0001 1 U.S. SECURITIES AND EXCHANGE COMMISSION 2 3 4 5 6 MEETING OF THE 7 FIXED INCOME MARKET STRUCTURE 8 ADVISORY COMMITTEE 9 10 11 12 13 14 15 Monday, April 15, 2019 16 9:29 a.m. 17 18 19 20 21 22 23 U.S. Securities and Exchange Commission 24 100 F Street, N.E., Washington, D.C. 25 Multipurpose Room 0002 1 PARTICIPANTS: 2 Jay Clayton, Chairman 3 Hester Pierce, Commissioner 4 Elad L. Roisman, Commissioner 5 John Roeser 6 David Shillman 7 Lizzie Baird 8 Brett Redfearn 9 Michael Heaney 10 David Dimitrious 11 Tom Eady 12 13 Ahmed Abonamah 14 S.P. Kothari 15 Matt Andresen 16 John Bagley 17 Horace Carter 18 Gilbert Garcia 19 Alex Ellenberg 20 Larry Harris 21 Scott Krohn 22 Ananth Madhavan 23 Lynn Martin 24 Amy McGarrity 25 Rick McVey 0003 1 PARTICIPANTS (CONT'D): 2 Suzanne Shank 3 Larry Tabb 4 Sonali Theisen 5 Brad Winges 6 Mihir Worah 7 Jude Arena 8 John Cahalane 9 Peg Henry 10 Justin Land 11 Chris Kendall 12 Anthony Liotti 13 Marshall Nicholson 14 Peter Sirbu 15 Tom Deas 16 Ed Fitzpatrick 17 David Knutson 18 Pat McCoy 19 Julian Potenza 20 Tom Wipf 21 Kumar Venkataraman(via phone) 22 Rachel Wilson(via phone) 23 Elisse Walter(via phone) 24 Carole Brown(via phone) 25 0004 1 C O N T E N T S 2 PAGE 3 Welcome and Opening Remarks 5 4 5 Presentation on Block Pilot and 16 6 Reference Data Service Proposal 7 8 Draft Recommendation on Pennying 35 9 in the Corporate Bond and Municipal Securities 10 11 Draft Recommendations on Certain 86 12 Principal Transactions with Advisory Clients 13 14 Updates from the Credit Ratings, ETFs and 159 15 Bond Funds, and Corporate Bond Transparency 16 Subcommittee 17 18 LIBOR Transition: Implications for the 163 19 Corporate Bond and Municipal Securities Markets 20 21 Adjournment 223 22 23 24 25 0005 1 P R O C E E D I N G S 2 MR. HEANEY: Good morning. I believe we have a 3 quorum, so we'll get the meeting started. In addition to 4 the members that we have here in the room Elisse Walter, 5 Rachel Wilson, Carole Brown, and Kumar, I believe, are 6 all on the phone. 7 Good morning. Are you all with us? 8 MS. WALTER: I am 9 MR. VENKATARAMAN: This is Kumar. I am here. 10 MR. HEANEY: Thank you, Kumar. 11 Rachel? Perhaps Rachel is joining us in a bit. 12 Thank you for joining us for the SEC Fixed 13 Income Market Structure Advisory Committee's first 14 meeting of 2019. I know it's been a little while since 15 we've all met in person as, unfortunately, we do not have 16 -- were able to meet in January. 17 A big thank you to all the members for 18 remaining so engaged since our last meeting in October, 19 and patient as we rescheduled this meeting for April. 20 I'll begin by welcoming Chairman Clayton and 21 ask him to make his opening remarks. 22 CHAIR CLAYTON: Thank you, Michael. Welcome 23 FIMSAC members, our guests for today, our staff, and 24 those of you in the audience. I'm going to depart from 25 my prepared remarks. I was going to talk about today's 0006 1 agenda and I know that Brett is going to go over the 2 agenda in some detail. I also have in my prepared 3 remarks what I'd like to focus on today, which is the 4 work of this committee to date. 5 I want to say that I'm extremely pleased. This 6 committee has already put forward five recommendations. 7 Today's agenda has the same theme as those 8 recommendations, which is they are pragmatic, thoughtful, 9 and actionable. That is all we could ask for from this 10 group of people. And it is incumbent upon us to then 11 look at those recommendations and make decisions, which 12 is what I intend for this commission to do. 13 I want to say I greatly appreciate that. 14 Thinking about that, and speaking with Michael, where do 15 we go from here? And I want the people on the committee 16 and those on the staff to know that I recognize that most 17 of you, all of you, have very important and taxing day 18 jobs and you've given a lot to this committee already. 19 My vision for this committee going forward is 20 for it to continue with the recommendations that it has 21 on tap. The time for this committee as it was originally 22 set was to sunset in two years. I think that's a bit 23 short-dated. I'd like your input on what the appropriate 24 date is. 25 That said is -- that said, things should not go 0007 1 on forever. You cannot ask people to give this much of 2 themselves for an indefinite period of time. But I am 3 entertaining thoughts on how long this committee will go 4 on, and I'm thinking some time into mid to late 2020, so 5 that you can see your recommendations through. But I 6 would like to float that idea, get your input, get the 7 input from my fellow commissioners. But all I can say is 8 I want to make sure you have a chance to see the work 9 you've initiated through, and I greatly appreciate it. 10 Thank you very much. 11 MR. HEANEY: Thank you, Chairman Clayton and 12 we'll now hear from Commissioner Roisman. 13 COMMISSIONER ROISMAN: Thank you to all the 14 FIMSAC members for contributing your time today, and 15 during the subcommittee meetings and discussions up to 16 this point. 17 I also want to thank the SEC staff who support 18 FIMSAC. Despite several months having passed since our 19 last meeting, the agenda today makes clear that -- remain 20 focused and busy. 21 I'll keep my remarks brief. So rather than 22 address each of the topics on the agenda, I will simply 23 highlight a few general themes that are important to me, 24 and I think you'll find relevant to the discussions 25 today. 0008 1 First, I'm keenly interested in ways we can 2 improve the transparency in all our markets, but 3 especially for fixed income, where opacity still seems to 4 be the norm, although I recognize that this is shifting. 5 Second, competition is very important to our 6 markets, and I'm always eager to find ways to remove 7 impediments that may discourage entry, participation, and 8 confidence. 9 Third, providing greater flexibility to market 10 participants is something I'm always open to considering, 11 especially when doing so helps the markets and investors, 12 and avoids impractical results. That said, you must 13 always be mindful of the potential for unforeseen risks 14 that even minor adjustments to existing rules and 15 protections may cause. 16 Finally, our U.S. markets are interconnected. 17 But as we all know, our markets are part of a larger 18 global financial ecosystem. I'm encouraged that we have 19 so many leaders and representatives of the markets in 20 this room, and that you are focused on how our domestic 21 and global markets interact. This is especially true at 22 a time when geopolitical and macro economic shifts are 23 occurring, thus creating new financial and systemic risks 24 and the need to understand and plan for contingencies to 25 assure stability and continuity. 0009 1 With that, I look forward to discussions today. 2 Again, thank you, Chairman Clayton, FIMSAC members, and 3 the Commission staff. 4 MR. HEANEY: Thank you Commissioner Roisman. 5 Next I'd like to turn it over to Brett Redfearn, director 6 of the Division of Trading and Markets and the 7 committee's designated federal officer. 8 MR. REDFEARN: Thank you, Michael. I would 9 also like to welcome everyone to our first FIMSAC meeting 10 of 2019. 11 Let me briefly introduce my colleague sitting 12 here with me today. To my right, from the Division of 13 Trading and Markets, we have Lizzie Baird, one of our 14 deputy directors. To Lizzie's right are Dave Shillman 15 and John Roeser, both associate directors in the Office 16 of Market Supervision. To my left we also have Ahmed 17 Abonamah, senior counsel to the director of the Office of 18 Municipal Securities; and S. P. Kothari, chief economist 19 and director of the Division of Economic and Risk 20 Analysis, or DERA. 21 Before we get started I need to remind everyone 22 that the views expressed here are those of the speaker, 23 and do not necessarily reflect those of the commission, 24 any commissioners, or any other members of the staff. 25 Like the chairman and Michael, I really just 0010 1 want again to personally thank all the members of the 2 committee for all of their hard work, and devoting so 3 much time to the committee, and coming up with so much 4 productive output. FIMSAC has been very productive, 5 producing five recommendations last year, as the chairman 6 mentioned, and today coming forward with three additional 7 recommendations. 8 Your recommendation so far have already 9 resulted in concrete actions by FINRA recently, the most 10 recent of which just last Friday. 11 Regarding your recommendation to conduct a 12 pilot program to assess the market impact of public 13 transparency for block-sized corporate bond trades, FINRA 14 just last Friday issued a regulatory notice soliciting 15 public comment on a pilot that would delay the 16 dissemination of trade information for certain large- 17 sized transactions and corporate bonds. I encourage 18 market participants to review this notice, and to 19 substantively respond to FINRA's request for comment. 20 FINRA also has filed a proposed rule change 21 with us, citing your recommendation to create a new issue 22 reference database for corporate bonds. The rule 23 proposal is now out for public comment, and interested 24 market participants should review the proposal and submit 25 any comments that they have on this important issue. 0011 1 This morning we will hear from FINRA staff on both of 2 these proposals. 3 Moving to the present, today we will consider 4 recommendations from the Technology and Electronic 5 Trading Subcommittee and the Municipal Securities 6 Transparency Subcommittee. These recommendations reflect 7 the FIMSAC Subcommittee's continued focus on ensuring 8 that the regulatory regime governing the corporate bond 9 and municipal securities markets provides the appropriate 10 protections for investors, and particularly retail 11 investors. 12 First, the Technology and Electronic Trading 13 Subcommittee will discuss a preliminary recommendation 14 concerning the use of last-look, often referred to as 15 pennying, in requests for "auctions" on electronic 16 trading platforms. An RFQ auction is commonly used -- is 17 a commonly-used trading protocol to facilitate retail 18 investor transactions in both corporate bonds and 19 municipal securities. 20 Second, the Municipal Securities Transparency 21 Subcommittee will discuss two preliminary 22 recommendations. 23 The first recommendation would facilitate 24 investment advisor trading in a principal capacity with 25 advisory clients to enable the client to purchase a 0012 1 negotiated new-issue municipal bond. The second 2 preliminary recommendation would facilitate investment 3 advisor trading in a principal capacity with advisory 4 clients when the client is liquidating a municipal bond 5 from its portfolio. 6 Your consideration of these recommendations 7 today is further evidence of this committee's continued 8 commitment to advancing recommendations to the 9 commission. 10 Additionally, this afternoon the committee will 11 hear from a panel of market participants on the 12 implications of the current global transition from LIBOR 13 for both the corporate bond and municipal securities 14 markets. Last year Chairman Clayton highlighted the 15 transition from LIBOR as a key market risk that we are 16 monitoring at the Commission. And at your last meeting 17 several of you noted that this was a topic that would be 18 worth highlighting for FIMSAC. I'm glad to see it on 19 today's agenda. 20 It is critically important for all bond market 21 participants -- dealers, investors, and issuers -- to 22 devote their time and effort to plan for this important 23 transition. 24 The Commission participates as an ex officio 25 member on the Alternative Reference Rate Committee, or 0013 1 ARRC, which is a group of private sector market 2 participants and official sector entities convened by the 3 Federal Reserve Board and the Federal Reserve Bank of New 4 York to help ensure a successful LIBOR transition. 5 Last week I participated in a roundtable 6 discussion on reference rates hosted by the Financial 7 Stability Board here in Washington, with representatives 8 from regulatory bodies and market participants from 9 around the world. The roundtable provided an opportunity 10 to reflect on the progress made so far, and to highlight 11 what additional work is needed to facilitate a successful 12 transition from LIBOR. 13 I look forward to the panel discussion on this 14 topic this afternoon, which will include individuals that 15 participate on the ARRC, as well as -- and hearing key 16 committee members' views on this important topic, 17 particularly as they concern our cash bond markets. 18 We're very pleased to be part of a much larger effort to 19 raise awareness and increase focus on this particularly 20 important issue. 21 Again, thank you for continuing to devote so 22 much time to this committee. This committee's input 23 continues to play an important role in our policy-making 24 process, and I look forward to today's discussions. 25 And with that, I turn it back over to you, 0014 1 Michael. 2 MR. HEANEY: Thank you, Brett. 3 Before we get started with today's agenda I 4 just wanted to take a minute to again thank all the 5 FIMSAC members for the engagement and the work on this 6 committee. You continue to devote a tremendous amount of 7 time to participate on subcommittee calls, which have 8 resulted in actionable recommendations to the SEC on 9 fixed income market structure. 10 I can recall sentiments expressed at our 11 inaugural meeting in January 2018 describing the U.S. 12 capital markets as "the most functional and efficient 13 globally," and yet FIMSAC was tasked to look for methods 14 to improve market structure, with the goal of enhancing 15 liquidity, transparency, and efficiency. 16 As we move through today's agenda and our 17 latest recommendations to the Commission, it's noteworthy 18 to take stock of FIMSAC's accomplishments: as previously 19 mentioned, five recommendations passed by FIMSAC and put 20 forth to the Commission for its consideration; an 21 additional three recommendations to potentially be voted 22 on today; and several more in the pipeline in 2019. 23 As we enter the homestretch of the committee's 24 tenure -- maybe not quite the homestretch, based on what 25 we heard today -- I encourage all of us to redouble our 0015 1 efforts, take this unique and important opportunity to 2 identify and debate issues within fixed-income markets, 3 so that ultimately we leave a legacy of fundamentally 4 enhancing and meaningfully improving the fixed income 5 markets in the United States. 6 FIMSAC members not only have the experience to 7 influence change within the subcommittees and at FIMSAC 8 but perhaps, more importantly, given the seniority of 9 this group, to influence change within the firms and 10 organizations that we all represent. 11 For our first agenda item we'll hear from 12 Jonathan Sokobin and Alex Ellensburg on two recent 13 proposals FINRA has put out for public comment that, as 14 Brett indicated, reference our recommendations. The 15 first is a notice outlining a potential block pilot for 16 corporate bonds, and the second is a rule filing with the 17 SEC to create a corporate bond new-issue reference data 18 service. 19 I want to take this opportunity to thank FINRA 20 for its receptiveness to our recommendations, and commend 21 Alex and Jonathan in particular on their tremendous 22 efforts. We all look forward to further engagement with 23 FINRA during the policy-making process. And I especially 24 want to thank Alex for subbing in last minute for Tom 25 Gira, who is dealing with some traffic delays out of 0016 1 Chicago. But, hopefully, Tom will be able to join us 2 later. 3 Before I turn it over to Alex and Jonathan, 4 just as a reminder for us, as FIMSAC members and people 5 around the table, for our discussions we ask that each 6 committee member continue the practice of raising the 7 name tag so that I know if there's an indication of a 8 question or a comment -- again, with the effort of trying 9 to get everybody to be able to participate. 10 So, Alex and Jonathan, thank you very much for 11 joining us today, and the floor is yours. 12 MR. SOKOBIN: Good morning, thank you very 13 much. Thank you for the opportunity to participate in 14 today's meeting. 15 As Brett mentioned, at the request of the SEC 16 FINRA republished this past Friday a proposal for a pilot 17 to study the impacts associated with the FIMSAC's 18 recommendations in block-sized corporate bond trades. 19 I'm here today to describe the pilot as outlined in the 20 regulatory notice, seek your feedback, and answer any 21 questions I can have -- I can at this point. 22 When we started the effort to write this pilot 23 we had a few goals in mind. Our first and foremost was 24 to address the research question raised by the FIMSAC 25 recommendation, which is to determine whether changes in 0017 1 transparency may improve market quality and, in 2 particular, increase the frequency or size of block 3 trades. We wanted to create a design which permits 4 reliable tests of causal inference, and as simple as 5 possible in design, implementation, and interpretation. 6 And finally we sought to describe potential impacts on 7 the various market participants, including central and 8 non-central dealers, large and small institutional 9 clients, retail investors, and those engaged in trading 10 derivative securities such as ETFs. 11 So let me turn to the pilot. The primary 12 features of the pilot are consistent with the FIMSAC's 13 recommendation. FINRA proposed a pilot with a duration 14 of one year, which can be terminated early, if necessary. 15 It includes all trace eligible corporate bonds including 16 144A and new issues, with a couple of minor exceptions. 17 It provides exogenous shocks to transparency by first 18 reducing transparency through a dissemination delay for 19 trades having a cap greater than or equal to 10 million 20 for investment grade bonds, and 5 million or greater for 21 non-investment-grade corporate bonds. 22 And second, it increases transparency by 23 raising the size cap from 5 million to 10 million for 24 investment grade, and from 1 million to 5 million for 25 non-investment-grade corporate bonds, such that the full 0018 1 size would be disseminated immediately for trades falling 2 below those caps. 3 There are, however, some differences from the 4 FIMSAC's recommendation. In the proposal we tried to 5 balance a tradeoff between good research design for 6 causal inferences and a desire to limit complexity. The 7 original FIMSAC recommendation incorporates two different 8 changes in the transparency regime, as we've mentioned, 9 that potentially have offsetting effects. 10 This choice of having these two offsetting 11 changes creates a few separate issues for us in pilot 12 design. Testing for both of those changes at the same 13 time is, in effect, a joint hypothesis at the specific 14 cutoff levels in the recommendation. It does not allow 15 us to understand the impacts of either hypothesis 16 directly, and typically researchers only test a change 17 indirectly when it is not possible, prohibitively 18 expensive, or unethical to do the direct test. 19 So, to assess the impacts more directly, we're 20 proposing a couple of changes. So the first is to add 21 two new test groups. The new test group one would apply 22 the impact on dissemination delay only at the current 23 reporting thresholds, and the new test group two would 24 increase the size cap -- the size dissemination cap with 25 no delay. There is a test group three which, in effect, 0019 1 reflects the FIMSAC's recommendation. 2 To permit to permit tests of causal inference, 3 we thought it important to propose a control group, 4 creating an effective baseline for measurement. We want 5 to employ a difference in difference methodology which is 6 a quasi-experimental technique used to understand the 7 effect of a change in policy. Key is that you need a 8 transparent, exogenous source of variation to determine 9 treatment assignment, and it helps because it can control 10 for omitted trends that are common to the to the entire 11 market in the cross-section and over time. 12 So to ensure that the control and test groups 13 are comparable, we need a criteria for assigning the 14 bonds to these different test groups. We are proposing a 15 stratification technique that had -- that uses a series 16 of dimensions, such as 144A versus non-144A, issue size, 17 age of issuance, and rating, which are all known to 18 affect the liquidity of a bond. 19 I will note that, in our conversation, SIFMA 20 has suggested an alternative assignment regime done 21 simply by assigning to test and control group based on 22 whether the CUSIP-assigned check digit is even or odd. 23 We ask about this and other alternatives as part of the 24 notice. 25 To maintain fairness of the pilot and avoid the 0020 1 possibility that only certain securities or issuers are 2 subject to transparency changes, we propose to rotate the 3 test and control groups halfway through the pilot. All 4 bonds initially assigned to the test group would become 5 control after six months, and vice versa, with the 6 exception of new bonds without sufficient trading 7 histories. 8 Finally, the proposal does not recommend a 9 change in the lagged release time of uncapped trade sizes 10 from six to three months, as was in the original FIMSAC 11 recommendation. We believe that implementing such a 12 change would have added complexity to the pilot, and that 13 the impact would be very difficult to evaluate, given the 14 pilot's proposed duration of one year with the rotation. 15 In developing this proposal we relied on a lot 16 of outside expertise. In the regulatory notice we cite 17 directly to the FIMSAC's recommendation, to peer-reviewed 18 academic literature, and to comment letters received by 19 the SEC as part of the FIMSAC process. In addition we 20 have received feedback from industry organizations, a 21 number of firms, and we've used FINRA's advisory councils 22 -- committees as a way to seek input from experts both in 23 fixed income markets and from a panel of academic experts 24 who could give us feedback on the design itself. 25 So what are our next steps? The published 0021 1 notice is out for a 60-day comment period. We welcome, 2 over that comment period and beyond, any opportunity for 3 feedback through comment letters, through meetings, 4 particularly any data or evidence that might inform 5 approach and design. We would then evaluate the comments 6 and feedback, including an assessment of the potential 7 economic impacts of the pilot on investors and other 8 market participants, whether the pilot will achieve its 9 stated goals, or if there are ways to improve the pilot 10 design. Based on that feedback, if there is a decision 11 to move forward, FINRA would then file a proposed rule 12 change with the SEC for approval, and with that would 13 come a second round of public notice and comment. 14 So I thank you for your time, and I'm happy to 15 take questions. 16 MR. HEANEY: Questions for Jonathan? 17 Lynn? 18 MS. MARTIN: How do you propose handling new 19 issues? 20 MR. SOKOBIN: So on the first day of -- after 21 secondary market trading, they would be assigned randomly 22 to one of the test groups or the control group. 23 MR. HEANEY: Sonali? 24 MS. THEISEN: Thank you. Obviously, a lot of 25 work has gone into putting forward a very thorough 0022 1 proposal here. And this is, you know, a topic that's 2 very near and dear to me, when -- I worked extensively on 3 the initial proposal to FIMSAC. 4 I have -- I do have a couple of questions about 5 the way in which the experiment groups would be 6 constructed. You mentioned a couple of different 7 features, which we agree with, about ratings, size of 8 issuance. In addition to that will you be taking into 9 account, like, the curve? So short end, and 10 intermediate, long end? 11 MR. SOKOBIN: Yeah. 12 MS. THEISEN: Okay. 13 MR. SOKOBIN: Though the goal here always is 14 going to be to do this as simply as possible. So there 15 are a lot of other variables we could have also included 16 -- 17 MS. THEISEN: As well as will you be sort of 18 dimensionalizing by sector? 19 So, like, TMT and financials, for example, 20 would be, you know, more liquid than utilities, 21 industrials, and pharma, retail -- look to kind of create 22 cross-sectional groups that represent -- I think the one 23 thing that, you know, when we were reading this over the 24 weekend that we were thinking about is that -- the idea 25 that there is, you know, certain cap structures that are 0023 1 obviously quite liquid and would have bonds that are 2 suitable. 3 Ideally, you'd love, you know, the same issuer 4 to have a bond that could fall into each of the four 5 groups. There is -- we think there's probably between 25 6 to 40, you know, cap structures that that's suitable for. 7 8 The other side of that, I think, the -- and the 9 control group that -- sorry, the experiment group, 10 rather, that would be the most tricky, particularly for 11 high yield in distressed names, we believe, is the 12 smaller cap structures falling into the 10 and 5 million 13 plus with no transparency bucket. Like, will -- you how 14 will you address kind of which size cap structures 15 volunteered each? 16 MR. SOKOBIN: So the goal is to have as good 17 and -- a distribution across the different test groups 18 and the control group, so that you can generate the kind 19 of results that permit some conclusion or not. 20 The first part of your question is whether we 21 had thought about distributing based on the sort of 22 industry of the issuer. We'd considered it. We did not 23 include it here because we just thought that that added a 24 layer of complexity in terms of the distribution. But 25 this is why the -- this is a regulatory notice out for 0024 1 comment, and we seek the feedback of people who live this 2 on a regular basis. 3 And so, just -- or more basically, the more 4 different ways you cut, the harder it is to get a sample 5 size that's large enough to generate statistical 6 significance. And so the balance is how do you find a 7 series of cuts that that are reasonably comparison and 8 give you enough observations that permit you to come to 9 some sort of reasonable conclusion about the impacts of 10 the different treatments. 11 MS. THEISEN: Thank you. If I could ask one 12 more question, when you mentioned that you proposed 13 halfway through the year flipping, is that all of the 14 experiment groups will go into the control and vice 15 versa? 16 MR. SOKOBIN: Yes. 17 MS. THEISEN: And so will you take some sort 18 of, you know, overlay of what has happened overall in the 19 market during those two times? In other words, if we are 20 in very different market conditions during those two 21 times, how will you account for that? 22 MR. SOKOBIN: So this is -- as I said, the 23 difference in different methodologies allows you to 24 compare the activities pre and during pilot across 25 control and test groups. 0025 1 So you can think of it as you would be -- if 2 the conditions were really different in the second half 3 of the year, you'd be comparing the control and test 4 groups in the second half of the year, basically -- 5 potentially separately from -- you could put a dummy in 6 just to sort of separate that from the first part of the 7 year, right? 8 We sort of -- we talk about in the pilot one of 9 the things that could potentially affect the first and 10 second half is how quickly market participants learn 11 about the change in the market structure due to the 12 pilot. If they learn very quickly in the first and the 13 second, then, you know, the difference won't be 14 significant, and we'll be able to compare across. 15 Effectively, there will be a series of statistical 16 techniques that will be used to ensure that the 17 comparisons are as good as possible. 18 MS. THEISEN: And one more question. The 19 actual dissemination for the prints that will be delayed, 20 how would those actually hit the tape, or how would they 21 be disseminated to TRACE? Would there be some sort of 22 identifier flagged that it was on a 48-hour delay, and 23 they would be able -- end of day or how -- like, what 24 would be the dissemination mechanism? 25 MR. SOKOBIN: Do you remember? 0026 1 MR. ELLENBERG: I think some of those details 2 are still open and subject to comment, but I think the 3 current thinking is that there would not be any 4 indicator; they would just arrive on the tape after 48 5 hour. 6 MS. THEISEN: Without a flag? 7 MR. ELLENBERG: Without a flag, but with time 8 of execution, so you -- 9 MS. THEISEN: With a -- with the original 10 timestamp -- 11 MR. SOKOBIN: Right. 12 MS. THEISEN: -- but without a flag? 13 MR. SOKOBIN: That's right. 14 MS. THEISEN: Thank you. Thank you. 15 MR. HEANEY: Other questions? 16 I would ask one if I could, Jonathan. Under 17 normal conditions -- and I get that there is a lot of 18 moving parts here, but what is the timing to get through 19 the comment period through to the SEC, and then start the 20 one-year trial? What's the average? Is it -- are we 21 talking about this a couple of weeks, or this is a couple 22 of months? 23 MR. SOKOBIN: Well, again, it depends upon the 24 comments. 25 I mean we understand the Commission's interest 0027 1 and your interest in doing this as expediently as 2 possible. Our efforts have been, I think throughout this 3 process, you know, to trying to meet that goal. Really 4 hard to give you a promise, particularly since I'm not a 5 lawyer. And as a non-lawyer at a lawyer's agency, it's 6 always dangerous to make promises about timing. But, you 7 know, I think we would do everything in our power to move 8 things along. 9 MR. REDFEARN: Michael, I'll just say that the 10 -- you know, the comment process is one where we 11 obviously encourage a lot of input on that. But we're 12 definitely talking months, not weeks, and I would suggest 13 that between FINRA's comment process and the SEC's 14 comment process, we're looking at least -- you know, at 15 least probably at least a six-month period before that's 16 all wrapped up. Probably longer than that. 17 MR. HEANEY: Appreciate it. Mihir? 18 MR. WORAH: No, no question from me. But I 19 just wanted to echo Michael and others that I really 20 appreciate the effort and work FINRA has put into this. 21 I read through the proposal. A lot of thought has gone 22 into it. And hopefully, after the public comment, we do 23 the pilot, you know, it gives us a conclusive answer on 24 changing the rules for large blocks, improve the 25 experience for participants in the fixed-income markets, 0028 1 improve transaction costs. So I look forward to moving 2 forward with this, and thank you for your work. 3 MR. SOKOBIN: Thank you. 4 MR. HEANEY: Other questions for Jonathan? 5 Scott? 6 MR. KROHN: So is it -- my understanding 7 correct that we won't proactively communicate what CUSIPs 8 are in a test group or not? 9 MR. SOKOBIN: No, we would. 10 MR. KROHN: We would? Okay. 11 MR. SOKOBIN: We would publish the set of 12 CUSIPs in each group. 13 MR. KROHN: Okay. 14 MR. SOKOBIN: So that everyone would know. 15 MR. HEANEY: I would just like to echo the 16 comments of others, and thank FINRA for all the time and 17 effort. 18 This group -- this subcommittee put a lot of 19 time into coming up with this proposal out of FIMSAC. 20 And as you read through the paper, it is -- I agree with 21 Mihir; it is incredibly well thought-out, in terms of 22 enhancing this and refining it for the better. 23 So we look forward to the start of the pilot. 24 I will just remind the FIMSAC members I did ask the 25 question of potentially when it could start. Just to 0029 1 kind of dovetail with Chairman Clayton's comments about 2 potentially going into 2020, so if this starts in six 3 months it will go -- the one-year period would run well 4 into 2020. And if we are looking to see things through, 5 that timing actually dovetails very well with what the 6 chairman was saying. 7 CHAIR CLAYTON: And I'll just say thanks, as 8 well. Terrific work. 9 MR. HEANEY: Alex, please. 10 MR. ELLENBERG: Good morning. Thank you for 11 the opportunity to be here. My name is Alex Ellenberg, 12 and I work in the office of general counsel at FINRA. 13 I'm here this morning to provide an update on our work 14 connected to the recommendation approved at the last 15 FIMSAC meeting that FINRA develop a centralized corporate 16 bond new issue reference data service. 17 On March 27th FINRA filed a proposal with the 18 SEC, based on your recommendation. That proposal was 19 published by the SEC in the Federal Register on April 20 8th, with a comment period that runs until April 29th. I 21 believe the proposal was also posted on the FIMSAC's 22 webpage as one of the background materials for today's 23 meeting. 24 The proposal we filed is substantially the same 25 as the FIMSAC recommendation. As recommended, the 0030 1 proposal leverages existing FINRA reporting requirements 2 for new-issue underwriters. Under the proposal, 3 underwriters must report additional information about a 4 corporate bond new issue before the first trade in the 5 security. FINRA will then disseminate that information 6 according to a fee structure that is intended to promote 7 broad distribution throughout the market. 8 Details of the proposal are included in the 9 filing that's been published for comment. So, for 10 purposes of this update, I thought I'd focus on a few 11 keynotes. 12 First, the FINRA proposal includes 23 of the 24 13 new issue data fields included in Schedule A of the 14 FIMSAC recommendation. The field we did not include is 15 calculation type, which is specific to one data vendor's 16 protocols and may not be readily available to all 17 underwriters that would be required to submit the 18 information to FINRA or to consumers of the data -- 19 although we note that we also think the same information 20 is available through the data fields that are included in 21 the service. 22 We also included seven additional data fields, 23 based on further outreach that FINRA conducted to a 24 cross-section of market participants. We included an 25 exhibit in our proposal that describes each of these 0031 1 fields in detail and our reason for adding them. But at 2 a high level we heard during outreach that these fields 3 are important for liquidity and risk assessment for new 4 issues. 5 Second, the proposal would require underwriters 6 to submit all of the new-issue data fields before a new 7 issue's first trade. We believe this will accomplish the 8 recommendation's goal for broad and uniform data 9 availability at the beginning -- at the commencement of 10 secondary trading for a new issue. 11 Third and last point I'll highlight today is 12 the proposed fee structure. We're proposing to provide 13 the data with unlimited use and redistribution rights for 14 6,000 a month, and for internal use only for $250 a 15 month. We thought this was a simple and equitable fee 16 structure that would support broad distribution of 17 information to all market participants. 18 As mentioned, details are available in the 19 proposal that is available on the Commission's website 20 and FINRA's website, and we encourage discussion here 21 today. We're happy to take any questions, and we also 22 encourage written comment to be submitted on the 23 proposal. 24 MR. HEANEY: Thank you, Alex. Questions? 25 Rick? 0032 1 MR. MCVEY: Just following on Mihir's comments 2 in the earlier recommendation, thank you to FINRA for 3 moving very quickly on this in a very practical way. 4 And I can tell you that, following the FIMSAC 5 deliberations last meeting, the feedback on this proposal 6 that I've received from investors, dealers, as well as 7 reference data providers has been universally positive. 8 So I think we're on to a very practical solution here 9 that will benefit all fixed-income market participants. 10 So thank you. 11 CHAIR CLAYTON: As you know -- I'll be a lawyer 12 now. This is currently in front of the Commission for 13 consideration, so I won't comment on the merits, but I 14 will comment on the process. 15 I'm grateful for how quickly you have moved on 16 this. Thank you. 17 MR. HEANEY: Any other comments or questions? 18 (No response.) 19 MR. HEANEY: Great. Thank you very much, 20 Jonathan. Thank you, Alex, for both presentations today. 21 Again, these proposals reflect diligent consideration 22 FINRA has taken on both of these two recommendations from 23 FIMSAC. We look forward to further engagement with 24 FINRA, as we have other proposals that may come to light. 25 At this point we would take a five-minute 0033 1 break, if we can. If you'd like to refresh coffee, and 2 then we will get going at 10:10. 3 (A brief recess was taken.) 4 MR. HEANEY: Thank you. We'll now turn to our 5 consideration of the recommendation from the Technology 6 and Electronic Trading Subcommittee on the use of 7 pennying in the corporate bond and municipal securities 8 markets. I'd like to turn it over to Rick McVey, this 9 subcommittee's chairman, to introduce the recommendation 10 and moderate the panel discussion. 11 Rick? 12 MR. MCVEY: Thank you, Michael, and good 13 morning, everyone. We're going to start with quick 14 introductions of the panelists today. But let me first 15 say thank you to the subcommittee for months' worth of 16 work deliberating on this topic that led to the 17 recommendation we're presenting this morning, and also 18 all of the outside guests that participated in our calls, 19 some of whom are panelists here this morning. 20 So let's start with Matt, and just get through 21 the introductions of the panel, and then I'll provide a 22 brief summary of the recommendation. 23 MR. ANDRESEN: I'm Matt Andresen from 24 Headlands. Just briefly, I think we're excited about 25 this recommendation because we have seen a lot of great 0034 1 improvement in the market structure in fixed income, 2 where -- you know, where there's only 3,600 U.S. 3 equities, and most of them don't trade like Amazon, you 4 can imagine the problems of price discovery when you have 5 a million different municipal securities. 6 But on the alternative trading system platforms 7 you actually have an auction process that works very well 8 when it works. And we actually see an average of over 9 seven respondents to a request for quote in municipal 10 securities. So that is a great result, it provides great 11 price transparency, and it also provides great best 12 execution for the initiating dealer. 13 Unfortunately, the results don't always work 14 perfectly. And specifically, the challenges being 15 addressed with this recommendation as around the practice 16 of last-looking the auction. So when a dealer initiates 17 a auction on behalf of a customer who's looking for a 18 trade, they will solicit this with a message that says, 19 "Hey anybody out here want to trade it?" Our firm and 20 others will respond to this. 21 And then there is a winner: the best price. 22 And Headlands responds to a lot of these messages, more 23 than anyone else, almost 15,000 of these a day. And we 24 win a lot, about a quarter of the time, which actually 25 is evidence of how competitive these are. 0035 1 Unfortunately, often these are pyrrhic 2 victories, because, even when you win you don't end up 3 getting the trade. We only get the trade about a third 4 of the time. And the other two-thirds of the time, 5 either there's no trade on the tape, or the initiating 6 dealer steps in front of our winning trade. And that -- 7 doing that and not participating in the auction 8 undermines the incentive for firms like Headlands to bid, 9 and therefore hurts the end customer. So we're excited 10 about this recommendation to address this practice and 11 maximize the value of these otherwise well-functioning 12 ATS auctions. Thank you. 13 MR. ARENA: Hi. My name is Jude Arena, and I 14 manage trading at Bank of America Merrill Lynch. I'm 15 happy to participate in the panel today. As a -- one of 16 the largest market-makers in the municipal securities, 17 and the largest underwriting desk, we think that there 18 are some recommendations that can be helpful. 19 I think that there is some confusion around the 20 notion of last-look versus the practice of pennying, so I 21 think parsing through some of that today will be helpful. 22 The act of pennying, where a dealer has a bid 23 wanted solely for the purposes of price discovery and 24 chooses to have an option to increase that bid 25 incrementally, I agree that that practice, if prevalent, 0036 1 would be damaging, or even the perception of that 2 practice would be damaging. I think it's not as 3 prevalent as some of the data might suggest, in that 4 large dealers, in particular, would like to reserve their 5 best bid for their own clients and their own retail 6 clients when there is a bid process. 7 So some of the data as, Mr. Andresen has 8 mentioned here, I think it's important to at least look 9 at the data from a perspective of dealers that want to 10 provide liquidity to their clients are always going to 11 give the best price to them. And if they're 12 participating outside of those six or seven bids that are 13 usually electronically or algorithmically generated, I 14 think it misses that. 15 The dealer community is standing behind its 16 customer base because it's good business to do so. And I 17 think that that's a large portion of the one-third of the 18 trades that you might be missing. 19 MR. BAGLEY: John Bagley, I'm the chief market 20 structure officer for the MSRB. The MSRB has been 21 looking at pennying for about a year. 22 We have a request for comment that was out in 23 September after significant outreach to broker-dealers, 24 ATSs, and broker's brokers, and buy-side people. The 25 request for comment really focused on pennying as a 0037 1 repeated practice and not an individual experience. We 2 received nine comment letters. The letters basically 3 said pennying, when it exists, is a problem. It said 4 that people don't really know how often it does exist, 5 and they also said that the acts of the MSRB should range 6 from anywhere from policies and procedures for the 7 dealers to forcing dealers to bid blind into an auction. 8 The MSRB is currently conducting analysis on 9 some data we received from FINRA as part of an inquiry. 10 We are also doing some additional analysis on data we 11 received from ATSs and broker's brokers. 12 I will say that the MSRB understands the 13 importance of the bid-wanted process, and we're looking 14 to make sure that -- we're focused on ensuring that any 15 potential guidance we might issue does not harm liquidity 16 and still enables dealers to meet their best execution -- 17 reasonable requirements. 18 MR. CAHALANE: Good morning. John Cahalane 19 with Tradeweb Markets. I run our retail distribution 20 business, both the ATS and the technology business, 21 providing front-ends to retail financial advisors, 22 investment advisors, and some of the retail self-directed 23 platforms. 24 Tradeweb welcomes the opportunity, and 25 continues our commitment to the future of fixed-income 0038 1 market structure. We believe in prudent regulation so as 2 to reinforce and promote great liquidity, robust and 3 orderly markets for traders and investors. We think this 4 is an important topic and one we look forward to 5 commenting on. 6 MS. HENRY: I'm Peg Henry, I'm deputy general 7 counsel at Stifel Financial. We are a very large 8 regional firm with a sizable retail component. 9 I want to say that I agree quite strongly with 10 my co-panelist from Bank of America. The -- I think it's 11 unfortunate that so much emphasis has been placed on the 12 word "last-looks," because I think a last-look is 13 actually mandated by MSRB rule G-30 to make sure that the 14 high bid and the bid wanted is, in fact, a fair price. 15 I didn't -- I was the key staff person on the 16 MSRB guidance from 2012 that the MSRB is now considering 17 extending to ATSs, and I'll talk a little bit later about 18 some of the thoughts I had going into that project. 19 However, I do want to draw one thing to your 20 attention. I think that the devil is in the details on 21 this project, and I think that it's important to focus on 22 exactly what is pennying, how you're going to measure 23 what -- you know, whether -- what the dealer's intent 24 was. But also, to more look more fundamentally is the 25 larger issue about the participants in this marketplace 0039 1 from the full service broker dealers, such as mine and 2 Bank of America, versus the algorithmic traders who have 3 no customers, and therefore no fair pricing duties. 4 Thanks. 5 MR. LAND: My name is Justin Land, I'm the 6 chief municipal strategist at Wasmer Schroeder and 7 Company. We are a $9 billion fixed income separate 8 account manager. About 80 percent of our assets are in 9 municipals. 10 The last 10 years have seen a lot of change in 11 the trading of municipal securities, in terms of their 12 growth in ATS, the growth in algorithmic bidding, the new 13 participation from foreign buyers, from hedge funds, and 14 so the nature of the market is growing. 15 As Matt mentioned, with over a million separate 16 securities, historically, it's been very difficult to 17 have efficiency of trading. However, with the 18 technological gains that we've seen over the past 10 19 years and we expect going forward, I look at it with the 20 end goal of having a fixed income market someday look 21 like equity market liquidity and pricing. And I'm 22 concerned that if certain participants, because of -- 23 whether it's pennying or last-look, lose confidence in 24 particular venues to trade, that my clients will not get 25 the best bid they might be able to get. 0040 1 MR. MCVEY: Thank you. And just a little 2 further background on the recommendation itself. The 3 subcommittee was concerned by the systematic use of last- 4 look on ATS platforms, known as pennying. So there was a 5 differentiation there of a type of last-look that we 6 viewed to be most problematic. 7 The role of the subcommittee is to consider the 8 impact of increased use of electronic trading on market 9 liquidity, fairness, and transparency. The systematic 10 use of last-look by certain bond dealers is an issue that 11 could impact market structure and fairness. 12 To follow on a few of the points that were made 13 already by the panelists, the committee also recognized 14 that broker-dealers executing orders on behalf of retail 15 customers have a best execution regulatory requirement. 16 And as a result, we were careful to differentiate 17 material price improvement for the customer from 18 systematic pennying. 19 Repeating pennying causes competitive harm, 20 even though it appears to benefit the customer with 21 improving price, in that it may deter aggressive pricing 22 by other dealers, and therefore have a negative impact on 23 competition. It also may limit responses by dealers who 24 don't want to serve as a price discovery function. And 25 it appears to give the submitting dealer unfair advantage 0041 1 in an auction. 2 As John mentioned, MSRB has expressed concerns 3 about this practice in the muni market. Both the 4 December 2012 notice to dealers regarding broker's 5 brokers using last-looks solely for price discovery and 6 the September 2018 requests for comment on draft 7 interpretive guidance on pennying by dealers on ATSs 8 dealt with this issue. 9 John will expand a little bit on the MSRB 10 findings and comments that were received to kick off our 11 panel discussion. 12 MR. BAGLEY: Sure. So the main thing I want to 13 talk about is in the MSRB request for comment we did 14 differentiate between last-look and pennying. And we are 15 we are looking at pennying, and we consider pennying to 16 be a systematic process whereby dealers are only 17 providing nominal improvement and internalizing trades. 18 Last-look, as Peg mentioned, as part of your 19 due diligence to make sure you meet your fair and 20 reasonable or best execution requirements before you hit 21 that button we understand. And we think that would lead 22 to probably a lot fewer internalizations than if they 23 did, probably internalizations with significant price 24 improvement, which is the responsibility. 25 And I think the one key thing that everybody 0042 1 has to remember is -- and Peg said it -- the dealer that 2 is putting the bid wanted out has the responsibility for 3 fair, reasonable, and best execution. And in the MSRB 4 we've been continually looking at that because we can't - 5 - if we do anything from a guidance perspective, we 6 certainly can't do anything that's going to make it more 7 difficult for them to meet that responsibility. You 8 can't have a rule -- a guidance going into place and, 9 order in order to meet that, you -- another one. So 10 we're taking this very seriously. 11 We're looking a lot of data because I think 12 that's the one piece that's missing, there's a lot of 13 incomplete data. But I think the commenters told us they 14 either bid less frequently, bid less aggressively if they 15 believe that pennying is prevalent on a platform. And 16 that just can't be good for investors, overall. So this 17 really will come down to continued outreach with 18 stakeholders, continued analysis on where we are to try 19 to come up with a solution that works for all of the 20 industry. 21 MR. MCVEY: Thanks, John. When we started this 22 deliberation and invited market participants to join the 23 subcommittee calls it became evident very quickly that 24 the same issues that the MSRB was concerned with on last- 25 look existed in the corporate bond market, as well. So 0043 1 that is why this subcommittee took up the issue and is 2 prepared to make the recommendation today. 3 I would like to point out that this is an 4 example of where greater regulatory coordination between 5 the SEC, FINRA, and MSRB would result in a more cohesive 6 regulatory framework with improved regulatory efficiency, 7 which was the reason for our first subcommittee 8 recommendation that we made last year. 9 I'd like to just read through the key elements 10 of the recommendation for the benefit of the full FIMSAC 11 committee. The recommendation is twofold. The FIMSAC 12 subcommittee requested the SEC make a statement 13 disapproving of the repeated use of last-look in either 14 the municipal or corporate bond markets on any electronic 15 trading venue, as the practice harms price discovery and 16 market efficiency. 17 The SEC should consider setting the clear 18 expectation that the use of last-looks should occur only 19 in the rare situation when the dealer does not receive 20 any reasonable responses to an auction request and/or 21 needs to use the practice to conform with its best 22 execution responsibilities. 23 The practice should not be part of the dealer's 24 everyday normal business practices. The SEC should 25 encourage dealers to have clear policies and procedures 0044 1 in place delineating when last-look may be used. 2 The SEC should encourage FINRA to publish a 3 request for comment on the use of last-look in the 4 corporate bond market similar to the September 7th, 2018 5 MSRB request for comment. 6 The SEC should encourage FINRA and MSRB to 7 expand its review of last-look to cover all electronic 8 trading venues, rather than limit specifically to ATSs. 9 Finally, the SEC should encourage FINRA and 10 MSRB to coordinate their respective final responses. 11 And with that, maybe, Peg, I could turn to you 12 -- since you have the benefit of a long history on this, 13 dating back to 2012 -- and ask for your thoughts on the 14 recommendation today dealing with the corporate bond 15 market and how you see it relating to the work that you 16 did in the muni market over the prior period. 17 MS. HENRY: Well, thank you. My history with 18 this actually goes back 3 years before 2012, because 19 that's how long it took to get that project off the dime. 20 As I mentioned, I staffed the 2012 guidance. It really 21 was somewhat ancillary to a larger project which was on 22 the regulation of broker's brokers, which would -- 23 resulted in the promulgation of the MSRB rule G43. 24 In the course of working on the project I 25 interviewed many market participants, including Mr. 0045 1 Cahalane, to my left, at Tradeweb, the other ATSs that 2 were in existence at that time, as well, as well as 3 numerous broker's brokers. And they all had a common 4 theme, and that was that there were some dealers -- not 5 all, but some -- that, on a regular basis, put something 6 out for the bid and then didn't hit the high bid, and 7 instead internalized the trade at a nominal amount. And 8 that was part of a pattern or practice of behavior on 9 their part. It wasn't just looking at it on a trade-by- 10 trade basis. 11 And I also did talk to investors who said, 12 "Well, I know that I was the high bid, and then I see it 13 trade away from me. And so why should I bid on that 14 platform?" So I understand and I agree that systematic 15 pennying, as we're calling it now, is a problem. 16 As I said, I think that it's unfortunate to use 17 the term last-look, because last-look is actually part of 18 your diligence responsibilities, as John mentioned, under 19 MSRB G30. 20 The notice, the 2012 notice, actually started 21 with a long discussion about the fact that both -- that 22 the NASD had found that there were a lot of instances 23 where the pricing wasn't fair because dealers were just 24 blindly taking the high bid and the bid wanted. So I 25 think that last-look is actually very important, from a 0046 1 fair pricing standpoint. That's different from pennying, 2 though. 3 And I do agree, if there is a systematic 4 pattern or practice of pennying, this internalizing at a 5 very nominal price improvement, that that is a problem 6 for the marketplace. I just encourage the regulators to 7 -- as I said, the devil is in the details -- to make sure 8 it doesn't have fundamental effects upon the provision of 9 liquidity and market-making by the full-service broker- 10 dealer firms. 11 MR. MCVEY: Great, thank you. And John 12 Cahalane, maybe I could turn to you as the operator of 13 one of the larger retail ATS systems to see what feedback 14 you've had from market participants on this topic. 15 MR. CAHALANE: Sure. Thanks, Rick. I'll start 16 by saying and agreeing with the other panelists that 17 Tradeweb does not believe that pennying is synonymous 18 with last-look. We think the establishment of clear 19 definitions of the two should be an important part of 20 this process. 21 Their particular markets, especially 22 municipals, where, given the limited liquidity of this 23 particular product and the inability of dealers to go 24 short in the municipal bond market, make look -- last- 25 look a very valuable protocol. 0047 1 Tradeweb doesn't, however, support the repeated 2 use of pennying. And in so, supports the subcommittee's 3 recommendation to try and correct any abuse, to ensure 4 that investors continue to have access to the broadest 5 liquidity pool possible. 6 There are, however, instances where the use of 7 last-look in what appears to be pennying can be valid. 8 And I think most often we see it in crossing of bonds 9 inside of a particular firm. A retail financial adviser, 10 while trying to establish the market for a thinly-traded 11 security to sell for a particular client, also needs to 12 establish a price where they may sell that bond back to 13 another one of their own clients. 14 So the use of venues, voice brokers, and other 15 market participants is particularly helpful in trying to 16 establish the correct price for that client across bonds 17 between the bid and the offer. Tradeweb believes it is 18 worthwhile that, during the examination of the costs of 19 pennying -- or this crossing of bonds be a part of the 20 examination of whether this practice is sometimes 21 actually not abusive, but saves the investor quite a bit 22 of money in the cost of the purchase of the securities. 23 Tradeweb also believes that any new regulation 24 should not only apply to electronic trading venues, but 25 also the voice brokering market, as well. We note that 0048 1 it's probably much more difficult to see systematic 2 pennying in the voice market than in the electronic 3 market, but should be looked at. And we do support 4 looking across -- broadly across asset classes in the 5 credit markets, as well as in municipal markets. 6 From a color or client reaction perspective, I 7 think, you know, we've heard from Peg and her firm, as 8 well as liquidity providers that, while a systematic use 9 of pennying would hurt liquidity, that there are valid 10 reasons why firms need last-look and potentially at times 11 to stand in front of a trade. 12 I think many of the retail firms believe 13 they've spent a significant amount of time and investment 14 acquiring these clients, they spend a significant amount 15 of time servicing these clients. And at times they 16 believe they'll have, at the other side of that 17 transaction, the right client to sell those bonds to. 18 And allowing all of those transactions, whether it 19 benefits our ATS or not, to happen against an external 20 liquidity pool versus internalizing and keeping the bid 21 to offer spread tighter, there is a lot of value for the 22 retail investor in the cost of the transaction from the 23 bid to the offer by internalizing some of those 24 transactions. 25 That being said, we would never want to have 0049 1 firms like Headlands not wanting to participate in these 2 auctions or providing offer-side liquidity to the markets 3 because that pool of liquidity is also extremely valuable 4 to marketplaces like ours. 5 MR. MCVEY: Great. Thank you. 6 Matt, maybe I can turn to you. You have been 7 keeping statistics on the auctions that you believe to 8 have had the best bid or offer, and where you actually 9 win the trade. 10 I guess what I'd ask you to expand on is do you 11 have a way of determining how much of that is a result of 12 systematic pennying, and how much is legitimate price 13 improvement or other reasons that may be taking place in 14 those auctions. 15 MR. ANDRESEN: Well, thank you, Rick. I'll 16 leave it to the listener to determine and make their own 17 qualitative judgment about what seems systematic and what 18 isn't systematic. Just from our experience, which -- 19 again, we respond to more of these auctions than any 20 other firms -- when we go and we win a trade, we only get 21 a trade about a third of the time. So the other times 22 either no trade goes up on the tape, which Peg talked 23 about from -- being addressed in other guidance. 24 But then let's just look at the times where we 25 win the auction and a trade goes up and we don't get it. 0050 1 Like, what happened to it. So it is suggested that 2 pennying some nominal amount of price improvement was 3 given for the away trade. But, in fact, almost half the 4 time the trade goes up away, it goes up at a worse price 5 than our winning bid. So I think, in those cases, the 6 clients would have killed to get pennied, and they 7 certainly would have been happy with our price or the 8 other winning prices from the other participants. But 9 instead, they end up actually getting a worse price. 10 So, given the scale of auctions we're talking 11 about here -- as many as 15,000 in a day, and you're 12 talking about a third of those being winners, and only a 13 third of those trading -- you're talking about a serious 14 market impact on the end client. So I believe that 15 nearly half would be a systematic application of the 16 process. 17 MR. BAGLEY: And can I just respond to that? 18 I think the one thing you have to take in 19 consideration is this isn't -- this is not potentially 20 rampant, not meeting your best execution requirement. 21 The piece of information that Matt doesn't have -- and 22 neither does anybody else -- is what sales credit a 23 financial advisor worked for on the sale and that client. 24 So, Matt, you agree that even if you guys were 25 high bid, the execution from the client could be at a 0051 1 lower price than your bid, right? 2 MR. ANDRESEN: They can make the price whatever 3 they like. I'm just pointing out that the client would 4 be better off if the dealer in that case acted as agent - 5 - 6 MR. BAGLEY: But -- 7 MR. ANDRESEN: -- and put the trade up at the 8 best price. 9 MR. BAGLEY: Right. I'm just pointing out that 10 financial advisors are entitled to compensation, and that 11 comes out of any bid, whether it's internal or external. 12 MR. MCVEY: Jude, maybe you can follow on Peg's 13 comments about the broker-dealer responsibility for best 14 execution with your retail clients, and the efforts that 15 you go to during the electronic trading process to comply 16 with those. 17 MR. ARENA: Yes, certainly. I do think -- I'm 18 happy that it's being pointed out here there is a big 19 difference between last-look, or having information 20 available to you at some point prior during a bid wanted 21 process that is in furtherance of our duty of best 22 execution. 23 So I am in agreement, as well, that systematic 24 pennying should be a violation of rules. 25 During our process we put up a bid to our 0052 1 client when we feel we've met the hurdle of our various 2 duties to that client. We don't think it ends there, 3 either. We don't think that we put up a bid to our 4 client, many of whom have told us they value flexibility, 5 and tell them, "We've met our duty, you have X amount of 6 time to respond to it or else we have to go through this 7 process again." So generally, we put up a bid to them 8 once we feel we've initially met the duty, but we think 9 reasonable diligence for best X goes beyond that. 10 So it could be market movement. It could be 11 bonds trading that are relevant to that bond, or -- which 12 happens occasionally. It could be an unsolicited 13 interest for that for that bond in that process that 14 happens after the formal process, in which case we now 15 acknowledge that a better bid is available for our 16 client. We're always going to give our client the 17 benefit of that better bid. 18 I think that -- to John's point about 19 responding and to the data that Mr. Andresen has -- is 20 that, yeah, comparing -- the bid that's being compared is 21 the best external bid less a fair desk spread that's 22 disclosed via the markup markdown rule. So there are 23 rules in place. You have the fair and reasonable, you 24 have best ex, you have markup/markdown that go onto the 25 client's statement. 0053 1 So I think that, as liquidity has actually 2 improved in the space pretty dramatically in the last 3 several years, it's not surprising to me that the covers 4 of the competitiveness has tightened up a lot. I would 5 expect that trend to continue. 6 I think that there is enough tools with respect 7 to G17 fair dealing -- I've long assumed that any sort of 8 systematic pennying is, common sense, a violation of fair 9 dealing. So any sort of any sort of systematic or even 10 intermittent pennying, I think, should be a rules 11 violation, and maybe some language about fleshing that 12 out would be helpful. 13 I think when you start to get into a bright- 14 line rules with respect to either process or with respect 15 to a safe harbor away from pennying, I think, is very 16 dangerous. From a prospect perspective, you lose some of 17 that flexibility that our clients want, and then you, you 18 know, get into this thing where you have a process, the 19 bid is only good for as long as that process allows, and 20 it takes away at least some of our ability to exercise 21 judgment when we're presenting a bid to a client and 22 updating it, if needed, which doesn't happen very often. 23 As far as the bright line with respect to a 24 safe harbor, I think that's very dangerous in that I 25 agree that pennying systematically is bad, it could have 0054 1 a chilling effect on the market. Codifying an amount 2 that we're going to agree is not pennying, I think, is 3 actually just as bad. If we say, "Well, if you improve a 4 bid by $.50 or $1, or even a quarter of a point, now you" 5 -- I think that would have more of a chilling effect on 6 businesses like Headlands, because now you know the 7 originating dealer has a free option. And then they get 8 negatively selected, the market rallies during the 9 process, and while the safe harbor says if it's a buck 10 fifty or two bucks then dealers will be left with the 11 free option that I think would have a bad effect on the 12 market. 13 So I think it's -- bright-line rules, I think, 14 are dangerous here on both process and on price, because 15 I think it will have unintended consequences, and I think 16 that there are tools that are already in place that can 17 be built around -- whether fleshing out a little bit more 18 about what constitutes pennying. But I think pennying in 19 any form is bad, but dealers need to have access to the 20 information to be able to fill -- fulfill our duties to 21 our clients that are not sophisticated muni market -- 22 MR. MCVEY: Thank you. 23 And, John, I know MSRB dealt with this bright- 24 line question, as well, and tried to -- tried your best 25 to have a definition for systematic pennying that we're 0055 1 all thinking of here. Maybe you could expand on that a 2 little bit. 3 MR. BAGLEY: Sure. I mean if -- the comments 4 were very clear. A lot of bright-line standards concern 5 people, just for the reasons Jude mentioned. So I think 6 that, you know, we have to -- that's why we need more 7 data, and that's why we're doing more analysis. So if we 8 can try to get to what this is -- but I think that the 9 more prescriptive we get, you know, could be more harmful 10 than good, in my opinion, as we go forward. 11 So I think it's really a process we have to go 12 through and understand what it is. And I think a rule 13 can be effective, even if it doesn't have a dollar amount 14 or a percentage, or things like that, because, you know, 15 I think everybody acknowledges that to internalize a 16 trade for various reasons -- and you couldn't come up 17 with an exhaustive list; there's plenty of reasons why 18 firms might need to do that. But if you can eliminate 19 the behavior where it's going on consistently, and where 20 you see it, you know, a large percentage of the time, I 21 think that would go a long way to help liquidity in the 22 marketplace and encourage more people a bit. 23 MR. MCVEY: Justin, expand a little bit on your 24 comments about ensuring competition and on electronic 25 venues for orders, and your observations of what you 0056 1 observe today in terms of what might fall under 2 systematic, which is -- versus what is true improvement 3 and price for best ex. 4 MR. LAND: Sure. And first, before I start 5 that, I do agree that systematizing, or putting some sort 6 of exact price on what pennying is does create a free 7 option, and that would potentially reduce liquidity over 8 time, because people will game the system. 9 I'm glad the comments about crossover trading 10 being potentially part of the statistics around pennying, 11 that, you know, Matt brought up -- was brought up -- our 12 firm does not engage in crossover trades. I know many of 13 my friendly competitors do. 14 I think one of the issues is that the 15 methodologies around determining the price for cross- 16 trades are all over the board. And I know, from our 17 point of view as a firm, if we're looking at an asset 18 manager we know has a high probability of crossing bonds, 19 that we generally will not bid those lists, just because 20 the likelihood of us taking the time and actually buying 21 securities is pretty low. 22 In a normal functioning market where things are 23 going well, it's probably not a big disservice to the 24 client. But in either illiquid securities or highly 25 volatile markets, that bond may being -- may be being 0057 1 crossed at a much lower price than if it had been truly 2 competitive. So there's a piece that, you know, having 3 better guidance around determining prices for crossing of 4 bonds would be helpful, I think, for everyone. 5 For us, you know, we increasingly trade on 6 electronic exchanges. We increasingly -- and talking to 7 many of our competitors -- are putting a lot of money 8 into technology around getting away from kind of the old 9 pick up the phone and call a guy to buy a bond. 10 The only way that works, though, in order for 11 us to lower our fees for our clients and improve 12 performance for them, is to have confidence that, when we 13 go out in the marketplace electronically and efficiently, 14 we can be pretty confident that if we're the best price 15 we get to buy bonds. Thank you. 16 MR. MCVEY: Maybe just one last question for 17 any of the panelists before we open it up to questions 18 from the rest of the FIMSAC committee. 19 But when we started out we talked about the 20 similarities in this practice in the corporate bond 21 market to the muni market that MSRB and John and his 22 colleagues had been working on. But if anyone has any 23 comments they'd like to make on how the practice is 24 similar and where there might be important differences 25 between the corporate bond market and the muni market, 0058 1 please feel free to speak up. 2 Good, how about for the rest of the committee? 3 Are there any questions from the rest of the FIMSAC 4 committee? 5 MR. HEANEY: Rick, if we can, can we just go to 6 the phone first for the four FIMSAC members on the phone, 7 knowing that it is just slightly challenging? Anybody on 8 the phone have questions for Rick or -- and/or the panel? 9 MS. WILSON: No, I'm all set. Thank you, 10 Rachel. 11 MS. WALTER: I'm fine, too. This is Elisse. 12 MR. HEANEY: Suzanne? 13 MS. SHANK: So it sounded like the panel was 14 recommending we change the language in the 15 recommendation, the use a last-look versus pennying. I 16 just wanted some clarification of that -- or no? Or -- 17 it looks like disagreement, shaking and nodding. 18 MR. REDFEARN: It seems to me that what we're 19 hearing is potentially changing the recommendation from 20 simply stating the use of last-look to something along 21 the lines of -- to the repeated systematic pennying in 22 the -- something like that might be more targeted and 23 specific to what is being discussed here. Thoughts on 24 that? 25 MR. MCVEY: I think we tried to make that 0059 1 differentiation, and if we can make it even more clear I 2 would personally be supportive. But systematic pennying 3 occurs because of the ability for dealers to exercise 4 their option in last-look. We viewed this as an 5 especially problematic use of last-look around systematic 6 pennying. And if it's possible to make that 7 differentiation more clear, I think the committee would 8 be supportive. 9 Larry? 10 MR. TABB: Thanks. So without a definition of 11 de minimi or a definition of systematic, you know -- and 12 will there be a penalty if we put this -- this just be 13 kind of like a community -- we won't do this, or we'll 14 know when we see it, or -- you know, will this have teeth 15 if we don't, you know, do anything more rigorous than 16 just say we don't like this process? 17 MS. HENRY: Well, I think that there's a good 18 bit in this proposal that would have to be left up to the 19 regulators to flush out. 20 I know that the MSRB is trying to do that right 21 now in the muni space, and I imagine that FINRA would 22 look at it, or the SEC would look at it, either way, in 23 the corporate space to try and provide more detail about 24 what's meant by systematic, how you would measure it, 25 and, you know, what the appropriate amount is, which will 0060 1 probably vary from maturity to maturity and credit class 2 to credit class. 3 MR. MCVEY: I know in our deliberations factors 4 like the frequency of pennying and the amount of price 5 improvement were important considerations to identify 6 dealers that are abusing the practice and not necessarily 7 using it on an occasional basis for price improvement and 8 best execution. 9 MR. HEANEY: Larry? 10 MR. HARRIS: When dealers and others provide 11 quotes or orders that provide options to trade, they 12 create value. They -- the options themselves may be 13 valuable, or they may reveal information is valuable. 14 The market failure that we have here is that, 15 when they do this and they're not rewarded, they're not 16 paid for the value that they created. And so that's a 17 market failure that's pretty classic. 18 The proposal tries to address this issue, and I 19 just want to ask whether there might be other ways that 20 the panelists think that we could address the issue. The 21 easiest way to do it, in principle -- but in practice I 22 think would be very difficult -- would be to simply some 23 -- find a way of compensating the dealers for the service 24 that they provide when they quote. I'm not clever enough 25 to find a way of doing that, practically, but perhaps 0061 1 others are. And so I'd like to ask whether any of the 2 panelists can find any other way to address this market 3 failure that -- short of the proposal that we're pushing 4 forward. 5 MR. BAGLEY: I mean I think there's certainly 6 other ways you can do it. I just think the repercussions 7 from those could be significant. 8 I mean you could put the best execution 9 requirement on the firm that's bidding and not the firm 10 that has the bid wanted. But I think that would have a 11 dramatic impact on liquidity, so I won't -- you know, not 12 saying that's good. 13 I think there's a lot of reasons, but pennying, 14 as a whole, sounds like a really simple concept. When 15 you read the request for comments people were all over 16 the map as to how to fix it and how big a problem it is. 17 The trade groups had a real problem. Even firms had a 18 problem coming out with a consensus because they had two 19 desks that do business the same way. 20 So I think what we in the MSRB and FINRA does - 21 - because you have to think about the same thing -- is 22 you have to really think about unintended consequences. 23 So you can fix the problem, but you got to make sure you 24 don't create a bigger problem with your solution. 25 MR. HEANEY: John, can I just jump in and 0062 1 reiterate what Brett asked, or the statement he made? Do 2 you think some of those comments that were all over the 3 map, to use your expression, is because it was defined in 4 the title as last-look, as opposed to systematic 5 pennying? 6 MR. BAGLEY: No, because this was the MSRB 7 comment, and we specifically differentiated between last- 8 look and pennying. 9 MR. HEANEY: Oh, this is your comment period. 10 Thank you. 11 MR. BAGLEY: This was the MSRB request for 12 comment periods. 13 MR. HEANEY: Thank you. 14 Horace? 15 MR. CARTER: I wanted to just -- I don't have a 16 question, I just want to make a quick comment. And it's 17 primarily for the benefit of the MSRB and FINRA and the 18 SEC. When in practice we go into reviews for enforcement 19 and things of this nature, I would encourage I would 20 encourage us to make sure that we're looking at markups 21 or price improvement in terms of basis points in yield, 22 and not dollar price, because what we've found is that 23 there is a concentration -- we process about 10,000 bid 24 wanteds a month, just in municipal retail. Maybe 40 25 percent of them actually end up trading. They are 0063 1 heavily concentrated in high-quality, front-end paper, 2 because most of the liquidations that we see are not 3 based on a decision to exit a certain bond; it's to raise 4 money. And so our clients are selling the bonds that are 5 easiest to sell. 6 So evaluating it based on a standard that 7 includes duration, I think, will be critical to what 8 ultimate enforcement looks like. Thanks. 9 MR. HEANEY: Thank you, Horace. 10 Ananth? 11 MR. MADHAVAN: Thank you. You know, I was 12 struck by Matt's comments about the number of times that, 13 you know, his firm won an auction and yet it wasn't a 14 print. I'm just trying to get a sense of -- from the 15 panel in general -- as to how extensive do we feel that 16 is, you know, across, you know, the time span of, say, a 17 year. 18 MR. LAND: I'll comment really quickly on that, 19 just from someone who does not trade things sometimes. 20 In a typical day we have cash requests every day from 21 clients. And again, it's normally not about exiting a 22 bond, it's raising cash to buy a house or whatever. 23 Especially in volatile periods in the market, 24 we will put out more bonds for the bid than we need to 25 raise cash because our certainty on what the price is 0064 1 received is less than, you know, in a kind of more 2 normalized market. So I may put out two bonds and only 3 end up selling one. And I know that a lot of the 4 separate account managers do need to do that. 5 And so we, on a normal day, sell 90 percent of 6 what we put out for the bid. But you look at very 7 volatile periods in the market, either prices going up 8 quickly or down, and that might drop to something like 60 9 or 65 percent for a short period of time. 10 MR. BAGLEY: And let me just add we -- Simon 11 Wu, our economist, presented a paper to the panel -- I 12 forget when it was, one of the other meetings -- and we 13 found that on the ATSs it was about a 25 percent hit 14 ratio for when they converted. We've got more data which 15 looks like that number is going up, which is a positive, 16 but realize you've got individual investors. They may 17 get a price they were not expecting, and they want to 18 sell. It may just be the client -- the broker can't get 19 a hold of the client because, by the time the bid comes 20 back, the client's not available. 21 So, you know, you're relying upon the client to 22 let you know, if it's a traditional brokerage account, 23 that they want to sell. And sometimes they will put out 24 more than they need, as was said. So I think those kind 25 of numbers -- 25 to 40 percent or something for brokerage 0065 1 accounts -- is probably normal. 2 MR. CAHALANE: Yeah, I will follow up on that, 3 that it's absolutely normal ratios in terms of 4 executions. I think in the case -- or to follow up on 5 Justin's comment here, we're talking about true municipal 6 market professionals who really understand where these 7 markets trade. 8 In many instances, a retail investor or 9 financial advisor, unfortunately, are not nearly as 10 sophisticated. And if they're looking to raise funds, 11 they may put an entire portfolio out for the bid on a 12 platform like ours, seemingly looking just for the 13 highest price of the securities they're putting out to 14 liquidate on behalf of the customer. 15 So, as much of a problem as systematic pennying 16 is, as one of the bigger retail platforms we'd love to 17 find a way to give better pre-trade price transparency to 18 retail investors and financial advisors so this practice 19 would actually decline, so that Matt's firm doesn't have 20 to bid quite as many RFQs in the municipal market before 21 trade could actually be consummated. 22 MR. MADHAVAN: Very helpful, thank you 23 MR. HEANEY: Lynn? 24 MS. MARTIN: In the second part of the 25 recommendation you make the statement the SEC should 0066 1 encourage FINRA and MSRB to expand its review of, I 2 guess, now repeated systematic pennying to cover all 3 electronic venues, rather than limit specifically to the 4 ATSs. 5 I was wondering if the subcommittee or the panel had 6 ideas around what surveillance protocols could be put in 7 place, either for the voice market or for the other 8 electronic trading venues, to surveil for this type of 9 activity. 10 MR. BAGLEY: There are requirements under G8 in 11 the municipal space that brokered -- that ATSs and 12 broker's brokers keep information like when they received 13 the bid, what was the bid they received, who they 14 received it from. So though -- all that impacted 15 information is documented. So if this were to become 16 interpretive guidance in the municipal space, and FINRA 17 were to go -- or the SEC were to go into a review, they 18 would have all that type of information from broker's 19 brokers or ATSs. 20 MR. HEANEY: Sonali? 21 MS. THEISEN: Thank you. Part of my question, 22 I think, was was answered with some of the comments 23 earlier around how often are there trades or requests to 24 trade that actually don't consummate a trade because of 25 the actual and investor. 0067 1 But I guess my question might be for John from 2 Tradeweb or maybe for Rick. Like, do you ever observe 3 and maybe comparing the retail space institution, do you 4 ever also observe that response rates to certain 5 counterparties are lower because there is a view that 6 there would -- that there is this practice going? Like, 7 is -- in other words is there a correction element in the 8 marketplace of certain counterparties being bigger 9 offenders, and then not getting responses? And does that 10 -- do you see that, as well, kind of -- 11 MR. MCVEY: I'm happy to take the part of the 12 question about the institutional market. And there is, 13 as you know, a bit of a self-policing mechanism in the 14 institutional market, because the majority of electronic 15 trading is done between disclosed counterparties. So if 16 a client in the institutional market consistently puts 17 requests out, receives lots of responses, and doesn't 18 trade it's very visible to the market-makers. 19 It's my understanding in the retail market that 20 that level of disclosure does not exist. So it's hard to 21 determine when there is an order from somebody that 22 abuses the system for a price discovery versus somebody 23 that legitimately has an intent to trade. 24 MR. ANDRESEN: And so that that problem is 25 especially prevalent, because you have -- because you 0068 1 don't know who the provider is that's coming in, 2 initiating the auction on the ATS. You have to bid the 3 level to the -- sort of the lowest common denominator. 4 You can't say, "Oh that's -- that person always you know 5 consummates the trade if I win. Let me bid more 6 aggressively and make sure I win that auction." Instead, 7 you have to price it for most toxic counterparty, who may 8 just be looking to find out what the price is they should 9 trade at, and then step in front of it at the end. 10 MR. CAHALANE: Rick was correct on our platform 11 and the other retail platforms. In general, the pre- 12 trade anonymous/post-trade give-up during the RFQ 13 process, we do have firms who choose to be named give up 14 prior to the auction going out. We don't see a 15 tremendous difference in the number of responses to those 16 RFQs versus the RFQs of pre-trade anonymous. 17 I would say that the firms that are choosing to 18 give themselves up generally are not taking the same 19 principal risk as those who are not. So you would almost 20 expect there to be a more aggressive market for their 21 RFQs because of less competition from an internal desk. 22 We generally don't see that. It's pretty even. The 23 number of responses on municipal RFQs in particular, as 24 John and I have done a little bit of work, even in times 25 of market distress we don't see a tremendous fluctuation. 0069 1 There's obviously some, but I wouldn't call it extreme. 2 MS. THEISEN: Thank you. I was asking this 3 question as a complementary offering. I was wondering if 4 there's any way, from a protocol perspective, that 5 there's ways to address some of these issues on venue, 6 i.e. giving, you know, even on anonymous trades, et 7 cetera, like, some sort of likelihood of trading or 8 historical average of trading. Or if someone's putting 9 out, you know, 10 bonds and only trades 5, if there's 10 some way, from a protocol perspective, to tackle some of 11 this, you know, as a complementary offering. 12 MR. REDFEARN: You do see scorecards in some of 13 the alternative trading systems and equity markets from 14 time to time, where they've done this kind of thing. 15 MR. HEANEY: Larry? 16 MR. TABB: Not that we want to go this way, but 17 on on the equity side there are two things that kind of 18 compensate Matt for for adverse selection. One is, well, 19 provide compensation. One is maker-taker, which I'm not 20 sure we want to go into; or second, there's the SIP, 21 another topic that we're, you know, fighting over on the 22 equity side. But there are things that could get back to 23 Larry's issue of trying to provide some sort of 24 compensation for adverse selection, or for quoting. 25 MR. HEANEY: Larry? 0070 1 MR. HARRIS: A quick response to Larry, and 2 then on my question. 3 The maker-taker doesn't provide compensation 4 for adverse selection, it just tends to narrow spreads. 5 So that's a different issue, and I don't think we should 6 join it at this point. 7 So Matt spoke about the fact that apparently 8 half of these transactions that do occur are at inferior 9 prices. 10 So those are trade-throughs. And John was -- responded 11 saying that it's important for us to remember that the 12 financial analysts who are the financial advisors need to 13 obtain compensation, and that can explain it. 14 The reason we're seeing that result is that if 15 the financial adviser is receiving some compensation, is 16 coming out of the markup -- and what's happening here is 17 that we have a system where people are trading net, which 18 is typical in fixed income. So what we have is of a 19 bundled service. We have, for those cases where the 20 financial adviser is getting some compensation, 21 compensation is coming out of the markup. But what's 22 being bundled is the execution services. 23 The question is this -- or I'll make the 24 following observation with a question. It seems to me 25 it'd be a whole lot simpler if the financial analysts 0071 1 were compensated out of a commission, because then we 2 would know what the commission is, and we'd know what the 3 price is, and then we could figure out what best 4 execution is, and we wouldn't, presumably, not have the 5 trade-throughs that Matt is observing. 6 So the questions is this -- or two questions. 7 The first one is are the markups known to the client 8 before the trade takes place, as they would be if there 9 were a commission? We know that they will have 10 ultimately been disclosed through the new regulations. 11 And the second is what harm would come to the markets if 12 we asked for this unbundling, and in particular said that 13 we will -- we really shouldn't be seeing these trade- 14 throughs, we expect that you'll be charging commissions, 15 and that the commissions would be, presumably, displayed? 16 So questions are, are the markups known in 17 these circumstances to the client ahead of time; and 18 secondly, if we required it, perhaps through a best -- 19 some standard of best execution that might prohibit 20 trade-throughs, what harm, if any, would come to the 21 markets? 22 MS. HENRY: I don't think that the markup is 23 necessarily going to be known to the customer at the time 24 of trade, because of the way that the markup rules work, 25 and they require an offsetting trade in the same day in 0072 1 order to trigger a markup disclosure. 2 So, depending upon what time of day the trade 3 occurred, the first trade occurred, you may not know what 4 the markup is at that point. 5 MR. HEANEY: I guess for the larger FIMSAC 6 group, this has come up, this unbundling topic has come 7 up in a different subcommittee, and has been addressed -- 8 not that it can't be readdressed or refocused at some 9 point in the future. It pertains, for sure, in some 10 aspects to this. But I would say that this, as a topic, 11 is probably better suited still for the other 12 subcommittee. 13 MR. MCVEY: Maybe I could ask Jude just to 14 expand, because it -- dealers representing retail orders 15 have -- are now required to report riskless markups, and 16 that's relatively new. But in terms of transparency to 17 the retail customer, has that made a difference? 18 MR. ARENA: I haven't seen a meaningful 19 difference. It's true that we markup -- it's disclosed, 20 a riskless markup. I'm not sure if they're aware prior 21 to the fact or after they receive it on their confirm. 22 I'm not sure of the timing of it. 23 MR. REDFEARN: Yeah, John, in the analysis that 24 you're doing, the discussion around the sales credit or 25 the markup, is there any way that when you're looking at 0073 1 data you can make some sort of assessment that will help 2 us have a better understanding of the extent to which 3 we're seeing -- you know, you're actually seeing sort of 4 trade-throughs and pennying, versus some other economic 5 aspect in the reported trades? 6 And secondly, after that I just wanted Matt to 7 just get your sense quickly about, given the aspect of 8 the prices that are being reported, including that in 9 them, you know, how do you sort of interpret that when 10 you think of the data you're seeing with respect to 11 trade-throughs from your quotes? 12 MR. BAGLEY: So we have two different data sets 13 we're looking at right now. The data set we got from 14 broker's brokers and ATSs would not have that type of 15 information. All we'd have is what was the high bid on 16 the ATS, what was the high bid on the broker's broker, 17 and we can obviously marry that to trade data and see if 18 there was a corresponding trade, and what price. We 19 wouldn't have any sales credit that a broker worked for, 20 or if there was any desk spread we wouldn't know that 21 information. 22 We have some other data that we have from FINRA 23 that we're looking at that may have that. But I think 24 it's too early to tell. But they got better information 25 on what the exact bids for the firms were. But that 0074 1 information looks like it has information on where the 2 firm bid the bond, which is something that's missing in 3 the other data set. 4 MR. REDFEARN: From any of the disclosed 5 markup, markdown -- you know, from any of that data, do 6 you think that that can be married at all with the 7 analysis that you have to get a better appreciation of 8 that part of this? 9 MR. BAGLEY: We'd have to do our own 10 calculation, because markup/markdown is just from a 11 confirm. We don't get that, so it's not reported. 12 MR. ANDRESEN: Before Headlines I started and 13 ran the largest equity market-maker. And this was a -- 14 in that business you are taking held orders, typically, 15 from retail customers, and you are either stepping up and 16 providing better liquidity than the market has, or 17 sending that order onto the market to get the market 18 price. 19 So not -- I didn't run into the -- into this 20 issue around trade-throughs; we were held to our best 21 execution obligation, and had to be held to the 22 statistical regime that the SEC put out 15 years ago or, 23 I guess almost 20 years ago now, to be able to measure 24 that best ex. 25 You know, in all of these discussions the -- I 0075 1 think we've heard some suggestions about what could be 2 done or what could be done to better measure, and I agree 3 with the -- my fellow panelists that coming up with a 4 bright-line definition does not make sense. There are 5 certainly scenarios where a dealer has to exercise 6 judgment and their -- there is sort of the last resort to 7 protect the end customer. But given the numbers that 8 I've shared in our previous comment letters, and then 9 today, we believe that the scale of this issue is 10 sufficiently large as to require some intervention by the 11 SEC. 12 MR. HEANEY: Larry? 13 MR. HARRIS: A potential internalizer could 14 always submit a blind bidder offer to the auction. And 15 if they were the highest bidder, then they would win the 16 auction and trade at that price. 17 The question is can our panelists identify any 18 circumstances where that requirement, if it were such a 19 requirement, would be disadvantageous. So we're not 20 proposing to require that. But if we did, what damage 21 potentially could be done? Are there situations where 22 it's important for an internalizer you to be able to 23 internalize after they see everything, and as opposed to 24 doing it blind as the highest bidder? 25 CHAIR CLAYTON: I think I shared earlier -- and 0076 1 what we see, statistically, as the most frequent reason 2 for internalizing or stepping in the middle is crossing. 3 So, should a firm want to cross bonds between either a 4 particular financial adviser's clients or two of the 5 firm's clients, despite the fact that they're bidding 6 during the auction, they may want to internalize a 7 transaction not being the best, and saving both clients, 8 you know, potentially, costs associated with the trade. 9 So while I think having the internal desk bid 10 at -- you know, or as part of the auction process is a 11 step in the right direction, I still believe there are 12 instances where internalization is necessary. 13 I will add, with very little resource put into 14 this so far, we have seen a decline in pennying in the 15 last five years. Again, very limited amount of data 16 we've looked at. We're happy to look at and work with -- 17 and have been working with -- the MSRB and the other 18 regulators to try and determine, on an undisclosed basis, 19 of course, you know, how prevalent it is. 20 I'd also like to add that the rate of 21 externalization on our platform, overall, has steadily 22 risen over the last five years. So while we're here 23 talking about last-look and internalization, I believe 24 platform participants -- both my liquidity takers have 25 been using external liquidity at a much greater rate over 0077 1 the last five years, and I think firms who provide 2 liquidity like Matt's have benefitted from the 3 opportunity to provide that liquidity. 4 MR. HARRIS: So would it -- would you be in 5 favor of carving out crosses in what we're doing here? 6 The crosses are very easy to identify. And, presumably, 7 if we carved that out we would add, if necessary, the 8 requirement that both parties participating in the cross 9 know that there was a cross, and perhaps even know the 10 origin of the price. 11 CHAIR CLAYTON: I think that's probably a 12 better question for liquidity takers, themselves. 13 Systematically, would there be a way for us to 14 potentially acknowledge that it's a -- you know, the 15 liquidity taker's purpose for initiating the auction as a 16 cross? Yes, we could probably do that with -- fairly 17 simply. 18 MR. LAND: And to add onto that, as I 19 mentioned, our firm doesn't cross, but I've talked to a 20 number of our competitors. Some do engage in their 21 crossing by using an ATS to blindly bid what they 22 potentially cross, so they don't have any sort of last- 23 look. So I know there are some participants doing that. 24 As I understand it -- and I'm not an attorney - 25 - they are required, as part of management agreements and 0078 1 disclosure and things like that, they need to tell their 2 clients that they do engage in crossing. Most firms 3 allow their clients to opt out of that, as well, as well 4 as the methodology. So it doesn't tell you on a per- 5 trade basis, but what methodology and documentation for 6 any cross-trading that they do. 7 CHAIR CLAYTON: And Larry, let me also add -- 8 and I don't -- we don't have -- I don't -- the MSRB 9 doesn't have an opinion on this, and neither do I, but 10 one of the things we look at is, if you were to require 11 something like that, how would a firm meet their best 12 execution requirement on a bid wanted where they didn't 13 get their bid in for whatever reason -- they were 14 inundated with bid wanteds that day, or something 15 happened. You've got to be concerned about that because 16 it doesn't matter whose bid the dealer puts up to the 17 client; the dealer is responsible for meeting all the 18 requirements for that. 19 So one of the things we're thinking about is 20 just that process, right, how do you not affect one rule 21 with another rule. That's why it's fairly complicated. 22 MR. HARRIS: I would note that the broker- 23 dealer faces a lot of regulatory risk when doing these 24 crosses, because if you're favoring one client to the 25 other, even if you're not and you seem to be, that's 0079 1 potentially problematic. So I would imagine it's not 2 exactly an attractive process, from that point of view. 3 MR. HEANEY: Horace? 4 MR. CARTER: I just wanted to briefly respond 5 to Larry's inquiry. What -- I would actually envision, 6 in that scenario, liquidity being reduced, and for the 7 following reasons. 8 So just following on what John said, it's not 9 only about, you know, our best ex requirements, it's also 10 about the sheer volume that we're seeing and the interest 11 that we may or may not have in a certain bond. So let me 12 just run through a scenario that we -- that could be 13 repeated, literally, hundreds of times per day at Raymond 14 James. 15 I receive a bond that's in for the bid. I 16 don't care to own it. We externalize more than 70 17 percent of our bid wanted flow, we put it out. For 18 whatever reason, we just -- we simply don't get 19 reasonable bids, and we go ahead and take the bonds, even 20 though we don't want them. We -- this is something that 21 happens all the time. 22 And so if we were unable to bid that bond, we 23 wouldn't -- it wouldn't be a best ex -- we would not 24 trade. We wouldn't trade the bond at all. And so you 25 would actually see liquidity decrease, and I think you 0080 1 would see those hit rates go down. 2 CHAIR CLAYTON: Could I just make sure I'm 3 hearing everybody correctly, if not consistently? 4 So there is an issue in that there is a lot of 5 what I'll call, for lack of a better term, good bidding 6 by the home -- I'll call it the home broker: times of 7 judgment, times when there's no liquidity crossing. And 8 then there is bidding that we should have a concern 9 about, which is systematic bidding that has the effect of 10 actually reducing liquidity. 11 The problem is distinguishing between the two. 12 Probably easier on an ex post, rather than an ex ante 13 basis, which is why you were saying don't go down the 14 road of trying to be highly prescriptive, because being 15 ex ante in this area, you're probably going to eliminate 16 a fair amount of the good bidding to get to the bad, or 17 not achieve your objective on the bad to get -- to permit 18 the good. 19 And I take Larry's point that there is some 20 apples and oranges comparison problems, due to the 21 differing ways that trades are executed, whether through 22 a commission or a markup. But I do get the sense that 23 this is something that you think the Commission should 24 continue to look at, as do we have consensus around that. 25 Fair enough, thank you. 0081 1 MR. HEANEY: Other questions or comments? 2 (No response.) 3 MR. HEANEY: Okay, we had a break for -- oh, 4 I'm sorry. 5 COMMISSIONER ROISMAN: Yes. I mean I want to 6 follow up on -- first of all, thanks again for taking the 7 time and kind of following on what the chairman just 8 mentioned. 9 It seems to be that everyone on the panel 10 seemed to agree that there is a problem of, if there is a 11 problem, it's systematic pennying. And I haven't heard 12 anyone disagree on that. 13 So, also hearing from John that there is a 14 whole host of reasons that people take a last look and, 15 you know, each one can have a different reason for 16 transacting the execution or filling the execution, the 17 problem, it seems to me, is that we just don't have 18 enough information for determining when it's systematic 19 pennying and when it's not. 20 And one of things, you know, I think that would 21 be helpful for us is identifying things that we should 22 look for to potentially understand when it's systematic 23 pennying. And this is along the lines of what the 24 chairman said. 25 I think -- and I'll tell you -- the thing that 0082 1 I just care about is how do we get the best execution for 2 a client. And, you know, the more information we have, 3 it seems, the better we can make that determination. So 4 if anyone has ideas, we would really appreciate it. 5 MR. MCVEY: We certainly talked about this in 6 the subcommittee deliberations. And to us, the two key 7 things were the frequency of pennying and the price 8 improvement from a particular dealer when they were 9 pennying. So I think it's fair to say that if there is a 10 high percentage of responses that involve using last-look 11 to penny the bonds, and the price improvement is 12 infinitesimally small, that that's a pretty good sign 13 that there is an abusive practice going on to use the 14 system for price discovery with every intent to continue 15 to internalize that trade. 16 COMMISSIONER ROISMAN: No, I appreciate that. 17 I mean, do we have that data? I mean -- 18 MR. MCVEY: The -- certainly, the ATS systems 19 have that data. 20 MR. CAHALANE: We have some of that data. You 21 know, depending upon which firm it is, and how they're 22 using our system and integrating with our system, we 23 don't necessarily know all of the bids that they receive 24 in an auction. We know all of the bids they receive in 25 an auction from our platform. But particularly over the 0083 1 last several years, following up on the best ex rule, 2 many of our platform participants started using multiple 3 ATSs, as well as voice brokers, as well as their dealing 4 desks to go out into the market to get pricing for 5 auctions. 6 So, while we absolutely will know all of those 7 responses in and from our platform, for some of our 8 platform participants we will have no clue where the best 9 bid actually is, or where it came from. 10 CHAIR CLAYTON: Just let me say I have to run 11 to another thing, but this is a -- this has been a 12 terrific discussion. I thank all of you. You know, this 13 is not a simple concept. And I agree with Commissioner 14 Roisman that, you know, we're all going for the same 15 objective: best execution. Keep putting our heads 16 together on how to get there, and I just thank you for 17 all the work here. 18 MR. BAGLEY: Can I just add? I think the two 19 pieces of data that are not readily available is the firm 20 that's bidding when they bid it, and what was the price 21 they bid it at. Because you could have a scenario where 22 the firm bid it first thing but didn't enter into the 23 auction, ended up being a high bid by a small amount. We 24 would say that's not pennying, because they didn't get 25 the information when they bid it. But that's not readily 0084 1 available. You'd have to go ask for that type of 2 information. 3 MR. ANDRESEN: I think one other indication is 4 -- you know, Horace talked about, at Raymond James, they 5 externalize 70 percent, which means the odds that any 6 given trade would be filled out of their inventory is 7 only -- is less than a third of the time. And so, again, 8 without trying to work ourselves around to any bright- 9 line standard, you can imagine that there is a 10 distribution of experience by different participants in 11 the marketplace. 12 And if you saw an externalization rate not of 13 70 percent, but of, let's say, 10 or 15 percent, that 14 would be something that would be -- it would be more 15 difficult to believe that there wasn't a systematic last- 16 look going on. 17 MR. HEANEY: So let me do this. We've got four 18 people on the phone, of which I think we've got to try to 19 make sure all are on. So let's -- we had an 11:30 break 20 scheduled. Let's just take it now. We'll come back to 21 this right after the break. So it's a 15-minute break. 22 So let's be back at 11:40. 23 On behalf of Rick -- and I don't mean to jump 24 in front -- I would -- just want to thank all the 25 panelists for coming down, making the trip to D.C., 0085 1 sharing an immense amount of insight and expertise for 2 FIMSAC, and we will reconvene afterwards on the 3 recommendation. 4 (A brief recess was taken.) 5 MR. HEANEY: Okay, Rick can I turn it to you 6 for for your thoughts here and your comment? 7 MR. MCVEY: Sure. You know, upon further 8 deliberation on the discussion this morning, it would be 9 my recommendation that we clarify in better terms in the 10 recommendation exactly what practice we think creates 11 negative implications for market competition, and tighten 12 up the recommendation to make it very clear that it is 13 the practice of systematic pennying, as opposed to broad- 14 based last-look, that we are concerned about. 15 And I believe, timing-wise, with a few 16 modifications that would be even more clear in the 17 recommendation, I would propose that the subcommittee 18 review new language in the next 48 hours that, Michael, 19 we could return to you, and ideally schedule a conference 20 call for the full committee so that we have a 21 recommendation that makes it very clear what we are 22 proposing, and also has the benefit of making it entirely 23 consistent with the MSRB comment letter. 24 So that would be my suggestion on next steps. 25 MR. HEANEY: So we can take that -- those -- 0086 1 the modest revision, we can re-circulate it after the 2 subcommittee -- and the SEC team, but the subcommittee 3 has reworded that. And then we can do a vote. We can 4 re-circulate that to the broader FIMSAC, and do a vote 5 within the next 10 days, either by email or by phone. I 6 think by phone might be the easiest, and we probably 7 should endeavor to do that. But there's no reason why 8 that can't be turned around quickly to do that. 9 I would make two other comments: one, I 10 thought that was a really good conversation, and I 11 appreciate how much effort went into this from the 12 subcommittee; and two, Rick, your leadership to say, you 13 know what, let's just -- let's make a few revisions, 14 let's tweak it slightly, let's get it to be as clean as 15 we can as we bring it forth to the SEC -- first to the 16 FIMSAC vote and then to the SEC, should it go that way. 17 But I think that's -- I wholeheartedly agree, and I 18 appreciate your stepping up to do that. 19 Any comments or questions? 20 (No response.) 21 MR. HEANEY: Okay, we will now turn to our 22 considerations of the recommendations from the Municipal 23 Securities Transparency Subcommittee on principal 24 transactions with advisory clients. The draft 25 recommendation developed by the subcommittee addressed 0087 1 two specific types of principal transactions with 2 advisory clients; one, purchasing negotiated municipal 3 under-writings; and two, liquidation of bonds from an 4 advisory client's portfolio. 5 Initially, these two recommendations were 6 compiled in a single document that you would have 7 received. The subcommittee determined to present each 8 one individually. This should help us organize our 9 dialogue in the consideration of the two distinct 10 recommendations. 11 Lynn Martin, who chairs this subcommittee, will 12 moderate the panel discussion on both the preliminary 13 recommendations, and then we'll take up each 14 individually. 15 And first we'll hear from our panelists 16 concerning the preliminary recommendation on municipal 17 underwritings. Again, after that we'll open up to 18 questions, and then a vote, and then we'll go to the 19 second. 20 Lynn? 21 MS. MARTIN: Thank you, Michael. I'd like to 22 welcome all the panelists to today's FIMSAC discussion. 23 And thank you very much for agreeing to participate in 24 this panel. 25 I'd like to first turn it over to Horace Carter 0088 1 to tee up the primary offering recommendation that we are 2 putting forth to the FIMSAC committee. I will then 3 proceed with a Q and A of the panelists, and I would ask 4 each of the panelists to introduce themselves and explain 5 what their role is in the municipal trading ecosystem as 6 part of their responses. 7 So, Horace, over to you. 8 MR. CARTER: Thank you, Lynn. 9 Currently, a broker-dealer that negotiates and 10 underwrites a new-issue municipal bond, or as a co- 11 manager, or is a member of the selling group is then 12 unable to sell bonds in the offering to its advisory 13 clients without meeting the disclosure and consent 14 requirements of the Advisors Act. The result of this is 15 that few or none of the underwriting dealer's advisory 16 clients buy bonds in the initial offering of negotiated 17 municipal bonds for which the dealer is a syndicate 18 manager or selling group member. 19 Advisory clients who wish to buy these bonds 20 will buy them after the deal is closed and the bonds are 21 free to trade, typically at a price that is higher than 22 the original offer price. 23 So the subcommittee recommends that the SEC 24 consider a rule that permits a broker-dealer that 25 negotiates and underwrites a new-issue municipal bond, or 0089 1 as a co manager or member of the selling group, to meet 2 the requirements of Section 206(3) of the Advisors Act 3 when acting in a principal capacity to sell new-issue 4 municipal bonds during the negotiated order period. 5 MS. MARTIN: Thank you. 6 The first question is for Anthony and Brad. 7 The draft recommendation would permit investment advisors 8 to sell underwritten new-issue municipal securities into 9 managed or advisory accounts. Some may be concerned that 10 permitting this to occur could result in harm to 11 investors, given certain conflicts of interest in 12 distributing the deal that the underwriter faces. 13 A couple questions to that regard: How does 14 this recommendation address those concerns, as well as 15 how would you anticipate the perception of double dipping 16 on the fee side? 17 MR. LIOTTI: All right, I'll start. Good 18 morning everyone. First I'd like to just say thank you 19 to everyone for allowing me to talk today about what I 20 think are -- is a very important topic within the 21 municipal market. 22 As an introduction, I worked at UBS for 11 23 years. I currently manage a team of portfolio specialist 24 that work well both within the tax exempt arena as well 25 as a taxed arena. These -- this team of portfolio 0090 1 specialists speak directly with the UBS financial 2 advisors on a regular basis, providing them with advice, 3 counsel, portfolio solutions, as well as trade and 4 execution on behalf of our clients. 5 So I would just like to say I look forward to 6 discussing these recommendations with all of you today, 7 as I feel that both of these recommendations support and 8 benefit the end retail client. This is, in my opinion -- 9 perhaps many others on the panel -- that these topics 10 should strongly be considered to ensure that all clients 11 within the retail space have the same access to the 12 markets as it relates to the full continuum of products, 13 best ex, and then, of course, the greatest amount of 14 liquidity. 15 So to answer your -- to get to your first 16 question, I think what's important here is really to 17 understand the changes that are taking place within the 18 marketplace as a whole. Your clients who used to work 19 off of a traditional brokerage model and executing off of 20 a traditional trade-by-trade execution are clearly 21 migrating into a more holistic approach of the fee-based 22 model. And I think that, as we look at the current rule, 23 it does allow for discretionary trading, albeit at a 24 trade-by-trade consent attestation by the actual client. 25 I think that what we're talking about here, 0091 1 really, is clearly more of a blanket attestation to 2 clients, which would really simplify a lot of things for 3 both the financial advisor and, of course, the end 4 client, as many clients who have decided to go into a 5 fee-based model don't want to really be involved on a 6 day-to-day basis with the trading and activity 7 recommendations being put forth by their financial 8 advisors. And so it truly is left up to the fiduciary 9 responsibility of that financial advisor to go and make 10 those particular purchases on the accounts. And the 11 process is quite onerous if we anticipate or expect that 12 financial advisor to call the client on those trades. 13 I think what a blanket recommendation does -- 14 or attestation -- is it allows the financial advisor and 15 the client to have full unadulterated access to -- 16 whether it be the secondary market and/or the primary 17 market, right? Which -- so it opens up and provides the 18 full product continuum for those clients, all right? So 19 it doesn't impede the end client to receiving these 20 securities. 21 The blanket consent would more easily allow 22 financial advisors and the clients in equal opportunity 23 to report to their -- to those securities within a fee- 24 based account. So whether they decide to transact in a 25 fee-based, it shouldn't be penalizing the client because 0092 1 they, and perhaps their lifestyle or the business in 2 which they conduct, it shouldn't impede them that the 3 firm in which they have assets with and they're working 4 with on a regular basis happens to be extremely active 5 within the primary municipal market, okay? 6 And I think that clearly -- that penalization, 7 what makes the primary market so appealing, and I think 8 to all the constituents in the marketplace, is that 9 everyone, whether it's a small, you know, retail client 10 looking to buy a five-bond piece, or that large 11 institutional client, one price clears the entire 12 marketplace. 13 And so, when we talk about and hear about best 14 execution here, the primary market is, in my opinion, and 15 perhaps many of those on this panel, the most 16 advantageous way for retail clients to buy bonds. And so 17 currently, right now perhaps, you know, I think expanding 18 this to where we have that blanket coverage opens up a 19 lot of opportunities for our clients. 20 MS. MARTIN: Brad? 21 MR. WINGES: So I'm Brad Winges, I'm the 22 president and CEO of Hilltop Securities. We're a large 23 municipal underwriting financial advisory firm. We have 24 wealth accounts on the brokerage side, institutional 25 side. 0093 1 I agree with Anthony's comments. I think, at 2 the end of the day, allowing an individual retail 3 investor to participate in a new-issue transaction is 4 essential to creating not only more liquidity, but more 5 access to the individual investor to the municipal bonds 6 that they ultimately want. 7 Let me give you an example of that. If you are 8 an investor with a particular brokerage firm, and you 9 have your money and a commission-based account, you can 10 buy a new-issue transaction if that firm is an 11 underwriter on the issue, and the underwriter will be 12 paid a takedown or underwriting fee for that transaction, 13 and you'll be able to purchase the security. If, in 14 fact, though, your money is in a managed account, as 15 Anthony was mentioning, you will not have the ability to 16 purchase that security. 17 And let's say in this case it's a local school 18 district that you'd like to participate in. You will 19 tell your advisor that you'd like to purchase a 20 transaction. And they will wait until the new issue has 21 been free to trade, the transaction is closed, and they 22 will pay a premium in the open market to get you the 23 ultimate security. And this can be viewed, and this rule 24 can be viewed, as a mitigant to the flipping that is 25 occurring in the marketplace because it's taking away the 0094 1 demand side of the component from the advisors, who 2 ultimately want the account -- or the bond wants the 3 account can by it at the new issue price, as opposed to a 4 premium at a later date. 5 So, knowing that flipping is something that is 6 a major concern also for the SEC, we deem this as a 7 wonderful opportunity to allow all retail investors, both 8 on the Commission side, but the managed money side, to 9 participate in that new issue at the original new-issue 10 price. 11 Secondly, as Lynn had asked, the double dipping 12 -- and that's what the reference is -- double dipping is 13 an acronym used where the underwriter is paid a fee for 14 the underwriting practice, or a take down. But, in 15 addition to that, they cannot be charged -- that 16 individual investor cannot be charged, then, commission 17 on that security. So that would be deemed double 18 dipping. 19 One of the mitigants to that issue that we 20 discussed was allowing that retail investor that's in a 21 managed account to get a credit to their account for that 22 take-down, so that the underwriter can, in fact, not 23 receive double payment. 24 The alternative to that would be, instead of 25 crediting the account, the take down, or the underwriting 0095 1 spread on that particular security, you could carve out 2 those securities for a fee-based charge for that one year 3 where the fee would normally be used. 4 So those were two issues that we saw that we 5 could address with a rules-based structure to avoid the 6 double dipping. 7 And secondly, we felt allowing the retail 8 investor, irrespective of what type of an account they 9 have, commission-based or fee-based, to have access to 10 that local municipal bond that they may desire. 11 MR. LIOTTI: If I could comment back on double 12 dipping, to add on here to what Brad was talking about, 13 within the syndicate process, whether the bonds are going 14 to be allocated to a wrap account, or, as you say, an SMA 15 or a mutual fund, let's say, the syndicate account will 16 retain the take-down that's embedded within the 17 syndicate. 18 The hypothetical that I would provide to 19 everyone is -- over the concern of double dipping is if a 20 client of ours comes to me and says, "I want to buy only 21 new issues for my account, and I have to, you know, defer 22 them away because I cannot, under this particular 23 platform, provide them with the sufficient amount of new 24 issues required to, you know, to provide them with their 25 investment solutions," they may have to go to a third 0096 1 party SMA manager, let's say, or a mutual fund to get 2 those new issues. 3 And so, while that concern over double dipping 4 is important, what you are ultimately doing by unintended 5 consequences is perhaps forcing them to go and pay an 6 additional, ongoing fee by that outside third-party 7 manager or fund or ETF over the life and course of those 8 particular bonds. 9 And so, while that take-down on these deals is 10 quite small, it's there. But the unintended consequences 11 of that could be a client going out and paying, 12 ultimately, a lot more for that one security over the 13 life of the bond. 14 MS. MARTIN: Thank you. 15 Brad, maybe I can direct this one to you. The 16 recommendation -- since you're on the Muni Subcommittee - 17 - the recommendation is really limited to negotiated 18 underwriting deals, rather than competitive deals. Would 19 you mind giving FIMSAC a view as to what got us through 20 that thought process? 21 MR. WINGES: Yeah. At the end of the day, even 22 though the competitive process is a underwriting 23 practice, it is viewed often times by -- as a secondary 24 market activity. And we viewed that the negotiated 25 environment was the one that really, at the end of the 0097 1 day, had the most opportunity to allow the retail 2 investor to come in and participate in those 3 transactions. The competitive market does enable that 4 investor to step in and buy it in the secondary market. 5 Different than a new issuer, the double dipping is 6 presumed to occur if the -- a client bought the asset in 7 the managed account. 8 MS. MARTIN: Thank you. I'm going to direct 9 this next set of questions to Chris and Anthony, with 10 your other hat on. 11 You represent the retail investor base directly 12 through advisory and brokerage services, and the 13 protection of investors is a pillar of securities laws. 14 Indeed, the restrictions on principal trading for 15 advisory accounts arise out of a concern of conflicts of 16 interest posed by principal trading. 17 Can you speak to the importance of an investor 18 obtaining a bond during the initial offer period versus 19 the secondary market? And how do these recommendations 20 actually benefit the retail investor? 21 Maybe Chris, we could start with you. 22 MR. KENDALL: Sure. Good morning, and thank 23 you to FIMSAC and the Exchange for having me here. 24 Again, my name is Chris Kendall. I'm a vice president of 25 fixed income trading for Charles Schwab. Our firm trades 0098 1 -- or my group trades for over -- does all the trading 2 for fixed-income products, whether it's done on a 3 secondary or a new-issue basis. 4 I'd like to maybe probably start by explaining 5 how we access new-issue municipals. So we have 6 distribution agreements with underwriters who will 7 underwrite these new-issue municipals. Through those we 8 can then make those offerings available to our clients 9 and take orders on a best-efforts basis. 10 Subsequently, we may get allocations on those 11 orders from the underwriters -- or hopefully, anyway. We 12 do make a portion of the concession, but the key there is 13 that we are not at risk on those bonds. So, in other 14 words, we're not an underwriter. 15 So now, if you look at it from the point of 16 view of an advisor that we -- in our firm, these 17 advisors, they have a fiduciary to the client, first and 18 foremost. They have to act in the client's best 19 interest. To that point, our advisors are not 20 compensated differently based upon whatever product, 21 fixed-income product, they may recommend. And again, 22 that actually goes as far as whether it's secondary or 23 new issue, as well. 24 And then, to Brad's point he -- the -- some of 25 these clients that are advised clients for us have 0099 1 regular brokerage accounts. So they may actually have 2 access to these new issues through their brokerage 3 account, but not through their managed accounts, and 4 that's an inconsistent experience for our clients, and a 5 pain point for us. 6 So, from our point of view, the fact that we 7 are not an underwriter and that our advisors are product- 8 neutral, it seems like it mitigates the conflict of 9 interest issue. 10 MS. MARTIN: Anthony? 11 MR. LIOTTI: Yes. So, as Chris mentioned, 12 fiduciary responsibility 100 percent upon the financial 13 advisor to ensure that they are meeting, you know, the 14 client's objectives within the securities laws. 15 I think this is a question about at least just 16 offering the ability for the client and the financial 17 adviser to purchase these securities. I mean, so it's 18 not -- it truly comes down to them, and that's what this 19 is, enabling them the ability to look across the entire 20 landscape and say is this the right investment for my 21 client at this particular time. 22 So I don't think that we're -- you know, the 23 perception of us pushing these securities on it is 24 probably unfair. 25 I think that, through a blanket client 0100 1 attestation in which the financial advisor and the 2 client, whether they sit down upon the initial opening of 3 the account -- as a hypothetical -- with blanket 4 attestation, would clearly need to state out what you 5 were or what the client would be signing up for, meaning 6 that UBS, or whatever firm, would be participating in 7 some of the securities in which that financial advisor 8 would be buying for that account, UBS or any of these 9 other firms could be acting within a principal capacity. 10 And then it's also allowing the client to step 11 out, and at some point opt out of that particular 12 relationship if they're unhappy. 13 But I think that, you know, in what you're 14 going to hear and have heard, with the pricing of 15 municipal securities in the negotiated market it 16 typically or usually is the best pricing to get within 17 the marketplace, giving -- all things being equal with 18 structure, callability, and, of course, ratings, that 19 most clients would probably end up preferring to buy more 20 of these securities within their portfolios. 21 MS. MARTIN: Chris, you mentioned your 22 fiduciary duty to clients. Do you believe there are 23 other areas within the federal securities laws that 24 actually protect from conflict of interest, as well? 25 MR. KENDALL: Sorry, are there areas of the law 0101 1 -- 2 MS. MARTIN: Yeah. 3 MR. KENDALL: -- that protect clients? I'm 4 sure. I'm not as familiar with all of that. I just know 5 that what we try to do is make sure that we structure our 6 business to remove any sort of conflict of interest on 7 the client's behalf. 8 So again, making sure that our advisors are not 9 rewarded differently one way or another, and making sure 10 that they are giving -- you know, acting in the client's 11 best interest at all times. 12 And then, on on my side, on the trading side, 13 of course, we have a best execution obligation. Again, 14 we are also somewhat blind to whether -- whatever type of 15 client it is. Of course, you know, currently we have to 16 act as agent in the case of a managed account. And with 17 new issues we have no way to currently facilitate that. 18 Like Anthony mentioned, the reality in that 19 case is that, the way the municipal bond market works, a 20 lot of times you're going to be better off getting the 21 issue in the primary offering. So if we can facilitate 22 that, it's almost serving our clients better. And 23 arguably, we're doing a disservice to clients by not 24 allowing them access to that paper. 25 MR. WINGES: That was one comment I was going 0102 1 to make, as well. Ninety percent of trading in the 2 secondary market on new-issue municipal security occurs 3 within the first thirty days, for the life of that bond. 4 5 You have to ask yourself why. What that's 6 telling you is that the end buyer is not getting access 7 on the new-issue market, and they're willing to pay a 8 different price. And this is one of the barriers to that 9 entry point for that retail investor. And we think, with 10 this recommendation, we'll start to see a change in some 11 of that secondary premium trading on new-issue 12 securities, because the entire retail market, whether 13 it's in a commission-based or a fee-based account. will 14 have access to the new deal without any without any 15 barrier to entry. 16 MS. MARTIN: Michael, I think we're ready to 17 open it up. 18 MR. HEANEY: Sure. 19 MS. MARTIN: Q and A. 20 MR. HEANEY: So let's open it up now to the 21 committee members for thoughts and questions on the 22 recommendation concerning negotiated offerings for 23 managed accounts. 24 Please, Suzanne. 25 MS. SHANK: What impact do you think this would 0103 1 have on retail order periods? 2 MR. LIOTTI: As far as -- further clarifying, 3 as far as a greater amount of orders received? 4 MS. SHANK: Yeah, yeah, a greater amount of 5 orders, or -- 6 MR. LIOTTI: I would anticipate there would be 7 -- yes, you would see that, a greater amount of orders 8 coming in. As I mentioned in my earliest comments, that 9 we're seeing the demographic shift here, especially 10 amongst the most active municipal buyers within the 11 marketplace into a fee-based structure, where if a 12 financial adviser has that blanket consent in which they 13 don't have to go and, you know, regularly speak to their 14 client, I would anticipate that there should be a 15 significant increase in orders. And, ultimately, that 16 could benefit the end issuer of bringing down and making 17 the better price entry for that end issuer into the 18 marketplace, as there is greater demand. 19 MS. SHANK: And I mention that because most of 20 the retail in recent years has been professional retail, 21 you know. So it's, I would anticipate, the same. But 22 any comment from Horace and Brad? 23 MR. WINGES: No, I would totally agree with 24 that. I think any time you bring a significant large and 25 increasingly growing number of assets that now have the 0104 1 ability to participate in that new-issue order period, I 2 have to believe the retail order period is going to see 3 significant more flow. 4 MR. KENDALL: I would second that, as well -- 5 or third, if you will. 6 For us, I mean, these are retail clients. They 7 may be high net-worth clients, but they are absolutely 8 retail clients, and we would see an increase. 9 MR. HEANEY: Gilbert? 10 MR. GARCIA: I'm on the -- this for anybody -- 11 I'm on the institutional side, but the benefits seem so 12 obvious. I guess my question is who would argue against 13 this, and why. 14 You know, I mean, what's the argument on the 15 other side? 16 MR. CARTER: I think the primary concern is the 17 double dipping question. You know, that -- it's -- I 18 think Brad addressed it -- addressed that pretty well. 19 But that would be where the main arguments really stem -- 20 MR. GARCIA: That's really it? 21 MR. HEANEY: Larry? 22 MR. BAGLEY: Can I just -- 23 MR. HEANEY: You want to refer to that? 24 MR. BAGLEY: Yeah. I think one of the things 25 the committee talked about was the reason we took out the 0105 1 negotiated -- I mean the competitive side of the market 2 was, on a competitive deal the dealer's at risk as soon 3 as the deal goes off. On the negotiated side, it's best- 4 efforts basis. 5 So if the deal doesn't happen, I'm not at risk. 6 But once you buy a deal and you're at risk, we're 7 worried about what would it look like if you were also 8 getting a bunch of managed account orders to reduce your 9 liability. So I think that, had we gone for the full 10 spectrum, that would have been a concern if we addressed 11 the competitive side, as well. 12 MR. HEANEY: Larry? 13 MR. HARRIS: So my concern is related. I'm 14 very sympathetic to the proposal, but I'm concerned about 15 potential unintended consequences. So if a deal is 16 under-subscribed, we don't want a situation where the 17 underwriter could push it or cram it down onto clients. 18 That's potentially problematic. 19 And then, a second situation that is 20 potentially problematic is if the deal is really small 21 and it goes only to the underwriter's clients. What 22 confidence do we have in the price? So I wonder if there 23 is -- are these well-placed concerns? 24 MR. LIOTTI: Yeah, I think it's a great 25 question, and I think -- I'd say it's been very well 0106 1 thought-out. 2 I think, to the point that John just made, 3 starting with the negotiated side the market, I think the 4 vast majority of -- I'll speak for UBS's clients -- are 5 traditionally buying high-grade paper. You know, at 6 least BAA or better. And when looking at these deals, 7 the unfortunate thing sometimes in the market is they are 8 priced very well, and they are often times oversubscribed 9 by a multiple of one, two -- I've seen deals five times 10 oversubscribed. So you bring a -- you know, a $100 11 million deal, you might have $1 billion in orders. It's 12 -- you can see that in some instances. 13 So that, I think, should hopefully alleviate 14 any concern about a deal not being priced well in the 15 marketplace. Ultimately, what happens is there is just 16 not enough bonds to go around to meet and satisfy that 17 demand, whether it be through wealth management clients, 18 retail clients, or even institutional clients. 19 The other question, again, I'm sorry, was? 20 MR. HARRIS: So the second question was suppose 21 it's a very small deal, and it goes only to the clients 22 of the underwriter. How do we know about the quality of 23 the price? 24 MR. LIOTTI: One thought -- I'll answer that. 25 One of the thoughts that we had at our firm is perhaps 0107 1 there's an exception, meaning that if it's a sole 2 underwriting deal, to to shed any concerns about, you 3 know, pushing those bonds aggressively, perhaps using the 4 term "dumping," to alleviate any of that concern, 5 perhaps, you know, there's an exclusion in which, if 6 you're a sole book running manager on a deal, that's one 7 consideration that could be made. All right? So it just 8 goes out to institutional clients, or -- 9 MR. HARRIS: That was my thought, as well, that 10 when we get a final draft, that we say that this applies 11 as long as no more than some X percent is going to those 12 clients of the deal, because we need to have some 13 confidence in the deal. And the fact that most of the 14 deals get oversubscribed doesn't give us confidence for - 15 - that there's some bad actors out there, or somebody 16 gets caught in an unusual situation. We don't want this 17 to be crammed down on the clients, and this is -- the 18 existing law provides -- or regulation provides some 19 protection against that kind of cramdown, although it's 20 probably pretty weak, as it is. 21 MR. HEANEY: Can I just follow up on Larry's 22 second concern? 23 Some of the smaller deals, whether they're 24 districts, states, cities with really small tranches, how 25 often do those go to these type of accounts? You know, 0108 1 the kind of two, three, five-million tranche deals. 2 MR. LIOTTI: I would say very little. 3 MR. HEANEY: Very little? 4 MR. LIOTTI: Yeah, to -- it's -- at least in my 5 experience at our firm, the traditional market which 6 we're talking about here would be the more active 7 borrowers within the marketplace in which there is 8 perhaps a larger degree of liquidity, as well. 9 MR. HEANEY: Okay. 10 MR. CARTER: One, I'd just like to quickly 11 address Larry's first question. In terms of getting -- 12 essentially, getting hung on a deal and forcing the bonds 13 down into these kinds -- generally speaking, the retail 14 order period is the first -- will be the first order 15 period. And so, after that, you can have -- and it 16 varies in length. But this only -- this proposal is only 17 -- is concern the order period itself. It does not -- so 18 we're not talking about bonds that have been taken down 19 by the dealer and underwritten. 20 To John's point earlier, in a competitive deal 21 we're essentially -- that would be a transfer of risk. 22 And in this scenario we're not talking about a transfer 23 of risk. So I think that's an important clarification, 24 and we went to a lot of trouble in the process to make 25 sure that that was the case. 0109 1 MR. HEANEY: Sonali? 2 MS. THEISEN: I just had a couple of questions, 3 just for my own understanding or education of the market. 4 Can you give, like, a general sense or estimate 5 of how much is managed versus competitive, in terms of 6 offerings? 7 And then also, another point that was raised is 8 about some of these flows potentially moving to SMAs and 9 mutual funds to get done. Can you quantify or estimate, 10 like, how much of that is happening today in practice? 11 MR. LIOTTI: I don't have statistics handy, so 12 I apologize about -- the actual balance between the two, 13 clearly the competitive deals are, in some instances, 14 less complicated. As was mentioned by the chairman, 15 sometimes smaller deals -- sometimes you'll have large 16 states that do go competitive, and some issuers have a 17 bit of a movement to go more to a competitive bidding in 18 certain instances. 19 I would say that the negotiating market 20 typically is the larger of the two. 21 At our firm, as it relates to -- I think the 22 question was the balance between advisory clients and 23 brokerage clients. Is that the question? We're roughly 24 -- we have about roughly 90 billion of assets of 25 municipal securities held at UBS and by our clients. 0110 1 There's roughly -- half of that is in transaction-based 2 accounts, and the rest of it is split between separately- 3 managed accounts, mutual funds, UITs, ETFs, and then the 4 advisory platform, which handles roughly about 13 billion 5 of assets under management. 6 MR. KENDALL: And just from Schwab's point of 7 view, I believe it's about 80 or 90 percent of the deals 8 that we get involved in are on a negotiated basis. So 9 it's definitely the majority. 10 In terms of clients with -- for managed assets, 11 in terms of the fixed-income world, for us, it's actually 12 a pretty small portion of the pie, probably less than 5, 13 10 percent of our trading on that. But keep in mind that 14 those are our high net worth or even ultra-high net worth 15 clients. And so their interest in individual municipals 16 is a lot higher than the average client would be. 17 MS. THEISEN: Thank you. 18 MR. HEANEY: Mihir? 19 MR. WORAH: Thanks. In one of your comments 20 you mentioned the possibility of allowing clients in 21 these managed accounts to opt out. But that's not any 22 recommendation. What do you what do you think of that, 23 or why is it not in the recommendation? 24 MR. LIOTTI: I am not part of the FIMSAC 25 committee, so those are just my general comments as 0111 1 something that could be considered at some particular 2 point. That's all. 3 MS. MARTIN: I mean I think, in general, we 4 tried to be less prescriptive on the recommendation. 5 There are a variety of ways that we can mitigate conflict 6 of interest as the rulemaking goes forward. So we tended 7 to be less prescriptive in the actual text of the 8 recommendation. 9 MR. HEANEY: Gilbert, did you have another 10 question? 11 MR. GARCIA: Just a quick follow-up to the last 12 question about the sort of distribution between 13 transaction-based accounts and managed accounts, if this 14 rule -- if this change was to go into effect, do you 15 think that that -- you know, the 50/50 split of 50 16 percent in transaction-based accounts versus 50 percent 17 in other -- do you think that this would have a material 18 impact on the -- that distribution? 19 MR. LIOTTI: It's tough to tell. I would 20 probably say no. I don't think this change would 21 materially move assets aggressively over to because of 22 that. I think, as I mentioned, there has been a move 23 into advisory fee-based accounts, and it continues to 24 grow. I think that this makes the ability -- it just -- 25 it works well with the clients who have that ability to 0112 1 once again talk to a client, say -- you know, you're able 2 to go and buy these securities in a fee-based account, 3 and you don't necessarily have to go externally to get 4 access to those solutions. 5 MR. HEANEY: Let me turn it to questions on the 6 phone, if I can. Elisse any questions? 7 MS. WALTER: None here. 8 MR. HEANEY: Thank you. Rachel? 9 MS. WILSON: One of the quick comments is that 10 there is a -- the advantage is that this really helps in 11 the aftermarket and kind of letting people not pay that 12 premium for how the municipal traded thereafter, balanced 13 against a threat of double dipping of fees, which, again, 14 as the industry is moving, seems like less and less of a 15 conflict. But regardless, is there any sense for quantum 16 of some of those average differences? And so that would 17 be a little helpful to hear. 18 MR. WINGES: So, Elisse, what exactly are you 19 asking for? Or Rachel, sorry. 20 MS. WILSON: A sense of scale or scope of the 21 difference between a double dip of fee for a particular 22 municipal security in the aftermarket, secondary, where 23 we are seeing that they generally appreciate -- the 24 difference of a fee versus the general average 25 appreciation in the aftermarket. So what is that risk 0113 1 versus benefit? 2 MR. WINGES: That's a good question. We'd have 3 to do some analysis around that. But, you know, on a 4 typical underwriting on a rated transaction you may have 5 a half-a-point. But we've seen in the secondary side 6 within, you know, days, securities trading up a point, 7 two points in some cases. 8 MR. CARTER: Yeah, definitely material is the 9 answer. 10 MS. WILSON: Yes. 11 MR. HEANEY: Carole or Kumar, any questions? 12 MR. VENKATARAMAN: Michael, I have a question 13 that follows upon Larry's question. It relates to the 14 primary offering that is a weak offering, there is 15 insufficient demand for the issue. And in that case, a 16 concern that that comes up is that investors in these 17 managed accounts might be adversely selected because they 18 receive a substantial allocation for the issue. 19 I think Horace responded that this is not 20 really a transfer of risk from underwriters to the 21 managed account holders. So I was wondering whether 22 someone can elaborate on that, and why it would not be a 23 concern if the primary offering happens to be a weak 24 offering. 25 MR. CARTER: Well, if the -- so, generally 0114 1 speaking, and certainly at Raymond James, retail orders 2 given during the order -- are highest priority orders. 3 So all retail orders are going to be filled -- at least 4 at Raymond James, all retail orders are going to be 5 filled prior to any other orders being filled. So they - 6 - generally speaking, they will get their full 7 allocation, regardless of the strength of the deal. 8 The instance where we sell 100 percent of any 9 negotiated deal retail would be rare, to say the least. 10 But that's -- at least at Raymond James, that is not our 11 experience. Our retail participation in the brokerage 12 side is less than five percent of most of the deals. 13 MR. WINGES: I think to to your question -- and 14 it brings up a valid point -- our proposal here is 15 allowing the managed accounts to participate in a new- 16 issue order period. I think, to your point, if the order 17 period is over and the underwriter is still long the 18 transaction, should there be a barrier against -- I don't 19 want to use the word "stuffing," but dumping, putting 20 those bonds then into managed accounts, I think that's a 21 topic that could be discussed to avoid that potential 22 conflict of the underwriter not having the deal sold, and 23 then using the demand in the fee-based accounts as an 24 opportunity to close out the transaction. I think that 25 would be a valid differentiator. 0115 1 But again, this proposal is addressing allowing 2 them to participate in a new-issue order period, and 3 doesn't address after the fact. 4 MR. HEANEY: I think that goes back to the 5 point of the data that could be looked at, should this 6 make it to the SEC and the SEC staff to try to look 7 through how many of these transactions actually have this 8 issue and how many do not, and should it be addressed. 9 Actually, a little bit of fact-finding and statistics 10 around that. 11 Yeah. Sorry, John. 12 MR. BAGLEY: And let me just point out -- 13 excuse me -- on a negotiated trade it's done on a best 14 efforts basis. So if a deal is weak, normally the 15 underwriter will go back to the issuer and negotiate for 16 higher yields to try to clear the market. If that 17 doesn't happen, the underwriter can walk away and they're 18 at no liability. So that's a huge difference in the 19 negotiated side versus competitive. 20 MR. HEANEY: Which -- correct me if I'm wrong, 21 reiterating what you said earlier -- is why competitive 22 bidding and the competitive deals we're taking out of -- 23 taken out of this recommendation. 24 Other questions or comments? Oops, sorry, 25 Larry. 0116 1 MR. HARRIS: I just had a very quick one. Much 2 of the premise for this is the notion that prices often 3 or very often rise afterwards. I was just wondering what 4 the issuers think about this. So if, indeed, that's 5 correct, aren't the issuers not doing so well? 6 MR. HEANEY: Well, if Carole's on the phone, 7 she'd be the one to perhaps address that. 8 I don't know if, Carole, you are there or 9 choose -- Scott could certainly tackle it on the 10 corporate side, as well. Carole, are you -- care to 11 comment? 12 (No response.) 13 MR. WINGES: Larry, I think to your point, any 14 issuer -- whether you're a corporate, the U.S. Government 15 issuing treasuries, or municipal issuers, any opportunity 16 to find more buyers of your security is a good thing. 17 MR. HEANEY: Yes, Scott? 18 MR. KROHN: So just wanted to share a 19 perspective on the corporate side regarding this agency 20 risk and double dipping. And maybe it's not applicable, 21 because the underwriting here is on a best-efforts basis. 22 On the corporate side it's firmly underwritten. 23 But when we deal with underwriters that also 24 have investment arms, it's frequently the case that the 25 investment arms' order cannot be put into the book until 0117 1 the deal works entirely away from that investment 2 advisor. So that could be one thing to look at as you're 3 looking at this rule is, okay, it's fine for the 4 underwriter and related investment arm to both be in the 5 deal, but only if the order book is sufficient before 6 considering those orders. 7 MR. HEANEY: Other questions or comments? 8 John, are you -- you're good? No, no. No 9 problems. 10 Suzanne? 11 MS. SHANK: So -- thank you. So, you know, 12 most firms have, you know, the wall between their 13 investment advisory arm on new issues. I guess it just 14 occurred to me it wasn't clear on whether that would 15 change as a result of this. That would not change. 16 Yeah. 17 MS. MARTIN: No, we're not recommending for the 18 -- 19 MS. SHANK: Okay, right. 20 MS. MARTIN: -- walls to be taken down between 21 the underwriting -- 22 MS. SHANK: Right. So that -- yeah, that's an 23 additional -- 24 MS. MARTIN: -- and the -- 25 MS. SHANK: Yeah, yeah. 0118 1 MR. HEANEY: Any other questions? 2 (No response.) 3 MR. HEANEY: Okay, at this point I will 4 entertain a motion to vote on this recommendation. 5 MS. MARTIN: Motion. 6 MR. HARRIS: Is there a -- did we want to 7 consider placing a limit on what fraction of the issue 8 could go into these accounts? 9 MR. HEANEY: My opinion, and as I think I 10 offered up earlier, I think that can be done with a 11 little bit of due diligence by the SEC staff, as opposed 12 to trying to do that statistical work from the 13 subcommittee's perspective. 14 So we're voting on it as is, as written. 15 Obviously, the staff, once they receive that 16 recommendation, can choose to do it with what they would 17 wish. But certainly, having heard that, I would guess 18 that that would be taken up. 19 But I'm sure there is -- I think that needs to 20 be evaluated in terms of any adverse consequences to 21 that, as well. So that's, I think, the work of the SEC 22 and MSRB. 23 Thank you, Gilbert. 24 Okay, so let's start. We, obviously, have 25 members on the phone. Let's start with those members in 0119 1 favor of this recommendation. Please raise your hands. 2 (Members voted.) 3 MR. HEANEY: Elisse? 4 MS. WALTER: Yes. 5 MR. HEANEY: Rachel? 6 MS. WILSON: In favor. 7 MR. HEANEY: Kumar? 8 MR. VENKATARAMAN: In favor. 9 MR. HEANEY: Carole? 10 (No response.) 11 MR. HEANEY: I wonder if we've lost Carole. We 12 have 18 for; no one voting against; and no abstentions. 13 So again, this has passed. 14 Thank you for all the work that's been done 15 both -- by the subcommittee. Thank you for the panelists 16 for joining and giving the insights. Obviously, we have 17 a second one, too, so sit tight -- the second one to talk 18 about, as well. 19 But thank you very much, as the recommendation 20 will now be sent to the SEC. 21 Okay, so second preliminary recommendation 22 concerning principal trading and liquidation of bonds. 23 Lynn, I'm going to hand it back over to you for 24 the discussion. 25 MS. MARTIN: And Horace, I'm going to turn it 0120 1 back over you to give a background on this one. 2 MR. CARTER: All right. Thank you. I realize 3 that I forgot to answer the easiest question all day, 4 which was in the process of introducing myself. So I 5 will go ahead and do that now. 6 (Laughter.) 7 MR. CARTER: I am Horace Carter. I'm the head 8 of trading and co-head of fixed income capital markets 9 for Raymond James. Raymond James is a large regional 10 financial services company. We've got a large retail 11 presence and a large public finance group, as well. 12 Generally, a top-10 municipal underwriter. 13 With that, I will introduce the second 14 recommendation. 15 There is a regulatory framework which exists today 16 requiring broker-dealers to comply with best execution 17 requirements. The typical process for the liquidation of 18 client bond positions involves very broad distribution 19 among broker-dealers using various venues, including but 20 not limited to alternative trading systems, which 21 provides a full audit trail of activity, or through 22 municipal broker's brokers or other bid wanted venues 23 which also have record-keeping requirements. 24 If the liquidating account is commissionable, 25 then the order-handling broker-dealer can also contribute 0121 1 a bid. In advisory accounts the broker-dealer is 2 required to sell the bonds to the highest bid away. This 3 is true even if the order-handling dealer would bid a 4 higher price. 5 We believe that the client could receive better 6 execution without undermining the process if a broker- 7 dealer handling such an order could submit a blind bid -- 8 a blind bid being defined as without having access to the 9 bids from other dealers, and requiring the dealer who 10 initiates the auction submitting their best price to that 11 auction at that time for advisory clients via a bid 12 wanted venue. 13 The subcommittee recommends that the SEC 14 consider a rule that permits a broker dealer to meet the 15 requirements of Section 206(3) of the Advisers Act when 16 acting in a principal capacity to sell certain client 17 positions within the normal liquidation process by 18 allowing dealers to submit a blind bid on a principal 19 basis against its advisory clients. 20 MS. MARTIN: Thank you, Horace. I'm going to 21 turn this question to Anthony, Pete, and Horace, 22 yourself, if you've got comments on this one. 23 We talked a lot about protecting investors from 24 conflicts of interest on the last panel. How do you 25 believe this recommendation benefits the retail investor 0122 1 with a view towards mitigating the conflicts of interest? 2 Pete, maybe you could start. 3 MR. SIRBU: Yeah. Pete Sirbu, I'm with 4 Ameriprise Financial, and I'd like to thank the committee 5 for allowing me to attend. I oversee trading and 6 strategies that includes the taxable and tax exempt 7 principal trading that occurs at our firm, all fee-based 8 advisory order flow, as well as the strategies group the 9 handles and helps build client portfolios. Ameriprise is 10 based out of Minneapolis, Minnesota. We have roughly 11 9,500 advisors at the firm. 12 So, to answer your question, the benefits to a 13 retail advisor and to a client is basically the 14 additional service we can offer them by being able to bid 15 blindly in the marketplace. 16 One of the things that we see, especially 17 during dislocated markets, high volatility, or an 18 illiquid securities, is a tendency for -- maybe you have 19 a number of bids that are still coming in the system, but 20 the quality of those bids starts to wane, and that just 21 makes sense. As volatile markets increase, spreads seem 22 to widen. And what we find is, as these bids are coming 23 into the system, my fee-based traders, who only have 24 access to the street -- not our principal traders -- at 25 times cannot send back a bid to the client. It may take 0123 1 them all day to send back a bid to the client. So 2 believe that the price is fair and reasonable under best 3 execution standards. 4 So the ability to incorporate a principal bid, 5 who has the best interest of the clients at heart because 6 it is our client, I think, is very beneficial. 7 We also do find at times of extreme duress in 8 the marketplace dealers backing away from the bids that 9 they provide. And what we have instructed our traders to 10 do is, if you do see a trader, back away. And the price 11 is fair or remotely fair, still bid those bonds and trade 12 them at that price. Again it's all about servicing the 13 end client. So that's, I think, the benefit. 14 How do we mitigate some of the risks that 15 obviously occur? I believe our proposal is only focusing 16 on the bid side of the marketplace. Therefore, any 17 concerns about dumping have mitigated. 18 And, of course, like I mentioned before, we had 19 the best execution mandates under G18 and FINRA 53-10. 20 We, as the broker-dealer with the end client, are 21 required to bid a price that's fair and reasonable. And 22 we need to follow the mosaic of data points that allows 23 us to make certain that we are doing that. It's very 24 difficult to do if you're a fee-based trader and don't 25 have access to commit capital. 0124 1 MS. MARTIN: Anthony? 2 MR. LIOTTI: I agree with everything that was 3 said. You know, the term "liquidity" has come up quite a 4 bit, and I clearly think that this provides that 5 additional layer of liquidity. 6 In most markets it may not be needed, but to 7 the point that was made by Peter, the concern that we 8 have is the markets over the last couple of years have 9 been fairly -- acting quite well, right? We haven't seen 10 those significant sell-offs. But when we do it has 11 dramatic, dramatic impacts, and a ripple effect across 12 the marketplace. 13 The inability of clients who have an advisor- 14 based account -- what typically happens is every firm has 15 advisory-based accounts, every firm is a receptor of 16 receiving bid wanteds. Most firms step away. They're 17 only then in those markets with this liquidity focusing 18 on bidding the bonds for their own wealth management 19 clients. And so the liquidity in the street dries up 20 dramatically. 21 Any bids that are given back to the client can 22 at times be extremely low. Most times financial advisors 23 and/or clients are evaluating that bid based on the 24 evaluation given out by an appraising service. They 25 decide not to be those particular bonds. And in the need 0125 1 of liquidity they typically go to the exchange markets to 2 get cash and/or liquidity. So, i.e. going and 3 liquidating ETFs, funds, things of that ilk, or 4 professional managers. And then there is, then, that 5 ripple effect, right? 6 And so in normalized markets, it -- perhaps 7 it's not as much needed. But clearly, I think having 8 that backstop and protection when those markets do go 9 awry is just as important. 10 MS. MARTIN: Horace, did you want to offer some 11 comments here? 12 MR. CARTER: Well, in the interest of not 13 repeating the same thing with which I agree that my 14 fellow panels have said, the only thing that I would add 15 is that the risk associated with this is more of a 16 general market structure risk, in not really -- it 17 doesn't really have to -- it doesn't apply on a trade-by- 18 trade basis. 19 And so the risk associated with it, because of 20 the price discovery that is available now in the market, 21 the original concerns around this prohibition aren't 22 really applicable anymore, and the rule in this context 23 is a little archaic because we can validate the price, 24 and that was really the point, I think, of the rule. 25 MS. MARTIN: Okay, thank you. 0126 1 Marshall, I'm going to turn it to you. 2 Obviously, the ETF infrastructure provides a pretty 3 robust regulatory framework for monitoring this type of 4 activity. I know you've looked at a lot of data in 5 particular around volatile periods, things of that 6 nature. Would you mind reminding the FIMSAC of the 7 protections in place, as well as what you've observed 8 during volatile periods around responses to bids wanted, 9 spreads, things of that nature? 10 MR. NICHOLSON: Sure. Thanks, Lynn. I'm 11 Marshall Nicholson from ICE Bonds. We operate three 12 ATSs, two of which provide municipal bond execution, 13 often with those executions being driven by retail 14 activity. 15 Just to kind of take a step back here, you 16 know, when you think about the electronic platforms, the 17 different types of regulatory frameworks from which these 18 platforms operate, I kind of think about the notion of 19 all retail bid wanteds being sent to a ATS, an 20 alternative trading system. It's a bit like going to 21 heaven, in that everybody thinks it's -- everybody wants 22 to do it, just not today. 23 And that's kind of been the approach that -- 24 I've been thinking about this, and if you guys would 25 indulge me for a second, I'm going to just read some 0127 1 prepared comments that should talk about some of the 2 questions that you proposed, Lynn. 3 So today we have this opportunity to add 4 incremental liquidity and better execution for the retail 5 investor. That's ultimately what our end goal is. It's 6 obvious to me that there's a good probability for both 7 with the surveillance that is done on ATS platforms as 8 part and parcel of operating a fixed-income marketplace. 9 10 While the benefits of this activity are 11 apparent, it should be done in a thoughtful way. ATS 12 platforms should be the foundation of this market 13 solution. ATSs are the most heavily-regulated of any 14 trading platforms in the fixed-income markets. Examples 15 include 15c3-5 compliance, post-trade transparency with 16 ATS reporting requirements, date regulators and market 17 surveillance through ATS data that are mandated to be 18 made available to them. 19 Furthermore, ATS platforms are a key source of 20 pre-trade transparency that now exists for retail 21 investors. And finally, ATS platforms are the most 22 transparent venues in that execution protocols are 23 required to be made available for subscribers, while 24 other platforms are not held to this standard. Mandating 25 this activity will result in the most clarity and the 0128 1 most consistency, resulting in credibility for all market 2 participants. 3 It is my belief that the most pragmatic 4 industry approach to handle principal trading and 5 advisory accounts with the primary goal of benefitting 6 the retail investor is to allow principal desks to 7 provide liquidity on these bid wanted requests from 8 advisory accounts, but require that the transaction take 9 place on an ATS at a blind auction. This will ensure 10 maximum dissemination of the bids wanted, and that there 11 is consistency in the auction process, both of which aid 12 in best execution. 13 An affiliate principle desk should bid as any 14 other participant in the competition for the advisory 15 order. This results in very little information 16 asymmetry. The best price wins, it's a level playing 17 field. There is a consistent order audit trail for each 18 auction submitted under this scenario, as described 19 above. The retail investor has the highest likelihood of 20 benefitting. 21 But allowing these auctions to take place via 22 other less regulated structures opens up variables that 23 may not ultimately benefit the retail investor to the 24 same extent. Inconsistent processes, non-standard 25 protocols, and less transparency for regulators and 0129 1 retail investors are all a distinct possibility. 2 And, Lynn, to get specifically to the question 3 that you asked me, we did do some analysis of what type 4 of -- in a volatile versus non-volatile markets, you 5 know, what is the impact on the amount of bid wanteds 6 received in the periods that we looked at. The number of 7 bid wanteds received was pretty consistent from volatile 8 and non-volatile times. Probably more importantly, the 9 number of bid wanted bids received back for these bid 10 wanteds was roughly in line -- very immaterial 11 difference. And I would say the distance to cover, 12 meaning that the winning bid price -- what was the cover 13 bid -- was slightly wider during a volatile period. 14 So, from the data that we looked at, it would 15 be hard to make any kind of observation that volatile 16 periods versus non-volatile periods would be impacted, 17 and the quality of bids that a retail investor might 18 receive. 19 MS. MARTIN: Thanks, Marshall. So you touched 20 on the concept of blind bid and, obviously, that's one of 21 the core components of this proposal. This has been a 22 topic of some debate amongst the subcommittee, 23 particularly in the last couple of weeks. So I thought 24 we would take a few minutes to just unpack this debate a 25 bit for the benefit of the broader FIMSAC, as well as the 0130 1 SEC staff. 2 So, as reminded, in the current recommendation 3 blind bid is currently defined as -- to mean the 4 submission of a bid by a dealer handling customer order 5 without having access to bids from other dealers at the 6 time the broker-dealer submits its bid to the bid wanted 7 venue. The broker-dealer would be aware that it's 8 bidding on its advisory customer's order, but it would 9 not have any information about the other bids submitted 10 to the auction for that customer order at the time the 11 broker-dealer submits its bid. 12 Anthony, Pete, and Horace, what are your views 13 on this aspect of the recommendation? Is blind really 14 required? 15 I would also love your views, Marshall, on 16 this, as well. 17 MR. SIRBU: I guess I can speak to that. I 18 understand the practicality of a blind bidding, and I 19 think it makes sense. However, traders -- at least in 20 our firm -- wear two different hats. 21 One is obviously a liquidity provider, and the 22 second is oversight for best execution. So, in order to 23 get to a blind bidding type of environment, that would 24 require our desk -- some pretty big changes, 25 operationally. We would, obviously, have to have another 0131 1 group that would be out there taking into account our 2 bid, as well as the street's, and then determining 3 whether or not the price is fair and reasonable. Today 4 we're not set up that way. 5 When I look at smaller firms who maybe don't 6 have a segregation between their fee base and their 7 principal desks -- unlike ours, ours does segregate -- I 8 can see that actually creating maybe an undue burden on 9 those firms, because they would need to bring in these 10 additional staff members to actually perform that task, 11 just given the fact that they wouldn't be able to have 12 that last-look to make certain that best ex was actually 13 being applied. 14 MR. NICHOLSON: From our perspective, the last 15 panel brought forth some issues that exist. You know, 16 assuming that you do this on a blind basis, you know, you 17 eliminate certain intricacies of the municipal bond 18 market around, you know, the pennying or -- you know, or 19 anything else. 20 Where -- you know, I touched on it in my 21 prepared remarks that the -- there is no information on 22 symmetry, so the best bid wins. It would be hard to 23 argue that anyone would feel that that process was not 24 fair for all market participants. 25 MR. LIOTTI: I would agree with that. So to -- 0132 1 not necessary to expand, I think that, you know, it's 2 putting your best foot forward the first time. 3 I think each market is different, though. And 4 the concern that I personally just have is, when you do 5 get in those markets, a complete dislocation, that that 6 last-look or best ex is extremely important. 7 As far as the pennying, you know, I look at it 8 as, you know, personally it ultimately hurts certain 9 firms because you lose that liquidity. You know, you've 10 sort of been identified as those, you know, typical folks 11 who do do that, the frequent folks who are involved that 12 market. And then, ultimately, I think that other dealers 13 begin to realize that, and they just step away from the 14 bidding process all together, and that ultimately will 15 hurt them and their clients. 16 MR. CARTER: So it's little bit embarrassing 17 that I'm about to speak against the blind bidding, since 18 I'm the one that originally put it in there. 19 (Laughter.) 20 MR. CARTER: But -- you know, full disclosure - 21 - but it's not the first time I've been embarrassed. 22 MS. MARTIN: It's been an interesting couple 23 weeks. 24 MR. CARTER: So -- yeah. But two things. 25 One, so Raymond James as an example, about 16 0133 1 percent of our clients or our advisory clients. And the 2 implementation of this new workflow would be very 3 expensive. And so, if this were to -- if we were to make 4 this recommendation, and then this alternative method to 5 apply it with 206(3) were in fact initiated, we wouldn't 6 do it because we would not be able to afford the 7 implementation costs of it. 8 I also believe that it's unnecessary. It's a 9 market structure question that we addressed in an earlier 10 panel. And the -- it does not have anything to do -- the 11 blind bidding aspect of this proposal does not have 12 anything to do with the actual execution price of that -- 13 of this bond at this time. We are still under the 14 obligation to have -- our best ex requirements are still 15 in place. The blind bid does essentially nothing to 16 improve that. And as Anthony points out, in -- when 17 markets become illiquid or distressed or dislocated, that 18 is when you will see what, for lack of a better term, 19 "predatory behavior," which it is our responsibility to 20 protect -- from which to protect our clients. 21 So that -- so I don't believe it's necessary. 22 And if it remains in the final recommendation, I believe 23 that it will result in lack of implementation by most of 24 the broker-dealer community. 25 MS. MARTIN: Go ahead, Chris. 0134 1 MR. KENDALL: Hi, thanks. You know, I just 2 want to echo what Horace said a bit. I think, given our 3 current systems and practices, this would be burdensome 4 and difficult to adapt to a bifurcated experience in 5 processing our bid requests. 6 And I don't honestly understand the value too 7 much of requiring it to be a blind bid, and it's already 8 getting discussed. I would prefer to defer what -- to 9 whatever suggestion or guidance the MSRB comes out with, 10 and make that a consistent experience. 11 From our point of view, you know, we have best 12 execution. In our model our advisor is actually blind to 13 whether it's a Schwab bid or not. They really don't 14 care. 15 And then, at the end of the day, the advisors themselves 16 have the fiduciary responsibility on the level. So any 17 liquidity we could provide would be good. 18 MS. MARTIN: Michael, I think we should 19 probably open it up to the broader FIMSAC, as I'm sure 20 there are some questions on this topic, in particular. 21 MR. HEANEY: I would imagine there will be, 22 since the blind bidding concept is basically -- 23 MS. MARTIN: The core of the proposal. 24 MR. HEANEY: At the foundation of the 25 recommendation. 0135 1 MS. MARTIN: Right. 2 MR. HEANEY: So why don't we -- we're heading 3 into uncharted territory at this meeting. 4 MS. MARTIN: Yeah. 5 MR. HEANEY: Ananth, why don't we start with 6 you, please? 7 MR. MADHAVAN: Sure. Horace, you've just 8 mentioned predatory behavior in times of market stresses. 9 I just wonder if you could just tease that out for the 10 committee a little bit more. 11 MR. CARTER: I knew I -- I knew I shouldn't 12 have used that word. 13 So a fallback example is -- the most recent 14 example, you've got the late '08, and then we also saw a 15 bit of it in the May and June of 2013 -- very stressed 16 markets, and there was a lot of retreat from the 17 marketplace, just in general. Broker-dealers essentially 18 provided little or no liquidity at all, even to their -- 19 even to their best clients. There was very little to be 20 done. 21 And so, instead of alternative solutions, which 22 you would see as bids -- I can think of specific 23 examples, some large blocks of taxable -- high-grade 24 taxable municipal bonds, where the bid versus the 25 expected price was down eight points. In the -- and 0136 1 we're talking about seven million bonds, you know, a real 2 block of bonds. And then the cover bid was down 10 3 points from that. So that's predatory, you know, for 4 just explanation purposes. That is taking advantage of a 5 market that -- and so, what did we do? We didn't sell 6 the bonds, of course. We -- that was an institutional 7 order, but we controlled that situation. 8 But the -- in an event where -- put a retail 9 client on the other side, and instead of 7 million bonds, 10 let's put 700,000 bonds on there. And they're afraid, 11 and they're an advisory client, and they say, "We want to 12 sell it." Well, what if those are our options? We were 13 selling this. We're selling those bonds. And so that's 14 what I mean by predatory. 15 MR. HEANEY: John? 16 MR. BAGLEY: Can we talk about why this would 17 be so difficult? Because I'm confused. 18 When a bid wanted comes into your firm, you 19 know whether it's advisory or brokerage, correct? And so 20 Marshall, as an ATS, you would like to get those bid 21 wanteds if at all possible. Couldn't you build a 22 functionality that, if someone tells you that this is a 23 advisory account, that you don't release the bid to the 24 firm until they, one, bid or, two, reply that they're not 25 going to bid? 0137 1 MR. NICHOLSON: Yeah, John, that's very 2 possible. I would say that the way that two of our ATSs 3 are set up is that they would, in fact, already follow 4 through what would be a process. 5 But then that process -- then, for example, if 6 something was for -- from the advisory side that was a 7 bid wanted, it just would never be sent, it would be 8 suppressed from going to the principal desk side. So 9 it's just really a configuration for the client because 10 they use our ATS, anyway. 11 MR. CARTER: This is going to be odd, because 12 I've got a question for another panelist. 13 So is it possible -- how much trouble is it to 14 set up two distinct workflows on your system, one where 15 the bids can be gathered prior to release and execution 16 and the bond bid process? 17 MR. NICHOLSON: Horace, I take it that question 18 is for me. 19 MR. CARTER: Yes, this one is for Marshall. 20 MR. NICHOLSON: Thanks, Horace. I would say 21 it's more -- you know, I can't speak to, you know, all 22 ATSs. You know, ours are fairly configurable to support 23 a variety of workflows. And if that was a workflow that 24 someone needed, I think it would be a configuration. 25 Now, I can't speak to what operational procedures would 0138 1 need to change on a trading desk or, you know, written 2 supervisory procedures or anything, but from a technology 3 and pragmatic approach to letting that flow happen, I 4 would say it would be a small project. 5 MR. SIRBU: I believe, from a technology 6 perspective, it probably wouldn't be too difficult to 7 change our processes today. But I take it one step 8 further and look at the potential for collusion between a 9 principal trader and a fee-based trader. I know in a 10 number of firms they sit very close together. At some 11 firms they're at the same desk. So putting some type of 12 a wall between those two so that communication couldn't 13 occur would be necessary. And Marshall mentioned -- some 14 WSPs, some type of supervisory procedures just to make 15 sure that those -- that that is not occurring between 16 desks would be required. 17 MR. HEANEY: So can I just follow up on that? 18 Because what I thought I heard was two different things. 19 That cost issue, Chris, as you were saying, to kind of 20 build the pipes to create this, I was going to ask about the 21 philosophical issue, which I -- Peter, you've just 22 touched upon, but the -- versus the philosophical issue 23 of bidding blind. 24 Away from the issue you raised, Horace, are you 25 saying you want to be able to take that out so that 0139 1 you're putting your capital towards those clients? So 2 philosophically you've got that issue? Because I'll get 3 back to the cost question in a second. I'm curious on 4 the -- away from cost, just the philosophical side, what 5 the panel thinks on it. 6 MR. CARTER: So my answer is yes, from the -- 7 for the same reason that was discussed in the earlier 8 panel about why there are legitimate reasons to step in 9 at the last minute and improve a price for your client 10 for those same reasons. From a philosophical standpoint, 11 yes, we would like to be able to do that. 12 MR. WINGES: Yeah, I'll just jump in here. I 13 think blind bidding actually improves the integrity of 14 the market, and I think it mitigates some of the concerns 15 around pennying and other issues. 16 If a firm is willing to commit their capital at 17 a certain price, that price should not change based on 18 what they're seeing in the open market after they get 19 visibility of all the other street bids. The price is 20 the price. And where you run into a problem is where -- 21 based on where other prices in the open market are 22 occurring, a firm will change their price. Hence, the 23 pennying and stepping in front. 24 So you avoid that in the bidding process. And 25 if a firm, you know, like Horace or others, say we don't 0140 1 want to add staff so we have blind bidding and not buying 2 bidding, well, they can choose not to bid. You know, 3 that's -- and then they follow their normal procedure and 4 go out to the open market and sell all the bonds at the 5 highest bid in the -- in that open architecture. 6 But philosophically, back to your question, I 7 think it's a necessary structural change that we need to 8 make in order to prevent some of the activity that was 9 referenced in the earlier committee meeting, and allow 10 the investor to arguably have the balance sheet of that 11 firm that is running their money stepping into that bid. 12 MR. LIOTTI: I'll simply add to that last-last 13 point. My viewpoint is those -- you know, these folks 14 who are looking for liquidity have been clients of our 15 firm for -- it could be 10, 15 years that they've held 16 those securities. And I think it's incumbent upon us to, 17 you know, to be able to at least provide them with that 18 extra level of liquidity. 19 You know, as it stands, there has been a 20 ongoing fee. And I look at it from the standpoint of 21 saying, whether it's a good market or bad market, I -- 22 from the ethical standpoint perhaps, or philosophical, 23 that us stepping up and having the ability to provide a 24 bid would be helpful. 25 MR. HEANEY: Rick? 0141 1 MR. MCVEY: I just want to follow up on this 2 point again. I think it's what Horace was getting at, 3 but I come at this the same way, that the more bidders 4 and market-makers in auctions, the better off clients 5 are, which is where we all want to be. 6 I guess I'm trying to square this with the work 7 that we did on the issue of last-look, where I think the 8 committee started with that point that fair auctions and 9 blind bidding or blind offering is the best way to create 10 a competitive and fair market structure. But the reason 11 that last-look existed was because of the retail broker- 12 dealer obligation around best ex on the behalf of the 13 client, and that it was critically important because of 14 the velocity of orders, and the fact that you needed that 15 last check before confirming that the price is compliant 16 or consistent with your view on best ex. 17 So what I'm wondering here is, from a practical 18 standpoint, if that velocity of orders for advisory 19 clients could grow to the same level where it's the same 20 issue, where what you're saying is that we can't 21 practically bid in a blind auction with our best ex 22 obligations. But there is an inconsistency here, in my 23 opinion, from where -- on where we're going with the 24 proposal on advisory clients versus a retail order that - 25 - I'm having a hard time squaring that circle. That's 0142 1 the first part of the question. 2 The second one was this all seems to be in 3 retail around bid wanted in comp, and I know that the 4 practice today is streaming offerings on what's available 5 in inventory. But is there also an opportunity to 6 improve competition through offer wanted in comp, where 7 multiple dealers are involved in the same bond, and the 8 market develops through competitive auctions on the 9 offered side? 10 MR. NICHOLSON: Rick, you bring up a very good 11 point. Just kind of taking a step away from municipal 12 bonds, we also looked at some data and we used a very 13 large advisory client that has a very large principal 14 trading desk, and we looked at when they came in to lift 15 an offer. And, of course, the price that was provided by 16 the principal desk was suppressed. 17 What did that mean? In about six percent -- 18 I'm sorry, eight percent of the time the principal desk 19 actually had a better level than where the corporate bond 20 offer side of the market was executed. 21 You know, munis are a little bit different. 22 It's hard to look at it the same way, because there is no 23 market depth. You know, you might have only one dealer 24 who is offering a given security, but, you know, at the 25 same time, a lot of these retailer orders are generated 0143 1 not CUSIP-specific, but by category searches. 2 So, for example, I'm interested in Virginia GO 3 bonds that mature in four years that have a coupon of 4 five percent, and then they use the ATSs to go find bonds 5 matching that criteria and, from there, make a investment 6 decision to have the trade executed. 7 I think there is a lot of value in the offer 8 side of the market for municipals, and then, you know, 9 also for corporate bonds, and is something that I would 10 encourage the committee to take a look at at a future 11 date. 12 MR. KENDALL: I'd like just to add to that. I 13 mean we've talked about the client sell side of the 14 market, but the bigger pain point for our clients, 15 actually, is the buy side. They'll see -- I mentioned 16 earlier we have advised clients who have regular 17 brokerage accounts. They'll see levels or bonds that are 18 available in that account that are not available in their 19 managed account. 20 And again, that's a significant pain point for 21 our clients, probably of all the things we've discussed 22 today, the biggest one, or the one that we hear the most 23 complaints about. 24 MR. SIRBU: Or to elaborate, they'll see our 25 principal desk's offering, and the offering in the 0144 1 street, and principle desk's offering is better, but they 2 can't lift that offering. 3 MR. KENDALL: Right. 4 MR. HEANEY: Our last question, let's go to 5 John. 6 MR. BAGLEY: Sure. I just want to bring up one 7 of my concerns about eliminating blind bidding is, the 8 way it works right now in a managed account, is you can't 9 bid it, but you collect a fee, so it's -- annually you 10 get a fee from this account. 11 If we put in a process where you can collect a 12 fee and you can last-look or bid -- you know, 13 internalize, does that become a very profitable business, 14 and now there's a conflict for the firm to try to move 15 more people into the managed account business? 16 To me, the blind bidding basically said you put 17 your best foot forward, and if you want to do that, good 18 for you. Here, if I can collect a fee and potentially 19 last-look, I'd be concerned about unintended 20 consequences, being a regulator for four-and-a-half 21 years, and that being the growth in managed account grows 22 maybe for not all the right reasons. 23 MR. CARTER: Well, John, that's a fair point. 24 I mean that's -- my I guess my counter-argument would be 25 that the -- one, managed accounts are growing, anyway, 0145 1 for reasons obviously having nothing to do with this. 2 But, in addition, what ultimately matters is the price, 3 is the price that the client receives. And the basis of 4 the recommendation is if we can validate, we can validate 5 our price. That's the -- that's, ultimately, what it 6 boils down to. 7 As I said at the outset, the blind bid element 8 is a market structure question, not an execution question 9 for the client. And so -- and Brad's point is right. We 10 can choose not to bid. That we can -- this -- assuming 11 this goes all the way through, and it becomes allowable, 12 we could easily just choose not to do it if it's -- if we 13 don't deem it in our best interest or our client's best 14 interest. But I think that if it goes through, it's 15 probably supposed to be in the client's best interest, 16 and I just don't think it's necessary. I think that the 17 validation of the price, the regulation around it, is -- 18 around best ex is robust. And that would be my -- but 19 it's a fair point. 20 MR. MCVEY: I just wanted to add one more 21 element here. 22 I think Horace brings up a very good point, 23 that, you know, in stress and certain times, that -- 24 knowing that it's your client, you know, you're willing 25 to step up when others wouldn't. 0146 1 That being said, I mean, you could always look 2 at this as having a rules-based system with certain 3 things have to be achieved before it's -- perhaps where 4 it starts off as blind and stays blind based upon the 5 number of responses that might be received. 6 So, for example, someone puts a bid wanted out. 7 I only get one response back. At that point you're 8 opening up a gray area. Well, does it make sense to 9 expose that order back to the principal desk, knowing 10 that there was only one bid? And what was that bid? 11 Because they may very well step up and provide liquidity 12 for their client. But if there's three bids, or four 13 bids, or five bids, at some point I think the price 14 formation is pretty well established there. And, you 15 know, maybe those are the types of scenarios where they 16 are just bidding blind. 17 So something, you know, to be thought about to 18 try to do, again, protect the retail investor, and 19 particularly when these markets get volatile. 20 MR. HEANEY: So, first and foremost, I just 21 want to thank again the panel for all the thoughts pre- 22 call, here at the meeting, that -- I know there was a lot 23 that went into this discussion. So thank you for taking 24 the time. Thank you for sharing it with us. 25 Then I'll come back to you and pose the 0147 1 question, given that this is somewhat of the foundational 2 aspect to this, the blind bidding, and what we've heard 3 both from the panelists and some of the subcommittee, 4 which -- it wouldn't be the first time that there hasn't 5 been unanimous thought process on the subcommittee -- 6 what you choose to do here with this, because it would 7 seem to me -- and I'll -- you certainly -- we can -- I'm 8 certainly, as subcommittee chairman, would like to hear 9 your thoughts. But it seems to me we're supposed to a 10 little bit go back to the drawing boards here and delve 11 into this. 12 MS. MARTIN: Well, I think there's a lot -- 13 been a lot of good questions raised, and I would not be 14 in favor of amending the recommendation on the fly, 15 particularly given all the dialogue that has occurred at 16 the subcommittee level. 17 Particularly concerning are the last -- two 18 panels ago, the Electronic Trading Subcommittee. Perhaps 19 we take this, reconvene the subcommittee, look at what 20 comes out of the Electronic Trading Subcommittee around 21 pennying and that process, and put some additional 22 clarifications into this recommendation, and then re- 23 present at the July FIMSAC meeting. 24 MR. HEANEY: I'm comfortable with that. 25 Okay, this has been an incredibly helpful 0148 1 conversation. And again, there are times when sometimes 2 further thought analysis is a good thing, and this is 3 actually one of those. So thank you for the time, thank 4 you, Lynn, for the leadership on this. 5 And we are going to abbreviate, unfortunately, 6 the lunch. Why don't we try to be back here at 1:45, and 7 we'll kick off with the afternoon session? Thank you 8 very much. 9 (Whereupon, at 1:07 p.m., a luncheon recess was 10 taken.) 11 A F T E R N O O N S E S S I O N 12 MR. HEANEY: Okay, welcome back. I know we're 13 running a little bit behind schedule. We'll try to make 14 some of that time up. But we're going to begin the 15 afternoon session from hearing from our three 16 subcommittees on the work they've done since the October 17 meeting. 18 We'll start by hearing from Amy McGarrity, who 19 is the chairman of the Credit Rating Subcommittee. As 20 you all remember, that's the newest subcommittee that was 21 formed in early February in response to what we learned 22 at the October meeting. I would like to thank Amy for 23 taking the leadership role on this subcommittee. We've 24 greatly benefitted from your dedication and leadership on 25 it. And as we go forward with some of these difficult 0149 1 topics, we look forward to your leadership in the future, 2 as well. So I'll turn it over to Amy. 3 MS. MCGARRITY: Thanks Michael. 4 As Michael said, at the October 29th 2018 5 FIMSAC meeting we heard from an expert panel who 6 presented on the role of credit ratings in the corporate 7 bond market. As a result of that panel, the FIMSAC 8 determined that we needed to establish a subcommittee to 9 further explore credit ratings. 10 In order to assist the subcommittee, SEC staff 11 provided background information on existing rules and 12 regulations, as well as the SEC's Office of Credit 13 Ratings, or OCR, annual report on ratings agencies. The 14 subcommittee has subsequently held four calls to further 15 discuss these topics. 16 On the February 21st, the committee discussed 17 what topics to focus on. Specifically, we considered 18 four potential topics: one, the use of credit ratings by 19 various market participants, and the implications of 20 rating changes for these market participants; two, the 21 costs and benefits of the current model for credit rating 22 issuance; three, the U.S. regulatory regime for credit 23 rating agencies registered as NRSROs; and, four, 24 issuances of unsolicited credit ratings and the 25 publication of commentaries. 0150 1 We began to concentrate our immediate attention 2 on the current ratings agency payment model structure. 3 As you know, rating agencies have various payment 4 structures. Ratings are paid for by the issuer of the 5 security, with the exception of unsolicited credit 6 ratings, which are not paid for. And investors pay for 7 access to ratings and research through subscriptions. 8 The subcommittee was interested in 9 understanding why this is the case. What is it about the 10 market structure that has caused this to evolve? And 11 does the issuer paid model present conflicts of interest 12 and encourage concentration in the NRSRO market? We also 13 proposed the idea of whether a disinterested third party 14 should be involved to administer the ratings process. 15 In our next call on February 27th, OCR staff 16 provided an overview of the U.S. regulatory regime 17 applicable to the credit ratings process. 18 In general, OCR provides oversight of NRSROs by 19 monitoring the activities and conducting examinations of 20 registered NRSROs, of which there are currently 10, to 21 promote compliance with regulatory requirements. These 22 requirements are generally designed to mitigate conflicts 23 of interest and promote effective internal control 24 structures for determining ratings, disclosure and 25 transparency of the ratings process, good governance, 0151 1 accountability and competition. 2 OCR publishes an annual report available on 3 their website which outlines generally, without naming 4 names, OCR's essential exam findings. The subcommittee 5 discussed the pros and cons of the current NRSRO model 6 and potential concentration issues amongst NRSROs, the 7 role of OCR in industry oversight, the BBB credit 8 universe and whether there may be a relationship between 9 the relatively large size of the BBB universe and the 10 current structure of the NRSRO payment model, and the 11 creation of a ratings oversight board. 12 On March 13th we hosted a panel discussion 13 including two smaller NRSROs, as well as a representative 14 from the academic community with expertise in ratings 15 agencies. Our discussion with the smaller NRSROs 16 revolved around competition and why a standalone 17 subscriber payment model has not been viable. 18 The smaller NRSROs believe the investor paid or 19 subscriber paid model has not been broadly adopted 20 because investors do not demand it. Specifically, 21 investors do not reward another rating on a given 22 security and investor guidelines, and index composition 23 require ratings from only the largest two or three 24 NRSROs. 25 In addition, due to no interaction with issuers 0152 1 ratings material -- sorry. In addition, due to no 2 interaction with issuers, material non-public information 3 is not disseminated to non-hired NRSROs, putting them at 4 a disadvantage versus hired NRSROs. 5 Among panel participants there was broad 6 consensus that meetings with management are a critical 7 component to effective ratings diligence, and without a 8 relationship with the company there was really no 9 incentive for issuers to engage with other NRSROs. 10 The academic contributor on the panel described 11 a 2009 SEC proposal. The proposal would have required 12 certain registrants to disclose preliminary ratings 13 received from certain rating agencies, and was intended 14 to address ratings shopping and potential conflicts of 15 interest. This proposed rule was ultimately not included 16 in the Dodd-Frank legislation. 17 OCR staff also provided an overview of the 18 SEC's rule 17G-5 program, which was intended to address 19 conflicts of interest, as well as promote competition and 20 improve the quality of credit ratings for structured 21 finance products. Rule 17G-5 requires hired NRSROs to 22 rate these products, to have a website listing each such 23 product they are in the process of rating, and the 24 website address where the issuer posts all information 25 given to the hired NRSRO for an initial or surveillance 0153 1 rating. This information is required to be posted on an 2 issuer's password-protected website at the same time it 3 is provided to the hired NRSRO, and to be made available 4 to other NRSROs that provide certain certifications. 5 The idea was that non-hired NRSROs would access 6 the data and develop ratings that could be used by the 7 market alongside the hired ratings. On the March 27th 8 call our objective was to further explore this rule 17G-5 9 program, and whether it could make sense to expand it to 10 other fixed-income products. We heard from a panel of 11 corporate issuers, which are also active issuers of 12 securitized products, as well as representatives from two 13 smaller NRSROs, the same firms which participated in our 14 prior call. 15 From the issuer perspective, while there was 16 general support of the mission and purpose of rule 17G-5, 17 the sense was complying with rule 17G-5 was unnecessarily 18 burdensome, requiring website maintenance and additional 19 staff time without much uptake by the market. 20 In addition, as discussed earlier, the 21 important conversations with management must be recorded, 22 or minutes must be taken and posted which may impair the 23 free flow of information and discourage transparency. 24 The smaller NRSRO's views were that rule 17G-5 may be 25 increasing transparency, access to information, and 0154 1 might, if changed, improve research and models. However, 2 competition has not increased, as unsolicited ratings 3 remain uneconomical. In other words, there is limited 4 willingness to pay for them. 5 In addition, rule 17G-5 is burdensome for a new 6 and NRSRO. Specifically, in order to have access to the 7 secure websites it must either access fewer than 10 8 issuer websites or rate at least 10 percent of the issued 9 securities for which it accessed 10 or more of those 10 websites. 11 There was also a sense that rating agencies, on 12 an unsolicited basis, implies some risk to the NRSRO for 13 which they may need compensation and may not be getting. 14 We also discussed unsolicited ratings and 15 comments that public information is sufficient to provide 16 a credible credit rating, but that they may be 17 conservatively biased due to a lack of engagement with 18 company management. 19 The overall takeaway from our study of rule 20 17G-5, incorporating feedback from issuers and smaller 21 NRSROs, is that it may not be working as intended. And, 22 as such, we do not believe there is a need to expand it 23 to corporates or municipal securities. 24 That concludes my general summary of the work 25 we have done so far. It has been a whirlwind two-and-a- 0155 1 half months, but there remain additional areas -- a lot 2 of additional areas -- of exploration. 3 Topics the subcommittee may review include the 4 2009 proposal I alluded to earlier regarding selective 5 ratings disclosure and ratings shopping; a broad review 6 of recommendations made in 2012 surrounding alternative 7 payment models; headline comments surrounding CLOs, the 8 state of that asset class, and implications of rule 17G- 9 5, and potential conflicts of interest. 10 We are also looking to put together a panel of 11 issuers that may be willing to speak freely about the 12 pros and cons of the current model and perceived or real 13 conflicts of interest. We hope to bring that panel to 14 the next FIMSAC committee -- FIMSAC meeting to inform 15 the entire FIMSAC on this very important topic. 16 With that I will conclude my remarks. 17 MR. HEANEY: Thank you, Amy. 18 MS. MCGARRITY: Thank you. 19 MR. HEANEY: And obviously, a lot of heavy 20 lifting by this subcommittee, many topics to -- that have 21 been discussed, and a lot more in the pipeline. So we 22 appreciate you taking the time to chair it, and your 23 leadership around it. 24 Let me open it up to questions and comments. 25 (No response.) 0156 1 MR. HEANEY: Anyone on the phone with questions 2 or comments? 3 (No response.) 4 MR. HEANEY: Okay, let's move forward to 5 Ananth. We're going to hear from the ETF and Bond Fund 6 Subcommittee, which is a report discussing the latest ETF 7 and bond funds and stress markets. This committee did an 8 enormous amount of work to get this report ready for us 9 at the FIMSAC meeting, so I just want to call out and 10 thank them: Ananth, Kumar, Lynn, Amy, and Rachel. A 11 large body of work, it got several revisions along the 12 way, but it's a tremendous effort and a tremendous 13 statement about the dedication of this subcommittee 14 amongst all subcommittees on putting forth what we 15 consider to be a really top-notch product. 16 So, Ananth, let me turn it over to you to take 17 us through the report. 18 MR. MADHAVAN: Yes. Thank you very much, 19 Michael. And indeed, thanks to all my fellow committee 20 members who did, you know, way above and beyond their 21 share in getting this report out. 22 So at the previous FIMSAC meeting I talked a 23 little bit about some of the work that our subcommittee 24 had done in terms of ETP classifications, in terms of 25 investor education. This is really the third piece of 0157 1 our mandate, which is to digest a lot of what we've heard 2 about the behavior of ETFs and bond funds in times of 3 market stress or high market volatility. And we heard 4 from a variety of different panelists, including some 5 regulators, some academics, and other folks from around 6 the industry. 7 And the purpose for our report is really to 8 synthesize what we had heard over the past few months, 9 and I think the detailed reports of who actually 10 presented it are on the SEC website, so I'll leave that 11 for anybody that's curious as to whom we heard from. 12 As I said, the report is intended to be a bit 13 of a synthesis. It is our views, and our views alone. 14 So I think I just want to make clear that, you know, the 15 subcommittee was unified in terms of the content of this 16 report and providing this report as a digest, really, for 17 a very complex topic. 18 The second point I want to make is that the 19 report is not meant to be an overwhelmingly comprehensive 20 document. It is meant to delve into the topics we were 21 largely tasked with, and where we had heard from 22 panelists. It's not meant to be a broad statement about 23 systemic risk in general, or the like. 24 And specifically, in terms of the report, which 25 is which is now part of the FIMSAC proceedings, I'll give 0158 1 you a very high-level tour of what we were trying to 2 delve into in the report. 3 So we began with an overview of the structure 4 of mutual funds and how ETFs might be different from 5 mutual funds, in terms of -- their particularly the way 6 these funds are redeemed, since that was one of the 7 aspects we were tasked with, is understanding how flows 8 and redemptions may affect market quality, particularly 9 in times of high volatility. 10 From that baseline we went on to talk a little 11 bit more in detail about what we actually knew. So our 12 report is -- tends to be fairly event-based, so we tried 13 to be as factual as possible, looking first at mutual 14 fund flows then turning to ETFs and some of the 15 implications for ETF behavior in stress markets. 16 We also then, you know, tried to lay in some 17 real-world examples to provide some context as to what we 18 had found. 19 And then lastly, we ended with a roadmap for 20 future academic research. Again, it's not meant to be 21 comprehensive, but these are the areas which we felt were 22 important, and where our knowledge or our empirical fact- 23 finding still showed that there were some gaps in 24 understanding. 25 So I'll leave it to the members to read the 0159 1 report, and we'd welcome any suggestions or feedback. 2 Thank you very much. 3 MR. HEANEY: Any questions or comments for 4 Ananth or the subcommittee? 5 (No response.) 6 MR. HEANEY: Okay. Last, but not least, we'll 7 turn it over to Mihir for a corporate bond transparency 8 update and some of the topics that they've been 9 considering recently. 10 MR. WORAH: Sure. So quick update on what the 11 corporate bond transparency subcommittee has been working 12 on and thinking about since our last meeting. 13 Start with the broad end with a very narrow 14 topic that we're focused on right now. So broadly, we've 15 been focused on the experience of the retail investor. 16 And by retail investor in the corporate bond market we 17 mean the investors who are buying corporate bonds 18 directly for their accounts, either self-directed or 19 through an investment advisor, a wealth management 20 service. Individual investors can buy ETFs and mutual 21 funds, but that's not something -- you know, the block 22 trade pilot that we have addresses some of the issues 23 around that. But what we've been focusing on for the 24 last couple of months are retail individual investors 25 buying directly for their accounts. 0160 1 The first -- you know, the first topic that 2 comes up, and is pretty obvious, is something that's been 3 talked about a lot in this room, which is the issue 4 around markups and disclosure around markups. It came up 5 this morning again. Question is should there be any pre- 6 trade transparency, post-trade transparency. 7 Clearly, some of the regulations that went into 8 effect just last May address some of the post-trade 9 transparency issue to the effect that, you know, 10 individual investors and the brokerage statements, to the 11 extent that bonds they're buying or selling have markups 12 or markdowns, get to see those, and they can make a 13 decision. Are the markups and markdowns too high 14 relative to the service they're getting? And they can 15 make informed decisions around the value that they're 16 getting from the advisors. 17 The question that came up again was is there a 18 way to put this in a more centralized way. Maybe in 19 TRACE. So some of the issues that, for example, we were 20 trying to grapple with this morning about pennying, et 21 cetera, there's more data around the actual price paid to 22 study impacts on the corporate bond market about actual 23 prices paid versus commissions versus riskless trades. 24 That's a topic we, you know, we spent some time thinking 25 about. We've tabled it. We should come back to it, but 0161 1 no firm recommendation from us yet. 2 Moving from the broad to the narrow, we spent 3 the last month or so on a pretty narrow but important 4 part of the corporate bond market. That's what's known 5 as InterNotes. It's a small part of the bond market. 6 There's about 22 billion outstanding corporate-issued 7 InterNotes compared to 5 trillion of the better-known 8 institutionally-issued corporate bonds, for lack of a 9 better word. So 22 billion versus 5 trillion 10 outstanding. There's about a billion-and-a-half of new 11 issuance every year, compared to about $1 trillion of 12 regular corporate bond issuance every year. 13 But the key thing is all of these are primarily 14 owned by individual investors. The key difference in 15 these InterNotes is corporations are issuing them 16 regularly, on a weekly basis, and they're underwritten, 17 they have -- they filed with the SEC, but they're placed 18 directly with individual investors. 19 There are certain characteristics of these 20 bonds that individual investors like. Typically, they 21 shoot at par, so they're simpler to -- you know, simply 22 to understand and put into your -- understand what yield 23 you're getting. They typically have callable features, 24 so -- which gives you some yield enhancement. And then 25 there's other features in these that make them attractive 0162 1 to individual investors, as well as to issuers. 2 We went through presentations and education by 3 a number of participants in the market. Incapital is the 4 biggest underwriter of these issues. Bank of America, 5 Merrill Lynch, and Morgan Stanley trade these. And then 6 Scott Krohn at Verizon, who is on our subcommittee, 7 helped us with the education part of it, as Verizon is 8 one of the larger issuers in this space. 9 So the concerns were, yes, there's a need -- it 10 serves a need in the market, both on the investor side 11 and the issuer side. But is there enough liquidity, is 12 there enough transparency, is there enough education on 13 the -- at the end user level. So that's a topic where 14 we're deep into right now. We hope, as we continue to 15 study it we hope to have a recommendation for the next 16 meeting. But you'll hear more from us on that, probably 17 at the next meeting. So that's about what we've been 18 working on. 19 MR. HEANEY: Thank you, Mihir. 20 I think this meeting today really represents 21 all the hard work that's been going on at the 22 subcommittee, so I do want to thank Mihir for his 23 leadership -- Rick, Amy, Lynn, and Ananth, and all of the 24 people for their active participation on these 25 subcommittees. It's really been some great work, and 0163 1 more in the pipeline. 2 We'll now turn to our final panel of the day to 3 discuss the LIBOR transition and implications on the 4 corporate bond market and municipal securities markets. 5 The transition from LIBOR is a broad-ranging 6 topic with global implications that has received in-depth 7 commentary from both market participants, as well as much 8 attention from the press. 9 Thank you to our panelists for coming down to 10 join us today with what I'm sure will be a great 11 discussion, an active discussion. I'll turn over to 12 Lizzie Baird, deputy director of the SEC Division of 13 Trading and Markets, who will moderate today's panel. 14 MS. BAIRD: Thank you. So the banks currently 15 reporting the information used to set LIBOR will stop 16 doing so after 2021, when their commitment to reporting 17 ends. 18 The Federal Reserve estimates that, in cash and 19 derivatives markets, there are approximately 200 trillion 20 in notional transactions referencing U.S. dollar LIBOR, 21 and more than 35 trillion will not mature by the end of 22 2021. There's approximately 1.8 trillion of outstanding 23 floating corporate and municipal bonds tied to LIBOR. 24 Since 2014. the Federal Reserve and other 25 federal financial regulators, including the SEC, have 0164 1 convened a diverse group of private market participants 2 to identify a risk-free alternative reference rate to 3 U.S. dollar LIBOR, and create and implement a plan for 4 its adoption. This is called the Alternative Reference 5 Rates Committee, or the ARRC, which, just last Friday, 6 designated our panelist, Tom Wipf, as its new chair. 7 Our final panel today will focus on the 8 transition away from U.S. dollar LIBOR to SOFR, the 9 Secured Overnight Financing Rate, and what the market 10 participants should be considering long before 2021. 11 Tom Wipf, you are a vice chair of institutional 12 securities at Morgan Stanley and, as I mentioned, on 13 April 12th was designated chair of the ARRC. You've been 14 involved with the ARRC since its inception. And if you 15 wouldn't mind, could you get us started by giving us a 16 brief overview of the committee, including what work has 17 been done to date in preparing for the transition away 18 from LIBOR? 19 MR. WIPF: Yes. Thank you very much. I can 20 just give a quick update on sort of where we've been 21 since 2014 and run through that. 22 I think, as many of you know, the ARRC began 23 with basically 15 SIFI banks coming together under the 24 auspices of the Federal Reserve Board of Governors to 25 solve several problems: one, which was pick a 0165 1 replacement rate for LIBOR; the second was to determine 2 how to best actually exit from LIBOR and enter into the 3 new rate; and at that point to sort of lay out how the 4 transition would occur, what would the rate be, and when 5 would that begin. That took place between -- and those 6 three questions took about three years to answer, and the 7 answers were that SOFR would be the rate, and the reason 8 that that was chosen is, as has been highlighted, is that 9 the underlying cash transactions that support LIBOR 10 really have really stopped taking place. 11 So the simple explanation of sort of how we got 12 here is post the financial crisis, interbank lending 13 pretty much ground to a halt. The 19 banks who report 14 their -- who report into the LIBOR panel really had no 15 observable transactions to look at before, so they would 16 basically use what they called "expert judgment." So 17 they would be picking these points on the curve, based on 18 other things that would happen, whether it be commercial 19 paper, euro deposits, but by no means had they exited the 20 litigation risks that they faced by actually not seeing 21 transactions with having to set these rates based on 22 estimates. 23 That put the banks in a situation where they 24 probably did not want to continue to do that, and which 25 meant that LIBOR itself was was at risk. And at that 0166 1 point in 2014 there was no alternative, nor was there any 2 language that existed in a cash product or any derivative 3 that effectively dealt with the permanent cessation of 4 LIBOR without significant disputes and challenges in the 5 market. That was what they were faced with. 6 So just on short numbers, back to the 200 7 trillion, if you think about 200 trillion assets and 8 liabilities, even at the most liquid point of the 9 interbank cash market, less than a half-a-billion traded 10 on any given day. So if you think about that, this 11 concept of the inverse pyramid is a small number of 12 transactions are driving a massive number of assets and 13 liabilities, and that reference rate, obviously, is not 14 durable. 15 So the determination was made that the highest 16 order of business for ARRC -- what we now call ARRC 1.0 - 17 - was to make a determination on what would be the most 18 durable rate. So SOFR, or the Secured Overnight 19 Financing Rate, which did not exist at the point that we 20 were embarking on this work, there was information that 21 led us to believe that overnight treasury repo, which is 22 reported by both the clearing banks and the information 23 is available to the Fed, administered by the Fed, has 24 about -- anywhere between, at that point, 750 billion and 25 currently a trillion in daily activity. 0167 1 So that was, far and away, the choice that we 2 made. Because if it even went to other rates that were 3 considered, like the overnight bank financing rate, 4 that's typically about 150 billion, Fed funds less than 5 100 billion. And if you -- even if you go down to 6 treasury bills or other things like that, you get well 7 under 100 billion. So the idea was if observable 8 transactions was the criteria, SOFR was the obvious 9 choice. 10 That left the market with two big questions. 11 LIBOR has a term component a credit component, and so far 12 has neither. So there's a few things that are taking 13 place now, we think, to bridge all that. 14 But we spoke -- several of us here spoke at a 15 panel at the FSB last week, and it's very clear that 2019 16 is the year for action. And this is the year we think 17 that several developments are going to take place that 18 will serve to accelerate this process through either 19 changes in market behavior, or just the realization of, 20 among market participants, that 2021 is not the day 21 people should be running a large book of LIBOR exposure 22 to that cliff. 23 So what we anticipate happening certainly this 24 year, ARRC 2.0, which was the second version -- so 25 following the original ARRC to the decision on the PACE 0168 1 transition plan, which was determined at the end of 2 decision to use SOFR, the Fed reconstituted ARRC, and the 3 ARRC's membership is now comprised of a wide variety of 4 private market participants, including banks, asset 5 managers, insurers, industry trade organizations, and the 6 official sector ex officio members, all the U.S. 7 regulators. 8 ARRC 2.0's mission is to ensure the successful 9 implementation of the PACE transition plan, to coordinate 10 and track planning across products as market participants 11 using USD LIBOR consider transitioning where appropriate 12 to the new rates, addressing risks and legacy contracts 13 like derivatives and cash products, and basically deal 14 with the robust fallbacks. Because, you know, I think 15 when we think about the rolldown of these books, it seems 16 like we have a lot that rolls down before that. But 17 unless people in the market begin using the new rates, 18 that rolldown will just continue to extend. 19 So when we think about the two kind of key 20 areas, certainly in the derivatives market, which 21 represents nearly 90 percent of these assets and 22 liabilities, that, for the most part, can actually be 23 repaired through a series of protocols. So ISDA is 24 working on protocols that will allow derivatives 25 participants to sign up and agree on a spread that would 0169 1 bridge us from SOFR to LIBOR. So if you had a derivative 2 trade on at the end of 2021, you could basically agree 3 with your counterparty that I'll agree that the credit 4 spread should be X, the term spreads should be Y, and you 5 will, to some degree, be able to sort of soften the blow 6 going through that. 7 The challenges in the cash markets are very 8 different. And I know that some of my colleagues here 9 will speak about that. But the -- and the short story 10 is, for the most part, floating-rate notes convert to 11 fixed at the point that LIBOR stops being printed. So 12 that's -- the cash markets, although certainly on a 13 overall scale, a smaller percentage of the issue, 14 actually face probably more challenging, and in some 15 cases intractable, problems in terms of the things that 16 you can do post-issue. 17 So what are we doing this year? In short 18 order, basically the -- we -- in July the ARRC hosted a 19 roundtable at the Fed. At this at this event Vice 20 Chairman Quarles gave an update, again calling out the 21 low number of cash transactions that reference LIBOR, 22 pointing people towards 2021 as sort of the end date of 23 this thing. And then, basically announced that the Fed 24 staff, the New York Fed staff, had begun working on a 25 backward-looking compounded rate that would serve some 0170 1 market participants who just need a term rate. Although 2 it wasn't forward looking, there was a belief that in 3 some parts of the market this could cover some ground. 4 So this was sort of the second leg to the 5 stool, which is we have overnight SOFR, which basically 6 is going to -- it can be used for derivatives and the 7 floating leg of other things like that, and floating rate 8 notes. We've seen about, I guess, over 80 billion in 9 issuance of SOFR notes since the inception of SOFR. SOFR 10 celebrated its first birthday two weeks ago, so there's 11 definitely been some pickup. Futures market are picking 12 up. 13 The second piece, I think, was to ensure that, 14 if there was a need for a term rate -- and certainly 15 everyone believes there is -- the question is could 16 market participants live with a backward-looking 17 compounded rate who are dealing with more operational 18 issues, and then those who need a forward-looking rate 19 would be the third piece of the puzzle. 20 So the announcement that the Fed in the 21 beginning of 2020 will begin again producing, as the 22 administrator, a SAFR, which is a secured average 23 financing rate, which will just be a compounded group of 24 SOFR daily observations, and that will be put forward. 25 So that was our news in last July. 0171 1 Since SOFR began its publication a year ago, 2 many ARRC members and others in the market have worked 3 hard to develop a SOFR futures market, which is moving in 4 the right direction, and a swaps market which I'd say is 5 moving, but certainly significantly slower. 6 There's been a lot of work, and I think for 7 this group in particular, in the cash markets, while 8 we've been sort of in between language that we know 9 doesn't work in a lot of existing cash products and bond 10 indentures and other things, ARRC will be producing very 11 shortly common language that actually allows people who 12 want to continue to use LIBOR to have much better 13 fallbacks when LIBOR stops. Nonetheless, in that interim 14 period, there have also been many, many, many bespoke 15 sets of language. So in every issuer's desire to be in a 16 better position than they are today, we now have many, 17 many kind of varieties of language and documents, which I 18 think presents certain challenges for both issuers and 19 investors. 20 In terms of -- in 2018, in September, the ARRC 21 released four market consultations asking the market on 22 how to develop fallbacks for floating rate notes, 23 syndicated business loans, bilateral business loans, and 24 securitization. Those things are in various degrees of 25 progress, but we've heard from the market and the ARRC 0172 1 will be releasing these things over the next -- fairly 2 shortly on two out of four, particularly the floating- 3 rate notes. 4 In February this year -- and I think this has 5 been very helpful -- the ARRC announced Federal Reserve 6 open office hours, so David Bowman, who is a senior 7 advisor at the Board of Governors, every Friday is 8 basically taking calls on this topic for people in the 9 market who have questions about this. The sessions have 10 been well attended. David's commentary has been very 11 critical in developing a broader market understanding. 12 And looking ahead to the remainder of 2019, I 13 think that the ARRC's key contributions to this work will 14 likely come in ongoing discussions with regulators to 15 clear up remaining uncertainties that could exist. And I 16 would certainly say at the meeting last week at the FSB 17 it was very clear that the official sector and the 18 private sector are now intersecting at a point. And I 19 think that the question that we all have, I think, is who 20 and how quickly can we provide clarity to the remaining 21 open issues. And I think what we've seen is tremendous 22 progress in the UK for good reasons. The rate that they 23 had was basically a reformed version of a rate that 24 existed called SONIA. And there's been really good 25 pickup, and it's been a lot of floating rate note 0173 1 issuance using a new rate. 2 So I think that what we're beginning to find 3 out is that there is going to have to be a little bit of 4 a bigger push in 2019. There are several risk factors 5 that I think are worth highlighting that the ARRC and 6 other groups are focused on. One is that ISDA will be 7 producing the final protocol for derivatives. The final 8 protocol for derivatives will determine when LIBOR 9 ceases, and we have a long-dated derivative that goes 10 beyond there. What are those two calculations we're 11 going to use, the term adjustment and the credit 12 adjustment? 13 Once that's available, we would certainly 14 assume, based on what we've seen, just on the original 15 consultation that ISDA did last year, is the market will 16 most likely think about converging to the point that that 17 protocol takes you to. So if the protocol says SOFR plus 18 X plus Y, pretty certain that the market will probably 19 try to find its way there over time, and begin to 20 converge around that. So those things are, you know, 21 certainly impactful. 22 The second thing is as the ARRC releases its 23 new language for floating rate notes, or common language 24 that people can use in the market, we could have a 25 situation where an issuer has old language in the market 0174 1 and it begins to use new language, and you would 2 certainly think that people would begin to think that 3 there's some price distinction between those two if you 4 have one that has, you know, terrible fallback language 5 and one that has significantly better fallback language. 6 And then the third thing is, obviously, for 7 derivatives at the central counterparties, or CCPs are -- 8 will be making adjustments to their calculations, will 9 again -- which should embed some risk in the market to 10 derivatives participants which should begin to flow 11 through and create again pricing dynamics that we haven't 12 seen. 13 So 2019, you know, definitely feels like we're 14 deep in the second act of a three-act play. There's a 15 lot that can be done right now. And when we think about, 16 you know, where we are, the ARRC also did send -- back in 17 -- last year -- basically, a letter really requesting 18 Title 7 regulatory relief around other things, and I 19 think, as we deal with this, what we are hearing from 20 large market participants is, one, is that clarity around 21 tax, which I think we've got some visibility to now, and 22 clarity around accounting, and clarity around margin 23 clearing and other regulatory requirements. 24 So I think in 2019, as I said, I think that the 25 -- certainly there's been an intersection between the 0175 1 official sector and the private sector, which certainly 2 should drive us to the point. And I think that, at the 3 best case outcome that we think about, is that people 4 will begin to address their inventories, deal with the 5 issues, and take the smallest possible amount of LIBOR 6 exposure into the end of 2021. 7 I'll stop there, and I'll pass it over. 8 MS. BAIRD: Thank you. Tom Deas, chairman of 9 the National Association of Corporate Treasurers, and 10 also an ARRC participant, could you please tell us what 11 the key concerns are facing corporate issuers with 12 respect to this transition, and what they should be 13 looking at with respect to LIBOR and SOFR? 14 MR. DEAS: Yes, Lizzie. Thank you very much. 15 So I've been at this for a while. I've been a chief 16 financial officer and treasurer of large publicly-listed 17 multinational companies for over 35 years. And I was 18 drafted into the original effort that preceded ARRC 1.0, 19 where the Financial Stability Board created this multi- 20 national group called the Market Participants Group to 21 look at what should -- you know, look at some of the 22 abuses that happened during the financial crisis, 23 involving rate setting and LIBOR, and to look at what 24 might come from that. 25 And one of the things that we did in our work 0176 1 was to survey corporate treasurers around the world in 2 order to identify just where they're using different 3 interbank offered rates within the company. And 4 everybody in this room well understands that LIBOR, in 5 this $200 trillion of instruments that we recounted, is 6 certainly used in floating rate notes, and it's used in 7 corporate credit agreements, it's used in multinational 8 credit agreements, and those corporate treasurers are 9 well up on and know where it is in the documents. 10 But these rates are marbled throughout our 11 corporate structure. And just to give you a feeling for 12 where they are, we use intercompany credit agreements to 13 manage liquidity within our group. So these are 14 agreements among 100 percent-owned related parties. We 15 have agreements with suppliers and customers in the form 16 of asset purchase and sale agreements, long-term supply 17 agreements, strategic capital goods purchases, and even 18 some employee benefit payment obligations, all of which 19 incorporate LIBOR to adjust for payment or timing 20 differences in the day-to-day administration of these 21 agreements. 22 And I'll also confess to you that sometimes the 23 purchasing people didn't call up corporate treasury when 24 they've got a new supply agreement. They just went to 25 the last one and they pasted in that late payments get 0177 1 figured out on LIBOR basis. So one of the important 2 things that corporate treasurers are doing right now is 3 to take inventory of where all these agreements were. My 4 colleague, Tom, mentioned that ISDA is working on a 5 protocol that -- for derivative counterparties. We can 6 all sign up to the protocol. 7 But under U.S. contract law, for all these 8 agreements that I just cited, the parties need to agree 9 to any amendments that are made to these agreements. And 10 as an example, the intercompany funding agreements I 11 mentioned -- let's say we've got a loan to 1 of our 30 12 international subsidiaries, and that subsidiary is in 13 Malaysia, and we priced it LIBOR plus 300 basis points. 14 And now, when we switch to SOFR, because SOFR is based on 15 a risk-free rate, the spreads going to be wider. So 16 we've got to go to the Malaysian Central Bank, the tax 17 authorities, and others and say, "I want to make this 18 loan that's LIBOR plus 300 into SOFR plus 350 basis point 19 loans." 20 And, of course, one of the things that we and 21 multinational corporations have been accused of is using 22 intercompany payments to manage the tax liabilities of 23 our worldwide group and move funds from high-tax 24 jurisdictions to low-tax jurisdictions. So we're going 25 to be involved in a negotiation and an explanation to the 0178 1 authorities in Malaysia about why this kind of a change 2 is needed. 3 And as Lizzie pointed out, the clock is ticking 4 and New Year's Day 2022 is going to be fast upon us. 5 When you consider what happened when we converted from 6 the individual currencies that made up -- that make up 7 what's now the euro, we had a 10-year-plus trial run, 8 where we created parallel trading, and we knew that on 9 launch day, which was January 1st, 2002, how to calculate 10 what the translation would be from French francs into 11 euros, and that was all agreed to well in advance. We 12 don't have that translation now. 13 We know that, right now, for many investment 14 grade companies, BBB companies were borrowing at LIBOR 15 plus 100. Well, we're going to be borrowing at SOFR plus 16 something else, and that something else, that other 17 spread, isn't now determined by any trading that's going 18 on. It's going to have to be first arrived at, and then 19 ultimately, through arbitrage transactions, derivative 20 transactions, various cash transactions, it'll have to be 21 determined and we'll have to translate our spread. 22 Let me just give you one example. For 23 instance, consider a corporate securitization of customer 24 receivables, where we take that into a special -- take 25 the receivables into a special-purpose entity and use 0179 1 those to secure a floating rate note. Well, in the 2 typical floating rate note indenture, the provision is 3 that if LIBOR ever ceases to be quoted, then the rate on 4 the floating rate note resets until maturity at the last 5 rate that was set. And so, if you were an investor and 6 bought this floating rate note and LIBOR went away before 7 the provisions were changed in that document, then you 8 just bought a fixed-rate note. And under almost all 9 indentures that I know of, changing the interest rate on 10 a debt instrument requires a 100 percent lender consent. 11 So all these note holders out there would have 12 to agree to that consent, and we're not going to get 13 that. And so that's what Tom talked about concerning the 14 rolldown, where these things are going to mature. 15 Fortunately, these credit card or other customer 16 receivables, secured floating rate notes, a lot of them 17 are issued for three years or five years, and we're going 18 to be able to deal with them. But we've got a lot of 19 work to do. 20 And what we require help from our colleagues in 21 the official sector at the SEC and the other regulatory 22 agencies is to give us certainty as to the accounting tax 23 and regulatory treatment of these conversions so that we 24 can know that now, and make these changes. Everybody is 25 aware that if you change the interest rate on a debt 0180 1 instrument by more than a de minimus amount, that that 2 change is treated for federal income tax purposes, as a 3 termination of the old loan and a deemed issuance of a 4 new loan, with the recognition of a gain or loss for tax 5 purposes, if that's what happens. And so, if a corporate 6 treasurer took the initiative and amended his deals from 7 LIBOR to SOFR, and that's what he got hit with, that 8 would be a very bad day for that corporate treasurer. 9 So we're talking to the Treasury Department, 10 we're talking to the CFTC on derivatives, and we're 11 talking to financial accounting standards boards and 12 other regulators to get their help, and that's what we'll 13 get, and we'll make this conversion. Thank you. 14 MS. BAIRD: Thank you. Pat McCoy is director 15 of finance for the New York Metropolitan Transportation 16 Authority. 17 Pat, the MTA recently issued municipal 18 securities that were tied to SOFR, as opposed to LIBOR. 19 The MTA is one of the few municipalities to do that. Can 20 you tell us how you arrived at that decision, and 21 speculate as to why other municipalities haven't embraced 22 SOFR yet? 23 MR. MCCOY: Sure, I'd be happy to. Thank you 24 for having me here. I'm really here representing the 25 Government Finance Officers Association in the United 0181 1 States and Canada's immediate past president. But I'm 2 happy to talk about our experience with our SOFR 3 transaction. 4 We, the MTA, has a large portfolio of 5 outstanding debt, approximately 40 billion. We like to 6 have exposure to natural variable rates, which we do. 7 Approximately five percent of the portfolio is in 8 variable rate. And then we have a similarly-sized 9 portion of our portfolio that is synthetic fixed rate, 10 which is -- obviously, we use the FRN product, together 11 with a derivative, to achieve a synthetic fixed rate, 12 which -- many of these transactions extend beyond 2021. 13 We knew this problem was looming, and is 14 looming. And last September we -- after evaluating the 15 transaction that was coming up for renewal, we opted to 16 go with a SOFR-based transaction, principally because we 17 know we're going to be there eventually. And we'd like 18 to get some experience under our belt with this new 19 index. 20 I've had the benefit of sitting on the ARRC as 21 a representative of GFOA, so I had the benefit of 22 participating in these discussions and learning about how 23 SOFR was formulated and how the markets were coming 24 together. 25 I think, as you have -- have all heard here, 0182 1 the different market participants are coming together to 2 really make this transition as painless as possible. We 3 know that it will be disruptive, and I think one of the 4 keys to avoiding that disruption is early planning and 5 early adoption to the degree that an entity can do that. 6 7 So, with the demise of the auction rate market 8 some years ago, the floating rate note became a very 9 popular product for municipal issuers. And this year, in 10 fact, I've got about eight transactions that I'll need to 11 re-market. Those transactions are a combination of 12 SIFMA-based, LIBOR-based, and now, of course, SOFR. So, 13 as we approach any of these transactions, we evaluate 14 what our alternatives are. We work with our advisor, our 15 municipal advisor, to come up with the best plan of 16 action, and evaluate those alternatives, obviously, to 17 the benchmark of what it -- what will the fixed-rate 18 markets deliver, as well. 19 For the municipal issuer community, nothing 20 could be more important than clarity around this issue 21 and having market participants work together to avoid 22 significant disruptions. Our members -- currently over 23 20,000 GFOA members around the country -- represent the 24 vast infrastructure that we rely on in this country. We 25 estimate that about 75 percent of core infrastructure in 0183 1 the United States has been built with the tax-exempt 2 bond. That's a great instrument. It's been a wonderful 3 way for this country to develop. 4 However, we do like access to the taxable 5 markets where there is a great depth and breadth of 6 market participation. And so, while it's not appropriate 7 for every municipal issuer to be issuing into a taxable 8 market as a percentage of, for an issuer like the MTA -- 9 we're one of the largest in the country -- it is 10 appropriate. And so, for the larger issuers out there, 11 we want to make sure that we're there. We're thankful to 12 Tom and David Bowman at the ARRC for giving us a seat at 13 the table and for allowing us to represent issuers there. 14 So I would be happy to take any questions 15 later. Thank you. 16 MS. BAIRD: Ed Fitzpatrick is a portfolio 17 manager in JP Morgan Asset Management. 18 Ed, as a fixed income portfolio manager, what 19 do you see as your challenges with the transition from 20 LIBOR to SOFR? 21 MR. FITZPATRICK: Thank you. Good afternoon 22 and thank you for having me here. 23 I think it's important to frame this transition 24 process in a framework that centers around three stages, 25 because each one of these stages has its own unique 0184 1 challenges. The first stage is addressing the legacy 2 security -- 3 MR. REDFEARN: Can you just go a little bit 4 closer to the mic, please? Thank you. 5 MR. FITZPATRICK: Is this better? 6 MR. REDFEARN: Better, yes. 7 MR. FITZPATRICK: So the three stages, the 8 first stage being the legacy securities that are the 9 former LIBOR regime; the second stage being the migration 10 or transition phase, which is a regime that has 11 potentially two or a dual reference rate regime; and then 12 the third stage being one where we're in the future 13 reference rate regime, which is principally a SOFR regime 14 in the U.S. 15 I know some of my other panelists will be 16 discussing the transition process. All focus mainly on 17 this future SOFR regime, but I will make two very brief 18 comments about the overall process. 19 I think the first, when investors are 20 considering this process, one of the concerns is about 21 investor awareness and market preparedness. And I think 22 Tom mentioned many of the steps that the ARRC has done to 23 inform the investment community, and communication and 24 education is a key principle in ensuring that this 25 transition continues, and that at each stage that 0185 1 education needs to to happen. And I think, so far, it's 2 been a fairly successful operation. 3 The second point I would make is just the 4 importance of this, this migration or transition phase in 5 the overall transition process. And that's where we are 6 relying on fallback language. The timeliness of this 7 process is paramount, because we believe that once the 8 industry is able to provide a credible recommendation for 9 fallback language, as well as credit spread adjustment 10 and trigger events, that will have a positive network 11 effect that will allow for a more robust SOFR market to 12 develop. 13 Now, focusing on the SOFR-only regime that 14 we're -- that this entire market is hoping to transition 15 to, I think there's three principal concerns that face 16 cash investors right now, and it's mostly centered around 17 convention. 18 I think the first one is about term reference 19 rates. The market has become very accustomed -- the 20 floating rate market has become very accustomed to 21 knowing what -- the interest they will accrue over the 22 subsequent coupon period payment date. This has two 23 benefits. One, it allows investors to manage their cash. 24 And second, it allows them to handle non-standard 25 settlement. 0186 1 So I think one of the challenges that we face 2 right now is that this term reference market, there's a 3 feasibility and timeliness associated with it. Right now 4 a specific owner or publisher of a term rate hasn't yet 5 been identified. I think this is one of the areas where 6 perhaps the New York Fed would likely be the right place. 7 But that still needs to be determined. 8 In addition to identifying the owner of that 9 published rate, the market would still need sufficient 10 time to evaluate and analyze how this term rate 11 translates into market pricing. Absent a term reference 12 rate, there is some uncertainty as to how comfortable and 13 how quickly the market will be able to transition to what 14 is dubbed interest accrual in arrears, or backward- 15 looking interest accrual. 16 I think one of the challenges, in addition to 17 the market adapting to it, is the fact that there's still 18 some uncertainties around the methodology, particularly 19 between the cash markets and the derivative markets. And 20 while these are subtle differences around calculations, 21 there is some concern that the investor community may 22 become disillusioned with the efficacy of hedges if there 23 was to be a different methodology going forward. 24 And then the final point I would make on the 25 eventual SOFR regime is the need for a more robust SOFR 0187 1 swaps market to develop. And the way this translates 2 through to the credit markets is that over the last 5 to 3 10 years there has been an increase in the amount of 4 variable rate notes that have been issued in the market. 5 They have a floating component to it, but part of the 6 structure of these variable rate notes also have a call 7 feature to it. And these call features tend to be struck 8 off of a swap rate. 9 And so you could run into a situation where the 10 floating rate component of these variable rate notes is - 11 - has a has a SOFR reference, while the actual call 12 feature that it struck off of is a rate that's other than 13 the SOFR curve. And it just adds to a little bit more 14 uncertainty as we move forward. 15 I would say that I think the market and -- the 16 SOFR market swaps market is developing, as Tom mentioned, 17 in a very positive manner, but it's still early days. 18 And I think this is, again, one of the areas where having 19 the transition phase be finalized and the market have a 20 clear picture of how both LIBOR and SOFR, as reference 21 rates, are going to evolve would improve the market depth 22 of the SOFR market. 23 The last point I would make before passing it 24 on is that the future reference rate regime isn't 25 isolated just to the United States, and the buy-side 0188 1 representatives up here are all international investors. 2 And right now, around five of the largest developed 3 markets are undergoing their own transition from LIBOR to 4 a different reference rate. And this is going to pose 5 challenges, because each of these various jurisdictions 6 are at different levels in the process. And I think it 7 would be a tremendous benefit for the U.S. to be able to 8 transition over to a SOFR-only regime and lead this 9 global transition process. 10 MS. BAIRD: Julian Potenza is a fixed-income 11 portfolio manager at Fidelity Investments. 12 Julian, SOFR experienced a small hiccup on 13 December 31st, when it jumped from 2.46 percent the prior 14 day to 3 percent, at that time the highest level since 15 the Federal Reserve began publishing the benchmark last 16 April. The move tracked a spike in the repo rate. 17 As a fixed-income portfolio manager, are you at 18 all concerned that SOFR could be unusually volatile and 19 hard to predict? 20 MR. POTENZA: Thanks. Thanks, Lizzie. And 21 thanks to the committee for giving us the opportunity to 22 come share some of our perspectives on the transition. 23 The potential volatility in SOFR is something 24 that we've thought and discussed a fair amount at 25 Fidelity. You know, I think a number of the fellow 0189 1 panelists have discussed how the market has become 2 accustomed to a forward-looking term rate, as far as a 3 reference rate for floating rate instruments. And I 4 think one of the natural advantages of a forward-looking 5 term rate is that it smooths out some of the volatility 6 inherent in an overnight structure. 7 And, you know, we've done some work and others 8 have done some work. If you if look at sort of the long 9 history of something like LIBOR, which has an overnight 10 out to, you know, a one-year term structure, and then a 11 swaps market that's developed off it, or something like 12 the federal funds rate -- obviously not as much of a 13 reference rate for instruments traded in the market, but 14 an important rate that market participants follow -- 15 overnight volatility of the types that we've seen in the 16 SOFR market, for example, at year end did not appear to 17 be unduly unique. 18 And some of the work on different approaches to 19 sort of smoothing out SOFR and translating it from an 20 overnight rate into a term rate show that whatever 21 methodology you choose, the volatility of a three-month 22 SOFR actually looks pretty comparable to, you know, to 23 some of the other reference rates that have a pretty 24 strong track record in the market. 25 So, you know, and then, as an investor, I guess 0190 1 you weigh that against when you're owning an instrument 2 referenced to something like SOFR -- and we have 3 purchased some of these in our portfolios -- and you get 4 a pop like that at year end, that is actually a little 5 benefit that accrues to your investor. 6 So, broadly speaking, at this juncture it 7 hasn't been -- at least my view -- that some of the repo 8 market dynamics that we're seeing at year-end, quarter- 9 end should be a significant impediment to the adoption of 10 SOFR. 11 Maybe I'll make just a few sort of more general 12 comments on sort of how we're approaching the transition, 13 and some of the issues that we're thinking of on behalf 14 of our fund shareholders. 15 And I think -- and one of these issues has 16 already come up, but we are very focused on the issue of 17 fallback language in existing LIBOR-based instruments, 18 particularly those instruments that mature past 2021, for 19 obvious reasons. This includes the new issue market and 20 the issue of documentation and the legacy securities in 21 the portfolio. And when I think about, you know, this 22 idea of LIBOR transition exposure, those two concepts are 23 linked. If you have a portfolio that owns existing SOFR 24 -- I'm sorry LIBOR-linked securities with various 25 fallback language, that's like a stock of exposure that 0191 1 you already have to this transition. But if you're 2 continuing to purchase new-issue LIBOR-linked securities, 3 that's a flow that's adding to that stock. And so, 4 clearly, focusing on both aspects of fallback are 5 crucial. 6 On the new issue front, our preference is for 7 the development of clear, consistent, and investor- 8 friendly fallback language across markets. And, our 9 perspective, some of the important components of investor 10 friendly fallback language include clear LIBOR succession 11 triggers, a minimal issuer or agent discretion, and 12 crucially value-neutral spread adjustments. So the idea 13 that, you know, the LIBOR transition process should not 14 be creating winners and losers between investors and 15 issuers. 16 And we're actually encouraged by the work on 17 fallback language that has been undertaken by the ARRC. 18 We're looking forward to the publication of the final 19 recommend language expected soon -- I think I heard very 20 soon from Tom. And we are we are hopeful that the 21 publication of that language will give the buy side a 22 common set of principles to rally around as we negotiate 23 for terms that are friendly to investors and in LIBOR- 24 linked issuance with issuers and underwriters. 25 On the legacy security front, our current 0192 1 understanding of the situation -- for some reasons that 2 have come up on the panel -- is that it is unlikely that 3 there is any sort of regulatorily-driven big bang 4 solution to the wide range of fallback language that 5 exists across the corporate and structured products 6 language. As investors, it would be preferable if there 7 were. I think it would reduce the risk of volatility 8 through the transition. But just given the nature of the 9 problem, it doesn't appear that that is likely to be 10 forthcoming. 11 And so our view is that it's incumbent upon 12 investors to assess and understand the exposure of LIBOR 13 fallback risk in existing funds. This is a relatively 14 painstaking process. It revolves -- involves parsing 15 through thousands of documents, but it is important and 16 crucial work to understand and manage exposure to the 17 risk. 18 As has been alluded to, you know, there 19 certainly is the possibility that the market will start 20 to apply valuation discounts to securities with inferior 21 fallback language. And, given that markets are, by their 22 nature, forward looking, that's a process that could 23 start to play out well before the end of 2021. So that's 24 something that we are actively involved in. 25 On top of the issue of fallback language in 0193 1 LIBOR-linked securities, we are also actively engaged in 2 the development of the SOFR market, having purchased 3 SOFR-linked securities in both bond and money market 4 portfolios. And as we discuss SOFR market structures 5 with issuers, we've been focused, really, on two high- 6 level principles. 7 The first is that it's important to structure 8 SOFR-linked transactions in a way that keeps the 9 economics of the instrument as close as possible with 10 current market conditions. And so I'm talking about 11 features like lockouts and look-backs, which can 12 introduce sometimes slight but still potentially 13 impactful lags between market rates and what happens with 14 the coupon of your security. And that has the potential 15 to create basis risk between instruments, depending upon 16 if monetary policy decisions and other things happen in 17 significant days. If you're in a lockout period, that 18 could have a meaningful implication for the valuation of 19 that security. 20 I would note that a lot of these features are 21 introduced as a way of dealing with some of the 22 operational aspects of using an overnight rate as a 23 reference rate. So that is further motivation for the 24 development of some type of a term SOFR, which I'll touch 25 on in a moment. 0194 1 The second high-level point on the development 2 of the SOFR market is that it is crucial that the market 3 trends towards consistent structural features, both over 4 time and across issuers. What we'd like to avoid is the 5 situation where many, on their face, very similar 6 securities actually have a range of unique structural 7 features, having to wade through that, slows our traders 8 down, and has the potential to impede market liquidity. 9 And then finally, we believe that the 10 development of a reliable, forward-looking term SOFR is 11 an important step in an orderly transition for the cash 12 market. It does appear that, for many investors and 13 issuers, an overnight reference rate would likely be 14 workable, operationally. A forward-looking term rate is 15 a more appropriate reference rate for most floating rate 16 instruments, as the term rate smooths out some of the 17 volatility inherent in an overnight rate, as we discussed 18 initially in response to the question about year-end. A 19 forward-looking term rate is also more straightforward, 20 operationally, as the frequency of interest rate resets 21 is reduced, the needs for lockouts and look-backs is 22 minimized, and base rates are known in advance. 23 The final point, the development of a full SOFR 24 term structure may help the transition of other asset 25 classes that currently price relative to the LIBOR-based 0195 1 swaps curve. So I'm talking mostly about ABS and CMBS, 2 asset-backed securities and commercial mortgage-backed 3 securities. While these markets may successfully 4 transition to pricing off the treasury curve, as has 5 largely happened in the agency CMBS market, the 6 development of a SOFR swaps curve would provide another 7 alternative pricing benchmark. 8 Thanks again for the opportunity to share some 9 of our perspectives. I'm happy to expand on any of these 10 points during the open discussion. 11 MS. BAIRD: Thank you very much. 12 David Knutson is head of credit research for 13 Americas at Schroder's Investment Management. 14 David, I believe that you've been involved with 15 the credit roundtable's LIBOR Alternative Working Group. 16 Could you give us a summary of the roundtable's proposal 17 to create standard LIBOR fallback language for the 18 corporate credit market? 19 MR. KNUTSON: Sure. Thank you very much. I'm 20 the co-leader of the credit roundtable. And just over a 21 year ago, the credit roundtable published an open letter 22 regarding SOFR, or what -- at that time we were 23 considering what would be a good benchmark rate. And so 24 we talked about a lot of the things that slowly the ARRC 25 has kind of come to address in terms of finding 0196 1 consistent structural features, as was just mentioned, 2 the credit spread adjustment that we've talked about, the 3 forward-looking -- all things seemed to come together, 4 not only in the cash market, but also in the derivatives 5 markets. 6 And just last week the credit roundtable held 7 its semiannual meeting here in Washington. We had a 8 chance to have a session with David Bowman. We did a 9 survey -- anecdotal, obviously -- but of the folks that 10 were -- or the firms that were gathered there. 11 And what we found was that approximately a 12 third, 30 to 40 percent, of the firms had bought a SOFR 13 note. I'm sure some of the positions were more, you 14 know, maybe experimental or temporary as opposed to a 15 large core position, but firms are coming into the 16 market, experimenting with the market, getting used to 17 the new rates, seeing how it reacts, seeing how it fits 18 in with the other risk exposures they have in their 19 portfolios. 20 Another survey question talked about the need 21 for hedge accounting, and I think that the SOFR 22 particularly fits the desire to find a good risk-free 23 rate that allows firms to hedge. 24 We also talked about hard triggers being an 25 important part of the change. I know that was just 0197 1 mentioned, that -- clear consistent, triggers so we know 2 when the formal change will occur. 3 We talked briefly about zombie LIBOR, which is 4 essentially LIBOR kind of continuing in kind of some 5 zombie-like state, and how we would address that has come 6 up a lot on the ARRC calls. 7 The -- it was just recently mentioned, but 8 essentially we talked a lot about minimum value transfer, 9 and that the transition between LIBOR to SOFR shouldn't 10 be an event that enriches one party at the expense of the 11 other party. So finding that perfect way to put it into 12 a contract is obviously the devil in the detail. But the 13 principle of minimum value transfer, we think, is an 14 essential principle for a successful transition. 15 One other topic came up, and that is how you 16 can adjust or modify the existing indentures. And as the 17 -- several panelists already mentioned, adjusting a 18 payment term in a contract requires 100 percent of the 19 investors to agree to it. And that's virtually 20 impossible -- most FRNs, or floating rate notes. And so 21 there was some discussion about perhaps creating a legal 22 or regulatory safe harbor to get in the states of New 23 York and Delaware that would eliminate the confusion and 24 uncertainty and disputes that would occur, especially if 25 there was an abrupt cessation of LIBOR. 0198 1 Those are some of the things that we talked 2 about. The other items that the credit roundtable in 3 particular has focused on is the desire to understand as 4 soon as possible, and so that's why we are very happy 5 that the ARRC recommendations will come out very soon. 6 We hope that'll be a point that the market can come 7 together on. We ultimately think the cash markets and 8 the derivatives markets will sync up, and there won't be 9 any kind of daylight between the two the two markets. 10 MS. BAIRD: Thank you. Does the committee have 11 any questions? 12 Nothing? Larry? 13 MR. HARRIS: The legacy fallback language 14 problem for existing contracts that go past 2021 seems 15 essentially intractable. A hundred percent, just not 16 going to happen. 17 The only two alternatives, seem to me, would be 18 to get some legal coverage which, I suppose, could be 19 grounded by the states. The other possibility, which I'm 20 sure you've explored but I'm curious to hear more about 21 it, is the question of essentially redefining LIBOR to 22 reflect the alternative. And so I don't know if that's 23 what you referred to by the zombie LIBOR, or something 24 like that, but we know that LIBOR has been -- its 25 collection method has been changed in the past. And so 0199 1 perhaps it's simply a matter of changing it once again, 2 but somewhat more radically this time. 3 MR. WIPF: So on the top, I think this concept 4 of the zombie LIBOR really is more along the lines where 5 the current administrator IBA would continue to publish 6 after perhaps the regulator, being the FCA, would declare 7 it not representative anymore. Right? And some 8 documents -- and that's this idea of kind of, you know, 9 synchronous triggers. Some documents would say as long 10 as LIBOR continues to be published you have to keep 11 referencing that, whereas some, there are triggers that 12 would say the minute that that first -- the determination 13 is made by the regulator, that you can actually begin to 14 trigger your fallback. So that's a little bit of a 15 different issue. 16 In terms of the original definition and sort of 17 the changes, I think that was -- when this work began way 18 back when, I think the original idea was can we change 19 the inputs. And I think, with the definitions of LIBOR, 20 this concept -- and I will defer to everyone else here -- 21 of contract frustration would mean that that just can't 22 be done. 23 And that's what led us down the road to having 24 actually select an alternative to get off LIBOR and get 25 on the new ones. But there was very significant legal 0200 1 issues around the idea of changing the inputs to LIBOR, 2 whereas the calculation and collection methods are still 3 collecting in the same way, but doing a bit of a 4 different calculation. But the definition itself made 5 that impossible, and certainly that would have been a 6 much easier way around this. 7 MR. HARRIS: Did the contracts refer to how 8 LIBOR is collected, or -- 9 MR. WIPF: Very specific on what LIBOR is, and 10 the definitions are very, very tight. And we -- and 11 there -- that was certainly -- I mean the idea that, you 12 know, to add inputs and to create observable transactions 13 was certainly the first port of call. But, you know, 14 across the entire legal community that was deemed 15 impossible. 16 MR. HARRIS: Okay, so one other question. It 17 seems clear to me, then, that it's almost essential that 18 there be legislation to handle this issue. I just -- 19 there's no other way around it. The question is, is it 20 sufficient to do it at the state level, or can it be. 21 Does it require national legislation? 22 MR. WIPF: Well, the reason why I mentioned 23 Delaware and New York is because a lot of these 24 securities, particularly the ones that the corporate 25 credit market looks at, and obviously there's other 0201 1 representatives for other markets here, but -- are issued 2 under New York or Delaware. And so changing those two 3 issuance platforms, I guess you could say, or creating an 4 environment where they can go out into the market and 5 establish a precedent -- I think there are others, 6 obviously, but I think that would go a long way towards 7 getting the market moving in a particular direction and 8 helping people focus on a particular solution. 9 MS. BAIRD: Sonali? 10 MS. THEISEN: Thank you all for your comments. 11 That was very enlightening. 12 I had a couple of questions around -- you know, 13 some of the same topics continue come up, but obviously 14 on value transfer there is, you know, the basis of the 15 credit component and the term structure to -- in order to 16 minimize those, at minimum, to two components. 17 So I guess my question would be how do you 18 actually envision in practice the market transitioning? 19 Is it going to be, like, a phase-in? Because, obviously, 20 those two components will float over time. Like, how do 21 you, in practice, see the market transitioning? 22 And then there was also the comment about 23 timeliness and about the market, like, opting in. So how 24 do you see this sort of phase-in working, while 25 mitigating, again, the arbitrage, if you will, of these, 0202 1 you know, two components and about -- and minimizing the 2 value transfer? 3 MR. KNUTSON: I'll start out. So I imagine 4 over the next 12 to 18 months you're going to see more 5 and more issuers issue into the SOFR market. And as an 6 investor I can then begin to trade, or I can assess the 7 risks relative to one unique issue, or an issue that's 8 very comparable to others, and see that SOFR difference. 9 You can look at it right now with maybe three- 10 month LIBOR and SOFR, and you can average that out to 11 maybe three eighths of a point over some time. But I 12 think, as the derivatives markets develop, and as the 13 term markets develop, I think -- and the swap markets 14 develop, which is absolutely fundamental here, you'll be 15 able to see the difference evolve. 16 And I think the markets are efficient enough to 17 find that price, which makes me indifferent whether or 18 not I have a SOFR or a LIBOR note. Obviously, as I 19 pointed out earlier the longer you hold these LIBOR 20 notes, the risk actually is increasing every day. So it 21 isn't really perceptible now and people continue to issue 22 into that market. 23 But whether it's a counterparty or an 24 issuer or, you know, a bank, if they're not making that 25 adjustment the market will begin to perceive a higher 0203 1 risk, I think. 2 MR. WIPF: We really -- yeah. And I think 3 that, in terms of the idea that there will be industry 4 fallbacks, there will be fallbacks for new cash product 5 for those who choose to continue to use LIBOR to some 6 degree, but -- and there will be at least an 7 understanding of what the fallbacks are that will be 8 significantly better than today. And derivatives, there 9 will be protocols that will allow us to agree, and those 10 things will be measured. There is a consultation that is 11 -- has -- ISDA's been doing a series of consultations, 12 and they will come back and then release sort of what 13 that is, what the period -- what the look-back period is, 14 the look-forward periods are, whatever the outcome is of 15 that, there will be determined industry-standard 16 protocols in place. 17 Nonetheless, you know, I think that we all 18 shoot for a minimal value transfer. But by no means 19 would any -- does any fallback really appropriately deal 20 with the fact that when we get to 2021 these things will 21 move around. There is nothing that gets us to zero here. 22 23 So, you know, the real approach, when we think 24 about it, is how do we how do we stop digging, right? 25 How do we begin using the new rates as soon as possible? 0204 1 How do we allow that fallback to be our friend in this 2 work, and take it forward? Because the idea that we're 3 all just going to, you know, just sort of bootstrap 4 everything with a series of protocols and ride this thing 5 into the end of 2021 and hope for the best -- the more 6 we've seen, I'd have to say that, you know, this has 7 really been a real journey in terms of education and 8 outreach and understanding. 9 But the piece that we've really gotten to was - 10 - last year, I would say, specifically -- everyone has 11 looked at what the fallbacks might look like. And 12 although there's no desire to create winners and losers, 13 at the end of it they really are seat belts, right? They 14 really are seat belts that will help us navigate this as 15 best we can. But by no means would we, you know, be 16 terribly reliant on a large book of business that relied 17 on those fallbacks. They put us in a significantly 18 better position, and they're the first step in risk 19 management both for issuers and investors, derivative 20 market participants, and others. 21 So they're the first step in sort of putting us 22 in a significantly better position than we're in today, 23 but by no means do we think that that's sort of the 24 answer to all the issues. With those, we suspect that 25 what will come out from ISDA through those market 0205 1 consultations, what's going to come out from the ARRC 2 through those market consultations will be reflective 3 and, most importantly, provide clarity, whatever they 4 are. 5 MR. FITZPATRICK: I would just add, from an 6 investor's perspective, it feels like the foundations are 7 there for a robust SOFR market to develop. And once we 8 have a formalized fallback language in place, we see that 9 as the evolution -- allowing the evolution of liquidity 10 to start moving from the LIBOR market toward the SOFR 11 market. And so we think that defining fallback language 12 puts a standardization within the market that allows for 13 then a smooth transition over. 14 MR. POTENZA: Sorry, maybe just one thing to 15 add from another investor perspective. I think, you 16 know, I would agree with the points that Tom made. As we 17 think about the fallback nature of value neutral, there 18 are a number of approaches that are currently being 19 debated. And ultimately the industry will coalesce 20 around one that's sort of a best-efforts estimate at that 21 time. 22 I think the reality of the situation is it's 23 just going to take a longer period to truly understand 24 the economics of investing in a SOFR-linked instrument 25 through market cycles, relative to a LIBOR-based 0206 1 instrument, secured overnight rate versus unsecured term 2 rate. 3 How is it going to behave in an easing cycle? 4 How is it going to behave in a financial crisis? 5 Fundamentally different than LIBOR, potentially. As an 6 investor, that's a lot. And so, to truly know, right, 7 what the right adjustment is, what the fair adjustment is 8 as an investor is something that we're going to have to 9 build an understanding of over time. 10 As of right now, if you're thinking about 11 investing in a SOFR-linked -- you know, a SOFR-linked 12 instrument, sort of kind of best practice is to look at 13 where an issuer could bring a similarly-structured LIBOR- 14 based transaction. Take a -- you know, they're not 15 particularly liquid, but take a look at the LIBOR/SOFR 16 sort of basis swap markets and see if you're getting any 17 concession. And I always say, kind of echoing the 18 earlier point, that doesn't tell you that that's the 19 right sort of economic decision over the long term. It 20 kind of tells you on that day you can't be arbitraged 21 against. You know, you're -- it's the best deal you can 22 find that day for your investors. It's going to take, 23 you know, a number of market cycles to really fully 24 answer the question about what the right spread to SOFR 25 is. 0207 1 MS. THEISEN: Thank you. That was very 2 helpful. 3 If I could ask one more question, Tom Deas, you 4 mentioned in your comments that, you know, certainty 5 around, you know, accounting, tax regulatory treatment is 6 obviously going to be very important. At what point will 7 there be enough information of how this is unfolding to 8 provide that certainty? And what do you think that the - 9 - again, the time period, if you will, will take after 10 there is that certainty to actually -- enough 11 information, rather, to then receive that certainty? 12 Because, obviously, all these pieces fit together. 13 MR. DEAS: Well, thank you. The tension 14 between issuers and our regulators is that we want this 15 transition to happen as soon as possible, and to 16 encourage everybody to convert. And the regulators -- 17 just if I could make this generalization -- don't want to 18 give a blank check to the market and give too broad a 19 dispensation, for fear that some people will take 20 advantage of that, and arbitrage the market in that way. 21 And so, in the meeting that Tom referred to 22 that several of us attended last week sponsored by the 23 Financial Stability Board, we were assured by the U.S. 24 regulators in the room that they were going to -- they 25 recognized the need to act quickly. I mean, as of today, 0208 1 that -- there is no provision that provides tax deferral 2 for conversion and change in interest rate, or that 3 anticipates that that would happen during a fallback when 4 -- if a fallback provision is executed, and give 5 dispensation then. 6 There is no provision under derivative 7 regulation that deals with the link between an interest 8 rate swap and. say, one of these floating rate notes that 9 we talked about. 10 Let's say that the protocol happens on 11 the interest rate swap side and we convert the swap that 12 we put on the floating rate note, the swap it fixed. And 13 that happens because the protocol happened and the 14 train's leaving the station, but the floating rate note 15 didn't change because it would require 100 percent note- 16 holder consent, and we didn't get that. 17 For accounting and even as an end user for 18 regulatory reasons, Dodd-Frank and the CFTC's regulations 19 provide that if -- for us to avail ourselves of the end- 20 user exemption for mandatory margining and clearing, the 21 hedge has to be an effective hedge and linked to the 22 commercial underlying risk. Is that still the case if 23 the underlying risk is LIBOR-based and the hedge is SOFR- 24 based, because we changed in the protocol? 25 We don't have any of those answers. And the 0209 1 first step in this is to convert these legacy contracts. 2 So we need the answers to do that. And the next step is 3 to start to issue new instrument securities derivatives 4 off the new index. 5 So the example I gave about the conversion to 6 the euro -- go back and look at the history. That was a 7 long trial run, and we don't have that trial run 8 available to us between now and year-end 2021. 9 MS. BAIRD: Horace, did you have a question? 10 MR. CARTER: No, I just -- a point of interest. 11 When you envision a term structure, is that -- would I 12 be right in assuming that that's essentially going to be 13 just an implied rate base on a series of overnight 14 forwards? 15 MR. WIPF: I think the -- you know, the way 16 that it's been sort of laid out is that, you know, we 17 have SOFR, we have an overnight rate. We would 18 anticipate that sort of by the beginning of next year 19 we'll have the Fed-administered backward-looking 20 compounded term rates that potentially have utility for 21 some market participants. 22 The piece of the puzzle really -- it becomes 23 sequential, really, with people in the market actually 24 developing the futures and the swaps markets and the 25 other things that will create enough observable 0210 1 transactions that the construction of that rate would be, 2 again, based on observable transactions, so we're not 3 just creating the same problem we have, right? 4 This whole -- this initiative is driven by the 5 fact that, you know, when LIBOR began in 1986 there -- 6 people would just take all their trades, put them on a 7 spreadsheet, and send them in, and those were the points. 8 And we evolved to a point where 19 banks -- or for the 9 most part -- you know, estimating what they think those 10 curves are. So to move to a place -- the definition 11 around us is really where the observable transaction is. 12 13 So it really kind of puts the third piece of 14 this heavily onto people in the market to begin using the 15 new rates for new production, creating the natural 16 hedging demand that's going to -- that will take place, 17 and giving people the opportunity to look at, you know, 18 real sizeable, measurable transactions that could, in 19 fact, create a forward-looking curve that would meet the 20 principles set forth, which is observable transactions, 21 you know, are the -- are sort of the -- you know, the 22 first box we need to check. 23 So, you know, the first two: you know, one 24 here already, one on the way. And the question is, as 25 market participants begin to look at their own exposures 0211 1 and their own types of business, we've heard, you know, 2 people, you know, can use SOFR. They may find utility 3 for SAFR, as they take it forward. And when we clear all 4 that, how much ground that covers will mean that the next 5 piece of the puzzle is very, very dependent on industry 6 take-up of SOFR on a go-forward basis for new production, 7 for new products, that will at least give us some -- at 8 least visibility into some transactions that potentially 9 could construct that forward-looking term curve. 10 MS. BAIRD: Larry, do you have a question? 11 MR. HARRIS: Yeah, I have a couple of very 12 quick ones. 13 The first one is -- so SOFR has been published 14 for the last year, roughly, right? Has it been 15 constructed back -- can it be constructed back so that it 16 can be analyzed in relation to LIBOR over the next -- 17 last 20, 30 years? 18 MR. WIPF: So the New York Fed, at the point 19 they released SOFR, went back to a series of data points. 20 So if we go back to 2012, there were collections of the 21 component pieces of SOFR that were taking place. And if 22 you can go all the way back, if we think about just 23 treasury repo as the -- as sort of the -- as the main -- 24 obviously, the main data point, you start with SOFR, 25 which has cleared activity from the DTCC, which has 0212 1 information from the two clearing banks. You go back to 2 -- the cleared activity at DTCC was actually available, 3 the GCF repo was available in 2000 -- I think 2007. 4 So if you go back, there are pieces of the 5 puzzle that are there that the Fed has put out. And then 6 you go all the way back, it goes back to what were called 7 the primary dealer calls that happened every morning 8 where they see where you're funding your treasuries. 9 So there's pieces of information and, you know, 10 the -- and people should determine the quality of that 11 data as they build their models out. But SOFR really is 12 the April 20, and then you can look back to points and 13 there's plenty of information on the New York Fed's 14 website on how they've gotten there. And it sort of 15 works its way all the way through. 16 MR. HARRIS: All right. So another legacy 17 question. I'm sure you've thought of all this stuff, and 18 I'm sort of curious. It's so challenging, obviously. 19 So if LIBOR has to be based on a particular 20 methodology, what are the possibility of encouraging the 21 banks to do interbank loans that are based on SOFR? And 22 then they'll just report what they're doing. But it's -- 23 but those loans, we know, are indexed to SOFR. You 24 thought of this already. 25 MR. WIPF: I think -- you know, I think when -- 0213 1 the problem really -- the problem that we -- that we're, 2 I think, trying to unpack is the problem is that the 3 banks currently don't do interbank lending, right? But 4 the repo market is, in fact, constructed of not just 5 interbank lending, but actual activity with clients and 6 short-term cash investors and non-dealers. 7 So the construction of SOFR, the overnight repo 8 market, really accomplishes that. It's just called SOFR, 9 because it's overnight, and the LIBOR because -- again, 10 back to the definitional question, the LIBOR definition 11 is very specific on what that is. 12 So those are, I think, things that we've -- we 13 look -- I mean certainly, you know, creating a brand-new 14 rate certainly was not the first port of call in this 15 work, and I think a lot of these things have been -- had 16 been investigated and continue to point us to the fact 17 that we -- our baseline risk management is that LIBOR 18 will be extinct or irrelevant by the end of 2021, and 19 that we need to be thinking about, A, what we should be 20 using today to reduce that risk as we head into 2021; 21 and, B, what are the things that we need to be looking 22 at, in terms of our positions that we know go beyond 23 2021, in terms of fallbacks and protocols and other 24 things we need to do to risk management. 25 MR. HARRIS: And the last question: We're a 0214 1 advisory committee to the SEC; what advice would you 2 suggest we provide to the SEC on this issue? 3 CHAIR CLAYTON: I'll jump in and say that I was 4 about to ask the same question, Larry, which is -- 5 (Laughter.) 6 CHAIR CLAYTON: And let me just frame it a 7 little. We've raised awareness -- at least I hope. And 8 I appreciate your being here to further raise awareness. 9 Identified some of the issues, Tom, I -- thanks, thank 10 you for the outline of issues that, you know, the market 11 can't solve on its own. That's kind of a list. 12 But do we -- do you feel like we have a handle 13 on the size of these issues, that -- let me put this way 14 -- that the notional number of outstanding instruments 15 that there is fallback language, the ones for which 16 there's not, is that an area where you feel like ARRC, 17 FSB -- can we help? Where can we help bring sunlight, 18 which is the thing we're best at? 19 MR. DEAS: If I could take a stab at that, one 20 of the things that I've discussed with some of your 21 colleagues -- and we have an SEC member who is -- you 22 know, attends our ARRC meetings, and is a member of the 23 ARRC. 24 And having written risk factors for prospectuses 25 10Ks, 10Qs for a number of years, I can tell you that 0215 1 there is some language, at least for non-financial 2 corporations, to describe the kinds of instruments that I 3 just referred to, and say what they feel the risks are, 4 and what their plans are for negotiating those risks 5 away. 6 And so I have an example here that I'll be 7 happy -- that I've shared with one of your colleagues, 8 and I'll be happy to pass around of what, you know, a 9 non-financial corporation might write. But it involves 10 that disclosure that we face these risks as an issuer, as 11 a filer, and we have -- we've identified what the risks 12 are, and we have a plan to address those risks. 13 But if LIBOR were to cease being quoted 14 tomorrow it could mean that some of these contracts for 15 long-term supply, or with customers, or, you know, other 16 contracts that we have are in dispute. And if those 17 disputes aren't settled, it could mean an interruption of 18 those supplies to us. 19 And so that's a commercial risk we face that 20 needs to be identified. And in -- we're saying, I think, 21 you know, underlying all this is the principle that 22 sunshine is the best medicine. And if we know what these 23 issues are, and we see what other people are doing in 24 their plan to address them, it will be a long way toward 25 fixing this problem. 0216 1 MR. WIPF: Yeah, and we continue -- you know, I 2 think we've -- certainly one of the big pieces of the 3 puzzle for ARRC is really outreach and how we think 4 about, you know -- and how we get this information out 5 there, right, and how we get everyone in the market who 6 may not even be aware that they have some LIBOR in their 7 book are actually looking at, looking at those risks, 8 looking at the maturities, you know. 9 And again, this concept of how we stop digging, 10 right, making sure that that new issue that's going into 11 the market -- that people are, you know, endeavoring to 12 use, ideally, the common language that will come out from 13 ARRC and utilize as best we can to, at this point, to 14 sort of reduce that -- you know, the fact that there's a 15 lot of bespoke language coming out now, and once there's 16 common industry language on new issue -- which, again, 17 can be the most intractable part of this -- that that's 18 important, that --ensuring that investors are looking at 19 their exposures that go beyond 2021, and thinking about 20 how to navigate that. 21 And really, when we think about, you know, 22 really, the cash markets, that it really comes down to 23 know what you have today, what your risk is, assuming 24 that LIBOR is gone -- and you should do that risk 25 management exercise at a lot of different points -- 0217 1 ensuring that issuers have an awareness that legacy 2 language doesn't work well in the cessation of LIBOR, and 3 beginning to, you know, in some ways use the tools that 4 are there. 5 I think when this work all began the idea was, 6 well, we can get some protocols, we can fix some things. 7 But there's just a series of tools that are available 8 this year that weren't last year. And potentially, you 9 know, people are using SOFR already, and there are, you 10 know, as we've heard today, finding that, you know, there 11 are investors for this product. 12 So I think, as we move through time, it's just 13 taking that outreach to the next level with the 14 information that we now know, and with the tools that 15 have now been sort of laid out over the last several 16 years, and asking people in the market to -- you know, to 17 take a harder look at their current exposures, the tools 18 available, and to really, in any way that they possibly 19 can -- we've certainly found that we -- there are uses 20 for SOFR that don't necessarily have to involve 30-year 21 swaps. 22 There are uses for, you know, internal transfer 23 pricing of cash. There are uses in repos and securities 24 loans where people use reference rates that have gone 25 from Fed funds open to Fed funds effective to OBFR, 0218 1 right? There is -- the changing of those rates has never 2 really presented a lot. 3 But the more familiarity we create with the new 4 rates, the more people have a greater understanding, can 5 do the backward-looking analysis that they need to do. 6 But, you know, any way in any form where participants can 7 stop digging and then, to the extent they can do that, 8 conduct their inventories -- but again, getting this 9 outreach and this awareness into parts of the market who 10 probably don't wake up going, "I think I have a problem 11 with LIBOR." 12 And I think that's a lot of ground to cover, 13 and we think that, you know, anything that -- you know, 14 the ARRC's -- one of the ARRC's challenges continues to 15 be outreach, awareness, education, and asking people to 16 look hard in their organizations on what risks they have 17 today. 18 CHAIR CLAYTON: Okay, well, let me let me say 19 this. I think we will -- we're going to continue to do 20 what we can, in terms of pushing disclosure and pushing 21 people -- financial companies to look at what their 22 exposure is, and make the market aware. 23 I don't mean to jump ahead of the committee, 24 but if you saw fit it wouldn't be bad to schedule these 25 guys to come back in six months and give us an update on 0219 1 whether, you know, we've been able to term out the market 2 and ARRC's work. But I'll just make that as a 3 suggestion. 4 MS. BAIRD: Lynn? 5 MS. MARTIN: Hi. Yeah, you guys have talked a 6 lot about the institutional market, and I guess a 7 question more to more to Tom is have you considered the 8 impact of the migration from LIBOR to SOFR on things like 9 the mortgage market and the student loan market, things 10 that actually really hit the -- that retail, not even 11 investor, the retail borrower in the market -- and what 12 does that transition really look like? Because, 13 obviously, there's quite a bit of paper tied to LIBOR in 14 those markets 15 MR. WIPF: Okay. 16 (Laughter.) 17 MR. WIPF: Where -- I mean I think the 18 interesting part is, as you go through all the asset 19 classes, is that in the -- in commercial mortgages and 20 other things of that nature, that the lenders actually 21 have a lot of flexibility to change those rates, right? 22 Now, with that comes the, you know, the concern about 23 conduct and franchise rates. So those are the real 24 concerns on that, is that the ability to change those 25 rates is fairly -- lays pretty much squarely with the 0220 1 lender. 2 But how you do that, and how you go through the 3 -- you know, again, the education, the understanding, you 4 know, explaining SOFR folks in the market who -- you 5 know, who may not be as familiar with what LIBOR is, 6 right, these are the things that have to happen as we get 7 deeper into the work. So there's definitely a work 8 stream at the ARRC, and I'm sure across many, many other 9 organizations on how to get that awareness up. 10 So it doesn't face the same challenges 11 contractually, but it does face a lot of different 12 challenges, obviously, in moving that rate, understanding 13 the impact, and making sure that, you know, that those 14 clients have a great understanding. Obviously, across 15 our firm, between investment management, the wealth 16 management, everything in the middle, these are the 17 questions we're asking ourselves, and I think it goes 18 sort of -- it begins at what's in the realm of the 19 possible, what can you do. And then when might you do 20 it, and what do you need to do before you do that. A 21 different type of risk management exercise, but 22 significantly as if not, in many cases, more important. 23 MR. KNUTSON: The only thing I would add is 24 that I think the principle minimum value transfer -- 25 because the student, the family that are making a 0221 1 mortgage payment or a student loan payment or whatever -- 2 car, or whatever -- if there is a significant jump either 3 way, it is going to raise questions. And, obviously, if 4 it goes against the borrower I think it raises lots of 5 issues, so that the point of minimum value transfer 6 finding something that, when it rolls into SOFR, it 7 doesn't result in a surprise when you open up their 8 monthly statement. 9 MS. BAIRD: Are there any committee members on 10 the phone who have a question for the panel? 11 (No response.) 12 MS. BAIRD: Okay. Thank you very much, panel. 13 Very interesting. 14 MR. HEANEY: Thank you, Lizzie. Thank you to 15 the panel, again, for making the trip to Washington, D.C. 16 Plenty for us to think about. And, as Chairman Clayton 17 said, we will see you again in six months. So thank you. 18 (Laughter.) 19 MR. REDFEARN: Thank you all. 20 MR. HEANEY: So we are at the end of our agenda 21 today. Again, another full day of FIMSAC. I just want 22 to thank all the FIMSAC members for your participation, 23 all the work that has led up to the -- to this meeting, 24 and the subcommittee work. We, obviously, are looking at 25 a wide variety of issues, as demonstrated today. 0222 1 And again, if there's other topics that this 2 committee feels like should be raised or discussed, let's 3 continue to talk about that. And please, let's continue 4 to bring them forward. 5 I do look forward to the work that gets done in 6 the subcommittee level between now and our next meeting 7 of FIMSAC, which is July 29th. 8 So at this point, if there's any -- no further 9 questions or comments -- 10 (Comment off microphone.) 11 CHAIR CLAYTON: Why don't we get the group's 12 view at the next meeting? But you heard my view. I'd 13 love to continue to benefit from your work. 14 MR. HEANEY: You know what, Mr. Chairman? We 15 will come back with perhaps a collective view, and I 16 think we can -- David and Tom and Ben, we can get the 17 collective view of FIMSAC, and we'll come back and report 18 on that in July, if that's okay. 19 COMMISSIONER PEIRCE: I'd like to add my voice 20 in support of continuing the work. We've really 21 benefitted a lot from your work. 22 MR. HEANEY: Thank you, Commissioner. 23 At this point I will entertain a motion to adjourn. 24 PARTICIPANT: So moved. 25 MR. HEANEY: All in favor? 0223 1 (Chorus of ayes.) 2 MR. HEANEY: Thank you very much for your 3 engagement. Safe travels, and we'll see you in July. 4 (Whereupon, at 3:26 p.m., the meeting was 5 adjourned.) 6 * * * * * 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0224 1 PROOFREADER'S CERTIFICATE 2 3 In the Matter of: SEC FIXED INCOME MARKET STRUCTURE 4 ADVISORY COMMITTEE 5 File Number: OS-0415 6 Date: Monday, April 15, 2019 7 Location: Washington, D.C. 8 9 This is to certify that I, Christine Boyce 10 (the undersigned), do hereby certify that the foregoing 11 transcript is a complete, true and accurate transcription 12 of all matters contained on the recorded proceedings of 13 the investigative testimony. 14 15 _______________________ _______________________ 16 Proofreader's Name) (Date) 17 18 19 20 21 22 23 24 25 0225 1 REPORTER'S CERTIFICATE 2 3 I, Kevin Carr, reporter, hereby certify that the 4 foregoing transcript is a complete, true and accurate 5 transcript of the matter indicated, held on 6 __4/15/2019___________, at New York, New York in the 7 matter of: 8 SEC FIXED INCOME MARKET STRUCTURE ADVISORY COMMITTEE. 9 I further certify that this proceeding was recorded by 10 me, and that the foregoing transcript has been prepared 11 under my direction. 12 13 14 Date: 4/15/2019 15 Official Reporter: Kevin Carr 16 17 18 19 20 21 22 23 24 25