0001
 1           U.S. SECURITIES AND EXCHANGE COMMISSION 
 2                                
 3                                
 4                                
 5                                
 6                 MEETING OF THE FIXED INCOME 
 7             MARKET STRUCTURE ADVISORY COMMITTEE 
 8   
 9                      AMENDED 4-27-2018
10   
11   
12   
13                    Monday, April 9, 2018 
14                          9:30 a.m. 
15   
16   
17   
18   
19   
20   
21   
22   
23   
24           U.S. Securities and Exchange Commission 
25             100 F Street, N.E., Washington, D.C. 
0002
 1   PARTICIPANTS: 
 2   COMMITTEE MEMBERS: 
 3   Michael Heaney, Committee Chair 
 4   Jay Clayton, Commission Chair 
 5   Kara Stein, Commissioner  
 6   Michael Piwowar, Commissioner  
 7   Hester Peirce, Commissioner  
 8   Brett Redfearn 
 9   John Bagley 
10   Kumar Venkataraman 
11   Larry Harris 
12   Tom Thees 
13   Rachel Wilson 
14   Larry Tabb 
15   Tom Gira 
16   Amar Kuchinad 
17   Gilbert Garcia 
18   Rick McVey 
19   Horace Carter 
20   Ananth Madhavan 
21   Dan Allen 
22   Lynn Martin 
23   Carole Brown
24   Amy McGarrity
25    
0003
 1    
 2   PARTICIPANTS (Continued): 
 3   Mihir Worah 
 4   Scott Krohn 
 5   Brian Archer 
 6   Joe Arcadi 
 7   Matt Berger 
 8   Sean Collins 
 9   Krishna Memani 
10   Gregory Peters 
11   Melissa Gainor 
12   Nick Goetze 
13   Gary Mottola 
14   Matthew Andresen 
15   Robert Colby 
16   Renzo Iturrino 
17   Kristin Maher 
18   Thomas Vales 
19   Rebecca Olsen
20   Jeff Harris
21   David Shillman
22   John Roeser
23   Tom Eady
24   David Dimitrious
25     
0004
 1                      C O N T E N T S 
 2                                                    PAGE 
 3   Welcome Remarks                                    5 
 4    
 5   Block Trade Dissemination Draft Recommendation    15 
 6    
 7   Liquidity Considerations for Bond ETFs            84 
 8    
 9   Lunch Break/Administrative Session               132 
10    
11   Retail Investor Disclosure and Education         132 
12    
13   Electronic Trading in the Retail Market          161 
14    
15   Discussion of Committee Next Steps               220 
16    
17   Adjourn                                          221 
18    
19    
20    
21    
22    
23    
24     
25    
0005
 1                     P R O C E E D I N G S 
 2             CHAIRMAN HEANEY:  Good morning, all.  I believe 
 3   we have a quorum, so I would like to call the meeting to 
 4   order. 
 5             Thank you for joining us for the second annual 
 6   SEC Fixed Income Market Structure Advisory Committee.  I 
 7   will begin by welcoming Chairman Clayton and asking the 
 8   Chairman to make his opening remarks. 
 9             CHAIRMAN CLAYTON:  Thank you, Michael. 
10             I am delighted to welcome you to the second 
11   meeting of the Fixed Income Market Structure Advisory 
12   Committee.  Our inaugural FIMSAC meeting in January 
13   focused on bond market liquidity and was extremely 
14   productive.  I look forward to a similarly constructive 
15   discussion today. 
16             I would like to recognize and thank all of the 
17   committee members, including Committee Chairman Michael 
18   Heaney.  We are grateful for your service and I am 
19   heartened by your efforts in the short time the committee 
20   has been operating. 
21             Following the first meeting, the committee 
22   members organized themselves into three subcommittees, 
23   each of which is focusing on a key topic area in 
24   corporate and municipal bond markets.  The first is a 
25   transparency subcommittee, which is considering issues 
0006
 1   relating to pretrade and posttrade transparency in these 
 2   markets.  The second is an ETF and bond fund 
 3   subcommittee, which is considering the impacts and 
 4   implications of the growing number of registered funds 
 5   that are active in these markets.  The third is a 
 6   technology and electronic trading subcommittee, which is 
 7   considering the impact of the growth of electronic 
 8   trading platforms and the increased use of other 
 9   electronic systems on the liquidity, efficiency and 
10   resiliency of these markets.  I am eager to hear updates 
11   today from each of these subcommittees.   
12             In closing, I will make a specific comment and 
13   a general observation.  My specific comment relates to 
14   pre- and posttrade transparency and, in particular, the 
15   recognition that reporting rules can substantially affect 
16   behavior of market participants.  In turn, those 
17   individual behavioral effects in the aggregate can have 
18   substantial market effects. 
19             I know that our panelists know this well from 
20   varying perspectives.  However, my sense is that many 
21   market participants and commentators do not understand 
22   the significant linkages among trade reporting, 
23   liquidity, volatility, efficiency, risk, cost of capital 
24   and fairness.  I ask our panelists to please explain 
25   their views on these linkages.  In particular, I ask them 
0007
 1   to focus on whether our reporting rules best serve the 
 2   long-term interests of our Main Street investors. 
 3             My general comment relates to the Commission's 
 4   continuous effort to identify emerging risks and issues, 
 5   including those that arise directly in the fixed income 
 6   markets, as well as developments in other areas that may 
 7   affect the fixed income markets.  We are keenly aware 
 8   that financial markets are interconnected and 
 9   interchanging and, as a result, the risk landscape is 
10   ever changing.  Today's fixed income trading markets are 
11   markedly different from those of a decade ago and, 
12   accordingly, we should assume that the risk landscape is 
13   different.   
14             The subcommittees you have formed and the 
15   initial topics you have identified reflect this view, and 
16   I thank you for focusing on market risk.  I hope that 
17   this committee will continue to help the Commission in 
18   this regard. 
19             Once again, thank you all very much. 
20             CHAIRMAN HEANEY:  Thank you, Chairman Clayton. 
21             I will now ask the commissioners to make their 
22   opening statements, starting with Commissioner Stein. 
23             COMMISSIONER STEIN:  Thank you, Michael.  I 
24   have a very brief opening statement.  I really just want 
25   to say good morning and welcome, everyone, to second 
0008
 1   meeting of our Fixed Income Market Structure Advisory 
 2   Committee. 
 3             I want to thank both the members of the 
 4   committee and our panelists for taking time out of very 
 5   busy schedules to be with us today and provide us with 
 6   your pro bono services.  I know that the committee has 
 7   been very busy since our last meeting and, indeed, that 
 8   one of the subcommittees, our transparency subcommittee, 
 9   already has a transparency recommendation regarding the 
10   reporting regime for block-size trades in corporate 
11   bonds.  So thank you for your hard work on this issue and 
12   I look forward to the discussion today. 
13             I am also looking forward to today's other 
14   discussions.  The two panels on retail issues should help 
15   inform us about the issues facing retail investors in 
16   these markets.  And the liquidity of bond ETFs is another 
17   very important topic that has implications for both 
18   retail and institutional investors.  A lot has been 
19   written on this subject and I look forward to hearing 
20   everyone's views.   
21             So once again, thank you for participating and 
22   coming to today's meeting and I look forward to it. 
23             CHAIRMAN HEANEY:  Thank you, Commissioner 
24   Stein. 
25             Commissioner Piwowar. 
0009
 1             COMMISSIONER PIWOWAR:  Thank you, Michael. 
 2             In keeping with today's theme, my comments will 
 3   be brief.  We have a packed agenda and we're all eager to 
 4   dive into the discussion.  However, I do want to address 
 5   my thanks once again to the members of the committee for 
 6   their willingness to invest their time and talents in 
 7   this endeavor. 
 8             As the members are now well aware, 
 9   participation in an advisory committee of this nature 
10   does not mean merely showing up once a quarter to sit at 
11   a long table in front of a big crowd.  As glamorous as 
12   these events may be -- we've even provided you with free 
13   coffee, which is a rare occurrence in this building -- 
14   they represent only a portion of the work our FIMSAC 
15   members have put in already.  I appreciate the many hours 
16   you have dedicated to the subcommittee meetings and other 
17   behind-the-scenes work, without which the impact of this 
18   committee would be severely limited. 
19             I would also like to join Chairman Clayton in 
20   thanking Michael Heaney for his leadership of this 
21   committee, as well as Commission Staff in the Division of 
22   Trading and Markets, Office of Municipal Securities and 
23   the Division of Economic and Risk Analysis, who have all 
24   been an integral part of this process. 
25             My final thanks go to the many industry 
0010
 1   representatives and academics who have participated in 
 2   conversations with the subcommittees, as well as those 
 3   who serve today as panelists.  Broad input from industry 
 4   and academia is a vital component of a successful 
 5   advisory committee.  We are not merely relying on the 
 6   expertise of these 23 FIMSAC members, even though the 
 7   knowledge around the table is unparalleled; we are also 
 8   pulling in viewpoints from across the industry and 
 9   academia and I have already heard numerous positive 
10   contributions from nonmembers of the subcommittee 
11   discussions. 
12             With respect to the topics on today's agenda, I 
13   understand that the subcommittees have each generated 
14   vigorous debates and I look forward to robust discussions 
15   among members of the full committee.  Thank you. 
16             CHAIRMAN HEANEY:  Thank you, Commissioner 
17   Piwowar. 
18             Commissioner Jackson. 
19             COMMISSIONER PEIRCE:  I think --  
20             CHAIRMAN HEANEY:  I apologize. 
21             COMMISSIONER PEIRCE:  Thanks, Michael.  And 
22   thank you for your work on the committee.  And thanks to 
23   all of you for being here today. 
24             As Commissioner Piwowar mentioned, there is a 
25   lot of work that goes on between meetings, I know.  And 
0011
 1   so thank you for all your work at subcommittee meetings. 
 2             We all know that there have been a lot of 
 3   changes in the fixed income markets due to 
 4   electronification, the rise of ETFs, many other factors 
 5   including regulatory factors, such as the Volcker Rule 
 6   and our impending muni markup rule.  And so these things 
 7   are all having an effect. 
 8             But I am also eager to hear from all of you 
 9   where we need to be engaged in rulemaking, or where our 
10   rules potentially are part of the problem in preventing 
11   incumbent ideas and incumbent technologies and incumbent 
12   companies from coming in to improve these markets.  So I 
13   look forward to the discussion today and in the upcoming 
14   months to working on recommendations that you all bring 
15   to us.  Thank you. 
16             CHAIRMAN HEANEY:  Thank you, Commissioner 
17   Peirce. 
18             Next, I'd like to turn it over to Brett 
19   Redfearn, Director, Division of Trading and Markets, and 
20   the committee's designated federal officer for his 
21   remarks. 
22             MR. REDFEARN:  Thank you, Michael. 
23             On behalf of the Division, I would like to 
24   welcome everyone to our second FIMSAC meeting.  Let me 
25   briefly introduce my colleague sitting here with me 
0012
 1   today.  From the Division of Trading and Markets, we have 
 2   Dave Shillman, John Roser, both associate directors in 
 3   the Office of Market Supervision; Tom Eady, senior policy 
 4   advisor in the Office of Analytics and Research; and 
 5   David Dimitrious, senior special counsel in the Office of 
 6   Market Supervision. 
 7             Also joining us on the other side are Rebecca 
 8   Olsen, acting director of the Office of Municipal 
 9   Securities, and Jeff Harris, the chief economist and 
10   director of the Division of Economic and Risk Analysis. 
11             Before we get started, I just want to say that 
12   any views by the Staff in this forum are ours alone and 
13   cannot be attributed to the Commission or the 
14   commissioners. 
15             So following our meeting in January, the 
16   committee formed these three subcommittees that Chairman 
17   Clayton mentioned earlier that will make presentations to 
18   the committee today.  I want to thank you all for your 
19   active engagement in the subcommittees and your 
20   willingness to solicit input from a variety of 
21   stakeholders, including those not represented on the 
22   committee. 
23             I know you all have significant demands on your 
24   time and I really appreciate your dedication and your 
25   important work to these subcommittees and to this 
0013
 1   committee.  We will all greatly benefit from your 
 2   insights and expertise. 
 3             I also want to offer special thanks to three 
 4   members of my staff in Trading and Markets, David 
 5   Dimitrious, Tom Eady and Ben Bernstein, who have been 
 6   working very hard to help ensure that this committee is a 
 7   success, including Michael Heaney, who is doing an awful 
 8   lot to keep this all moving. 
 9             The subcommittees have begun exploring a range 
10   of important issues with the corporate bond and municipal 
11   securities markets, both in the institutional and retail 
12   segments, and many of the issues that are coming up will 
13   be -- will be highlighted today.  The first item that we 
14   will be discussing is a preliminary recommendation from 
15   the transparency subcommittee concerning a block trade 
16   dissemination pilot.  The subcommittee's presentation of 
17   a preliminary recommendation today is a testament to 
18   their genuine commitment to producing actionable 
19   recommendations for the Commission's consideration, 
20   consistent with Michael's challenge to the committee at 
21   the first inaugural meeting last January. 
22             Transparency was a prominent discussion topic 
23   at FIMSAC's inaugural meeting, and the optimal level of 
24   posttrade transparency in the block market was a key 
25   aspect of that discussion.  Transparency, as we all know, 
0014
 1   is critical for robust market structure and striking the 
 2   right balance is important.  On the one hand, 
 3   transparency can promote efficient markets through lower 
 4   search costs and greater price competition.  On the other 
 5   hand, too much transparency could impair liquidity and 
 6   market quality in certain market segments if it increases 
 7   risk in the provision of capital or the likelihood of 
 8   market impact.  The goal is to ensure that the level of 
 9   transparency is appropriately calibrated to accommodate 
10   the unique characteristics of that market segment. 
11             As we will hear shortly, the subcommittee's 
12   preliminary recommendation today proposes a pilot to 
13   assess whether a more calibrated approach to today's 
14   transparency regime for large trades may improve the 
15   market for corporate bond investors.  In particular, the 
16   proposed pilot would raise the current TRACE caps for 
17   investment grade and high-yield bonds while also delaying 
18   the public dissemination of trade reports above those 
19   caps for 48 hours and shortening the time until the full 
20   size of trades above the caps are unmasked. 
21             Based on 2017 data, by proposing to raise the 
22   caps, the recommendation would enhance transparency, 
23   into the precise size of over 20 percent of trades for 
24   investment-grade bonds and over 25 percent for high-yield 
25   bonds.  At the same time, by delaying public 
0015
 1   dissemination for the larger trades, the recommendation 
 2   is designed to assess whether this change in the 
 3   dissemination protocols ultimately enhances market 
 4   quality and liquidity for corporate bond investors. 
 5             I look forward to the full committee's 
 6   consideration of the preliminary recommendation.  And as 
 7   Mihir will explain in more detail this morning, key 
 8   decision points were the subject of debate and at times 
 9   differences in opinion.  A goal of today's meeting is for 
10   the full committee to hear from the transparency 
11   subcommittee about these decision points and to fully 
12   deliberate those issues along with the preliminary 
13   recommendation. 
14             Next, I just want to spend one minute on 
15   process, given the presentation of the subcommittee's 
16   preliminary recommendation to the full committee.  It is 
17   possible that the full committee may reach sufficient 
18   agreement on the draft recommendation today and vote to 
19   make a formal recommendation to the Commission.  In the 
20   course of today's discussion, however, if the committee 
21   decides that a recommendation needs further consideration 
22   and refinement beyond what might be achieved during 
23   today's discussion, then the subcommittee would meet as 
24   necessary to revise its recommendation in light of the 
25   feedback received from the full committee.  A new 
0016
 1   recommendation would then be presented at one of the full 
 2   committee's quarterly meetings we've scheduled, or we 
 3   could hold an ad hoc public meeting telephonically to 
 4   vote on a revised recommendation. 
 5             Finally, I want to emphasize that the 
 6   interested parties may also submit comments -- any 
 7   interested parties can submit comments on the work of the 
 8   committee via the FIMSAC webpage on the SEC's website.  
 9   The preliminary recommendation that the committee will 
10   consider today is available to the public on FIMSAC's web 
11   page, and all preliminary recommendations brought to the 
12   committee for consideration will likewise be publicly 
13   available prior to consideration at a FIMSAC meeting.  So 
14   I encourage members of the public to review these 
15   materials and participate in the policymaking process by 
16   offering their input. 
17             Thank you all again today for your willingness 
18   to serve on this committee.  I look forward to the 
19   discussions.  And now, Michael, I will turn it back over 
20   to you. 
21             CHAIRMAN HEANEY:  Thank you, Brett. 
22             Again, thank you all for joining us for our 
23   second meeting of the FIMSAC.  Again, just to briefly 
24   recap, and you've heard it a little bit, at our inaugural 
25   meeting in January, we engaged in a robust discussion 
0017
 1   concerning the state of liquidity and the drivers of 
 2   liquidity conditions in the U.S. corporate bond market. 
 3             The Industry Research Panel helped us bring the 
 4   years' long debate about liquidity conditions up to date. 
 5    The Market Participant Panel provided insight into the 
 6   practitioners' views of liquidity conditions and 
 7   behavioral responses to the ongoing evolution of 
 8   liquidity supply and demand. 
 9             Our discussion at the first meeting helped 
10   highlight several aspects of the corporate and municipal 
11   bond markets that we believed was ripe for further study. 
12    In particular, we discussed the impacts of current 
13   transparency regime on liquidity and potential 
14   opportunities for further action on both pretrade and 
15   posttrade transparency.  Also the implications of the 
16   growth products such as open-ended mutual funds and ETFs, 
17   and finally the changes in intermediation model and the 
18   increasing electronification of markets. 
19             Following our January meeting, we formed three 
20   subcommittees, transparency, ETFs and bond funds, and 
21   technology and electronic trading.  This was to drill 
22   down on these topics in greater detail. 
23             The subcommittees have all held multiple calls 
24   to prioritize near-term and long-term initiatives for 
25   full committee consideration and have included non-FIMSAC 
0018
 1   member industry participants to help inform their work.  
 2   Minutes of these calls are available on the FIMSAC 
 3   webpage. 
 4             I want to thank all the members for devoting so 
 5   much of your time -- again, reiterating what others have 
 6   said -- but so much of your time on participating in 
 7   these calls.  I also want to thank the nonmember 
 8   participants that have joined these subcommittee calls 
 9   and that are participating in today's meeting for helping 
10   this committee advance its agenda.  I believe the active 
11   engagement and frank discussion that took place in these 
12   subcommittees will continue to enhance our ability to 
13   develop consensus around actionable recommendations to 
14   the Commission.   
15             I also want to thank Mihir, Ananth and 
16   Rick for their willingness to serve as subcommittee 
17   chairs. 
18             Over the course of the day, we will hear 
19   updates and presentations from each of the three 
20   subcommittees.  Each chair will have the opportunity to 
21   outline the progress to date of their respective 
22   subcommittee and moderate a panel discussion on a 
23   selected agenda item. 
24             As you can see, the panelists include market 
25   participants, staff from FINRA and the Commission, as 
0019
 1   well as several FIMSAC members.  Our panel sessions will 
 2   include opportunities for discussion among all of us 
 3   sitting at the table.  Members will be able to pose 
 4   questions to panelists or to each other, and to offer 
 5   their views on the subjects that we will be discussing. 
 6             For our first panel, the discussion will focus 
 7   around the consideration of a recommendation to conduct a 
 8   pilot program to evaluate the impact of current posttrade 
 9   dissemination regime on block-size trades in corporate 
10   bonds.  The panel will also include a discussion on 
11   whether such a pilot program should be recommended for 
12   municipal securities. 
13             Following a short break, Brett will moderate a 
14   panel that will discuss matters concerning liquidity 
15   considerations for fixed income ETFs, a topic that has 
16   been a subject of a fair amount of press coverage in 
17   recent years.  Panelists will address matters considering 
18   liquidity, particularly in those times of stress, and any 
19   potential impacts on the market for underlying bonds. 
20             After lunch, Ananth will provide us with an 
21   update from the ETF and bond fund subcommittee and 
22   moderate a panel addressing retail investor disclosure 
23   and education.  This panel will address current 
24   disclosure and sales practices, legal requirements around 
25   the sale and marketing of bond funds and ETFs, as well as 
0020
 1   a discussion of relevant studies and outreach efforts 
 2   that will provide insight into retail investor 
 3   understanding of these products. 
 4             Following a short afternoon break, Rick will 
 5   update us on the technology and electronic trading 
 6   subcommittee and will moderate our final panel discussing 
 7   several issues they are working on regarding electronic 
 8   trading and transparency in the retail corporate bond and 
 9   municipal securities market. 
10             During our discussions, in order to make this 
11   an opportunity for everybody's voice to be heard, I would 
12   ask that each committee member please turn their name on 
13   its side as an indication to me that you'd like to speak. 
14    This would really ensure that we have a robust and open 
15   and a fair dialogue amongst all the FIMSAC members. 
16             We will now turn it over to our first 
17   presentation from the transparency subcommittee.  Mihir, 
18   who serves as the subcommittee chair, will present a 
19   recommendation for a pilot program, as I mentioned, to 
20   assess the impact on market quality of a delayed public 
21   dissemination of trade reports for block trades in 
22   corporate bonds.  As Mihir will describe in more detail, 
23   FINRA rules govern the dissemination of trades reported 
24   to TRACE.  Pursuing this pilot program would necessitate 
25   FINRA amending its rules for disseminating trades 
0021
 1   reported to TRACE. 
 2             As I am sure many of you know, FINRA, as a 
 3   self-regulatory organization, must file changes to its 
 4   rules with the SEC for approval.  Part of that approval 
 5   process involves soliciting public comments on the 
 6   proposed changes to FINRA rules.  So any changes to the 
 7   FINRA rules that would facilitate conducting the pilot 
 8   that the subcommittee is recommending would be subject to 
 9   public comment and SEC approval. 
10             With that, I will pass it to Mihir. 
11             MR. WORAH:  Thank you, Michael, and thanks to 
12   the commissioners for taking the time to get our input on 
13   all of these topics in a changing fixed income market. 
14             As we saw the transparency subcommittee, our 
15   main goal as we saw it was to consider pretrade and 
16   posttrade reporting and its interaction with liquidity in 
17   the corporate and municipal bond markets.  As you've 
18   heard, we have a proposal today for modifying some of the 
19   TRACE dissemination rules that we think achieves the 
20   balance of increasing transparency while also potentially 
21   increasing liquidity in the corporate bond markets. 
22             The bulk of our meetings, the bulk of our 
23   discussion and debate was spent around this topic, and I 
24   will get back to it in a few minutes.  But let me just 
25   take a few minutes to update the Commission and the rest 
0022
 1   of the committee, the FIMSAC, on some of the other topics 
 2   that were considered by our subcommittee. 
 3             So the first thing we did was -- one of the 
 4   first things we did was divide -- consider topics to be 
 5   studied and divided them into longer term and shorter-
 6   term topics.  And we formed working groups around each of 
 7   these longer and shorter-term topics.  Some of the 
 8   longer-term issues to study were implementing changes in 
 9   -- implementing a pretrade reporting regime in the United 
10   States, as well as complementing some of the academic 
11   studies, some of the issues, you know, that were raised 
12   around academic studies, do they account for behavioral 
13   changes, do they properly account for the fact that we've 
14   had a long bull market in corporate bonds, and other such 
15   issues that have been raised.  So those are some of the 
16   long-term topics that we studied.  Not a lot in the three 
17   months that we've had, not a lot to report on the long-
18   term topics.  Like I said, the bulk of our time was spent 
19   debating the proposal that we will make to the Commission 
20   and the rest of the committee today. 
21             One of the short-term topics that we considered 
22   was around the municipal bond market.  While we had a 
23   sense, including from panel presentations at the first 
24   FIMSAC meeting, that most of the corporate bond market, 
25   most market participants on the buy side, sell side, 
0023
 1   including and issuers as well were open to some changes 
 2   in the trade reporting requirements in order to enhance 
 3   liquidity.  The situation or the sentiments in the 
 4   municipal bond market weren't clear to many of us on the 
 5   transparency subcommittee. 
 6             So we had -- so we invited noncommittee 
 7   members, both buy side and sell side, to present to us a 
 8   couple of times.  However, after listening to the seven 
 9   or eight municipal market participants, we still didn't 
10   feel there was a consensus on what the municipal bond 
11   market wanted or whether there was a need for a change to 
12   the transparency -- to the trade reporting requirements 
13   in the muni bond market.  So we asked the municipal bond 
14   working group in our subcommittee to reach out, offline, 
15   in addition to having them present to us, to reach out 
16   offline to municipal bond participants and canvass them 
17   for whether there was a consensus for even considering a 
18   change in the current reporting requirements.  And as we 
19   posted in the public document, the sense we got was that 
20   the municipal bond market at this point was comfortable 
21   with the way the municipal bond market was functioning 
22   regarding trade reporting requirements.  So at this 
23   point, we are not proposing any change to trade reporting 
24   requirements for the municipal bonds and focusing solely 
25   on corporate bonds. 
0024
 1             As I turn over to the corporate bond proposal 
 2   that we have, before I do that, let me ask John Bagley 
 3   from the MSRB, who was a member of our municipal bond 
 4   working group, to quickly report on their findings. 
 5             MR. BAGLEY:  Sure.  Thanks, Mihir. 
 6             So when we did speak to market participants, 
 7   both buy side and sell side in munis, really a couple of 
 8   market structure issues that are different from munis and 
 9   corporates were the reasons cited for not thinking a 
10   pilot was necessary.  And there being not nearly as many 
11   large blocks of munis as there are in corporates because 
12   of the nature of the way the muni deals come to market, 
13   and also the difficulty in hedging municipal bonds made 
14   people feel that dealers would not spend more of their 
15   capital in a down market, which was the only time people 
16   were worried about liquidity.  As a general rule, people 
17   thought liquidity was pretty good in munis and they 
18   thought the transparency rules worked for them the way 
19   they are now and did not think that changes were 
20   necessary. 
21             We'll be open to more points of view as people 
22   come to us and see where we are.  But at this point, the 
23   market certainly did not think that changes to the 
24   transparency rules were necessary. 
25             MR. WORAH:  Thanks, John. 
0025
 1             So let me now turn to the second short-term 
 2   topic that we discussed, and which resulted in the 
 3   proposal that we're bringing to the rest of the committee 
 4   today.   
 5             So there was a general consensus, we felt, 
 6   within the committee and also through outreach to other 
 7   market participants that maybe the corporate bond market 
 8   -- restraints in the corporate bond market, especially 
 9   for larger blocks, wasn't working as it should and was 
10   perhaps being hindered by some of the TRACE reporting 
11   requirements.  So we had a working group that came up 
12   with a proposal that was discussed and debated within the 
13   subcommittee that we feel -- that we're presenting to the 
14   rest of the committee today -- that we feel enhances -- 
15   strikes the right balance, as some of the commissioners 
16   talked about, between enhancing transparency as well as 
17   improving liquidity in the corporate bond markets. 
18             So before I go there, you know, a lot of us 
19   are, you know, are deep in the depths of these reporting 
20   requirements and the minutiae.  But let me just recap 
21   what the current reporting, TRACE reporting regime is, 
22   for those who don't, you know, think about these topics 
23   all the time. 
24             So broadly, the current trace reporting regime 
25   for corporate bonds is for every trade, every corporate 
0026
 1   bond trade to be reported to the FINRA within 15 minutes 
 2   of pricing.  So every trade gets reported to the FINRA 
 3   within 15 minutes of pricing.  We are not proposing any 
 4   change to that reporting requirement. 
 5             In terms of what information is disseminated to 
 6   the rest of the market, there's a cutoff defined, and 
 7   let's call it a block size.  Trades below the block size, 
 8   all information including the exact size of the trade is 
 9   disseminated to the rest of the market at this 15-minute 
10   mark.  Currently, the block size cutoffs are $5 million 
11   for investment-grade bonds, and $1 for below investment 
12   grade.  So all trades below $5 million investment grade, 
13   and below 1 million for high-yield bonds, all details are 
14   reported right away to market participants.  If you're 
15   above this block cutoff size, then the only detail that's 
16   reported is that there was a trade done whether, you 
17   know, it's a customer buying or a dealer selling to a 
18   customer, a dealer buying from a customer, a dealer-to-
19   dealer trade, and the fact that a block trade went 
20   through with a size 5 million plus for the investment-
21   grade market and 1 million plus for the high-yield 
22   market, with no further information.  And then six months 
23   later, all of this information is released to the 
24   markets. 
25             So what we are proposing is the block size 
0027
 1   cutoff, two changes to the dissemination regime.  One is 
 2   that the block size cutoff for investment-grade bonds 
 3   moves from $5 million to $10 million.  What this achieves 
 4   is in the previous regime, on average, and this is all 
 5   data that we've posted that the FINRA has made available 
 6   to us and we have posted on the FIMSAC website, on 
 7   average, over the last five years, from 2013 to 2017, 
 8   96.6 percent of all trades were below $5 million.  And so 
 9   all information was reported.  Moving the cutoff from 5 
10   million to 10 million means now the unmatched trades 
11   where all information is reported, in terms of the number 
12   of trades, moves from 96.6 to 98.8 percent, so almost 99 
13   percent of trades, the information gets disseminated 
14   right away. 
15             In terms of volumes of trade, and I think Brett 
16   touched upon it in his introduction, in terms of volume, 
17   44 percent of trades were covered by this 5 million 
18   cutoff previously.  This 44 million would move to 67 
19   million.  So moving the block size cutoff from 5 million 
20   to 10 million, it means about 99 percent of all trades in 
21   terms of the number of trades, all information would be 
22   reported.  And 67 percent in terms of volume, all 
23   information would be reported. 
24             What we are proposing is trades above the block 
25   size do not get disseminated for 48 hours, allowing 
0028
 1   efficient risk transfer between buyers and sellers before 
 2   the rest of the market can react to this and use this 
 3   often, as we find as market participants, in in predatory 
 4   fashion.   
 5             So in investment grade market, move the block 
 6   cutoff from 5 million to 10 million.  And if it's above 
 7   10 million, do not disseminate this information to the 
 8   market.  While it still gets reported, it doesn't get 
 9   disseminated for 48 hours. 
10             In the high-yield market, we're proposing 
11   moving the cutoff for block trades from 1 million per 
12   trade to 3 million.  And again, looking at the last five 
13   years' statistics, what this accomplishes in terms of 
14   transparency, 81 percent of high yield or below-
15   investment-grade trades were below the 1 million cutoff. 
16    Moving the cutoff to 3 million means you cover 93 to 94 
17   percent of all trades, in terms of instant dissemination. 
18             In terms of volumes of trades, size, 15 percent 
19   were below the prior threshold of a million dollars.  
20   Moving the threshold to $3 million covers 43 percent of 
21   all trades in the below-investment-grade market.   
22             We think, and there's been significant outreach 
23   to market participants, both buy side and sell side, we 
24   think this change, what we think is a relatively modest 
25   change in the trades reporting rules, increases 
0029
 1   transparency on a number of trades while masking 
 2   dissemination for 48 hours and has the potential to 
 3   significantly increase liquidity for block-size trades in 
 4   the functioning of the corporate bond market. 
 5             So this is the proposal that we have, and that 
 6   we're going to discuss and debate.  I wanted to talk 
 7   about two decision points that we had before that we 
 8   debated within our subcommittee in coming up with this, 
 9   with the final proposal.  And both of these are outlined 
10   in the public document as well. 
11             The first one was around the design of this 
12   study.  The way we have it is TRACE reporting rules for 
13   all corporate bonds would be changed from 5 to 10 for 
14   investment grade and for 1 to 3 million for the high-
15   yield bond.  And over a proposed study period of a year, 
16   we would study the functioning and the impact of the 
17   corporate bond market, a number of factors that we can 
18   talk about in detail, to see whether liquidity in the 
19   functioning of the corporate bond market was improved or 
20   not by these changes. 
21             An alternative design for the study would be to 
22   not change the reporting requirements for all bonds, but 
23   to have a control group where the reporting requirements 
24   did not change and a pilot group where the reporting 
25   requirements did change, simultaneously.  So we would go 
0030
 1   forward for the next year, after all the public comments, 
 2   et cetera, where there would be a group of bonds where 
 3   the reporting requirements were like they were 
 4   historically, 5 million for IG, 1 million for high yield, 
 5   and a group of bonds where the reporting requirements 
 6   would change, 10 million for IG and 3 million for high 
 7   yield. 
 8             There are pros and cons to both approaches.  
 9   The pro to the approach where there is a control group 
10   and the pilot group is that you have a control 
11   environment side by side, at the same point in time, 
12   where you can compare what's happening to a group of 
13   bonds with certain rules and a group of bonds with a 
14   different set of reporting rules.  The con, and the 
15   majority of the subcommittee felt that it would be a con, 
16   is manyfold.  Primarily what it does is we feel -- we 
17   felt it created winners and losers, both in terms of 
18   issuers, in terms of bonds owned by asset managers, bonds 
19   owned by individuals and bonds owned on dealer balance 
20   sheets.  Certain bonds would be subject to a certain set 
21   of rules and another set of bonds to another set of 
22   rules. 
23             And we felt practically as market practitioners 
24   and even reaching out to market practitioners outside the 
25   subcommittee, that while in terms of an academic 
0031
 1   experiment this wasn't the ideal setup, by changing the 
 2   rules and having the proposal as it is for all bonds, we 
 3   could still tease out the information that we needed in 
 4   terms of did this proposal, did this study impact the 
 5   functioning of the corporate bond market in a significant 
 6   way, without creating winners and losers amongst issuers 
 7   and owners of bonds. 
 8             There's issues around do some names fall into 
 9   one group, into the control group versus the pilot group, 
10   which creates winners and losers?  And there's also the 
11   issue depending on how the control group and pilot group 
12   is designed, whether the same issuer, some bonds fall 
13   into the control group and some into the pilot group.  
14   And amongst market practitioners, the consensus was in 
15   such a situation, they felt quite sure that liquidity 
16   within the same issuer would flock to the bonds which 
17   were above the 10 million masking threshold and you'd 
18   have some orphan bonds which would not trade at all. 
19             So I will discuss the pros and cons, and there 
20   are certainly valid reasons, and I am going to give the 
21   floor to Kumar in a second to talk about why, you know, 
22   having -- designing the study as a control group and a 
23   pilot group may be a better way, the subcommittee debated 
24   and, while not unanimous, there was a broad consensus 
25   that the proposal we have in front of you was the right 
0032
 1   way to go. 
 2             So before I go to the details, before we move 
 3   to the panel to discuss the proposal and how we feel it 
 4   might help the corporate bond market, let me have Kumar 
 5   say a few words about the alternative proposal that we 
 6   did not go forward with. 
 7             MR. VENKATARAMAN:  Thank you, Mihir, for this 
 8   opportunity to share my point of view. 
 9             So as Mihir said, we have discussed two 
10   research design options.  The first one would be where 
11   the delayed reporting of block transactions for all the 
12   bonds, and the alternative is what is called a randomized 
13   control design, where we have treatment bonds which are 
14   subject to this transparency change, and then control 
15   bonds which are subject to no change. 
16             So I would like to motivate the pros and cons 
17   of this -- of these two research design options using an 
18   example of a clinical trial of a drug.  So suppose a drug 
19   company invents a new drug.  It claims that the drug can 
20   cure a disease.  And if the regulator is certain that the 
21   drug will cure the disease, then there is no need for a 
22   clinical trial or a pilot.  All patients should be 
23   administered the drug because otherwise it's unfair, it 
24   will create winners and losers.  Patients who don't 
25   receive the drug will be disadvantaged. 
0033
 1             But suppose a scenario is slightly different.  
 2   Suppose there is genuine uncertainty within the medical 
 3   advisory committee about whether the -- about the 
 4   efficacy of the drug.  Some people genuinely believe the 
 5   drug is effective, others genuinely believe it is harmful 
 6   or simply don't know.  So in that case, the question of 
 7   fairness or creating winners and losers is moot because 
 8   it is not clear whether the new drug helps or hurts the 
 9   patient.  In that case, the standard approach is to 
10   design a clinical trial to learn the truth.  So people 
11   participating in the trial are randomly allocated to 
12   either the group receiving the new drug or to a control 
13   group receiving standard treatment or a placebo 
14   treatment.  Then you examine the outcomes for patients.  
15   And so this is the gold standard in research design, the 
16   randomized control trial. 
17             So in the case of the pilot, I would argue that 
18   we are in the second scenario.  There is genuine 
19   uncertainty on whether the pilot helps or hurts 
20   liquidity.  Some people on the subcommittee believe that 
21   delayed block trade reporting will help block 
22   participants.  Other people believe it will hurt the 
23   nonblock participants, such as retail investors and 
24   smaller institutions.  These players rely on the price 
25   discovery which is facilitated by timely disclosure of 
0034
 1   all transactions.  And block transactions are pretty 
 2   clearly important because they convey valuable 
 3   information on the demand and supply imbalance in a 
 4   security.  And there is plenty of academic evidence 
 5   suggesting that timely disclosure of transaction 
 6   information in the corporate bond market and in the 
 7   municipal bond market has improved liquidity. 
 8             So if we agree that there is genuine 
 9   uncertainty on how this experiment is going to play out, 
10   then the objective is really to understand whether the 
11   pilot helps or hurts liquidity.  The point of creating 
12   winners and losers is moot because we don't know what the 
13   impact of this pilot is going to be. 
14             So if we agree on this, then the objective of 
15   the pilot is to learn the truth.  So in this regard, the 
16   current proposal where we introduce the delayed reporting 
17   for all bonds has an inherent weakness, which is the 
18   following. 
19             Let's say we go with the pilot and suppose one 
20   month into the pilot there is a stress event.  This could 
21   be a big change in interest rates, a geopolitical event, 
22   a big stress event.  And suppose everyone agrees that 
23   this stress event will impact liquidity.  Then under the 
24   current research design, it will be very difficult for us 
25   to attribute the change in liquidity to the pilot.  It 
0035
 1   will be very difficult to decompose the impact of the 
 2   change in liquidity into the component that is due to the 
 3   pilot versus the component that is due to the stress 
 4   event. 
 5             However, if you use the alternative research 
 6   design, which is the randomized control design, the 
 7   stress event, whatever that is, will affect both the 
 8   treatment and control bonds, where the pilot only affects 
 9   the treatment bonds.  So we would implement what is 
10   called a difference in difference.  That is, we would look at 
11   a change in liquidity for the control bonds and compare 
12   it to the change in liquidity of the treatment bonds.  
13   And the difference between the two would allow us to have 
14   clear implications for the impact of the pilot. 
15             So the pilot will yield useful data under the 
16   randomized clinical control design for a broad range of 
17   market conditions.  So those are the advantages of going 
18   with a randomized control design, which is something that 
19   I would recommend that we consider strongly. 
20             Thank you, Mihir. 
21             MR. WORAH:  Thanks, Kumar. 
22             So what -- you know, there was actually one 
23   more decision point that I wanted to address before I 
24   turn it over to the panel.  So one was the design of the 
25   study and the second one, again, outlined in the document 
0036
 1   is the size of the cutoff for the below-investment-grade 
 2   market.  As far as -- as far as our subcommittee, as well 
 3   as interaction with noncommittee members and market 
 4   participants for the investment-grade corporate bond 
 5   market, there was fairly clear consensus that moving from 
 6   5 million to 10 million would accomplish the two parallel 
 7   goals, A, increase transparency, including for, in 
 8   particular, for retail and small-sized investors 
 9   significantly while improving liquidity in the -- in the 
10   block-size market.  On the high-yield side, we didn't 
11   have as much consensus on where the size of the cutoff 
12   should be.  There were two alternatives considered.  
13   Should the cutoff be at 3 million, as we've proposed, or 
14   a higher cutoff on the high-yield bond market at 5 
15   million for below-investment-grade? 
16             There was essentially the -- the majority of 
17   the committee came down on the 3 million proposed cutoff 
18   that we have in our current proposal.  Primarily, 
19   thinking in terms of parallels to the investment-grade 
20   bond market.  We were increasing the size of the cutoff 
21   in the investment-grade bond market by a factor of two, 5 
22   million to 10 million.  Moving the high-yield market from 
23   1 million to 2 million didn't seem to accomplish much.  
24   Moving it to 5 million would be too far, even though that 
25   was a valid consideration, and the committee decided on 
0037
 1   moving the high-yield cutoff to 3 million per trade, 
 2   defining block trades that would not be disseminated for 
 3   48 hours versus trades where all the information would be 
 4   disseminated within 15 minutes. 
 5             So that's the bulk of the proposal.  A one-year 
 6   study period.  And obviously the Commission and FINRA, as 
 7   they go through public comment if they do choose to 
 8   implement this, part of the proposal is if you find that 
 9   it's having counterintuitive impacts on the corporate 
10   bond market, we could shut this down.  But a one-year 
11   study period where we would study if changing TRACE 
12   dissemination rules, 5 million up to 10 million for 
13   investment-grade bonds, 1 million up to 3 million for 
14   high-yield bonds, and where trades above this block size, 
15   information would not be disseminated for 48 hours.  On 
16   the other hand, we would move dissemination of all 
17   information from a period of six months down to a period 
18   of three months, study it for a year and see if it 
19   improves liquidity and functioning of the corporate bond 
20   market. 
21             That's our proposal.  I am going to turn it 
22   over to our panelists.  That's Brian Archer from Citi 
23   representing the sell side, Scott Krohn from Verizon 
24   representing issuers, and myself, representing the buy 
25   side, in terms of what we think this proposal does for 
0038
 1   the corporate bond market. 
 2             MR. ARCHER:  Good morning, everyone.  I think 
 3   it is important to think about the backdrop of this in 
 4   terms of trying to enhance liquidity for the marketplace 
 5   and, in doing so, consider what a typical trader on the 
 6   sell side faces every time he's asked to trade a block or 
 7   bid or offer a block on behalf of a client.  And 
 8   basically, the equation is we have a very asymmetric risk 
 9   profile when we bid or offer a block of securities to a 
10   client, in the sense that if we are right in our 
11   assessment of where the market is at that time -- and 
12   it's not always a very defined market -- we have the 
13   ability to make a small amount of money on that trade.  
14   If we are wrong in our assessment of the market, we have 
15   the potential to lose a significant amount of money.  And 
16   obviously, that equation gets larger, the larger the size 
17   of the block.  And that's why we're proposing to consider 
18   the change in the dissemination of the information around 
19   blocks while still reporting the blocks to the 
20   regulators. 
21             Now, that's all anecdotal.  So we decided to 
22   try and look at some of the data that FINRA kindly 
23   presented to the group.  And does everybody have the 
24   slides that I sent out? 
25             So if we turn to slide 2, just to analyze some 
0039
 1   of the data, what we're trying to show here on the top is 
 2   a 10-year history of the U.S. investment-grade market, 
 3   data by size of trade for 2007 and data by size of trade 
 4   for 2017.  The darker -- the brighter blue line is 2017, 
 5   the darker line is 2007. 
 6             And basically what it shows is transactions 
 7   migrated toward smaller-size transactions from 2007 to 
 8   2017.  And although it's not shown here, just to 
 9   summarize the data, trades under 5 million in 2007 
10   cumulatively were 26 percent of the market, in 2017 they 
11   were 43 percent of the market, so a significant move to 
12   smaller size trades in the market. 
13             If you look at the trades over 10 million in 
14   that period, in 2007 it was 49 percent of the market and 
15   in 2017 it's 34 percent of the market.  Again, so a move 
16   away from big trades to small trades in the marketplace. 
17             If you look at the same data for U.S. high 
18   yields, again, a migration of trades to less than 1 
19   million, which is the block size reporting for TRACE.  
20   Cumulatively, in 2007, under 1 million was 5.9 percent of 
21   the market, 2017 it's 16.1 percent of the market.  Again, 
22   a migration to smaller tickets in the marketplace. 
23             If we turn to the following page, page 3, this 
24   is a demonstration of what the change in dissemination of 
25   volumes would be in the marketplace.  So if you look at 
0040
 1   the top, from 2007 to 2017, the market is now -- or as of 
 2   2017 has been reporting 43 percent of volume in the 
 3   marketplace.  And these statistics were already suggested 
 4   by members here today.  Under a proposed cap to 10 
 5   million, we would move to 65 percent of the market.  So a 
 6   significant increase in transparency in the marketplace 
 7   in terms of transactions that are taking place. 
 8             If we flip to slide 4, it's the same kind of 
 9   view in terms of nominal for the U.S. high-yield market. 
10    Again, under the current regime, under 1 million, 16.1 
11   percent is actual transparency provided to the market.  
12   If we raise that cap to 3 million, it moves to 43.8 
13   percent of the market.  I believe -- and Mihir mentioned 
14   we discussed going to 5.  I believe that would raise that 
15   amount to somewhere around 60 percent.  Again, we -- both 
16   proposals would suggest increasing a significant amount 
17   of transparency in the marketplace. 
18             Slide 5 again shows data that was already 
19   mentioned.  This is ticket count for the U.S. investment-
20   grade market.  The top shows under the current 5 million 
21   cap or the current regime, 96.9 percent of tickets are 
22   fully transparent.  Under the proposal for 10 million, 
23   98.8 percent of transactions will be fully transparent. 
24             And the same data on page 6 for high yield 
25   would show a migration from 83.2 percent of full 
0041
 1   transparency to 94.2 percent if we move to a $3 million 
 2   cap. 
 3             And then interestingly, what we did was use 
 4   some of our own data on page 7.  This is all Citi 
 5   internal data.  And it's a continuous time series of 
 6   4,000 trades.  It shows how we are able to manage our 
 7   risk in the marketplace in a period of time.  And it's 
 8   not put here, but the data represents about 60 billion of 
 9   transactions.  And it shows on T plus zero, on average, 
10   we are able to recycle 30 percent of block trades.  And 
11   block trades here, we're using 10 million as a block 
12   trade.  And by T plus two on average, we're able to 
13   recycle about 50 percent of that risk.  So you can see 
14   the curve in shifting risk is pretty steep in the first 
15   two days.  And that's why I think it makes sense that 
16   we're considering the 48 delay in dissemination of 
17   information.   
18             And then over the following days on average, 
19   that slope is just a little flatter or that curve gets a 
20   little flatter.  And by T plus six, on average, we've 
21   moved about 66 percent of our trades.  So even after T 
22   plus six, and this puts you in the role of a market 
23   maker, that data set, we still own about 20 billion of 
24   the risk that we took on behalf of clients. 
25             So it's just a demonstration of why dealers are 
0042
 1   very cognizant of what they do in the marketplace and 
 2   it's not all obvious at all times where the liquidity for 
 3   us as market makers is in a marketplace.  So only 50 
 4   percent under the current regime are we able to recycle 
 5   on T plus two. 
 6             We believe this is a pretty big data set.  
 7   Again, it's only our data so other dealers may have a 
 8   different experience or a similar experience.  We did run 
 9   -- and we got to it a little too late -- we did run a 
10   similar exercise for our data in high yield.  And if we 
11   move to the $3 million cap, the data set will look pretty 
12   similar in the averages and the shape of the slope. 
13             I didn't specifically speak to any 
14   counterparties.  If I speak to specific counterparties in 
15   the marketplace, there are some legal considerations for 
16   me to do that.  But SIFMA has been having some calls over 
17   the last number of months to consider different proposals 
18   as well.  And I would say by and large, there's pretty 
19   good consensus from the sell side on those calls with 
20   SIFMA for a proposal similar to this.  Not exactly this. 
21    I don't think there was exact consensus on what the 
22   proposal should be.  But by and large, there was 
23   consensus that something of this nature would make sense 
24   to most market participants on the sell side. 
25             MR. KROHN:  I think someone mentioned earlier 
0043
 1   this topic is probably pretty well understood in this 
 2   room but good luck explaining it to your neighbors.  I 
 3   will try to make a few real-world connections at least to 
 4   the corporate world as to how this proposal would be 
 5   beneficial. 
 6             Verizon is supportive as both a large 
 7   investment-grade issuer of this proposal, as well as of a 
 8   pension plan sponsor, which I think is also very 
 9   important.  My favorite statistic that has come out of 
10   this exercise is that in the last 10 years, the 
11   investment-grade corporate market has doubled from 3 
12   trillion to 6 trillion.  But as the Federal Reserve has 
13   shown, primary dealer balance sheets have decreased by 90 
14   percent.  So from well over -- you know, almost 300 
15   billion in '07 to now somewhere in the range of about 30 
16   billion. 
17             We think this proposal is a step forward in 
18   transparency and potentially we hope in narrower bid/ask 
19   spreads on block trades, because it gives dealers the 
20   chance to not have their positions exposed, as Brian 
21   mentioned.  You know, we look forward to, you know, 
22   positive results from the study.  And we also think that 
23   the timing of transmission proposal is very appropriate 
24   because we've had about 10 years with the wind at our 
25   back in fixed income markets, so we really haven't yet 
0044
 1   had to test those really low dealer inventory balances as 
 2   to how secondary trading would work in the next Lehman or 
 3   next, you know, financial crisis whenever that does 
 4   occur. 
 5             We also support this from a pension and 
 6   fiduciary perspective.  Just as a reminder, Jay mentioned 
 7   linkages.  You know, the way that risk management is 
 8   analyzed for corporate pension plans is based off of 
 9   corporate investment-grade bonds.  So the discount rate 
10   for all those promise liabilities for your employees is 
11   discounted at an index of corporate bonds.  So by 
12   definition when you hear corporations talking about 
13   derisking their pension plans, that means they are buying 
14   more investment-grade corporate bonds.  And, as we think 
15   about 2018 on the heels of tax reform, where basically 
16   technology and health care are repatriating billions of 
17   dollars and really not having a need to issue at a time -
18   - and you buttress that with pension plans, which 
19   actually in the S&P 500 had a 5 percent improved funded 
20   status last year from 80 to 85 percent, and many of which 
21   contributed to their plans -- again because of tax reform 
22   -- if you contribute this in '17 or also in '18 by 
23   September 15, you can deduct at 38 percent, not at 21 
24   percent. 
25             So there is this very material improved funded 
0045
 1   status amongst pension plans, at a time when new issuance 
 2   from corporates is less.  Therefore, the ability to 
 3   source liquidity in the secondary market from block trade 
 4   sizes from managers like PIMCO, from Vanguard, et cetera, 
 5   the type of managers that Verizon hires and many others 
 6   in the S&P 500 hire, I think is very, again, appropriate, 
 7   given some of the other market structural considerations. 
 8             So those were a few comments that I wanted to 
 9   share.  And I also just wanted to reiterate to Mihir's 
10   point something that we like to say around Verizon 
11   finance is simple, understandable, executable.  And I 
12   think this approach of the pilot applying to all issues, 
13   all CUSIPs, all issuers is a much more manageable process 
14   than trying to carve the market up, whether it's Verizon 
15   versus AT&T or even if we make it eligible to all issuers 
16   but only certain CUSIPs.  You know, to implement it I 
17   think would create some complexity in the markets as 
18   well. 
19             MR. WORAH:  Thanks, Scott. 
20             So let me just wrap up this panel with the buy 
21   side perspective.  Some of it is, you know, my 
22   perspective representing PIMCO and some of it is -- I 
23   might share with you my interactions with other buy side 
24   participants. 
25             So the first thing is, as asset managers, you 
0046
 1   know, we -- again, not all, but the consensus among asset 
 2   managers is liquidity in block size is being impaired and 
 3   there should be some consideration, some movement toward 
 4   maybe trying to increase the liquidity, changing some of 
 5   these TRACE dissemination rules. 
 6             And again, to be clear, as an asset manager, we 
 7   don't own assets; we manage them for pension funds, for 
 8   individuals, for retirees.  So when we try to increase, 
 9   when we say we think this proposal increases liquidity, 
10   probably reduces transaction costs in the corporate bond 
11   market, it's looking for a helping of the end user, the 
12   folks who invest in our funds, not helping us but 
13   helping, you know, the individuals and institutional and 
14   individual investors who have invested in our fund. 
15             In terms of market consensus, so I have, as 
16   part of the committee and in my role as chair of the 
17   subcommittee, socialized this, you know, some version of 
18   this, with many large asset managers.  And again, not 
19   universal.  From the handful that I spoke to, there was 
20   certainly, I should highlight, one exception who wanted 
21   no change in the current reporting requirements.  But the 
22   majority of others supported a change.  And in fact, 
23   supported a change such as this in terms of even the 
24   design of the study.  And, as Scott said, all bonds 
25   versus a control and a pilot group. 
0047
 1             I have also spoken to, as Brian said, 
 2   representatives from the SIFMA, and they are supportive 
 3   of this proposal as well.  Again, not universal but the 
 4   consensus seems a proposal such as this should be 
 5   considered to address some of the concerns folks have in 
 6   terms of liquidity in the corporate bond market. 
 7             And finally, you know, Kumar raised some points 
 8   in terms of the design of the study, the valid points.  
 9   The consensus of the committee was it would, you know, 
10   complicate -- you know, create these winners and losers 
11   and also create complications in terms of 
12   implementations.  But when Kumar was giving his example 
13   about the clinical trials, I thought of another example 
14   to support the way the proposal is written.  Right now, 
15   you know, finance and economics works in the real world. 
16    If you could control everything, that's great.  But it's 
17   our job to look at how, you know, economies and markets 
18   are functioning and use statistical analysis to tease 
19   out, you know, inferences and conclusions.  We don't 
20   always have the luxury of a controlled environment.  And 
21   the example I thought of is the Fed raising interest 
22   rates for half of the United States and not raising it 
23   for the other half so they could figure out the exact 
24   impact of interest rate hikes.  Doesn't work that way.  
25   They do it for everyone and then we figure out whether it 
0048
 1   worked or not. 
 2             So I think while there is definite benefits to 
 3   having a control group versus the pilot group, we think, 
 4   you know, this serves the bulk of the purpose and we hope 
 5   that this proposal -- you know, we're open to discussion 
 6   and questions from the rest of the committee, but we 
 7   think this moves the ball forward in terms of trying to 
 8   enhance -- address the issues around liquidity and 
 9   transparency in the corporate bond market. 
10             So I will stop here and we will take questions 
11   and let's have a debate. 
12             CHAIRMAN HEANEY:  So we will open it up to the 
13   committee members and of course to the commissioners.  
14   And let me please start with Commissioner Piwowar. 
15             COMMISSIONER PIWOWAR:  Yes, just a quick 
16   question. 
17             First of all, it's clear you all put a lot of 
18   thought into this.  And it makes it clear we also picked 
19   a great group of people to be on this committee, so I'll 
20   take a little bit of credit for that. 
21             A lot of discussion between the test group 
22   versus control group, whether or not to have that, was 
23   there any discussion about differential treatment among 
24   the test group?  So for example, in Brian's PowerPoint 
25   presentation, he shows there is some support for going 
0049
 1   for 48 hours for the dissemination there, because that's 
 2   where the curve is the steepest.  Any discussion about 
 3   giving differential treatment some 24 hours versus 48 
 4   hours or doing along some other dimensions?  And again, 
 5   to Scott's point, there's a tradeoff here between 
 6   simplicity and complexity, but there's more stuff we can 
 7   possibly tease out from this sort of thing. 
 8             MR. WORAH:  Right, so early on the -- you know, 
 9   so there was a proposal where, you know, I've simplified 
10   it into two competing proposals, the pilot and control 
11   versus all bonds being subject to this change.  And 
12   again, you know, one way of thinking of it, in this 
13   version, everything prior to the change is the control 
14   group and you compare it to pilot group post change.  So 
15   I have simplified it to two broad versions but there were 
16   -- there were versions where, again, in terms of the 
17   pilot and control, where you change in terms of issuer 
18   size, you know, bonds with a big outstanding base might 
19   be subject, might not, 48 hours versus 72 hours, having a 
20   1 million, 5 million, 10 million -- three different 
21   cutoffs.  So there were, you know, considerations around 
22   all of those.  But we kind of coalesced around the simple 
23   approach. 
24             MR. ARCHER:  If I can just comment on that?  I 
25   think, you know, the first reaction from many 
0050
 1   participants would be they be longer, and I just don't 
 2   think that's appropriate.  I think 48 hours strikes a 
 3   nice balance in terms of 24 versus 48. 
 4             CHAIRMAN HEANEY:  Larry. 
 5             MR. HARRIS:  Michael, I have a procedural 
 6   observation and then some questions.  And I would like to 
 7   reserve the opportunity to opine afterwards. 
 8             So the procedural observation is that we have 
 9   two issues in front of us.  The first one is whether to 
10   adopt this proposal.  And the second one is, if we do 
11   adopt the proposal, how to design the pilot study. 
12             I would suggest that discussions about designs 
13   of the pilot study are premature.  If we don't adopt the 
14   proposal, then we've been wasting our time talking about 
15   the pilot study.  So I would suggest that we confine our 
16   topics to the question of whether it's even sensible to 
17   do this. 
18             So now, I have a couple questions.  And I don't 
19   want to opine until I hear the end of these questions.  
20   And I am more than happy to let others ask questions 
21   first before I opine. 
22             The first question is this.  What is the 
23   incremental benefit of the additional transparency that 
24   we hope to obtain by displaying these large trade sizes? 
25    Transparency by itself, as you have observed, is not 
0051
 1   necessarily a -- the policy goal.  The goal is more 
 2   liquidity.  So the question is, what do people learn 
 3   that's of value of going say from 5 million to 10 million 
 4   in displaying the size?  Isn't it sufficient to simply 
 5   know that there is a very -- a large trade took place and 
 6   this is the price?   
 7             And then the second question is, what is the 
 8   exact mechanism by which we think liquidity is going to 
 9   be enhanced by not displaying a price for 24 hours?  So 
10   we've heard that there are some concerns about protecting 
11   position exposure.  But if our concern is about 
12   protecting position exposure, we recognize that presently 
13   a very large trade is only reported up to 5 million.  If 
14   that's a problem, shouldn't we simply reduce the block 
15   cutoff size, perhaps down to 1 million, to provide 
16   additional position exposure? 
17             And then the third question which perhaps you 
18   guys are a little less qualified to opine on but I think 
19   I would still like your answer, perhaps Mihir best would 
20   be able to opine on this.  We know these prices are used 
21   by mutual funds to price portfolios and it is very 
22   important that portfolio prices for open-ended mutual 
23   funds be correctly stated so that redemptions and 
24   deposits can be done on a fair basis.  If we delay the 
25   reporting of prices and particularly of the largest 
0052
 1   positions, isn't that quite problematic? 
 2             So those are my three main questions.  And 
 3   after others have a chance to ask questions, I hope that 
 4   you will permit me to opine on these issues. 
 5             MR. WORAH:  So let me take a stab at -- you 
 6   know, I may have lost track of all three.  The first one 
 7   was how does the exact size --  
 8             MR. HARRIS:  The first one is, what's the 
 9   incremental benefit of knowing size?  So if we go from 5 
10   to 10 million, I mean, what benefit do people get knowing 
11   that the trade was bigger than 5? 
12             MR. WORAH:  The incremental benefit is it opens 
13   up a -- so firstly, the reason why the masking in my 
14   opinion hurts the market is knowing that a size above 5 
15   million has gone through, there's conjecture, is it a 5 
16   million trade, is it a 50 million trade?  If it's a 50 
17   million trade, I want to jump in front of it.  I don't 
18   care who's doing it, why.  But it's a big trade, 
19   someone's got a lot of risk, and I can jump in front of 
20   it and profit from it. 
21             And so this closes -- what we think our 
22   proposal does is close that conjecture loop in terms of 
23   predatory behavior in the markets.  Moving from 5 to 10 
24   just increases the amount of transparency so we want to 
25   reduce -- our approach was to reduce this predatory 
0053
 1   behavior, enable better risk transfer between buyers and 
 2   sellers without significantly hurting transparency.  And 
 3   we felt the market today had evolved to a spot where 
 4   moving from 5 to 10 and then masking for a period of time 
 5   would accomplish both goes. 
 6             MR. HARRIS:  Would moving from 5 to 1 
 7   accomplish the same goal? 
 8             MR. WORAH:  It would -- it would reduce 
 9   transparency in the markets. 
10             MR. HARRIS:  What benefit is there of that 
11   transparency?  How does the market benefit from knowing 
12   that there was more volume traded than just a million 
13   that was reported or 5 million reported? 
14             MR. WORAH:  Because the way we have it is you 
15   get full -- whether you go below 1 million, again, you 
16   know, what the proposal does, reducing down to 1 million, 
17   wouldn't you say it shrinks transparency in the market? 
18             MR. HARRIS:  Oh, there's no question that less 
19   information is disseminated.  But the question is, what 
20   is the value of that information that's being 
21   disseminated? 
22             CHAIRMAN HEANEY:  Larry, I'll take a shot at 
23   that and I'll speak from the FIMSAC perspective.  I think 
24   this first meeting in January talked all about increasing 
25   transparency.  I don't think it was ever the charge or 
0054
 1   the consensus or the conversations that we had, and I 
 2   will take that as summarizing, that we're trying to 
 3   create less transparency in the marketplace.  And I would 
 4   be surprised if that was --  
 5             MR. HARRIS:  But isn't that exactly what this 
 6   proposal is? 
 7             CHAIRMAN HEANEY:  No.  If I can finish?  Thank 
 8   you. 
 9             So I think the whole goal is to increase 
10   transparency.  There's two things going on with this 
11   proposal.  One is the block, so you are trying to 
12   increase more risk taking and the facilitation of bigger 
13   blocks to trade, and that's obviously the 5 million to 10 
14   million.  The second is, the amount of transparency that 
15   therefore gets increased, and they've gone through the 
16   numbers, from 5 million to 10 million size trades that 
17   ordinarily wouldn't be 15-minute dissemination. 
18             So there is no real argument about how much 
19   transparency in those amounts of trades that are now 
20   going to be subject to 48-hour reporting, increased price 
21   discovery, increased transparency, all which benefits 
22   institutional and retail investors.  So I don't think you 
23   make the argument by moving the block size up we're 
24   getting less transparency; we're getting monumentally 
25   more transparency for the biggest belly of the curve of 
0055
 1   trades that take place --  
 2             MR. HARRIS:  But, Michael, let me then ask you 
 3   the same question I've asked Mihir and the others.  Of 
 4   what benefit is the additional transparency?  You said 
 5   that we're trying to increase transparency and yet we 
 6   have a proposal here to reduce it for the larger trades. 
 7    So clearly, that's not -- the object here is not 
 8   transparency.  The object is to improve welfare for the 
 9   investing public.  So what is the benefit that the public 
10   obtains from knowing large trade sizes above, say, a 
11   million or 5 million? 
12             CHAIRMAN HEANEY:  The goal of this is increased 
13   liquidity. 
14             MR. HARRIS:  Of course. 
15             CHAIRMAN HEANEY:  And increased transparency, 
16   without question, increases liquidity. 
17             MR. HARRIS:  In which case, shouldn't we 
18   display all trade sizes? 
19             CHAIRMAN HEANEY:  Let me, if I can, we've got 
20   plenty of people that need to be able to jump in. 
21             MR. THEES:  If I could just give you the 
22   opinion that what we're -- the design of this -- first of 
23   all, I experienced 1996 through 2002 with the SEC and 
24   Arthur Leavitt and the original TRACE panel, on which I 
25   was a Bond Market Association participant for the sell 
0056
 1   side.  And so what you've accomplished in a couple 
 2   subcommittee meetings took six years and didn't go 
 3   anywhere nearly as dramatically forward as you have with 
 4   this proposal.  So first of all, I commend you.  Because 
 5   if you had asked me as a betting corporate bond 
 6   experienced trader where you'd come out, I wouldn't have 
 7   thought you wanted to go from 5 to 10, I thought you 
 8   would say status quo but just delay the larger size. 
 9             So the fact that you're willing to say that the 
10   right thing to do is to increase transparency for 5 to 10 
11   million is a statement of confidence in the market's 
12   performance and ability for liquidity be increased with 
13   that statement.   
14             Number two, to Larry's point, why would you do 
15   this?  It's clear, the opinion of the buy side most 
16   especially affecting this opinion and Richie Prager 
17   enunciated this in his January presentation, I think the 
18   quote was nobody was promised liquidity, you're not owed 
19   it.  And this objective here achieves an incentive for 
20   increased liquidity by market makers.  Those market 
21   makers at this stage in the game could be buy side 
22   players because plenty of buy side players make markets 
23   and make prices.  They could be sell side players.  But 
24   the significant change here is to increase the incentive 
25   and willingness of participants in the marketplace to put 
0057
 1   a price on a security above 10 million, knowing that 
 2   yesterday someone else would have known that price within 
 3   15 minutes and today I get 48 hours to reduce my 
 4   liquidity.  In Citi's perspective, I get to reduce my 
 5   risk by 50 percent.  That is a significant appetizer for 
 6   a participant in the marketplace to want to participate 
 7   in a risk-taking activity with a buy side or sell side 
 8   participant that will incentivize them to do that.   
 9             And we don't know if it will actually occur.  
10   But behaviorally, in my experience, that is exactly the 
11   end result that will occur and we will look forward to 
12   the pilot program achieving that. 
13             CHAIRMAN HEANEY:  Chairman Clayton. 
14             CHAIRMAN CLAYTON:  Just -- I may -- this is 
15   always dangerous.  I may show a lack of understanding.  
16   But I want to just follow up on this point and ask a 
17   question, Brian, about the chart that you provided and 
18   the 48 hours. 
19             Now, you take down a chunky bond, okay.  And 
20   what this chart shows me is that over T plus two you on 
21   average have worked 50 percent of that position into the 
22   market.  I expect that most of those trades themselves 
23   are reported. 
24             MR. ARCHER:  That's correct. 
25             CHAIRMAN CLAYTON:  So it's not as if there's 
0058
 1   this chunky bond that's taken down and the market doesn't 
 2   see the effect of that; the market actually sees then 
 3   your derisking your position by making sales into the 
 4   marketplace that I would expect -- I don't know -- that 
 5   would be below the $10 million cap? 
 6             MR. ARCHER:  Correct.  So if I give an example, 
 7   if we took down 20 million and we sold that into the 
 8   market in ten $2 million increments, there would be ten 2 
 9   million prints. 
10             CHAIRMAN CLAYTON:  And because you're in the 
11   business of making money, those would be around the price 
12   or just slightly above the price at which you bought 
13   them, or around the price. 
14             MR. ARCHER:  The reality is the dissemination 
15   of information is going to limit -- like our clients will 
16   hold us to a reasonable revenue on that transaction and 
17   if it's too much, they will never do business with us 
18   again.  Right?  So you will get the information just a 
19   little bit delayed to allow us to recycle risk.  And I 
20   think the objective is not to sustain a 50 percent risk 
21   recycling within T plus two, but to move that entire 
22   slope higher, then the more recycling in risk we can add, 
23   then we can provide more and more liquidity with it. 
24             CHAIRMAN CLAYTON:  But I'll use my lay way of 
25   saying it.  Which is, the market is not blind to the 
0059
 1   block trade that you made.  The market will see your 
 2   dissemination of that block trade? 
 3             MR. ARCHER:  That's correct. 
 4             CHAIRMAN HEANEY:  Rachel. 
 5             MS. WILSON:  Thank you.  I think that the 
 6   design of the study here is relevant because this isn't 
 7   in theory, right?  This is in practice.  And so I think 
 8   it is part of a relevant discussion for us to really 
 9   think that through.  And there are some eloquent comments 
10   about a control group.  But also as we think about 
11   transparency and liquidity, there's also pricing.  So, 
12   you know, back to the have and have nots, I can't think, 
13   as an issuer, the thought of which group am I in and 
14   what's that going to do with my pricing and what's that 
15   going to do the trade of my indentures.  I'm just curious 
16   some of the feedback you got around that, because I think 
17   Kumar did a very eloquent explanation of better 
18   understanding, you know, some of the shocks or what could 
19   happen having a control group.  So I think we do have 
20   good articulation there but curious to hear, you know, a 
21   little bit more on some of that parity and fairness.  
22   Thanks. 
23             MR. WORAH:  So let me take that.  So our 
24   understanding, in talking to Kumar, talking to other 
25   academics, talking to quantitative analysts, most states 
0060
 1   of the world you could tease out information on whether 
 2   this was helping, whether liquidity was going up, 
 3   comparing the prior regime to the proposed regime.   
 4             Obviously, not having the side by side control 
 5   groups, there are one of the two states of the world 
 6   including -- I think volatility, interest rate changes, 
 7   all of those can be accounted for there.  One example I 
 8   can think of is if there's a sudden regulatory change, 
 9   Dodd-Frank gets rolled back and dealer balance sheets are 
10   allowed to blow up again, then it's hard.  In a situation 
11   like that, it will be hard to disentangle increased 
12   liquidity because you could still -- you could still try 
13   to do it but it would be a little bit harder. 
14             So most -- most states of the world, changes in 
15   time, changes in interest rate regimes, changes in 
16   volatility, you could actually tease out these are the 
17   impacts, but not in a hundred percent of the changes in 
18   the world.  And then before, so that's that. 
19             And then just to answer Lawrence's last 
20   question on fund pricing, you know, all of us as fund 
21   managers have a fiduciary responsibility to price funds 
22   appropriately.  Today, there's lots of bonds that don't 
23   trade every day, that don't trade for weeks.  They're 
24   priced.  You know, there are pricing services that 
25   provide prices and if we disagree, it's our duty to 
0061
 1   provide a fair market price.  And again, as this 
 2   experiment goes on -- I don't expect it, but if we 
 3   thought that the pricing services are continuously 
 4   mispricing bonds because of this change in regime, 
 5   obviously we would raise it with the Commission and 
 6   others.  But I don't see that being a problem. 
 7             MR. HARRIS:  Mihir, would you agree that the 
 8   pricing is better when they have more information?  The 
 9   pricing services are able to do a better job with more 
10   information?  Has it improved since TRACE was adopted? 
11             MR. WORAH:  I have not studied pricing service 
12   pricing. 
13             MR. HARRIS:  Can you imagine if it was 
14   otherwise? 
15             MR. WORAH:  I can. 
16             CHAIRMAN HEANEY:  Let me just jump in again.  
17   And I will offer an answer to that, that pricing services 
18   probably did get better.  And therefore, in this 
19   proposal, as we take investment grade, for an example, 
20   from 5 to 10 million, we go from 43 percent to 65 
21   percent, so we are increasing that transparency on that 
22   pricing. 
23             MR. HARRIS:  That's volume transparency, not 
24   price transparency. 
25             MR. TABB:  Just a couple of things in the 
0062
 1   determination or clarification of the pilot.  For those 
 2   of you who haven't seen the equity side or the options 
 3   side of the world, these pilots tend to go on forever.  
 4   If it's a one-year pilot, it should sunset. 
 5             And the other issue is, if you don't have two 
 6   different tranches then can you really, you know, go 
 7   back?  A lot of these issues, people program and let's 
 8   just say you go from, you know, five to 10 and can you 
 9   actually roll it back to five? 
10             MR. WORAH:  So we think, and again, that this 
11   was debated within the subcommittee, the length of the 
12   program, the side by side, et cetera.  And we felt like 
13   one year -- you know, something like three months would 
14   be two short.  One year was sufficient.  And if it's 
15   adopted, we give the regulators and the Commission, the 
16   way it's designed and worded, the ability to either 
17   extend it or to roll it back. 
18             MR. REDFEARN:  Larry, as you know, there are 
19   various pilots that we have done in the past and those 
20   which we continue to consider, whether or not it's here 
21   or elsewhere.  And I think the thinking is that the idea 
22   is to make sure they are properly constructed and that 
23   they have clear deadlines and goals to do that.  So I 
24   think in this particular case, we're trying to think 
25   about it in a very distinct way so that we'll do it, 
0063
 1   we'll study it, we'll measure it and we'll figure out 
 2   what's next.  But it won't just linger on, you know, 
 3   eternally. 
 4             MR. TABB:  And kudos on the access fee pilot 
 5   which actually has a definitive roll-off date. 
 6             MR. GIRA:  And if I could just jump in for a 
 7   second on Larry's point? 
 8             This might be unique in that I think in terms 
 9   of the impact on the industry.  As the -- as the pilot 
10   was described, this won't impact the industry.  What 
11   FINRA would do is delay the dissemination of the reports. 
12    So there literally wouldn't be any impact on the 
13   industry. 
14             A query whether -- you know, going back to the 
15   pilot as a whole, it seems like we're moving potentially 
16   three different dials.  And I think one other debate that 
17   we also had was, you know, could these dials potentially 
18   counteract each other?  And so there was some discussion 
19   of, you know, do you structure the pilot maybe so some 
20   but not all of these dials are moved, so that you can see 
21   what maybe had a better impact on enhancing liquidity?  
22   Was it the enhanced size dissemination?  Was it the 24 
23   hours?  Was it the three months versus the six months?  
24   So that could make it more complicated but that was a 
25   debate that was going on as well. 
0064
 1             CHAIRMAN HEANEY:  Amar. 
 2             MR. KUCHINAD:  Two questions for the 
 3   subcommittee.  Were you able to look at data, TRACE data 
 4   from 144(a) transactions prior to 214 and post 2014?  
 5   Because it seems like a lot of what's in this pilot was 
 6   already inadvertently done with nondissemination of 
 7   144(a) trades and then FINRA's rule change disseminating 
 8   those trades?  And there, there was inadvertent control, 
 9   because issuers opted in and we didn't have to decide 
10   winners as losers; issuers opted into issuing 144(a) or 
11   not and there's a handful of issuers that had both 
12   outstanding. 
13             So that was my first question.  My second 
14   question, have you considered what anti-manipulation 
15   surveillance or regimes would be needed if we did 
16   implement this pilot?  Because, just off the top of my 
17   head, as a former trader, I can think of a couple things 
18   that I could do to help myself.  As a trader, if there's 
19   a 20 million trade to do, I do 19 million at one price 
20   and then print 1 million at another price, and only the 1 
21   million gets disseminated for the first 48 hours.  So, I 
22   mean, there are things that I think that we need to think 
23   about that can't be done right now, given the way TRACE 
24   is set up, that we would probably need to figure out for 
25   the pilot.  I don't know if you guys considered that? 
0065
 1             MR. WORAH:  So the answer to your first 
 2   question is, we did not look at the 144(a) data.  On the 
 3   second one, I believe -- so the way, you know, we want to 
 4   move forward is, if this pilot is adopted, clearly in any 
 5   regime there can be market manipulation and, you know, 
 6   those issues would have to be considered and designed 
 7   around.  But I don't think it's particularly changes the 
 8   ability relative to the current regime.  It's a new 
 9   proposal and we would have to think about, you know, 
10   during the comment period, et cetera, I am sure we can 
11   think about those issues.  But I don't think it's 
12   particularly better or worse. 
13             CHAIRMAN HEANEY:  Gilbert. 
14             MR. GARCIA:  I was on the committee and I have 
15   one comment and then one question for Kumar if I could.  
16   The statement I have is, I would certainly feel a lot 
17   better, and my expectation would be a lot different, if 
18   I'm trying to sell say a $5 million block, if I see a 
19   print for 5 million to 10 million than if I just saw a 
20   print for 1 million.  So, in other words, it gives me 
21   greater confidence on my expectations of what transaction 
22   price I should get.  So that's one reason why I favor 
23   increasing the limit.  I think that's a good thing. 
24             But Kumar, in the spirit of making sure I'm not 
25   missing anything, how would you propose that these test 
0066
 1   bonds are chosen?  And could issuers choose to get in or 
 2   get out?  Or could you just elaborate a little bit on how 
 3   you see that happening? 
 4             MR. VENKATARAMAN:  Sure.  Thanks, Gilbert. 
 5             Ideally, we would like to not have any self-
 6   selection in the data.  So I would defer to the SEC.  
 7   they have plenty of experience designing these types of 
 8   pilots.  Most recently they had the access fee pilot but 
 9   earlier they had the tick size pilot.  So they have 
10   expertise in allocating part of the sample to the 
11   treatment and part of it to the control. 
12             But ideally, you don't want to have self-
13   selection, because then you worry that the selection 
14   itself may result in differences in outcomes that may not 
15   clearly give you information on the pilot. 
16             And if I may just take a minute to respond to 
17   Mihir's research design example. 
18             So with respect to the Federal Reserve, it is 
19   true that they do not do a natural sort of control 
20   experiment where half the economy is subject to it.  But, 
21   as you know, people have been writing about the impact of 
22   monetary policy 50 years ago and they are still writing 
23   about it, and that's the reason why we don't have a clear 
24   understanding of how monetary policy might actually 
25   impact the economy in a very precise way. 
0067
 1             So if we wrote a pilot where we're not really 
 2   clear on the research design, we may be having this 
 3   conversation 50 years from now where we talk about 
 4   whether, you know, delayed reporting of transactions, 
 5   whether it improves liquidity or hurts liquidity. 
 6             The other observation I have is that if you 
 7   look at the distribution of trade sizes, 60 to 70 percent 
 8   of the trades in the corporate bond market are less than 
 9   $100,000.  So there are a large number of nonblock 
10   participants in this market and smaller stuns as well, 
11   and they rely on timely reporting of transactions.  And 
12   there's plenty of evidence that introducing transparency 
13   has improved market quality. 
14             So I appreciate the concerns of the large 
15   institutional investors regarding delayed reporting of 
16   trades.  But if we do that, we should probably make sure 
17   that what is being delayed is really block transactions. 
18             So for example, if you take the case of the 
19   high-yield, noninvestment-grade proposal, we are looking 
20   at delaying 5.8 percent of trades under the 3 million 
21   cutoff, which represents 56 percent of the volume for 48 
22   hours.  So my question to the committee is if they will 
23   feel comfortable delaying such a large proportion of the 
24   trading volume by 48 hours or whether we should consider 
25   thresholds where we feel comfortable that these are 
0068
 1   really representing block transactions. 
 2             CHAIRMAN HEANEY:  Kumar, can I just follow up 
 3   then and make sure I heard you correctly about the 
 4   transparency?  The increased transparency and the price 
 5   prints that would be generated either through high yield 
 6   or IG, and in this case I will talk about IG and then 
 7   follow up on high yield, you feel like it's obviously 
 8   additive to the retail holder of bonds? 
 9             MR. VENKATARAMAN:  Yes, it is.  So if you look 
10   at the transaction data itself, what may be surprising is 
11   even if you look at insurance company trades, and these 
12   are all insurance companies so they're all institutions, 
13   on the same day for the same bond traded by the same 
14   dealer on the same side with two institutions, you see a 
15   difference of 25, 30, 40 basis points in terms of the 
16   transaction prices.  So I think surge costs are very 
17   relevant in this market because it's highly 
18   decentralized.  And so to the extent that common 
19   information is provided to a lot of market participants, 
20   it really benefits the market. 
21             So on the one hand, we have the challenge of 
22   making sure that large transactions do get done.  But on 
23   the other hand, these trade prints are important and to 
24   the extent that we can, you know, keep the balance in 
25   mind, I think it will be helpful. 
0069
 1             CHAIRMAN HEANEY:  Mihir raised it earlier.  And 
 2   just for the group at large, the 3 million, 5 million 
 3   high yield was definitely one that came under pretty good 
 4   debate.  So I would say if there are people -- and think 
 5   about this as we go forward here, running close to 11:15 
 6   or 11:30, if the 3 million to 5 million in high yield was 
 7   something that was of importance to think about amending, 
 8   we can certainly amend that should we choose to. 
 9             MR. McVEY:  Thanks, Michael.  Just a couple of 
10   observations and a few questions for the panel as well.  
11   But first of all, we're big fans of transparency and I 
12   personally think that TRACE has been a great thing for 
13   U.S. corporate bond markets.  The level of participation 
14   in our market has never been greater.  There are new 
15   market-making models emerging everywhere.  And investors 
16   and dealers have both benefitted from the increase in 
17   price information that's available, the understanding on 
18   liquidity and the risk characteristics of our corporate 
19   bond market. 
20             On the back of all that, we've been able to 
21   issue record levels of corporate bonds each year for the 
22   last five or six years in a row.  So transparency, in my 
23   opinion, is working very well for our market. 
24             Having said that, I support this pilot because 
25   information leakage on large transactions in illiquid 
0070
 1   bonds is a real issue.  And I would concur with the 
 2   comments that we have heard from the panel.  There is 
 3   widespread support, in my opinion, from both dealers as 
 4   well as investors on this pilot program to see if we can 
 5   re-incent capital for market making and also for the 
 6   growing base of ALL-TO-ALL trading that's taking place in 
 7   the corporate bond market. 
 8             We heard from the panels in January that 
 9   investors have never had better access to data than they 
10   do today in the U.S. credit markets.  That's one of the 
11   things that's allowing them to become very active as the 
12   price provider in our markets.   
13             So I think this is an issue that's relevant to 
14   both investors and dealers and I think the pilot would 
15   help to accomplish a better understanding of whether this 
16   relatively short delay on a very, very small percentage 
17   of our corporate bond trades would in fact improve 
18   liquidity in the market. 
19             Which then turns me to a couple questions.  I 
20   personally would come out in favor of the 5 million 
21   threshold on high yield.  I don't think that information 
22   leakage concern is between 3 and 5 million in high yield. 
23    And I think you get the symmetry of benefits and the 
24   increase in transparency in high yield in the say way 
25   that it's been proposed for high grade.  So I prefer to 
0071
 1   see fewer trades subject to the delay.  And I think 
 2   moving high yield from 3 to 5 would accomplish that 
 3   without disincenting the capital providers on those 
 4   trades. 
 5             Then I have two other questions.  One is, it is 
 6   very clear that new-issue bonds are far more liquid in 
 7   the first three or four weeks of trading than bonds are 
 8   after that.  I wonder if the committee considered any 
 9   differences for the first month of trading in new issues, 
10   where there's a meaningful part of trace volume taking 
11   place in those newly issued bonds.  And in my opinion, 
12   the information leakage concern there is very different 
13   than it would be on seasoned bonds that trade 
14   infrequently and could that potentially be an area where 
15   perhaps it's a 24-hour delay instead of a 48-hour delay? 
16             And then one final comment is that I think it's 
17   especially important, and again, when new issues are 
18   active, to get an accurate read on the day's volume and 
19   turnover.  So while I support the delay in reporting the 
20   blocks, I would like to suggest that FINRA continue to 
21   report in the aggregate the day's full volume when it is 
22   traded, so the market does not lose that information on 
23   the volume and turnover that has taken place on the 
24   actual trading day, rather than seeing that reported two 
25   days later. 
0072
 1             CHAIRMAN HEANEY:  Rick, I can tell you that the 
 2   pilot -- and David and Tom, please correct me if I'm 
 3   misstepping here, the pilot talked about holding back the 
 4   idiosyncratic risk, the CUSIP-level data, for 48 hours.  
 5   Nothing that would change necessarily the total reported 
 6   volume for the day, which would -- could still come out 
 7   to that concern. 
 8             MR. WORAH:  And on the first question, new 
 9   issues are very liquid.  They trade.  Should there be a 
10   different rule for new issues for the first three days, 
11   five days?  I don't believe this was considered by the 
12   whole subcommittee, but it was considered by the working 
13   group that was designing the study and then presenting to 
14   the rest of the subcommittee.  And we just opted for 
15   simplicity.  That point is understood, but we just opted 
16   for simplicity. 
17             CHAIRMAN HEANEY:  Horace, please. 
18             MR. CARTER:  Thank you.  I know that we're 
19   over, so I will be very brief. 
20             I just wanted to address Dr. Harris's concerns 
21   with a couple points, and Brian touched on it very 
22   modestly earlier.  But in a perfectly transparent market, 
23   what we've seen is that sell side, a dealer, can make a 
24   certain spread and that spread is essentially fixed.  
25   Which is fine for most normalized markets.  But if 
0073
 1   markets become stressed and volatility increases so that 
 2   the implied risk of taking on these positions increases, 
 3   what you see is essentially an inflection point where 
 4   it's not that the -- that the dealers tend to back away a 
 5   little bit or they start to fade their bids; they quit 
 6   participating.  And that's the real underlying risk, I 
 7   think, in terms of liquidity. 
 8             So really this, this proposal, I think, is very 
 9   modest.  If I have one complaint about it, I would say 
10   that it favors larger dealers over smaller dealers who do 
11   more block trades.  But that's okay.  I do think it's an 
12   improvement. 
13             That's my comment.  Thank you. 
14             CHAIRMAN HEANEY:  Ananth. 
15             MR. MADHAVAN:  Yeah, one quick comment and then 
16   a question for Brian.  So the comment is really just to 
17   be super clear on how greater transparency actually 
18   reduces bid/offer spreads and increases liquidity, which 
19   is I think the used car analogy is often used.  If you're 
20   trying to sell your 2012 Toyota and if you don't know 
21   what cars like that go for, then you don't get a good 
22   price.  So it's very straightforward, I think the link 
23   between greater transparency and the liquidity in the 
24   markets. 
25             The question for Brian is really, Brian, you'd 
0074
 1   referred a couple of times or alluded to, you know, 
 2   speculative activity around some of the larger trades and 
 3   some of the possible manipulative activity.  I was 
 4   wondering if you could go into a little bit more detail 
 5   on that, because I think that's very important for this 
 6   48-hour consideration. 
 7             MR. ARCHER:  Actually, I don't think I made 
 8   that statement about speculative activity.  I think maybe 
 9   you did, Mihir.  So I think it boils down to a couple 
10   things.  
11             One, this asymmetry of what you can make on a 
12   trade versus the risk you're taking.  And I think that, 
13   to me, is the biggest thing.  It's not what other people 
14   in the market might do with that information.  It's just 
15   time to digest that risk in the marketplace. 
16             Certainly, certain participants in the 
17   marketplace may see a print go up and look to shoot 
18   against you as a dealer, knowing you have that risk and 
19   try and force your hand.  I don't think that's anything 
20   that's illegal or whatnot but it's participants using 
21   their capital trying to gain from what you just did. 
22             I don't know if you want to make --  
23             CHAIRMAN HEANEY:  Dan Allen. 
24             MR. ALLEN:  Look, I would say, from a couple of 
25   observations, I agree with the proposals.  I happen to 
0075
 1   think 5 million makes more sense in the world of high 
 2   yield as well.  Although I understand 3 million could 
 3   easily be introduced and accomplish many of the same 
 4   goals. 
 5             Also, you know, Scott brought up the size of 
 6   the marketplace that obviously has grown substantially.  
 7   And I think a lot of folks talk about liquidity as a bear 
 8   market phenomenon.  And it goes both ways.  When you're 
 9   trying to put money to work, trying to source the paper 
10   and the bonds, it's a challenge.  And when the tide goes 
11   out, I think it's going to be a bigger challenge.  And we 
12   have been in a functioning market for some time, as we 
13   all know, and the focus on the tide going out is going to 
14   be really important here at some point in the coming 
15   months or years. 
16             That said, TRACE is an important component of 
17   understanding and improving liquidity.  But obviously, 
18   we've highlighted some of the other issues, including 
19   balance sheet, Dodd-Frank, the costs and the ability to 
20   hedge using derivatives and accessing, you know, shorts 
21   to -- even in the cash market as well, there's active 
22   versus passive growth.  There's ETFs and it's the foreign 
23   buyer which has been a big percentage of the marketplace, 
24   which has changed the liquidity profile as well.  So 
25   these are all things to consider. 
0076
 1             That said, you know, talking about 5 million or 
 2   if we're talking about 10 million, I think it does 
 3   incentivize the dealer community to take risk again, 
 4   something that has not been there for some period of 
 5   time, and give the dealer community a bit of time to 
 6   digest that risk and distribute it as Brian said. 
 7             It's not a given fact that you're going to buy 
 8   20 million bonds.  As Brian, you know, mentioned four, in 
 9   over two days he's going to work out of it at a profit.  
10   You're taking market risk, you're taking credit risk but 
11   having a little bit extra time will incentivize that 
12   risk.  And on the way down and when the tide does shift, 
13   I do think that's going to be an important factor. 
14             CHAIRMAN HEANEY:  Larry. 
15             MR. HARRIS:  Thank you.  We are talking about 
16   two types of transparency here.  We're talking about 
17   price transparency and also quantity transparency.  The 
18   argument about not reporting the larger trades is about 
19   quantity transparency.  It's about concerns about 
20   position exposure, it's about concerns about leakage of 
21   information about quantity.  
22             We know that the price information is 
23   incredibly important.  Careful studies have shown that 
24   American and other investors in our markets have saved a 
25   billion dollars a year from the reporting of the TRACE 
0077
 1   data.  That billion dollars has benefitted not only the 
 2   investors but also the issuers, because issuers can issue 
 3   securities at higher prices when people know that they 
 4   are going to trade in liquid markets.  And transaction 
 5   cost is your ultimate measure of liquidity.  So I think 
 6   what we're really talking about here is the importance of 
 7   price transparency.  I don't think that the quantity 
 8   transparency is really all that important.   
 9             So the proposal is that we're going to increase 
10   quantity transparency for smaller large trades -- so 
11   they're definitely large trades, but they're the smaller 
12   large trades -- in exchange for a complete lack of price 
13   transparency until two days later.  While, at the same 
14   time, we're told that we're going to see total quantity 
15   transparency for the day but we won't be seeing the 
16   prices. 
17             So it's very hard for me to listen to this 
18   debate without thinking that really what people are 
19   asking for is that the dealers don't want their customers 
20   to see the prices at which they acquired a block.  So 
21   when I suggested that if they're worried about front 
22   running or quantity and stuff like that, let's just not 
23   report quantity all the way down to, say, 1 million.  And 
24   while that seems to be a reduction of quantity 
25   transparency, it's a means of serving the concerns that 
0078
 1   were raised by most people at the table here. 
 2             So I would like to close with just a quick 
 3   observation about asymmetry.  So Brian spoke about the 
 4   asymmetry that dealers face.  But Ananth and -- at other 
 5   occasions, I'm sure people understand the asymmetry that 
 6   the customer faces.  So the customer -- the customer 
 7   really -- an uninformed customer can only make one type 
 8   of mistake.  If they overbid for something or if they see 
 9   an offer that is too high and they take it, the dealer 
10   isn't going to complain.  But if the customer doesn't 
11   step up as much, the trade just doesn't get done.  So the 
12   asymmetry there is the customer easily can overpay but 
13   rarely is the customer going to walk away with a deal. 
14             The whole purpose of price transparency is to 
15   level the playing field to some extent.  The empirical 
16   evidence says that it's helped tremendously and I would 
17   be reluctant to see us break that down. 
18             One final quick point.  By making a margin -- 
19   drawing a line at say $5 million or $10 million, we're 
20   going to change the incentives to do large trades versus 
21   small trades.  And some of the trades that are currently 
22   taking place at $5- to $10 million are going to end up 
23   going over 10 and we are going to lose even more 
24   transparency than we imagine.  Which I think would be 
25   quite problematic. 
0079
 1             CHAIRMAN HEANEY:  Thank you, Larry. 
 2             MR. ARCHER:  Michael, can I just comment on 
 3   that last -- Larry, just to address your last comment, I 
 4   do think, contrary to what you're saying, there's 
 5   significant investment being made by dealers to service 
 6   small trades much more significantly, and there's a lot 
 7   of work being done to do that via algorithm today, to 
 8   service that small set of trading.  And that, in fact, of 
 9   course explains the empirical results that you presented 
10   earlier. 
11             The shift to smaller trades reflects the fact 
12   that larger trades are being broken down, not only for 
13   retail but also for institutional.  So frankly, the 
14   market structure has changed.  The traditional dealers 
15   have become less important and other dealers operating 
16   out of hedge funds and proprietary trading firms have 
17   stepped in. 
18             A lot of that problem has nothing to do with 
19   the SEC, it has to do with the regulation of capital 
20   requirements that takes place completely away from this 
21   forum. 
22             MR. HARRIS:  Yeah, I would argue that the work 
23   being done on that small trading has been very recent and 
24   has not been something that has been in place for a very 
25   long time, at least from our perspective. 
0080
 1             CHAIRMAN HEANEY:  So can I ask others who 
 2   haven't raised issues or questions and would like to 
 3   before we move forward? 
 4             (No response.) 
 5             CHAIRMAN HEANEY:  Mihir, as subcommittee chair, 
 6   can I ask your view on pivoting here, given what we heard 
 7   from certainly a handful and I know there was obviously a 
 8   debate on the subcommittee itself, whether you believe we 
 9   should pivot to a 5 million -- to have the recommendation 
10   on the pilot changed to 5 million on high yield, keeping 
11   10 million for IG? 
12             MR. WORAH:  Actually, given what we heard and 
13   given the data that Brian presented, while the 
14   subcommittee put forward a 3 million proposal, I would 
15   actually support moving that, the high yield cutoff, to 5 
16   million. 
17             CHAIRMAN HEANEY:  Any issues by FIMSAC members 
18   if we were to pivot there? 
19             (No response.) 
20             CHAIRMAN HEANEY:  Okay.  Look, I think we've 
21   heard a lot of thoughts on this topic.  I think we heard 
22   a lot of data on this topic and there was certainly a 
23   back and forth as there was at the subcommittee level.  I 
24   think the transparency, I will call it in the belly of 
25   the curve, the price transparency, but in these 5 to 10 
0081
 1   and now I'll say this 1 to 5 million high yield is worth 
 2   considering.  I think we've gotten enough feedback both 
 3   in the subcommittee and around the table to take this to 
 4   a vote.  And at this point, I will ask a motion to vote 
 5   on the recommendation. 
 6             MR. THEES:  I'll move the motion. 
 7             PARTICIPANT:  I'll second. 
 8             CHAIRMAN HEANEY:  Thank you.  At this point, 
 9   the easiest way to do this will be as a show of hands and 
10   we will record what that is.  So I will ask all voting 
11   members in favor to raise their hands. 
12             (A show of hands.) 
13             CHAIRMAN HEANEY:  Fifteen. 
14             All opposed, please raise your hand. 
15             (A show of hands.) 
16             CHAIRMAN HEANEY:  And that's four. 
17             Anybody abstaining? 
18             (No response.) 
19             CHAIRMAN HEANEY:  Okay.  I thank everybody for 
20   their, again, spirited debate, in subcommittee and in 
21   total, and the panelists, I appreciate your time on this. 
22    This is one that has taken a lot of work and a lot of 
23   reasonably really good teamwork to get where we are 
24   today.  And from where we were at the subcommittee 
25   -- first subcommittee call to where we ended up today, I 
0082
 1   will tell you, was a lot of good debate, good discussion 
 2   and a lot of hard work. 
 3             CHAIRMAN CLAYTON:  May I?  Look, okay, I've 
 4   been doing this a year now.  And I want to say that in 
 5   the space of an hour and a half, there was an incredible 
 6   amount of substance here, which I wish occurred more 
 7   often.  And, no, I want to thank all of you for being 
 8   willing to share your views and experience with reference 
 9   to data, with reference to risks in the marketplace, 
10   changes in the marketplace, the points about how much our 
11   fixed income markets have changed in the last 10 years, 
12   the fact that we're at a shift in government policy.  The 
13   fact that when TRACE came in, we recognized that there 
14   would be a balancing of liquidity and transparency.  We 
15   already have a block and nonblock component and the 
16   question of where does that calibration go from here is 
17   an important question.  And we have the right people 
18   around the table. 
19             And Amar, thank you for reminding us what we 
20   should always be reminded of which is, are we creating an 
21   opportunity for manipulation or insider trading?  You 
22   know, we should have that at the fore of our minds.  So I 
23   thank you for that. 
24             Let me just say it this way.  There was not a 
25   point made today that wasn't valuable to consider as we 
0083
 1   move forward.  And I really appreciate that.  So thank 
 2   you all. 
 3             MR. REDFEARN:  Yeah, I also just wanted to echo 
 4   the Chairman's comments as well.  A fantastic debate.  It 
 5   is very good to hear this kind of discussion and thank 
 6   you very much for bringing data and bringing all of your 
 7   insight to this discussion. 
 8             I just want to close with just a couple of 
 9   process points.  So first of all, so people understand 
10   exactly what's happening here, the way this process will 
11   go is so this is a recommendation, this comes to the 
12   Commission.  At this point in time, the Commission itself 
13   will consider this recommendation.  Ultimately, it would 
14   be FINRA that would write a rule proposal that would take 
15   this forward.  And then that would have a comment period. 
16             So this is by far there is still some ground to 
17   cover between here and this going into effect.  
18   Ultimately, at some point, if we were to put this rule 
19   out and the comment period would be had, there would be 
20   plenty of opportunities to continue to get additional 
21   perspectives out there.  And I am certain that on a topic 
22   like this, we would hear plenty of it.  So all of that 
23   would have to take place before this is ultimately 
24   crossing any line. 
25             So again, let me just thank you all again for 
0084
 1   this conversation and I think with that, we're ready to 
 2   move on. 
 3             CHAIRMAN HEANEY:  We are at break time.  In 
 4   fact, we are into our break. 
 5             So for the benefit of the outside speakers who 
 6   have come to join us at 11:30, if we can just hold this 
 7   break to 10 minutes and come back at 11:35, please? 
 8             (Recess.) 
 9             CHAIRMAN HEANEY:  Thank you.  I will now turn 
10   it to our next panel, which will be discussing liquidity 
11   considerations regarding fixed income ETFs.  And I will 
12   turn it over to Brett, who will moderate. 
13             MR. REDFEARN:  Thank you very much, Michael.  
14   And again, thank you all for being here for this 
15   discussion. 
16             Today, an issue that was referenced earlier 
17   about what has often been referred to as a liquidity 
18   mismatch between bond ETFs and the underlying assets is 
19   part of the topic of our discussion.  There has been a 
20   lot written over the past several years, academic 
21   studies, blog posts, media about this so-called liquidity 
22   mismatch between relatively liquid bond ETF shares and 
23   the less liquid bonds that they hold.  This was brought 
24   to light recently in the Wall Street Journal article that 
25   had the headline, Could ETFs Fall into a Liquidity Jam, 
0085
 1   one of the latest stories to raise questions about the 
 2   prospect of this liquidity mismatch and its potential 
 3   impact.  And it specifically mentions fixed income ETFs. 
 4             So the ETFs and bond funds subcommittee, led by 
 5   Ananth Madhavan, has begun to consider this question as 
 6   one of its discussion topics, although that discussion is 
 7   still relatively in the early stages.  Since the topic 
 8   and the question continues to draw continuing attention 
 9   from market observers, we thought it would be helpful to 
10   begin to explore this today at this meeting. 
11             Just as some context, there has been 
12   significant growth in bond ETFs in recent years.  As of 
13   February 2018, there were approximately 164 corporate and 
14   municipal bond ETFs with approximately 215 billion in 
15   aggregate assets.  These 164 ETFs however currently 
16   represent less than 2 percent of the total corporate and 
17   municipal bond issuance currently outstanding. 
18             We are very fortunate today to have a 
19   distinguished panel of market experts to share their 
20   insights on this topic.  We have included perspectives 
21   from two buy side credit portfolio managers, a sell side 
22   credit product specialist, an ETF authorized participant 
23   and market maker, and the chief economist for an industry 
24   association for regulated funds, which includes mutual 
25   funds, ETFs, closed-end funds and UITs.  I will introduce 
0086
 1   them specifically in just a second. 
 2             The topics that we intend to discuss include an 
 3   exploration of the role of bond ETFs in the market and 
 4   how they are used by investors, the so-called liquidity 
 5   mismatch between ETF shares and underlying assets, 
 6   liquidity considerations in normal times versus times of 
 7   stress, the market impact of potential herding behavior 
 8   by investors, and the potential for price dislocations or 
 9   anomalies between the ETF shares and underlying asset 
10   markets in times of stress. 
11             So with that, let me introduce our panelists.  
12   First, we have Joe Arcadi, head of North American credit 
13   index trading at JP Morgan Chase; Matt Berger, global 
14   head of fixed income and commodities at Jane Street; Sean 
15   Collins, chief economist from the Investment Company 
16   Institute; Krishna Memani, the chief investment officer, 
17   head of fixed income, at Oppenheimer Funds; as well as 
18   Gregory Peters, managing director, senior portfolio 
19   manager at PGIM Fixed Income. 
20             So before we get into the panel discussion, we 
21   thought it might be a useful level-setting exercise for 
22   the committee to get a brief primer on bond ETFs and how 
23   they compare to other fund products like open-ended 
24   mutual funds.  And with that, Ananth, would you mind 
25   briefly providing some background on the structure of 
0087
 1   bond ETFs?  And, in particular, describe the dynamics of 
 2   bond ETF share liquidity and liquidity in the underlying 
 3   cash bond market. 
 4             MR. MADHAVAN:  Thanks, Brett.  Yeah, at the 
 5   most basic level, bond ETFs are portfolios or baskets of 
 6   bonds.  These funds share elements of both open and 
 7   closed-end funds.  Like closed-end funds, ETF shares can 
 8   be bought or sold intraday on organized, lit exchanges at 
 9   market-determined prices, so like closed-end funds.  
10   Similar to open-ended funds, the shares of ETFs are in 
11   open supply, in the sense that the ETF shares can be 
12   created or redeemed by institutional trading firms called 
13   authorized participants or APs.  That allows the shares 
14   of the ETF to adjust to supply and demand.  That process 
15   is known as the arbitrage mechanism and that keeps the 
16   ETF's price in line with the value of the underlying bond 
17   portfolio. 
18             Bond ETFs typically offer access to diversified 
19   fixed income exposure at low costs with high degrees of 
20   pre- and posttrade transparency because they trade on 
21   organized exchanges.  There is another dimension of 
22   transparency which is worth mentioning, which is that the 
23   holdings of bond ETFs are also available on a daily 
24   basis. 
25             Turning to Brett's question about the dynamics 
0088
 1   of liquidity, unlike traditional open-ended mutual funds, 
 2   shares of ETFs are primarily redeemed in kind.  What that 
 3   means is that the assets do not need to be converted to 
 4   cash by the fund to make redemptions, and that's a very 
 5   important distinction. 
 6             Investors in the bond fund own shares of the 
 7   ETF, similar to how single stock investors own shares of 
 8   a company's outstanding stock.  Those shares trade on the 
 9   exchange just like a company stock does, and changes in 
10   ownership are really what is manifested by ETF secondary 
11   market trading.  Okay, the trading of ETF shares does not 
12   necessarily result in flows in or out of the ETFs.  So 
13   instead, ETF flows occur through the arbitrage mechanism 
14   I had alluded to before, through the creation or 
15   redemption by the authorized participants. 
16             And finally, just in closing, a comment about 
17   the liquidity of these ETFs.  Many observers have noted 
18   that the bid/offer spreads of ETFs can be much lower than 
19   the bid/offer spreads of their constituent bonds, and 
20   that's true for strong economic reasons.  We have heard 
21   from the previous panel about the inventory 
22   considerations at the single bond level, considerations 
23   about asymmetric information.  All these are mitigated 
24   when trading bond portfolios.  So I'll just stop there. 
25             MR. REDFEARN:  Thank you.  So to better 
0089
 1   acquaint the various folks here with your individual 
 2   perspectives on bond ETFs, I think it would be great if 
 3   we could start by hearing from each panelist about your 
 4   direct experience with these products, including the role 
 5   that you play in the bond ETF market and how you use 
 6   these products.  So maybe Joe, start with you? 
 7             MR. ARCADI:  Thank you, Brett.  So my name is 
 8   Joe Arcadi.  I run the North American credit index 
 9   trading desk for JP Morgan.  I will also refer to this 
10   desk as the macro credit trading desk.  Essentially, 
11   we'll trade any sort of credit access product off of this 
12   desk.  It's going to be CDX indices, iBoxx total return 
13   swaps, credit ETFs and options across all these products. 
14             So historically, ETFs have remained in the 
15   realm of the equity world and equity participants.  
16   However, we saw a shift in demand from our institutional 
17   clients several years ago.  We brought our block trading 
18   of these instruments up to the credit floor of JP Morgan. 
19    So with this, our desk has also participation in the 
20   creation/redemption process, attempting to bridge the gap 
21   between the prices of the underlying bonds versus the 
22   prices of the shares.  However, this year, we have 
23   transitioned this closer to our existing bond market 
24   making franchise.  So we think it's going to be better 
25   utilized within our firm versus existing inventory, and 
0090
 1   also we may get into this later, part of facilitating 
 2   portfolio style transactions. 
 3             So within the credit markets, there has been a 
 4   genuine need for instruments that could accurately track 
 5   the performance of the corporate bond market.  So this 
 6   has led to the evolution of the iBoxx TRS product set and 
 7   credit ETFs as active traded instruments.  So these 
 8   products have been adopted by asset managers to 
 9   effectively manage their market exposure and diversify 
10   sources of liquidity in their portfolio.  And because our 
11   desk is uniquely suited to trade across the spectrum of 
12   these products, we're well positioned to advise clients 
13   on suitability for various products for their needs, and 
14   we're also set up so that we can offer liquidity for 
15   clients in a specific product while accessing all of the 
16   access products that we need to be able to hedge 
17   ourselves and provide the liquidity that the client 
18   needs.  But that's my role in this whole ecosystem. 
19             MR. REDFEARN:  Great.  Thanks, Joe. 
20             Matt. 
21             MR. BERGER:  Thanks.  My name is Matt Berger.  
22   I'm at Jane Street.  The firm, for those of you who don't 
23   know, we're a proprietary trading firm.  We have about 
24   700 people in New York, London and Hong Kong, and we have 
25   really grown up with ETFs.  We started trading in 2000.  
0091
 1   This past year, we traded $5 trillion of ETFs.  It's a 
 2   huge part of our business.  Thanks a lot for having us.  
 3   We really care deeply about these issues. 
 4             I joined the firm in 2001 and I have been a 
 5   market maker in ETFs in many different asset classes, 
 6   also in New York and in Europe.  We are currently lead 
 7   market maker on exchanges in 120 fixed income ETFs and we 
 8   do the majority of our trading on exchange.  So we're 
 9   usually facing anonymous counterparties.  But in 2014, we 
10   started an institutional client-facing broker-dealer and 
11   that business has been kind of a growing part of our 
12   business.  And it's interesting, in fixed income ETFs, 
13   kind of bigger ticket sizes have become more common over 
14   time.  So it's not unusual at this point to see 100 
15   million, 200 million at a clip going through in fixed 
16   income ETFs. 
17             I think our most important role in the ETF 
18   ecosystem, though we are a lead market maker on 
19   exchanges, I don't think the value that we provide being 
20   at the top of the book is the most important thing.  Like 
21   spreads are really tight, they're often a penny wide.  
22   But we think where we provide the most liquidity is when 
23   the order flows is very much one sided and we're willing 
24   to take risk and commit our balance sheet to buying -- 
25   you know, buying when we are not going to be able to sell 
0092
 1   that day or selling when we're not going to be able to 
 2   buy that day.  So we're, you know, creating and redeeming 
 3   very frequently.  We are usually using an agent to affect 
 4   that creation or redemption but we are willing to take a 
 5   lot of risk kind of at the end of the day to provide that 
 6   liquidity. 
 7             MR. REDFEARN:  Great.  Thank you, Matt. 
 8             Sean. 
 9             MR. COLLINS:  So I am Sean Collins.  I am the 
10   chief economist at the Investment Company Institute, 
11   which is the trade association for mutual funds and ETFs 
12   in the U.S.  We also have a global presence. 
13             So I thank the committee for inviting me here 
14   today and I think the objective of trying to inform 
15   FIMSAC about, you know, what ETFs are, how they operate 
16   and who uses them is a good objective.  To that end, we 
17   track assets and flows at ICI of ETFs, mutual funds, 
18   including bond ETFs, including bond mutual funds.  And 
19   we, along with ICI research staff and ICI member firms, 
20   can be a resource for regulators and others to try to 
21   answer questions about how ETFs operate, their assets and 
22   other issues.  And so just to that point, I want to talk 
23   a little bit if I could about liquidity mismatch, since 
24   it seems to be the subject of the meeting. 
25             So there has been a lot of discussion in the 
0093
 1   past year or two about liquidity mismatch in ETFs.  And 
 2   personally, I don't know of any evidence that's 
 3   convincing that liquidity mismatch is a problem.  But if 
 4   you think it is, I think there are two really important 
 5   things to keep in mind about ETFs.  One is, if there is 
 6   any liquidity mismatch issue, they solve the problem in 
 7   two ways.  One is, most of the trading of ETF shares 
 8   occurs on the secondary market.  There is no liquidity 
 9   mismatch there.  They trade the same as a stock. 
10             The second way they solve the problem is by 
11   doing in-kind transfers.  So you transfer shares of an 
12   ETF from an AP to the fund.  That transfer usually takes 
13   place in kind.  So again, there, there's no liquidity 
14   mismatch.  You're handing in shares of the ETF, you're 
15   getting back bonds of the underlying security. 
16             Thank you. 
17             MR. MEMANI:  Good afternoon.  My name is 
18   Krishna Memani.  I am the chief investment officer of 
19   Oppenheimer Funds.  In that role, I supervise our 14 
20   investment teams, managing roughly $250 billion in 
21   assets.  And I personally manage a few bond funds myself. 
22             Over the last few years, much like other 
23   derivative instruments available in the marketplace, we 
24   have been extensive users of ETFs, especially credit 
25   ETFs, to get in and out of exposure in a very efficient 
0094
 1   and timely fashion.  We find the liquidity conditions in 
 2   these markets to be quite good and extraordinarily 
 3   helpful for asset managers for all types of liquidity and 
 4   risk management -- risk management purposes. 
 5             The liquidity mismatch was a narrative that was 
 6   created in the marketplace when the -- when these large 
 7   or these generic passive ETFs got going in early part of 
 8   the decade.  And it came from the dealer community which 
 9   kind of -- bond dealer community which sort of 
10   misunderstood the ETF product. 
11             By its very definition, the ETF products are 
12   meant to be traded amongst ETF investors as opposed to 
13   accessing that liquidity in the native -- in the native 
14   market.  That is not to say that all bond ETFs are very 
15   liquid and provide that level of liquidity that we so 
16   crave.  But for the most part, the larger ones, the 
17   generic ones, the passive ones are very useful tools in 
18   the marketplace. 
19             MR. REDFEARN:  Thank you, Krishna. 
20             Greg. 
21             MR. PETERS:  Hi.  I'm Greg Peters.  I'm a 
22   portfolio manager for PGM Fixed Income.  We're a 700 
23   billion active fixed income manager, both institutional 
24   as well as retail fund, so we cover the lot. 
25             From my perspective, I look at ETFs as a 
0095
 1   potential way versus other options, which I'll speak of 
 2   in a moment, to manage my portfolio inflows and outflows, 
 3   as well as to modulate risk.  And so the previous panel 
 4   talked a lot about liquidity and so portfolio managers, 
 5   particularly ones that have sizeable assets under 
 6   management to help manage for their clients, they turn to 
 7   other alternatives due to the illiquidity in the 
 8   marketplace. 
 9             So there's really four options that I see.  The 
10   first option that you have as a portfolio manager in 
11   fixed income is the use of total return swap.  That could 
12   either be an off-market or tradeable instrument.  The 
13   second is in synthetic form as well, the CDX index.  
14   There's multiple uses there, from investment-grade 
15   corporates to high-yield, for example.  The third is an 
16   ETF itself.  You know, these are passive instruments that 
17   I'm talking about.  Then fourth, really just options on 
18   those.  And so you could use options on CDX, TRS as well 
19   as ETFs.  And so that's, you know, what we evaluate. 
20             There's clearly pros and cons on each of those. 
21    But for me, as a portfolio manager, I actually evaluate 
22   ETFs relative to these options as acting as a fiduciary 
23   for my end clients as just part of the evaluation 
24   process. 
25             MR. REDFEARN:  Thank you. 
0096
 1             So the first question we were going to get into 
 2   next was more on this question of the liquidity mismatch. 
 3    And let me just say a couple things and try to solicit a 
 4   couple more thoughts on this if possible. 
 5             I mean, clearly you have a marketplace in the 
 6   equities markets where it's easy to get in and out and 
 7   things can trade pretty quickly, versus a relatively -- 
 8   often relatively difficult to trade or even difficult to 
 9   price underlying securities.  There are arguments that, 
10   to a certain extent, this can provide a greater level of 
11   price formation in the product.  On the other hand, there 
12   are concerns that sometimes if the supply and demand gets 
13   out of whack effectively in the equities market that 
14   there can be increasing dislocations through, you know, 
15   when you look at sort of correlated trading. 
16             So let me just ask before we move on to 
17   something else, any other thoughts about how, you know, 
18   given the different nature of the liquidity between the 
19   ETF and the underlying products, if there are any other 
20   considerations pertaining to this, whether it's for 
21   retail or institutional traders?  Or is there anything 
22   that in this landscape that investors should be concerned 
23   about vis-a-vis the liquidity in these products?  And I 
24   will just open that up to anybody. 
25             MR. BERGER:  So I think under normal 
0097
 1   circumstances, the ETFs are adding a lot of liquidity.  
 2   If you want to be buying and selling, you're crossing 
 3   really tight spreads.  Oftentimes, it's a basis point 
 4   wide.  I think that's, you know, a great advantage of 
 5   ETFs.  And I think if investors are using them in a smart 
 6   way, that they provide a lot of benefits. 
 7             One thing that's important, I think, is client 
 8   education.  So we spend a lot of time educating people 
 9   about, all right, this is how you should use ETFs.  You 
10   should trade them during the periods where they are most 
11   liquid and where the underlying is most liquid.  You 
12   shouldn't use market orders, you shouldn't use stop 
13   orders.  Kind of normal equity market structure ideas. 
14             I think if people are using them responsibly, 
15   they are going to have a very good experience. 
16             When we talk about dislocations, I think there 
17   are a couple different things.  One, there is kind of a 
18   perceived dislocation.  If you look at NAVs for fixed 
19   income ETFs, they are not necessarily going to update 
20   right away.  I mean, we heard a lot about transparency in 
21   the last panel.  But oftentimes if it appears that an ETF 
22   is trading at a discount or at a premium, it's actually 
23   the ETF has the correct information and the bonds just 
24   kind of haven't gotten there yet because they're not 
25   trading as rapidly and as quickly.  So I think oftentimes 
0098
 1   there may be a perceived premium or discount that doesn't 
 2   actually exist. 
 3             You know, there are firms like us and others 
 4   that are, you know, kind of competing for that business 
 5   and keeping that premium and discount as small as 
 6   possible. 
 7             I think to your client education portion, it's 
 8   also important for people to realize if you buy a high-
 9   yield bond ETF, at the end of the day you own high-yield 
10   bonds.  It's just as Ananth said, you have shares in this 
11   fund that holds high-yield bonds.  There's not a magical 
12   way to translate that portfolio into something that is 
13   more liquid than what it otherwise is.  And on a given -- 
14   on a normal day, there's a lot of liquidity. 
15             If, for example, you know, there's continuous 
16   outflows and, as I've talked about, then we're going to 
17   be creating or redeeming.  We're going to take those 
18   bonds onto our balance sheet and we're going to take the 
19   risk, you know, for those that we can't sell.  But at the 
20   end of the day, if that outflow continues, we're going to 
21   be relying on the market to absorb those bonds and we're 
22   going to hit our risk on this.  We're going to need to 
23   hedge in order to buy more. 
24             So, you know, I think it is enhanced liquidity 
25   during most times and a great product and leads to a lot 
0099
 1   of transparency.  But investors should understand that at 
 2   the end of the day, they own a share of an asset.  You 
 3   know, it might be very liquid or it might be less liquid. 
 4    It might be treasuries, it might be illiquid high-yield 
 5   bonds.  And investors should just understand what they're 
 6   buying when they buy. 
 7             MR. COLLINS:  I underscore what Matt just said 
 8   about investors should understand what they're buying and 
 9   that education is really important.  And I would also 
10   underscore his point that sort of at least it was 
11   implicit in there that ETFs are adding to price 
12   discovery. 
13             So you may have a relatively illiquid or 
14   underlying security that's not trading at certain parts 
15   of the day, and that could be in the ETF's basket.  The 
16   ETF is providing liquidity by bundling those securities 
17   up and allowing investors, retail investors, could be 
18   institutional investors to trade them in a more liquid 
19   market.  So it's providing liquidity and price discovery. 
20             MR. PETERS:  Here, I'll jump in --  
21             (Laughter.) 
22             MR. PETERS:  Look, I mean, there's a couple 
23   other things to consider though.  And so the first and 
24   foremost is that the performance does lag.  And so if you 
25   look at, you know, ETFs relative to active management, 
0100
 1   there is a lag.  And that is something to consider.  And 
 2   while I wouldn't say that's an ETF problem in and of 
 3   itself, it's more of an index construction issue, it 
 4   still, I think, speaks to passive management within fixed 
 5   income. 
 6             The second is the cost.  And so there is much 
 7   more cost associated, too, than just the management fee, 
 8   you know, implies.  And I think that's a byproduct of the 
 9   bid/offer/execution that's being inside that structure.  
10   And so it does manifest itself in a tracking error vol 
11   which is much higher in fixed income ETFs relative to 
12   equity ETFs.  And so me, as a manager who puts that into 
13   one of my funds, I'm actually introducing volatility into 
14   my fund, all else being equal relative to my benchmark. 
15             And then another big issue I think is 
16   important, is from a fixed income, active management 
17   perspective, I don't have complete price transparency.  
18   And so we talk about the redemption process, so on and so 
19   forth, but that's a sliver, right?  And so, you know, the 
20   other 900 securities within that structure, I am 
21   personally not a hundred percent comfortable with me 
22   getting best execution on that and thereby not getting 
23   best execution for my clients, acting as a fiduciary. 
24             And then last, unlike CDX, for example, there's 
25   no real ability to short it, right?  So it's a one-way 
0101
 1   market.  And so the borrowability is quite squirrely.  So 
 2   it's really tough to borrow.  And so I think that's 
 3   something to always consider when it's just effectively a 
 4   long-only product. 
 5             MR. MEMANI:  I like that, from the last panel. 
 6             So I think ETFs, from my perspective, add one 
 7   element into the corporate bond trading arena that is 
 8   extraordinarily important.  Corporate bond traders, every 
 9   single one of them in different economic regimes all head 
10   one way.  And the problem that you have in that 
11   environment is that there's nobody to take the other side 
12   of the risk because of the aggregation problem that we 
13   have.  So if you -- if you are a bond trader and you 
14   wanted to buy Verizon sevenths of 36, then you have to 
15   buy Verizon sevenths of 36.  On the other hand, through 
16   an ETF mechanism, now you bring in all sorts of other 
17   players in the marketplace that can take the other side 
18   of that risk in an aggregated way, who would never play 
19   in Verizon sevenths of 36.  So that brings in other pools 
20   of capital into the marketplace that would not exist 
21   otherwise.  It's the same reason why index trading, in 
22   the derivative market, is far deeper than individual CDX 
23   trading.  So that's one important consideration. 
24             The second important consideration is the issue 
25   of stale prices.  So when people say that there is an 
0102
 1   arbitrage opportunity that exists and that arbitrage 
 2   opportunity at various points is huge, it's because you 
 3   have different marks that have been completed at 
 4   different times of the day.  ETFs, because they are 
 5   actively traded instruments, actually provide you 
 6   realtime marks, again, on an aggregated basis.  And 
 7   therefore, they are more timely with an investor buying 
 8   and selling at the same time.  So it is a much more 
 9   reliable information that you can -- that you can get. 
10             The third issue in that regard is I think the 
11   price transparency, every level of disclosure with 
12   respect to trading on a daily and a kind of minute-by-
13   minute, second-by-second basis, that provides a great 
14   deal of information to people to manage their overall 
15   risk profile in the market. 
16             Having said that, I think that much like other 
17   ETFs, whether they are stock ETFs or bond ETFs, sometimes 
18   because of market shock, supply and demand overwhelms it, 
19   finding ways of managing that dislocation in the 
20   marketplace from a regulatory standpoint will build 
21   confidence in these types of products. 
22             MR. REDFEARN:  So one of the questions I have 
23   is, as we mentioned before, there are a lot of ETFs out 
24   there now that are trading in the bond market.  One of 
25   the things that we've seen with the data is that today, 
0103
 1   over 75 percent of the total volume in corporate and 
 2   municipal bond ETF trading occurs in just 10 ETFs, and 
 3   that these 10 ETFs comprise 65 percent of assets under 
 4   management by corporate and municipal bond ETFs.  So 
 5   maybe, Joe, I'll throw this one at you.  Which is, you 
 6   know, when your clients are trading in bond ETFs, is the 
 7   focus on just a few of these things or do you trade 
 8   across the entire spectrum of 164 corporate and muni bond 
 9   ETFs? 
10             MR. ARCADI:  So you're right when you notice 
11   there are a lot of ETFs on the market.  Most of them have 
12   been utilized as investment vehicles by a handful of 
13   market participants, potentially some of which are 
14   retail.  Some of them have been new products that issuers 
15   have provided to try and generate client interest. 
16             The institutional clients that we've seen are 
17   mainly interested in a handful of tickers.  Now, these 
18   are ETFs that have bridged the gap from an investment 
19   vehicle to a trading instrument.  And we generally will 
20   be focusing on those handful of tickers that can support 
21   institutional-size trades. 
22             So I think a lot of the comments that we made 
23   here today, we should think about, okay, are we talking 
24   about these trading instruments, are we talking about the 
25   ETF as an investable instrument and the problems that may 
0104
 1   come along with either one of those. 
 2             MR. REDFEARN:  Anybody else on that particular 
 3   one in terms of the scope of trading --  
 4             MR. MEMANI:  I think Joe's point is an 
 5   extraordinarily good one.  Which is, there are ETFs, bond 
 6   ETFs, that are liquidity vehicles in the marketplace.  
 7   And then there are investment vehicles.  So I think 
 8   investment vehicles are where investors are looking for a 
 9   type of risk structured in a particular way with a 
10   particular investment philosophy.  Those people are 
11   interested in particular, specific ETFs.  And those tend 
12   to be because they are so structured and so unique and 
13   customized, end up being less liquid than otherwise.  And 
14   in that regard, they are no different than mutual funds 
15   or very specific stylized mutual funds. 
16             The other side is these 10 ETFs that you talk 
17   about, and I think those have gotten to a large enough 
18   size that end up being very liquid and are the liquidity 
19   vehicle in the marketplace. 
20             MR. REDFEARN:  So how do we -- I'm sorry go 
21   ahead, Sean. 
22             MR. COLLINS:  I would add two things to that.  
23   One is, I think, you know, maybe part of what you're 
24   seeing is that, I think, you know, in some sense, 
25   liquidity is a virtuous circle.  So people come to the 
0105
 1   ETFs and there's trading and the increased trading in a 
 2   particular ETF ticker symbol begets more demand to come 
 3   use that ticker symbol, so that's what we see in the SPY. 
 4    And there's a lot of trading in it and there's a lot of 
 5   demand for trading in it, because it's very deep in 
 6   liquid security. 
 7             The other thing I would say is, to the extent 
 8   that there are a lot of smaller bond funds, bond ETFs and 
 9   perhaps have smaller trading volume on the U.S. stock 
10   exchanges, that doesn't necessarily mean that they are 
11   not providing liquidity.  For example, you could have an 
12   emerging market bond ETF.  And trading volume on bond 
13   markets in, let's say, Indonesia is very limited.  The 
14   ETF may, through a market in the U.S., be providing 
15   liquidity to people that want to express a position in 
16   Indonesian securities.  So I would be a little bit 
17   careful about assuming that because an ETF is big and 
18   that's where the trading volume is, that the smaller guys 
19   are not adding liquidity to the market. 
20             MR. REDFEARN:  So we have looked at a lot of 
21   the data in terms of what we see in secondary market 
22   trading.  And, for example, we have seen that 70 percent 
23   of corporate and municipal bond ETFs trade less than 
24   100,000 shares a day.  Which, from an equity perspective, 
25   generally we'd say that's a pretty illiquid range to be 
0106
 1   trading in. 
 2             Appreciating your points that you made, Sean, I 
 3   guess the question is when we think about these other 
 4   issues in terms of the mismatch and in terms of the 
 5   investors that are potentially trading in these products, 
 6   are there different concerns for investors and different 
 7   concerns for the liquidity mismatch issue for some of 
 8   these products that really are fairly thinly traded. 
 9             MR. BERGER:  I guess I can talk about that.  So 
10   we make markets in all of the ETFs in the U.S.  And I 
11   think if you are using these vehicles, again, a lot of it 
12   does come down to how are you using it.  Somebody calls 
13   us up and they say, I want to buy -- I'm looking for an 
14   offer on $50 million of this symbol that only trades $5 
15   million a day.  That doesn't mean they're going to get a 
16   bad price.  That means that they're going to be priced at 
17   where kind of the arbitrage is and it might not matter 
18   whether that stock trades a lot in the secondary or not, 
19   because we're going to hedge it in the underlying market. 
20             For example, if it holds a basket of IG bonds, 
21   even though the ETF doesn't trade much, it's going to 
22   have a tight spread when quoted with, you know, one of 
23   the ETF specialist market makers.  So getting in and out 
24   won't be that expensive but it does require, you know, 
25   understanding what you're doing.  If you just were to 
0107
 1   send a huge order into the open or the close of the 
 2   exchange, I would recommend against doing that or 
 3   something like that.  But if you call around, call a 
 4   bunch of people, get a bunch of competitive prices, 
 5   you're going to do all right. 
 6             MR. COLLINS:  Can I just follow up on that one? 
 7    So I think one way to think about it is that it's no 
 8   different than a small cap stock.  So a small cap stock 
 9   is not going to be as liquid as a large cap stock.  Do we 
10   raise considerations about that for systemic reasons?  I 
11   don't think so. 
12             So that's exactly what's going on here.  There 
13   are some ETFs that may have a lower trading volume.  You 
14   can think of it the way you would think of trading a 
15   small cap stock.  It's perhaps less liquid.  There's 
16   perhaps a wider bid/ask spread.  But I don't see that it 
17   necessarily implies some kind of a liquidity mismatch 
18   problem. 
19             MR. BERGER:  I guess I would suggest that 
20   liquidity for a lot of these ETFs, a small cap 
21   consideration might be true for like the bid/ask.  If you 
22   want to buy or sell 100 shares, it might be wider.  But 
23   oftentimes, for a larger size, you know, as somebody who 
24   is going to do a creation or redemption, we're able to 
25   tap into the underlying market.  And I think a new issuer 
0108
 1   that comes out with an S&P 500 ETF, they're still going 
 2   to be able to get a tight market on $200 million of that 
 3   ETF because it's drawing upon the liquidity of the S&P 
 4   500. 
 5             MR. COLLINS:  Yeah, just to be clear, I was 
 6   talking primarily about liquidity for a retail investor. 
 7             MR. REDFEARN:  So, Matt, you mentioned the 
 8   arbitrage process.  And, obviously, the arbitrage process 
 9   is critical in terms of making sure that the prices 
10   remain somewhat aligned.  We do have examples.  In 
11   particular, August 24, 2015, when there was a significant 
12   dislocation around the open where many people would say 
13   on that particular morning and day that the arbitrage 
14   process did completely break down for a number of 
15   different products. 
16             I guess when we think about in particular among 
17   some of these less liquid products or products where the 
18   underlying tends to be less liquid, you know, what are 
19   your thoughts there about ensuring that those prices 
20   remain aligned, the prices don't necessarily break down 
21   and, in particular, during times of stress, how do we 
22   make sure that investors who are trading those products 
23   are getting the right prices. 
24             MR. BERGER:  I think that's an excellent 
25   question.  August 24, I guess a couple points just to 
0109
 1   start off.  I would say it was primarily an equity market 
 2   structure issue and then kind of the ETFs secondarily on 
 3   top of that.  And fixed income ETFs performed much better 
 4   than equity ETFs that day.  It really was the equity 
 5   market that was having the struggle.  I think a lot of it 
 6   was because stocks weren't opening at the normal time, 
 7   futures were down.  There were lots of things going on. 
 8             We've been working very closely, our firm and 
 9   many others, on market structure changes.  They're 
10   putting in limit up, limit down bands, working with 
11   exchanges to change their rules around opening.  There's 
12   a lot of work that's being done to, you know, ensure that 
13   something like that doesn't happen again.  But again, in 
14   fixed income, we didn't really have that issue the way we 
15   had it in the equity market. 
16             But I think, again, it's also investor 
17   education.  You know, using limit orders, not using stop 
18   orders, some of the same things I talked about I think 
19   are very important for investors to have good results. 
20             MR. REDFEARN:  So in terms of trading the 
21   underlying product when you're trying to keep them 
22   aligned, when the underlying product is not particularly 
23   liquid, you don't see that as an issue in terms of being 
24   able to maintain --  
25             MR. BERGER:  I think it's a good point.  You 
0110
 1   know, again, when people are trading ETFs, I think they 
 2   should try to trade as best they can the ETF when the 
 3   underlying market is at its most liquid.  So if you're 
 4   trading corporate bond ETF at 4:30 in the afternoon, you 
 5   should think twice.  You probably want to trade during 
 6   the most liquid time because market makers such as 
 7   ourselves are relying on the underlying market at some 
 8   point. 
 9             If you want to trade a small amount after 
10   hours, we're able to model the bonds and we're able to 
11   make good markets.  If you want to trade a couple hundred 
12   million, that's not really an issue.  But if we talked 
13   about huge, you know, outflows during a time where the 
14   bond market was less liquid, that, you know, again, it 
15   could become a concern.  But we've seen historically when 
16   there's been big outflows, that the bond market has been 
17   there and we've been able to work out of our risk via 
18   dealers, via electronic platforms, via all the different 
19   ways that we connect to the Street. 
20             MR. PETERS:  I would say that the ETF 
21   infrastructure has really benefitted from the bull market 
22   here.  If you go back just a couple years ago, I used to 
23   see really wild swings, particularly in high yield where, 
24   you know, ETF bonds would trade in a vacuum and prices 
25   would drop 10 points or so.  You actually don't see that 
0111
 1   today.   
 2             And so while I don't necessarily think they're 
 3   a great fit for my fund specifically, I would say the 
 4   infrastructure itself looks much better behaved today 
 5   than it has in the past.  We haven't experienced another 
 6   real downturn in a meaningful way, but I would say that 
 7   the infrastructure looks a lot better today from a bond 
 8   perspective than it did. 
 9             MR. COLLINS:  So just to add a data point on 
10   Matt's issue, so Matt pointed out that, you know, when 
11   there are big outflows there could be -- there could be 
12   issues.  But it's important to keep in mind what big 
13   might mean for a market maker is probably different than 
14   what big means for a regulator.  So, for example, in 
15   early February, we had big outflows from high-yield bond 
16   funds, second largest on record on one particular day, 
17   $23 billion in outflow. 
18             In terms of assets under management in high-
19   yield bond funds, that was only 1.1 percent of high-yield 
20   bond fund assets.  So just keep in mind that there can be 
21   a difference of perspective here. 
22             MR. REDFEARN:  I think, Krishna, we have to 
23   reward you for using the methodology there of turning 
24   your thing -- so, please. 
25             MR. MEMANI:  No worries.  So two observations. 
0112
 1             First, I think from a liquidity standpoint, 
 2   even for less liquid ETFs, the right comparison in that 
 3   regard is smaller mutual funds.  That is, at the end of 
 4   the day, the liquidity environment that both mutual funds 
 5   face and ETFs face is basically a matter of hours.  That 
 6   is, if you couldn't liquidate that particular position in 
 7   one format and you're challenged with an ETF, you have 
 8   the same challenge with a daily liquidity mutual fund 
 9   that you have to address in some way.  So in that regard, 
10   they are really not that different. 
11             And the issue that you raise about the August 
12   2015, I think Matt's point is a good one.  That is, it 
13   was a structural issue with ETFs in general, rather than 
14   a liquidity issue with bond ETFs in particular.  So 
15   whatever regulations or what market facilitation 
16   mechanisms that you put in place for regular ETFs I think 
17   would help bond ETFs just as much.  They just don't need 
18   to be any more specific with respect to bonds. 
19             MR. REDFEARN:  Yeah, that's a great point.  And 
20   looking at a lot of those names, part of the reason the 
21   arbitrage broke down is there were a lot of underlying 
22   names that just weren't even open in the morning because 
23   they opened late or they halted.  And that was one of the 
24   reasons why you saw the dislocation. 
25             In this particular case, though, we have a 
0113
 1   situation where, whether it's a certain time of the day 
 2   where there's not a lot of liquidity in the underlying 
 3   bonds or whether it's just not a bond that trades 
 4   particularly much, there are some similar dynamics that 
 5   could potentially overlap in terms of the pricing 
 6   mechanism sort of staying in line. 
 7             So I think our concern is probably it sounds 
 8   like most of the time, everything is pretty good.  It's 
 9   the concern is what happens when, you know, we have an 
10   adverse market event when the underlying bonds aren't 
11   particular liquid, when something is potentially going 
12   on, to just ensure that there is no real investor risk 
13   during those times.  That's what we're -- I'm sorry, 
14   Matt, did you want to jump in again? 
15             MR. BERGER:  I kind of want to jump in and talk 
16   a little bit about 2008.  You know, people talk about 
17   fixed income ETFs as being a new product, and they are a 
18   new product, but they were around during 2008.  They were 
19   around when, you know, oil was selling off and high yield 
20   sold off a lot in the beginning of 2016.  There have been 
21   a few tests and I think they've performed pretty well. 
22             In 2008, people talked about really big 
23   discounts.  But I know we were long the ETFs and we were, 
24   you know, having kind of a hard time selling some of the 
25   bonds.  And at the end of the day, when everything 
0114
 1   converged, the ETFs had been the fair price.  They were 
 2   the accurate reflection of what was fair.  And I know, 
 3   because we didn't make money. 
 4             (Laughter.) 
 5             MR. BERGER:  So I think the ETFs are, you know, 
 6   actually a pretty good reflection of where the bond 
 7   market is. 
 8             MR. REDFEARN:  That's good to know. 
 9             So we want to talk for a second about the fact 
10   that right now currently less than 5 percent of bonds are 
11   held by bond ETFs.  And even though the bond ETF market 
12   has been growing by about 20 percent a year over the last 
13   five years, we would like to kind of get your sense on 
14   where is this market going, what is the outlook for this? 
15    You know, what do you foresee coming in the future in 
16   terms of where this -- how this market evolves and grows? 
17             MR. ARCADI:  Just to comment on that a bit, I 
18   think it is important to look at which of these ETFs have 
19   attracted assets.  You talk about under 4 percent 
20   -- I think you might have mentioned 5 percent of bond 
21   assets are within ETFs.  Of the space I look at, 1 
22   percent of the high-grade universe falls under an ETF and 
23   4 percent is in high yield.  So it's fairly small and the 
24   big one is high yield.  High yield is the one that has 
25   become the trading instrument of choice. 
0115
 1             So a lot of what we're talking about here, not 
 2   to be nuanced about, we're talking about a couple tickers 
 3   that are high-yield bond funds that are extremely 
 4   popular.  Why are they popular?  Well, 4 percent of 
 5   assets under management are with the ETFs of high-yield 
 6   bond funds.  About 20 percent are reporting mutual funds, 
 7   are held in reporting mutual funds.  So these mutual fund 
 8   assets under management are volatile also.  So mutual 
 9   fund managers, there are times where, over a two-week 
10   period, their assets under management may change by as 
11   much as 6 to 8 percent. 
12             So as a mutual fund manager, you're going to 
13   need to keep large cash balances, have very liquid 
14   instruments in your portfolio that you can liquidate 
15   quickly, or you can use some of these access products 
16   that have come to the market to help you keep your bond 
17   exposure that you've promised to investors and manage the 
18   liquidity profile that you need when you have potential 
19   for daily redemptions. 
20             So there has been a need in the market to -- 
21   for these instruments and it's one of the reasons why 
22   these ETFs have come into this tradeable instruments 
23   sphere because they actually are solving a real problem 
24   for the market. 
25             And you look at the high-yield market also.  
0116
 1   It's very idiosyncratic and there are many different 
 2   bonds, many different issuers.  They can be very 
 3   volatile.  Just modern portfolio theory.  Any time you 
 4   break up a diversified portfolio into the risk of the 
 5   individual components, it's going to be more difficult to 
 6   trade, it's going to be more costly to transact. 
 7             So by aggregating all these together into a 
 8   market-wide index portfolio, you can reduce the risk of 
 9   the portfolio and it provides a useful instrument for 
10   investors, traders, anybody who is managing high-yield 
11   risk.  It's a tool in their toolbox that has become very 
12   useful over the past couple years. 
13             MR. REDFEARN:  Anybody else on that one? 
14             MR. MEMANI:  Our expectation is that the ETF 
15   market will continue to grow, and probably more so on the 
16   active space than it is or -- or more so in active than -
17   - active equity ETFs in the bond space than otherwise.  
18   And the reason for that is relatively straightforward. 
19             In the equity market, disclosures, the daily 
20   disclosures, is a really big issue for portfolio managers 
21   and it's not such a large issue for bond portfolio 
22   managers.   
23             Having said that, I think the growth of ETFs in 
24   general in bonds is probably not going to be as great as 
25   it has been in the passive equity space for one very 
0117
 1   important reason, which is the ability of active 
 2   portfolio managers, such as my friend Greg Peters, to 
 3   deliver extraordinarily good results over a long period 
 4   of time.  And therefore, people's attraction towards 
 5   mutual funds is not going to be taken off just because of 
 6   liquidity availability in the marketplace. 
 7             MR. REDFEARN:  So I won't ask you, Greg, if you 
 8   can deliver extraordinary results over a long period of 
 9   time.  But I will ask this.   
10             The related question though that we're thinking 
11   about here is, is there a scenario where the growth of 
12   bond ETFs gets to a point where there starts to become 
13   almost a tipping point, when you look at the size of 
14   what's happening in the ETFs relative to the underlying 
15   cash bond market that starts to create some sort of 
16   structural risk or destabilizing scenario in the future? 
17             MR. PETERS:  So just to follow up on Krishna's 
18   point, if there's growth in fixed income ETF land, I do 
19   think it is on the active side.  I think there's limited 
20   capacity for the passive side to get much, much bigger, 
21   just given how complex fixed income is relative to 
22   equities.  And so we have -- let's say we have one issue 
23   of Verizon.  There's 45 different securities to choose 
24   from.  You have an index that is debt weighted.  So 
25   there's a perversity function there which is you have to 
0118
 1   buy more of a company that is worse off.  And so that 
 2   leads to just inferior results on the passive side and, 
 3   quite frankly, the index itself. 
 4             So I think in order for ETFs to really expand, 
 5   it has to be on the active side.  And to me, it's all 
 6   about pricing.  So there's really very little different 
 7   on an active mutual fund, active ETF.  It's really the 
 8   prevailing theme of fees are going down. 
 9             So, in short, yes, it can grow. 
10             MR. REDFEARN:  So again, back to the other 
11   question though about the potential tipping point and 
12   whether or not there's a scenario for destabilizing 
13   market event in terms of the growth relative to the 
14   underlying cash bond market, does anybody have any 
15   thoughts on the extent -- Sean, please. 
16             MR. COLLINS:  So you said at the outset that 
17   the bond ETFs account for less than 3 percent of the 
18   corporate bond market.  So I would say, you know, the 
19   first thing is, look, if there is a tipping point, and 
20   personally I'm dubious about that because there will 
21   always be guys like Greg that can come in as active 
22   traders and take advantage of dislocations in the market. 
23    So there is always going to be a place in the bond 
24   market for active trading, as Greg says. 
25             But let's say there is a tipping point.  
0119
 1   Hypothetically, we're at 3 percent right now.  So if 
 2   there's a tipping point, I don't know, it's over the 
 3   horizon, it's nowhere in sight.  Maybe it's, you know, on 
 4   the other side of the earth or maybe it's on the dark 
 5   side of the moon.  But at this point, it seems like this 
 6   is not really a particularly useful question to be 
 7   addressing. 
 8             You know, there are questions that could be 
 9   addressed.  But this is a very small portion of the 
10   market. 
11             MR. MEMANI:  I would concur with Sean.  Just 
12   given the size and you can attribute whatever growth rate 
13   you want on top of it, just it's -- if you start from 
14   puny, you'd get to little, not too far away from there. 
15             The other issue to be mindful of in that regard 
16   is the point that I made earlier.  That is the liquidity 
17   requirements of ETFs -- and I think if you think in 
18   boundary condition terms, the liquidity requirements of 
19   ETFs is really not that different from liquidity 
20   requirements of open-end mutual funds, where you provide 
21   daily liquidity.  So if there is a tipping point, given 
22   the size of mutual fund market, I would worry about 
23   mutual funds far more than I would worry about bond ETFs. 
24             MR. REDFEARN:  I have a few more questions 
25   here.  But I think, given the time, I think we'd like to 
0120
 1   open it up to the rest of the committee for any 
 2   respective questions. 
 3             CHAIRMAN HEANEY:  I'm just trying to stay in 
 4   order.  So Amar was up first, then Larry, then Ananth. 
 5             MR. KUCHINAD:  This question is probably best 
 6   answered by Joe or Matt, or maybe both of you. 
 7             Just from your experience, the panelists that 
 8   have talked a lot about the ETF create/redeem process and 
 9   how that's an additional layer of protection for 
10   liquidity.  But from your experience, are there any 
11   inefficiencies or lack of transparency in that process?  
12   I know, Matt, you mentioned that the I-NAV that's 
13   disseminated at times can be pretty far off from what you 
14   can experience as a trader in those instruments.  Are 
15   there things that you think warrant Commission attention 
16   or things that we should think about for improvement? 
17             MR. BERGER:  Yes.  So -- I guess, yeah, the 
18   funds benefit a lot from the creation and redemption 
19   process.  That's where we're able to access -- you know, 
20   the fact that the bond liquidity is -- generally, there's 
21   a lot more trading than there is in the ETFs.   
22             You know, the I-NAV, I think, is not accurate. 
23    One benefit of the ETF creation/redemption process as 
24   compared to mutual funds is that people have talked about 
25   how it's in kind.  So if we're pricing the ETF, we're 
0121
 1   thinking about, all right, what is this actual basket of 
 2   bonds worth.  Whereas, you could imagine if, you know, a 
 3   cash redemption, the early exiters kind of get the 
 4   benefit and the people that leave last are left with the 
 5   worst kind of slice of the fund.  In ETF world, that 
 6   doesn't happen and the fund managers make really hard to 
 7   make sure that, you know, the baskets -- the bonds that 
 8   are going into and out of the basket are chosen to, you 
 9   know, to represent the interest of the existing 
10   fundholders. 
11             MR. ARCADI:  Certainly, I'd echo Matt's point. 
12    You know, the ETF structure is very beneficial to 
13   remaining holders, if there's any redemptions.  I think 
14   the one place as far as inefficiencies may come is in 
15   periods of market stress, these redeem baskets tend to 
16   get very large.  So there could be additional transaction 
17   costs or just uncertainty of pricing in periods of market 
18   stress that may cause ETFs to trade at a bigger discount 
19   than you may have expected.  And that may be because the 
20   baskets have been broadened or it's just a lot more 
21   difficult to transact in a huge portfolio of bonds versus 
22   a handful of bonds to raise cash. 
23             MR. BERGER:  And I would say that the reason 
24   for those increased basket sizes and for some of the 
25   costs is to protect -- it's exactly to protect the 
0122
 1   remaining shareholders. 
 2             CHAIRMAN HEANEY:  Larry. 
 3             MR. HARRIS:  Here's a quick answer to Brett's 
 4   question about is it possible to be too big.  Imagine 
 5   that ETFs controlled say 99 percent of the bonds 
 6   outstanding.  Price discovery in the bond market then, of 
 7   course, would be in the ETFs and not in the underlying 
 8   securities.  And we wouldn't really know the values of 
 9   the individual underlying securities very well, but we'd 
10   certainly know the values of the ETFs quite well. 
11             Now, what's the damage of that?  The damage is 
12   almost nothing.  These securities exist.  They -- people 
13   aren't making decisions based on the values of those 
14   securities.  Presumably, the companies will either pay or 
15   they won't pay.  And that's that. 
16             But what about the primary market without that 
17   information?  Well, the primary market might be a little 
18   bit worse off because you can't see the values of the 
19   individual credits.  But the authorized participants are 
20   all very, very sophisticated people and they will be the 
21   primary players in the primary market and we'll be just 
22   about as well off as we would otherwise.  We'll see a 
23   little less information but I don't see that the loss of 
24   information is serious. 
25             Now, what's the probability of this happening? 
0123
 1    Almost zero.  Most of the securities are going to 
 2   continue to be held by institutions and others, and there 
 3   will be markets in the individual securities.  So I'm not 
 4   particularly concerned about it. 
 5             The main concern, I think, is what happens when 
 6   you allow people to hold lots of ETFs?  Do they tend to 
 7   be faster traders than they would be if they were holding 
 8   the underlying.  So the reason that lots of people hold 
 9   ETFs is because the underlying markets are kind of poorly 
10   structured.  That's why we're here, to try to figure out 
11   how to improve liquidity in the underlying markets.  
12   Improve liquidity in the underlying markets and we'll get 
13   people out of the ETFs or that we'll arrest the growth of 
14   the ETFs and people will hold the individual securities, 
15   where they probably have a lower propensity to trade them 
16   in the event of a crisis.  Certainly, probably retail 
17   that's the case. 
18             So the tradeoff is if you're worried about 
19   people rushing for the exits, then you need to get them 
20   into the individual securities where they're less likely 
21   to trade.  But to do that, you've got to lower the 
22   transaction costs and in the individual securities. 
23             MR. MADHAVAN:  So just one comment on the 
24   notion of the tipping point.  So we've heard the stats 
25   that, you know, the ETF share of the corporate bond 
0124
 1   market and the muni market is just -- is tiny or puny.  I 
 2   think the other consideration that I would sort of urge 
 3   people to think about is it's not just the share of the 
 4   market capitalization or the market value of bonds that 
 5   are held in this particular form of wrapper, but it's 
 6   also the turnover. 
 7             So these are passively managed indices.  They 
 8   have relatively low turnover compared to, say, an active 
 9   manager or the typical active manager.  So it's the money 
10   in motion that also matters in terms of our thinking 
11   about the tipping point.  So we are very far away from 
12   any notion that, you know, ETFs or even passive products 
13   might dominate the market.  That's really the only 
14   comment. 
15             And I know Sean and his team have done a lot of 
16   work on the secondary market activity relative to primary 
17   market.  I don't know if you want to just comment on 
18   that, Sean? 
19             MR. COLLINS:  More than 80 percent is in the 
20   secondary market. 
21             CHAIRMAN HEANEY:  Kumar. 
22             MR. MEMANI:  Just one quick comment on that.  
23   So I think as far as price discovery is concerned, again, 
24   the issue there is nothing unique to bond ETFs.  That is 
25   generic to ETFs in general.  Having said that, there is a 
0125
 1   large institutional component of investors that could 
 2   certainly help, and large asset managers that service 
 3   that institutional asset owners, where there will be 
 4   ample opportunities to come up with -- come up with price 
 5   discovery. 
 6             As far as ETFs, ETFs growing to such a level 
 7   that they will cause a problem in the marketplace, again, 
 8   I would come back to the boundary condition.  In the 
 9   boundary condition, ETFs are no different than open-end 
10   mutual funds.  It's eight hours of the day.  That is the 
11   only difference between the two. 
12             MR. VENKATARAMAN:  This is a question for Sean. 
13    In terms of understanding the risk management from the 
14   perspective of a market maker, so suppose there is a 
15   stress event where there's a change in sentiment and so 
16   you're feeling selling activity -- seeing selling 
17   activity in the underlying.  And so the dealer is net 
18   long, given the imbalance.  And the same dealers also are 
19   participating as APs in the ETF world.  And in the ETF as 
20   well, there is an imbalance and so the AP is picking up a 
21   long position and they're getting in-kind redemption.  So 
22   how will the risk management -- how will that work out?  
23   Because it looks like there would be a systemic imbalance 
24   both in the underlying and in the AP community. 
25             So typically you would find that if you don't 
0126
 1   have obligations to participate, APs typically will 
 2   withdraw because, you know, it's not profitable to make 
 3   markets under these circumstances.  So how would this 
 4   play out?  Where would they go and, you know, hedge their 
 5   positions?  I'm just curious. 
 6             MR. COLLINS:  This is probably a question that 
 7   is better addressed to Joe or Matt.  But I would say, 
 8   look, if APs, dealers, market makers pull back from the 
 9   market, then spreads widen in the secondary market.  I 
10   think that that would be the outcome. 
11             MR. BERGER:  Yeah, I mean, one thing about ETFs 
12   is you have brought in a bunch of new participants, such 
13   as ourselves and others, who are willing to take, you 
14   know, again, some amount of risk.  Like we're willing to 
15   take a pretty good amount of risk but at some point, 
16   maybe we can't sell more. 
17             One nice thing about the ETFs and about 
18   transparency in general, we talked about, all right, 
19   people are now going to see the ETFs trading at a 
20   discount or trading cheaper.  It allows a lot of 
21   different types of fundamental players or hedge funds to 
22   come in and say, you know, now is a good opportunity for 
23   me to buy.  These things are cheaper than they 
24   fundamentally should be.  Whereas, if they were just 
25   trying to trade the bonds, that would be a very difficult 
0127
 1   thing to do.  They are going to have to cross very wide 
 2   spreads to put on that position.  Whereas, this allows, 
 3   you know, kind of more new players into the market. 
 4             MR. ARCADI:  I'll comment on that a bit, too. 
 5             On the market failure questions, I think one of 
 6   the biggest protections we have against market failure is 
 7   to have a diversified set of participants and it's going 
 8   to be liquidity providers, it's going to be authorized 
 9   participants, it's going to be arbitrage funds and, most 
10   importantly, it's going to be a diversity of client set. 
11             So ETFs, you're right to be pointed out that it 
12   tends to attract short-term capital.  So it is a trading 
13   vehicle.  You tend to get short-term investors in it.  
14   But anything that can be done that broadens the client 
15   type of ETFs and is more broadly marketed towards clients 
16   that will have different investment objectives, different 
17   mandates that could potentially offset some of the flows 
18   of the short-term investor crowd.  
19             And one last thing on the passive versus active 
20   while we're on it, also, because I think that's sort of 
21   getting lumped in with the vehicle selection portion of 
22   it, is that I think the passive versus active is probably 
23   an argument that's going to go on for some time.  I think 
24   that where it balances out is if the market gets overly 
25   passive, you're going to have active managers that are 
0128
 1   having excess performance that are going to reallocate 
 2   capital back towards -- away from passive back towards 
 3   active. 
 4             So you just hope that transparency in the 
 5   market and transparency of returns and fee competition is 
 6   going to balance that out to the point where passive 
 7   versus active stays pretty dynamic and it doesn't get 
 8   overly passive. 
 9             CHAIRMAN HEANEY:  Lynn. 
10             MS. MARTIN:  So my question is a little bit of 
11   a variant on the last question.  Do you all have a 
12   concern about the amount of APs in the market, 
13   particularly since you have frequently mentioned the 
14   create/redeem process and challenges within the 
15   create/redeem process?  Particularly, are there any 
16   recommendations you would have for us to consider, given 
17   the impact on dealer balance sheets and folks stepping 
18   into the role of an AP that we should be thinking about? 
19             MR. ARCADI:  I think Matt's very upset that 
20   there are more APs than him. 
21             (Laughter.) 
22             MR. ARCADI:  I'm not exactly sure how to 
23   answer.  I think more is going to be better.  And as far 
24   as like dealer balance sheet concerns or balance sheets 
25   of the APs, this tends to be a fairly clean process, 
0129
 1   where you're exchanging shares for bonds.  And the APs 
 2   are generally going to be risk light in the middle of it. 
 3    You know, I imagine, you know, Matt probably has a 
 4   pretty complicated book but, you know, in general the 
 5   goal is going to be to completely be clear after the 
 6   transaction. 
 7             MR. BERGER:  Joe's right.  Being an AP is also 
 8   a competitive business.  And, you know, if there weren't 
 9   enough, the existing APs can charge more.  It is a 
10   marketplace.  You know, for example, we are entering 
11   becoming an AP, probably many others are working on it. 
12             I think that, you know, right now we see 
13   sufficient.  We have multiple different APs that we use 
14   and then we are also adding the ability to use ourselves. 
15    It's not that hard of a process to become an AP.  It's a 
16   hard process to understand the full creation/redemption -
17   - everything that goes into it.  But actually signing up 
18   to be an AP is not a big process. 
19             MR. REDFEARN:  So we are pretty much out of 
20   time.  But as a regulator, I just need to ask one more 
21   question here which is, when we look at the current 
22   regulatory framework for bond ETFs, I just want to -- is 
23   it optimal as a regulatory framework, optimal?  Or is 
24   there anything that you think that we should be thinking 
25   about in terms of considering for the bond ETF market, a 
0130
 1   regulatory framework? 
 2             MR. MEMANI:  I think anything that we can do to 
 3   bring mutual fund regulation and ETF regulations closer 
 4   and make them fungible, I think, would facilitate 
 5   liquidity in the marketplace.  Effectively, you will be 
 6   providing vehicles for people to bring in other types of 
 7   investors just -- not just credit investors.  You know, 
 8   what you accomplished through the futures exchanges, 
 9   where you brought in speculative capital that didn't 
10   exist before. 
11             MR. REDFEARN:  Anything else? 
12             Sean, of course. 
13             MR. COLLINS:  There's a rumor that the SEC is 
14   working on an ETF rule.  So I would say that's a, you 
15   know, obvious first step.  You know, help in streamlining 
16   the approval process I think would be a good thing.  It 
17   would bring competition to the industry. 
18             And also to the extent that there's anything 
19   that can be done about custom baskets, that would help 
20   level the playing field and also bring further 
21   competition into the business, I think. 
22             MR. REDFEARN:  Joe? 
23             MR. ARCADI:  It may or may not be in the SEC's 
24   jurisdiction, but anything where regulatory changes will 
25   allow ETFs to be held by accounts and get accounting or 
0131
 1   capital treatment as if they were holding the underlying 
 2   asset class would be beneficial to diversifying the 
 3   investor base that can actually trade these products.  
 4   And that could also include being accepted as collateral, 
 5   as if it was the underlying instrument also.  You know, 
 6   so it's sort of a fungibility.  The ETF is a wrapper.  
 7   Anything that makes that wrapper decompose in the same 
 8   sort of risk and treatment as the underlying instruments, 
 9   I think, is going to be beneficial to the ETF ecosystem. 
10             MR. REDFEARN:  Thank you.  Matt? 
11             MR. BERGER:  A couple things.  I agree with 
12   Sean on the ETF rule.  I think a streamlined process 
13   would be a good thing for the industry.   
14             The other thing I wanted to talk about some, 
15   just to raise as an issue is -- it would be nice to have 
16   better collateral treatment on ETF redemptions.  This is 
17   a bigger problem outside of fixed income, it's a bigger 
18   problem in international ETFs.  But we would be very 
19   happy to talk anyone through the details of that.  I 
20   think it's hard to get into the details in this room 
21   exactly.  But the collateral treatment for redemptions is 
22   pretty punitive and I think that it would be great to 
23   find a way to work on that. 
24             MR. PETERS:  Then just last, just to follow on 
25   that point, and I'm not sure the SEC has jurisdiction, 
0132
 1   but the borrow is a real problem, to have a two-way, 
 2   long/short market.  And so I think something has to be 
 3   done on the borrowability and the margin that allows it 
 4   to trade more freely. 
 5             MR. REDFEARN:  Okay.  I think that we have run 
 6   over our time.  So with that, I would like to thank you 
 7   all very much for this engaging and informative 
 8   discussion.  Thank you very much. 
 9             We are now going to take a break for lunch 
10   until, I guess, 1:30.  All right, thank you very much. 
11             We'll break now. 
12             (Whereupon, at 12:42 p.m., a luncheon recess 
13   was taken.) 
14             A F T E R N O O N  S E S S I O N 
15             CHAIRMAN HEANEY:  Let's try to get started.  
16   And if there are people outside, hopefully they hear this 
17   and come on in. 
18             Why don't we get started. 
19             Our next panel, we're gong to turn to talking 
20   about the same basic topic on ETFs but the retail 
21   investor disclosure and the education issues surrounding 
22   both ETFs and bond funds, and some of the underlying cash 
23   bonds. 
24             We will begin by having Ananth provide a brief 
25   overview of ETFs and the bond funds subcommittee discussions 
0133
 1   and then we will break into the panel. 
 2             MR. MADHAVAN:  Thank you, Michael. 
 3             So, dear FIMSAC members, I wanted to take this 
 4   opportunity to update you on the activities of the bond 
 5   ETF subcommittee and bond fund subcommittee that I've had 
 6   the honor to chair. 
 7             So since our inaugural meeting, the 
 8   subcommittee has met four times, including two occasions 
 9   where we have had experts share their knowledge on issues 
10   that we've identified of special relevance.  There are 
11   three areas of focus that the subcommittee has 
12   identified. 
13             First, issues concerning retail investors, 
14   particularly around education and disclosure.  Second, 
15   the previous panel has covered volatility in bond funds. 
16    We also in our discussions covered liquidity, risk 
17   management and redemption risk.  And the third stream is 
18   the bond ETF ecosystem.   
19             We have met as a group on items one and two, 
20   item three is still to be scheduled.  And, as you will 
21   note, the subcommittee has gathered a considerable amount 
22   of relevant information that we would like to share with 
23   the broader FIMSAC group. 
24             So let me give you a quick recap of what we 
25   have done vis-a-vis the two streams so far and then we 
0134
 1   will head over to our panel on retail education. 
 2             The March 9 subcommittee meeting had a panel to 
 3   cover retail investor disclosure and education efforts.  
 4   We had a number of panelists from the retail sales side, 
 5   from the legal perspectives and requirements, as well as 
 6   perspectives on ETP and ETF naming conventions.  And we 
 7   also had Marc Sharma, the SEC Office of Investor Advocate 
 8   come to us to speak about investor education. 
 9             On the volatility in bond funds, I don't want 
10   to cover the same ground that we just did.  You've heard 
11   many of the same themes that were echoed by the previous 
12   panel explored by the subcommittee.  I just want to give 
13   you a highlight of who we had through and the topics.  So 
14   the liquidity risk management rules, we had the SEC's 
15   Division of Investment Management speak to us about that. 
16    We had a number of bond fund and ETF portfolio managers 
17   speak to us about fund risk management practices.  We 
18   also had Sean Collins from the ICI speak to us about 
19   their industrywide findings.  And we had an academic 
20   perspective with Adi Sunderam from Harvard Business 
21   School and Dan Li from the Federal Reserve Board sharing 
22   some of their perspectives about those issues. 
23             The string that we have not yet covered is the 
24   bond ETF ecosystem, meeting to be scheduled.  Some of the 
25   topics that we want to explore there are the role of the 
0135
 1   ETF sponsor, the role of the authorized participant, the 
 2   role of the ETF market maker and the role of the 
 3   exchange. 
 4             Right now, we are going to switch to our panel 
 5   on retail investing and investor disclosure and 
 6   education.  So let me kick it off by just introducing our 
 7   panelists.  And first of all, thank you very much for 
 8   taking the time to attend today and, you know, we are 
 9   looking forward to hearing your views on this. 
10             So starting off, we have Bob Colby, the chief 
11   legal officer of FINRA, who has also spoken to the 
12   subcommittee; Melissa Gainor, senior special counsel, 
13   Division of Investment Management, who has also again 
14   addressed the subcommittee; Nick Goetze, managing 
15   director, fixed income services, Raymond James; and Gary 
16   Mottola, research director of FINRA's educational 
17   foundation, who has again also spoken to the 
18   subcommittee.  So welcome, thank you all. 
19             All right, so let me just kick this off with a 
20   couple of preliminary questions that we explored at the 
21   subcommittee level and then I will direct some questions 
22   to the panelists.  So at the subcommittee, the topics 
23   that we looked at were, do retail investors understand 
24   the differences between '40 Act registered bond ETFs and 
25   other types of exchange-traded products?  Do retail 
0136
 1   investors understand what they are buying?  Are current 
 2   disclosures sufficient to ensure that retail investors 
 3   understand how bond funds and bond ETFs meet their 
 4   objectives?  Are there any improvements in fund naming 
 5   and/or marketing conventions to highlight some of the 
 6   differences that are there for retail investors in 
 7   particular? 
 8             We also talked a little bit about the 
 9   predominant uses of bond ETFs and bond funds by retail 
10   investors.  For example, retirement savings, college 
11   savings, day trading, et cetera.  And then finally, I 
12   think the subcommittee was informed about possible 
13   improvements or enhancements to existing disclosures or 
14   other types of communications with retail investors. 
15             So that is a little bit of the background on 
16   the subcommittee and its activities to date.  With that, 
17   let me start with our panel. 
18             So first question and let me preface it with a 
19   little bit of a, you know, preamble here.  So today 
20   retail investors have a variety of options to invest in 
21   corporate bonds or municipal securities.  They can, of 
22   course, buy those products directly and, you know, do so 
23   through their broker or financial adviser.  
24   Alternatively, they can invest in the United States bond 
25   markets by purchasing a mutual fund or ETF shares that in 
0137
 1   turn invest in these instruments and are designed to 
 2   provide investors with broad exposures to bonds. 
 3             So let me kick it off with a question for Nick. 
 4    So, Nick, as a retail intermediary, your firm is on the 
 5   front line of investor education and interaction and you 
 6   are very well positioned to understand the needs of your 
 7   retail investors.  How do you go about thinking about how 
 8   those needs might be addressed and what considerations 
 9   does your firm take into account when you are assessing 
10   how best to achieve a good result for your investors? 
11             MR. GOETZE:  Is that all?  Okay. 
12             MR. MADHAVAN:  That's it. 
13             MR. GOETZE:  Okay.  Well, quickly, my 
14   background, I started in the business as a retail 
15   financial adviser.  So I had clients.  I had sold them 
16   mutual funds, individual bonds.  I went on to become a 
17   trader.  I understand the capital markets side of it.  
18   And now currently in my role, I do a lot of the education 
19   for both the financial advisers and the end clients.  So 
20   in terms of, you know, my background, that gives you an 
21   idea of my area of expertise. 
22             You know, we try to educate advisers and 
23   particular clients on what's important about fixed income 
24   as kind of the beginning of the process when you decide 
25   what product they should be in, whether it's packaged or 
0138
 1   individual bonds.  And the first and foremost thing we 
 2   always work on with clients is the realization, the 
 3   single most important thing about a bond is that you get 
 4   your money back.  That's really why you buy them. 
 5             And so that starts with credit quality.  And 
 6   obviously, it goes into diversification quickly, and 
 7   obviously through a package like a mutual fund you can 
 8   achieve diversification in smaller dollar amounts. 
 9             And then we talk about yield to worst and yield 
10   to maturity.  Obviously, easy follow-ons.  We talk about 
11   coupon cash flow structure, which can be achieved through 
12   a money manager obviously selecting high-coupon bonds to 
13   produce more cash flow.  We do the same thing with 
14   individual buyers. 
15             Duration and maturity comes into the 
16   conversation.  It's an education point that I think is 
17   extremely important, I will get into that in a second, 
18   when it comes to the terms like effective duration and 
19   effective maturity and what that means.  And then how you 
20   pay for your investment.  Obviously PMP is coming out, 
21   the markup disclosure, May 14.  That is a significant 
22   change in our industry.  But I think it's really 
23   important that we focus on educating the advisers and the 
24   clients on the fact that, you know, how you pay for your 
25   bond can greatly influence returns.  And I say just 
0139
 1   briefly if you buy a 10-year bond and you buy and hold 
 2   the bond until maturity and you pay a 1 percent sales 
 3   charge, that's 10 basis points a year average cost over 
 4   the life of the investment.  If you pay 2 percent, it's 
 5   20 basis points a year.  And we have to look at those 
 6   comparisons to other options they have in the fee world 
 7   and the packaged product world in terms of the cost 
 8   that's associated with them and really educate folks on 
 9   what we're coming out with. 
10             And so as we continue kind of what I think 
11   about education and how we determine what's best for the 
12   client, it starts with diversification, get your money 
13   back.  If you can't afford to diversify in individual 
14   bonds, you need to buy an instrument like a mutual fund 
15   or an ETF.  We would also say that, depending on the 
16   structure and strategy of your portfolio, if you are 
17   actively investing for total return, you are going to 
18   have more success rebalancing most efficiently through 
19   either trading pure treasuries or an ETF.  That is going 
20   to be a more efficient model if you are a total return 
21   trader.  If you are simply a buy and hold investor, which 
22   is a lot of our investors, it's kind of -- you know, 
23   they've been successful in life and now they want to 
24   protect it, we look at that individual portfolio. 
25             When I think about education though, the truth 
0140
 1   is we are drowning retail with education.  And there's 
 2   only so much they can take.  And there's only so much 
 3   that even if I sat down with someone and I spent a half 
 4   hour right now and I thoroughly educated them on 
 5   duration, effective maturity, all the different things 
 6   they need to know, in two weeks when they come back in 
 7   here, 3 percent retention, maybe.  I think that one of 
 8   the key pushes though we have at our firm and at every 
 9   firm is the opportunity for mandatory firm element 
10   education.  You have to read it, you've got to take a 
11   test, it's -- you can take the test as many times as you 
12   want until you pass it.  But it requires the financial 
13   adviser to be up to speed on all the important things 
14   relative to the investments for their clients. 
15             And when it comes to ETFs, mutual funds, 
16   generally it's in a fee-based discretionary model.  
17   Oftentimes, the people making the decisions are the 
18   financial advisers with fiduciary responsibility for 
19   their clients.  They need to be thoroughly educated and I 
20   feel that's more of the target area we go after.  And 
21   it's on topics like effective maturity or, in your world, 
22   duration.  You know, these are things that on the surface 
23   you look at it and you say, okay, the duration of this 
24   fund is three.  Well, if it's all long premium dated 
25   bonds, that's because the expectation is the bonds get 
0141
 1   called on the call dates.  If rates go up relative to the 
 2   coupons in the portfolio, you have an extension that 
 3   takes place once you cross a barrier and your duration 
 4   goes from three to 13.  Well, the volatility in that 
 5   portfolio is much greater potentially than you realize on 
 6   face value.  And those are the kind of nuanced 
 7   educational points, I think, that are really important 
 8   that we get out there, so that people truly understand 
 9   what they own and know what they own. 
10             I think I answered your question, but --  
11             MR. MADHAVAN:  No, you did.  Thank you very 
12   much, Nick. 
13             And then sort of a follow-on question that I 
14   think is relevant before we turn it over to Melissa and 
15   Bob to comment on the regulatory side.  But could you 
16   just speak a little bit to some of the uses of bond funds 
17   by retail investors? 
18             MR. GOETZE:  The uses of bond funds, it's to 
19   achieve many different goals.  There are some folks who 
20   have a total return mindset and those are the people that 
21   say, I just want to look at my statement and see it 
22   higher every time I look at it.  And they're looking for 
23   fixed income, in that case, as an allocation based on 
24   their determination of market movements and they think 
25   that bonds will produce a return in excess of some 
0142
 1   benchmark.  And they are buying -- they are allocating 
 2   their bonds for that reason. 
 3             Other folks buy bond funds long term because 
 4   they just feel they want to be more conservative.  And 
 5   over a long period of time, regardless of interest rate 
 6   moves, they feel like they'll ride up and down and 
 7   they'll produce a constant cash flow. 
 8             And I would say a third, obviously, is cash 
 9   flow, people looking for income.  You are going to 
10   generally see a higher income coming out of bond funds 
11   than other types of funds.  And I think those are kind of 
12   the three biggest reasons we see it. 
13             But ultimately, though, for us, a lot of times 
14   the decision comes down to can you diversify in 
15   individual securities or are you trying for a total 
16   return, or you're not trying for total return but you 
17   can't afford to diversify and you need that 
18   diversification first and foremost in our process. 
19             MR. MADHAVAN:  Okay, very helpful context. 
20             Okay, let me turn this over to Melissa and Bob. 
21    We've heard from Nick about how retail investors may 
22   gain access to the markets.  We'd very much appreciate 
23   your perspective on, you know, how the regulatory 
24   environment applies to the various intermediaries that 
25   are interfacing with these investors.  And anything you 
0143
 1   can shed on that would be helpful. 
 2             MR. COLBY:  Thank you, and thank you for the 
 3   chance to be here.  FINRA is here in force, between Gary 
 4   and myself.  Gary works on investor education and 
 5   understanding how investors think about these products 
 6   and I'm going to be talking more about the regulatory 
 7   side, how we deal with brokers and how they deal with 
 8   clients. 
 9             First, let me just say FINRA, I hope, is well 
10   known in this room but not necessarily in the rest of the 
11   world.  We are a not-for-profit self-regulatory 
12   organization set up -- the predecessor organization set 
13   up in 1939 to assist the SEC in overseeing securities 
14   firms and the markets.  And that's what we try to do all 
15   day long, 3,500 employees, budget of north of 900 
16   million.  The vast majority of this is watching over 
17   brokers and watching over the markets. 
18             So we've thought a lot about disclosure, and 
19   this is Gary's world, disclosure to investors and how you 
20   educate them.  And it is a major tool.  But our 
21   experience has been that, in the securities firm world, 
22   the broker-dealer world, stocks are sold and not bought. 
23    That's the old adage.  And so we focus on the broker-
24   dealer advise provider.  Our sense is that investors that 
25   are seeking for advice, they're not self-directed, are 
0144
 1   looking for information about the bond, information about 
 2   where their account should be and how it should be 
 3   allocated between classes.  And so the role of the 
 4   registered representative that provides advice is key.  
 5   And so we have a series of requirements that are trying 
 6   to address that. 
 7             So the first is that the broker is supposed to 
 8   carefully -- and the firm -- carefully review what's 
 9   being sold so that they know the nature of the instrument 
10   and they've thought about whether this instrument -- 
11   there's at least someone in their clientele that this 
12   instrument is appropriate for.  And if they can't get to 
13   that level, then they should not be approving the 
14   security to go on their platform. 
15             But that's just the start.  And that requires 
16   looking into it, understanding that the different aspects 
17   of the bond or how the bond fund works and what its 
18   duration is, so that they're prepared to analyze it. 
19             Second comes, they're supposed to understand 
20   the customer and what the customer needs.  So that means 
21   they need to gather financial information about the 
22   customer, understand what their investment objectives 
23   are, and then they need to make a reasonable 
24   determination that if they're recommending a fund or a 
25   bond to that customer, that that bond is suitable for the 
0145
 1   customer, it's appropriate.  This is not a best interest 
 2   standard, this is not saying it's the best possible 
 3   instrument, it's just saying that this is in the zone 
 4   that's right for that customer. 
 5             So in order to be able to do that, there's a 
 6   number of things that are required.  So the broker is 
 7   required to, in a presentation to the customer, in 
 8   materials that they use, in materials that they provide, 
 9   the materials have to be fair and balanced.  If they are 
10   completely one sided, if you've got a sales script that's 
11   rah, rah, rah about this bond but doesn't mention that it 
12   has serious, serious illiquidity problems and risks under 
13   volatile markets, then those materials aren't fair and 
14   balanced.  Some of these materials have to be prefiled 
15   and precleared with our advertising department, who have 
16   come at this process with an eagle eye, trying to make 
17   sure that what's said is presenting the product fairly 
18   and adequately. 
19             Now, to do this, as Nick talked about, you have 
20   to train these registered representatives.  So we have 
21   tests that they have to pass, but that's just the 
22   beginning of the process.  The firm has to do continuing 
23   education each year, and they're expected to one way or 
24   another make sure that their advice providers know about 
25   the products, they know about these conditions so the 
0146
 1   customer may only retain 3 percent, but the broker is 
 2   supposed to know it.  And if they don't know it, then 
 3   they're not appropriately qualified in order to be making 
 4   recommendations. 
 5             And then finally, the firm has to adequately 
 6   supervise the brokers.  They have to have written 
 7   policies and procedures that are designed to make sure 
 8   that the products are being reviewed, that the 
 9   recommendation is suitable, that the broker knows what 
10   they're doing, and that they have ways to check that 
11   these things are actually happening.  And we examine and 
12   look for both problems with the broker but also whether 
13   they have the policies and procedures that work. 
14             Now, that goes for every product.  If you get 
15   into a complex product, then we expected heightened 
16   supervision.  We expect each of these factors to be done 
17   more deeply.  So if there is complexity, there need to be 
18   formal written procedures that talk about how those 
19   complex products get reviewed and what the process is.  
20   There has to be a careful control of what products get 
21   recommended to what customers.  And they have to think 
22   hard about which sort of customers are appropriate for 
23   the particular complex product.  And we expect there to 
24   be a consideration of whether there is a cheaper or a 
25   less complex product that would serve the same purpose 
0147
 1   for that customer. 
 2             So with that, I'll stop. 
 3             MS. GAINOR:  And I will just briefly note that, 
 4   you know, with respect to investment advisers, investment 
 5   advisers have a fiduciary obligation to act in the best 
 6   interest of their clients.  You know, today I am really 
 7   going to talk a little bit more about the rules that 
 8   would apply to the funds themselves and ensuring that the 
 9   funds and the disclosures that investors are receiving 
10   help to provide the investors with information about the 
11   risks and investments that are part of that fund. 
12             MR. MADHAVAN:  Great.  Gary, I think we will 
13   head over to you then.  I know the FINRA Foundation has 
14   done a ton of very interesting work on understanding how 
15   retail investors approach this complex marketplace, what 
16   they do actually follow and what they don't follow.  And 
17   I was wondering if you could share, you know, at a high 
18   level some of the findings of the FINRA Education 
19   Foundation, particularly vis-a-vis, you know, bond 
20   products. 
21             MR. MOTTOLA:  Sure.  Thanks for the opportunity 
22   to speak today. 
23             I will mention two studies that we have done 
24   over the last couple years that might shed some light on 
25   investor knowledge and investor behavior about bonds and 
0148
 1   a little about disclosure as well.  So as part of our 
 2   bond pricing transparency initiative, we did some testing 
 3   on how to communicate markup and markdown language on a 
 4   bond trade confirmation statement.  And we also explored 
 5   the knowledge of investors regarding bond investments.  
 6   And what we found was that investors really don't have a 
 7   high knowledge of bonds and bond terminology.  And we 
 8   talked to and surveyed people who were invested in 
 9   individual bonds and bond mutual funds. 
10             Just to give you an example, we found that less 
11   than half of bond investors could define what a markup 
12   was.  And nearly a third of bond investors thought that 
13   the term markdown meant that they were buying a bond at a 
14   discount. 
15             In addition, half of bond investors could 
16   define the term yield.  Only about a third could define 
17   the maturity date or call date.  And 20 percent admitted 
18   that they didn't even know what the term credit quality 
19   meant. 
20             So the knowledge is a little on the low side.  
21   But I would caution that just because investor knowledge 
22   around bonds is low doesn't necessarily mean disclosure 
23   can't help.  We'd be going beyond the data, at least the 
24   data that we polled in this study, because for this study 
25   we were just specifically looking at investor knowledge 
0149
 1   of bonds and how to communicate effectively.  And it's -- 
 2   we weren't looking at testing disclosure effectiveness.  
 3   And in fact, the work outside of FINRA done by academics 
 4   shows that disclosure and increasing transparency can 
 5   have benefits to the investor independent of investor 
 6   knowledge.  So just that one caution about a takeaway 
 7   from low investor knowledge and disclosure. 
 8             Another point I will make is we also did 
 9   research -- now this is not related to bonds, but I think 
10   it serves as a good example about educating investors and 
11   what they know and don't know and how to communicate with 
12   them.  We did some research on basically -- when a broker 
13   changes firms, they're given disclosures saying here's 
14   the issues that you need to consider when your investor 
15   changes firms.  And what we tested was how best to 
16   communicate that in a one-page document.  And what we did 
17   for this particular project is essentially FINRA staff 
18   gave the text of the disclosure or at least the important 
19   components of the disclosure to researchers and 
20   importantly communications professionals.  And then the 
21   communications professionals wrote the disclosure in 
22   plain language and then it was tested, it was refined 
23   based on the investor testing.  And what we found was 
24   that investors indeed found the more plain language, 
25   short, professionally designed disclosure more effective 
0150
 1   than traditional disclosure. 
 2             So I think approaching disclosure and 
 3   communication in education in this fashion is important. 
 4    Yes, there absolutely are inherent limitations to 
 5   investor testing.  It's expensive, it can be time 
 6   consuming.  But getting the voice of the investor in 
 7   terms of what they need to know and how to communicate 
 8   that is an important part of what we do at the FINRA 
 9   Investor Education Foundation. 
10             MR. MADHAVAN:  Great, that's very helpful.  I 
11   have a follow-on question.   
12             So in the previous panel, we heard from Matt 
13   Berger of Jane Street, who talked a lot about the need to 
14   educate investors about how best to trade, particularly 
15   around exchange-traded products.  Things like, you know, 
16   avoiding trading in less liquid after-hours markets and 
17   the use of limit orders versus market orders, those types 
18   of things.   
19             How does FINRA think about education in terms 
20   of trading for investors? 
21             MR. MOTTOLA:  That's an interesting question.  
22   About -- when you think of the investor population, 
23   essentially, you can break American households into a 
24   third of households have no investments whatsoever, 
25   retirement or taxable.  A third have taxable investments. 
0151
 1    And a third have retirement investments.  So really only 
 2   a third of households that really have taxable 
 3   investments that they are trading in regularly.  And I 
 4   think Nick had pointed out that some topics are quite 
 5   complex and can be very difficult to communicate.  
 6   Trading is one of them.  Risk tolerance, by the way, is 
 7   another one that we have great difficulty educating 
 8   investors on. 
 9             So we haven't done a tremendous amount of work 
10   in that space.  And the work we have done has indicated, 
11   similar to the work we've done in bond trading, that 
12   investors aren't particularly knowledgeable in that area. 
13             MR. MADHAVAN:  Okay, all right.  That's very 
14   helpful. 
15             I will turn back to Nick and actually to the 
16   entire panel.  So, Nick, you're a retail intermediary 
17   and, you know, Bob, you're a regulator.  Do you think 
18   there are aspects of retail -- of bond fund products that 
19   could be improved in order to provide a better retail 
20   experience and further protect retail investors that are 
21   looking to invest in the bond markets?  In particular, 
22   are there any improvements that are needed about fund 
23   naming and/or marketing about the different types of ETPs 
24   that are out there?  What would be your thoughts on 
25   those? 
0152
 1             MR. GOETZE:  Should I go first?  First, on the 
 2   naming side, you know, we have a phrase, "Know what you 
 3   own."  And I think that when you have names of funds like 
 4   the So-and-So Unconstrained Fund, which is really code 
 5   for the manager can buy whatever they want to, but it's 
 6   the Fixed Income Unconstrained Fund, that's a naming area 
 7   where I think we can do a better job letting the adviser 
 8   and the client know really what it is they're buying. 
 9             I think that we can -- I also think that the 
10   education that goes into around the features of the bond 
11   fund, whether it's the strategy that's being employed, 
12   whether it's the type of securities being bought I think 
13   that the up-front education, particularly to the 
14   financial advisers, I think we could improve the 
15   messaging coming from the fund families to the advisers, 
16   you know, whether it's in the national meetings that the 
17   firms have or whether it's the branch visits that they 
18   get from wholesalers.  I think that, you know, opportunity 
19   to educate and define the product better would serve 
20   people well. 
21             MR. MADHAVAN:  Melissa?  Bob? 
22             MS. GAINOR:  I can go next.  You know, I think 
23   it might help to provide a little bit of background 
24   regarding the rules that apply to labeling currently for 
25   funds.  You know, in addition to our disclosure rules, 
0153
 1   which are intended to provide investors with key 
 2   information about -- that are key to an investment 
 3   decision, the Investment Company Act also gives the 
 4   Commission rulemaking authority to find names or titles 
 5   that are materially deceptive or misleading for an 
 6   investment company to adopt as part of its name.  Really, 
 7   our concern that investors may focus on the name to 
 8   determine a fund's investments and risks. 
 9             So with this authority, the Commission adopted 
10   Rule 35d-1 in 2001.  It very generally requires a fund, 
11   including an ETF, with a name that suggests a focus in a 
12   particular type of investment or a particular industry, 
13   to invest at least 80 percent of their assets in that 
14   type of investment.  So, for example, a fund that has the 
15   word "bond" in its name has to have at least 80 percent 
16   of its net assets plus borrowings for investment purposes 
17   in bonds. 
18             The rule however doesn't incorporate certain 
19   things that could connote an investment strategy.  So, 
20   for example, it wouldn't apply to a fund that has 
21   "growth" or "income" or "total return" in its name, would 
22   apply to a fund that has "fixed income" in its name. 
23             So the rule generally establishes certain 
24   constraints on fund name but provides a good deal of 
25   flexibility.  And an investment company seeking maximum 
0154
 1   flexibility with respect to its investments is free to 
 2   select a name that doesn't connote a particular type of 
 3   investment emphasis. 
 4             Similarly, the rule allows funds to define 
 5   terms that are used in the fund name, so long as those 
 6   definitions are reasonable.  A fund that's an emerging 
 7   markets fund, for example, would define in its prospectus 
 8   what emerging markets would be.   
 9             And the rule generally requires investment 
10   companies to comply with this standard under normal 
11   circumstances.  This standard essentially permits 
12   investment companies to take temporary defensive 
13   positions, positions to avoid losses in response to 
14   adverse market, economic, political or other conditions 
15   and in other limited appropriate circumstances, such as 
16   unusually large cash inflows and outflows. 
17             It doesn't create a safe harbor, however.  So a 
18   name that can be materially deceptive or misleading, even 
19   if the fund complies with this requirement.  Index funds 
20   generally are expected to have more than 80 percent of 
21   their assets in investment connoted by the applicable 
22   index. 
23             And just to sort of close that out, there has 
24   also been some Staff guidance that particular relate to 
25   types of bond funds.  So a high-yield fund generally must 
0155
 1   have a policy that it would invest at least 80 percent of 
 2   its assets that are below investment grade.  One nuance 
 3   to note though is that a fund isn't required to have 
 4   high-yield in its name, even if it has a significant 
 5   emphasis in those types of securities. 
 6             So a fund could, for example, have a broad 
 7   investment purview suggested by its name, like a fixed 
 8   income fund, and have a significant asset allocation to 
 9   high-yield securities.  This, of course, should be set 
10   forth in the fund's prospectus. 
11             I would note that the rules are designed to 
12   address names that are likely to mislead an investor 
13   about a company's investment emphasis.  But it is 
14   important to note that an investor shouldn't really rely 
15   solely on the name as the sole source of information 
16   about a fund. 
17             Now, with respect to the particular question 
18   about ETPs versus ETFs, it's really something we are 
19   thinking about.  Today, the term ETF is used to describe 
20   lots of different investment companies with different 
21   investment strategies, as well as a number of products 
22   that aren't investment companies or even funds.  We see 
23   in news reporting that, you know, especially in February, 
24   that the term ETF is used to describe a number of 
25   different types of ETPs. 
0156
 1             Director of Investment Management, Director 
 2   Blass, has noted that the naming and marketing of these 
 3   products has sometimes failed to draw clear distinctions 
 4   among these products.  And investors may have to work 
 5   harder to identify important differences and risks.  And 
 6   so in this regard, the Division of Investment Management 
 7   is welcoming thoughts from investors, funds, others about 
 8   whether addressing ETP nomenclature would be helpful to 
 9   investors and the markets.  We're interested in hearing 
10   whether the differences in risks, investment strategies 
11   and investor protections among ETFs, commodity pools and 
12   other exchange-traded notes are clear when the term ETF 
13   is often used for all three. 
14             Do, for example, people assume, when they see 
15   ETF, that they're looking at a fund that is registered 
16   under the Investment Company Act?  These are all things 
17   that we're really interested in getting more feedback on 
18   from everybody in the industry.  And if anybody in this 
19   room has some thoughts, we'd also like to hear from you. 
20             MR. MADHAVAN:  Okay, thanks.  I guess we want 
21   to make sure there's time for questions from the 
22   audience.  So let me just close it out with maybe a 
23   question for Gary. 
24             So, Gary, the FINRA Foundation's work on the 
25   status of investor education has been truly enlightening. 
0157
 1    The question then is, how do we continue that education 
 2   effort in your view?  How do we best communicate?  How do 
 3   we actually move the dial with the average investor? 
 4             MR. MOTTOLA:  Yeah, moving the dial is 
 5   difficult.  You probably got a sense of that from this 
 6   panel.  Educating investors is tough. 
 7             What we do know is that information needs to be 
 8   presented frequently and needs to be presented at the 
 9   right time, at the time the investor is making a 
10   decision.  And it needs to be presented in a concise 
11   manner, in plain language.  So those are some of the key 
12   elements that you need to address. 
13             But it's going to be a challenge.  And I don't 
14   mean to imply that it's easy to move the dial, because it 
15   can be very, very difficult to actually educate investors 
16   on particularly important topics or complex topics.  But 
17   basically, it's just a matter of using, say, employer-
18   provided retirement plans as a channel for investor 
19   education, using -- some investors' education actually 
20   takes place in high schools these days, in colleges, as 
21   well as all the agencies, federal agencies and nonprofits 
22   that have all sorts of online educational tools for 
23   investors as well. 
24             MR. MADHAVAN:  Okay.  Big, big challenge ahead. 
25             I think with that, we've got time for some 
0158
 1   questions from FIMSAC members.  I see Professor Harris 
 2   has his -- is following protocol there. 
 3             MR. HARRIS:  Thank you, Ananth. 
 4             Gary, I have two related questions for you.  
 5   And if you have solid results, delighted to hear it.  If 
 6   not, I'd be real curious about your professional opinion. 
 7             So the first one is, what fraction of retail 
 8   investors do you think can distinguish between brokers 
 9   and dealers?  And a second related question is what 
10   fraction of investors know that their brokers are trading 
11   with them as dealers? 
12             MR. MOTTOLA:  We have some data on the first 
13   question.  We have asked in the past if investors could 
14   distinguish between brokers and advisers.  And for the 
15   most part, they can't.  In fact, we've kind of shifted in 
16   our survey methods, we've shifted from approaching the 
17   very question that way.  We just talk about a generic -- 
18   use the generic term "adviser" as opposed to a broker-
19   dealer or an adviser. 
20             So, yeah, I think your intuition is correct 
21   that investors have a hard time distinguishing between 
22   that. 
23             MR. HARRIS:  And strictly speaking, my question 
24   was about the difference between brokers and dealers as 
25   opposed to advisers.  So when you say adviser, what do 
0159
 1   you mean? 
 2             MR. MOTTOLA:  Oh, okay.  I'm sorry.  Very 
 3   similar.  They have a hard time distinguishing between 
 4   brokers and dealers as well.  I thought you were talking 
 5   about broker-dealers and advisers. 
 6             MR. HARRIS:  Right, right.  May I ask another 
 7   question? 
 8             MR. MADHAVAN:  Go ahead. 
 9             MR. HARRIS:  Sure.  Melissa, we know that bond 
10   funds, like other funds, can augment their returns by 
11   selling out-of-the-money calls and puts.  Under normal 
12   circumstances, that will usually allow them to beat their 
13   benchmark and most of their competitors.  But under 
14   extraordinary circumstances, it can lead to disastrous 
15   results. 
16             I was wondering what regulations prevent that 
17   and -- doesn't really prevent it, but how do we disclose 
18   it?  And do you think the disclosure standards are 
19   adequate in this area? 
20             MS. GAINOR:  Well, one would expect the fund to 
21   disclose.  If it's a principal investment strategy, it 
22   should be disclosed within the summary prospectus, the 
23   first four pages of the prospectus.  One of the 
24   undertakings that the Division is currently undergoing is 
25   this analysis of retail investor experience and whether 
0160
 1   they understand enough about the disclosures that they're 
 2   getting from the prospectus and from other disclosure 
 3   documents.  So it will be interesting to see, as that 
 4   project moves forward, how those particular issues get 
 5   addressed. 
 6             MR. HARRIS:  And so how do we define what a 
 7   principal investment strategy is?  So, for instance, a 
 8   fund such as I just described could be 95 percent 
 9   invested, say, just in an index.  But the options 
10   contracts that I just described, representing less than 5 
11   percent, could still be a dominant determiner of 
12   performance, but only once every 15 years. 
13             MS. GAINOR:  So very generally, there is Staff 
14   guidance out there that says a principal investment 
15   strategy is 5 percent or more.  But, so to the extent 
16   that it's not set forth within that summary prospectus, 
17   it should be within the broader discussion of the 
18   investments of the fund within the prospectus and the 
19   SAI. 
20             MR. MADHAVAN:  Any other questions from anybody 
21   on FIMSAC?  Otherwise, I will thank our panelists and 
22   turn it over to Mike. 
23             CHAIRMAN HEANEY:  I just want to, the same, 
24   thank you all for taking the time.  And interesting 
25   panel, for sure.  Ananth, thank you for chairing the 
0161
 1   subcommittee. 
 2             We are right on time at 2:15 for our final 
 3   break before we head into the final panel, which Rick 
 4   will chair in 15 minutes, so back at 2:30, please. 
 5             (Recess.) 
 6             CHAIRMAN HEANEY:  All right, so for the last 
 7   panel of the day, we will be discussing electronic 
 8   trading in the retail market.  I will turn it over to 
 9   Rick McVey and again thank Rick for chairing this 
10   subcommittee of technology and electronic trading.  And 
11   again, promises to be an interesting panel for sure.  And 
12   I again would like to also thank the panelists for 
13   participating.  Rick. 
14             MR. McVEY:  Thank you, Michael. 
15             The E Trading and Technology Committee met four 
16   times during the quarter.  The committee prioritized 
17   electronic trading in the retail market first and invited 
18   industry experts to join our calls.  We explored 
19   important issues, including how technology is improving 
20   price transparency and bond choice for retail investors, 
21   the expansion of fixed income retail E trading on ATS 
22   venues, ways to increase market access and availability 
23   of data for retail investors, best execution challenges 
24   for retail brokerage firms operating on ATS venues 
25   particularly in the context of posting limit orders, and 
0162
 1   protocols such as last look that may inhibit fair and 
 2   competitive electronic markets. 
 3             The committee invited outside guests from TMC, 
 4   Wells Capital Advisors, Charles Schwab, FINRA, Fidelity 
 5   Capital Markets and Wilmington Trust to provide input and 
 6   advice to the committee on opportunities and challenges 
 7   in today's retail E trading market.  While we do not have 
 8   any specific recommendations for the full FIMSAC 
 9   committee today, we do believe our committee is much 
10   better informed on the evolving market structure in 
11   retail trading and our work will continue in the quarters 
12   ahead. 
13             Notably, the subcommittee spent time 
14   considering whether the retail market would benefit from 
15   any central posting or national best bid and offer 
16   mechanism.  And most members believe that the retail 
17   brokers and ATS systems were already sufficiently 
18   aggregating the available data for the benefit of the 
19   retail participant. 
20             In addition, a very small percentage of 
21   outstanding issues have live two-sided markets during the 
22   trading day.  Thus, for the time being, the subcommittee 
23   does not consider building out this infrastructure to be 
24   a cost-effective way of improving pretrade transparency 
25   for the retail market. 
0163
 1             The committee is now turning its attention to 
 2   the institutional market.  We are beginning to explore 
 3   issues and opportunities including improvements to market 
 4   liquidity, industry standards for all E trading venues, 
 5   trading protocols and impartial access to bond reference 
 6   data. 
 7             Today, we have a panel of retail market experts 
 8   to share their views on important matters in the 
 9   electronic trading markets for retail investors.  I would 
10   like to join Michael in thanking the panelists for 
11   participating not only today but on many of the calls 
12   that we have had on the retail market to date. 
13             Matt Andresen is the founder and CEO of 
14   Headlands Technologies, a global quantitative proprietary 
15   trading firm and a committee member of FIMSAC.  Bob Colby 
16   is the chief legal officer of FINRA and is playing a 
17   double header this afternoon, after appearing on the last 
18   panel.  Renzo Iturrino is the head of electronic fixed 
19   income product development and trading for Fidelity 
20   Capital Markets.  Kristin Maher is a managing director at 
21   Wells Fargo Advisors and is the head of fixed income 
22   services in capital markets trading.  And last but not 
23   least, Tom Vales is chairman and CEO of TMC, one of the 
24   leading electronic trading venues in the retail fixed 
25   income space. 
0164
 1             So rather than doing introductory comments 
 2   across the panel today, we agreed we would just dive into 
 3   the topics that we have been debating with the committee. 
 4    And we hope you find it informative.  And we do plan to 
 5   leave time at the end of the session today for questions 
 6   from the full FIMSAC committee. 
 7             So Kristin, if you don't mind, maybe we could 
 8   start with you on how transparency has improved the 
 9   information available to retail investors and how you go 
10   about the process of aggregating that data both across 
11   corporate bond and municipal bond trading. 
12             MS. MAHER:  Just a brief on my background.  I 
13   have spent the last 25 years in retail fixed income 
14   trading.  And from my perspective, the data and 
15   technology that's available today has transformed the 
16   retail client experience.  The ATSs play a critical role 
17   in our both offering and bid side liquidity and it is a 
18   very important part of the daily -- a component of our 
19   daily business. 
20             So we execute 4,000 to 5,000 retail customer 
21   orders.  On average, the size could be 25,000 face 
22   amount.  And those are all with our retail customers in 
23   the fixed income markets.  And a significant component of 
24   those trades depend on the liquidity which is efficiently 
25   accessed through the ATSs.  So without that in today's 
0165
 1   market, it would be hard to provide that many retail 
 2   executions. 
 3             And when I think about the TRACE and MSRB data, 
 4   it plays an important role for all of us that make up 
 5   this market.  It's very important for our clients as they 
 6   are looking at the investment and they are determining 
 7   the price that they're paying for their security and 
 8   whether it's fair.  And for our financial advisers, when 
 9   they do a search, they could have 40,000 line items to 
10   choose from for a particular asset class.  So in 
11   assessing which security to purchase for their clients, 
12   they also depend on the data when they get to a certain 
13   level in that analysis. 
14             And then for our own trading teams, we have 
15   best execution responsibility for all of our executions 
16   with our retail clients, whether or not we are the 
17   liquidity provider in the execution, and we depend on the 
18   data that's available in the marketplace to perform that 
19   analysis.  So important for everyone. 
20             We feed the ATS to our retail adviser desktop. 
21    And we allow -- our financial advisers get to see a full 
22   display of offerings when they are making decisions for 
23   their customers.  So they are seeing sometimes 40,000 or 
24   50,000 line items in an asset class.  And equally, on the 
25   bid side, when our clients are looking to sell, we submit 
0166
 1   it in a competitive process out to the ATS and the best 
 2   bid goes back to our retail customer.  So it has been 
 3   critical and transformative. 
 4             MR. McVEY:  And, Kristin, if you had to think 
 5   back to four or five years ago in terms of how many 
 6   offerings might have been available to your retail 
 7   customers before this growth in ATS trading, what might 
 8   that have looked like then? 
 9             MS. MAHER:  I think it was a couple hundred and 
10   they all had to be typed on the screen.  So technology 
11   has certainly played the role.  But to access liquidity 
12   from another broker-dealer, it was a lot of voice, 
13   brokers, brokers, and it was just a very, very bulky 
14   process and one that couldn't be performed on a per 
15   transactional basis like it can today. 
16             MR. McVEY:  Great, thank you. 
17             And Renzo, maybe you can comment from Fidelity 
18   Capital Markets on the process that you go through to 
19   aggregate price information on behalf of your clients? 
20             MR. ITURRINO:  Sure.  So, first of all, thank 
21   you to the Commission and thank you to the FIMSAC for 
22   having me here and discussing this important topic.  By 
23   way of background again, my name is Renzo Iturrino and, 
24   in my role, I'm responsible for fixed income trading 
25   infrastructure, electronic product development efforts 
0167
 1   and our approach to fixed income market access for the 
 2   benefit of Fidelity's retail customers.   
 3             So just to lay the framework on Fidelity's 
 4   disposition on the issues, with regards to fixed income 
 5   market structure, Fidelity has buy side and sell side 
 6   expertise as a provider of financial services to millions 
 7   of retail customers.  While we're a significant 
 8   participant in the markets, we are not an investment 
 9   bank, nor do we have a large proprietary trading desk or 
10   market making business.  So that's our general 
11   disposition. 
12             As it relates to how we aggregate markets for 
13   the benefit of our retail investors, we do this through 
14   electronic connectivity into a proprietary aggregation 
15   infrastructure and platform.  So we connect to some of 
16   the major ATSs, and also we have strategic direct 
17   connectivity with some dealers that provide significant 
18   liquidity and potentially price improvements for our 
19   customers. 
20             We look at our architecture and we call it an 
21   open architecture, or we deploy an open architecture 
22   model.  It is an environment designed to put the 
23   interests of our customers first.  So we permission north 
24   of 200 dealers to have access to our flow.  We have north 
25   of 115,000 live items at any given point, give or take, 
0168
 1   depending on the day.  Our view is that by deploying our 
 2   resources and focus towards scalable technology and 
 3   infrastructure, it is possible to provide a competitive 
 4   market that retail can interact with.  We believe that a 
 5   competitive marketplace is an adequate framework for the 
 6   retail investor. 
 7             Today, we have achieved depth in approximately 
 8   80 percent of our corporate and 60 percent of our 
 9   municipal line items, which substantiates our engagement 
10   efforts.  And by depth, I mean we have different prices 
11   on the same quote and we're able to rank those prices and 
12   allow retail to interact with the top of book. 
13             We augment that competitive market with 
14   important standards and processes internally.  And this 
15   is very important.  We present the market as we see it 
16   from our liquidity providers to our customers.  We do not 
17   mark up the bonds for the initial display to our 
18   customers.  So as we aggregate, we present that market as 
19   we see it.  And we have, like I said, internal ranking 
20   mechanisms. 
21             So, electronic trading, the growth of 
22   electronic trading, the development of technology 
23   particularly from an infrastructure and platforming 
24   perspective has been a tremendous asset for us in our 
25   approach toward fixed income. 
0169
 1             MR. McVEY:  Thank you. 
 2             Tom, maybe you can comment, too -- you've been 
 3   in the ATS retail space from the beginning -- about the 
 4   growth that you've seen in available quotes and breadth 
 5   of bond offerings that you see on your system and taking 
 6   place more broadly on ATS. 
 7             MR. VALES:  Sure, thank you.  For many of you 
 8   that don't know, TMC used to be known as the Muni Center. 
 9    And so we used to joke around that whenever we were 
10   front of a municipal group, we'd advertise ourselves as 
11   the Muni Center, and TMC Bonds in front of a taxable 
12   mixed muni group as treasuries, munis and corporates.  
13   And so that's generally the background. 
14             In just listening to Kristin and Renzo, there's 
15   two things that might not be as apparent.  And that is 
16   when transparency went live, most of the broker-dealer 
17   firms initially didn't make that available to their 
18   adviser networks, meaning advisers or brokers.  It took a 
19   little bit of a transition period. 
20             Fast forward to today, we have approximately 60 
21   firms that have private labeled our application.  Every 
22   single one of them makes a realtime or a delayed price 
23   feed available to their network.  And that's a 
24   significant change. 
25             Firms like Fidelity in their online brokerage 
0170
 1   unit have been taking that information and passing it 
 2   down to retail.  And so for years, I've heard this theme 
 3   or this narrative of how much information is retail 
 4   getting?  They are actually getting it then from two 
 5   perspectives.  One, they're getting it directly if 
 6   they're with an online firm.  But, secondly, they're 
 7   getting it indirectly now through their brokers and 
 8   through their traders having access to this information. 
 9    And, more importantly, having it delivered in a context 
10   that's easily analyzed.  And I think that's really 
11   important to keep in perspective. 
12             In terms of growth in the market, as most of 
13   you know, both TRACE and MSRB required an ATS indicator 
14   to be included I think it was about a year and a half ago 
15   for trade reporting.  And while we haven't had the 
16   opportunity to see the data from TRACE yet, the MSRB had 
17   sent out a report.  And it looks like approximately 25 to 
18   30 percent of the market now is trading through the ATSs. 
19    And that's quite significant, especially if you were to 
20   put that in context of the odd lot market.  If you look 
21   at the odd lot market, it looks like approximately 60 
22   percent of the market now is trading -- for interdealer 
23   trades, is trading through the ATSs. 
24             And so the fragmentation that used to exist in 
25   the marketplace has really been centralized and these 
0171
 1   ATSs have really kind of filled the void of an exchange. 
 2             Just to give you a few more numbers to kind of 
 3   put into context, at TMC, we trade in munis with about 
 4   318 different firms a day.  We will have close to 700 
 5   different firms logging into our marketplace. 
 6             It was interesting.  I looked at the blue list 
 7   the last day that it was in print, which was August 16 to 
 8   2001, and there were 450 million par amount of bonds 
 9   offered.  Last week, we had 18.8 billion in munis 
10   offered, so it was 102,000 line items.  So the amount of 
11   content that is now available to the market and the 
12   amount of information is just phenomenal. 
13             MR. McVEY:  Thank you.  And, Matt, maybe for a 
14   slightly different angle, you seem to be one of those 
15   rare birds that keeps moving down in market liquidity 
16   through your career, having started in equities --  
17             MR. ANDRESEN:  It's been a long journey to the 
18   middle. 
19             (Laughter.) 
20             MR. McVEY:  -- moved on to treasuries and now 
21   landing in municipal bonds.  But maybe you can comment a 
22   little bit on both how important transparency is to your 
23   world in being a relatively new market maker in the 
24   municipal bond market and also the role that the ATS 
25   systems play in allowing you to compete as a market 
0172
 1   maker. 
 2             MR. ANDRESEN:  Sure.  So for context, you are 
 3   correct that munis or many of the challenges that we all 
 4   associate with fixed income, the huge number of 
 5   individual issuers and issues and the relatively small 
 6   number of trades are all on steroids when it comes to the 
 7   municipal markets.  There are about a million different 
 8   individual CUSIPs but only about 40,000 trades a day. 
 9             We've been trading munis since 2014 and we 
10   couldn't have and wouldn't have thought to enter this 
11   market were it not for a few of the innovations that we 
12   just heard about first, having the trade feed.  Without 
13   that information, it wouldn't be practical for someone to 
14   be able to model all the prices that we do on a daily 
15   basis.  So that was sort of the regulatory difference. 
16             And then there was a market-based difference 
17   because you had these ATS platforms, the alternative 
18   trading systems like TMC, that popped up and started 
19   providing a platform for people to meet.  Because really, 
20   you're only ever going to be able to trade by appointment 
21   in something that has a million issues and 40,000 trades. 
22    So if you have a more efficient appointment scheduler 
23   who can bring people together, that can solicit interest, 
24   you actually can have a very vibrant market. 
25             And we were actually surprised when we got into 
0173
 1   the market that it was as competitive as it was.  We were 
 2   expecting, because it's so hard to model these bonds, 
 3   that we would -- once we went into one of these auctions 
 4   that solicits interest in municipals, that we'd probably 
 5   be alone a lot of the time.  And, in fact, we see an 
 6   average of seven bidders in these auctions.  So these ATS 
 7   platforms provide excellent transparency for the end 
 8   customer, because they get to interact with a lot of 
 9   people that would never be standing bids and offers but 
10   can be trading by appointment. 
11             So that has helped us grow our business.  We 
12   have approximately $400 million of inventory that we show 
13   out, about 13,000 line items every day in munis.  And we 
14   are quite bullish over what's going to happen over the 
15   next couple of years. 
16             You know, if you look back in December 2015, 
17   there were only 7,000 ATS auctions for municipals.  And 
18   in December just past, it was 20,000.  So the market has 
19   exploded in only three years.  And I think as we look 
20   forward, I think we're going to have -- should have a 
21   catalyst for a lot better prices and a lot more 
22   competition for these individual auctions. 
23             MR. McVEY:  And maybe you can expand on that.  
24   And I know others have views and comments on the 
25   different protocols at work because in terms of the live 
0174
 1   markets it does seem to generally be more heavily 
 2   weighted toward offer wanted.  But you've talked a lot 
 3   about the auction process and how much more efficient 
 4   that's gotten over the years, in terms of generating best 
 5   bids for retail orders as well.  So maybe you can kick 
 6   that off and others can follow. 
 7             MR. ANDRESEN:  Sure.  So generally, because 
 8   again there aren't really standing bids out in the 
 9   municipal market, when someone looks to sell a bond, 
10   their broker will typically initiate an auction.  Which 
11   is just saying, hey, this thing hasn't traded in six 
12   months, would anyone like to bid for it.  And they put 
13   out the solicitation.  And typically in munis, it's about 
14   four hours is the window that you have to collect a bid. 
15    So just like if you were on eBay and you had a four-hour 
16   auction.  You are hoping that someone will go on there 
17   and buy the tchotchke or whatever. 
18             In municipals, you have about a four-hour 
19   window to collect these bids and then, at the end of 
20   that, there's a winner.  And this is where -- you know, 
21   I've been very sort of optimistic so far in my comment, 
22   this is where the bad news is -- often, there's -- 
23   winning doesn't mean what it means on eBay. 
24             When you win in municipals, we win a lot of the 
25   time, but of those times that we do win, we only actually 
0175
 1   trade about a third of the time.  And so two thirds of 
 2   the time that we're actually winning an auction, we're 
 3   not actually transacting. 
 4             MR. McVEY:  So you're welching the rest of the 
 5   time? 
 6             MR. ANDRESEN:  No, we would trade every one of 
 7   them.  What happens is about a third of that time, no 
 8   trade ends up even happening.  And so some percentage of 
 9   that time, that's legitimate.  Somebody wanted a price, 
10   they didn't like the price, they decided not to trade.  
11   But other times, we suspect that people were just using 
12   the auctions to try to see what the market was out there 
13   to have a better idea of what they held was worth without 
14   a real intent to trade. 
15             But then the part that is more problematic is 
16   about a quarter of the time that we win, it trades away. 
17    So we won the auction but we don't get the trade.  It 
18   actually just trades to someone -- the person that 
19   initiated the auction who didn't participate in the 
20   auction.  They waited, got our price and then matched it 
21   away.  And, it would be humorous if it wasn't money, 11 
22   percent of the time, we actually just get traded through. 
23    So they generate the auction, we win the auction and 
24   then, through the winner's curse, you know, we won it at 
25   let's say 100, it then prints at 99, as an example. 
0176
 1             So there are real opportunities for improvement 
 2   in the marketplace.  So we have, I think, excellent 
 3   structures that, when working well, do generate 
 4   competitive auctions, that do generate good prices, but 
 5   we still have a lot of wood to chop to make these 
 6   auctions as efficient as possible in generating rewards 
 7   for those that bid and win.  So we have a better 
 8   incentive scheme. 
 9             MR. McVEY:  Thank you.  And, Kristin, I know 
10   you have some perspective on the orders that do not 
11   trade.  So maybe you could follow on on Matt's comments 
12   there. 
13             MS. MAHER:  Yes.  So our experience with our 
14   customers is, where price discovery is required, so in 
15   the municipal asset class more important than maybe in 
16   corporates, our customers will look at a portfolio and in 
17   the discussion with their financial adviser, they will 
18   come to the market for four bids, when they intend just 
19   to sell two of the securities. 
20             So our own statistics show that 40 percent of 
21   the time after a bid wanted, our customers actually sell. 
22    And it's interesting that in the corporate space, where 
23   there is more quoted markets, that percentage is much 
24   higher.  So the requirement of price discovery certainly 
25   leads to certain retail behaviors. 
0177
 1             MR. VALES:  Let me just chime in on that, as 
 2   well.  So it's interesting.  If you take a look at the 
 3   bid won at auctions, there are a couple things.  One, one 
 4   of the unintended consequences of best ex was that market 
 5   participants felt to comply that it was conservative for 
 6   them to put the bid wanted out in multiple places.  And 
 7   so part of the growth in the industry has been 
 8   duplication, as opposed to unique bid wanteds.   
 9             Having said that, clearly there is a part of 
10   the industry too that, for best ex, if you're not 
11   familiar with it, in the muni world and in the corporate 
12   world, it's not about best price but it's about exposing 
13   an order to a competitive process so that there is a 
14   portion of that market that I'm going to say, let's say, 
15   20 to 30 percent of it that is growth from now opening up 
16   orders to a competitive process. 
17             To the point that both Matt and Kristin are 
18   making in terms of the bid wanteds is that munis are 
19   typically just much more difficult to assess in terms of 
20   their value.  So this is interesting. 
21             I had looked at the statistics of, well, if 
22   that's the case then, what we should see is that during 
23   an auction then, if it's really that much more difficult 
24   to ascertain the price of a municipal even after getting 
25   the bid, right, then you should see a larger time horizon 
0178
 1   or time frame for that decision to be made. 
 2             So in corporates on bid wanteds, 90 percent of 
 3   the time an auction decision is made within 30 minutes.  
 4   In munis, in 30 minutes, 2 and a half percent of the 
 5   time, an auction decision is made.  And so you can see, 
 6   when you put these items out, it's not just a question of 
 7   getting the price, but then it's also a process of 
 8   looking at the information and, to Kristin's point, you 
 9   might even have to compare four or five different assets 
10   to see which ones are really on market and which ones are 
11   the most appropriate ones to sell. 
12             MR. ITURRINO:  I think -- so, you know, I agree 
13   actually with all points that have been made, certainly 
14   by Kristin and Tom.  And we do have a process in place at 
15   Fidelity that is yielding us very positive results as it 
16   relates to, in this case in particular, municipal 
17   auctions.   
18             So first and foremost, our slots are open to 
19   taking bids from any dealers out there.  So, however, in 
20   the case of munis, we recognize that that's not 
21   practical, given the number of CUSIPs, so on and so 
22   forth.  We have found the auction process to be fairly 
23   orderly and robust.  And we have deployed sort of an 
24   electronic approach into managing the auction process.  
25   So we are a firm that reaches out to multiple ATSs and 
0179
 1   destinations with our municipal bid wanted flow. 
 2             Those bids are obviously collected under a 
 3   prescribed auction period, as we've all alluded to.  And 
 4   at the end of the auction, the best bid wins the auction 
 5   and it is ripe to receive the order.  So hopefully, Matt, 
 6   you don't see any trade-throughs in our destination. 
 7             That being said, there is an argument out there 
 8   obviously about false liquidity and things of that 
 9   nature.  We don't see that evidenced in the data.  So 
10   today, 97 percent of our auctions receive a bid.  We 
11   average, for municipal bid wanteds, approximately 20 bids 
12   per auction.  And no, most of those are not duplicates.  
13   About 16 to 17 of them are unique, meaning that we feel 
14   we're reaching out to the market and getting a fair 
15   market for our retail investors. 
16             Of those bids that get presented back to the 
17   customer with the option to interact with the top of 
18   book, 66 percent of them end up trading.  And obviously, 
19   the fill rate at that point is 99 percent. 
20             So why am I sharing all of this?  I do believe 
21   that there are some challenges in the auction process.  
22   There are some inefficiencies, perhaps, inherent in it.  
23   But it is possible to yield positive results from retail. 
24             MR. VALES:  Just a question I had.  How do you 
25   determine what your collection period should be by item? 
0180
 1             MR. ITURRINO:  By item or by asset class? 
 2             MR. VALES:  Yeah, by asset class.  Do you say 
 3   that you need at least an hour to work on it or for the 
 4   collection period, and then how do you reach back out to 
 5   retail to let them know what the results are? 
 6             MR. ITURRINO:  So we do that also 
 7   electronically through electronic notification as soon as 
 8   the auction is over.  But as far as how we determine what 
 9   the current time should be, we study how bids come in.  
10   So there are certain protocols, right?  So, Tom, you send 
11   us all the bids at the end of the current time.  Others 
12   send us the bids as they're coming in.  And so we study 
13   those time frames, essentially, right?  We study the data 
14   that we get or that we see and that's how we determine 
15   should it be an hour, should it be 90 minutes, should it 
16   be 30 minutes?  And it definitely varies by asset class. 
17             We also take into consideration the retail 
18   experience.  So it is an online experience, by and large, 
19   and they are not in front of their computers all day.  
20   And so there is that element of consideration.  So that's 
21   how we determine, that's all. 
22             MR. McVEY:  And on that last topic of those 
23   that might be around their computers all day, we talked 
24   with the subcommittee about the ability for retail 
25   clients to leave limit orders through brokerage firms to 
0181
 1   be worked on the ATS venues.  And you both had experience 
 2   and perspective on that that I think would be valuable to 
 3   share. 
 4             And, Bob, this is also where you will come in. 
 5    Because the broker-dealer best execution rules get to be 
 6   complicated with securities that trade infrequently and 
 7   limit orders and how that broker-dealer has determined 
 8   it's the best price for the client.  So can you talk a 
 9   little bit about that issue and how you both think about 
10   that? 
11             MS. MAHER:  Just to add one thing to the prior 
12   discussion.  I mean, I think sometimes it's hard to 
13   identify between the various ATSs if you're actually 
14   getting unique liquidity or not.  I mean, I've had 
15   situations where I even got different prices from 
16   different ATSs and it turned out to be the same liquidity 
17   being provided, so I think that's a tough analysis. 
18             As far as limit orders, currently our customers 
19   are not asking us.  And I think we had a little bit of a 
20   talk of whether that's the chicken or the egg.  But if it 
21   was available, would they use it more? 
22             I think the challenge that we see is that our 
23   customers would have prices in mind that they would like. 
24    And how do we meet our best execution responsibilities 
25   if they leave that order out in the marketplace and then 
0182
 1   they come back to execute it.  You know, was it fair and 
 2   reasonable at that time. 
 3             And so with the current tools and technology 
 4   that's available in the current regulatory structure -- 
 5             MR. McVEY:  So your practice then generally is 
 6   the ATS systems have plenty of live and competitive 
 7   offerings.  The bid side is usually conducted through 
 8   some form of an auction.  But typically, you're not 
 9   accepting direct limit orders from clients to be posted 
10   in the systems? 
11             MS. MAHER:  Correct. 
12             MR. McVEY:  Correct. 
13             MR. ITURRINO:  Yeah, so as far as limit orders 
14   are concerned, I will stick to the strict definition of a 
15   limit order.  We do allow, within constraints, for our 
16   customers to put in an order that might be a little off 
17   the quote, if you will.  And so they see a quote and they 
18   say, you know, I want to try to maybe buy these bonds a 
19   little cheaper.  We allow for that to happen under a 
20   specific period of time.  And the one that the order is 
21   routed to, they can either accept or decline that order. 
22    And then we get notified of that. 
23             MR. HARRIS:  How long is the time period? 
24             MR. ITURRINO:  So it depends, right?  So it can 
25   be 90 seconds, up to 20 minutes. 
0183
 1             MR. HARRIS:  And you require full execution or 
 2   partial? 
 3             MR. ITURRINO:  Full. 
 4             MR. McVEY:  And, Bob, you were kind enough to 
 5   join our calls.  But I think you provided the FINRA 
 6   regulatory perspective on this issue, which I think, Tom, 
 7   I'm right in saying of the 1.1 million or 1.2 million 
 8   muni CUSIPs, you have a market in about 80,000 on the 
 9   offered side and 8,000 or 10,000 on the bid side.  So 
10   it's a small subset.  But you provided the regulatory 
11   context on the broker-dealer best execution rule, and I 
12   think that would be helpful to the whole committee as 
13   well. 
14             MR. COLBY:  So limit orders are a device that 
15   were very common in the equity world, but they are not 
16   unique to equity active trading.  So there's been a 
17   progression on how limit orders get addressed in the 
18   equity world that transfers over to the fixed income 
19   world. 
20             So a limit order is a type of order that you 
21   don't have to offer to customers.  But if you do, then 
22   there are certain obligations that go with it.  And you 
23   have different set of obligations if a customer comes to 
24   you and says, I want to enter this order at this price at 
25   this time.  Then, you accept it, then you have to 
0184
 1   represent it as requested for the time period that you 
 2   agreed.  And if the market moves and the customer has 
 3   chosen that, then that's the customer's responsibility. 
 4             If you recommend a limit order in a certain 
 5   situation, then you need to be able to explain to the 
 6   customer why you're recommending it for them, how it is 
 7   going to work in the market, so that the customer 
 8   understands what the recommendation means and what you're 
 9   going to do in the context of that limit order. 
10             And there are some default conditions, right?  
11   So it developed that a limit order is a particular type 
12   of agency order that comes with agency duties, and that 
13   you have to handle it in a specific way, in accordance 
14   with those agency duties.  For instance, you can't trade 
15   ahead of it.  As a fiduciary, you can't do that.  And 
16   there was a lot of development of that in the equity 
17   markets as we went along.  But it's established policy.  
18   If you accept it, you can't trade ahead of it at the same 
19   or a better price without filling your customer. 
20             Now, this is tricky, because you have to take 
21   these conditions and apply them, you know, in an evolving 
22   electronic market.  So I'll stop there.  That's with 
23   respect to limit orders. 
24             Is that sufficient? 
25             MR. VALES:  Just a couple other I think maybe 
0185
 1   interesting points.  One is, if you're comparing the 
 2   corporate market against the municipal market, you might 
 3   be surprised to hear that in corporates, there are 
 4   actually more quoted bid sides in a collab type of 
 5   marketplace then there are offers.  And so we run almost 
 6   two to one in terms of number of bids that we have in 
 7   credit versus the offered side. 
 8             To Rick's point earlier, we generally run 
 9   somewhere around 5 to 7 percent of the munis have quoted 
10   bid sides.  And typically those munis are the new-issue 
11   munis.  They are new-issue munis where there is a 
12   syndicate that's backing them and there are depths to the 
13   market and they are supporting the deal when it comes to 
14   the market.  So it's a very different environment. 
15             One of the reasons and one of the major reasons 
16   why you can look at it and say, well, would limit orders 
17   be helpful for retail, you would also ask, well, then why 
18   aren't more dealers participating with limit orders in 
19   the muni market?  And the reason for that is that in 
20   munis, if you think that you want to buy something, you 
21   can post a bid.  But you can't post a bid -- it's very 
22   difficult in a market where there's a million to 2 
23   million different CUSIPs, it's very difficult to post a 
24   bid and then hope that somebody would find it.  And so 
25   then even within that, then you have to think about 
0186
 1   quantity constraints and then constantly updating your 
 2   price. 
 3             That's the other part, is that the easier part 
 4   about munis is they tend to update much less frequently 
 5   than credit.  We do on an average day about 60- to 70 
 6   million price updates; 90 percent of those price updates 
 7   are for the taxable product area, not the municipal area. 
 8             MS. MAHER:  I would just add that I could see 
 9   in a quoted marketplace with time constraints limit 
10   orders working.  But when I think of retail in regards to 
11   fixed income, most of the time those aren't the 
12   securities where they are focusing their activity.  And 
13   the activity that we see in the more actively traded 
14   issues, I think, is for more professional type of 
15   accounts in the majority. 
16             MR. McVEY:  And Bob, just sticking with the 
17   regulatory theme, we've got an important change coming up 
18   in May with the markup disclosure rules.  And maybe you 
19   can kick that conversation off with the development of 
20   those rules and have the market participants react to 
21   whether the industry is ready and how we see this going 
22   as that rule takes place next month. 
23             MR. COLBY:  Well, we're coming to the end of an 
24   18-month implementation period for a markup disclosure 
25   requirement that has been proposed at least twice before, 
0187
 1   once in '78 and once in '94.  So we're finally getting to 
 2   an end.  It's a little different. 
 3             So this is disclosure of the markups on trades 
 4   where there was -- the other side took place on the same 
 5   day.  So it started off with an idea of a riskless 
 6   principal trade.  But then as we thought more about it 
 7   and we looked at the numbers, it turned out that if you 
 8   set a very short time period, it could be very easily 
 9   evadable.  So when we looked at the numbers, it looked 
10   like there was a natural break point at the same day.  
11   And so if you're doing a trade and it's offset the same 
12   day for a noninstitutional customer, in May, you're 
13   supposed to show your markup and you're supposed to give 
14   a link to TRACE and also the precise time of trade.  So 
15   someone that wants to go into TRACE and try to identify 
16   their trade and the offset trade can do it. 
17             I should say that I'm speaking in terms of 
18   TRACE but the MSRB under the leadership of Mike Post did 
19   the exact same thing.  So we're working in parallel here. 
20    And so we are very hopeful that we will see 
21   confirmations to retail customers showing their markup in 
22   same-day trades. 
23             MR. McVEY:  Do the rest of you want to react to 
24   whether the industry is ready and the benefits to your 
25   retail clients on the back of it? 
0188
 1             MS. MAHER:  Well, it's been extraordinarily 
 2   complex.  We will be ready, even if I work 24 hours a day 
 3   and put it in by hand. 
 4             You know, it's going to take, for our 
 5   organization, four different vendor coordinations.  And 
 6   so it has just been incredibly complex and it goes from 
 7   our position management system to our FA front end, to 
 8   our back office vendor to ultimately our confirm, which 
 9   is in house.  And so we have four different vendors that 
10   need to communicate and we have not been presented with, 
11   as of today, a fully developed solution that can be 
12   tested end to end. 
13             So they all have target dates, and they say 
14   they will be ready. 
15             MR. ITURRINO:  So I can completely relate to 
16   the complexities of this.  I think we chose to go in a 
17   different direction.  We are building the solution in 
18   house.  We, too, will be ready and be compliant, 
19   obviously, by May 14. 
20             But I do think, and I just want to -- and I 
21   think everybody in this room at some point or another has 
22   sort of talked about markups quite a bit.  I do want to 
23   pivot to the fact that we view this as, you know, part of 
24   a portfolio of initiatives that, you know, from a 
25   regulatory perspective, that have brought -- whether more 
0189
 1   attention, more transparency -- have consequently led to 
 2   more electronic trading, more participation in ATSs.  And 
 3   so we do view this as just part of a set of regulatory 
 4   initiatives that have been beneficial or are beneficial 
 5   to the retail customer, as it's generated sort of this 
 6   interconnectivity and participation in broad markets, 
 7   particularly in electronic venues. 
 8             So we view that as an overall positive.  
 9   However, I am not going to trivialize how difficult it 
10   is. 
11             MR. VALES:  I should say as a vendor, we better 
12   be ready.  And I hope my guys are watching right now, 
13   that they get that message. 
14             But the -- what I've stated is that, to echo 
15   the sentiments of the other members up here, that it is 
16   incredibly complex to do.  And so what I think, as an 
17   industry, is that from the regulatory front, to be able 
18   to put a number on a confirm and satisfy the regulatory 
19   obligation of having that methodology in place, we'll all 
20   be ready. 
21             I think what everyone is waiting for is that 
22   there really -- I think most firms, and to Kristin's 
23   point about the complexity and having the 
24   interconnectedness between back offices, front end 
25   trading systems and accounting, that until we get into 
0190
 1   some different types of markets to actually see how the 
 2   modeling will work, that's where people are more 
 3   concerned.  So it's not necessarily the regulatory risk, 
 4   it's the business risk, and I think that's what's going 
 5   to take a little bit of time to flesh out. 
 6             MR. McVEY:  Great.  And then one final one on 
 7   the regulatory landscape, and then we'll open it up to 
 8   the rest of the committee for any questions that you may 
 9   have. 
10             But most of the retail E trading venues are 
11   regulated as ATS venues.  The institutional platforms 
12   tend to be as broker-dealers, because of the RFQ protocol 
13   being the dominant protocol.  But are the ATS rules 
14   working suitably for retail fixed income and is there 
15   anything any of you would like to see changed about the 
16   ATS rules? 
17             MR. VALES:  I was hoping to do cleanup on this 
18   in case there was anything that people were unhappy with. 
19    But I am going to say in general that I think the SEC's 
20   ATS guidelines have been very helpful in that they're 
21   more requirements to keep the SEC current on the type of 
22   business that you're doing, the business procedures that 
23   you have in place and quarterly reporting.  And in doing 
24   that, they have allowed for a lot of flexibility and 
25   creativity in terms of the types of models that ATSs can 
0191
 1   function under.  And so if you look at the marketplace, 
 2   and especially in credit, you've seen a number of the new 
 3   entrants that have come into the marketplace.  I think 
 4   that's a function of us having the flexibility and the 
 5   opportunity to experiment and try different things. 
 6             They're all happy with the ATS, I think. 
 7             MR. McVEY:  Matt, you used to know these 
 8   guidelines a lot better than you do now. 
 9             MR. ANDRESEN:  You're wearing me out today. 
10             (Laughter.) 
11             MR. ANDRESEN:  The ATS construct, I think, has 
12   served the fixed income well, because it does allow for a 
13   -- ease of entry into the marketplace for new models.  So 
14   your company, TMC, and the other ATS platforms have 
15   provided a degree of innovation between the different 
16   types of protocols offered and the different market 
17   structures, but within the context of something that can 
18   be tracked and well understood.  So I don't think we're 
19   concerned about that as a construct. 
20             MR. VALES:  I guess the one part that I would 
21   encourage is that there have been a number of provisions 
22   like SCI and CATs that have originated on the equities 
23   side that have been examined on the fixed income side.  
24   And fortunately, I am going to say that I think that the 
25   direction so far, we've been going the right way.  
0192
 1   Whereas, opposed to getting burdened down on fixed income 
 2   with SCI, they recognize that 15c3-5 was sufficient 
 3   enough.   
 4             So my one word of caution would be that, you 
 5   know, to keep in perspective that you've heard the 
 6   numbers before that, in the muni market, there are 40,000 
 7   trades a day; in the credit market, there are 45,000 
 8   trades a day.  We are trading annually what trades in one 
 9   day -- in the fixed income market, we're trading what 
10   trades in one day in the equity market. 
11             And so to kind of get overburdened with the 
12   ability to track items -- and I always think about, Rick, 
13   you once made a comment that a liquid bond in credit is 
14   22 trades a day.  It's just not that difficult to kind of 
15   trace where it started and where it ended.  And so I 
16   would be just cautious on that front. 
17             MR. McVEY:  Good.  We'll open it up to the 
18   committee.  And Tom, I see you have a question. 
19             MR. THEES:  Just what -- I didn't -- I'm not 
20   clear, Bob, on what the definition of an offset is inside 
21   of a firm.  So, for instance, if a firm executes a trade 
22   and it's a $25,000 trade, does it have to be $25,000 on 
23   the other side to trigger the disclosure and the markup? 
24    Or is it any trade that happens in that bond triggers 
25   the disclosure? 
0193
 1             MR. COLBY:  Yeah, you're seeing the 25,000 sale 
 2   to a retail customer?  You have to offset it that amount 
 3   that they --  
 4             MR. THEES:  That amount that they -- so, for 
 5   instance --  
 6             MR. COLBY:  Having said that, you know, 86 
 7   percent gets offset.  And, Tom, I'm trying to remember 
 8   this, in an hour.  And something like 60 percent gets 
 9   offset within 15 minutes. 
10             MR. THEES:  Okay, so the majority of the market 
11   is going to get offset.  So therefore this rule is going 
12   to actually trigger a markup disclosure on the majority 
13   of trades? 
14             MR. COLBY:  Oh, for the great majority, yes. 
15             MR. GIRA:  Yeah, I think it's 87 percent -- 
16   sorry.  I think the offset is actually more like 87 
17   percent in half an hour.  Thanks. 
18             MR. THEES:  And I had one question for Matt. 
19             When it trades through, do you get an 
20   explanation as to the -- I mean, because it's a disclosed 
21   marketplace.  You know who you're dealing with and you 
22   know that they just traded through you, right?  I'm not 
23   saying it's ultimately disclosed --  
24             MR. ANDRESEN:  So actually the degree of 
25   transparency we have into what happened varies widely by 
0194
 1   market.  So some markets will give us what we call color 
 2   information.  They'll say, hey, you won this bid.  Here's 
 3   your distance to the second-best bid.  Or if we lost, 
 4   here's what your distance was to the best bid.  So you 
 5   get like proximity information.  You don't always get 
 6   that, and you certainly don't know who it was that you're 
 7   doing that.   
 8             And I think in the case of what you're asking, 
 9   we will, if we're bidding on a given ATS platform, we win 
10   the auction and then we see -- we have to go to EMMA and 
11   look and see, well, hey, did this result in a trade?  And 
12   one thing about me, it's not that hard, right, because 
13   it's so disparate.  There's the 47 bonds in that CUSIP.  
14   Like, that's the 47 bonds in the CUSIP.  And so we draw 
15   an inference that that is, in fact, the same piece of 
16   inventory.  We call the ATS and we say, hey, what 
17   happened?  And typically, you know, the explanation will 
18   be, well, that was like a desk credit or they will have 
19   another euphemism for it, for not having actually  
20   -- and I think the ATSs see it our way.   
21             I think they would rather that their auctions 
22   be sacrosanct and their ability to sort of enforce that 
23   is not a hundred percent.  So I think they probably see 
24   it our way.  But for us as a bidder, we are providing 
25   price discovery to the marketplace with -- in auctions 
0195
 1   where we're not actually being rewarded.  And, for us, 
 2   our being someone who generates the prices from a 
 3   quantitative model and is able to respond to auctions in 
 4   an automated way, we're not -- the incremental costs of 
 5   us, in fixed costs, of providing another response to an 
 6   auction is indistinguishable from zero.  But for a lot of 
 7   other entrants, if they're saying, I could go into that 
 8   auction but even if I win, you know, I might not win and 
 9   get the trade, that undermines the incentive to 
10   participate in that.   
11             And that's the same challenges you had in other 
12   asset classes around just straight limit orders that 
13   we're talking about.  Is you wanted -- that limit order 
14   was there.  You wanted people to say, I'm going to 
15   actually interact with that if that's the best price.  
16   Here, it's a similar challenge in a more diffused 
17   marketplace where you have an auction and if you win it, 
18   you know, let the winner win.  And if you want to trade 
19   it yourself, then go ahead and compete in the auction on 
20   a level playing field. 
21             MR. VALES:  And traditionally, in the old voice 
22   world, if you had an item, if you wanted to do an RFQ, 
23   you went to a single broker's broker.  And that broker's 
24   broker, if you were going to bid the bonds, you bid what 
25   was called in comp.  And before you would get the results 
0196
 1   back, the dealer would have the responsibility to give 
 2   him, the broker's broker, the bid. 
 3             In the new world, and especially with best ex, 
 4   is that these bid wanteds are going out to multiple 
 5   places.  And so the idea that we would know the firm's 
 6   bid prior to the auction time doesn't exist anymore.  And 
 7   so that's one of the issues that the industry is faced 
 8   with, is that you're kind of operating and believing that 
 9   the auction is being run fairly and that the dealer 
10   that's conducting the auction or putting the item out is 
11   acting fairly with all the other participants. 
12             MR. McVEY:  Larry. 
13             MR. HARRIS:  When Rick introduced this session, 
14   he told us that a very small fraction of bonds trade in 
15   markets with two-sided quotes.  That's undoubtedly true 
16   in the muni markets where there are a million issues. 
17             There were 17,000 bonds, separate bonds, that 
18   reported trades in TRACE up through March of 2005.  Using 
19   data that were given to me by Interactive Brokers, I 
20   counted the number of times that bonds traded in markets 
21   with two-sided quotes and I found very, very different 
22   results.  Let me tell you what the results were and then 
23   I will ask a question. 
24             So first of all, the data.  Interactive Brokers 
25   allows their clients to trade bonds as though they were 
0197
 1   stocks.  They look exactly the same with two-sided quotes 
 2   when they're available. 
 3             To this end, they collect the best bids and 
 4   offers from a variety of electronic systems, many of 
 5   which are represented here today, and a few others, and 
 6   from dealers as well.  And they adjust those quotes also 
 7   to reflect any fees that the various venues may charge. 
 8             So here is a quick result.  For those 17,000 
 9   bonds, some of which only traded once in TRACE, ever, I 
10   counted the fraction of the time between 8:00 in the 
11   morning and 5:15 in the evening, which is more than a 
12   normal trading day, that a two-sided quote was present in 
13   that bond.  And then I counted the number of bonds and 
14   classified the results. 
15             For 25 percent of all those 17,000 bonds during 
16   the first quarter of 2015 -- and I'm sure the results are 
17   much stronger now -- there were two-sided quotes 95 
18   percent of the time.  That is, 95 percent of the time 
19   from 8:00 to 5:15.  And for the median bond in the sample 
20   was on the order of 60 percent. 
21             Now, we might suspect that these quotes aren't 
22   very strong, that maybe they're far from the market.  But 
23   43 percent of all TRACE trades in the first quarter of 
24   2015 traded through these quotes.  That means that these 
25   quotes were good enough that when a trade took place, the 
0198
 1   trade was outside the quote, it traded through.  Ninety 
 2   percent of all the trades, by the way, were in two-sided 
 3   markets.  So if you just count the trades and look to see 
 4   whether there was a two-sided market, it was 90 percent. 
 5             Of these trades, 50 percent were riskless 
 6   principal trades.  So those are trades where you can see 
 7   the buy and then the sell or the sell and then the buy 
 8   within -- typically within a second.  The mean markup for 
 9   those trades was 71 basis points.  So that's the markup 
10   on the trade, that is not even the total transaction 
11   cost.  The total transaction cost would typically be 
12   measured from the middle of the bid/ask spread and this 
13   markup is being measured from the -- what's the 
14   difference in the two trade prices, one of which is 
15   typically the quoted price. 
16             And then of these riskless principal trades, 41 
17   percent of them are -- I'm sorry, of the trade-throughs, 
18   41 percent of them are riskless principal trades. 
19             So what does this say to us?  It says that many 
20   of our -- and, of course, these are counting trades.  So 
21   that means that these results primarily apply to retail 
22   transactions and not to institutional transactions.  They 
23   primarily have relevance to the retail. 
24             So what it says is that currently in the 
25   market, and this is as of a couple years ago, almost 
0199
 1   three years ago now, that the retail was largely being 
 2   served by brokers and not by dealers.  Of course, they 
 3   were broker-dealers, but their service was essentially 
 4   brokerage and not dealing.  The markups, of course are 
 5   pretty high. 
 6             So the question is, given these types of 
 7   results, should we be making greater distinctions between 
 8   broker trades for which, in every other market, we see a 
 9   commission schedule up front, and dealing trades, for 
10   which we understand that we normally see markups and the 
11   markup reflects the risk that the broker-dealer faces?   
12             And, secondly, the other question I'd like to 
13   ask is that, in markets that look like this, like thinly 
14   traded equities, it's common to report bids and offers to 
15   the customer before they decide to trade.  Wouldn't it be 
16   sensible for us to recommend that retail broker-dealers 
17   when servicing their clients, especially when they're 
18   providing essentially brokerage services, that they 
19   provide the bid and offer the best prices available at 
20   time of -- before the trading decision is made?  So those 
21   are my two questions. 
22             MR. ITURRINO:  I could try to address the 
23   professor's questions to the best that I can, to the 
24   degree of what it is that we do internally to address the 
25   core issues, I think, that you're implying.  So the 
0200
 1   questions that you're asking, I'm sure, are going to be 
 2   debated by this committee. 
 3             As far as what we do, north of 75 percent of 
 4   our corporate quotes have a bid side.  So, as I said at 
 5   the beginning, our slots are open.   
 6             Now, that being said, it's not -- and then you 
 7   quoted sort of a 90 percent number in there.  There are 
 8   controls that we have in place to make sure that our 
 9   customers are interacting with what we would consider are 
10   an adequate bid.  Right?  So if we see an offer that's 
11   par or 100 and the bid is 20, maybe we're not going to 
12   promote that.   
13             And so I think that it's a decent point.  I 
14   think, as far as, you know, an open architecture 
15   environment, we do promote the bids.  The corporate world 
16   is a different world than the municipal world.  We just 
17   don't see that type of liquidity in the municipal world. 
18             MR. HARRIS:  We're not talking munis here. 
19             MR. ITURRINO:  But that's essentially what we 
20   do internally.   
21             I do not know how to answer sort of --  
22             MR. HARRIS:  But you do show a two-sided market 
23   to your customers in equities, right? 
24             MR. ITURRINO:  As we do in most corporates, 
25   yes. 
0201
 1             MR. HARRIS:  Okay.  Well, are you showing the 
 2   two-sided market before they trade? 
 3             MR. ITURRINO:  Yes. 
 4             MR. HARRIS:  And how about Fidelity?  I'm 
 5   sorry, is Fidelity here?  Wells Fargo, Kristin. 
 6             MS. MAHER:  Wells Fargo will send that bid 
 7   wanted side of the market back to our financial adviser. 
 8    So all of our client interest comes to us through RFQ 
 9   right now. 
10             MR. HARRIS:  Will the financial adviser even 
11   see the two-sided market, much less the customer? 
12             MS. MAHER:  The financial adviser will; it's 
13   right on the platform, via the ATS. 
14             MR. HARRIS:  They'll see the two-sided market 
15   as best you can see it? 
16             MS. MAHER:  Correct.  Which I think speaks to 
17   why our data was so different between the two asset 
18   classes.  So there's not a need for as much price 
19   discovery in the corporate market, and so we see a much 
20   higher conversion rate, along the lines of 70, 75 
21   percent.  And so our advisers are able to get those 
22   levels back to our customers, our customers experience 
23   that day to day, and therefore they trust the pricing and 
24   they can just -- they don't have to request pricing on so 
25   many securities because they know what they're seeing is 
0202
 1   something that's available. 
 2             MR. HARRIS:  Is your model essentially an 
 3   advisory model as opposed to a discount brokerage model? 
 4    If the Commission decided to impose some requirement 
 5   that a best bid and offer be displayed, would it be 
 6   reasonable to carve an exemption for your model so that 
 7   it gets displayed to the advisers but not necessarily to 
 8   the customers? 
 9             MS. MAHER:  I'm sorry, your question to me is, 
10   is that possible? 
11             MR. HARRIS:  Virtually everything is possible. 
12             MS. MAHER:  Right. 
13             MR. HARRIS:  The question is, if the Commission 
14   were to decide that the best bid and offer should be 
15   displayed to the market, should there be a carveout for 
16   people who use models such as yours, which is an advisory 
17   model as opposed to a -- your clients aren't actually 
18   seeing the market, your clients are seeing their 
19   advisers. 
20             MS. MAHER:  Correct.  So we would -- it would 
21   go directly to our financial adviser.  Otherwise, we 
22   would have to change the entire model. 
23             MR. HARRIS:  Right, right.  So we're not asking 
24   for that. 
25             MS. MAHER:  I think Wells Fargo would be 
0203
 1   supportive of that. 
 2             MR. VALES:  And, Larry, I could add just a 
 3   couple things.  One is -- and I don't know what the data 
 4   was that you used in your study.  But as I mentioned 
 5   before, with us, it's actually two to one.  And so we see 
 6   many more bid sides than offer sides.  So I'm going to 
 7   agree with you that it sounded understated from what 
 8   you're seeing. 
 9             MR. HARRIS:  My results show that, as well.  We 
10   all know the reason why, is because people who are 
11   providing these quotes, it's easier for them to buy 
12   something that they don't own than to sell something that 
13   they don't own. 
14             MR. VALES:  Correct.  The second part of that 
15   is then one of the reasons -- and I'm not sure if you 
16   looked at quantity.  But typically what we see, when we 
17   see two-sided markets, we see in credit bid wanteds go 
18   out for two reasons.  First, I should mention we do have 
19   a number of firms that if there is a quoted two-sided 
20   market and they can validate the market, they will give 
21   the adviser or the client a warning saying, hey, this was 
22   the quoted market, here is what the bid is, do you still 
23   want to put it out for the bid. 
24             If you pass through that, if you think about 
25   it, once you put it out for the bid, now you're asking 
0204
 1   for a time premium, right?  You're asking for somebody to 
 2   give you a bid that now has to be firm for some period of 
 3   time.  And even though I used the statistics before that 
 4   about 90 percent of the bid wanteds in credit trade 
 5   within 30 minutes, there is that time premium that you 
 6   have to account for and, as a dealer, you would want to 
 7   get paid for. 
 8             The second part, if you didn't look at 
 9   quantity, that's what we see a lot.  We see a lot of two-
10   sided markets but the bid side will be for a reasonable 
11   quantity.  Whereas, the offered side, if they're long the 
12   bonds already, that they'd be willing to trade smaller 
13   amounts.  So that might be something interesting to look 
14   at as well. 
15             MR. HARRIS:  It certainly would be.  And just 
16   for clarity, the quotes that I have referred to are 
17   standing quotes, not requests, and of course they vary in 
18   size, and that's why these results are most relevant to 
19   the retail and not to institutional. 
20             MR. VALES:  Yeah.  And even retail, there's a 
21   big difference between, as you know, like the three and 
22   four bonds versus the tens and twenty-fives. 
23             MR. HARRIS:  Understood. 
24             MR. McVEY:  Horace. 
25             MR. CARTER:  Just a couple of points.  One, 
0205
 1   first is a question.  Do you have any idea how reliable 
 2   the quotes that are being posted are? 
 3             MR. HARRIS:  Yes.  Strictly speaking, the 
 4   quotes that are being posted, almost all of them are 
 5   actually legally indications as opposed to actionable 
 6   quotes.  But as a practical matter, they're almost all 
 7   reliable.  And the reason why is that both the venues 
 8   that are supporting them and the traders that are behind 
 9   them only survive when they provide service.  And so 
10   unless the market moves extraordinarily, the expectation 
11   is that they'll trade. 
12             Interactive doesn't complain about failing to 
13   hit their quotes. 
14             MR. CARTER:  Okay.  It's been my experience 
15   that that can be a little more unreliable than they seem. 
16    But I accept your answer. 
17             My second question is, do you think that -- it 
18   seemed like a lot of the problem that originates from 
19   this will be solved with the markup disclosure.  Because 
20   you were talking about trade-throughs, which are almost 
21   certainly markups or markdowns that were priced against 
22   the quote.  If we have -- you know, when we have markup 
23   disclosure for instantaneous trade, that is materially 
24   the same thing as agent trading with commission 
25   disclosed.  Won't that go a long way towards fixing this 
0206
 1   problem? 
 2             MR. HARRIS:  So in every other market where we 
 3   have brokers, we see that clients choose their brokers 
 4   based on their commission schedule.  And so the 
 5   competition takes place according to that schedule.  
 6   There's no question that if people would remember the 
 7   markups that they're seeing, they'll start asking and 
 8   they may move from one broker-dealer to another broker-
 9   dealer. 
10             But to know what the markups are only if you 
11   trade seems to be a strong burden on competition.  So I 
12   can only know that -- Interactive doesn't even mark up.  
13   But, you know, you've got Broker A versus Broker B, okay? 
14    So I see that I'm trading with A and I'm getting a 71 
15   basis point markup.  That's pretty extreme.  I think, 
16   well, maybe I should trade with B.  But I don't even know 
17   -- I have to move my account to B and start trading with 
18   them before I can even find out what his markup is, much 
19   less negotiate for a better markup. 
20             So until this information -- these are people 
21   who are acting as brokers doing riskless principal 
22   trades.  So until that information is posted ex ante as 
23   opposed to ex post, I don't think that we're going to get 
24   as much improvement as we hope. 
25             But there is no question that the ex post 
0207
 1   posting information is going to help.  It's definitely a 
 2   step in the right direction. 
 3             MR. CARTER:  I would just push back a little 
 4   bit on the idea that 71 basis points, by which I assume 
 5   you mean .71 percent of dollar price, not of yield, I'd 
 6   push back on the idea that that's an excessively large 
 7   markup, especially when you're talking about potentially 
 8   very small trades and the enormous cost of the 
 9   infrastructure that goes into executing these trades. 
10             And there's more to the execution of a bond 
11   than just the actual execution of a trade.  There's 
12   customer education, something we've gone over pretty 
13   thoroughly today.  There's a lot more that goes into it 
14   than just the execution of that specific trade.  And so I 
15   would push back on that concept just a little bit. 
16             MR. HARRIS:  Well, to be completely clear about 
17   this, I never said that it was excessive; I only 
18   indicated that it seems large to me relative to 
19   commissions.  But you're right. 
20             All that I would be asking for is that we have 
21   an open competition among broker-dealers so that we find 
22   out what the true costs are.  Because surely, they're not 
23   going to go below their costs.  And those costs, of 
24   course, include their normal return for the 
25   entrepreneur's time and for the shareholder's capital and 
0208
 1   so forth. 
 2             You know, in competitive markets, that's what 
 3   we're looking for.  And if that's what it turns out to 
 4   be, then so be it. 
 5             MR. CARTER:  I guess my final comment would be 
 6   that that's a large commitment in structural change for 
 7   what amounts to a very small percentage of fixed income 
 8   transactions.  It's a small subset of one asset class.  
 9   You know, we trade mortgages, we trade agencies, 
10   treasuries, we trade lots of different bonds.  And I 
11   think that that would be a large commitment for such a 
12   small subset. 
13             MR. HARRIS:  And yet almost every broker-dealer 
14   is capable of doing this in equities, which are just 
15   other securities.  And I would say that the numbers that 
16   I presented to you are not a small fraction by any means. 
17    And the total amount of money that's traded through 
18   amounts to a significant fraction of a billion dollars 
19   per year.  This is real money we're talking about. 
20             MR. McVEY:  Tom. 
21             MR. GIRA:  Yeah, just to -- maybe there are 
22   some recommendations in here in the future.  But to 
23   Larry's point, and he makes a very valid point, I do 
24   wonder if this -- if part of this is driven by the fact 
25   that with TRACE, the markup is inclusive in with the 
0209
 1   price that is disseminated?  And I'm familiar with 
 2   Larry's study and I think in some cases, part of that is 
 3   because unlike equities, the markups are part of the 
 4   reported prices.  So I think it can give sort of a false 
 5   sense of trade-throughs. 
 6             I do think there's also a lot to be said for 
 7   maybe the -- if I get my words right -- the exe ante 
 8   disclosure or information about markups.  And I can tell 
 9   you that we are, at FINRA, starting to look at sort of, 
10   on an aggregate basis, sort of a grid that could be 
11   disclosed that takes average markups across different, 
12   you know, high-yield, investment grade, different sizes, 
13   different ratings and things like that, so that people 
14   can see that before they're trading. 
15             MR. HARRIS:  Tom's point is very well taken.  
16   And let me reinforce it. 
17             Suppose somebody does a riskless principal 
18   trade when there's a standing offer, say 100.  And he 
19   marks it up.  The markup is essentially his commission.  
20   And so that markup will be 100-point-something more.  
21   Okay?  That would appear as a trade-through in my data. 
22             For counting trade-throughs, I required that 
23   the trade-through be more than 10 basis points beyond the 
24   quoted price.  The reason why was that discount retail 
25   brokers are typically charging 10 basis points in 
0210
 1   commission.  So, now a full-service broker, of course, 
 2   might charge more and that may explain a good portion of 
 3   the 71 basis points that we discussed.  And what you're 
 4   being charged is not only for execution services but 
 5   you're also being charged for advisement and other 
 6   things.  All of these things are things that can be 
 7   identified or eventually identified in competitive 
 8   markets.  So I'm not criticizing these prices.  But I 
 9   think the point is very well taken. 
10             But the numbers I gave you already account for 
11   at least a 10 basis points commission.  They also in 
12   fairness note there are a number of markups for which 
13   there are no commissions at all because people are doing 
14   wrap accounts in other situations or they're transferring 
15   positions from one place to another where it wouldn't be 
16   reasonable to have a markup. 
17             MR. McVEY:  Anyone else?  Tom, did you have 
18   another question or are you just still vertical?  Tom. 
19             MR. THEES:  Matt, given we're talking about 
20   market quotes, you talked about -- did I get it right?  
21   You said you priced 13,000 securities?  Sorry, that was 
22   line items? 
23             MR. ANDRESEN:  No, no, that's the line items of 
24   inventory. 
25             MR. THEES:  So therefore you certainly priced 
0211
 1   13,000, because those are your line items. 
 2             MR. ANDRESEN:  A lot more than that. 
 3             MR. THEES:  So you have two-sided quotes, 
 4   potentially out there in the -- in the ecosystem, you 
 5   have two-sided quotes that you are either ready to 
 6   provide or you do publish actually? 
 7             MR. ANDRESEN:  On munis, you can't short. 
 8             MR. THEES:  So it's a quote. 
 9             MR. ANDRESEN:  So the offerings that we have of 
10   the million CUSIPs and munis were limited to like 13,000, 
11   ones that we can make a two-sided market for. 
12             MR. THEES:  But you theoretically could quote 
13   an offered side without having -- it's not an actionable 
14   item, but you could quote it? 
15             MR. ANDRESEN:  We could quote -- I think of a 
16   quote as something where I'm holding myself out.  I can't 
17   do that. 
18             MR. THEES:  No, no, I'm sorry.  Let me 
19   distinguish between a quote of price, which is not 
20   actionable --  
21             MR. ANDRESEN:  Could we generate a platonic -- 
22   a sort of platonic price, if we had happened to own it, 
23   what would we look to sell it at? 
24             MR. THEES:  Yes. 
25             MR. ANDRESEN:  Yeah. 
0212
 1             MR. THEES:  Well, in fact, on the last look 
 2   platform, a lot of those prices are sitting there.  
 3   They're not actionable and they're last look and 
 4   therefore they're not live, and therefore they turn into 
 5   a DK, right?  The bond was attempted to be traded at the 
 6   price and the individual on the other side of that trade 
 7   said, no, thank you, I can't trade it, or I won't trade 
 8   it at that price.  But I'm just saying, you publish those 
 9   quotes, if you choose to, on ATSs?  Bid side. 
10             MR. ANDRESEN:  I've got Bob Colby like next to 
11   me -- sharing a microphone --  
12             (Laughter.) 
13             MR. THEES:  So let's talk about your line 
14   items, your 13,000 line items.  Would you theoretically 
15   be willing or are you already publishing those 13,000 
16   line item, bid side, offered side, as you referred to it, 
17   quotes today? 
18             MR. ANDRESEN:  So we show all 13,000 line items 
19   out on multiple platforms.  So we do that today.  We can 
20   generate bid prices on over 700,000 prices, but we're 
21   waiting for someone to want to sell them. 
22             MR. THEES:  Someone asks you for it, then you 
23   do it. 
24             MR. ANDRESEN:  We have them ready should 
25   someone ask.  And so we respond to many thousands a day. 
0213
 1    But we don't know which ones.  You know, there's about 
 2   -- as I said, there's over 20,000 auctions a day.  We 
 3   don't know what they're going to be in.  When they 
 4   happen, we generate responses and then we're -- but we're 
 5   at the mercy of the sellers to say, one, when are they 
 6   going to sell; two, are they going to have an auction.  
 7   If they have an auction and we win it, are they willing 
 8   to trade. 
 9             MR. THEES:  I guess my only point, and more 
10   referring to the corporate bond market experience, but I 
11   think it applies to munis as well, there are, as many buy 
12   siders would attest to, there are many people who quote 
13   bonds at 8:00 in the morning and at 8:01 in the morning 
14   when you call the person who just quoted that bond and 
15   say, I would like to trade with you, they say, oh, that 
16   was a quote and I've got to go to the restroom, I'll be 
17   back.  So it's not real.   
18             And I only say that in the context of looking 
19   at quote data and having that be a reference point is, I 
20   think, in both markets, the muni and the corporate 
21   market, is subject to interpretation, I guess, is what I 
22   would say.  Extreme interpretation, in my experience.  It 
23   is not -- live, executable prices are 100 percent.  And 
24   many people publish those and you're publishing them on 
25   your 13,000. 
0214
 1             MR. ANDRESEN:  Any time that we bid in an 
 2   auction, that's a live price, that's a firm price. 
 3             MR. THEES:  That's a live price.  Right, no, I 
 4   understand. 
 5             MR. ANDRESEN:  And any time that we offer 
 6   something out, it's a firm price. 
 7             MR. COLBY:  And the rest is bogus prices, 
 8   right, Matt? 
 9             (Laughter.) 
10             MR. THEES:  I have no more questions. 
11             MR. KUCHINAD:  Matt, not to keep you on the 
12   spot while you're sitting right next to Bob.  But is 
13   there a reason why your firm is willing to post thousands 
14   of prices on equities even when you're not sure if 
15   somebody is interested but yet, in munis and corporates, 
16   while you can generate a price on the bid side or the 
17   offer side, but let's say the bid side where you don't 
18   even have to worry about going short, you don't post all 
19   those prices, you wait until somebody is indicating some 
20   interest to sell that security?  Is there a reason that 
21   you can think of, incentive wise, market structure wise? 
22             MR. ANDRESEN:  Oh, I see what you're saying.  
23   That's the market structure. 
24             I guess to make a point that I think we're sort 
25   of dancing around the edges of, I do not believe that a 
0215
 1   retail customer represented by a retail broker who 
 2   generates a bid wanted request -- they're getting a very 
 3   vibrant price discovery process.  So municipals don't 
 4   trade continuously throughout the day and people's 
 5   opinions don't change throughout the day like they do in 
 6   equities. 
 7             It tends to be one sort of round of trading a 
 8   day, a four-hour window, a collection period, and then 
 9   there's a denouement where they decide whether to trade 
10   or not trade at the very end.   
11             For this, what's unique about fixed income 
12   versus equities is that lack of change of direction to 
13   that.  So -- and what's great about this process is I 
14   don't think, if you're in a -- if you're in a competitive 
15   auction, you're getting a great price.  And I would argue 
16   potentially, some academic work that I'm sure the 
17   academics here could opine on, that one-off call markets 
18   like that where you have a collection period and a single 
19   price is in some ways better than the sporadic price 
20   discovery because of the compression of time down to one 
21   moment at the point of sale. 
22             And so I do think that we should be excited 
23   about what happens with these auctions in fixed income, 
24   because you are getting such -- I mean, seven bidders.  
25   Like if I could get seven bidders for your house or your 
0216
 1   car, you'd be getting a great price.  Like you'd be 
 2   selling through ask. 
 3             And so this is a great result.  And what we 
 4   advocate is wider use of these.  So oftentimes trades are 
 5   happening without there ever being an auction generated. 
 6    Or when the auction does happen, the winners aren't 
 7   being rewarded for winning.  So I think there are things 
 8   to be improved.  But I don't think it is necessarily a 
 9   worse market structure for municipals and corporates to 
10   have an auction-based model than a continuous bid and 
11   offer.  I don't think it should be thought of in that 
12   sort of pejorative sense. 
13             MR. VALES:  And just to take some of the heat 
14   off of Matt, the RFQ process is incredibly effective.  
15   And if you look at credit, it's really interesting.  
16   Because you go back to this analogy, if there are more 
17   quoted bid sides than items coming out for the bid, 
18   right, why doesn't everyone just hit the bid side?  And 
19   what we see is that, in credit and corporates, it tends 
20   to be high yield or different quantities, being in 
21   smaller quantities, threes, fives, sevens, where there is 
22   not going to be a quoted market to support it.  And by 
23   the RFQ process, you're asking people to sharpen their 
24   pencil. 
25             So just by putting up a quoted bid side doesn't 
0217
 1   necessarily mean now all of a sudden those bid sides are 
 2   going to be hit.  And that's why the RFQ process is so 
 3   effective in a market where there are so many different 
 4   types of credits. 
 5             MR. McVEY:  One more, Larry? 
 6             MR. HARRIS:  A quick comment and then an 
 7   unrelated question for Renzo. 
 8             So the quick comment about these RFQs is that 
 9   this is a really important process.  It takes time and 
10   money to respond to these quotes.  If we don't protect 
11   the people who are making an effort to provide this 
12   liquidity, then we're going to lose that liquidity.  So 
13   allowing last looks is not a good thing.  People are 
14   basically producing information for free and then they 
15   get hurt by -- well, they get hurt just by virtue of the 
16   fact that they spent effort to do it.  And so I think 
17   there's definitely room for improvement here somehow. 
18             The question to Renzo is this.  When we 
19   discussed whether Fidelity will offer -- allow its 
20   customers to offer standing limit orders to the market, 
21   you said yes.  But then indicated that those orders could 
22   only stand from 15 seconds to, I forget, I think you said 
23   perhaps 20 minutes, and that you would only accept all or 
24   none offers.  Now you, of course, have a business model 
25   and can choose whatever you want to do.  But I was a 
0218
 1   little curious as to knowing why these restrictions are 
 2   placed on your customers' ability to offer limit orders? 
 3    And I presume that your customers know it, but you may 
 4   want to confirm that there's some mechanism by which they 
 5   know that their orders are only good for 15 seconds to 20 
 6   minutes and that they're only all or nothing, as opposed 
 7   to what they might have otherwise have assumed? 
 8             MR. ITURRINO:  Yes, so the short answer is, 
 9   yes, our customers know that those are the parameters.  
10   As far as why we don't let our customers sort of, you 
11   know, do anything other than all or none, you know, that 
12   could be something that we revisit.  But we find that 
13   when it comes to trading fixed income, the all or none 
14   protocol is more suitable for retail.  You know, there 
15   are chances that -- or there are scenarios where you can 
16   execute part of the order and you don't ever get to 
17   execute the other part of the order.  Right?  So the all 
18   or none protocol is just -- it's better served for the 
19   retail experience. 
20             MR. HARRIS:  And if you would, but you don't 
21   have to, can you tell us the rationale behind such a 
22   short period of time for which the orders will be good, 
23   from 15 seconds to 20 minutes? 
24             MR. ITURRINO:  Sure.  So that's 90 seconds and 
25   20 minutes, sorry. 
0219
 1             With that being said, yeah, it depends on asset 
 2   class, right?  And it depends on where it goes.  So these 
 3   are things that we've agreed on with either multiple ATSs 
 4   or a single ATS.  So these are business decisions. 
 5             MR. HARRIS:  And when the firm is offering 
 6   liquidity, do you impose similar restrictions on your own 
 7   dealers?  I would presume not; is that correct? 
 8             MR. ITURRINO:  When the -- you mean us? 
 9             MR. HARRIS:  Yes.  When you -- I presume you 
10   post regularly, right? 
11             MR. ITURRINO:  Yeah.  We are on the exact same 
12   playing field that I described.  So that's part of our 
13   open architecture environment, right?  So, no, our deal -
14   - us, as a broker-dealer, we don't have any sort of 
15   advantage over our liquidity providers. 
16             MR. HARRIS:  I wasn't suggesting you had an 
17   advantage.  I was just wondering if you have a market 
18   maker who is posting liquidity, does it only stand -- I 
19   mean, is the system canceling the order automatically? 
20             MR. ITURRINO:  I'm not quite sure I understand 
21   your question.  So I believe the all or none protocol is 
22   also adopted, right, by our traders.  Or the way they 
23   quote, right?  So if they have a larger position, they 
24   may have a minimum quantity, things of that nature. 
25             MS. MAHER:  Just two quick comments.  In the 
0220
 1   conversation about last look, I think it would be 
 2   important to discuss whether or not or examine whether 
 3   the auction process could serve to meet the broker-
 4   dealer's best execution responsibilities, if in fact it 
 5   was put into an auction and the dealer didn't 
 6   participate. 
 7             And I think it's interesting, Tom, if I 
 8   understood your comments, I think it would be interesting 
 9   to quote fixed income without the financial adviser 
10   commission.  And it might become a place where the data 
11   in the marketplace makes a lot more sense for retail 
12   investors.  They could understand what the broker-
13   dealer's markup was and then the commission could be 
14   analyzed in a separate line item. 
15             MR. McVEY:  Great.  Well, join me in thanking 
16   our panelists for a great discussion today.  And clearly, 
17   technology and E trading have a much bigger role to play 
18   in retail fixed income than ever before and there are 
19   still some open issues and areas for improvement.  So 
20   thanks very much for helping the committee with that. 
21             CHAIRMAN HEANEY:  Thank you, Rick.  And I would 
22   like to thank Rick again and reiterate our thanks to both 
23   this panel and the panels that we had previously in the 
24   day. 
25             Just a couple of housekeeping items.  We talked 
0221
 1   earlier about the subcommittees having short-term and 
 2   long-term not only deliverables but conversation pieces 
 3   and deliverables.  So between now and July, we will have 
 4   more work on the subcommittees. 
 5             I did want to raise just one other topic which 
 6   is, as FIMSAC members, if you think about topics that we 
 7   haven't tackled yet in these three subcommittees, things 
 8   either adjacent, related or something very different, 
 9   let's start communicating as a group perhaps on what we 
10   think July and September could look like in terms of 
11   other items.  We don't have to open that for business 
12   now.  
13             But even if it's through email, let's start 
14   thinking about the topics by which after meeting one we 
15   created these subcommittees.  And this has been a 
16   phenomenal day and phenomenal work.  But just going 
17   forward, as we think about July and September topics, as 
18   well. 
19             And as you know, July 16 is our next meeting.  
20   Hopefully, the snow will stop in D.C. and everywhere else 
21   by then and we'll actually have warm weather for that. 
22             Any other topics, any other comments before a 
23   movement to adjourn?  Any other comments? 
24             (No response.) 
25             CHAIRMAN HEANEY:  All right, just a motion to 
0222
 1   adjourn our FIMSAC meeting.  Anyone second? 
 2             Have a great day. 
 3             (Whereupon, at 3:52 p.m., the meeting was 
 4   adjourned.) 
 5                           * * * * * 
 6    
 7    
 8    
 9    
10    
11    
12    
13    
14    
15    
16    
17    
18    
19    
20    
21   
22   
23   
24   
25   
0223
 1               PROOFREADER'S CERTIFICATE 
 2                                 
 3   In the Matter of:  SEC FIXED INCOME MARKET STRUCTURE    
 4                      ADVISORY COMMITTEE 
 5   File Number:       GC-0409 
 6   Date:              Monday, April 9, 2018 
 7   Location:          Washington, D.C. 
 8    
 9        This is to certify that I, Christine Boyce, (the 
10   undersigned) do hereby swear and affirm that the attached 
11   proceedings before the U.S. Securities and Exchange 
12   Commission were held according to the record, and that 
13   this is the original, complete, true and accurate 
14   transcript, which has been compared with the reporting or 
15   recording accomplished at the hearing. 
16    
17    
18   ____________________________    _______________________ 
19   (Proofreader's Name)                (Date) 
20    
21    
22    
23     
24   
25