bernie is suggesting. 
 9    At his end of the market, in the world of Morgan 
10   Stanley, we are dealing with higher net worth 
11   individuals.  Maybe that's the case.  
12             But the kind of folks that we interact with all 
13   the time, the broad cross-section of the American public, 
14   it's a canard to suggest that these folks have 
15   sophistication and experience to be able to effectively 
16   purchase securities in this marketplace the way that the 
17   state of the art is at present.  
18             And I think that they're outclassed.  They're 
19   outgunned.  They're out-sophisticated by the industry.  
20   And the situation definitely needs the kind of 
21   improvement that this panel today is really describing of 
22   greater transparency in the marketplace.  
23             MS. MARTIN:  Okay.  I want to turn it over to 
24   the brokers to weigh in with their views.  
25             MR. HEANEY:  Because we're always in the middle 
0175
 1   of these things. Right? 
 2             (Laughter.)  
 3             MS. MARTIN:  To weigh in with their views on 
 4   this topic.  So Chris and David, would you mind 
 5   commenting on how you make information available to your 
 6   customer base?  Specifically again, what available 
 7   securities there are to be traded in the market, any 
 8   information around pricing, and does that vary based on 
 9   the type of participation you're informing, be it the 
10   retail or the more institutional type investors.  
11             Chris, please.  
12             MR. FERRERI:  Great.  Well, thanks.  And thanks 
13   for inviting me to participate in this.  It's a -- 
14   thanks, Ric.  
15             It's interesting.  I spent 35 years as an 
16   inter-dealer broker.  And when I came to munis, I found 
17   that they were called broker's brokers.  Right?  It was 
18   more of a defined role of what a municipal bond broker's 
19   broker is.  
20             HTD, Hartfield, Titus and Donnelly, has five 
21   offices around the country.  We're a national broker's 
22   broker.  We have about 80 human brokers that intermediate 
23   trades, not only between dealers, which I think is why 
24   we're not quoting to dealer brokers, but between SMMPs, 
25   sophisticated municipal market professionals.  
0176
 1             We have about 400 firms that we do business 
 2   with.  And back in the day, we distributed our prices 
 3   through closed circuit television, green screens.  These 
 4   were sometimes live prices, sometimes the aged prices, 
 5   sometimes indicative of prices.  There was no requirement 
 6   to get them up there.  
 7             What I've learned since coming to the muni 
 8   market is the tremendous role of the MSRB, as played, in 
 9   making this market stronger -- things like best 
10   execution, to your point, Bernie.  Requirements of the 
11   broker's broker, you  have levels that would be expected 
12   to be on a certain bond -- predetermined levels that a 
13   broker would say, "This might be with the bonds you td."  
14             It's a fine balance between talking to a client 
15   and not coaching them on a bid walk to process.  But 
16   again, it's unique to the muni market where a seller of 
17   bonds sees bonds out for bid.  No obligation to td.  No 
18   requirement from people to respond.   
19             But that process generates a series of bids for 
20   the product.  The voice brokers enhance that by knowing 
21   they get paid when they get a trade done.  So they try to 
22   get as many bids as they can.  In 2002, HTD actually 
23   posted, at the end of the td, the depth of bids when it 
24   resulted in a trade.  So there was pre-trade information 
25   on our screens, on those little green screens.  As 
0177
 1   technology evolved from closed circuit TV to YouTube, we 
 2   went on to  a web-based product called HTD Online.   
 3             It's important for us to underscore something. 
 4    I'm also here in another capacity.  We took our trading 
 5   platform and in '24 scene, decided to license it.  So we 
 6   formed a separate company, separate management, and 
 7   developed from the MuniBrokers.  
 8             So unique to the fixed income market in the 
 9   United States, there are 15 competing broker's brokers  
10   on one platform.  So someone like Varney can log in once 
11   and see the markets generated by 15 competing broker's 
12   brokers.  There's no other product of fixed income where 
13   this exists.  
14             That platform allows traders to go into the 
15   system, see live indicative and aged prices, so the 
16   system will show you prices that come in electronically. 
17    MuniBrokers has more than 60 electronic feeds of bid 
18   wanteds and live executable prices.   
19             We have, on average -- just throw some numbers 
20   at you again because I think it's important for your 
21   question, Lynn -- around 5600 bid wanteds on a daily 
22   basis come into the system.  More than 200,000 offerings 
23   are in the system at any given time.  So there's a 
24   plethora of market data that comes into these systems.  
25             And we don't tweak it or customize it for a 
0178
 1   certain trader's view.  There are all sorts of search 
 2   capabilities, alert capabilities.  It's tied to EMMA. 
 3   It's tied to MSRB.  MuniBrokers does the post-trade 
 4   reporting.  There's a robust G-43 compliance regimen 
 5   built into it. If you want to be in the broker's 
 6   brokerage space, you just license MuniBrokers and you're 
 7   up and running.  
 8             To the point of the information being 
 9   available, MuniBrokers is a $200 per month subscription. 
10    So this is not limited to high net worth individuals who 
11   may say, "I need to know where the muni market is."  We 
12   recognize the shift to ECFs.  In 2017, we actually 
13   contracted an ETF database so you can look at ETF 
14   holdings against the inventory of bid wanteds and 
15   offerings in our system.  
16             So it's remarkable to me.  I'm a Treasury guy; 
17   I spent 30-odd years in the U.S. Treasury market.  And 
18   coming to the municipal market and seeing the advances in 
19   the muni market on probably the least liquid product in 
20   fixed income in the U.S., probably the least frequently 
21   traded as a percentage of CUSIPs outstanding -- and the 
22   industry found a solution to that problem.  
23             This may not work for corporates.  It wouldn't 
24   work for Treasuries.  It's a different market with 
25   different problems.  And I think that's one of the key 
0179
 1   things I'd like to bring out as we talk to the committee. 
 2    I echo your comments, Lynn, that it's great that this 
 3   has been separated out because just recognizing at a 
 4   highest level fixed income in equities -- got it -- and 
 5   Treasuries aren't, corporates aren't, munis.  
 6             So I think it's really important to underscore 
 7   that and to see the advances companies are making.  I 
 8   think that answers your question.  
 9             MS. MARTIN:  Yes.   
10             MR. FERRERI:  Thank you.  
11             MS. MARTIN:  David?  Would you also mind in 
12   your answer talking on the bids wanted process, too?  
13   Because it's come up a couple of times.  
14             MR. UMEDA:  Sure.  Sure.  Well, first let me 
15   start by saying thank you to both the FIMSAC and the 
16   committee for invitation to be here today and discuss 
17   this important topic.  As a whole at Schwab we believe 
18   that transparency benefits retail investors, and 
19   certainly are in support of efforts to advance the fixed 
20   income market structure in that direction.  
21             My name is David Umeda, as Lynn said, and I'm 
22   responsible for a part of the fixed income trading 
23   business at Schwab.  More specifically, that includes the 
24   liaison trading desk, whose primary focus is on customer 
25   facilitation in some of our core fixed income products as 
0180
 1   well as our ATS relationships.  And I also recently took 
 2   over the fixed income risk management and operations 
 3   businesses.  
 4             I've been in this business for the past 13 
 5   years, the last 10 of that being in various fixed income 
 6   roles at Schwab.  And I think to help kind of describe 
 7   the retail investor perspective, it would be helpful to 
 8   first provide just a little bit of context about Schwab 
 9   as a whole and then, more specifically, our fixed income 
10   offering to start to drill down into some of the more 
11   specifics.  
12             But Schwab services a wide range of 
13   comprehensive financial products and services to millions 
14   of customers.  We service both retail investors as well 
15   independent RIAs.  Our fixed income platform supports 
16   wide access to the market and allows clients to interact 
17   with us completely on a self-directed basis online or 
18   through the assistance of one of our fixed income 
19   professionals or some hybrid there between.  
20             With regards to individual bond transactions, 
21   we're a significant participant in the retail fixed 
22   income markets and execute somewhere between 3- and 4,000 
23   trades a day in the secondary space, about 20 percent of 
24   that or so related to municipal securities.   
25             In certain instances we have a trading desk 
0181
 1   that also facilitates customer order flow, and in small 
 2   cases, may make markets in various securities to augment 
 3   the liquidity received from external sources.  But it's 
 4   important to note we're not an investment bank and we 
 5   don't have a large proprietary trading business.  
 6             So more specifically, around our platform, our 
 7   platform seeks to provide a transparent view of the fixed 
 8   income markets through connections with many of the 
 9   leading retail ATS trading platforms as well as 
10   connectivity with individual dealer inventories and our 
11   own inventory.  
12             This architecture provides a wide access to 
13   bond quotes with significant breadth, and where it 
14   exists, depth to our customers, and provokes price 
15   discovery and price competition for their benefit.  On a 
16   typical day, we provide access to tens of thousands of 
17   unique quotes from over 200 different dealers.  
18             A significant portion of that aforementioned 
19   liquidity is available directly for customers to interact 
20   with online.  This permits clients to look at individual 
21   muni offerings on a given day, to search for available 
22   markets, to compare the research, and in many cases place 
23   orders versus those quotes.  
24             As clients search those bonds, our system 
25   automatically returns the best quote for a particular 
0182
 1   security based upon the client's parameters.  And for 
 2   clients that wish to see depth, they can also expand that 
 3   view to see depth of market as it exists.  
 4             At Schwab, we also seek to be as transparent as 
 5   possible about our fees, and that's why we've published a 
 6   straightforward, low-cost, fixed income fee schedule for 
 7   quite some time.  To that extent, we've also chosen to 
 8   display our quotes that are available on a net basis so 
 9   customers know up front the total price of the 
10   transaction.  
11             On top of our wide quote architecture, we 
12   provide additional pre-trade transparency in the form of 
13   individual bond trade history information via access to 
14   EMMA data, a robust set of descriptive data, evaluated 
15   pricing, important issuer documents, bond rating 
16   information, all of which can be useful in helping 
17   individual investors value and price individual 
18   securities.  
19             Given our wide access, we also augment this 
20   system with tools that seek to identify potentially 
21   egregious or off-market prices and remove them from 
22   clients' view.  And we also alert clients at time of 
23   order entry if a particular bond price deviates from any 
24   recently reporting trade history by a significant margin. 
25    And certain transactions may actually be routed to a 
0183
 1   trader for additional review prior to us executing the 
 2   transaction.  
 3             So in terms of the question about how that data 
 4   is available, everything I've described that is made 
 5   available online is really available to both the RA 
 6   population and the end clients.  It doesn't differ 
 7   significantly between the two universes.   
 8             We also provide our fixed income sales 
 9   professionals with some additional tools to assist in 
10   valuing, such as Bloomberg, comparable bonds, information 
11   and access to the trading desk.  Traders have an 
12   additional set of information such as things that we 
13   referred to earlier in the panel -- NMD scales, things of 
14   that nature.  
15             In regards to the auction process, we have 
16   found that to be a pretty well-functioning part of 
17   today's market.  So for clients that wish to sell a long 
18   position prior to maturity, while we've described in this 
19   panel and in the previous panel much of retail bond 
20   transactions or typically buy and hold, there are 
21   instances where investors would like to liquidate a 
22   position.  
23             We take a bid wanted or RFQ framework and send 
24   those requests online either initiated by the customer or 
25   through the assistance of one of our sales professionals 
0184
 1   to a wide range of ATS platforms -- in instances, 
 2   broker's brokers, individual dealers.  Potentially our 
 3   own desk may be willing to provide a price, and on 
 4   average, we see somewhere between 10 and 15 bids per 
 5   request.   
 6             Given some duplication that may by its very 
 7   nature occur across platforms, there's probably some 
 8   duplicity there.  But we also observe a significant 
 9   amount of unique liquidity through that wide distribution 
10   as well.   
11             To the extent there's any posted bid side 
12   liquidity coming in from those sources as well, we'll 
13   aggregate that in addition to the manual responses we 
14   receive.  And as we collect that information, we display 
15   both the best price, the cover bid, and the total number 
16   of bids received back to the customer.  And if those 
17   levels meet our best execution standards, we will 
18   actually enable the price for the customer to interact 
19   with or the salesperson directly online.  
20             Overall, we see a pretty high conversion rate 
21   on those responses.  But in terms of the actual data 
22   available on the bid wanted option process, it's very 
23   similar to what customers see on the purchase side as 
24   well.  
25             MS. MARTIN:  Okay.  Thank you.  Turning to Ron 
0185
 1   to provide something near and dear too many heart, which 
 2   is the vendor perspective, I guess we've always had the 
 3   strategy that transparency really adds liquidity to 
 4   markets, and tools for transparency that vendors can 
 5   provide really will beget liquidity in markets.  
 6             And in my seat, I have the privilege of leading 
 7   ICE Data Services, which is a leading provider of pre-
 8   trade transparency tools around evaluated pricing 
 9   reference data, as well as various analytics which enable 
10   folks to transact in the U.S. municipal securities 
11   market.  
12             Ron, what are you seeing from the vendor 
13   perspective as to what is available in the market for 
14   firms such as those we've heard from today to either get 
15   information about reference data, get information about 
16   value of securities, particularly given the amount of 
17   securities in this market?  
18             MR. VALINOTI:  Well, thank you, Lynn.  Thank 
19   you for having me.  My name is Ron Valinoti.  I am a 
20   founding partner of Triangle Park Capital Markets Data, 
21   which is a small information company.   
22             More importantly, I'm here today to talk about 
23   a company that I manage on behalf of 10 inter-dealer 
24   brokerage NATS partners who formed a consortium company 
25   back in 2013 to address the concerns raised in the SEC 
0186
 1   white paper to Congress in 2012 when then acting Chairman 
 2   Walter had highlighted the need for greater market price 
 3   transparency and the availability or the lack of 
 4   availability of pre-trade data.  
 5             So Lynn, to answer your question with regard to 
 6   the vendor community and what's going on with regard to 
 7   this type of data, I could say that half of the customers 
 8   that we've been able to acquire over the last few years 
 9   are information providers.   
10             Triangle Park and MBIS are small businesses.  
11   Our strategy to be able to make the data broadly 
12   available was to work through partnerships, and those 
13   partnerships are with firms like ICE Data Services, who 
14   integrates our data into their manufacturing of 
15   evaluation opinions.   
16             And they distill what is a big, complex data 
17   set that we aggregate on a day in/day out basis into -- 
18   synthesized into opinions of value that they produce and 
19   market broadly to the municipal investor and dealer 
20   communities.  
21             I think there was a comment earlier about the 
22   state of market transparency since some key regulations 
23   have gone into effect, in particular the one Bernie cited 
24   regarding best execution.  We aggregate from our broker's 
25   broker and ATS partners almost 400,000 quotes on a daily 
0187
 1   basis, and that makes up thousands of secondary market 
 2   offerings and the bids on bid wanted auctions that are 
 3   conducted by those platforms.  
 4             And we've seen that grow almost double in size 
 5   since these regulations became effective.  Both that and 
 6   the newest markup disclosure, I think, have two qualities 
 7   -- one, have enforced greater canvassing of the markets 
 8   for best price by the dealer community who participate on 
 9   these platforms.  
10             I also think that they're also providing 
11   greater efficiency in the trading of fixed income and 
12   municipal securities.  One of the challenges that the 
13   market uniquely has for municipals is it's geographically 
14   dispersed.   
15             There are a great degree of specialization 
16   around the country in the regional firms.  Those firms 
17   don't necessarily have access or have had access to 
18   either systems or technologies to be able to digest and 
19   utilize all this data, whereas the wire houses, the large 
20   New York-based investment banks, have had the technology 
21   to be able to put in place the infrastructure.  
22             What we're observing is that changing as things 
23   progress, as firms have no choice, especially in trading 
24   a municipal retail book of business, but to employ 
25   technology to make those traders more effective and 
0188
 1   efficient at what they do.  And we see it in the pricing. 
 2    I think that we've seen observable spreads contract in 
 3   the data sets that we aggregate from bid to ask to being 
 4   much more reasonable than early day.  
 5             So with regard to that, we package that data up 
 6   in its raw form for firms who have the infrastructure and 
 7   the technology to consume it.  But the big issue is, many 
 8   more firms in the municipal market need assistance.  They 
 9   need somebody to help package that solution to fit within 
10   a workflow.  
11             And one of the key areas of work flow where 
12   we've helped work with firms who are trying to deliver 
13   solutions is in municipal trade compliance.  There are a 
14   number of firms who are trying to offer solutions for 
15   best execution, markup disclosure, and a variety of other 
16   obligations, to perform with more diligence and more 
17   demonstrable or documented diligence so they can 
18   basically present to those who care, an auditor or 
19   regulator, risk management in their firms, that they are 
20   following solid business practices as it relates to the 
21   trading or underwriting or sale of municipal securities 
22   to clients.  
23             With that, we think there is a number of 
24   solutions that are in the development cycle, I'll say, 
25   that will help prefer greater transparency in the 
0189
 1   marketplace.  I think the challenges for a lot of firms 
 2   who have individual investors as a key part of their 
 3   client base is the packaging of that content.   
 4             It's a big, complicated data set.  I've heard 
 5   people use the term "comparable securities," which is not 
 6   a trivial task.  Just as a matter of background, I've 
 7   been on the information side of the business for very 
 8   long time, more than 18 years, Bernie -- more like double 
 9   that -- almost exclusively on the information side of the 
10   business.   
11             And most of my background is in modeling these 
12   securities, in particular munis.  So there are these 
13   third party valuation services, providers; they provide 
14   their best effort to take what is a difficult data set 
15   that is only now just becoming available to the 
16   marketplace.  MBIS has only been around the last five 
17   years.  
18             In 2008, when the financial crisis hit, I was 
19   at Bloomberg.  And our project there, BVAL, was in 
20   desperate need of more market data.  It was a highly 
21   algorithmic approach to valuing municipal bonds, and it 
22   required lots of data to feed those algorithms.  
23             And I went out and convinced management to give 
24   me a checkbook, and we went out and we bought $15 million 
25   worth of data from municipal bond dealers.  And the two 
0190
 1   things that were telling is 2008 highlighted the need for 
 2   greater transparency and more dealer input because I like 
 3   to say the only ones who really price bonds are traders 
 4   on dealer desks.  Everybody else responds and reacts to 
 5   those decisions that they make.  And with that, we've 
 6   tried to harness and harvest that data.  And what our job 
 7   is now is to be able to provide it in useful packages to 
 8   the marketplace.   
 9             We have done some interesting work.  We 
10   attempted to create a benchmark out of the data, and what 
11   that involved was the creation of a mini-aggregate index 
12   from which we could distill and fit a yield curve through 
13   each day.  We take an automated rule set and we apply it 
14   to our market data.   
15             We gather about 25,000 of those 400,000 quotes, 
16   and it applies to a basket of around 4,000 securities.  
17   And we're able to fit a yield curve through what I will 
18   refer to as the actively quoted municipal investment-
19   grade universe.   
20             And that yield curve produces very good 
21   results.  We believe it tracks the changes in rates in 
22   munis throughout the day.  We believe it comports with 
23   the principles for benchmarks that are being established 
24   at this time by an industry standards organization called 
25   IOSCO.   
0191
 1             And we think that that kind of transparency, 
 2   where we can show our work as to why we think the long 
 3   end of the curve, adjusted by a few basis points -- we 
 4   can show all that detail and provide all of that in the 
 5   services that we packaged up.  And unlike 400,000 quotes, 
 6   this is a much easier package for most firms to consume.  
 7             So we're working with the industry trade 
 8   journal, The Bond Buyer, to make this information and the 
 9   bid and offer data that we have available to what they 
10   call the Bond Buyer Workstation, which is making this 
11   data available to a large number of subscribers who don't 
12   have access to trading desk information, bid wanted and 
13   offerings.   
14             And the price for that subscription, as Chris 
15   pointed out, we're very reasonably priced.  We charge 
16   $100 a month for that subscription to the yield curve and 
17   the market data throughout the day.  So we think we're 
18   making a lot of progress towards broadening market 
19   transparency.  The real challenge is to make it available 
20   for the individual investor in a consumable form.  
21             One of the byproducts of transparency, a lot of 
22   folks in the dealer community -- and I'm not saying 
23   anybody here in the room has represented this -- but 
24   transparency and more information flow has never helped a 
25   trader make a living or a salesman make a living.  It's 
0192
 1   always been bad for the fixed income markets from a 
 2   perception perspective.  
 3             MS. MARTIN:  Okay.  Thanks, Ron.  I want to 
 4   just have one more question, and I know we're running 
 5   short on time.  
 6             MR. VALINOTI:  Sure.  I'm sorry.  
 7             MS. MARTIN:  Thank you for your perspective.  
 8             Before I turn it over to my FIMSAC co-
 9   panelists, the Commissioners, as well as the SEC staff, I 
10   just wanted to hear -- and this is really for anyone on 
11   the panel -- are there any specific sources of data, 
12   including pre-trade price information, which in your 
13   believe would improve the liquidity of muni securities if 
14   they were made publicly available?  And again, that's 
15   open to anyone on the panel.  
16             MR. FERRERI:  I'll just -- again, a perspective 
17   here -- is that it is publicly available.  This is 
18   information that, in a very small market -- Hartfield's 
19   also an ATS.  We're a broker's broker.  This registers 
20   the NCS because we publish two-sided markets.  A very low 
21   price point to get that data shows that it is publicly 
22   available.   
23             I think the question, Lynn, becomes:  To whom 
24   is that data useful?  If I am going to buy 10,000 munis 
25   through Ric Edelman, is a quarter of a point the right 
0193
 1   markup to pay for that?  And the markup disclosure rule 
 2   is not part of this discussion, but that's another key 
 3   element, from a regulatory perspective, of this sort of 
 4   data that's important.  
 5             But I do think the availability of the data, 
 6   the pre-trade transparency, other than the muni market 
 7   exists and is robust.  And I think it's important for 
 8   everybody here to realize that.  And maybe because we 
 9   haven't done a great job of marketing that product 
10   because it is such a small part of the industry, it might 
11   serve to be a good model for a regulator that gives a 
12   non-prescriptive mode of regulatory rules:  Here's what 
13   we're trying to accomplish.  
14             The broker's brokers were not thrilled that we 
15   would be the ones to come up with the parameters for 
16   pricing.  The broker doesn't want to get in the middle of 
17   a bad trade for either party.  But we took that on, and 
18   it's part of our process.  And I think it's worked well. 
19    I think the MSR has been a good partner in that.  
20             MR. EDELMAN:  I agree with Chris.  I think that 
21   the issue here isn't so much about data partly because 
22   data simply gets sale, simply because of the lack of 
23   trading of these million issues that exist.   
24             So the issue is not so much one of data 
25   accessibility, although that clearly is an issue to some 
0194
 1   degree.  The bigger issue is the rules around which the 
 2   trading itself occurs and the disclosures of that 
 3   trading.   
 4             It to me seems inappropriate for a consumer to 
 5   be purchasing a security and not knowing how much money 
 6   the broker or the dealer is earning in that trade, not 
 7   knowing how much they're paying for it, and certainly not 
 8   knowing in advance.  It's kind of astonishing that in 
 9   many cases they never know how much they ended up paying 
10   for the trade, even after the fact.  
11             Second is the fact that these are negotiable 
12   securities, and most consumers are unaware of the fact 
13   that prices can be different on the same issue at the 
14   same time from two different resources.  And therefore, 
15   we need, I think, parameters over which those markups can 
16   be.   
17             Is there a limit so that it is more -- it is 
18   less likely for the consumer who went to, I'll say, the 
19   "wrong" broker to execute the trade that they ended up 
20   paying substantially more than if they had gone to a 
21   competitor?  
22             So I would argue for -- the very nature of this 
23   conference -- transparency in advance of the trade and a 
24   limitation as to the variance of what those trade costs 
25   might be from one organization or another.  Those two in 
0195
 1   place helps to provide a substitute for data which may or 
 2   may not be available.  
 3             MS. MARTIN:  Okay.  Thank you.  
 4             MR. UMEDA:  Lynn, could I just jump in real 
 5   quickly, follow on on that topic? 
 6             MS. MARTIN:  Absolutely.  
 7             MR. UMEDA:  I think, from the trade 
 8   perspective, as I highlighted in some of my comments, 
 9   there is and has been a fair amount of pre-trade 
10   transparency from our perspective for our retail 
11   customers in municipal securities, cost being one of 
12   those factors.  We've published that fee schedule for 
13   quite some time so customers know in advance.  
14             And I think the other key point here being that 
15   experiences differ across, obviously, the street in 
16   different broker dealers that service retail investors, 
17   but the ATSs have done a significant job pooling retail-
18   size municipal offers and facilitating, for the bid 
19   process, to where customers have a pretty significant 
20   degree of pre-trade transparency in the form of breadth 
21   of quoting markets to draw comparisons to help in price 
22   formation process.  
23             Certainly there's some ideas that have come up 
24   about other ways to help in how you value those illiquid 
25   parts of those securities.  But from our perspective, 
0196
 1   clients have a pretty significant degree of access in 
 2   current day.  
 3             MR. HEANEY:  So at this point I'd like to open 
 4   it up to the committee for any thoughts for the panelists 
 5   and any questions.  And we'll start with Gilbert, please.  
 6             MR. GARCIA:  Thank you.  I think this is for 
 7   you, Ric, if you don't mind.  
 8             If you look, is there anything -- ignoring 
 9   diversification, is there anything in the whole MiFID II 
10   framework that you would like to see here that would lead 
11   you to start utilizing muni bonds more, and to get away 
12   from ETFs?  
13             MR. EDELMAN:  Credit quality is an issue.  It's 
14   related to diversification.  Munis are, of course, very 
15   widespread over their credit quality, and part of the 
16   concern we have is the state of the underlying issuers, 
17   and particularly in the pension underfunding situation 
18   that we have in so many states around the country these 
19   days.  
20             So we're concerned.  I guess it doesn't go much 
21   beyond diversification.  We're concerned about the 
22   ability for some of these bonds in the revenue side to be 
23   able to deliver as promised, and were concerned about the 
24   importance increasingly of paying attention to issuers, 
25   whereas that isn't as much of an issue, or isn't 
0197
 1   perceived to be as much of an issue, as I think it really 
 2   is.  
 3             So for us, the primary issue, I guess, does 
 4   come down to diversification.  And it's rather ironic 
 5   because buying the individual muni has a huge single 
 6   advantage.  It has a maturity date.  And the funds don't, 
 7   so there's always market risk in the funds which you can 
 8   eliminate by buying individual munis and laddering them, 
 9   for example.  
10             So we do like the idea of buying munis.  But 
11   for the small retail client who's only going to be buying 
12   20- or $50,000 worth of them or $100,000 worth of them, 
13   it's problematic because of the concentration risk that 
14   exists.  And I'm not sure how to get around that other 
15   than dramatically reducing the minimum size of the bond, 
16   which doesn't become realistic for most issuers.  
17             MR. SMELSER:  I think that's an important, too, 
18   regarding the diversification and the minimum to 
19   nomination of 5,000 generically across most of these and 
20   the diversification's available.  As separate account 
21   managers, we do transact primarily in these odd lots and 
22   more retail size.   
23             So in terms of what we're doing on a day-to-day 
24   basis in terms of the percentage of number of trades that 
25   we're doing, the vast many of them are more of this 
0198
 1   retail and odd lot size.  When we look at it in terms of 
 2   the breadth of the market and the liquidity that's 
 3   available there and we've built a business on the fact 
 4   that it is strong in the number of bids that we're 
 5   getting with the access to the market, and the number of 
 6   dealers is very strong whether it's through these ATSs.   
 7             And I think something else that hasn't been 
 8   mentioned that's an important point of what the dealer 
 9   community is providing is more of the algorithmic 
10   trading, where they're setting valuations based on 
11   algorithms rather than a muni trader at, say, Morgan 
12   Stanley or Citigroup going through thousands.   
13             In David's case, he said it was 5,000; in our 
14   case, we say it's more of the bid wanted that we're 
15   seeing, having a trader go through and make a valuation 
16   determination on each one of those line items.  The fact 
17   that there's algorithmic trading and the growth of that 
18   has helped substantially provide liquidity to these 
19   smaller odd lot size positions in which we're primarily 
20   transacting.   
21             MR. HEANEY:  Bernie, did you want to respond?  
22             MR. COSTELLO:  Just one followup with respect 
23   to credit.  While we may respectfully disagree about the 
24   amount of pre-trade price transparency with respect to 
25   pricing and markets, we 100 percent agree with Ric where 
0199
 1   -- and I think it's just a function of the differences in 
 2   our market -- the amount of available information around 
 3   credit is nowhere near where it is in other markets.  
 4   Just the function of the amount of issuers.  I heard 
 5   36,000 was one.  I thought it was closer to 25.  
 6             I mean, the numbers are exponentially larger 
 7   than what you see in corporates.  Municipal issuers are 
 8   not beholden to some of the same quarterly reportings.  
 9   These aren't CFOs of organizations.  In some instances, 
10   they're elected officials.   
11             So just the amount of credit information that's 
12   available to the municipal market participant is 
13   decidedly less than it is in other markers.  So yes, 
14   while we may have our respectful back-and-forth on the 
15   amount of pricing available that's in there, but these 
16   differentiated disparate credits, that information is -- 
17   and I don't know what the solution is there, but it's 
18   definitely decidedly less than you see in some other 
19   markets.  
20             MR. HEANEY:  John Bagley? 
21             MR. BAGLEY:  Thanks, Michael.  I wanted to ask 
22   anyone on the panel about what really constitutes pre-
23   trade data.  Most of the conversation we've had today has 
24   focused on offerings and bids.  But I just wrote down a 
25   few things that I would potentially consider pre-trade, 
0200
 1   such as new issue scales.  
 2             Bernie, you mentioned some of these benchmark 
 3   indices and changes to them, Treasury market evaluated 
 4   prices.  And my question is -- on bids, I'm not talking 
 5   about an individual client's bid.  I'm talking about a 
 6   client's access to bid information on things that are not 
 7   there so they didn't put out.  
 8             And my question is really two-part.  Of those 
 9   things, what do you consider to be pre-trade?  And what 
10   do you think buy-and-sell side participants have ready 
11   access to?  And what do you think retail clients have 
12   ready access to?  
13             MR. SMELSER:  I'll just start from the buy side 
14   in terms of what is pre-trade information.  The things 
15   that you listed, I would all of them, and it's exactly 
16   what we're considering.  So in the instance of, we have a 
17   market that is close to a million -- or more than a 
18   million -- CUSIPs, if there is, and there is often, a 
19   CUSIP that comes out that's available for sale that 
20   hasn't been transacted in a good period of time, the 
21   question is:  What is pre-trade information that's 
22   relative to that bond?  
23             New issue scales are certainly important.  The 
24   MMD in terms of where the triple-A scale -- and this is 
25   sort of the benchmark for where most of the market trades 
0201
 1   on various securities -- offerings that are available 
 2   through the market, potentially bid wanted items that 
 3   might be coming for sale, whether or not they've already 
 4   been traded, is all relevant information.   
 5             And it's something where, getting back to the 
 6   fact that this a market that is still driven by heavily 
 7   tax-exempt, it's the individual's sort of tax status and 
 8   what is going into the relative value of that compared to 
 9   other securities.  
10             MR. HEANEY:  Chris?  
11             MR. FERRERI:  John, you're raising a good 
12   point.  All that information is available in disparate 
13   sources.  I think what's interesting, and what we've seen 
14   in muni brokers, for example, those 60 feeds of 
15   electronic markets do offer a lot of duplicity.   
16             There are some dealers that feel that to 
17   fulfill their best execution requirements, they 
18   distribute every primary everywhere.  The broker's 
19   brokers' markets are not 15 disparate markets.  It's the 
20   inter-dealer market.  
21             So although that information is widely 
22   available, someone also has to manage that information 
23   with a client to say that this is the true value of these 
24   markets.  Clearly the bids aren't distributed for someone 
25   else's bid wanted; that's given to the person that issued 
0202
 1   the bid wanted.  But post-trade, that information is 
 2   available.  
 3             Now, I do think the combination of pre-trade 
 4   pricing, evaluated pricing, the tools that are out there, 
 5   give great insight into the value of a certain bond.  But 
 6   I will tell you as a market operator, it's the dealer 
 7   that creates the interest in a particular bond.  
 8             It's important to talk about the primary level 
 9   of a bond.  But in determining which bond to buy is a 
10   determination of:  What's the best value at the moment.  
11   More often than not, it's triggered by the activities of 
12   our customers, who say, "At this point in time, this is 
13   the bond I'm interested in."  And you go to work and you 
14   make that happen.   
15             We don't trade a million bonds, a million 
16   CUSIPs a year.  But we do trade tens of thousands.  And 
17   we do that as neutral intermediaries whose only objective 
18   is to match the buy and the seller.  So when you look at 
19   the information that's available; the information on ATSs 
20   like Hartfield; the fact that on a platform like 
21   MuniBrokers, MuniBrokers has more notional volume than 
22   the other ATSs combined.   
23             Not more individual trades because there are so 
24   many smaller trades that happen, and through the 
25   efficiency of the more resale-oriented customers.  But in 
0203
 1   sheer volume of the market, that activities helps 
 2   determine where the prices really are.  
 3             MR. HEANEY:  Ric?  
 4             MR. EDELMAN:  When you go to buy a car, you 
 5   look at the sticker price and you see a piece of data 
 6   there that talks about the EPA mileage on the car.  And 
 7   we all know that that is not what you're going to get 
 8   driving the car because you might be driving lots on the 
 9   highway; you  might be driving entirely in the 
10   neighborhood.  
11             But it doesn't matter because every car has 
12   that number, and it's a good way for you to compare 
13   apples to apples, based on how the EPA did it so that you 
14   can effectively decide, do I want a gas-guzzler or do I 
15   want to be more efficient with my automobile's mileage?  
16             By that notion, I want to raise something I 
17   would love to see in pre-trade transparency that is a 
18   very different nature of what we're discussing here.  
19   Because we're talking, I think, so far today at the firm 
20   level.  We're talking about the information data, the 
21   broker trading desk and the information being provided at 
22   the corporate level.  
23             But at some point, all the trading that's being 
24   done, Bernie, is being then given not to the end client 
25   at Morgan Stanley but to the Morgan Stanley FA, who is 
0204
 1   the intermediary to the client.  And at that point you 
 2   lose quality control.  We all do.  I mean, I've got 160 
 3   advisors to your 16,000, but we all lose quality control 
 4   at that point.  
 5             What did the FA say to the client?  And what we 
 6   have often found is that when we look at a brand-new 
 7   client coming in and they show us their existing 
 8   portfolio and they own a bunch of munis, they live in 
 9   Virginia but they own munis from New York and 
10   Pennsylvania and Wyoming and Illinois and so on.  
11             And the reason is, that FA was trying to 
12   provide diversification to the client.  But the end 
13   result was they bought a muni where the client, living in 
14   Virginia, did not get the state tax break because they 
15   bought an out-of-state bond.  
16             So I would love to see an EPA styled "Apples to 
17   Apples Comparison," even though it might not be a 
18   legitimate -- or totally accurate, I should say -- 
19   methodology, is a taxable equivalent yield.   
20             Assuming a given tax bracket at the federal 
21   level, and you'd have to assume it because we're all 
22   going to be different, and you assume, say, the top 
23   bracket at the state level, if we can provide to every 
24   muni in advance, "Here's the taxable equivalent yield; 
25   let the client understand that buying an out-of-state 
0205
 1   bond isn't necessarily going to be as profitable for them 
 2   as an in-state bond."  
 3             That level of disclosure I don't think either 
 4   one of us can promise is always being conveyed to the 
 5   client, or if it was, I think we would see a lot fewer 
 6   out-of-state bonds being purchased by consumers who don't 
 7   really fully appreciate that.  So I'd love to see taxable 
 8   equivalent yield added to the pre-trade transparency.  
 9             MR. COSTELLO:  I believe the question was how 
10   will we define pre-trade transparency, and I think that's 
11   the right exercise because I think when you go from 
12   market to market, sector to sector, the market 
13   conversations and protocols are different and people 
14   interpret things as being differently.  
15             So from my perspective, when I think about pre-
16   trade price transparency or all the things that we look 
17   at before setting the price for a security, and I said in 
18   my opening remarks, I think for the most liquid of 
19   sectors and balanced markets and stable interest rate 
20   environments around modest trade sizes, it consistently 
21   is, and I think, the best -- it's post-trade history.  
22             Post-trade history for me, moreso than markets, 
23   is more valuable because I think the inform is more 
24   consistent across all markets in how it's reported.  
25   There's definitely differences in terms of masking 
0206
 1   requirements and what's shown and times.  But when you 
 2   think about when you're using post-trade history to 
 3   analyze a product, pretty much across all sectors that 
 4   have realtime reporting, you're seeing the exact size, 
 5   for the most part; the level; who's on the side of the 
 6   trade; and at what price that risk cleared the market.  
 7             Whereas in markets, for a lot of the committees 
 8   that I heard -- I mean, Larry, you spoken of it 
 9   previously -- there's nuances of what's a market.  Is it 
10   live?  Is it quartered?  Is it indicative?  Are there 
11   walk-away issues?  Things like that.  
12             So when you start to talk about some of the 
13   things that we want to do from other asset classes, and 
14   then maybe potentially provide them equally in munis, the 
15   sheet amount of data that you're talking about, when 
16   you're talking about our end user, I think could be more 
17   noise.  
18             To me, what I like a post-trade data as kind of 
19   the holy grail is it's not real subject to much 
20   interpretation.  And with respect to some comments about 
21   it being, well, it's old and it's stale, I respectfully 
22   disagree.  It's just a time stamp.   
23             You can very easily reconstruct exactly what 
24   was going on in the market at that time based on all the 
25   post-trade history of everything else that you're looking 
0207
 1   at in your reference securities at that time.  I don't 
 2   think it makes the information less valuable.  
 3             But if I start showing things other than where 
 4   bonds actually traded, but markets that are today, to 
 5   clients, some of which you may be able to transact on, 
 6   may not be able to transact on -- I'd rather just deal 
 7   with the facts.   
 8             I mean, I just -- in my collective experience, 
 9   and I know we're not saying that here, but if you gave me 
10   my choice of making markets based purely and solely on 
11   post-trade data or making my markets based purely and 
12   solely on current markets today, I would choose the 
13   former -- I would rather -- and again, that's not what 
14   we're doing here -- the combination of both.  
15             But I can tell you without question the first 
16   thing just about every trader looks at when they see a 
17   bond and an offering is, well, where did they trade last, 
18   or where did something similar trade last?  So again, 
19   that's my two cents.  I think the integrity of the data 
20   is very, very important.  And if we're going to introduce 
21   new rules, we obviously need to maintain that integrity 
22   in doing so.  
23             MR. HEANEY:  Larry's got the last question, and 
24   he promises it's a short one.  
25             MR. HARRIS:  Oh, it's the shortest question and 
0208
 1   answer of the day.  
 2             Ric, what are the annual expense ratios that 
 3   your clients indirectly pay when you put them into ETFs 
 4   and bond mutual funds?  
 5             MR. EDELMAN:  On average, in our firm, it's 
 6   about 30, 35 basis points.  The ETFs are getting pretty 
 7   cheap these days, so for some of the more exotic emerging 
 8   markets stuff, those might be 50 to 75 basis points.  But 
 9   they're a very tiny portion of the portfolio.   
10             So on a blended basis, clients are typically 
11   paying under 40 BPs for their investments.  And the bonds 
12   -- the bond ETFs are among the cheapest that you're going 
13   to buy, so under 20 basis points.  So it's -- and that's 
14   part of the reason that yields -- that the spread issue 
15   is a big conversation.  
16             A long time ago we had to argue or justify or 
17   rationalize why you were better off buying a bond fund as 
18   opposed to bond, because the fund's expense ratios were 
19   so high.  Today, they're so low, and within a couple of 
20   years I'm convinced they're going to be zero.  Pretty 
21   soon I think they're going to be giving you toasters to 
22   open a fund account the way that we used to.  
23             (Laughter.) 
24             MR. EDELMAN:  So that's becoming -- making it a 
25   lot easier for us.  
0209
 1             MR. HEANEY:  So I just want to take this 
 2   opportunity to thank Lynn for moderating what was an 
 3   extremely interesting panel.  We knew this would be an 
 4   enlightening and interesting one going in, and it lived 
 5   up to it.  So thank you all very much for coming to this 
 6   meeting as well.  
 7             We are going to make a decision on the fly 
 8   here.  So we are through our break.  Our break will be 
 9   4:00.  Let's jump right into the electronic trading and 
10   technology discussion.  Brett is going to moderate the 
11   next panel.   
12             So if you do need to step out, I think, while 
13   the panel is getting set up, now is the time to do it.  
14   But we will go right into this panel, and the thought 
15   process is, we have a recommendation to put forth to 
16   FIMSAC.  That'll happen at 3:45.  
17             (A brief recess was taken.) 
18             MR. HEANEY: Okay.  Thank you all for sticking 
19   with us and for joining us for this next panel.  I'm very 
20   excited to have this distinguished group here to speak 
21   with us today.  And thank you, Michael, for everything so 
22   far.  I'm going to provide a few introductory remarks, 
23   introduce our panelists.  We're also going to be talking 
24   about the recommendation that is before the Technology 
25   and Electronic Trading Subcommittee.  
0210
 1   PANEL IV: ELECTRONIC TRADING VENUE REGULATION --  
 2   DRAFT RECOMMENDATIONS 
 3             So we do have a recommendation that calls for 
 4   the SEC to form a joint working group with FINRA and the 
 5   MSRB to review the framework for the oversight of 
 6   electronic trading platforms for corporate bonds and 
 7   municipal securities.  Rick McVey of MarketAxess will be 
 8   going over those recommendations momentarily.  
 9             But before that, just a few comments.  So at 
10   the heart of fixed income trading and markets regulation 
11   is the oversight of market intermediaries.  This 
12   oversight is absolutely critical for the protection of 
13   investors and the promotion of fair, orderly, and 
14   efficient markets, which is kind of what we're doing 
15   here.  
16             There are two categories of market intermediary 
17   regulation under the Exchange Act, one for trading 
18   platforms and one for firms that act as brokers or 
19   dealers.  Trading platforms, as you know, provide 
20   facilities for bringing buyers and sellers together and 
21   are regulated under the Exchange Act as either registered 
22   exchanges or ATSs.  
23             And again, broker dealers or firms engaged in 
24   the business of buying or selling securities for the 
25   accounts of others, or for their own accounts, in the 
0211
 1   case of dealers, and are regulated as such under the 
 2   Exchange Act.  
 3             While the Exchange Act and Commission 
 4   regulations distinguish between these two types of 
 5   intermediaries, that distinction has not always been 
 6   clear as technologies have developed and new market 
 7   participants and business models have emerged.  Indeed, 
 8   the blurring of these lines have given rise to the 
 9   Commission's adoption of Regulation ATS in 1998.  
10             At that time, the Commission provided a new 
11   interpretation of an exchange to account for advances in 
12   technology and emerging business models for trading 
13   securities that until then were regulated solely as 
14   broker dealers or not at all.  
15             The subcommittee's preliminary recommendation 
16   today recognizes this longstanding challenge of 
17   appropriately categorizing market participants and 
18   business models for optimal market regulation.  At the 
19   core of the preliminary regulation is an acknowledgment 
20   that fixed income electronic trading occurs through a 
21   number of trading protocols that compete with one another 
22   yet are regulated differently, and that regulation should 
23   be tailored to the fixed income market.  
24             Coincidentally, later this week the Commission 
25   will consider whether to adopt the first set of 
0212
 1   amendments to Regulation ATS since its adoption in 1998. 
 2    While these amendments were not proposed to apply to 
 3   fixed income ATSs, the Commission recognized the evolving 
 4   and increasing news of electronic venues to execute 
 5   trades in corporate bonds and municipal securities.  
 6             In that proposal the Commission solicited 
 7   public comment regarding the regulation of fixed income 
 8   trading platforms that go to the heart of today's 
 9   preliminary recommendations.  The Commission asked 
10   whether commenters believed that the current conditions 
11   to the exemption from the definition of an exchange are 
12   appropriate for fixed income ATSs.  
13             Further, the Commission requested comment on 
14   whether its potential feature regulation of fixed income 
15   ATSs would place them at a competitive disadvantage with 
16   respect to non-ATS trading venues that trade fixed income 
17   securities.   
18             Finally, in recognition of the differences 
19   between equity and fixed income ATSs, the Commission 
20   requested a range of comments designed to identify the 
21   relevant differences between equity and fixed income 
22   platforms that could inform a regulatory approach 
23   tailored to the fixed income trading models.  
24             As I mentioned earlier, we appreciate this 
25   committee's expertise in considering whether our rule set 
0213
 1   fixed income electronic trading platforms is optimally 
 2   designed to protect investors and protect the fairness, 
 3   efficiency, reliability, and resiliency of these 
 4   electronic markets.  
 5             We will begin by asking Rick McVey, chair of 
 6   the Technology and Electronic Trading Subcommittee, to 
 7   introduce their recommendation and discuss the 
 8   subcommittee's deliberations.  So Rick, over to you.  
 9             MR. MCVEY:  Thank you, Brett.  And let me just 
10   start by reiterating my thanks to the entire E-Trading 
11   and Technology Subcommittee.  We had active participation 
12   in the recommendation that we're making today, very good 
13   discussion, and broad debate.  
14             As Brett mentioned, the genesis of this is the 
15   regulatory framework for fixed income markets, which we 
16   believe today is fragmented.  And what that has resulted 
17   in is the retail markets, which tend to trade closer to 
18   click-to-trade and a little bit closer to equity markets, 
19   tend to be regulated primarily as ATS systems.  
20             The institutional markets, where RFQ technology 
21   and protocols have been popular, are regulated primarily 
22   as broker dealers by FINRA.  And there is at least one 
23   example of an active e-trading participant in both retail 
24   and institutional markets that is not regulated by either 
25   FINRA or ATS regulations today.  
0214
 1             The differences in regulatory status, the 
 2   committee believes, is arbitrary.  It is based primarily 
 3   on trading protocols and not business models, and it also 
 4   to a certain extent is based on revenue models, which 
 5   again, we believe are arbitrary distinctions that do call 
 6   for a review of whether the regulatory categorizations in 
 7   fixed income e-trading are appropriate today.  
 8             We believe that the reason for these 
 9   differences, as Brett pointed out, are primarily because 
10   of the history of ATS rules that were built primarily 
11   around the electronic trading developments in equity 
12   markets.  And there are important distinguishing factors 
13   in today's fixed income markets.  The primary difference 
14   is the popularity of RFQ trading protocols, which, as I 
15   mentioned, have been the predominate source of electronic 
16   trading volume in fixed income electronic markets.   
17             Those RFQ protocols have gone through a series 
18   of evolutions, which greater automating in RFQ trading 
19   protocols today than ever before, with market-making 
20   algos at work, as well as investor auto-execution 
21   becoming increasingly popular, some of the same steps 
22   that we saw take place in equity markets historically.  
23             Given the growth in fixed income electronic 
24   trading overall and the size of the fixed income markets, 
25   the committee believes that a more consistent regulatory 
0215
 1   is wanted for electronic trading venues active in U.S. 
 2   fixed income markets in order to ensure fair and 
 3   effective fixed income electronic markets:  investor 
 4   protection from systemic risk, e-trading system outages, 
 5   and client confidentiality.   
 6             The avoidance of regulatory redundancy across 
 7   multiple regulatory bodies -- one great example of that 
 8   is John Bagley has been very helpful to the committee by 
 9   relaying the work that MSRB is doing on looking at "last 
10   look" protocols at work in fixed income markets, and the 
11   pros and cons of last look.   
12             In the committee's opinion, that's a great 
13   example of work that could be shared more broadly across 
14   markets beyond just munis, where corporate markets and 
15   other fixed income markets have similar issues that need 
16   close examination.  We also want to make sure that the 
17   regulatory work encourages the promotion of innovation 
18   and investment in new e-trading protocols that could 
19   improve liquidity and efficiency in our markets.  
20             In order to accomplish these goals, the 
21   subcommittee recommends that the SEC, FINRA, and MSRB 
22   form a working group to study these issues carefully to 
23   determine if modernizing the fixed income e-trading 
24   regulatory framework could improve market resiliency, 
25   stability, and safety for fixed income market 
0216
 1   participants.  
 2             And with that, I will conclude, Brett.  
 3             MR. REDFEARN:  Thank you.  Thank you, Rick.  
 4             Okay.  So now I'd like to introduce our 
 5   panelists.  And I'm going to ask each of them to offer 
 6   their thoughts on the recommendation that we're going to 
 7   be considering today.  And I think in so doing it would 
 8   be good if everybody could provide just a little bit of 
 9   information on your background and how it relates to the 
10   recommendation.  But first, to introduce our panelists.  
11             I'll start just next to Rick there on the left 
12   with Horace Carter, head of fixed income trading at 
13   Raymond James, member of the committee; Doug Friedman, 
14   general counsel of Tradeweb; we have Ben Macdonald, 
15   global head of enterprise products and president of 
16   Bloomberg SEF, Bloomberg; Alex Sedgwick, head of fixed 
17   income market structure and electronic trading at T. Rowe 
18   Price; and Bill Vulpis, head of BondPoint, ICE BondPoint.  
19             So with that, again, thank you all very much 
20   for being here.  We look forward to this discussion.  
21   Before I turn it over to Horace, I will just say this:  
22   Because of our timing, I think that we have approximately 
23   40 minutes of discussion and then we are going to call a 
24   vote just because we're looking for a quorum and we know 
25   some people are going to be needing to move on to other 
0217
 1   places.  So that's kind of the timing that we have here, 
 2   which is one of the reasons why we're moving this along 
 3   quicker initially.   
 4             So Horace, over to you.  
 5             MR. CARTER:  Okay.  I promise to be brief.  So 
 6   I'm the head of fixed income trading at Raymond James.  
 7   Raymond James is a large regional dealer, regional as 
 8   defined by being headquartered outside of New York.  It's 
 9   headquartered in St. Petersburg.  I'm responsible for all 
10   trading, taxable and municipal trading, institutional and 
11   retail.  
12             My opinion about the recommendation is that for 
13   what it ideally would look like would be a large 
14   framework that, as Rick mentioned, would not favor or 
15   responsible or narrow the different protocols that might 
16   be developed going forward.  I think that the electronic 
17   trading universe is going to need to continue to see 
18   these innovations in order to improve liquidity.  That 
19   would be the key element that I would recommend.  
20             It needs to address five different things:  
21   system integrity, obvious; access, fair and open access 
22   to the different platforms; data, critically important -- 
23   what are the baseline rules around data goes across these 
24   platforms; trade reporting -- I think that we need to 
25   improve trade reporting over the platforms.  I think 
0218
 1   trades need to be, for example, reported net of fees, and 
 2   I think that the way many trades are reported are 
 3   misleading terms of volume; and then "integrity," which I 
 4   didn't know we were going to beat to death in the earlier 
 5   panels, so I'll just leave it there.  
 6             MR. REDFEARN:  Okay.  Doug?  
 7             MR. FRIEDMAN:  Good afternoon, and thank you 
 8   for the opportunity for us being here.  We appreciate it.  
 9             Trader markets as a global leader:  In 
10   electronic marketplaces, we have trading platforms om 
11   over 25 markets, asset classes, and fixed income 
12   derivatives and equities.  And we're the only trading 
13   company that has both the institutional, the whole sale, 
14   and the retail segments, and with over 500 billion 
15   average daily volume.  So we think we bring some 
16   perspective.  
17             We have three broker dealers, two of which are 
18   ATSS.  We have some SEFs in there, MTS and the like.  So 
19   we appreciate being here to talk on this topic.  
20             And over our 20-year history, we have seen 
21   massive benefits in electronic trading.  We think is 
22   brings greater transparency and efficiency in reduction 
23   of risk, and I think the increase in electronic adoption 
24   across all marketplaces is reflective of that.  
25             We're supportive of regulation here, but one 
0219
 1   that's appropriately sized for the type of market that's 
 2   largely principal-based, and that it's also sized to meet 
 3   the policy objectives.  I think -- we heard this morning 
 4   about the size of the paperwork for MiFID and Dodd-Frank, 
 5   and I think those are recent examples of just the massive 
 6   regulatory overhaul.  And that's forced broker 
 7   participants to prioritize people and technology to 
 8   implement those initiatives.  
 9             And so I think there are a lot of lessons to be 
10   learned in that process.  The regulation must recognize 
11   the differences in the markets -- how markets trade, the 
12   impact of trading and reporting on liquidity, the time 
13   and cost of resources associated with any such 
14   implementation, and also the opportunity cost because you 
15   have to prioritize regulatory and compliance over what 
16   might be commercial innovation, potentially.  
17             So ultimately we think it's got to allow for 
18   the flexibility of the different marketplaces that exist 
19   today.  But we do think it's important to have 
20   established baselines and consistency across similar 
21   platforms.  
22             And the recommendations in particular, I think: 
23    We think the idea of a working group is very 
24   constructive.  They should aspire towards consistency.  
25   We think a re-review of Reg ATS makes perfect sense, for 
0220
 1   reasons we'll talk about in a little more detail later.  
 2   And obviously, as I said before, I think the cost-benefit 
 3   analysis is critical here.  
 4             And finally, we just -- we encourage the 
 5   working group to work very closely with the marketplace, 
 6   both venues and users alike, because feedback will be 
 7   critical in how they do their work.  
 8             MR. REDFEARN:  Thank you.   
 9             Ben?  
10             MR. MACDONALD:  So I'm responsible for 
11   enterprise products at Bloomberg as well as president of 
12   Bloomberg SEF.  I'll just make a couple of comments on 
13   the recommendation.  
14             Clearly, the focus should be investor 
15   protection.  And I think any recommendation that 
16   ultimately gets made, should you really balance the risk 
17   and reward around the objectives of investor protection 
18   versus the cost that's overload on the industry?  
19   Obviously we've got a lot of experience in both MiFID II 
20   and Dodd-Frank.  We operate an MTF in Europe, and we 
21   operate a SEF here in the U.S.   
22             I would make the argument, and somebody made 
23   the comment earlier on, the -- it went all of a sudden 
24   down.  The cost to the industry may have been around $2.5 
25   billion to implement MiFID II, and all of that to arrive 
0221
 1   at pre-trade transparency for about 400 bonds.  So the 
 2   return on -- the cost-benefit of that, I think, is worth 
 3   bearing in mind.  
 4             With that said, I think what MiFID II does very 
 5   well is lay the foundation of creating an ecosystem of 
 6   trading that ultimately provides a funnel for post-trade 
 7   transaction reporting, which is why I think the valuable 
 8   outcome of MiFID II ultimately will be that post-trade 
 9   reporting, which we've talked about a lot today.  
10             I would make the argument that in the U.S., 
11   that exists today with TRACE, and similarly, under Dodd-
12   Frank, with the swap data repository ecosystem that has 
13   been put in place.  And so I think the U.S. has a little 
14   bit more luxury to be sensitive of the ecosystem that is 
15   fixed income.  
16             In the institutional market -- we don't really 
17   operate in the retail market, which is more of a central 
18   counter party anonymous marketplace; we operate in the 
19   institutional space -- that's very much still a 
20   negotiated market, whether that's negotiated over the 
21   phone, message, or electronic trading platforms, or 
22   message -- transaction message systems like ourselves.  
23             But I think whatever recommendations are being 
24   made, I think that plurality of how the market operates 
25   is massively important.  And I think if you look at the 
0222
 1   competition which is starting to happen today, you have 
 2   more and more large sell side firms that are interacting 
 3   directly with larger buy side firms through trade 
 4   messaging ecosystems, et cetera.  
 5             The regulation, if it to be regulated, needs to 
 6   be more forward-looking with regard to where the 
 7   technology is going, as against backward-looking in 
 8   regards to where technology has been.  Thank you.  
 9             MR. REDFEARN:  Thank you, Ben.   
10             Alex?  
11             MR. SEDGWICK:  Hi.  My name's Alex Sedgwick, 
12   and I run fixed income market structure and e-trading for 
13   T. Rowe Price.  I've been there for 4 years, and then 
14   prior to that I ran -- or I worked at MarketAxess for 7 
15   years, culminating in a role as head of research.  
16             The first thing I want to do is thank the 
17   committee for the invitation to be here today.  I think 
18   overall we believe that the evolution of e-trading is at 
19   an important role in supporting best execution and 
20   reducing trading costs for investors.  
21             One of the more important aspects of my role at 
22   T. Rowe Price is to evaluate electronic trading platforms 
23   for our fixed income trading desk.  And so as I think 
24   about how we look at it from a regulatory perspective and 
25   the ways in which it could be supportive of that process, 
0223
 1   I think ultimately the kind of guiding principle should 
 2   be creating a fair playing field for all the e-trading 
 3   platforms -- specifically, greater transparency into 
 4   operations and supportive overall operational resiliency.  
 5             There's been significant organic growth in this 
 6   space in the last several years, and I think well-
 7   designed regulation can ultimately be a further catalyst 
 8   for that growth.   
 9             To get very specific in terms of the areas that 
10   I would request the review sort of evaluate would be, 
11   first of all, transparency in business operations.  We've 
12   seen a large number of new protocols come to market prior 
13   to, say, 2015, prior to 2008.  The RFQ protocol was the 
14   dominant protocol in this space, and over the last 
15   several years, we've things like dark pools, auctions 
16   come to the market, all-to-all trading, and crossing 
17   networks.  
18             So I think one thing that would be helpful is 
19   consistency in reporting how those protocols work, what 
20   liquidity you're interacting with, and as Tim Morbelli 
21   pointed out on an earlier panel, even if you have buyers 
22   and sellers coming to the market at the same time, the 
23   price formation process is a critical piece of whether or 
24   not that interest actually culminates in a trade.  So I 
25   think additional transparency on price formation.  
0224
 1             I think another element to this is as we see 
 2   these platforms evolve over time, we want to ensure that 
 3   there's uniform and consistent information of any changes 
 4   in those protocols with all market participants.  So I 
 5   think that's a key element that we can import in a lot of 
 6   respects from our experience in other markets.  
 7             Reporting around fee structure, I think that's 
 8   important.  Is that consistent amongst users?  Are there 
 9   discounts?  If there are discounts, some disclosure 
10   around them?  
11             Moving on, I think there's also -- we had 
12   brought up the topic of auto-execution.  I think this is 
13   another example of an emerging protocol in this market.  
14   Historically, any issues with trades and trade 
15   consummations was typically the -- was the responsibility 
16   of the trading counter parties to resolve.  
17             I think as we see trading platforms take more 
18   responsibility for execution of orders, there also has to 
19   be greater responsibility around how they're going to 
20   handle and communicate any issues with the trading 
21   protocols, any problems around technological glitches 
22   associated with that.  So that's another area that I 
23   think additional transparency in reporting could be 
24   helpful.  
25             The final piece -- I'll just echo another item 
0225
 1   that had come up earlier -- is data, data protection.  
 2   That's a function of not only the technology around it 
 3   but also, internally at some of these platforms, who has 
 4   access to trade data in an attributed format.  Further, 
 5   how is that data used in terms of monetizing it through 
 6   any market data?  A better understanding of that would be 
 7   very helpful.  
 8             I think ultimately the kind of -- to sum all of 
 9   this up, I think what would be helpful is just knowing 
10   that you have data that's easily comparable across 
11   different e-trading platforms because ultimately, what 
12   we're looking at is a large number of them and trying to 
13   figure out:  Where is there currently activity?  Where 
14   are we seeing momentum in that activity?  And do we have 
15   a strong sense of really understanding how these 
16   platforms work?  
17             Now, for a larger asset manager like T. Rowe 
18   Price, you may have somebody like me that spends their 
19   time getting answers to all flood those questions. But I 
20   think one thing that would help with the adoption of e-
21   trading, particularly as you get down into smaller asset 
22   managers, is making that data more available in a public 
23   format where it can be easily consumed.  Thank you.  
24             MR. REDFEARN:  Thank you, Alex.  
25             Bill?  
0226
 1             MR. VULPIS:  Good afternoon, everyone, and 
 2   thanks for having me here.  Bill Vulpis, head of ICE 
 3   BondPoint.  BondPoint is a multi-asset-class fixed income 
 4   registered ATS.  It was established in 1999, and was 
 5   recently acquired by Intercontinental Exchange earlier on 
 6   this year in January.  
 7             BondPoint's subscriber clients consist of sell 
 8   side firms, buy side firms, wealth management, and retail 
 9   brokers, so all the people you've heard from today and 
10   many of the panels that in front of you.  BondPoint 
11   supports both the click-to-trade, as we spoke about 
12   through ATSs, as well as an RFQ protocol.  The click-to-
13   trade protocol is supported by a limit order book.  
14             In terms of the recommendation, we would be in 
15   support of the recommendation because we believe it would 
16   help better define a set of rules for fixed income 
17   electronic trading, provide additional transparency, 
18   provide a framework, as Alex was just mentioning, about 
19   actually knowing what is trading electronically.  How 
20   about a definition of what an electronic trade actually 
21   is?  
22             A common set of rules to operate from, and 
23   obviously an advancement to support the advancement of 
24   electronic trading in fixed income.  A regulatory 
25   framework should apply to all market participants, 
0227
 1   regardless of the platform protocol or business model.  
 2             It should be designed to promote the 
 3   advancement in market structure through transparency, 
 4   fair access, and transparency around market center 
 5   operations.  Trade reporting should also indicate where 
 6   these trades and how these trades are actually being 
 7   executed.  
 8             The integrity of the liquidity, which we've 
 9   heard throughout the day, must consistently be measured 
10   or mandated, meaning participants need to stand up to 
11   pricing.  System resiliency, especially around latency 
12   standards -- these are not only latency standards around 
13   the electronic platforms in terms of the ingestion of 
14   market data that we have from our 250 liquidity 
15   providers, but also latency in terms of how that 
16   liquidity and how that market data is being absorbed by 
17   clients who are taking that market data in.  
18             In addition, we've heard about many different 
19   types of platforms.  It's not only vendor platforms.  
20   It's not only regulated ETFs' platforms.  We did hear 
21   from the previous panel regarding the municipal 
22   transparency about retail aggregator platforms, other 
23   platforms that are out there.  
24             So we're looking at a framework that should 
25   apply to all various types of platforms phthalate are out 
0228
 1   there in the market.  Thank you.  
 2             MR. REDFEARN:  Okay.  Thank you very much.  So 
 3   I think the next thing we're going to do is have another 
 4   fairly general question for all of the panelists.  This 
 5   one is looking specifically in the context we have about 
 6   how the trading protocols and fixed income platforms have 
 7   indeed continued to evolve.  Some is occurring in a more 
 8   automated manner.  But we are seeing volume, more volume, 
 9   going to electronic platforms.  
10             With that in context, if you were starting with 
11   a clean slate, how would you design a regulatory 
12   framework for fixed income electronic trading?  And I 
13   want, if possible, to get specific about a few areas in 
14   addition to what's already been discussed here today.  
15             So, for example, who would the framework apply 
16   to?  What would be the key regulatory objectives of the 
17   framework?  And again, some of the themes we've mentioned 
18   -- should this framework in fact be supporting 
19   operational transparency systems, integrity resiliency 
20   standards, and so on?  
21             So I think this is just an opportunity to 
22   elaborate on what we've said a little bit.  But again, 
23   clean slate.  What would you do, Horace?  
24             MR. CARTER:  Well, as I said earlier, it would 
25   be -- it would be probably applicable.  But it would 
0229
 1   leave enough room for innovation.  That would be the 
 2   critical point.  And the reason is, as you point out, 
 3   electronic trading is growing, and it's going to continue 
 4   to grow.  
 5             So one of the problems that I see with the 
 6   growth of electronic trading is that it could -- it can -
 7   - in normalized markets, it's fine.  But if you get 
 8   stressed markets or distressed markets, then you can find 
 9   that the lack of that sort of relationship business, 
10   which is the traditional -- it's built with a traditional 
11   -- a voice trading model, which is still where most block 
12   target occurs.   
13             As electronic trading becomes more prolific, 
14   and it will increase, there is -- our protocols have made 
15   it very clear they like it and they're going to continue 
16   to use it more and more.  And we're going to need to find 
17   protocols for large blocks of bonds, and it's going -- I 
18   don't know what the solution looks like exactly.  But it 
19   will be some version of customizable by the buyer in 
20   terms of what's available for sale.  
21             So in terms of specific regulations that I 
22   would -- specific aspects of a framework, I'll say it 
23   again:  I think that trades need to be reported net of 
24   fees on these systems.  I understand that that's not as 
25   simple as I'm making it sound.   
0230
 1             But for lack of a better solution, you could 
 2   report what the buyer pays, the price that should be 
 3   reported as the price that the buyer pays because when we 
 4   talk about market structure problems and liquidity 
 5   problems, we're not worried about being able to find 
 6   bonds.  We're worried about being able to turn bonds into 
 7   money.   
 8             So I would say, in terms of reporting, we 
 9   report what the buyer pays.  And we need to look at the 
10   various protocols in terms of how volume is reported 
11   because I think that there's a lot of duplicative 
12   reporting in terms of volume based on whatever the 
13   current protocol might be.  
14             MR. REDFEARN:  Thanks, Horace.  
15             MR. FRIEDMAN:  I think when we think about a 
16   clean slate, obviously you got to start with what are the 
17   objectives?  What are the markets?  And what are the 
18   products?  And I think some of the struggle with the ATS 
19   is that that -- what it was originally designed for has 
20   sort of morphed, or the market has changed so much, since 
21   that started.  
22             So I think it sounds simple, and obviously we 
23   would encourage any regulation to be kept simple.  But 
24   ultimately, it's what's the scope?  What's this supposed 
25   to cover?  And what are we trying to achieve?  And I 
0231
 1   think, actually, we can use an ATS example, maybe, to 
 2   illustrate the point.   
 3             If you look at Exhibits F and G of Form ATS, 
 4   it's described the manner and the operation of the 
 5   platform, and then what are the system safeguards and 
 6   capacity and scalability.  Those are very good core 
 7   requirements that every electronic platform should have 
 8   to demonstrate and describe.  But I think the problem is 
 9   that it was done in the context of regulation that was 
10   sized for central limit order books and exchanges, and 
11   not for all of the various markets and protocols that we 
12   see today.   
13             And so I think, again, if you think about it at 
14   a high level, think about the manner and operation.  What 
15   are the core technologies, core system requirements you 
16   want?  And then try to get that broad-based across what 
17   might be varying protocols in various trading platforms.  
18             And so, again, in the other report, the point 
19   is that you have to think about the context of global 
20   markets now and how do they fit with existing regimes?  
21   And what's the cost-benefit analysis of undoing sort of 
22   where you are and what you have to implement anew?  
23             MR. MACDONALD:  So I think when you think about 
24   the fixed income market in particular, I think you have 
25   to break it down into two different components.  There's 
0232
 1   what I'd characterize as trade messaging, and what I'd 
 2   characterize as services.  And I think they're very 
 3   different in the sense that trade messaging is nothing 
 4   more than the automation of the telephone.  
 5             The ownership of the decision belongs to both 
 6   the counter parties.  They enable each other to 
 7   communicate with each other, and ultimately any 
 8   transactional decision that's made as a result of that 
 9   messaging is done by the counter parties.  And as I 
10   pointed out, you're all -- when there's an issue, the 
11   counter parties decide how to resolve that issue.   
12             I think when you talk about services and you 
13   start getting into things like central counter parties 
14   and anonymous trading and generally platforms where the 
15   decision is automated, I think there's no debate that 
16   there needs to be a layer of regulation around that.  
17             And I think that for all the reasons that 
18   people have pointed out, systems safeguard integrity.  
19   Transparency around the protocol is understanding what 
20   you're signing out for, et cetera, et cetera.   
21             So then you have to re-look at that.  And I 
22   think that the competitive landscape is actually evolving 
23   down those two different paths as well.  On the trade 
24   messaging front, what we're seeing increasingly is larger 
25   institutions establishing direct activity and building 
0233
 1   their own technology on either side of that trade 
 2   messaging.   
 3             And so increasingly after, you're going to see 
 4   a world where there's insourced technology and outsourced 
 5   technology, outsourced technology being people who don't 
 6   have the same technology budgets to be able to do it for 
 7   themselves.  
 8             In the services area where a lot of -- there is 
 9   also a lot of innovation.  And I think that's generally a 
10   very good thing.  But I think also that you have to look 
11   at equity market structure and understand some of the 
12   issues that have happened over time and learn from that.  
13             But I do think that that differentiation in 
14   fixed income is incredibly important because if you're 
15   not careful, you could land in a place where the less 
16   sophisticated investor is obliged to outsource that 
17   technology in order to be able to provide that 
18   connectivity with counterparts.  
19             And they could end  up, if that part of the 
20   market is regulated, versus the insourced technology part 
21   of the market wasn't regulated, then you'd be crying an 
22   unfair advantage for one part of the market versus the 
23   other.  
24             So I think fixed income is a bit of -- it is a 
25   very challenging market, so good luck.  I do think that 
0234
 1   the transaction reporting and the TRACE reporting is 
 2   incredibly important is this market.  And regardless of 
 3   how people trade today or transact or communicate, 
 4   ultimately all of that information does land in the same 
 5   place, which does actually provide quite a robust 
 6   framework around catching things going wrong.  C 
 7             MR. SEDGWICK:  I think in my earlier answer, I 
 8   provided some specifics.  But I do think it's important 
 9   to kind of take a step back and think, as you're looking 
10   at this market, what are the guiding kind of questions 
11   you should be asking yourself?  So I would chalk that up 
12   to, really, two.  
13             The first is, how do we ensure that these 
14   platforms, given their importance in the market, have a 
15   strong set of internal controls and end up being 
16   resilient, given the liquidity that they provide?  So I 
17   think a lot of the focus there is really on transparency 
18   around that data protection, around protocols.  And I 
19   think ultimately you'll actually learn quite a bit in 
20   terms of best practices that already exist in the market. 
21    And I think that's enormously helpful.  
22             As I think about my own role, I think there's 
23   also an element of the supporting that process of users 
24   in the market making a choice and having confidence that 
25   the choice that they made about fulfilling a need that 
0235
 1   they have was in fact the right one.  
 2             So I always describe it as looking at the 
 3   trading desk.  You're really looking for tools that you 
 4   have in a toolkit.  And we've talked a lot about 
 5   liquidity in credit markets, and the reality is there are 
 6   different types of liquidity.   
 7             There are on-the-run bonds; they're very small 
 8   sized trades that you might have sitting in your order 
 9   book.  Those might be perfectly appropriate for a central 
10   limit order book.  Okay?  Well, then the question would 
11   ultimately be:  Do we have one that we could use?  
12   Another part of the market is better for an RFQ platform. 
13    Do we have one of those?  
14             And so I think ultimately we're looking at this 
15   as, there's an opportunity set for different types of 
16   protocols and different types of trading platforms.  And 
17   really, what we want to be able to know is:  As we go out 
18   to the market and we make those choices, do we have a 
19   high level of confidence that we have all the information 
20   to have made the right choice?  And so I think that 
21   should really be the kind of two guiding principles as 
22   you go forward.  
23             MR. VULPIS:  I may have jumped the gun a little 
24   bit in terms of my opening.  But I'll reiterate a couple 
25   of things.  
0236
 1             Number one, I think that we need a set of rules 
 2   designed specifically for fixed income and not adopted 
 3   for fixed income.  Around transparency, not only knowing 
 4   a trade report the way we have today in terms of TRACE or 
 5   MSRB, that trades are happening.  
 6             But how are they transacting?  Where are they 
 7   transacting?  And all the various ways in which you can 
 8   execute electronically.  Because the transparency around 
 9   the marketplace is very important, we've heard that time 
10   and time again.  
11             How do you bring that transparency from a post-
12   trade basis into the pre-trade markets?  Something that I 
13   think we kind of all aspire to, but very, very tricky, 
14   and look to a lot of engagement with that, with the 
15   market participants because I think that one's going to 
16   be held up to be very important or the holy grail out 
17   there in terms of where we are, something that certainly 
18   we needed to have our eyes opened to.  
19             And lastly, as I mentioned, we do have multiple 
20   participants, whether they be participants represented on 
21   this panel right here or throughout the day.  And all 
22   rules and regulations need to apply to all participants.  
23             MR. REDFEARN:  Okay.  Thank you.  So moving on, 
24   and I wanted to address this one specifically to you, 
25   Ben, Bloomberg's obviously a very important player in the 
0237
 1   U.S. fixed income markets.  You're facilitating sell side 
 2   distribution of quotes and indications of interest in 
 3   corporate bonds broadly to institutional investors, and 
 4   enabling that information in a manner that helps to 
 5   identify trading opportunities and implement informal 
 6   decisions.  
 7             I just wanted to get your sense -- and I think 
 8   you made reference to some of this already -- but 
 9   elaborate just a little bit in terms of how Bloomberg's 
10   services in fixed income are different from sort of the 
11   ATSs and RFQ trading protocols.   
12             And do you believe these differences sort of -- 
13   what is the sort of way in which we should think about 
14   that in terms of sort of a different regulatory regime 
15   around this versus sort of the ATS regime?  
16             MR. MACDONALD:  Sure.  So I'm going to 
17   apologize to the people who had to listen to me on the 
18   first panel.  But if you think about how a business 
19   evolved, our purpose was not really the electronic 
20   trading business.  It was really bringing transparency to 
21   the market.  
22             And so, really, the business that we have today 
23   is the result of a single dealer franchise.  And so all 
24   the single data pages were were really a set of message 
25   boards that allowed the sell side to leverage us as a 
0238
 1   distribution mechanism to get onto the desktop of their 
 2   clients.  
 3             They control who they show pricing to.  They 
 4   are on a bilateral -- regardless of the format the 
 5   trading takes, in the end the conversation's bilateral 
 6   between two of our clients.  We don't get into the middle 
 7   of that.  
 8             And so if you think of the RFQ business or 
 9   request for quote business as pertains to Bloomberg, 
10   really all it is is giving a buy side investor the 
11   ability to launch a series of bilateral messages in an 
12   aggregated way to their counter parties they've 
13   permission to trade.  
14             So we don't do anything more than that.  We 
15   don't make the decision.  We provide a lot of analytics 
16   around that.  People have talked about the ALLQ page on 
17   Bloomberg; really, what ALLQ is is it allows people to 
18   look at the prices that are being made available to them 
19   by their counter parties in an aggregated way.  
20             And so, in essence, the experience their client 
21   can get is different depending on who they are and who 
22   they trade with.  And so there is no one size fits all in 
23   the market, which actually I think is a fair reflection 
24   of fixed income.  
25             We're non-ATS.  We day-to-day match orders.  We 
0239
 1   don't play a role in the trade.  We're not a counter 
 2   party to the trade.  It's name-disclosed.  Nobody trades 
 3   with Bloomberg; they're trading with whoever the counter 
 4   party is on the trade.  And the dealer and the client are 
 5   at all times in control of their decisions.   
 6             We're simply the trade messaging service.  
 7   We're not a broker dealer.  We don't get paid as a 
 8   broker.  We have no per-trade fee in the U.S.  And we 
 9   don't charge anything beyond what people pay for their 
10   terminal subscription.  And we don't settle or clear 
11   trades.  
12             So I do think it's incredibly important because 
13   we don't have a problem with regulation, and in fact, 
14   we've been regulated in Europe.  We've been regulated in 
15   the U.S.  And it's not, in 2018, that big a deal, I 
16   think, for anybody.   
17             What I think is important is actually how 
18   technology's evolving and bifurcating between platforms 
19   that operate, algorithms and technology that makes 
20   decisions on behalf of investors regardless of whether 
21   it's outsourced or insourced.  And trade messaging 
22   services, which whether you look at fixed network or 
23   whether you look at trade messaging services ourself are 
24   really just a transport layer.   
25             And so I think these two things are 
0240
 1   fundamentally different.  With that said, it's very hard 
 2   to argue that things like resiliency, client data, et 
 3   cetera, are important.  And in fact, I tend to take the 
 4   view, given the plurality of RFQ systems in the corporate 
 5   market in the U.S., that the ultimate measure of 
 6   resiliency is competition.   
 7             I mean, in effect, if we weren't operating a 
 8   resilient platform, as many of my peers who are sitting 
 9   on the desk, people wouldn't use us because they have 
10   choice.  And so I'm not sure.  With regards to 
11   resiliency, I think, again, there's a lot of nuance to 
12   that.   
13             And I think if you're talking about resiliency 
14   of technology that's making medications on behalf of its 
15   clients, I think that's one thing.  If you're talking 
16   about resiliency with regards to trade messaging, that's 
17   a different topic.  
18             MR. REDFEARN:  Thank you.  So Doug and Bill, 
19   you both operate fixed income ATSs.  Doug, I know you 
20   operate both a platform that's an ATS and one that's not 
21   an ATS.  I guess the question now is just sort of a 
22   contrast here.  
23             So is there anything, any of the aspects 
24   specifically of this ATS regime, that you think benefit 
25   or harm the fixed income markets?  And just to elaborate 
0241
 1   a little bit on the current Reg ATS regime as it applies 
 2   to fixed income securities, do you feel that it 
 3   adequately sort of covers the range of platforms out 
 4   there?   
 5             Would there be negative consequences of 
 6   expanding it or limiting it?  Are there any notable gaps 
 7   in the substantive regulation of ATS that you think that 
 8   regulators should be looking?  So maybe we'll start with 
 9   you, Doug.  
10             MR. FRIEDMAN:  Sure.  Thank you.  I think it's 
11   largely definitional.  So the ATS regime was meant to be 
12   an alternative of exchanges.  It was meant to have 
13   multiple buyers and sellers interacting on a 
14   nondiscretionary basis on orders.  And that is not what 
15   RFQ systems do.  At least most of them are fully 
16   disclosed, discretionary trading between two parties, and 
17   the venue's not necessarily intermediating.   
18             And so with that said, the ATS, there's a lot 
19   there that's good.  So it's not as if it's all bad.  I 
20   think what really turned the tide on the ATS regime was 
21   actually 15c3-5.  So when the -- known as the market 
22   access rule, not Ric's rule but a different rule -- was 
23   born largely out of concern over non-broker dealers 
24   entering into dark pools, essentially, with sponsorship 
25   by a broker dealer and borrowing their MBID to play in 
0242
 1   the dark pool.  
 2             And the idea would be that if you were 
 3   sponsoring that non-broker dealer, you had to have 
 4   controls and had to understand the financial limits that 
 5   they could -- in which they could participate.   
 6             And 15c3-5 as applied to RFQ systems, 
 7   particularly those that are a fully disclosed, where the 
 8   venue itself its not involved, really turned the tide, 
 9   wanted that it really work.  It was hard to, sort of on a 
10   per-trade basis, to have limits, and the venue itself 
11   wasn't the counter party.  It wasn't sponsoring that 
12   client into its RFQ system.  
13             And so it just didn't work.  And it was 
14   extremely burdensome, extremely costly, and it actually 
15   created a problem because you didn't -- as a venue 
16   operator, you didn't necessarily have the transparency 
17   you needed to comply with 15c3-5 vis-a-vis your non-
18   broker dealer participants.  And so that's actually a 
19   thing where the thing has got difficult.  And Reg SCI is 
20   sort of added to that a little bit to some degree.  
21             So I think, ultimately, the ATS regime, like I 
22   said, their only error, describing the manner of 
23   operation, having a system, describing your participants. 
24    Things like fair access, all those make perfect sense.   
25             I think it's a question of, definitionally, do 
0243
 1   you have, one, the right definition of who it covers?  
 2   And then two, if you have -- if there's something like 15 
 3   Series C35, does that make sense for all the markets 
 4   which fall under that umbrella?  And it may not be the 
 5   case.   
 6             And I think that's -- I think -- and again, 
 7   TRACE reporting, for example, if the venue's not in the 
 8   trade, it's not a counter party to the trade, should they 
 9   really be participating in the reporting of that and the 
10   costs associated with that?  
11             So there are a number of factors that I think 
12   go into that where having the label of ATS was easy, but 
13   it was burdensome in a lot of other respects.  
14             MR. REDFEARN:  So Bill, then on to you.  What 
15   aspects do you think of the ATS regime are benefitting or 
16   harming the fixed income markets?  And what are your 
17   thoughts about if there's any notable gaps in the 
18   substantive regulation of ATSs?  
19             MR. VULPIS:  So I'll make this -- kind of move 
20   this along quick because everything that Doug just 
21   mentioned before, given the fact that we're a registered 
22   ATS as well, I agree with.  
23             First, I think Reg ATS, while it's a set of 
24   operating rules for a platform like BondPoint to operate 
25   by, it doesn't adequately cover all the various protocols 
0244
 1   that are in the marketplace today.   
 2             Doug mentioned before some of the disadvantages 
 3   of the ATS is subject to today, and the cost involved, 
 4   regardless of whether the transaction is happening on the 
 5   platform and your broker dealer is sitting in between the 
 6   two clients, or whether it's a bilateral transaction.   
 7             There's reporting obligations and costs 
 8   associated with it.  And those are not necessarily bad.  
 9   I'm just saying it's a disadvantage to what the 
10   electronic marketplace is today because not everybody is 
11   playing by the same set of rules.  
12             One of the things that I'll add on to 15c3-5 
13   and market access is just onboarding a client.  As it is 
14   a registered ATS, it needs to be operated by a broker 
15   dealer.  And therefore, just onboarding of a client and 
16   all of the things that go forth with that in terms of 
17   your risks and setting risk limits and monitoring risk 
18   limits and all of that, as well as disclosures to that 
19   client, makes that onboarding process not just a 
20   paragraph sheet of paper and an agreement.   
21             It certainly makes that a process long, about 
22   three to four months, in terms of onboarding this 
23   particular client, going through risk controls and all of 
24   that.  So that's what I would have in addition to what 
25   Doug had mentioned before.  
0245
 1             MR. HEANEY:  So one more question and then I 
 2   think we're getting to the time when I'm going to turn it 
 3   back over to Michael for proceeding with this 
 4   recommendation.  But for Alex and Horace.  
 5             So the question I have for you is, as 
 6   institutional investors, fixed income broker dealers 
 7   obviously active participants in electronic trading 
 8   platforms of all types.  Are there material differences 
 9   to you as a user between platforms that are regulated as 
10   ATSs and other trading platforms?  Do you think that 
11   investors generally understand this distinction?  And if 
12   so, do you -- do they consider it to be meaningful?  
13             MR. CARTER:  As a broker dealer, candidly, we 
14   do not use one platform over another based on the 
15   regulation.  So I don't detect any difference, at least 
16   as it -- that we can attribute directly to one set of 
17   regulations over another.  
18             Largely, at least institutionally at Raymond 
19   James, we are reactive, and we are -- we have platforms 
20   that we prefer based on cost or protocol.  But we are 
21   reactive to what our clients want to use.  And so that's 
22   what's directing us.  
23             I don't think that our clients necessarily 
24   fully apprehend the differences between the two.  And 
25   then it seems that a 100 percent, or at least 99 percent, 
0246
 1   of their use is based on recordkeeping, retention of 
 2   recordkeeping for best ex, for their own audits, for 
 3   their own regulators.  
 4             So no, I don't say that.  I don't think that I 
 5   could specifically point to a set of regions that does or 
 6   does not favor a platform.  
 7             MR. REDFEARN:  Alex?  
 8             MR. SEDGWICK:  Sure.  I don't think there is 
 9   any particular benefit to whether or not somebody in the 
10   market is regulated as an ATS or not.  However, I would 
11   agree:  I don't think many investors or market 
12   participants truly understand the difference as to why 
13   one platform is regulated one way and another is not.  
14   And even in some cases, you have the same sort of 
15   provider or venue with multiple platforms, some of which 
16   are regulated as an ATS and some of which are not.  
17             When I think about the benefits beyond the ones 
18   that I've already enumerated, another area that I would 
19   point to is TRACE data.  I think it's a secondary 
20   benefit, but many of us have looked at trading volume on 
21   TRACE.  There's now an ATS flag.  I think there are 
22   people who would love to see that as some sort of 
23   authoritative estimate of electronic trading activity.  
24             However, there's a large amount of volume that 
25   goes across platforms that are not ATSs, and therefore 
0247
 1   are electronic but don't show up with that flag.  Maybe 
 2   that's a reason to have a different flag associated with 
 3   it.  But it would certainly be a downstream benefit.  
 4             MR. REDFEARN:  Okay.  So we have, because of 
 5   the timing considerations that we've had today, we have 
 6   had to try to move through this a little bit quicker than 
 7   we initially anticipated.   
 8             But I know that Michael and I think the 
 9   committee is eager to go ahead and have a vote on the 
10   recommendation.  So I'm going to turn it back over to 
11   Michael.  
12             MR. HEANEY:  So just prior to doing that, if 
13   there's any questions by committee members about the 
14   recommendation or particularly to the panel?  Larry?  
15             MR. TABB:  I'd just kind of echo the sentiments 
16   of the panel that these protocols are developing.  And I 
17   think what you want to try to do is be more disclosure-
18   oriented around the rules and protocol-specific or 
19   defining, we can only use RFQ or dark pools are banned or 
20   anything like that -- I think open but disclosed.  So I 
21   think that's what I would be in favor of.  
22             MR. HEANEY:  Others?  Rachel?  
23             MS. WILSON:  Yes.  Just as a practical matter, 
24   as this is a recommendation. It's really about working 
25   together amongst the other groups.  I was sort of 
0248
 1   curious, as we get into the recommendation phase of this, 
 2   how that's been thought through as to how that would 
 3   actually come together, if there's further thought around 
 4   this?   
 5             Because this is a base recommendation.  If we 
 6   should consider it together, that's -- to be honest, your 
 7   kind of wish was fairly noncontroversial.  But I think a 
 8   little bit more meat on the bone would be great.  
 9             MR. HEANEY:  Okay.  Let me just add to that 
10   some of the points at least I heard from the panelists, 
11   just to throw it out here for the record because I think 
12   they were all interesting, which is:  Recognizing 
13   principal-based regulation; time and cost, including 
14   opportunity cost; consistency across similar platforms.   
15             Simplified solutions; investor protection 
16   versus the cost associated; forward-looking technological 
17   regulation; well-designed regulation that could fuel 
18   further growth in markets; price formation transparency; 
19   fee structures; consumption of data that's easily mined; 
20   greater coverage of increased protocols in today's 
21   market; and then the ATS flag, which I think is very 
22   relevant.  
23             So these are some of the things, as we think 
24   about taking this forward for a group of three to work 
25   together, these are some of the themes hopefully we can 
0249
 1   incorporate and deliver to that.   
 2             I don't want to sidestep Rachel's question.  
 3   I'm going to pass it along to the SEC folks to answer the 
 4   question of how the three would work together and how 
 5   that actually -- kind of the mechanisms for that.  
 6             MR. REDFEARN:  I think that there is a lot -- 
 7   there's still some -- there are some things that we would 
 8   be working out there.  But we have worked together on 
 9   other things, and so I think the general idea would be to 
10   simply get together and take all of the information that 
11   we've gotten from this particular committee, from all the 
12   substance that has come up here, and sit around a table 
13   and just start to hammer out some of those details.  
14             And I think it probably would make sense for -- 
15   at some point in time for us to kind of come back and 
16   talk to the committee about some of the things that are 
17   coming up, but -- and continuing to get guidance that we 
18   can possibly get from this group in that process.  
19             But it's kind of hard to prejudge exactly those 
20   details, except I will say that I think that this has 
21   been a pretty rich set of discussions that we've had here 
22   today.  And I think Michael had a very good -- a good 
23   starting LIS.  I have a lot of different notes that I 
24   think that myself and our team will be digesting to help 
25   guide us.  
0250
 1             But I think that -- I think there's certainly a 
 2   lot to do.  My concern would be some of the comments that 
 3   have been made here, including making sure the cost-
 4   benefit of this is right, making sure that it's not 
 5   overly prescriptive, making sure that it's not 
 6   potentially restricting competition but encouraging 
 7   competition, and perhaps more of a disclosure-oriented 
 8   approach to a certain extent.  
 9             So those are some of the themes that I've 
10   heard.  And I think that the most important thing in this 
11   process is effectively, do no harm.  We're not trying to 
12   create a regulatory regime, a regulation for no purpose 
13   whatsoever; to be very focused in a value-added way.  And 
14   I think that's -- those would be some of the guiding 
15   principles that we would be keeping in mind.  
16             But obviously, we would be having that 
17   discussion and sort of working out those details as we 
18   go.  
19             MR. MCVEY:  Can I just add onto that, Brett, 
20   too, just to provide a little bit of color on some of the 
21   committee's debates on this?  On the path that we're on 
22   right now, I think one of the primary concerns is that 
23   we're going to have three regulatory agencies focusing on 
24   the same regulatory oversight issues, which we've thought 
25   was not good from a cost-benefit standpoint or even a 
0251
 1   consistency standpoint in the approach.  
 2             So we felt the time was not to have the 
 3   regulatory bodies working together, talking about the 
 4   issues together, and determining who has oversight over 
 5   which issues.  So you do right now have most of the 
 6   electronic trading regulatory experience within the SEC 
 7   ATS Division.  You have the traditional broker dealer 
 8   responsibilities sitting within MSRB.  But they all have 
 9   a piece in the oversight and regulatory responsibilities 
10   for the development of fixed income, e-trading as it 
11   stands today.   
12             So we didn't know exactly where the working 
13   group will end up on that.  But we thought it was a very 
14   bad path to be going down, to start to see the overlap 
15   and redundancy growing on three different regulatory 
16   bodies focusing on the same issues.  
17             MR. HEANEY:  I'll just echo, having 
18   participated on that subcommittee, that was where the 
19   broad support from the subcommittee came from, that kind 
20   of generation of thought.  
21             So I think at this point I'll entertain a 
22   motion to vote on the recommendation.   
23             (Motion made.)  
24             MR. HEANEY:  If I could have a second.   
25             (Second.) 
0252
 1             MR. HEANEY:  Thank you.  
 2             Can I just check who's on the phone?  Elisse, 
 3   are you on the phone?  
 4             MS. WALTER:  Yes, I am.  
 5             MR. HEANEY:  Matt, are you on the phone?  
 6             MR. ANDRESEN:  Yes.  
 7             MR. HEANEY:  And Suzanne, are you on the phone? 
 8             MS. SHANK:  Yes, I am.  
 9             MR. HEANEY:  Thank you.  And Lynn, are you on 
10   the phone?  
11             MS. MARTIN:  Yes, I am.  
12             MR. HEANEY:  Okay.  Great.  So I would now ask 
13   for all voting members to raise their hands if in favor 
14   of --  
15             (Hands raised.) 
16             MR. HEANEY:  Let me just go to the phone, 
17   please.  That was 12 in favor in the room, just so those 
18   on the phone are aware.  
19             Elisse, can I ask if you are in favor or 
20   opposed?  
21             MS. WALTER:  Yes.  In favor.  
22             MR. HEANEY:  Thank you.  Matt?  
23             MR. ANDRESEN:  In favor.  
24             MR. HEANEY:  Thank you.  Suzanne?  
25             MS. SHANK:  I'm in favor.  
0253
 1             MR. HEANEY:  Thank you.  And Lynn, please?  
 2             MS. MARTIN:  In favor.  
 3             MR. HEANEY:  Thank you.   
 4             Do I have anybody who is opposed?  
 5             (No response.) 
 6             MR. HEANEY:  Thank you.  And any abstentions?  
 7   I don't think we have any.  
 8             (No response.) 
 9             MR. HEANEY:  So the recommendation has been 
10   approved by the committee.  Rick, I thank you very much 
11   for your leadership on the subcommittee, and all those on 
12   the subcommittee who got us to this point.  We'll be 
13   taking this back to, obviously, the commission and will 
14   work on this from hereon going forward.  
15             I'd also like to take this opportunity to thank 
16   the panelists.  This was, again, another rich discussion 
17   of, really, the particulars and the details and things 
18   that this committee very much needed to listen to here 
19   and learn.  So I thank you for your time.  I thank you 
20   for coming down to D.C. as well for this.  And again, 
21   thank you all for the vote.  
22             MR. REDFEARN:  Likewise.  Thank you all very 
23   much.  We appreciate it.  I'm sorry that we were a little 
24   rushed in discussion.  But I think -- so we're going to 
25   take a break now, and then we're going to pick up the 
0254
 1   last one.  
 2             MR. HEANEY:  We have a five-minute break.  
 3   Five-minute break.  Then we go to ETFs and Bond Funds, 
 4   and then we'll adjourn.  
 5             (A brief recess was taken.) 
 6             MR. HEANEY:  Okay.  The lovely distinction of 
 7   wrapping up the day.  Rachel and Kumar.  
 8             So we have the ETF and Bond Fund Subcommittee 
 9   update.  And Rachel and Kumar have graciously offered to 
10   do it.  And so we're going to hear about the plans of 
11   initiatives over the next couple of months.  I think, 
12   Rachel, you're going to kick off?  Thank you.  
13             MS. WILSON:  Happy to.  I will note that we are 
14   a considerably smaller crowd at this point.  And I think 
15   next time the strategy will be to bring a panel.  
16             (Laughter.) 
17             MS. WILSON:  Or go first.  And don't go on 
18   Amazon Prime Day.  That was another bad mistake.  
19             We're the ETF and Bond Fund Subcommittee.  And 
20   just to kind of wrap up with the group and give you an 
21   update with what we've been focused on, we've really been 
22   focused on the topic of classification schemes for 
23   exchange-traded funds.  We chose a guiding objective, and 
24   that is really truth in labeling.  All right?   
25             So we chose that as our objective, truth in 
0255
 1   labeling.  And we defined our goals to provide enhanced 
 2   clarity for self-directed investors as well as those with 
 3   financial advisors.  So we really tried to consider the 
 4   breadth of all investors out there.  
 5             So we had -- during our May subcommittee, we 
 6   did ask about an SEC rule that was proposed in 2008 
 7   addressing ETF classification that was not adopted at 
 8   that time.  And I was told that some version of an ETF 
 9   proposal rule would be published shortly, which indeed it 
10   was, on 6/28, post our subcommittee meetings.  
11             We were informed that our further research and 
12   work around this topic would be added to this effort, and 
13   this was confirmed in the SEC's 6/28 press release, which 
14   noted that they will have "continued consultation with 
15   our Investment Advisory Committee and Fixed Income Market 
16   Structure Advisory Committee," with today being another 
17   example of this continued joint exploration.  
18             As a first order question and as part of the 
19   truth in labeling objective, our subcommittee started to 
20   consider if the ETF label should be reserved.  And we had 
21   a nice debate around this concept of reserved.  For 
22   example, does a retail investor truly understand what an 
23   ETF is?   
24             There are complex ETFs that might kind of 
25   change that definition for some; and also something about 
0256
 1   the exchange mechanism and really understanding, and also 
 2   how price can move from NATIVE.  So really some thoughts 
 3   around:  What is the importance of reserving what ETF 
 4   means?  
 5             And for example, we looked at two sub-
 6   definitions for this reservation concept.  One was:  
 7   "Should the ETF label be limited to non-complex funds 
 8   registered under the 1940 Act, UCITS, or other 
 9   commensurate regulatory regime?"   
10             The second was an ETF could be defined as any 
11   investment vehicle that uses an arbitrage mechanism?  
12   Specifically per Henry Hu and John Morley in their 
13   publication, "Regulatory Framework for ETFs," an ETF 
14   could be defined as a "pooled investment vehicle that 
15   publicly trades on an exchange and that permits an 
16   investor to create and redeem shares in exchange for 
17   assets or combination of assets in cash and values equal 
18   to the NCDS net asset value."  So again, two different 
19   kind of concepts for reservation.  
20             The currently proposed SEC new ETF rule, which 
21   is 6c-11, which is in the public comment period right 
22   now, the proposed rule has in essence defined an ETF 
23   through its application.  6c-11 has limited its 
24   application to those ETFs organized as open funds, the 
25   structure for the vast majority of ETFs today.  
0257
 1             ETFs organized as a unit in investment trust, 
 2   ETFs structured as a shared class of a multi-class fund, 
 3   and leveraged or inverse ETFs would not be able to rely 
 4   on the proposed rule.  This application lines closely to 
 5   the reservation complex that we discussed as a 
 6   subcommittee around the non-complex funds, or the first 
 7   one that I just described.  
 8             In addition to defining ETFs that would be 
 9   eligible for the potentially streamlined approval process 
10   versus seeking a special order from the SEC today to 
11   operate, which is a lengthy process, the proposed rule is 
12   intended to create a level playing field by allowing all 
13   sponsors to use custom baskets.  
14             Ultimately, the rule is intended to facilitate 
15   greater competition and innovation in the ETF marketplace 
16   by lowering barriers to entry while with standing 
17   conditions designated to create a level playing field 
18   amongst most ETFs.  Investors are similarly expected to 
19   benefit with proposed disclosure amendments that would 
20   enhance information provided to them.  
21             6c-11 also imposes certain conditions around 
22   transparency, custom basket policies and procedures, and 
23   website disclosure.  Specifically and per the press 
24   release, these include:  Transparency -- ETFs would be 
25   required to provide daily portfolio transparency on their 
0258
 1   websites.  Custom basket policies and procedures -- an 
 2   ETF relying on proposed rule 6c-11 would be permitted to 
 3   use baskets that do not reflect a pro rate representation 
 4   of the fund's portfolio, or that differ from other 
 5   baskets used in transitions on the same business day, 
 6   a.k.a. custom baskets, if the ETF adopts written policies 
 7   and procedures setting forth detailed parameters for the 
 8   construction and acceptance of custom baskets that are in 
 9   the best interests of the ETF and its shareholders.  The 
10   proposed rule would also require an ETF to comply with 
11   certain recordkeeping requirements.  
12             To help create a consistent ETF regulatory 
13   framework, the proposal recommends rescinding exemptive 
14   relief previously granted to ETFs that would be able to 
15   rely on the rule.  This includes exempting relief 
16   permitting ETFs to operate in a master feeder structure, 
17   which remains an uncommon structure.  However, ETFs in 
18   the current -- who are currently using this structure 
19   would be grandfathered.  
20             Finally, the Commission is proposing several 
21   form amendments to N-1A and Form N-8B-2 to provide more 
22   useful ETF-specific information to investors who purchase 
23   ETF shares on an exchange.   
24             This information would include costs borne by 
25   ETF investors that are applicable to mutual funds, such 
0259
 1   as trading costs.  
 2             The proposal is specifically seeking comment 
 3   around some of the following themes, which we will 
 4   similarly seek to discuss in subcommittee, such as:  Do 
 5   retail investors understand that an ETF does not sell or 
 6   redeem individual shares?  That shares can only be bought 
 7   or sold on secondary markets?  If not, what disclosure 
 8   might help clarify this?  
 9             Should the rule consider other limitations 
10   regarding ETF sales literature?  To further prevent 
11   investors from confusing ETFs with mutual funds, should 
12   the rule require an ETF to include the identifier "ETF" 
13   in its name?  This is something we spent a lot of time on 
14   in our discussions as well.  
15             To further prevent investors from confusing 
16   ETFs with mutual funds, should the rule require an ETF to 
17   explicitly disclose in its sales literature that 
18   shareholders may pay more than NAV when buying shares and 
19   may receive less than NAV when selling ETF shares?   
20             Should the rule impose any additional 
21   considerations or require any additional disclosures to 
22   help investors distinguish ETFs from other ETPs, such as 
23   exchange-traded notes or commodity pools that are not 
24   subject to the Investment Company Act?  Should the 
25   Commission consider proposing naming conventions based on 
0260
 1   these or other distinctions in a future rulemaking?  Are 
 2   naming conventions useful to investors?  
 3             Prior to the rule publication, as a 
 4   subcommittee we had already been discussing choices in 
 5   classification systems.  When tested in practice, some of 
 6   these distinctions draw sharper lines amongst ETFs while 
 7   others may provide greater truth in labeling.  
 8             Our subcommittee reviewed a few academic papers 
 9   on this topic, but this remains a relatively less-studied 
10   area.  We intend to explore these classification options 
11   in further detail, but some of the classification choices 
12   include:  passive versus active investing strategy or 
13   rules-based versus discretionary; long, short, and 
14   leveraged exposures; creating of redemption mechanisms. 
15             We're additionally addressing how 
16   classification systems may play out in practice.  Sales 
17   and marketing metrics, prospectus, required fund name 
18   change, and how actually some of these factors might be 
19   applied retroactively or not.  We had a good debate about 
20   the difficulty in real life of imposing some of these 
21   changes in practice.  
22             Given the now specified areas for comment in 
23   the proposed Rule 6C-11, we will not only continue with 
24   our efforts around exploring classification options and 
25   applying them in practice, but also take a deeper look at 
0261
 1   the questions raised for public comment in this proposal.  
 2             So that's a good summary, I think, of where 
 3   we've focused.  
 4             MR. HEANEY:  Kumar, anything to add?  
 5             MR. VENKATARAMAN:  Thank you, Michael.  Rachel 
 6   did a really good job of describing the main discussions 
 7   in the committee.  I just had one minor thing to add, 
 8   which is that we've also spent quite some time talking 
 9   about liquidity in the ETF space and how liquidity from 
10   ETF investors might have an impact on the underlying 
11   bonds.  
12             And we've had some conversations regarding what 
13   would be the impact for ETF investors in a period of 
14   stress when the "create redemption" process might be 
15   under some strain, and what might be the impact on the 
16   bid/ask spreads and whether, in terms of truth in 
17   disclosure, it might be appropriate to have some 
18   information provided to investors that the cost at which 
19   they may be able to buy or sell the ETFs may be different 
20   from the NAV, particularly during periods of stress.  
21             So these are some of the additional topics that 
22   we have discussed with respect to the liquidity of the 
23   security.  Thank you.  
24             MR. HEANEY:  Thank you both, and we look 
25   forward to more of the work from the subcommittee at the 
0262
 1   next meeting.  I just want to take this opportunity to 
 2   thank everybody for what has been a full day, more full 
 3   for the people still at this table than others, but for a 
 4   full day and with a lot that got done.  So thank you, and 
 5   for all of FIMSAC.   As I said early in my opening 
 6   comments, for the amount of work that's being done by 
 7   this group with full-time jobs, it is admirable.  
 8             I just want to remind everybody that the next 
 9   meeting is October 29th, that we had moved that back one 
10   week.  And we'll send out another email that just reminds 
11   people of that.  
12             MR. HEANEY:  So at this point I'll entertain a 
13   motion to adjourn.  
14             (Motion made.)  
15             MR. HEANEY:  All in favor?  
16             (A chorus of ayes.)  
17             MR. HEANEY:  Safe travels.  Thank you. 
18             (Whereupon, at 5:17 p.m., the Fixed Income 
19   Market Structure Advisory Committee was adjourned, to 
20   reconvene October 2, 2018.) 
21                           * * * * * 
22     
23   
24   
25   
0263
 1           PROOFREADER'S CERTIFICATE
 2   
 3   In the Matter of:  FIXED MARKET INCOME ADVISORY MEETING
 4   File No.:          OS-0716
 5   Date:              Monday, July 16, 2018
 6   Location:          Washington, D.C.
 7   
 8        This is to certify that I, Christine Boyce,
 9   (the undersigned) do hereby swear and affirm that the
10   attached proceedings before the U.S. Securities and
11   Exchange Commission were held according to the record,
12   and that this is the original, complete, true and 
13   accurate transcript, which has been compared with the
14   reporting or recording accomplished at the hearing.
15   
16   
17   _____________________________   ________________
18   (Proofreader's Name)               (Date)
19   
20   
21   
22   
23   
24   
25