0001 1 U.S. SECURITIES AND EXCHANGE COMMISSION 2 3 4 5 6 MEETING OF THE FIXED INCOME 7 MARKET STRUCTURE ADVISORY COMMITTEE 8 9 AMENDED 11-8-2018 10 11 12 13 Monday, October 29, 2018 14 9:30 a.m. 15 16 17 18 19 20 21 22 23 U.S. Securities and Exchange Commission 24 100 F Street, N.E. 25 Washington, D.C. 0002 1 PARTICIPANTS: 2 Michael Heaney, Committee Chair 3 Jay Clayton, Commission Chair 4 Elad Roisman, Commissioner 5 Robert Jackson, Jr., Commissioner 6 Kara Stein, Commissioner 7 Dan Allen 8 Matt Andresen 9 John Bagley 10 Carole Brown 11 Horace Carter 12 Gilbert Garcia 13 Tom Gira 14 Larry Harris 15 Jessica Kane 16 Amy Edwards 17 Rebecca Olsen 18 Brett Redfearn 19 Lizzie Baird 20 David Shillman 21 John Roesner 22 Craig Parmelee 23 Adam Richmond 24 Amar Kuchinad 25 Anath Madhavan 0003 1 PARTICIPANTS (Continued): 2 Lynn Martin 3 Amy Mcgarrity 4 Rick McVey 5 Larry Tabb 6 Sonali Theisen 7 Kumar Venkataraman 8 Elisse Walter 9 Rachel Wilson 10 Brad Winges 11 Mihir Worah 12 Frederic Demesy 13 Spencer Gallagher 14 Bob LoBue 15 Ola Persson 16 Alex Sedgwick 17 Richard McVey 18 John Bagley 19 Simon Wu 20 Eric Beinstein 21 John Bender 22 Daniel Gates 23 Van Hesser 24 Brian Kennedy 25 Tom Murphy 0004 1 PARTICIPANTS (Continued): 2 Suzanne Shank (via teleconference) 3 Scott Krohn (via teleconference) 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0005 1 C O N T E N T S 2 PAGE 3 Welcome and Opening Remarks 6 4 5 Draft Recommendations on ETF Classification 22 6 and Education Data 7 8 Draft Recommendation on Collections and 64 9 Dissemination of Reference Data 10 11 Corporate Bond Transparency Subcommittee Update 66 12 13 Municipal Securities Transparency Subcommittee 70 14 Update and Presentation of the Pre-Trade 15 Transparency Analysis 16 17 Corporate Credit Markets: The Role of Credit 138 18 Ratings in a Higher Leverage World 19 20 Future Topics for Committee Consideration 205 21 22 23 24 25 0006 1 P R O C E E D I N G S 2 CHAIRMAN HEANEY: Good morning. I believe we 3 have quorum, so I would like to call the meeting to 4 order. 5 In addition to the members in the room, we have 6 Suzanne Shank and Scott Krohn joining via teleconference. 7 So, good morning, Suzanne and Scott. 8 MR. KROHN: Good morning. 9 MS. SHANK: Morning. 10 CHAIRMAN HEANEY: Thank you for joining us for 11 the fourth meeting of the SEC Fixed Income Market 12 Structure Advisory Committee. I would especially like to 13 welcome Sonali Theisen and Brad Winges, who both joined 14 the committee in September and have already made large 15 strides and valuable contributions in the subcommittee 16 work to this point. 17 We will miss Tom Thees and Brian Archer. We look 18 forward to continuing to work with Brad and Sonali, and 19 we're glad to have you both has FIMSAC members. So 20 welcome. 21 I will begin by welcoming Chairman Clayton and 22 asking him to make his opening remarks. 23 CHAIRMAN CLAYTON: Thank you, Michael. And 24 right on cue, welcome to Commissioner Roisman to your 25 first FIMSAC meeting. 0007 1 COMMISSIONER ROISMAN: Thank you very much. 2 CHAIRMAN CLAYTON: And I note your enthusiasm 3 for our examination of fixed income markets and some of 4 the topics on today's agenda. 5 Also welcome to Sonali and Brad. And as Michael 6 mentioned, a thank you to Brian and Tom. 7 We have an excellent agenda today. And I am eager 8 to hear updates from our subcommittees, including the ETF 9 and Bond Subcommittee; the Technology and E-Trading 10 Subcommittee; the Municipal Securities Transparency and 11 Corporate Bond Transparency Subcommittees, which I think 12 you've now effectively divided. 13 COMMISSIONER ROISMAN: We have. 14 CHAIRMAN CLAYTON: Excellent, excellent. 15 I understand there will be recommendations 16 today. I want to say that I am delighted to see that we 17 have credit rating agency issues on today's agenda. I am 18 going to come back to that, the vital role they play. 19 But I want to note that you have had a very successful 20 first year, three insightful meetings. I expect today's 21 to be insightful as well. 22 There have been two recommendations. We will 23 have some updates today, and there are more 24 recommendations to come today, productive by any measure. 25 So, I want to thank all of you. Thank you, 0008 1 Michael. I want to thank the panelists today for their 2 participation. And then I want to make a remark on 3 pricing in our marketplace. 4 We have had a lot of information about data 5 pricing in the marketplace. And I recently returned from 6 meetings of the Financial Stability Board and meetings of 7 the International Securities Commissioners at IOSCO. And 8 I just want to comment how the important the work you are 9 doing here is, because the importance of pricing and 10 accurate pricing in our marketplace is hard to overstate. 11 We see it day-to-day in trading, but the number 12 of decisions that are made based on the price signals 13 that our market send is orders of magnitude greater than 14 the trading decisions. 15 And it is important, what this group is doing, to 16 examine whether the inputs to those prices are correct, 17 what we would want them to be. And, in particular, 18 whether the models that we follow, maybe we follow with 19 too much commitment. Just said another way, we have a 20 dynamic economy. And if you are following the same model 21 for a long time in a dynamic economy, it may be outdated. 22 And I just want to thank this group for 23 exploring those types of issues as well. 24 Back to you, Michael. 25 CHAIRMAN HEANEY: Thank you, Chairman Clayton. 0009 1 We will now hear the opening comments from the 2 Commissioners, and we will start with Commissioner Stein. 3 COMMISSIONER STEIN: I am going to make a few 4 quick opening remarks. And I am also going to welcome 5 Commissioner Roisman to his first meeting. And thank 6 Sonali and Brad for joining the committee. And I also 7 agree with Chairman Clayton, you have had a really 8 productive first year. And it has meant a lot, to a lot 9 of us, that you are taking the time out of busy schedules 10 to make recommendations to us. So, thank you again. 11 As always, you have a very busy day ahead of 12 you. And the agenda is covering a wide range of issues 13 that affect our fixed income markets. ETFs, corporate 14 bond reference data, municipal bond pre-trade 15 transparency and credit ratings. And all of these topics 16 are of great importance. And I look forward to hearing 17 your thoughts on all of them. 18 In particular, I am interested in thoughts on 19 ETF classification. What is the right classification? 20 Are there other categories? And if a classification 21 system were adopted, how should it be used to ensure 22 truth in labeling? 23 I am also interested in hearing your ideas on 24 the collection, on the dissemination of reference data 25 for corporate bonds. In particular, I would like to hear 0010 1 your thoughts on how a new issue reference data service 2 for corporate bonds relates to electronic trading. Would 3 it facilitate more fair and efficient trading? And if 4 so, how? 5 So, once again, thank you to everyone for 6 coming to today's meeting. And I look forward to your 7 comments. 8 CHAIRMAN HEANEY: Thank you, Commissioner 9 Stein. Commissioner Jackson? 10 COMMISSIONER JACKSON: Well, thanks so much. I 11 am delighted to be here. I want to join my colleagues in 12 welcoming Elad to his first FIMSAC meeting. So, looking 13 forward to working together on these important issues. 14 And it is apt I think that we are meeting this 15 morning on a day where the Red Sox have become baseball's 16 champion for the fourth time in 15 years, which shows 17 that America has a lot of work to do. 18 I think that the chairman is right, that this 19 group's intense activity over the last year has been 20 extraordinarily impressive, and I want to begin just by 21 congratulating you and noting how important that work is. 22 As he pointed out, this is your fourth meeting. You have 23 come away with two very thoughtful recommendations. And 24 I want to just open the meeting by thanking those who I 25 know I have been especially committed to the mission, 0011 1 starting with you, Michael, and your investment in this 2 work. 3 Tom Eady, David Dimitrious, Ben Berstein, you 4 all have been absolutely extraordinary. 5 We know how hard this work is. And we are very 6 grateful for the time and dedication you have brought to 7 it. 8 Like my colleagues, I am looking forward to 9 today's discussion. I always learn a lot from these 10 conversations. 11 Thank you to all of you who are investing all 12 this time. And I look forward to today's conversation. 13 CHAIRMAN HEARNEY: Thank you, Commissioner 14 Jackson. And finally, Commissioner Roisman, again we 15 would like to welcome you to the first meeting as well 16 following the swearing-in in September and look forward 17 to your opening comments. 18 COMMISSIONER ROISMAN: Thank you very much, 19 Michael. So, thank you first of all, Chairman Clayton, 20 for your work to constitute this committee. And thank 21 you all, the Commissioners, for backing this. And 22 members of the committee for your contributing time and 23 expertise for our meeting today. 24 As I mentioned before, thank you all the staff 25 and the different divisions and Offices of the SEC for 0012 1 your efforts to make this committee an important part our 2 work to improve the fixed income markets. 3 I am pleased to be a part of this fourth and 4 final FIMSAC meeting of 2018 and, as you said, my first 5 as a Commissioner. And I look forward to this group's 6 continuing its important work in 2019. 7 One of the areas in which I plan on focusing 8 during my time in the Commission will be the steps the 9 SEC can take to further improve the fixed income markets 10 for investors and for the overall U.S. capital markets. 11 The Commission oversees several important 12 markets that are interconnected and increasingly global 13 in nature. The U.S. fixed income market is a significant 14 and vital part in our capital markets. One that directly 15 affects U.S. corporations, our cities, states and towns 16 and individual investors. Despite this, the Commission 17 has historically focused on the fixed income market less 18 than it has for example on the equities markets. This 19 has occurred even the U.S. fixed income market outsizes 20 the U.S. equities market and has large direct retail 21 participation, certainly in the muni space. 22 I comment Chairman Clayton for establishing 23 this committee and for giving the fixed income market the 24 attention it warrants. While FIMSAC's role is advisory 25 in nature, it is an incredibly important one, providing 0013 1 the Commission with a form to draw from the members and 2 panelists insights and varying perspectives on fixed 3 income market structure and operational matters. 4 The committee's prior feedback has already 5 included very thoughtful discussions and ideas, including 6 changes to post-trade transparency -- sorry, post-trade 7 reporting for block trades and corporate bonds and the 8 further work that can be done to improve transparency in 9 the corporate and muni markets. 10 I am eager for the Commission to continue its 11 consideration of how we can contribute to making the 12 using fixed income market as favorable to investors as 13 our equity markets. This includes exploring efforts to 14 achieve reliable pre- and post-trade transparency in our 15 fixed income markets, a topic your committee has 16 discussed before and continues to deliberate. 17 I also hope to learn more about your views on 18 best execution and other topics that receive considerable 19 attention in the equities context. Additionally, I am 20 very interested in your new draft recommendations, as 21 Commissioner Stein mentioned, including regarding 22 collection and dissemination of reference data. 23 I look forward to hearing each of your 24 perspectives on these important topics. And I will close 25 by reiterating my appreciation to all of the 0014 1 Commissioners and this Committee's members and staff and 2 my eagerness to work with my fellow Commissioners to 3 consider these topics in a meaningful and balanced way to 4 help retail investors and corporate America as well as 5 our capital markets. 6 Thank you. 7 CHAIRMAN HEANEY: Thank you, Commissioner 8 Roisman. Next, I would like to it over to Brett 9 Redfearn, Director of the Division of Trading and Markets 10 and the committee's designated federal Officer for his 11 opening comments. 12 MR. REDFEARN: Thank you, Michael. And I too 13 would like to welcome everyone to our fourth FIMSAC 14 meeting. I think that this meeting and the commitment of 15 all the participants that you see here today is a 16 demonstration of the increased focus that we have had on 17 the fixed income markets. 18 I would like to briefly introduce my colleague 19 sitting here with me today. First to my right from the 20 Division of Trading and Markets, we have Elizabeth or 21 Lizzie Baird. She is one of our deputy Directors, and it 22 is Lizzie's first participation in this FIMSAC meeting. 23 Lizzie recently joined the division and has been a great 24 asset us to us in fixed income. She was a bond trader 25 for nine years, so I think that experience will prove 0015 1 helpful. 2 To Lizzie's right are Dave Shillman and John 3 Roesner, both associate Directors in the Office of Market 4 Supervision. 5 Down to the left are Rebecca Olsen, Director of 6 the Office of Municipal Securities and Amy Edwards, 7 assistant Director in the Division of Economic and Risk 8 Analysis. 9 Also, I would like to welcome the two new 10 members to the committee, to FIMSAC at the table with us 11 today, Sonali Theisen from Bank of America Merrill Lynch 12 and Brad Winges from Piper Jaffray. So, welcome to you 13 both. Thanks for agreeing to support the important work 14 of this committee. 15 Before we get started, the general disclaimer, 16 the views expressed here by myself and other staff are 17 ours alone and do not necessarily reflect those of the 18 Commission, the Commissioners or other members of the 19 staff. 20 So today the committee will consider three 21 recommendations, two from the ETF and Bond Fund 22 Subcommittee and one from the Technology and Electronic 23 Trading Subcommittee. 24 As you will see from these recommendations, the 25 FIMSAC subcommittees are focused on ensuring that 0016 1 investors and other market participants have the 2 information necessary to productively participate and 3 assess the effectiveness of bond investment options in 4 our bond markets. 5 For example, the ETF and Bond Fund Subcommittee 6 will recommend to the committee a naming convention for 7 exchange-traded products designed to better inform 8 investors of available investment products. This 9 subcommittee is also recommending to the committee the 10 creation of a standardized, widely available database of 11 key information that would help investors compare ETFs. 12 The Technology and Electronic Trading 13 Subcommittee will present a recommendation to the 14 committee designed to provide timely, market-wide access 15 to key new issue reference data that the subcommittee 16 believes will increase the efficiency of pricing and 17 electronic trading of corporate bonds. 18 Your consideration of these recommendations 19 today will conclude an active first year for the 20 committee. You made two recommendations already that 21 have been mentioned. At your April meeting, you 22 recommended conducting a pilot program to assess the 23 market impact of public transparency for block size 24 corporate bond trades. 25 As I have noted before, the SEC staff is in 0017 1 active consultation with FINRA staff concerning the 2 design and implementation of that block pilot. 3 And, as you know, at your last meeting in July, 4 you recommended conducting a review of the regulatory 5 framework for electronic trading platforms used in the 6 corporate bond and municipal securities markets. 7 As you may know, the Commission indicated in a 8 release adopting amendments to Regulation ATS for equity 9 trading platforms that it would review the regulatory 10 framework for fixed income electronic trading platforms. 11 And the staff is currently engaged in that review. 12 As we approach the conclusion of this first 13 year and of your two-year term, today's agenda provides 14 an opportunity for this committee to discuss new topics 15 for potential consideration and to take stock of the work 16 that has been considered thus far. 17 For example, later this afternoon, the 18 committee will receive updates from two transparency 19 subcommittees on their work and what they expect to focus 20 on going forward. And the municipal securities 21 transparency update will include the presentation of a 22 timely analysis of municipal securities pre-trade data by 23 MSRB staff, including the committee's own John Bagley. 24 Additionally, the FIMSAC will hear from a panel 25 of experts on the role of the credit ratings in corporate 0018 1 bond markets. 2 There has been much discussion about higher 3 leverage in the corporate bond market and the 4 creditworthiness of certain issuers in the investment- 5 grade market. Some have suggested an economic downturn 6 can lead to significant credit downgrades. Others 7 believe this concern may be overstated. The FIMSAC's 8 engagement on this timely topic at today's meeting 9 highlights the important role this committee serves. 10 I look forward to the panel discussions and 11 hearing committee members' views on whether there are any 12 issues or concern that merit further committee focus over 13 the next year. 14 Finally, committee members will have the 15 opportunity to express thoughts on other topics that they 16 would like the FIMSAC to consider over the next year 17 beyond the broad subject matters defined in the current 18 subcommittee mandates. This could lead to formation of 19 new subcommittees. 20 As part of this discussion, I would love to get 21 any reaction from members about the current global 22 transition away from LIBOR. As many of you know, the 23 U.S. industry participants and regulators are working on 24 facilitating broader options of the secured overnight 25 financing rate or SOFR as an alternative reference rate 0019 1 to LIBOR. I am interested in hearing your views on the 2 implementation of SOFR and the extent of its impact on 3 the corporate bond and municipal securities markets and 4 whether this topic merits any further consideration by 5 FIMSAC. 6 Like this committee, I am also concluding my 7 first year here at the Commission. Tomorrow is actually 8 my one-year anniversary. And every year, every day here, 9 we pursue policies designed to improve our securities 10 markets in furtherance of the public interest and the 11 protection of investors. This work requires the diligent 12 consideration of many factors and oftentimes the 13 balancing of competing interests. Healthy debate and 14 consideration of competing points of view are necessary 15 for the formulation of good policy. 16 As we pursue our policy agenda, we will choose 17 an issue that advance this agency's mission to protect 18 investors, maintain fair, orderly and efficient markets 19 and facilitate capital formation. 20 This committee's input is a very important part 21 of that process. 22 Thank you for continuing to devote so much of 23 your time to this committee. And I look forward to 24 today's discussion and the committee's continued work 25 over the next year. 0020 1 With that, back to you, Michael. 2 CHAIRMAN HEANEY: Thank you, Brett. And, 3 again, thank you all for joining us today. As we have 4 mentioned, this is FIMSAC's fourth meeting and the final 5 meeting of our first year. 6 As Brett noted, over the course of the day, we 7 will hear updates on the work and priorities of our 8 subcommittees. We will consider three recommendations, 9 two from the ETF and Bond Fund Subcommittee and one from 10 Technology and Electronic Trading Subcommittee. 11 In addition to the consideration of the three 12 potential recommendations, we will also receive updates 13 from the two transparency subcommittees and learn about a 14 recent MSRB staff study on pre-trade transparency. 15 As we begin to look forward to the second year 16 of this committee's tenure, we will also have an 17 opportunity to discuss and debate potential future topics 18 and subcommittee structures that we feel may be required 19 to address issues not currently being contemplated by the 20 four current subcommittees. 21 In particular, our afternoon panel, as Brett 22 mentioned, on the role of credit ratings and the open 23 discussion that follows will provide us with additional 24 time to begin that conversation and explore other 25 potential topics that we may choose collectively to 0021 1 pursue in 2019. 2 As you all know, we have already scheduled the 3 remaining meetings of the FIMSAC, which will be held th th th , 4 January 28 , May 6 , July 29 and October 28 2019. 5 And, finally, on a personal note, I would just 6 like to take this opportunity to sincerely thank all the 7 FIMSAC members for the continued hard work and engagement 8 on the range of issues that we have had during the course 9 of 2018. 10 As is reflected in the FIMSAC subcommittee web 11 page, I can personally attest having participated on 12 these calls, we have all dedicated a tremendous amount of 13 time in exploring issues critical to the fixed income 14 markets. In doing so, you have invited the input of 15 outside market participants to supplement your 16 considerable expertise, offering others the opportunity 17 to participate in the debate and provide their valuable 18 perspective on these important issues. 19 The collaborative approach is resulting in key 20 recommendations being discussed and debated by this 21 committee and provided to the Commission for its 22 consideration. 23 And I would just like to say with all the 24 amount of time being spent, this is while many of you are 25 working full time running companies, running businesses, 0022 1 running portfolios, running the finances of cities, 2 municipalities or corporations, so, again, a warm and big 3 thank you. 4 As always, we look forward to a thoughtful and 5 a frank a conversation and discussion over the course of 6 today's meetings. As a reminder, during the discussions, 7 I would ask that committee members please turn their 8 cards if they have comments or questions just to ensure 9 that we can get all to be able to participate and engage 10 in the conversation. 11 Okay, we will get our business started and turn 12 over it to our first presentation from the ETF and Bond 13 Fund Subcommittee. As I mentioned, Anath, who serves as 14 the subcommittee chair and the other members of the 15 subcommittee will be presenting the two recommendations 16 for our members to consider. 17 Anath, I think it is probably prudent if you 18 and the subcommittees mention both recommendations 19 consecutively, and then we will take our time to go in 20 turn and have discussion and questions on each of them. 21 But I turn it over to you. 22 MR. MADHAVAN: Okay, thank you very much, 23 Brett. 24 First of all, let me start by reiterating my 25 thanks to the entire FIMSAC Subcommittee on ETFs and Bond 0023 1 Funds. We have to my left, Lynn Martin, Amy McGarrity, 2 Kumar Venkataraman and Rachel Wilson, as well of course 3 as Michael, who has been participating in all of our 4 calls. 5 So just again by way of reminder, the Fund 6 Subcommittee was tasked by FIMSAC as follows, and I will 7 just read an excerpt from our mandate: 8 "The subcommittee is charged with possessing 9 the consequences of the increased presence of fixed 10 income mutual funds and ETFs, including the current and 11 possible future impacts on the liquidity and pricing of 12 the underlying bonds under a variety of scenarios, as 13 well as investor understanding of these products. 14 Topics that may be considered include the 15 interaction of fixed income prices and fund prices, 16 including through the ETF arbitrage process an assessment 17 of the impact of redemptions from funds in stressed 18 conditions, an assessment of the potential impact of the 19 rebalancing process on underlying markets, whether funds 20 help to diversify sources of fixed income liquidity, the 21 role of index providers and retail investor education." 22 So, a large mandate. Since the formation of 23 the Fund Subcommittee, a considerable amount of 24 discussion has taken place among the members. And we 25 have numerous presentations from various industry 0024 1 participants across the entire ecosystem from academics 2 and from regulatory bodies as well. 3 We have had active engagement and participation 4 in the unanimous set of recommendations that we are 5 making today. Overall, a very discussion and broad 6 debate. 7 In terms of our charter, we have identified 8 three main work streams. The first is a draft comment 9 letter for FIMSAC to review prior to sending to the 10 Commission regarding the appropriate classification and 11 labeling of exchange-traded products. Subcommittee 12 member Lynn Martin will go into the draft letter in 13 detail in a moment. 14 The second major work stream relates to our 15 conclusions and recommendations regarding education and 16 data, particularly as they apply to the proposed new Rule 17 6(e)11 under the Investment Company Act of 1940. 18 Subcommittee member Kumar Venkataraman will discuss our 19 recommendations to the broader committee for 20 consideration and discussion. 21 And, as Michael has indicated, we will do those 22 consecutively and then take questions because one topic 23 may actually be the subject of more debate. 24 The third topic in terms of our mandate 25 concerns stress markets. The subcommittee is still in 0025 1 the process of reviewing evidence and will be writing a 2 formal report summarizing our findings for presentation 3 at the January FIMSAC meeting. 4 And with that, I will turn it over to Lynn 5 Martin for an overview of our thinking on ETF 6 classification and other issues specific to Rule 6(e)11. 7 MS. MARTIN: Thank you, Anath. Thank you, 8 Commissioners and staff from the SEC. 9 As Anath mentioned, many of us have been 10 actively working on a comment letter for the recently- 11 proposed 6(e)11 rulemaking. 12 First, I would like to say that we commend the 13 SEC for its forward thinking with its proposed rulemaking 14 as we believe that the proposed rulemaking will increase 15 standardization in this important instrument. 16 As background and for those on FIMSAC who are 17 less familiar with this type of product, ETFs have become 18 an increasingly popular investment vehicle for investors, 19 particularly retail who looks to gain exposure to a 20 certain asset class using a low-cost liquid vehicle. 21 I think this is probably best evidenced by the 22 growth of ETFs over the past few years. Primary flows 23 across asset classes have grown steadily with cumulative 24 annual growth rates of more than 10 percent since the 25 introduction of this asset class. This year alone we 0026 1 believe that there will be at least $2.4 trillion in 2 primary flows. 3 That said and as markets have developed, so 4 have the complexity of these instruments, particularly 5 when you look at the superset of the class called the 6 exchange-traded products. 7 As complexity has increased, the ETPs have 8 embedded certain risk characteristics which deviate from 9 the initial more standard ETFs that were initially 10 introduced in the market and that initially gained 11 popularity amongst the retail segment. 12 I think this is probably best evidenced from 13 back in February when volatility spiked and some of these 14 ETFs or ETPs rather lost about 90 percent of their value. 15 Our comment letter proposes to introduce a 16 streamline classification system. And by doing so, we 17 believe that that will help educate the investors with 18 the risks associated with each of those products. 19 In fact, in the proposed rulemaking, 20 classification buckets are proposed. However, given the 21 complexity in the markets and things such beta investing, 22 we believe some of the proposed classifications may be a 23 bit blurred. 24 So, instead, with our proposed classification 25 system, we attempted to take into account the realities 0027 1 of how this market has evolved and also plan for 2 potential future growth in this asset class and future 3 diversity of underlyings. 4 Importantly, when we were developing this 5 comment letter, it was important to us to not be 6 prescriptive when it comes to the structure of an ETF, 7 into which bucket it would fall into. Specifically, 8 certain of the very liquid ETFs are registered as unit 9 investment trusts as opposed to registered investment 10 companies. We believe that any ETF today who falls into 11 this designation should be grandfathered in to the term 12 "ETF" given the fact that it is not complexity of 13 underlyings, it is more a choice around the structuring 14 of the investment vehicle that was made at the time of 15 issuance. 16 So, more specifically, our recommendation calls 17 for the following buckets: 18 I would say the superset that we are 19 recommending continues to be the ETP classification, 20 exchange-traded products. 21 Underneath that, there are four subsets or four 22 sub-classifications, the ETF classification, which would 23 be an instrument that complies with Rule 6(e)11 except 24 for the aforementioned ETFs that we wish to grandfather 25 in. Those underlyings could be in this instrument, could 0028 1 be a stock bond, fixed income instruments, futures, 2 basically anything very liquid and exchange traded. 3 Importantly, ETFs in our proposal would not 4 include funds with any embedded leverage or inverse 5 characteristics. 6 The second category we are proposing today is 7 the ETM category, which is a debt-based instrument which 8 provides an index-based return. And those are not 9 subject to 6(c)11. 10 The third bucket, with its own unique 11 characteristics that we are proposing, is exchange-traded 12 commodities. Again, these products also do not comply 13 with 6(c)11. And as the name would suggest, their 14 underlyings are either physical commodities or commodity 15 derivatives. 16 And then finally a bucket called exchange- 17 traded instruments, which is an open-ended registered 18 investment company. These tend to cover the products 19 that provide leverage and inverse to the market, as was 20 seen in the February challenges. 21 So, as mentioned above, and in addition to the 22 existent classifications in the market, we are really 23 attempting to introduce three new -- two new 24 classifications, the ETC bucket and the ETI bucket. 25 We are happy to answer any additional questions 0029 1 that you may have. I have quickly gone through that, but 2 we believe that this classification system coupled with 3 the proposed educational system will add additional 4 transparency when it comes to the investing public. 5 With that, I will turn it over to Kumar who 6 will comment on the education platform. 7 MR. VENKATARAMAN: Thank you, Lynn. Good 8 morning everyone. Good morning, Commissioners. I would 9 like to thank all the members on my subcommittee as well 10 as the Commission staff, who have been extremely helpful 11 as we put together these recommendations. 12 As Lynn pointed out, we have we believe 13 identified some recommendations regarding classification 14 of ETPs that will help address many of the educational 15 elements of the subcommittee's charter. 16 However, as we were collecting information, 17 there were some other distinct issues that came up that 18 falls outside the classification effort. And I would 19 like to detail that. 20 In our discussions with the industry 21 participants, academics and regulators, we asked to what 22 extent do financial advisors and retail investors 23 understand the ETF products that they are buying. In 24 particular, we recognize a special need to place greater 25 emphasis on investor trading knowledge, particularly 0030 1 around ETFs. 2 And in this context, we heard back that many 3 participants felt that it is important to educate 4 investors, particularly how the educational needs of 5 retail investors differ from those of institutional 6 investors and how information aimed at retail investors 7 should be delivered in a different manner and in distinct 8 plain language. 9 Further, there is a general view among 10 participants that education will be more impactful if it 11 is provided from a cross-section of academics and 12 industry groups. 13 So, to this end, we recommend formation of an 14 industry-wide group that includes ETF issuers, academics 15 and others that will further investor education. We 16 recommend that the Commission in this industry-wide group 17 work towards a robust and detailed training module for 18 financial advisors that fits within their existing 19 required certification, as well as their continuing 20 education efforts. 21 Items that would be covered under this training 22 module include understanding the trading process of ETFs, 23 including appropriate usage of orders and types of 24 transactions costs, a recommendation that ETF premiums 25 and discounts can be misleading when NAV is still when 0031 1 transactions may occur at prices different from intrinsic 2 value, depending on market volatility, liquidity and time 3 of day. And the correct usage of terms such as ETFs as 4 specified by our classification scheme. 5 Some education for retail investors should be 6 completed at the platform level, focusing on improved 7 disclosure of ETF risks, particularly regarding trading 8 and management of investor expectations. 9 The Commission's proposed Rule 6(c)11 request 10 comment outlined specific potential areas for consistent 11 disclosures, which we believe would be effective in 12 improving investor awareness. 13 We support the Commission's proposal to expand 14 disclosure of historical beta spreads associated with 15 ETFs. However, we found that it is frequently difficult 16 for investors to compare even structurally similar ETPs 17 because various market data services, broker-dealers and 18 ETF issuers have historically each reported basic 19 metrics, such as yield spreads, aggregation, beta 20 spreads, et cetera using their own proprietary methods. 21 This has made it difficult to compare fixed income ETFs 22 through other fixed income instruments as well as with 23 each other. 24 So, in this context, we asked industry 25 participants, academics and regulators for suggestions on 0032 1 how we can improve the ETF ecosystem. In particular, 2 data that is needed to support and further investor 3 education. 4 Their feedback indicates that it is important 5 to improve the availability, quality and quantity of data 6 related to the ETF market. As this market continues to 7 grow, it will be critical for regulators to have 8 consistent readily available data to analyze. Feedback 9 received from academics highlighted that data on ETFs is 10 hard to find and not centralized or standardized. This 11 comment also applies to advisors and others, including 12 sophisticated investors who know the difficulty in 13 comparing even basic data, such as bid/ask spreads across 14 asset managers. 15 Other important data, such as the ratio of 16 primary to secondary market volumes or number of 17 authorized participants is difficult to find. 18 We believe that data and interface 19 standardization is a first step. We recommend that the 20 industry-wide group create a primer analyzing ETFs, 21 focusing on identify key data aspects along with common 22 definitions for the use of investors, regulators and 23 academics and interfaces that are standardized and 24 consistent. 25 We recommend that these data include portfolio 0033 1 information, such as ETF characteristics and intrinsic 2 values and trading data, such as bid offer spreads, ETF 3 holdings, among others. 4 A detailed list of the data that we recommend 5 is provided in our memorandum. 6 Market-makers play an important role in the 7 smooth functioning of the ETF market by their 8 participation in the credit redeem process and the 9 provision of secondary market liquidity. However, there 10 is a lack of suitable data and common standards which 11 makes it difficult to assess whether the arbitrage 12 process works well, particularly under stress conditions, 13 and to identify potential problems that need to be 14 addressed. 15 To this end, we recommend that transactions 16 data that identifies market-maker trades create redeem 17 order information, ETFs daily creation and redemption 18 basket information be collected in a centralized manner. 19 We recognize the sensitive nature of the data and 20 recommend that some information can be made available 21 with a delay, say one year or 18 months, for academic and 22 industry usage on a non-discriminatory basis. 23 Thank you. 24 CHAIRMAN HEANEY: First, and as the committee 25 members can hear or have heard, there is a large body of 0034 1 work that this subcommittee went through to come to the 2 two recommendations. So, first, I would like to commend 3 them all on the hard work. 4 I think, Anath, at this point let's open it up 5 to the comment letter recommendation that Lynn present. 6 And we'll open it up to the FIMSAC members for any 7 comments or any questions on that initially before we 8 move. 9 Elisse? 10 MS. WALTER: Thank you, Michael. And thank you 11 for all of your work. I am very impressed by the output, 12 and I couldn't agree with you more. 13 I just had one question, which doesn't affect 14 my support for the recommendation, which is how much 15 difficulty do you think there will be trying to determine 16 with any particular product what category it is going to 17 fall in? And how much need do you see for interpretative 18 advice in the future? Is this going to be more or less 19 self-executing or are we headed for some rough waters 20 until this settles in? 21 MR. MADHAVAN: Thank you, Elisse. Yes, indeed, 22 I think we have talked a little bit about the practical 23 aspects of the truth in labeling effort that is the 24 subject of the first letter. 25 For the most part, I think we have given some 0035 1 deep thought as to how these categories would be defined 2 and what particular funds might fall into these 3 categories. 4 I suspect as we evolve this, we will learn by 5 doing. And there may be some funds that you know perhaps 6 cross the boundaries where additional guidance from the 7 Commission would be required. 8 I don't know, Lynn, if you wanted to add to 9 that? 10 MS. MARTIN: Yeah, no, I agree with that. We 11 tried to be as prescriptive as possible without being 12 overly prescriptive. And, as you can see from our body 13 of work, we did try to take into account the 14 practicalities of some of the instruments that are 15 currently in the market. It is why for the ETF 16 classification, for example, we didn't require it to be 17 all registered investment companies. And we also allowed 18 for the grandfathering in of certain unit investment 19 trusts to be part of that. 20 CHAIRMAN HEANEY: Mihir? 21 MR. WORAH: My question was very much along 22 Elisse's line. So how do you -- so, firstly, again, I 23 think -- I think what you guys are doing is a good 24 recommendation, and we should go forward with it. But 25 just some of the finer points around it. 0036 1 So, if I had an actively managed ETF trading 2 only liquid products, how would you see whether it was 3 equities or fixed income, and I decided to substitute 4 some of my you know energy producers with crude oil 5 exposure, say 5 percent or so, is this an ETF? Is this 6 an ETC? Are there some percentage limits? How do you 7 guys see that? 8 MR. MADHAVAN: Let me take that one. It's a 9 great question. It is very similar to Elisse's question. 10 We did consider alternative ways of classifying. In 11 particular, one of our early discussions centered on the 12 idea that maybe make a cut between active products and 13 index products. 14 I think, as was touched upon, in earlier 15 remarks, the boundaries between active and index 16 investing are increasingly blurred. So, Mihir, you gave 17 an example of a fund that is perhaps tracking a custom 18 benchmark. It may not necessarily market cap weighted. 19 The portfolio manager may have some degree of flexibility 20 in terms of sector weights, in terms of industry weights, 21 in terms of even the types of products or assets that are 22 in the fund. 23 Again, in our mind, the main objective is truth 24 in labeling. As long as the fund is in very liquid 25 instruments, be they stocks, bonds, futures contracts and 0037 1 the like. Liquid, easy to value. Instruments does not 2 have embedded leverage and is not basically an 3 uncollateralized debt instrument. Then it would be 4 classified as an ETF. 5 So, in your example, unless there were some 6 very unusual structural features that would require 7 attention by the Commission staff, I would say that it 8 would fall into the ETF category as would all products 9 that are currently actively managed ETFs. 10 So, our cut is not about active in the index 11 where the field of boundaries are very, very blurred and 12 getting more blurred but more about the kinds of holdings 13 of these funds and their goals in terms of returns over 14 specific intervals of time. 15 MR. WORAH: I think -- again, just to fine- 16 tune. So, obviously the recommendation goes forward, but 17 there will be a lot of fine-tuning. But some kind of, 18 you know, in the naming convention, some kind of 19 percentage, like the SEC has right now for 40 Act mutual 20 funds, if you have a name, then you have got to have 80 21 percent of your assets in that kind of security. 22 The specific example I had was crude oil. If I 23 have 5 percent crude oil in an ETF, do I become an ETC or 24 can I still market it as ETF? You know, so I think there 25 should be some thought around those kinds of things. 0038 1 CHAIRMAN HEANEY: And, Anath, correct me if I'm 2 wrong, this is a letter recommendation to the SEC where 3 some of that fine-tuning can take place? 4 MR. MADHAVAN: Yes, exactly. Yeah, Michael, 5 the goal would be that we would get the letter. This is 6 a letter that we -- a draft letter that we have written 7 for FIMSAC to approve and then go on to send to the 8 Commission where I think a lot of this fine-tuning and 9 you know interpretation would be then done at a later 10 stage. 11 CHAIRMAN HEANEY: Larry? 12 MR. HARRIS: I think this classification is 13 very important, and I appreciate the work that everybody 14 has done on it. 15 I wanted to ask only one question to further 16 explore the relationship or the importance of active 17 versus passively managed funds. So, the classification 18 is not -- the proposed classification is not going to 19 identify that difference. We do not identify that 20 difference presently with respect to open-ended mutual 21 funds, although we're careful about how they name 22 themselves so that they are representative. 23 I am concerned only about the ETFs because 24 traditionally most people have associated ETFs with 25 indices. And so the question is at this juncture when we 0039 1 have an opportunity to do something, would it be very 2 simple to add some language about active versus passive 3 in how we title things just to make sure that things are 4 more clear. 5 Alternatively, it might just be sufficient that 6 the educational component that has already been proposed 7 have enough stuff in it that we hope that people will 8 understand that there is a difference between active and 9 passive. 10 But when you think about the risk 11 characteristics of funds, this is a very important risk 12 characteristic. As long as we are addressing those 13 characteristics, I wonder if we shouldn't be thinking 14 more about this particular one. 15 MR. MADHAVAN: Let me just comment on that 16 really quickly. And, again, I know our subcommittee 17 members also have views on this. 18 But the active and passive classification or 19 distinction is increasingly hard to make. So, Mihir gave 20 an example. I mean another one might be an international 21 equity fund, market capitalization weighted. You could 22 argue that is an index product, but what if it has layer 23 of smarts where it hedges some countries or some regions 24 and not others based on the cost of getting that currency 25 hedge on. Is that an active product or a mostly active 0040 1 product or a partially active product? 2 So, these are the kinds of discussions we had. 3 And I think, Larry, your point is well taken that there 4 is room in the education component for making these 5 distinctions and for making investors aware that some 6 products are attempting to beat an index, others are 7 attempting to track an index. 8 But I think again what we are seeing in the 9 marketplace is a blurring of the lines. And many index 10 products have features that could be considered active, 11 such as intelligent currency hedging. 12 Let me turn over to the other subcommittee 13 members for their views too. 14 MS. MARTIN: Yeah, I would agree with you, 15 Anath. And, again, we are trying to -- when we came up 16 with the suggested framework, we were trying to almost 17 future proof it as well as the lines between active and 18 passive continue to blur. We did not want to be too 19 prescriptive as to active versus passive, although it is 20 a great point. And I agree with you that it could very 21 much be folded into the education plan that we have 22 developed. 23 MS. WILSON: Yes, just kind of further to that, 24 there is one idea of constructing more by what does it 25 hold and what is its intention. And just being careful 0041 1 if it is leveraged or not, and trying to make sure there 2 are some distinction or commodity. The other is more of 3 some of these, how is it being managed, right? And that 4 is when we talked about passive versus active. We looked 5 at if it is rules-based versus discretionary. If it was 6 long/short and leveraged, exposures and using derivatives 7 and the creation and redemption arbitrage mechanisms, 8 right. 9 So, you kind of -- there obviously could be a 10 matrix between all of that. But as we thought about what 11 would be the more clear-cut way to cut some of the 12 nomenclature, we felt that more of what does it hold or 13 constitute seemed easier than looking at some of these 14 factors that tend to blur. 15 And I know a lot of us are talking about 16 passive versus active. That is a very clear one for a 17 lot of us as to where that blurring is starting to occur. 18 MR. HARRIS: So, I hear a lot of comment about 19 the difficulty of distinguishing between active versus 20 passive. And I understand it is -- there are so many 21 different types of judgment that go into the construction 22 of these products. And, of course, the smart betas could 23 be argued as an active strategy. 24 That said, it seems to me there is a very 25 simple way to discriminate between active versus passive. 0042 1 So, active funds are funds where there is ongoing 2 judgment as to what is held in the portfolio. Passive 3 funds are funds where the criteria for construction the 4 portfolio can easily be enumerated and probably should be 5 enumerated publicly. 6 And for those funds that have well-defined 7 criteria, then they should qualify as being passive funds 8 regardless of the complexity of that criteria because 9 they are following a specific set of well-founded rules. 10 And for those funds that are engaged in 11 judgments, whether those judgments are human judgments or 12 even machine-based judgments, that would be an actively 13 managed fund. 14 And the difference is very important for policy 15 and for investors because in one case the risk is fully 16 disclosed, though may be difficult to analyze. In the 17 other case, the risk is disclosed but certainly not 18 easily analyzed because of the judgment issue. 19 So, I am of mixed mind about where to put this. 20 I am just drawing attention to the fact that I think it 21 is very important. And I can see an argument for at this 22 point perhaps using a single word in titles or something 23 like that to distinguish between those products that are 24 -- that follow mechanical rules and those products that 25 allow more investment judgment. 0043 1 COMMISSIONER JACKSON: Michael, I am so sorry, 2 I do not mean to interrupt. I just have one follow-up 3 question for you, Larry. What I hear you say -- these 4 are very important issues, but I hear you saying that you 5 are two minds and in fact open to the notion that this 6 could be addressed in the education portion of this 7 proposal as opposed to the naming convention, do I have 8 that right? 9 MR. HARRIS: Yes, and the trade off here really 10 has to do with where we think education is going to be 11 most effective. So, in an ideal world, we would have 12 everybody read everything that we ever write. And in a 13 real world, we know that never happens. 14 COMMISSIONER JACKSON: Really? What? 15 MR. HARRIS: So, when we have a simple 16 opportunity to draw a line between something that is 17 really easily well-defined, so mechanical versus human 18 judgment or, you know, some other type of judgment, might 19 the insertion of a single word in a title be sufficient 20 to provide the -- if not the education, the impetus to get 21 educated? 22 And so I hate to push things, and yet, when 23 things are really, really easy, then they should be 24 pushed. 25 CHAIRMAN HEANEY: I -- look, Larry, I think that's -- 0044 1 it's a good point. And, Anath, correct me if I am wrong, you 2 are seeing this -- you, Lynn -- subcommittee -- view this 3 as more in the education portion of it though, correct? 4 Not -- not dismissing it, just not including it in this portion? 5 MR. MADHAVAN: That's right. I think some of 6 the earlier questions that Mihir and others have had about 7 how would this actually work in practice, get even more 8 complex when we are talking about -- we'll say, in Larry's idea, 9 maybe rules-based versus more PM discretionary-based. Still, 10 it's -- it's very difficult. 11 Again, you can imagine a fully rules-based 12 system that, you know, classifies sectors based on some 13 textual analysis or something like that. You can write 14 it down; but you could argue that, you know, something that is 15 rotating between sectors based on some, some artificial 16 intelligence kind of criteria might be actually quite 17 active, even though it's rules-based. 18 So, I just think the practical aspects of this 19 are such that perhaps it fits better in the education 20 part of it. 21 CHAIRMAN HEANEY: And I just want to draw 22 people -- thank you -- to the comment letter recommendation 23 in Tab B. I am sure people have read it. But it does, to 24 the point being mentioned, it does have the opportunity 25 for more granularity, more classifications, as it is 0045 1 submitted to the SEC. 2 So, I just think that is important for us as 3 FIMSAC members to think about as we head into the vote. 4 Other comments, please? 5 MR. WORAH: I agree. I think Larry's point is 6 an important one. In the fine-tuning process, whether it 7 is in education or whether it is the naming convention, 8 but you know just as you guys fine-tune if this 9 recommendation is adopted, again the active versus 10 passive, typically talking about smart beta, which is 11 where the lines are getting blurred and which is the 12 point Anath made. 13 You could be rule-based. So, I will speak for 14 ourselves. We have smart beta strategies, but you are 15 constantly doing research on them. What you define as 16 value changes. And so even smart beta rule-based 17 strategies, there is constant research going on. They are 18 not stuck forever. 19 So, it is an important point, but there's nuances 20 there that should be considered. 21 CHAIRMAN HEANEY: Others? Please, I am sorry, 22 Amy. 23 MS. EDWARDS: All right; thank you. I wanted to 24 touch on the IIV portion of the comment letter a bit. 25 I note that it looks like the costs that were 0046 1 considered by the subcommittee were similar to what was 2 in the economic analysis, and so I wanted to get a little 3 bit more of your thoughts on the costs of the IIV portion 4 of the proposal. 5 But, also, I think you are recommending that 6 there should be continued some mandated IIV. And so I 7 wanted to get your thoughts on what market failure exists 8 that should require the mandated IIVs? 9 MS. MARTIN: So, I will speak to why I think 10 IIVs are good for the market, and I will let Anath or 11 others talk about the specific cost because, as you can 12 imagine, there are bunch of different perspectives on 13 this issue. 14 In my mind, IIVs are good for the market for 15 the retail sector of the market because it does 16 constitute a value that the retail can point to 17 throughout the course of a day. 18 Now, a participant that is very sophisticated, 19 such as a market-maker, an authorized participant, makes 20 a living out of coming up with tools that may produce 21 different values than what the IIV in the market does. 22 However, particularly for the more liquid instruments, 23 the U.S. equities markets, I think the role of IIVs has 24 worked incredibly well for the markets. And for retail, 25 they very much provide transparency to retail throughout 0047 1 the course of the day. 2 MR. MADHAVAN: Okay, so, Amy, to your question, 3 I think the committee, the subcommittee felt that IIVs 4 indeed valuable particularly to retail investors. I 5 think the challenge that we had was how do we produce 6 meaningful numbers? So, for example, you know a foreign 7 equity market might be closed and the IIV that we maybe 8 seeing may simply be the last price adjusted for currency 9 movements. That is not really representative of the true 10 value of the portfolio. 11 So that was our consideration of terms of you 12 know just the kinds of information. So, there's a 13 balance of perhaps some of the information we provide may 14 be misleading. Some of it may actually be quite helpful. 15 How do we strike that balance? And that was a topic that 16 we discussed quite a bit. 17 And I don't know, again, Amy, Kumar, Rachel, if 18 you want to add to that? 19 MR. VENKATARAMAN: Yes, I agree. I think it is 20 a really good question, and we talked about it quite a 21 bit. It really for when the underlying securities are 22 actively traded, the IIV is quite useful. So that takes 23 care of a slice of the market where the information will 24 be valuable. 25 When the underlying securities are liquid, you 0048 1 worry about the pricing being a bit stale and whether it 2 is sufficiently informative. So, we felt that it is 3 useful to provide the information but with the caveat 4 that the prices may be stale, so one needs to be careful 5 for certain types of underlying securities. So, that's 6 part of the education recommendation that we have as well 7 where we have recommended that the training module for 8 financial advisors, as well as the platform-based 9 education for retail investors include some information 10 about what has been produced in the process with the 11 understanding that the pricing may be stale. 12 So, it is a really good question, we struggled. 13 But we felt it will still be useful in many cases and 14 perhaps under some market conditions, it may be less 15 useful. But it is still useful to have that out there. 16 CHAIRMAN HEANEY: Larry? 17 MR. HARRIS: There is a very simple measure of 18 the value of an IIV. It is just look and see whether the 19 inter-day correlation and the changes to the IIV and to 20 the market prices, how closely they are correlated. If 21 they are very closely correlated, then it is valuable. 22 If it is not very closely correlated, then we are 23 presuming that the market prices are pretty accurate, 24 then it is not all that useful. 25 So, I think it is very important that we have 0049 1 IIV published for all ETFs because they are trading 2 inter-day. But it is equally important that if an 3 analysis can simply be done that would characterize the 4 importance of the IIV, that that information be produced 5 by somebody and revealed to the public. It is very 6 inefficient to have everybody doing it and most people 7 can't do it. 8 So, this would be a sensible disclosure item 9 that we might ask of the sponsors of the ETFs. 10 CHAIRMAN HEANEY: Tom? 11 MR. GIRA: Sticking with what Larry said, I 12 think from a surveillance standpoint, while we do look to 13 see if the IIVs diverge from the market, NAV and things 14 like that. So, while it is not a be all or end all, it 15 can be informative that there might be a disconnect in 16 the marketplace. 17 CHAIRMAN HEANEY: Others, questions, comments? 18 Please, Jay? 19 CHAIRMAN CLAYTON: I want to thank the 20 subcommittee for this recommendation. Some of the points 21 made, in particular whether we should have some type of 22 basket to deal with hedging or with commodities, they are 23 things that we should think about. But this is an issue. 24 And I welcome this comment letter, and I think it goes a 25 long way toward addressing an issue that exists by retail 0050 1 investors. 2 As Larry note, people often do not read beyond 3 a title, and if we can do something to substantially 4 improve the understanding of our retail investors as to 5 what their investing in, we should do it. 6 So, I thank you very much. 7 CHAIRMAN HEANEY: Any other questions or 8 comments? Thank you, Chairman. 9 Okay, so at this point I would enter a motion 10 for the comment letter recommendation? 11 MS. WALTER: So moved. 12 CHAIRMAN HEANEY: Thank you, Elisse. Thank 13 you. I would ask all voting members in favor to raise 14 their hands, please. Suzanne, may I ask your vote on the 15 phone, please? 16 MS. SHANK: In favor. 17 CHAIRMAN HEANEY: Thank you, Suzanne. Scott? 18 MR. KROHN: I support. 19 CHAIRMAN HEANEY: Scott. Any voting members 20 opposed or voting members abstaining? Okay, the 21 recommendation has been approved by the committee 22 unanimously. 23 Thank you all. Thank you, Anath, and the 24 subcommittee members for the hard work and, again, the 25 body of great deal and great work that has gone through 0051 1 to create this comment letter recommendation. 2 Let's now stay with this and turn to the 3 education and the availability of data concerning ETFs 4 and open that up to FIMSAC members for questions and 5 comments? Sonali? 6 MS. THEISEN: Thank you again for all of the 7 very hard work. I think this is again a very 8 comprehensive proposal and recommendation. 9 I really have just two questions for clarification 10 from the proposal that you have put forth. 11 First, around the education point. Do you envision 12 education to encompass ETFs only or also products -- 13 other products, such as ETF options? That is my first 14 question. 15 MR. MADHAVAN: Initially, our scope is really 16 about ETF and bond funds. So, we have limited it in our 17 language to the mandate that we were given. 18 I could see at a later point perhaps you know 19 this effort if successful could be expanded to the rest 20 of the ecosystem. But I think that actually goes beyond 21 what we were specifically asked to do, were tasked with. 22 MS. THEISEN: My second question is in relation 23 to there was a point in here regarding participation 24 measures to track Aps, I believe in Section H under, 25 "Trading Data." I wanted to just understand a bit more 0052 1 about how you envisioned this working, i.e., would the 2 information be given by the ETF sponsor, would it be made 3 available publicly and with what frequency, what data 4 points did you have in mind? 5 MR. VENKATARAMAN: So, in our survey of trying 6 to understand how the ETF ecosystem is function, we 7 realized that there is a shortage of useful data to help 8 us understand whether the process is working well. So, 9 information on creation redemption, how often does it 10 happen? How many APs are involved? Does that change 11 over time based on market conditions? 12 So, these are all important aspects of the 13 process to understand the functioning of the market. 14 So, what we suggested here is a way for us to 15 collect this information, perhaps provided by the ETF 16 issuer in this case because they would have the 17 information. And Anath can activate given his expertise. 18 But so with this data being collected, it would 19 help us understand and the regulators understand whether 20 the process is working well. But at the same time, we 21 understand that some of this information could be 22 sensitive. And so we could delay availability of this 23 information for a year or 18 months, along the lines of 24 what FINRA has been doing for say corporate bond 25 transaction data with the dealer identification, which is 0053 1 typically delayed by a three-year period. 2 So, it will help facilitate research and our 3 understanding of how the market works. But at the same 4 time, it addresses potential concerns of sensitive 5 information being provided. 6 MS. THEISEN: So, then would the idea be to 7 collect the data and provide it in a kind of timely real 8 time basis to the regulators but then have it be publicly 9 disseminated with a delay, is that -- am I understanding 10 correctly? 11 MR. VENKATARAMAN: That is correct. 12 MS. THEISEN: Okay, thank you. 13 CHAIRMAN HEANEY: Horace? 14 MR. CARTER: I would like to offer my thanks 15 again and congratulations to the subcommittee for their 16 work. I just have a comment or call it a suggest. 17 When we talk about education in terms of -- I 18 am particularly thinking of the exchange trade of the 19 instruments, we need to do everything that we can to make 20 sure that the explanations are as simple and 21 straightforward as possible. 22 Many of us in this room I think take it for 23 granted that we understand when somebody says 24 "leverage" or "inverse," we understand what the 25 implications of that are. And our retail clients do not 0054 1 necessarily automatically understand that. 2 And so I think we should include things like 3 costs and particularly something that I would to consider 4 is appropriate horizon for investment. So that when a 5 retail investor invests in one of these instruments, they 6 understand the appropriate time to hold such an 7 investment. What are you attempting to do by investing 8 this? And I think that that should be quite prominently 9 in the education material. 10 Thank you. 11 CHAIRMAN HEANEY: Please, Anath, could you 12 comment on that? 13 MR. MADHAVAN: Yes, I will fully concur with 14 Horace's point. I think it is a great point. 15 And we saw this early on with the leverage and 16 inverse products where there is sort of two issues, one 17 is a question of systemic risk with same direction 18 rebalancing. The other one is for investors, the decay 19 over long horizons that just comes from the compounding 20 of facts. 21 And, actually, since some of that earlier work 22 done by myself and Minder Cheng, you know actually the 23 prospectuses have become notably better with clear 24 examples. But I think we still have quite a way to go in 25 terms of education about the appropriate horizon. And I 0055 1 certainly think that again as we fine tune this 2 educational document, and we refine the recommendations, 3 we would probably try to get some of that in there as 4 well. 5 But your point is very well taken. 6 CHAIRMAN HEANEY: Larry? 7 MR. HARRIS: I support the proposal, but I have 8 a comment and a suggestion I think are both important. 9 The production of this educational product by 10 an industry or by the government is useful because it 11 puts in a single place work that many, many people might 12 be doing separately. And so it is efficient. 13 The problem is that in many cases the brokers 14 or the registered investment advisors already have a 15 responsibility to provide this education. And they have 16 a potential liability if they don't do a good job of it. 17 And so the creation of this product moves the 18 liability away from them to another entity, which means 19 it is very important that that entity be well controlled, 20 and in particular that it is not overly influenced by 21 people who might want to minimize risk or something like 22 that. 23 So, as we think about how to design these 24 products, it is important that people who are -- who 25 truly represent the interest of individual investors and 0056 1 also institutional investors but who are further away 2 from the industry than the people who are typically 3 serving those customers, that they be well-represented in 4 the design of this curriculum. 5 The second comment is much simpler. Quite a 6 bit of work has been done by educational researchers into 7 what is effective. Increasingly, we are starting to rely 8 on that research in the design of our materials. But 9 this is a particular place where it would be useful to 10 ensure that people who really understand how people learn 11 are involved in the development of these materials so 12 that they are as effective as possible. Sometimes just 13 making sure that it is written in a simple way or that 14 the font is right or something like that can make a big 15 difference. 16 And so if we are going to make some effort to 17 design these materials, we should really design them well 18 by deferring to people who actually know this stuff. 19 And, fortunately there are a lot of such people. 20 CHAIRMAN HEANEY: Elisse? 21 MS. WALTER: I would like to second what Larry 22 just said, and also that we should give some thought to 23 actually going through a testing process. Having been 24 involved in this in a variety of different scenarios, I 25 am always shocked the things that I think are completely 0057 1 understandable that people just don't understand and 2 things that confuse them. So, testing I think is very 3 important as well. 4 CHAIRMAN HEANEY: Other questions and comments? 5 MR. WORAH: Just a final point. This is 6 separate from the point Larry and Elisse made, which is 7 in terms of the standardization of data and risk metrics 8 to give some thought to just the fact that bond funds are 9 different. Most bonds, many bonds have embedded options, 10 whether they are corporate bonds or mortgage-backed 11 securities in terms of yields, durations, are they option 12 adjusted? Are they not option adjusted? So, either 13 allow the flexibility. Whose models are you using? Or 14 be prescriptive or allow flexibility. I think some 15 thought needs to go into that as well. 16 MR. MADHAVAN: I think it is a very fair 17 comment. And, again, I think our intention was again 18 just to move forward and to establish education data 19 standardization, open access as important principles. 20 And I totally agree again with Mihir that there 21 are a lot of subtleties at OAS, option-adjusted spreads. 22 There's a lot of details that need to be carved out. I 23 would imagine in practice we would give the issuers quite 24 a bit of flexibility in terms of trying to capture those 25 features. 0058 1 CHAIRMAN HEANEY: Chairman Clayton? 2 CHAIRMAN CLAYTON: Just two thoughts. Horace, 3 I want to thank you for your comment. And it kind of 4 dove-tails with Larry's because I can tell you that when 5 I have to write something for this group, maybe it takes 6 an hour so I know that I am going to be understood. If I 7 have to try and write something for someone who is not 8 familiar with finance and all the terms we're familiar, 9 it may take a day because it is harder to write for that 10 audience. And we should recognize that. 11 To Larry's point on whether we would be 12 absolving or limiting the responsibility of financial 13 professionals, I think there should be a clear 14 distinction between describing how something works and a 15 recommendation as to whether it is appropriate for a 16 person. I don't think we're looking at making 17 recommendations with this educational subgroup, but 18 trying to clarity into how these instruments work and 19 things like appropriate horizon as a generic matter. And 20 so I hear your caution, and I have the same view. 21 And as far as testing goes, I am very 22 supportive of it. But I do believe that our enforcement 23 actions and other actions show that education does need 24 to come to this market. If you asked professionals in 25 this marketplace, whether leveraged products, whether 0059 1 they were inverse or following the market or pro-cyclical 2 when they had to rebalance, I doubt many would get that 3 question right. And so we need to -- we do need to bring 4 some education to this marketplace. 5 CHAIRMAN HEANEY: And, again, as the comment I 6 mentioned in the first recommendation, this investor 7 education and data in Tab C, as Brett and I just 8 discussed too, can be tweaked, amended, worked on if 9 FIMSAC so chooses to pass this recommendation to include 10 definition of terms, to include acceptable holding 11 period. 12 So, there is ways to continue as the SEC, as we 13 move it to the SEC, should we do that for some of these 14 modifications as being talked about as Chairman Clayton 15 mentioned for all these to be still amended, added. 16 Tom? 17 MR. GIRA: Just curious, I have always had this 18 thought, and I was wondering if you all thought about 19 this. You know, we're talking about education. And I 20 wonder if another way we could think about this is, no 21 pun intended, but leveraging another regime that we 22 already have options where there is a pre-approval, 23 there's sort of education that goes on with those 24 investors that is separate and apart from the plain 25 vanilla equities. 0060 1 And maybe this is an issue for another day, but 2 I was curious if this came up in the discussions. What 3 if for those really complex, inverse leveraged, things 4 that we have now and will probably occur in the future, I 5 think in The Netherlands they trade turbo ETFs, you know 6 whether that could be a regime to maybe bring it to the 7 mix. 8 MR. VENKATARAMAN: I think that is a very good 9 idea. Clearly, there are some ETFs which are more 10 straightforward, but complexity in some other products is 11 very high. And, clearly, some participants will not 12 understand it too. So, to the extent that there is some 13 kind of a module which tests them before they are allowed 14 to participate in these types of products, I think 15 clearly will be helpful. 16 CHAIRMAN CLAYTON: Yes, let me just say I 17 support exploring this question because there are some 18 ETFs that when I tried to explain them, I start by 19 explaining how an option works and how this is just a 20 little different. And I seem to think that in many of 21 these cases, it is much closer to an option than it is to 22 a straight cash investment. And I think Tom's point is a 23 very good one. 24 CHAIRMAN HEANEY: I think it would be incorrect 25 to say this is scratching the surface on education. This 0061 1 is a thorough document, but there is plenty to do once 2 this door is cracked a bit. And I think specificity for 3 sure going down the line could help. 4 I would refresh the FIMSAC committee members to 5 some of the panels we have had in the past, including 6 those that we talked about education to the retail truly 7 individual level and how the wealth management advisors 8 saw it as their responsibility to do the education or the 9 education stop there. Something that probably we can 10 continue to pursue as a question as to whether that 11 should be the case or not, to your point, Tom. 12 Amy? 13 MS. EDWARDS: I guess in following with some of 14 these statements. I wonder if you could elaborate on the 15 current state of the education of the financial advisors 16 who are providing this information to retail investors 17 and the variation in what they know relative to the 18 recommendation? 19 MR. MADHAVAN: Well, let me start by just 20 saying that, yes, I think we have heard from many people 21 in the space. And it was a little eye opening, I think 22 we realized there is quite a bit of variation. 23 One category that Kumar mentioned that we 24 haven't touched upon in the Q&A yet is about either the 25 kind of knowledge about trading ETPs that advisors need 0062 1 to have. So, to be an ETP investor, you need to make a 2 trade. And there are basic things about market orders 3 and limit orders and stop orders and so on where, again, 4 there seems to be some sort of gaps in terms of knowledge 5 that the hope is that this educational document would 6 help advance that. 7 th 8 We saw it very clearly on February 6 of just 9 this year when we saw these products, the volatility 10 products, you know some of them go to zero in a day. 11 They performed exactly as described in the prospectus, 12 but many, many financial advisors had no idea, certainly 13 many retail investors in these products had no idea that 14 that could happen. Again, it gets to the point of can we 15 explain in clear, clear language to advisors and then to 16 the people that are buying these products ultimately what 17 they are designed to do. 18 And I would go again to Chairman Clayton's 19 remarks about some of these products look a lot more like 20 an option. It is much easier to explain them as options 21 than as funds. 22 CHAIRMAN HEANEY: Elisse? 23 MS. WALTER: Just one brief footnote perhaps to 24 what Larry said before. And not to volunteer Tom, but I 25 will volunteer Tom. The FINRA investor ed folks and the 0063 1 FINRA Investor Education Foundation, on whose board I 2 sit, has a lot of experience in these matters. And I am 3 sure would be very, very happy to lend some assistance to 4 the educational efforts. 5 MR. GIRA: Happy to be volunteered. 6 CHAIRMAN HEANEY: Thank you, Elisse. Other 7 comments and questions? 8 Okay, at this point, I would like to open it 9 for a motion on the vote for the investor education and 10 data recommendation as found in Tab C. 11 MS. SHANK: Support. 12 CHAIRMAN HEANEY: Second? 13 MR. GARCIA: Second. 14 CHAIRMAN HEANEY: Thank you, Gilbert. All 15 those voting members in favor, I would ask you to please 16 raise your hands. 17 All voting members opposed? 18 Voting members abstaining? 19 Suzanne, I would like to ask for your vote as 20 well, please? 21 MS. SHANK: I support. 22 CHAIRMAN HEANEY: Thank you, Suzanne. Scott? 23 MR. KROHN: In favor. 24 CHAIRMAN HEANEY: Thank you, Scott. 25 Again, the recommendation has been approved 0064 1 unanimously. I would like to thank again Anath and the 2 subcommittee not only for the two recommendations today 3 but for their work all throughout the year. An 4 incredible amount of time spent, an incredible amount of 5 progress made, an incredible amount of education for all 6 of us on FIMSAC in terms of this topic, so thank you very 7 much. 8 We are way ahead of schedule, as should happen 9 with any board. So, why don't we take our 15-minute 10 break now and reassemble here at 11:05 if that is okay. 11 And, Rick, we will jumpstart a little early. Great. 12 Thank you. 13 [Break.] 14 CHAIRMAN HEANEY: Okay, welcome back. We will 15 now turn to our consideration for the recommendation from 16 Technology and Electronic Trading Subcommittee on the 17 collection and dissemination of corporate bond new issue 18 data. 19 I will turn it over to Rick McVey, who will 20 introduce this and introduce the panel as well. And we 21 will go from there. Thanks, Rick. 22 MR. MCVEY: Thanks, Michael. And let me kick 23 off by thanking all the members of the Technology and E- 24 Trading Subcommittee. We have been fast at work again 25 over the past quarter. This has been our primary topic 0065 1 around potential improvements to fixed income reference 2 data that would benefit U.S. fixed income market 3 participants. 4 We have also spent considerable time over the 5 last quarter on last look practices on e-trading venues, 6 especially with a focus on retail. And we will continue 7 those deliberations in the coming months but working very 8 hard on potential improvements to market structure there. 9 I would also like to thank the panelists that 10 we have assembled today to help us think through the 11 current state of fixed income reference data, and the 12 recommendation that we have put forward. All five of 13 them have helped the committee on our various calls. 14 And what we identified through that process is 15 that there are indeed gaps in corporate bond fixed income 16 reference data, both in the timing of when that data is 17 available with different reference data providers, as 18 well as sometimes the accuracy. 19 And we consider both of those to be significant 20 issues. One, because if e-trading venues are using 21 different reference data providers, and some of them get 22 the new issue data later than others, then you can run 23 into situations where bonds are available on some systems 24 and not others during the active period following a new 25 issue. 0066 1 Secondly, there is significant manual effort 2 today in getting new issue information into various 3 databases. And that is prone to error. Reference data 4 errors lead directly to trading errors. 5 So, we thought this work was important. And, 6 again, lots of good debate and discussion went into the 7 recommendation that was unanimously approved by the 8 subcommittee to bring forward to the full FIMSAC 9 committee today. 10 So, let me stop there and just ask all the 11 panelists to introduce themselves and a little bit about 12 their responsibilities. And then we will get started 13 with Q&A. 14 MR. SEDGWICK: I am Alex Sedgwick. I represent 15 T. Rowe Price where I run fixed income market structure 16 and electronic trading. 17 MR. PERSSON: Good morning. I am Ola Persson. 18 I am senior vice president of transparency services with 19 FINRA. So, we are the department which is tasked 20 operating the trade reporting facilities and, in this 21 case, I think we are focused on TRACE. 22 MR. LOBUE: Bob LoBue. I work at J.P. Morgan. 23 I run our global fixed income syndicate desk, and I'm 24 responsible for distributing all fixed income products 25 globally. 0067 1 MR. GALLAGHER: Spencer Gallagher, Director of 2 product management at ICE Data Services, about 25 years 3 in the industry. I have built a variety of reference 4 data databases and products. And I have also been 5 involved in the building of fixed income indices. 6 MR. DEMESY: Frederic Demesy from Refinitiv. 7 We are formerly known as the financial and risk division 8 from Thomson Reuters. So, we have just been renamed the 9 st 10 1 of October. This is quite important. I run the 11 reference data globally for Refinitiv. I am based in 12 Singapore. So, it is a long trip for me. And thanks for 13 welcoming me here. 14 And so reference data covers all our asset 15 classes or financial asset classes, fixed income to 16 equities, options, et cetera. 17 MR. MCVEY: Thank you. When we started a 18 deeper dive into this topic, one of the things that we 19 learned is that a central database does currently exist 20 for municipal bonds. And it was interesting to note that 21 the needs database that is operated by DTC has solved 22 many of these issues in municipal bonds that exist in 23 corporate bonds. 24 And, Spencer, maybe I can start with you since 25 you have had experience with both, but just describing 0068 1 some of the challenges for corporate bond reference data 2 providers today and also comparisons of where we are in 3 corporates today relative to municipal bonds? 4 MR. GALLAGHER: Sure, thanks, Rick. 5 At ICE Data Services, we initiated an 6 investment program in reference data that has been active 7 over the last several years. The result has made our 8 reference data services stronger and more competitive in 9 the market. We have expanded the use cases we serve, and 10 in the process added a fixed income index business to our 11 portfolio products. 12 The index completeness -- requirements for 13 completeness and timeliness are very extreme. And it was 14 part of that investment that we made to make sure that we 15 could serve that index environment. 16 I could go on at length about the investments 17 we have made and the subsequent accomplishments, but the 18 point here is that there is no -- there is one area that 19 no investment or no level of ingenuity can solve and that 20 is equal access to new issue reference data at or prior 21 to first trade execution. 22 The preliminary recommendation, in my 23 experience that access and timeliness to fixed income 24 reference data has a significant impact on the efficiency 25 and inter-operability of the corporate bond markets. 0069 1 This is especially apparent in new issue market. 2 We find that many of our clients required 3 important data for pre-trade activities, such as 4 compliance, as soon as the data is solidified for a 5 particular new issue. This generally occurs at the 6 pricing of the deal. 7 Pre-trade activities begin at pricing. Trades 8 occur. Trade confirmations are queued. I have looked at 9 over 100,000 U.S. securities and found that the reference 10 data was made widely available only about half of the 11 time at the event of pricing of a security. 12 The underwriters start to distribute this 13 content after pricing but this is highly inconsistent and 14 inefficient. This involves e-mails to various vendors. 15 For the market, at pricing, the clock starts 16 ticking. And the need for information is immediate. 17 Even an hour of delay is impactful and costly for many 18 corners of the market, be it trading, buy side retail, 19 and the venue community that feed all of these processes. 20 Underwriters are the primary source for this 21 information. Generally, this makes it to us via e-mail 22 from the underwriter. Again, clearly this is 23 inefficient. Distribution is not consistent in both 24 completeness of the content or timeliness of the 25 delivery. 0070 1 We endeavor to secure data directly from every 2 underwriting desk in the market. When you are successful 3 at this, there is no consistency in the bond information 4 that is disseminated by the underwriters. The timeliness 5 of submission is not standard, stats being supplied are 6 not standard. 7 We receive this information at the will of the 8 many underwriting desks. Often you will see this at the 9 end of the day on the day of pricing. The deal prices at 10 11:00 a.m. Trading has begun, and our clients, and the 11 same for any vendor's clients, same with Refinitiv, would 12 find it difficult to complete pre-trade activities. 13 We cannot tell our clients that we can depend 14 on having the data before trading begins. We can't make 15 that commitment. 16 This often leads to the buy side scrambling to 17 fulfill their pre-trade responsibilities. This is 18 clearly not contributing to straight through process in 19 the industry. Unnecessary costs in this environment 20 begin to mount. 21 The buy side and the value-added companies and 22 applications are a part of the pre-trade and trade 23 workflow have to resource for this gap and suffer from 24 trade delays or lost opportunities. 25 Conversely, in the muni market, we do not have 0071 1 this problem. We can clearly state when a reference data 2 is available on municipal new issues. The award date and 3 time is established and the data is made available prior 4 to the first execution. For municipals, new issue 5 reference data dissemination is mandated. 6 There are other sources for new issue data for 7 the vendors. There are publishing services dedicated to 8 corporate bond new issue market. They are not timely or 9 complete enough. There are issuer websites. But, again, 10 it is not timely and the trick here is you need to know 11 that that issue is coming to market at that date and 12 time. There is no comprehensive calendar that covers 13 announcements of deals set to price on a given day in the 14 corporate market. There are calendars out there but none 15 of them are comprehensive. 16 All said, none of the avenues, underwriter e- 17 mails, new issue publishing announcement or issuer 18 websites provide a comprehensive coverage in a timely 19 manner. We piece all of this together as available to 20 us. 21 On the few cases where we see no information, 22 we will see the data on Edgar, usually via prospectus. 23 But that is well after the pricing event and clearly not 24 sufficient for pre-trade and trade workflows. 25 To service a variety of clients in the variety 0072 1 of use cases, we need to be complete. Prospectus level 2 content may not be needed at the time pricing, but we do 3 need to supply content that will allow our clients to 4 perform pre-trade compliance calculation yields, transact 5 trades, provide trade reporting and queue trade 6 confirmations. So, we don't need a 100 data items to get 7 this started, but I think in this recommendation we will 8 come to agreement of what is sufficient for that initial 9 new issue. 10 I talked to various participants since I was 11 invited to participated on the panel. They were 12 immediately interested because of the gap they see for 13 reference data in newly issued securities. I would 14 expect an active and productive comment period. There 15 was no disagreement. And from the parties I talked to, I 16 did see a significant opinion that the recommendation is 17 on the right path. 18 There is a comment that was made in the July 19 FIMSAC municipal session that the municipal market has 20 good new issue disclosure and maybe that has not been 21 marketed well. The SEC, MSRB, guided some compelling 22 advances on disclosure. I mention that now in the 23 corporate discussion as these municipal efforts could be 24 a model for the corporate bond new issue dissemination. 25 I point to two particular regulations in the 0073 1 municipal market, MSRB Rule G34 and SEC Rule 15c2-12. 2 G34 required the sell side, the underwriters, to take 3 certain steps to make information available to allow 4 complete and timely transaction reporting, produce trade 5 confirmations and to aid clearance and settlement. 6 This effectively required underwriters to 7 disseminate reference data information upon pricing prior 8 to first trade execution. 9 The gap of timing in new issue information was 10 resolved as the DTC worked with regulators and market 11 participants to create what is called NIIDS, as Rick had 12 talked about. NIIDs stands for New Issue Information 13 Dissemination Service. It includes a regulated minimum 14 set of reference data that is available and distributed 15 to market participants, specifically the vendor 16 community, prior to first trade. 17 This was a positive transformation in the way 18 municipal content was made available. We re-tooled our 19 products to make sure our clients had increased access to 20 data to fit the more efficient new issue dissemination 21 and trade reporting requirements. This had a significant 22 impact on the muni market as it could now depend 23 sufficient content to execute pre-trade and trade 24 activities without a scramble to secure the required new 25 issue reference data. 0074 1 In the corporate market, there is no such 2 analogous distribution of new issue content. The 3 recommendation outlines a process for the corporate bond 4 market to coalesce around. It calls for dissemination of 5 a minimum set of content for new deals. I believe during 6 the comment period the various market participants can 7 agree on reasonable standards of disclosure and 8 distribution. I do not say this without acknowledging 9 that this could put a burden on the underwriting 10 community. The caveat is that the do this type of 11 dissemination of new deals today, it is just not 12 standardized. 13 Possibly, the centralization will work out in 14 their benefit as underwriters are distributing through 15 just one pipe instead of the multiple pipes that they do 16 today. 17 Another caveat is that often we will improve 18 the workflows for other divisions of the firm as many 19 sell side shops also have a buy side segment. 20 And when I looked at questions from our 21 clients, my experience shows there is often a client on 22 the buy side asking us for new issue details of a 23 security underwritten by their firm on the sell side. 24 So, looking at the recommendation, new issues 25 is primary. There was also a segment for future 0075 1 considerations about corporate actions, and that is where 2 SEC, the municipal rule, SEC Rule 15c2-12 comes into 3 play. In the secondary market, there is a similar gap of 4 corporate action distribution for corporate bonds. 5 Calls, bankruptcies, non-payment defaults, ratings, 6 current rates, M&A and other material events are not 7 centrally disclosed and distributed in any timely manner. 8 There are 8-Ks but you have to go through an ocean of 8-K 9 comments to filter to the really materially events. 10 Here again the municipal market may be 11 somewhere we can look for guidance. SEC Rule 15c2-12 12 requires continuing disclosure of certain important 13 events. This data is made available to market 14 participants in the vendor community. The MSRB's Emma 15 Website, which is a fantastic website, publicly displays 16 this continuing disclosure information. The corporate 17 market should also look at a more centralized, 18 standardized and timely manner of disclosing critical 19 corporate actions. 20 Continued disclosure may not be the highest 21 priority in this recommendation, but it is still critical 22 to the market. And maybe not on the top of our list now, 23 I fully agree new issues is the biggest issue. But if we 24 don't address it now, we will revisit corporate actions 25 as well. 0076 1 Thank you. 2 MR. MCVEY: Thank you, Spencer. Frederic, 3 maybe you could follow-on from the Refinitiv perspective 4 on some of the comments that Spencer made. 5 MR. DEMESY: Thank you, Spencer. I think you 6 described quite well the situation most of the vendors 7 are facing with new issues. The sourcing part, so from 8 Refinitiv's standpoint, we saw some of the data from 9 three avenues. One is contribution from our customers, 10 so the sales desk that contribute the information to us. 11 We also have sell-sourcing capacity, so we've 12 got teams that sourced information from websites or from 13 dealmakers or we have a division called IFR and a deals 14 team that is very close to the sell side market and look 15 at what new issues are coming to the market. And we 16 collection that information. 17 And there's an additional source for this data 18 is third party vendors. And so Refinitiv, we are very 19 open in working with other vendors in the market, 20 especially vendors. So, in the European markets, for 21 example, we source data from Euroclear. But globally, we 22 have agreements with other vendors to collect their 23 information and store that information on a database. 24 And obviously from the sourcing, one thing that 25 is very, very important for a fixed income market is the 0077 1 normalization of data. So, we are an international 2 vendor, and we provide information to global 3 participants. So, it is very important to have the data 4 normalized, so all of the -- all our clients globally can 5 understand the data that is available. 6 It is very, very complex in the fixed income 7 space as well as there is no rules or standards in the 8 way that prospectuses are created. And then with 9 distribution of that data to our customers, which is as 10 important as the rest, to ensure that the distribution is 11 done on time. 12 And I think I have been in this industry for 20 13 years, or 24 years actually, and we see a transformation 14 in the bond markets where in the past market participants 15 were expecting the data to be available at the end of day 16 or the timeliness was not as important as it is now. 17 Now, a market participant wants to have the information 18 when the bond prices to set up their platforms to be able 19 to trade. They want to have updates intraday, and that 20 is a very big difference from what happened maybe two, 21 three or five years ago where end of day updates was 22 enough for them to operate. Now, the market participants 23 want information intraday. And that forces market 24 vendors I think to rethink the way we distribute the 25 reference data. 0078 1 And obviously the more the bond trades 2 electronically, the more market participants would want 3 to have this information on time. And at the moment, we 4 see that there are some market anomalies where some of 5 the vendors have access to information much earlier than 6 other vendors. And that creates basically competitive 7 advantage on certain platforms, which is in my view not 8 ideal for having a transparent market. 9 It also incurs higher costs for our customers. 10 The first one would be on vendors. Market participants 11 will have to source the data from multiple vendors to 12 ensure that all the information is available, so 13 duplicating costs. 14 There is also an operational cost related in 15 terms of data quality. So, when you onboard multiple 16 feeds, ICE Data Service and Refinitiv data is not 17 automatically in the same format. So, the customer has 18 to develop operational efficiency tools to standardize 19 the data on their platform. 20 And third is when the market participant gets 21 things wrong, it can have a huge impact, missing trade 22 opportunities but also reputational risks that would be 23 the worst. 24 That's why I think either a look at the 25 recommendations, and I think the recommendation that Rick 0079 1 and the subcommittee is putting forward is extremely 2 valid, and we are very supportive of that recommendation. 3 And that's it. 4 MR. MCVEY: Thank you, Frederic. Bob, you were 5 very helpful with the subcommittee in describing the 6 underwriter process in new issue reference data. And if 7 you could please share that with the rest of the FIMSAC 8 committee. 9 MR. LOBUE: Sure. Thanks for having me. 10 I would say first and foremost what we try to 11 do is ensure the information around transactions is 12 accurate. And it is going to participants that are 13 evaluating a transaction on behalf of the issuer. 14 So, the majority of our communications are 15 targeted toward the investor community at the onset. And 16 what we want to ensure we accomplish, and one of the big 17 sensitivities of the street is that we don't have 18 multiple constituents communicating that information 19 while we are undertaking the distribution. Inaccuracies 20 along that way would likely complicate the financing for 21 the borrower. So, we are very sensitive to how the 22 information goes out, that it is compliant with the 23 various regulatory regimes around the globe, that it is 24 consistent and accurate. So that would be my first 25 comment. 0080 1 As we go through the process, when we talk 2 about the dissemination of information at the end, so 3 let's call the end pricing where I think all of the 4 trading venues, the compliance groups, other vendors, 5 there is a desire for that information. 6 But I want to just highlight what our focus is 7 because one thing was mentioned, pricing at 11:00 a.m., 8 it is not actually when we price deals. So, we normally 9 price deals somewhere between 3:00 and 5:00 p.m. We have 10 a 15-minute window post-pricing to deliver the pricing 11 information of FINRA for trace eligibility. 12 And we could talk about, Rick, while we are 13 delivering to FINRA, I think both FINRA and ourselves 14 would say we could probably populate that a little bit 15 deeper. And I think that would help achieve one of the 16 agendas of the board. So, we have that 15-minute window. 17 The second item that we need to do is ensure it 18 is set up in trading venues around the globe, and whether 19 that is OMS systems for investor clients. But as 20 importantly, the trading venues that the dealer community 21 uses, Bloomberg obviously is one of those significant 22 venues. 23 And I think the Refinitiv team and the ICE team 24 intimating a competitive advantage for Bloomberg, there 25 is no question that we do undertake getting our 0081 1 securities set up on the Bloomberg trading platform 2 because that is what the industry predominately uses to 3 book our tickets. We do need to get our tickets in by 4 the end of the day. And usually by 6:00 p.m., we can no 5 longer get tickets into the various trading venues. And 6 that is 6:00 p.m. Eastern. That would complicate our 7 risk positions, our VAR reporting from a regulatory 8 standpoint at the end of the day. 9 So, our priority, post-pricing, is to 10 communicate to the regulatory authority, to communicate 11 to the trading venues the data that people need to 12 provide the transactional information. 13 We do undertake some communications, and 14 various dealers do it differently. I can comment on JP 15 Morgan. We tend to not disseminate data to third party 16 vendors off the corporate platform. I think the point of 17 inaccuracies is the reason for that. So, we tend to use 18 Bloomberg as our let's ensure it is accurate, and then 19 people can source that information from that venue. My 20 fear is that everything is so manual that doing it ad hoc 21 is not a good process. 22 And then we also, the last step is once we have 23 a legal term sheet of an offering, we disseminate that to 24 all clients, investor clients. I guess I should be 25 clear. And we cannot make any communication until that 0082 1 is signed off by the issuer community, the legal 2 community and the underwriters of that transaction. That 3 is usually about an hour after pricing. 4 So, if you think about that two-hour time 5 period, we are doing a lot with data. But it is targeted 6 toward the buyers of the transaction, the underwriters 7 and the regulators. 8 MR. MCVEY: Great. Thanks, Bob. And, Alex, 9 maybe we could turn to you for just the investor view on 10 how you collect new issue data today and how it feeds 11 into the systems at T. Rowe Price. 12 MR. SEDGWICK: Absolutely. And I think, Bob, 13 you kind of articulated it and set this up very well for 14 me to kind of describe our desk and how T. Rowe Price 15 consumes and what we ultimately do with a lot of the data 16 related to new issues. 17 So, one of the things that T. Rowe Price has is 18 we have actually built a proprietary piece of software 19 that manages the information flow on new issues from the 20 desk to our portfolio managers and also to our research 21 analysts. 22 And there are a variety of reasons why that is 23 very helpful in terms of keeping all of the information 24 about pricing, all of the updates on the deal kind of in 25 one place where everybody who is interested in that deal 0083 1 can see it. 2 I think one thing that I would point out is 3 there is one key kind of manual process in the entire 4 daisy-chain of events once you get off the desk and that 5 is really entering the information into that system. So, 6 typically, what we do is we have somebody on the desk who 7 takes the information, usually off of Bloomberg and hand 8 enters it into that product. 9 So, when I think about what are the things, 10 particularly in this proposal that would be very, very 11 helpful. You know, as part of my role looking at 12 electronic trading and technology on the trading desk, I 13 am also thinking about things like operational risk, how 14 do we minimize that. 15 So, to the extent that we could get accurate 16 information that is distributed in electronic format 17 where you could consume it into the type of product that 18 we've built, we now take at this point a manual process 19 and replace it with something more automatic. 20 I also think it is important to think about 21 some of the other areas in which this information is 22 being consumed ultimately constitutes sort of secondary 23 dependencies on the trading desk. So, the things that I 24 would point to would be how we access liquidity. I would 25 reiterate some of the comments made about secondary 0084 1 trading systems. 2 Historically, we have noticed cases where a new 3 issue does take time to get set up on some of our 4 electronic trading platforms, and that means that we 5 can't necessarily go and use those electronic trading 6 platforms right away. So, we have to trade them via 7 voice or another venue. 8 I think the other thing to think about is we 9 have increasingly seen electronic market-makers become 10 important sources of liquidity on our trading desk. 11 Electronic market-makers ultimately need this information 12 to provide accurate pricing and accurate valuation for 13 the prices that they are pushing out to the market. If 14 this information is not available, that ultimately means 15 that there are liquidity providers that may not be able 16 to provide liquidity to us when those new issues are free 17 to trade. So, that's another kind of secondary 18 dependency that we think about. 19 The final piece, back to kind of the discussion 20 around evaluated pricing is another area that I spend 21 time focused on is thinking about not only calculating 22 NAVs for our mutual funds and the securities that we have 23 in our portfolios, but also, I would go a step further in 24 looking at things like trade cost analysis. 25 So, when we are going, and we are trading on 0085 1 the desk, we need to be able to measure our execution 2 against benchmarks. If it takes more than a couple of 3 hours or even more than a day for those benchmarks to 4 become available, that is an area where we may not be 5 able to do accurate trade cost analysis. And that is a 6 very important sort of supporting piece of information as 7 we think about best execution on the trading desk. 8 So, once again, I think from our perspective, 9 we are supportive of the proposal. Our focus is 10 primarily on the automated delivery of accurate and 11 timely data and ultimately minimizing secondary 12 dependencies on the desk. 13 MR. MCVEY: Great. Thanks, Alex. And, Ola, 14 maybe I could turn to you, you were also helpful in the 15 subcommittee understanding of the current process. And 16 the reason that you are here is that we learned along 17 that journey that the underwriters do have a requirement 18 to make sure that new bonds are set up within 15 minutes 19 of trading starting with FINRA for TRACE reporting. So, 20 maybe you could talk about that process and your thoughts 21 on the recommendation. 22 MR. PERSSON: Happy to, of course. So, TRACE 23 is obviously a transaction reporting facility, as I think 24 everybody is aware. But in order to enable the trade 25 reporting to begin with the security needs to be set up 0086 1 on TRACE so we are ready to receive the reports. 2 So, to accomplish that, we have one of the 3 rules as part of the TRACE is something called obligation 4 to provide notice, Rule 67-60, that in essence says that 5 an underwriter has an obligation to give us basic details 6 of that security prior to execution of the first 7 transaction. 8 It is really geared towards enabling TRACE 9 reporting. So, it is a limited set of fields that we 10 have. And enough -- the information that we collect 11 really falls into one of three categories: One is to 12 identify the firm and the role in the deal for that firm. 13 Then we have information we collect so we know how to 14 properly disseminate that bond. So, we know which feed 15 to put it out on. We know which caps to put on it. And 16 then we have information that we collect that is 17 fundamentally helping us identifying the bond. 18 We have tried to minimize the amount of 19 required fields since the most important thing is that 20 the security gets on the system. So there are a handful 21 of required fields, and then there are some other 22 optional fields that are on there. 23 Looking at the delta that we currently take in 24 versus what the proposal has, it is about 10 fields that 25 is a delta. And, as I said, some of the other ones that 0087 1 are part of the proposal may be optional on TRACE today. 2 But the delta is not huge from a number's perspective 3 today. 4 So, and then we distribute this information to 5 two constituents really. It is a limited set of 6 information we distribute, but it goes to the TRACE 7 reporting firm so they can themselves set this up for 8 TRACE reporting. And it also goes to subscribers of the 9 data feed so they can consume the feeds on there. 10 MR. MCVEY: Great. And as a result of the work 11 that FINRA is doing, the subcommittee felt that FINRA was 12 best positioned to expand the set of fields to provide a 13 central database for corporate bond new issues. But 14 maybe all of you can follow on by saying whether you and 15 your colleagues share that view as well. And whether you 16 thought at all about the potential costs involved in 17 building such a database. 18 MR. PERSSON: So, starting with the 19 feasibility, it is -- I think we would take a look -- I 20 think the proposal is a very sound proposal. And one of 21 the things I think is also mentioned, and maybe go out 22 and talk to different user groups of reference data to 23 Alex's point, some primary and some secondary. And 24 understand if there is a bigger scope or if this would 25 satisfy a majority of the constituents that would consume 0088 1 it. 2 Speaking for FINRA, not the effort on behalf of 3 the underwriters, but speaking for FINRA, we would have 4 some work to do. The technology today does not lend 5 itself very well to this. We would need to create the 6 ability for underwriters to come in, give us partial 7 information and have the ability to edit their own 8 records, et cetera. Today, that is a -- as I said, it is 9 a bit of a one-way street. It is set up on TRACE and 10 anything that changes from there, we either source from a 11 vendor or the underwriter calls us up to correct it. So, 12 we would need to do that. 13 We would also need to create a separate 14 distribution channel for this. And the reason being, 15 today, since the only thing that really matters is that 16 the security gets on TRACE, we actually do have contracts 17 with vendors that allows us to take certain records or 18 certain elements of records and incorporate those into 19 the database and distribute that. That also explains 20 where we can only today grant very limited usage rights 21 to the data we distribute. 22 So, this would have to be a service that would 23 be a service that would be entirely sourced from 24 underwriters we know common link vendor data, and then we 25 would have to build that obviously, the amounts of 0089 1 fields. 2 I think one thing to consider, depending on how 3 many fields we end up with, there may still -- obviously 4 timeliness of TRACE reporting can't be compromised. So, 5 if there is a larger set of fields, it is potential we 6 can create a structure where, similar to NIIDS actually, 7 certain trade eligibility fields are set up on a very 8 timely basis but then complemented later with more 9 descriptive fields if that were to make sense. 10 MR. MCVEY: And, Bob, you had given that a lot 11 of thought as well in terms of the right set of fields. 12 And the subcommittee felt that it was appropriate with 13 the information that we had to recommend a set of fields 14 for final deliberation with the SEC and ultimately FINRA, 15 which we assume then would be available during a comment 16 period for all market participants to weigh in. 17 But, Bob, you had given this a lot of thought 18 in terms of the fields that would be required and what 19 would also be practical for the underwriters to be able 20 to provide. 21 MR. LOBUE: Sure. And FINRA does incorporate 22 many of them today. There are probably some additional 23 fields, and I have reviewed what you put out in your 24 pack, that would assist in delivering some of the things 25 that Alex talks about, ensuring accuracy on the 0090 1 calculation of price and being able to value a security. 2 I think some of the fields that FINRA does not have today 3 would be additive to that agenda item. And I don't think 4 would cause a big stir with the underwriting community in 5 providing detailed pricing information, which we are 6 providing to investor clients today to trading venues and 7 a broader set of providers. 8 I think the big question that the underwriting 9 community is going to have is as that list expands, where 10 is there issuer confidentiality that is breached or an 11 inappropriateness or intellectual property from the 12 underwriter side where it is detrimental to their 13 business advantage or insight to borrowers. So, I think 14 that is one consideration for the committee. 15 The second would be to whom you are 16 disseminating the data. Where is too far? When we think 17 of munis, retail is an extraordinarily large component of 18 distribution. In corporates, it is not. The media. We 19 have some restrictions on ensuring we don't trigger a 20 need for disclosure to the SEC. If we are unsold on 21 certain positions, which does happen, we have some 22 restrictions on 144As in terms of being able to 23 disseminate and comment. 24 And so to whom we disseminate would be a second 25 consideration. But when we think the populace of data 0091 1 fields, I think we could work with the regulatory 2 community as well as the market participant community to 3 better round out that set of data fields. 4 MR. MCVEY: Great. Thank you. And back to 5 Spencer and Frederic with respect to the recommended 6 fields. Do you think that that creates a valid starting 7 point? And with respect to the experience with munis and 8 NIIDs, do you see similarities? Are there lessons to be 9 learned there from the fields that are required to be 10 delivered to DTC? 11 MR. GALLAGHER: Right. Clearly, this is a good 12 time for a content period to go through the fields. I 13 looked at the fields. They were fairly sufficient of 14 what we need to produce for clients. One of the things 15 that was missing, but it was actually in the text was 16 issue amount. That is a pretty basic field that we would 17 add. 18 We talked about some sort of debt ranking, 19 junior lien, senior lien, things of that sort that may 20 help with some risk issues as well. 21 So, we do want to go through it you know quite 22 carefully. It is not as expansive as needs by any means. 23 It is a smaller set but, again, we are not asking for 24 prospectus-level content at this point. 25 MR. DEMESY: I haven't seen the list of fields 0092 1 yet. But I would basically be aligned with ICE, with 2 Spencer's comments. We do not need -- I don't think we 3 need the full terms and conditions to be able to set up 4 an instrument on a trading platform. And we have to 5 consider also the challenge on the issuer side to be able 6 to provide the information. So, we have to find a 7 balance between the two. 8 Once I have got the list of fields, I am going 9 to review that with my team in the data operations. And 10 see what is the complexity for us to be able to process 11 these fields. And if they have got any other 12 recommendation as well. 13 MR. GALLAGHER: I just wanted to add one more 14 thing. ICE Data Services, as well as Refinitiv and 15 probably other data vendors do service a lot of third- 16 party providers of content, value-added content. And 17 there are software or applications that do things like 18 risk compliance. And their needs would want to be met as 19 well. 20 A standard use case that I have had to deal 21 with recently is a trading desk that is trying to run 22 pre-trade compliance on a security without the reference 23 data that they need to run that. And, of course, they 24 have to pass on that security because they don't have 25 enough content to do that. 0093 1 So, those types of applications that are part 2 of the pre-trade process, we definitely want to make sure 3 their opinions are brought into this. I don't think they 4 are going to ask for a lot of content. I have seen their 5 list, and it is actually quite skinny. But I just want 6 to make sure that those are addressed as well. 7 MR. MCVEY: Great. Just one more question from 8 me, and then we will open it up to other committee 9 members with any questions that you may have. 10 But, in your opinions, are the challenges with 11 corporate bond reference data largely in new issues today 12 or are there also challenges in seasoned issues? 13 MR. GALLAGHER: From my perspective, new issues 14 is the number one challenge. Seasoned issues are a 15 challenge. I did mention that in my comments that we put 16 corporate actions as a secondary requirement. 17 It is a global problem. So, I mean we are 18 talking about the U.S. market here. I am going to have 19 these problems continuously, even if we do solve this in 20 the U.S. 21 But that time frame of getting a security up on 22 a new issue is what our clients are focused on. We do 23 service electronic trading platforms of course, and that 24 was one of the issues mentioned. They are plumbed into 25 our content. And if they are not seeing that content, 0094 1 well, that bond is not available to trade. 2 So, that new issue, part of this is the number 3 one issue as I would see it. I do think corporate 4 actions are very important, especially when I think about 5 say the index use case. And you have criteria whether a 6 bond is included in an index or not. Issue amounts, if 7 there are paydowns or taps or something on that security, 8 that is very important whether that bond will be included 9 as constituent in the index. And that could be 10 detrimental if that information is late on to the able as 11 well. But I will stick with it, new issues are our 12 number one problem. 13 MR. DEMESY: Same comments. New issues are the 14 most important. Corporate action is definitely also a 15 big challenge for our customers, especially in the index 16 -- index space. 17 One other thing, so if I may comment also on 18 the recommendations. We have also to be conscious about 19 the limitation of the recommendations that are put 20 forward, especially if the issuer, U.S. issuer is issuing 21 the instrument abroad. I think that is also a point to 22 take into account if the issue is issued in, for example, 23 in the Cayman Islands, for example. The regulations that 24 or the recommendation put forward won't solve the 25 problems for the new issues. That is also something to 0095 1 take into account I think in the recommendations. 2 MR. LOBUE: The only thing I would mention on 3 seasoned issues is that there are a lot of errors in the 4 data quality outstanding. And many of the fields that 5 some of the data providers, and more importantly 6 probably, index providers, many of them are opinion- 7 based. 8 I will use the country of jurisdiction as an 9 example there where you may have an issuer domiciled in 10 one place but the subsidiary is issuing here. And you go 11 down the list, and it seems that many of the data fields 12 are randomly chosen. And we need to be careful that we 13 don't put the underwriting community in a position to 14 have to opine on things. Once again, we tend to view it 15 as the investor needs to make the right decisions for 16 their clients. We want to ensure our information is 17 accurate for them to make those decisions. 18 But I always get fearful of asking the 19 underwriters to fill in every field. The data providers 20 then provide it. And there is some subjective license in 21 there that we should avoid. 22 MR. GALLAGHER: I will second that. Countries 23 of risk is a field that has subjectivity to it. We would 24 not expect the underwriters to come with a clear 25 statement on that. 0096 1 MR. DEMESY: Countries of risk comes up quite 2 often in discussion with customers. And our position at 3 Refinitiv is to have complete transparency on how 4 countries of risk is derived. And there are multiple 5 models. There is also obligation depending on the 6 jurisdictions that a bond trades in to take into account. 7 So, it is a very complex subject. And our view is really 8 to provide transparency so that the investors or our 9 customers can really demonstrate the countries of risk 10 used is in line with what can be expected. 11 MR. MCVEY: Great. Carole? 12 MS. BROWN: Thank you. And thank for being 13 here today. I was fascinated by this because the muni 14 market is so different in terms of reference data. And 15 it led to two questions that I had. The first is part of 16 the reference data, and Spencer alluded to it with 15c2- 17 12 and the reference data that is NIIDS has a credit 18 component to it and that munis report. 19 And I was wondering if you considered that at 20 all for the corporate recommendations including a credit 21 component, whether it is rating or other metrics? And 22 the second question is what do you view as the issuer 23 responsibility in this or is it all borne with the 24 underwriter? Again, in the muni market it is a little 25 different, the issuer bears responsibility really for the 0097 1 15c2-12 disclosures. And is very hands-on in terms of 2 the reference data as it is disseminated to the market. 3 MR. PERSSON: Thank you. If I can address 4 that. So, when an underwriter comes in to set up 5 security for TRACE reporting, that actually matters to us 6 if it has a Moody's rating or S&P rating, what that is 7 because that rating will determine if the security is 8 considered an investment grade or high yield, and that in 9 turn affects the dissemination cap. 10 So, we ask for that, but we cannot transmit 11 Moody's and S&P ratings to third-parties. Our contracts 12 simply don't allow us to do that. But we use that 13 information to provide the grade for TRACE purposes. 14 And to your second question, the issuer, so 15 FINRA's jurisdiction is the broker-dealer community, so 16 the underwriter. So, we don't have a nexus with issuers. 17 So, any obligation that we create would be an obligation 18 on behalf of our members. 19 MR. LOBUE: One additional comment on the 20 issuers. Many of the transactions in the corporate 21 market are SEC registered, so the pricing details would 22 be filed with the exchange post-pricing by the issuer and 23 their legal representative. 24 So, the real question would be the 144A 25 securities where we do not have that, difficult for the 0098 1 issuer community unless the exchange changes their rules 2 to disseminate that data in like way to the SEC register. 3 I think it is very difficult for the issuer community to 4 have the responsibility to connect with all market 5 participants because they just don't have that 6 connectivity. So, I do think the burden does lay with 7 the underwriters or the legal community or the regulatory 8 community. 9 MR. MCVEY: Larry? 10 MR. TABB: Thank. A question about the 11 process, I guess more to Bob and to Ola. How much of 12 this is kind of input multiple times or this kind of, you 13 know, there are platforms like IPREO or you have to 14 register with CUSIP or you talk about the credit rating 15 agencies? Is this something that that kind of flow 16 straight through from those guys or is this all "sneaker 17 net" and everybody has to kind of input things to 12 18 different screens? 19 MR. LOBUE: Sure. Current process is fairly 20 horrific and antiquated. We have multiple underwriters 21 on each transaction. The underwriter that purchases the 22 securities from the issuer is the one that would 23 disseminate to FINRA. And that varies. In many of the 24 corporate transactions, if we raise $50 billion for 25 someone, you are likely going to have 10 or 12 individual 0099 1 securities, and you may have four or five banks 2 disseminating that data just to the regulatory authority. 3 We do have an endeavor undergoing with 10 banks 4 right now that we hope to turn into an industry protocol 5 that I think is the right venue to connect with a 6 regulatory body to standardize and disseminate 7 electronically. 8 I don't know that using a third-party mechanism 9 is solution today because each dealer has its own venue 10 or technology, some is in-house, as T. Rowe highlights, 11 ours is in-house. Many use third-party vendors, like 12 IPREO. IPREO tends to have more of a monopolistic 13 presence in the municipal market, which is why I think 14 you use that tool to connect with the MSRB 15 communications. And credit is much more disseminated 16 than that. 17 But I do think there are items on the horizon 18 in early 2019 that could provide us that efficiency. And 19 if we could -- once again we are at 10 now, if we could 20 ensure that the industry used this as an acceptable 21 protocol, I think it would create for some of these 22 issues. 23 MR. MCEVY: Larry? 24 MR. HARRIS: The object of our recommendation 25 is the creation of machine-readable, reference data that 0100 1 are readily available to everybody and accurate and 2 timely. 3 Our proposal focuses on FINRA because it was 4 our understanding that FINRA is best equipped to solve 5 this problem. But we can easily imagine other ways of 6 obtaining these data. 7 So, for instance, the SEC could ask issuers to 8 be responsible for reporting into a system like Edgar or 9 some other similar system. It's not a particularly 10 difficult problem. It is a limited number of fields. 11 And you don't even have to edit the past reports. You 12 could just produce a new report every time. And any 13 vendor then could extract these data. 14 So, the question to you all is what would be 15 your preference? Would you like to see FINRA do this or 16 would you like to see it reported directly to the SEC? 17 And consider also that, as Ola has said, FINRA has 18 limited jurisdiction. Presumably, maybe this should lay, 19 the responsibility should lay perhaps with the issuers, 20 which they would then delegate to their lead underwriters 21 in solving the problem. 22 So, there are different ways of doing this, and 23 my question is are we doing it the best way? 24 MR. LOBUE: I will pick up the grenade. So, 25 one complexity is that I think putting the burden on the 0101 1 issuer community is quite challenging. 2 We are sitting here talking about the U.S. 3 market predominately when we talk about the SEC. I run a 4 global business where each regulator has different 5 requirements. We have issuers across the globe. In many 6 cases, we are pricing a security at 5:00 p.m. It is 4:00 7 in the morning for the issuer client. So, to me, I don't 8 envision the issuer community as being the right 9 community. 10 We have the additional complexity of we are 11 dealing with one regulatory authority here. We want to 12 create, when I say "we," the underwriting community. I 13 will talk on behalf of us even though they haven't 14 authorized me. Would like to standardize a practice that 15 works from a SEC standpoint. It works from an FCA 16 standpoint and for every jurisdiction on the globe that 17 creates sound market environments for our clients. And 18 when I say "our client," I tend think more around 19 investors, the dealer community and issuers. Ensuring a 20 sound and liquid stable market. 21 It probably puts the burden on the underwriters 22 to work with each regulatory authority, and hopefully we 23 can create consistency. I don't think this is so topical 24 that there should be multiple views as to sharing a level 25 of detail around security information and creating 0102 1 centralized databases. But I think the burden should be 2 with the underwriting community. 3 The question is what is the best way to 4 disseminate it? I don't know that I have the right 5 answer there, whether it is the dealer community 6 maintaining the database as an industry, not as a firm, 7 and making it accessible. That might be one alternative. 8 Or do we put that burden of maintenance and accessibility 9 on the various jurisdictional regulatory bodies? But I 10 find it very difficult to not have the underwriting 11 community with the responsibility of we are creating the 12 data, delivering the data in a standardized way. 13 MR. HARRIS: And I can't imagine why the 14 underwriting community wouldn't be delegated by the 15 issuers if the burden was with the issuers. I think the 16 issue is really a regulatory requirements issue. And if 17 the issuers are required, it will go the underwriters 18 anyway, so. 19 MR. LOBUE: Right. 20 MR. HARRIS: So, I don't really care about 21 that. 22 MR. GALLAGHER: Another comment on that. If 23 you look at the muni regs, it is not a monolith. So, it 24 is not one reg. There's G34 and 15c2-12, which do have, 25 to Carole's point, 15c2-12 is of the issuers, and that is 0103 1 the continuing disclosure of a security which the 2 underwriters may not be a part of at that point in time. 3 When we are talking new issues, I could not 4 disagree with Bob at all, that it is the underwriter 5 community that would probably be the most likely supplier 6 of that information. 7 MR. MCVEY: Scott and Suzanne on the phone, we 8 realize you are at a disadvantage about being able to put 9 your name card up, but if you do have any questions or 10 comments, especially you, Scott, from the issuer's 11 perspective, feel free to speak up. 12 Sonali? 13 MR. KROHN: No disagreements to what Bob has 14 said. 15 MS. THEISEN: Great, thank you. So, look, I 16 think all of the comments -- I wholeheartedly agree with 17 everything that Bob and Spencer and Frederic have raised 18 around balancing the robustness of the data with it being 19 objective, structured, succinct. 20 And so my question is really for Ola. If we 21 were to expand the list, it sounds like roughly 10 22 additional fields, give or take. And you had mentioned 23 today you have this issue of commingling data with if you 24 had to then go strip it out. If FINRA was the vehicle 25 for dissemination, provided that that data arrived at 0104 1 FINRA at again in a structured format via ATI, what do 2 you think would be the latency, if any, of then 3 disseminating that? 4 MR. PERSSON: So, in terms of completeness of 5 the new issuance, that is in really good shape today, 6 what we get from the underwriters. There are 7 occasionally actually when we get an alert from a vendor 8 that a CUSIP has shown up on the vendor fees that we 9 still didn't get from the underwriter, very rare 10 instances today. It is very few and far between. 11 We do complement some of the facts on the 12 record with vendor data if we didn't get it from the 13 underwriter. So, if we were to create the service, it 14 would be exclusively underwriter data. So, clearly, what 15 some of the fields we consider optional today couldn't be 16 considered optional going forward. So, it would increase 17 the burden a little bit on the underwriters for sure. 18 As I said, the question is how many fields we 19 land up with. And if the amounts of fields would 20 jeopardize the timeliness of the reporting, that would be 21 a problem. So, then we would have to think of it maybe 22 in two tiers. 23 Having said that, personally it doesn't look to 24 me like the delta is big enough that it would create an 25 issue, but I don't want to pre-judge what the outcome is. 0105 1 MR. LOBUE: One comment on that. The current 2 process is it is rare that I am keying in the data to 3 FINRA. So, usually I am communicating with someone 4 internally at JP Morgan, or any other bank, to someone 5 who has connectivity to the FINRA system. Munis, I think 6 is two hours. That feels very long to me. I don't think 7 the data fields should cause us to elongate that, but I 8 would like to have a conversation with FINRA as to 9 whether we can comply in the 15-minute window, depending 10 on the length of the data. 11 MS. THEISEN: I would fully agree. And I think 12 you know within the subcommittee we had robust debate 13 around the actual fields for this very reason of making 14 sure that we balance objectives that we have all stated 15 with making sure that we can book trades and do 16 everything that Alex mentioned in a timely fashion. So, 17 you know, it would certainly be our view that proposal as 18 it is, unless there is glaring omissions, that we 19 continue to try to keep that list as succinct as possible 20 to satisfy these objectives. 21 MR. McVEY: Elisse? 22 MS. WALTER: Thank you. Bob mentioned before 23 difficulties in dissemination presented by the 24 restrictions of 144A. I think, as you know, from time to 25 time over actually much longer than people know, the SEC 0106 1 has thought about alleviating some of those restrictions 2 through looking at the private placement offering 3 exemption and perhaps failing to restrict, deciding not 4 to restrict offerees but only to restrict purchasers. 5 And, putting aside the question of whether the SEC can do 6 that alone or there is any need for legislation, how 7 important do you think that is to the functioning of the 8 marketplace? Is it important? And is it something that 9 you think this Committee should put on its agenda as 10 something to look at? And not just Bob, but it was 11 directly related to what you had to say. 12 MR. LoBUE: Yes. I will make a very light 13 comment. There was no question we disseminate that data 14 much more narrowly today. So if I had to guess, it is 15 more challenging and I will defer to my peers for 16 them to source that information. 17 Second comment would be there is no question it 18 is much more difficult for the secondary market to source 19 those documents. So we only disseminate the final 20 prospectuses to purchasers today. Those are private 21 documents. And so if you think about an SEC-registered 22 security, a secondary market participant could source 23 that information. Even though they didn't participate in 24 the primary distribution, there is clearly an 25 inefficiency of information amongst 144A securities. 0107 1 MR. GALLAGHER: I clearly second that. It is a 2 problem. How much of a problem, it depends because the 3 interest should be also as narrow as the distribution is. 4 And it is something to consider. I am not sure if I am 5 well-versed in all the legalities of distributing a lot 6 of private placement securities. I do know that other 7 vendors have knocked on our door and asked us for our 8 completion of private placement securities as well 9 because they have a gap in their population. We have a 10 gap. I am sure you have a gap. It is consistent across 11 the industry. And we find it very hard to source that 12 information. 13 MR. DEMESY: Just a compliment on the private 14 placement. It is also another topic of the Committee's 15 around pricing. So we see because we don't have access 16 to these documents. So it is very difficult for us to 17 provide evaluative pricing for our customers. And that 18 is a big, big challenge that we face on these types of 19 securities and the market, actually, faces on these types 20 of securities, actually. 21 MR. GALLAGHER: Yes. It is a manual effort. 22 When a client comes in to us and holds a private 23 placement security and they want us to evaluate it we 24 have an evaluative pricing business we often have to 25 use that holder of the security to go back and get that, 0108 1 help source that information for us. So it is a problem. 2 MR. PERSSON: Just so everybody is clear, 144A 3 securities are reportable to TRACE and go through exactly 4 the same process that we described here today. So unless 5 there are any sort of restrictions for us including that 6 information in this type of service we are looking at, 7 then they are already covered under TRACE. 8 MR. McVEY: Rachel? 9 MS. WILSON: And also an issuer perspective. 10 You know, living that post-3 o'clock time, it is usually 11 a rushed one. Okay? And so, you know, sometimes, Bob, 12 we don't agree on price. Right? And we are going to be 13 battling over an eighth as an issuer and with our 14 underwriters or whatever it is. Right? So that is a 15 compressed period. 16 And I know, Sonali, you were actually raising 17 this kind of same point. How many metrics can we really, 18 you know, get? On this list as an underwriter, when I 19 look through it and schedule A also. I mean, it is a 20 no-brainer. These are not hard. Okay? They are not 21 expensive for us to certainly know or produce and 22 certainly for the underwriters. 23 The issue to me is much more what would be 24 those kind of tier I most important in keeping the 25 discipline that this is a very focused list. So to the 0109 1 very, very minimum of what would be needed, you know, 2 looking, you know, to Spencer and Frederick of your 3 constituents to be able to trade, but, again, I think 4 this that time is very precious and you want to often 5 get that first trade off. So, you know, that is a 6 challenging dynamic. 7 So I just, you know, would mark a little 8 caution that it is not so easy. Those are very precious 9 moments. 10 MR. HEANEY: Larry? 11 MR. HARRIS: Reference data has two purposes in 12 our market. We spent most of our time talking about the 13 first purpose, which is to allow people to reliably 14 settle trades so that they know what they are trading and 15 can figure out the final payments and so forth. 16 The second purpose is that it allows people to 17 value bonds. And this is an extremely important purpose 18 because dealers and investors need to know what a bond is 19 worth so that they can confidently trade it. When they 20 can confidently trade a bond, then they will bid more for 21 bonds because there is less uncertainty and capital 22 formation overall in our economy improves and issuers are 23 able to issue at lower cost. So, to that end, ultimately 24 we would like to have readily accessible, cheap reference 25 data that is accurate and so that everybody who is in the 0110 1 business of valuing bonds, for whatever purpose, can do 2 so. Anything that we can do to lower the cost of valuing 3 bonds will make these markets more efficient and more 4 beneficial to the economy as a whole. 5 The information necessary to value bonds 6 includes all that we have discussed but also includes 7 other information that is not particularly expensive to 8 acquire; so, for instance, the time of the first call, 9 the terms of that call with the conversion and so forth. 10 So these information are readily available by just 11 simply asking investors how they value bonds. What do 12 they need? 13 So here is the question to Ola. Your 14 organization has primarily been focused on the first 15 purpose and, in particular, just making sure that the 16 TRACE is going to be useful for reporting purposes. 17 FINRA, of course, is a very capable organization. And 18 the question is, do you feel confident that FINRA could 19 act as a faithful agent to the second purpose, which is 20 to ensure that these data are collected? It doesn't have 21 to be quite as timely, though most of this data could be 22 delivered equally quickly. So, again, are you 23 comfortable with that additional mantel? 24 MR. PERSSON: So I understand your question 25 correctly, you are saying collect a deeper dataset here? 0111 1 For example, your example is a call feature. So if it 2 has a call feature, it would collect the entire call 3 schedule, et cetera, et cetera, so a deeper set than was 4 being asked here? 5 MR. HARRIS: That is correct. 6 MR. PERSSON: Yes. I think we would simply 7 follow the recommendation that this Committee would make 8 and evaluate. So I think we would probably try to stay 9 within scope on the recommendation. I think it comes to 10 Sonali's point that everybody who has worked their 11 reference data can easily envision another 20-30 fields 12 that would probably be nice to have but would 13 significantly increase the burden on the underwriters 14 providing this information. So I think we would simply 15 take a look at the recommendation and obviously am happy 16 to be working with the Committee and the SEC on what the 17 scope would be but probably would stay within that scope. 18 MR. HARRIS: So I think it is really important 19 to recognize here that the burden, though it might be 20 twice as large, is still extremely small and very, very 21 small in comparison to the value of these data. So in 22 our subcommittee deliberations, we specifically decided 23 that we did not want to be the ultimate arbitrator of the 24 content on the list that appears in appendix A. And I 25 think it is very important that when that list is 0112 1 ultimately determined, that we have investors, 2 sophisticated investors, who are valuing bonds help us 3 understand what data should be included. 4 MR. McVEY: Thanks, Larry. 5 Mike, I will turn it back to you, I think. 6 MR. HEANEY: Thank you. 7 I just want to follow up, actually, on that 8 last point. I guess it is back to you, right? The 9 subcommittee discussed, debated schedule A at length, 10 came up with this set of criteria or inputs, to Sonali's 11 point and to Larry's, thinking that going any further 12 and it is a question going any further would be 13 burdensome perhaps to making this process work. Is that 14 correct? That is my understanding, having been on the 15 subcommittee, that this actually was the criteria, 16 subject to a common period, subject to input from 17 institutional investors, but this was the thought going 18 forward. 19 MR. McVEY: We did put considerable work into 20 the list and investor and underwriter views both 21 represented in the list that we presented. But we did 22 want to open up for other opinions as to move forward. 23 MR. HEANEY: Understood. Okay. 24 MR. HARRIS: My clear understanding was that 25 the list was not final. 0113 1 MR. HEANEY: Any other questions, comments for 2 the panel for Rick? 3 (No response.) 4 MR. HEANEY: Again, a very good discussion, 5 good input from the panel. So thank you very much. 6 I think, again, at this point, we probably have 7 gone through this process in great enough detail and I 8 think enough Q&A to entertain a motion to vote on the 9 recommendation if I can. Thank you. Again, same format. 10 If I could ask all voting members to raise their hands 11 that are in favor of the recommendation? 12 (Show of hands.) 13 MR. HEANEY: Suzanne? 14 MS. SHANK: In favor. 15 MR. HEANEY: Thank you. 16 Scott? 17 MR. KROHN: I support. 18 MR. HEANEY: Any voting members opposed, please 19 raise their hands. 20 (No response.) 21 MR. HEANEY: Any voting members abstaining? 22 (No response.) 23 MR. HEANEY: Okay. Again, I would like to 24 congratulate the subcommittee on an incredible amount of 25 good, hard work, progress. The recommendation has been 0114 1 approved unanimously by the Committee. Rick and team, as 2 I have said to the other subcommittees, this year has 3 been a marathon, not a sprint. This is some of the 4 culmination of many months of work. So thanks for all 5 your hard work and to the subcommittee as well. 6 At this point, unless there are other comments 7 or questions, we are going to break early for lunch and 8 will resume back here at 1:30. Thank you. 9 (Whereupon, at 12:20 p.m., a luncheon recess 10 was taken.) 11 A F T E R N O O N S E S S I O N 12 MR. HEANEY: Why don't we try to get seated 13 I know there are some people still outside just to 14 keep us on time. All right. So we still have a pretty 15 full afternoon. Welcome back, everybody, from the lunch 16 break. We will start the afternoon session by hearing 17 the updates from the transparency subcommittees, and we 18 will start with Mihir on the update from the Corporate 19 Bond Transparency Subcommittee. 20 Just I'm sorry. One housekeeping item I have 21 been asked to mention. If you are talking in the mike, 22 the people who are creating the transcripts that get sent 23 out publicly are having trouble hearing some people 24 around the table. So just if you could make sure to have 25 the microphone close? 0115 1 MR. WORAH: Thanks, Michael. 2 So quick update from me on where the Corporate 3 Bond Transparency Committee has been working on for the 4 last couple of months. The topic we focused on was pre- 5 trade transparency, seeing if there were any 6 recommendations we could come up with that would help 7 liquidity and functioning in the corporate bond market as 8 it relates to pre-trade issues. 9 We don't have a recommendation today. That is 10 not because we didn't do any work, but it is because the 11 other subcommittee stole our work. 12 (Laughter.) 13 MR. WORAH: So the ETF classification is, you 14 know, pre-trade transparency on what kind of exchange- 15 traded product he would be trading? And the reference 16 data is some pre-trade transparency around new issues. 17 So, though they stole much of our thunder, we decided to 18 keep plowing on and see if there is any recommendation we 19 could come up with. So I just want to go over a couple 20 of the while we don't have a recommendation, I think 21 it might be educational to go over some of the topics we 22 considered. 23 The first thing I think the FIMSAC and others 24 in this room are familiar with is we decided to study the 25 European experience with method 2 in terms of pre-trade 0116 1 transparency. And the general consensus was onerous, 2 expensive, didn't accomplish much, early days but don't 3 have expectations that it will really improve market 4 functioning, and the Europeans would say that the U.S. 5 market without any of those regulations is probably far 6 more transparent and functioning far better than the 7 European market. 8 So, not wanting to go down the method 2 route, 9 a number of us, you know, on the subcommittee, we felt 10 let's look at what is happening in the retail sector of 11 the corporate bond market. That is something we don't 12 have a lot of people on the subcommittee didn't have a 13 lot of familiarity with. So we invited participants from 14 the retail market. SEC staff helped us set that up. So 15 we had the reps from Fidelity, Styfo, Interactive 16 Brokers, and the retail arm of Morgan Stanley update us 17 on what is going on in the retail corporate bond market. 18 And there are a couple of issues, but, you know, there 19 are different models, discount brokers, full-service 20 brokers, but the general feel was liquidity was there and 21 transparency was there, at least at the broker level, 22 depending on the model. There were dual markets, and 23 people were able to get execution where they wanted. 24 There was probably an issue that we could follow up on 25 around Commissions and mockups and disclosure around 0117 1 those. There is disclosure in the back end in broker 2 statements about whether those should be disclosed up 3 front or not. So that is a topic for further discussion. 4 But generally the retail market, from at least the folks 5 we spoke to, seemed to be functioning well. 6 And then the final topic we considered was 7 whether market makers, other brokers should be mandated 8 or required to post levels on SIPs, or S-I-Ps, and either 9 be mandated, you know, up to a certain size, 10 incentivized, mandated to stick by those levels or allow 11 execution at the levels they were posting. We thought 12 that recommendation didn't make a lot of sense. There 13 wasn't a lot of incentive for market makers to do that. 14 And the second one was whether okay, if they 15 are not mandated to trade at those levels, whether there 16 would at least be some kind of reporting, say, on a 17 monthly basis in terms of levels that were posted on an 18 SIP and what trades got executed inside those levels and 19 what didn't. 20 And, you know, once again, the consensus was 21 that the market was really evolving with the growth of 22 electronic trading, with the growth of SIPs or the growth 23 of even R-to-R markets in the corporate bond market, 24 that, really, we didn't want to force the issue. There 25 really wasn't a lot for us to do in terms of pre-trade 0118 1 transparency because the market was solving itself. 2 So that is an update of where we are today. We 3 don't stop. You know, there are still topics for us to 4 consider in terms of, can we improve liquidity and 5 transparency in the corporate bond markets? But as of 6 today, no recommendation for the FIMSAC from our 7 subcommittee. 8 MR. HEANEY: Thank you, Mihir. 9 Any questions or comments for Mihir? 10 (No response.) 11 MR. HEANEY: Okay. Next we will move to the 12 Municipal Securities Transparency Subcommittee, and we 13 will hear some of the work that has been done, recent 14 work, by the MSRB. I will turn it over to Lynn right 15 now. 16 MS. MARTIN: Thank you, Michael. 17 As the broader FIMSAC will recall, the 18 transparency subcommittee focused on municipal securities 19 was a bit of an offshoot from the original transparency 20 subcommittee, which focused on both corporate and 21 municipal securities. This subcommittee was formed late 22 June. And we presented an initial panel of the state of 23 pre-trade transparency for municipal securities at the 24 last FIMSAC meeting. 25 Following that meeting, we have been busy at 0119 1 work, although we also don't have a recommendation to 2 present today. I take your point about the other 3 subcommittees really focusing on the transparency point 4 and transparency. As I was listening particularly to the 5 corporate bond reference data panel this morning, I was 6 very much thinking back to our own work and how available 7 securities are very much an important part of pre-trade 8 transparency. 9 So on the muni side, following the last FIMSAC 10 committee meeting, we have been focused on I would say 11 three real functional areas and looking at those topics 12 in-depth. The first topic that we have explored was a 13 review of the SEC's 2012 white paper issuing some 14 recommendations for adding transparency to the muni 15 market. And we have been following up on the state of 16 those recommendations, specifically those recommendations 17 which have not yet been implemented. We are going to 18 continue to focus our efforts to ensure that the SEC's 19 good work in this regard from 2012 continues to be 20 brought out and if any outstanding recommendations are 21 still to be implemented, that they do get implemented. 22 The second topic was a white paper or comment 23 letter that got issued I would say late summer 24 recommending the implementation of XBOL disclosures in 25 the municipal securities market as a method of improving 0120 1 transparency. Given the cost to the marketplace of 2 implementing such a standard, while there are benefits, 3 we have decided to adopt more of a wait and see on this 4 recommendation as we do not want to impose undue burden 5 on the market from a cost perspective. 6 The third area, where we have actually spent 7 most of our time, is really around potential improvements 8 to the primary offering process. Most importantly, how 9 do we get more retail participation in primary issues? 10 We have had a variety of presentations at the 11 subcommittee level from participants, including the 12 syndicate desks from Citigroup, Morgan Stanley, Bank of 13 America, Merrill Lynch. And then we have also had a 14 presentation from Charles Schwab also giving us a status 15 on the amount of retail participation in the markets as 16 well as the function of things such as the retail order 17 periods during the issuance process. 18 We have agreed as a subcommittee to start 19 working in this regard on a best-practices document for 20 retail participation in the primary offering process. 21 And that is one of the things we are really looking to 22 bring forward to this Committee I would say probably in a 23 subsequent FIMSAC Committee meeting, possibly as early as 24 January of next year. 25 That said, while we have been able to coalesce 0121 1 around a variety of topics, we haven't really lost sight 2 of the fact that there are many things that we can be 3 doing to improve pre-trade transparency in the municipal 4 securities markets. To that end, we received a white 5 paper from last week issued by MSRB giving their views on 6 the state of pre-trade transparency in these markets. So 7 today I am happy to introduce Simon Wu, who is chief 8 economist at MSRB, who is going to present their 9 findings. And Simon is accompanied by our own John 10 Bagley, who can add his commentary as well on the white 11 paper. 12 So thank you, Simon and John, for being with us 13 today. And we look forward to hearing your white paper 14 findings. 15 MR. WU: Thank you, Lynn. Thank you for 16 inviting me to present our newly released research report 17 on pre-trade paper. 18 As you may be aware, the MSRB has been 19 collecting and disseminating post-trade municipal 20 securities since 1995. The MSRB also provides other 21 pricing information to the public, such as a yield curve, 22 municipal market indices, and a new-issue pricing scale. 23 In the broadest sense, these are pre-trade data as well. 24 However, the focus of our report is on quote data and 25 those quote data signaling trading interest in electronic 0122 1 platforms is currently only accessible to ATS 2 participants and other market professionals if in fixed- 3 income markets, which obviously includes the municipal 4 bond market. With the recent focus on the usefulness of 5 pre-trade information to promote market transparency and 6 we assess the post-quote data as well as the request for 7 quote data collected from two ATS systems that were 8 prominent in the municipal bond trading for a 4-month 9 period in 2015, the assessment is an additional step 10 towards determining whether quote information on 11 electronic platform may be useful for the broader market 12 price discovery and, therefore, may provide value for 13 investment to other market participants. 14 So before I go over our findings, I want to 15 just quickly remind you of the unique qualities of the 16 municipal bond market. There are estimated over 50,000 17 issuers in muni securities and more than 1 million total 18 municipal bonds outstanding. And that compares only to 19 55,000 issuers in corporate bonds and the 50,000 issuers 20 in the corporate bond market. 21 The municipal bond market is also primarily a 22 retail market. So, you know, what I mean by in the 23 retail market is that individual investors directly hold 24 about 40 percent of the market. And if you are counting 25 indirect holding via ETFs and mutual funds, the total, 0123 1 you know, holdings by individual investor could account 2 as much as 66 percent of the total municipal bond market. 3 And most municipal security investors are buy-and-hold 4 investors. So what you see in the typical trading 5 pattern for a municipal security is that when the 6 security is first issued, you will see the rapidly 7 frequent trading in the initial period after the 8 issuance, but then it will follow by very infrequent and 9 sporadic trading for the rest of the life of the 10 security. 11 Of the 1 million outstanding municipal bonds, 12 only about 1 percent of those bonds trade on a typical 13 day. We are talking about on average about 39,000 trades 14 reported to the MSRB. And if you are looking to the PAR 15 value, that is about $11 billion PAR value trade every 16 day. 17 In contrast to other markets, including other 18 fixed-income markets, the directly liquid nature of the 19 municipal bond market, and the mostly buy-and-hold 20 investor positions make the ability to locate a 21 counterparty to trade more difficult. 22 Furthermore, as you may be aware, municipal 23 market participants cannot cost effectively short 24 municipal securities, for various reasons, including the 25 tax regulation by the IRS and the difficulty of locating 0124 1 bonds for borrowing. So, therefore, you know, any type 2 of shorting strategy, including offering a bond on ATS, 3 without actually possessing the inventory is very 4 difficult. So there is also currently no central 5 exchange in the municipal market through which pre-trade 6 quote data are made broadly available to the public in a 7 consolidated manner. 8 So Lynn just mentioned, you know, SEC 9 presented, SEC Division of Economic Analysis, DEA, 10 presented, a white paper during the July 2018 FIMSAC 11 meeting. I just want to point out one main difference I 12 methodology between our paper and the SEC paper. The SEC 13 paper focused on trades where there were at least one 14 like quotes and one response to the request for bid 15 quotes available contemporaneously. And then they are 16 looking for those trades. They assess the transaction 17 cost, you know, using ATS data as a benchmark. 18 Our report, you know, what we did is we traded 19 like quotes and the request for bids separately because 20 we think, you know, these are two different processes. 21 And we focus on the price discovery and the usefulness of 22 ATS quotes as pre-trade information. 23 MR. BAGLEY: Simon, could I just add in? 24 MR. WU: Sure. 25 MR. BAGLEY: And in the muni space, there are 0125 1 very few true two-sided markets. So we thought that the 2 RFQ responses were significantly different enough from 3 the offering responses. So we traded them individually 4 because less than 1 percent of the offerings have two- 5 sided markets. 6 MR. WU: Yes. And okay. So now we talk about, 7 you know, the results of what we found. So overall I 8 paint a picture, you know, for muni ATS usage by 9 interdealer, interdealer subcategory of muni bond trading 10 that does not involve customer trades. About 60 percent 11 interdealer trades were SQ, not ATS'. And that number 12 has been fairly constant during the past two years. If 13 you are counting by the PAR values, that is about 30 14 percent. But the majority, 90 percent, of ATS trades has 15 less than $100,000 PAR value, which is usually a 16 guideline people use, say, "Hey, that is retail size 17 below 100,000 and an institutional size above 100,000." 18 You know, as John mentioned, the majority of 19 quotes are majority live-posted quotes, which are the 20 quotes an ATS participant can't see are live-offered 21 quotes. Very few are live-bid quotes. So very few are 22 in two-sided markets. Because there are so few bid 23 quotes, you know, when an ATS participant wants to sell 24 the bond, you know, the participant will have to request 25 for bids in order to sell the bonds. So, therefore, if 0126 1 you look at a request for quote data, 99 percent of those 2 are requests for bids. So let me talk about the request 3 for bids data results first. 4 During the 4-month period in 2015, there were 5 700,000 requests. And what we found is 2.7 million 6 responses to those requests across a true ATS platform we 7 used. And so that averaged about 3.9 responses per 8 request, with a median about 3. But 11 percent RFQ did 9 not receive any response. Eighty-nine percent did. 10 Overall, the only 25 percent of requests for quotes 11 actually resulted in execution. So if you think about 12 it, the trade rate is only 25 percent. And that trade 13 rate varies quite a bit depending on how many responses 14 you get. So it varies from 9 percent, a trade rate, when 15 you only get 1 respondent to about 40 percent trade rate 16 when you get 20 respondents. 17 Still, 75 percent do not result in trade. But 18 for those 25 RFQ that result in trade, a vast majority of 19 those, over 99 percent of those, actually did trade at 20 the highest bid price. So that is good to know. 21 Okay. Then now we look at the live quotes. 22 During the same period, there were 8.1 million live-quote 23 updates on the 2 platforms we looked at. In order to get 24 a sense how deep the market is, what we did is we took a 25 snapshot. We picked 10:00 a.m. every trading day during 0127 1 the 4-month period. We looked at how many people are 2 quoting for every municipal bond, a 10:00 a.m. snapshot. 3 What we found is 90 percent municipal bonds only have 1 4 or 2 market participants who are quoting at a time. So 5 the vast majority of those, you know, very much there is 6 only one quote, you know, maybe a few percentage or two 7 quotes. So 90-some percent is only 1 or 2 market 8 participants who are quoting. 9 MR. BAGLEY: Simon, let me add. 10 MR. WU: Yes. 11 MR. BAGLEY: So the key thing and this was a 12 market-structure issue with municipals is you really 13 do have to own the bonds in order to offer them. So if 14 someone has got an offering up on the platform, unless I 15 have that exact bond, I can't decide to offer the bond. 16 So that is why you see such a huge percentage. 17 And after pulling out some data that we didn't 18 think was added, it was really like 90 percent of the 19 offerings only are truly unique, with 1 dealer offering 20 the bonds. And that is going to be a significant 21 difference in the municipal structure than you will see 22 on the corporate side. 23 MR. WU: Yes. So I think we expect, you know, 24 much deeper you know, if we did the same for Treasury 25 or corporate, we expect a much deeper end market. 0128 1 Then, finally, we look. You know, we match the 2 market trade price with best-offer quotes. So, again, 3 only one-side quotes are best-offer quotes. We match the 4 market trades with offer quotes. So what we found that 5 is interesting, the interdealer trades, not surprisingly, 6 the median difference between interdealer execution price 7 and the best-offer quotes was exactly zero, you know, and 8 that is maybe not surprising because 60 percent 9 interdealer trades were traded on the platform. And you 10 should expect them to trade at a best-offer quote. 11 But when we matched the customer trades as 12 you know, the customer trades are investor trades that 13 mostly trade off the platform here is what we found. 14 You know, prices for customer buy and customer sales are 15 actually symmetrical around the best-offer quotes when we 16 are looking at the customer buys, about 75 basis points 17 above the best-offer quotes. If you think about bonds 18 quoting 1,000 PAR, you know, that is $7.50. When you 19 look at customer sales, it is about 73 basis points below 20 the best-offer quote or $7.30 for 1,000 PAR. 21 So these results, you know, those two gaps are 22 nearly the same. They reflect the compensation to dealer 23 and financial advisers. But this result is a little bit 24 surprising to us because think again. You know, most 25 other markets we look at, like Treasury, equity, and 0129 1 corporate bonds, have two-sided quotes. So normally you 2 think a customer buy would be related to the best-offer 3 quotes and customer sell would be related to the best-bid 4 quote. So maybe the price may be symmetrical around the 5 bid-ask midpoint, but here we only see the one-sided 6 quote because only one side is visible and live on the 7 platform. And what we found is the quotes actually are 8 symmetrical around the best-offer quotes. But that does 9 show that live-offer quotes provide a good indicator to 10 the market, even though they only represent one side of 11 the market. 12 So, in conclusion, our research shows this pre- 13 trade information could be useful for market price 14 discovery purposes. However, more with more recent data 15 in this area would be needed to reach a firmer 16 conclusion. 17 So the data, as I mentioned, are from 2015. We 18 have learned since then a lot of propriety trading firms 19 and other trading firms are becoming heavy users of ATS 20 in recent years. And institutional buy-side traders may 21 have increased the usage of ATS as well. As a result, we 22 were told that, you know, if we did the same analysis 23 today, you know, we would see, you know, ATS platform 24 have significantly more offering of bonds for live 25 quotes, perhaps 2 to 3 times more than what we saw in 0130 1 2015, even for requests for quotes responses. In our 2 study, we see about three responses, median three 3 responses, per request. But now we are seeing perhaps 4 double that number, you know, responses to each request. 5 And, finally, you know, in 2016, MSRB 6 implemented a new best-execution requirement. And that 7 may have altered our dealer use of ATS as well. And we 8 are also seeing a continued decline transaction cost and 9 for customer trades. So we expect if we did the same 10 analysis, you know, matching trades with quotes, we will 11 see a narrow what I call gap between customer trade price 12 and the best-offer quote. 13 So this is it. And thank you again for 14 inviting me to present this paper. And I will be happy 15 to take any question from the Committee. 16 MR. BAGLEY: Let me just add and then we 17 will happily open up a couple of things that I thought 18 were surprising. So what Simon was talking about is how 19 symmetrical the customer buys and sells were around the 20 midpoint. So for any retail-sized customer trade, 21 basically within $7.50, that was 50 percent of the trades 22 happened with less spread than that. And that was 23 substantially less than I would have thought. And more 24 surprisingly is on the sell side, it was still within 25 $7.30 for 50 percent of the trades. So it certainly 0131 1 seems like the offering price had some impact on where 2 bonds actually traded in the market. 3 Obviously the 90 percent unique offering is a 4 significant thing that we have to think about as we go 5 forward. And I will say that the data that we have, the 6 firms basically had between 50 and 60 thousand offerings 7 on a daily basis. We are now hearing that number is 8 anywhere from 150 to 185 thousand. So they have seen 9 that business grow dramatically. We have also seen a 10 huge uptick in institutional business that we have heard 11 in this platform. But, to be clear, that is almost all 12 on the offered side of the business. Institutional RFQs 13 are still done on a separate platform. So based upon the 14 size of these differences and the ATS platforms, our 15 thought process was probably getting new data and looking 16 at it and seeing what has changed would be the most 17 prudent step. 18 And I do want to make it clear that this was a 19 paper by the MSRB staff, not the MSRB board. 20 MR. HEANEY: Any other comments or questions? 21 Elisse? 22 MS. WALTER: As many of you know, this is an 23 issue near and dear to my heart. In the roll-up to the 24 2012 SEC report, as you know, we held field hearings and, 25 by hook or by crook, actually managed to get real retail- 0132 1 investor participation in those hearings. I think the 2 biggest complaint that we heard from retail investors was 3 that on those occasions when they did choose to sell 4 municipal bonds, recognizing the buy-and-hold nature of 5 the marketplace, they really had no idea what they were 6 worth. And so they had no idea whether they were getting 7 a fair price or not. Presumably, people were complying 8 with their obligations or for the most part were and 9 doing their best to get people the best price. But I 10 think there is and have thought for a long time that 11 there is a screaming need for better information. And I 12 thought that and I continue to think now that to the 13 extent there is information like that which you analyzed, 14 it is important to get it out there and more widely 15 disseminated for the benefit of retail investors. 16 I was very interested in the study, as I have 17 told John privately. I will tell everyone publicly. I 18 thought it was very well-done. I would really very much 19 like to see this Committee move forward with a 20 recommendation in this area. 21 MR. HEANEY: Thank you, Elisse. 22 Larry? 23 MR. HARRIS: Simon, you had said let me make 24 sure I quoted you correctly that when there is a 25 request for proposal and there were 20 responses 0133 1 received, that ultimately the trade only took place 40 2 percent of the time. Is that correct? 3 MR. WU: Yes. 4 MR. HARRIS: What was happening the other 60 5 percent of the time? You got 20 different quotes from 6 different dealers? Is that right? 7 MR. WU: Yes. 8 MR. HARRIS: Do you have any idea what was 9 going on? 10 MR. WU: Well, all we I mean, the platform 11 data, we identify ATS, all we see, it just did not result 12 in an execution. You know, I don't know exactly why. 13 John may add something from a trader's perspective, but 14 from an economist's perspective, the data tell me that 15 they did not result in execution. 16 MR. HEANEY: Horace, do you want to add to 17 that? 18 MR. CARTER: Larry, a lot of times, clients 19 will be trying to discover what their bonds were worth. 20 And in order to do that, they will put their bonds out 21 for the bid. When bonds are submitted, retail client 22 wants to bid on their bonds, whether it is because simply 23 perhaps the adviser is simply trying to find out how much 24 they are worth or the client is. It goes through the 25 entire bid-wanted process, which means it goes through 0134 1 open architecture out to all of the dealers that are on 2 our platform. And they will bid it. And often they 3 don't have any intention of trading it. They're just 4 trying to price-report. 5 MR. HEANEY: Other thoughts, questions, 6 comments? Yes, Amy? 7 MS. EDWARDS: Yes. On that same point, are you 8 looking only at the trade occurring on the platform that 9 had the request for quote or are you looking across the 10 entire market? 11 MR. WU: Yes, it is only on the platform. 12 MS. EDWARDS: So it could be that there is a 13 trade off the platform, correct? 14 MR. WU: Yes. Yes, that is true. 15 MR. HEANEY: Kumar? 16 MR. VENKATARAMAN: Simon, you mentioned that in 17 the municipal bond market, you cannot offer something 18 that you do not own. Is there an exception made for 19 market makers or dealers? 20 MR. WU: No. So I think it you know, so 21 part of the reason, obviously, you know, first, it is 22 very difficult to borrow bonds, but the other reason is 23 the IRS rule because a lot of muni we are talking 24 about tax-exempt muni bonds. And I am not a tax expert, 25 but the rule basis says you cannot create a position and 0135 1 take advantage of tax-exempt status twice because if when 2 you borrow in short, only one of you can be tax-exempt on 3 interest. 4 MR. BAGLEY: I am not a tax expert either, but 5 I did stay at a Holiday Inn Express. 6 (Laughter.) 7 MR. BAGLEY: So I will try to. One of the 8 issues is when you short munis, you end up creating 9 phantom tax-exempt income. So it is possible for a 10 dealer to go out to most likely an institutional client 11 and borrow bonds in order to go short. But now that 12 client can't claim that interest as tax-exempt. So they 13 have to be compensated for it. So it is not that you are 14 not allowed to. It is just that it can be prohibitively 15 expensive in order to try to short munis. And most of 16 the time when someone is short, it is because they made a 17 mistake, not because they were physically really trying 18 to. 19 MR. HEANEY: Brad? 20 MR. WINGES: Yes. A comment I will add to 21 that, you can see the difference. You will see a fair 22 amount of shorting happening in the taxable municipal 23 market, which gets to the point. It proves out that it 24 is related to the phantom income that you don't want to 25 double-create tax-exmepting. 0136 1 MR. HEANEY: Any other comments or questions? 2 MR. HARRIS: I am just going to add to this. 3 So the payments in lieu of interest are fully taxable as 4 ordinary income, and that is the problem. What is 5 interesting is in the equity market, you have a similar 6 problem, but it doesn't seem to be quite as disruptive, 7 perhaps because it is not such a large issue. So if you 8 are holding a security that pays a dividend and you held 9 it long enough, the dividend is given special treatment. 10 But if the position has been lent out and, instead of 11 dividend, you receive payment in lieu of dividend, that 12 is ordinary income. 13 And so people often receive dividends that they 14 think are privileged, and they are not. They don't seem 15 to complain as much as I would expect and perhaps because 16 it is the difference between 40 percent and 20 percent 17 versus the difference between 40 percent and 0 percent, 18 which we see in the municipal market. 19 MR. HEANEY: Horace? 20 MR. CARTER: Just one final comment. When we 21 are talking about pre-trade price transparency in 22 municipal bonds, achieving something along the lines of 23 equity quotations is not I think it has been laid out 24 fairly clearly for us. That is not really feasible. I 25 think that through the process that the dealers engage, 0137 1 we can, in fact, prove pre-trade price transparency, that 2 it is, in fact, provided. Our clients come to us. They 3 ask us for bids. We expose that RFQ extremely broadly. 4 We come back with prices, which we then reflect to our 5 clients. So our clients are getting pre-trade price 6 transparency. 7 In addition, PMP, or prevailing market price, 8 which is now policy, will, in fact, post-trade tell all 9 of our clients what the prevailing market price was at 10 the time of the trade and disclose what the dealer made 11 on it. 12 So, all in all, I think the transparency is 13 fairly robust. 14 MR. HEANEY: Kumar, did you have another 15 question or comment? Okay. Thank you, John. Thank you, 16 Simon. 17 MR. BAGLEY: Thank you. Appreciate it. 18 MR. WU: Thank you. 19 MR. HEANEY: Appreciate your time and your 20 work. I do agree with Elisse. This is a topic of 21 particular interest, I am sure, for us as we go forward 22 in the Transparency Subcommittee. And I just want to 23 take the opportunity to thank Mihir and Lynn and both 24 transparency subcommittees again for the amount of work, 25 time, dedication, and effort that they put forth 0138 1 throughout the year. Thanks very much. 2 We are going to take the shortest of breaks, 3 three, four minutes, just as we set up the panel, so 4 three, four minutes. And then we will come back here and 5 have our last panel of the day. 6 (A brief recess was taken.) 7 MR. HEANEY: Why don't we take our seats as we 8 start this final panel? So our final panel of the day 9 will focus on the role of credit ratings in the corporate 10 bond market. Some of the topics I think we will start 11 with and I am sure this will progress. 12 We have a fairly large panel that based on 13 the couple of the panel calls that we had, there is a lot 14 of territory to cover here but certainly the implications 15 of the increased M&A corporate leverage, where we see it 16 in 2018 and beyond, the impact of a potential economic 17 downturn and what would that mean to the creditworthiness 18 of the BBB credits, credit ratings themselves. Are there 19 tools and methodologies that are currently used for 20 assessing credit versus what could be used versus 21 assessing credit? And then ideas for improving 22 application and reliability, are there model changes that 23 could take place? And this includes modification to 24 investment guidelines as some of our buy-side 25 participants can address that will alleviate or change 0139 1 perhaps the investment-grade high-yield ratings breaks. 2 So just, again, a little bit of a background. 3 The corporate bond market has experienced unprecedented 4 growth, as we all know, over the past decade, aided by 5 both the strong economy, low interest rates, low 6 borrowing costs, corporate debts widely held by 7 individual retail investors and a range of institutional 8 investors, including mutual funds, pension plans, hedge 9 funds, insurance companies, ETFs. 10 The rapid increase of debt issuance has led 11 some to be concerned about the financial capacity of many 12 corporate bond issuers to repay their debt, again 13 particularly in the BBB-rated space. This could be 14 exacerbated in the event of an economic slowdown, and we 15 have panelists that will address some of those concerns. 16 Just last week, a Federal Reserve official 17 highlighted this issue and questioned whether banks are 18 maintaining appropriate risk controls around the 19 increasing use of leveraged loans that are being used, 20 particularly as a means to finance the current M&A 21 activity. 22 In this panel, we will have the opportunity to 23 hear about the current state of the investment-grade and 24 non-investment-grade markets, the credit outlook from a 25 cross-section of market experts, including sell-side 0140 1 strategists, buy-side investors and portfolio managers, 2 and representatives from the credit rating agencies. I 3 think here is, you know, a good example of a portion of 4 the fixed-income market, a sector, not in crisis 5 currently. And this discussion could lead to 6 subcommittee formation perhaps, perhaps not, but this 7 could also be somewhat of a preemptive discussion about 8 modifications that could be used to improve certain 9 sectors of the market. 10 Before we begin the panel, Jessica Kane, the 11 Director of the SEC Office of Credit Ratings, will 12 provide the Committee with a brief overview of the 13 current regulatory structure for credit rating agencies. 14 Jessica? 15 MS. KANE: Great. Thank you, Michael. And 16 good afternoon, everyone. I congratulate you on running 17 early. I sprinted down here from the eighth floor. So 18 forgive me if I am a little bit out of breath. 19 I really appreciate the opportunity to provide 20 a brief overview of the U.S. regulatory structure for 21 credit rating agencies. First, let me just give you the 22 standard SEC disclaimer, which you have probably heard at 23 least once already today. The views I express today are 24 my own and do not necessarily reflect the views of the 25 Commission, any of the Commissioners, or any of my 0141 1 colleagues on the staff of the Commission. 2 The Office of Credit Ratings, which was 3 established in 2012, is responsible for administering the 4 Commission's rules regarding credit rating agencies that 5 are registered with the Commission as nationally 6 recognized statistical rating organizations, or NRSROs. 7 Today's NRSRO regulatory structure essentially began with 8 the Credit Rating Agency Reform Act of 2006, which 9 established a formal registration process and oversight 10 program. In 2010, the Dodd-Frank Act augmented the NRSRO 11 regulations and mandated the creation of OCR. 12 Under the authority of these acts, the 13 Commission has adopted a variety of rules relating to 14 NRSROs. Today I would like to highlight a few of the 15 requirements that may be of interest to this Committee. 16 NRSROs must have an effective internal control 17 structure governing their policies, procedures, and 18 methodologies for determining credit ratings. NRSROs 19 must address and manage conflicts of interest arising 20 from their business. For instance, NRSROs must separate 21 analytical from sales and marketing activities and 22 prevent sales and marketing considerations from 23 influencing ratings. An NRSRO's board of Directors must 24 oversee four specific regulatory areas, including 25 policies and procedures for determining credit ratings. 0142 1 The boards, which must comprise a certain number of 2 independent Directors, are also required to approve the 3 NRSRO's procedures and methodologies for determining 4 ratings. And an NRSRO must designate a Compliance 5 Officer to administer its policies and procedures and 6 ensure compliance with the securities laws. 7 With each rating action, an NRSRO is required 8 to provide disclosures, the 14 major items of 9 information. These include the version of the 10 methodology used to determine the credit rating, a 11 description of the types of data relied on, an assessment 12 of the quality of information considered, an explanation 13 of the potential volatility of the rating, and 14 information on the sensitivity of the rating to the 15 NRSRO's assumptions. The disclosures must be available 16 to the same persons who can receive or access the 17 relevant credit rating. 18 NRSROs are required to make certain public 19 disclosures on form NRSRO, including information such as 20 performance measurement statistics, consisting of 21 transition and default rates for ratings classes. NRSROs 22 must have standards of training, experience and 23 competence for staff that determines ratings. 24 It is important to highlight that pursuant to 25 statute, the Commission is not permitted to regulate the 0143 1 substance of credit ratings or the procedures and 2 methodologies by which an NRSRO determines ratings. For 3 example, the Commission may not prescribe components for 4 an NRSRO to include in its ratings criteria. This would 5 include whether or how an NRSRO factors into a corporate 6 bond rating a company's leverage ratio or expectations 7 regarding de leveraging, the triggers used by the NRSRO 8 for a rating downgrade, and how the NRSRO should surveil 9 outstanding ratings. But that still leaves a lot of room 10 for what we can regulate. 11 OCR is required by statute to conduct an 12 examination of each NRSRO at least annually, covering 13 eight specific review areas. One of the eight areas is 14 whether the NRSRO conducts business in accordance with 15 its policies, procedures, and ratings methodologies. The 16 policies and procedures are not limited to those 17 specifically required by NRSRO regulations. Past exams 18 have found that NRSROs did not properly apply or adhere 19 to their policies and procedures for determining ratings, 20 such as those for surveilling outstanding ratings. 21 The purpose of the NRSRO exams is to generally 22 monitor compliance with the NRSRO regulations, identify 23 potential regulatory violations, and encourage remedial 24 action. The exams are informed by a robust risk- 25 assessment process. This process involves monitoring and 0144 1 analyzing information from a wide array of sources, 2 including media articles, industry publications, and 3 tips, complaints, and referrals. 4 OCR prepares an annual public report 5 summarizing the essential findings of the required 6 examinations. These reports are available on OCR's 7 website through the SEC's website. 8 In addition to the public exam reports, 9 information publicly disclosed by NRSROs under the 10 regulations can assist ratings users in evaluating the 11 quality and accuracy of the ratings as well as the risks 12 that may be associated with particular ratings. 13 So, again, I want to thank you for the 14 opportunity to provide a brief overview of the current 15 NRSRO regulatory structure. And I look forward to 16 hearing the panel talk about the topics that Michael 17 identified and turn it back over to you, Michael. 18 MR. HEANEY: Thank you, Jessica. 19 I would like to introduce the panelists, if I 20 could: Adam Richmond from Morgan Stanley; Eric Beinstein 21 from JP Morgan; Brian Kennedy from Loomis Sayles; Tom 22 Murphy from Columbia Thread Needle Investments; John 23 Bender, Legal and General Investment Management; Craig 24 Parmelee from Standard and Poor's; Daniel Gates from 25 Moody's Investor Services; and Van Hesser from Kroll Bond 0145 1 Ratings. I would like to have each of the panelists kick 2 off with a few introductory remarks. I think we will 3 start with Adam and Eric. They are the corporate bond 4 research strategists from Morgan Stanley and JP Morgan, 5 respectively. 6 Adam, why don't you kick off call it four or 7 five minutes and then, Eric, on just the state of play 8 and how you see the market as we had laid it out. 9 MR. RICHMOND: Sure. Thank you. And thank you 10 very much for the invitation to speak on this panel. 11 So for those who follow our work, you will know 12 that we have a fairly cautious view on credit markets. 13 And part of that is the idea that we think we are quite 14 late in this cycle. And that is when credit starts to 15 experience certain problems here and there. Credit 16 cycles it is important to remember are always going to be 17 different from one to the next, but the two rules of 18 thumb that we try and use when we are thinking about 19 credit cycles are, one, the excesses or the 20 vulnerabilities are always going to be hard to spot when 21 markets are going up and only obvious after the fact. 22 And we very much think that those excesses are out there 23 this time around, at least looking in financial markets. 24 25 And, second, when you are searching for the 0146 1 problems down the line, follow the debt growth. Follow 2 the leverage. It often leads you at least in the right 3 direction. You know, where was the leverage last time 4 around? It was in the financial system. It was in 5 housing. It was on consumer balance sheets, to name a 6 few. Many of those areas seem much healthier to us this 7 time around. In our view this time, the leverage has 8 really sort of found itself, at least in part, on non- 9 financial corporate balance sheets. 10 If I give you a couple of numbers and this 11 is going to be based on the indices that we follow 12 investment-grade credit has grown by and PAR outstanding 13 has grown by about 142 percent since 2009. BBB debt 14 outstanding has grown by about 227 percent. A lot of 15 investors like to strip out financials, which is 16 reasonable. Non-financial BBBs have grown by 181 17 percent. So that is 650 billion in PAR in 2009 versus 18 1.9 trillion today. I know we are going to spend a lot 19 of time on BBBs, I would imagine, but it is not just a 20 BBB phenomenon. 21 The leveraged loan index has nearly doubled in 22 this cycle. And a lot of that growth has come in the 23 last two years with higher and higher leverage levels 24 deteriorating, covenant quality, weaker and weaker 25 structures. The high-yield market has also nearly 0147 1 doubled in this cycle. 2 What has driven all of this? And Michael 3 alluded to some of this up front, but in a nutshell, 4 right, nearly a decade of ultra-low rates, in fact, you 5 know, rates at or below zero in many parts across the 6 globe for so long has incentivized companies to issue 7 debt to buy back stock for M&A and the like. And they 8 have responded accordingly, as one would expect. 9 We all probably slice our leverage numbers a 10 little bit differently, but we would argue at Morgan 11 Stanley that leverage is quite high almost anywhere we 12 look, at least in the non-financial space. And I think 13 you have to be careful when you are comparing leverage 14 today with leverage in past cycle peaks, so, for example, 15 in 2009, as those leverage peaks in the past typically 16 happened after recessions, when earnings had collapsed. 17 Right? So the fact that we are at or above these type of 18 peak leverage levels in a healthy economy when earnings 19 growth is so strong, if anything, in my view is more 20 concerning, not less. 21 We focus on BBBs for a second. And we show in 22 both of the notes that we handed out or that we submitted 23 ahead of time that many companies in investment-grade 24 have leveraged levels that would historically have been 25 consistent with lower ratings. Now, why is this? And I 0148 1 am sure others on the panel will talk about their views, 2 but we would argue that these companies are being given 3 credit for other factors. 4 Now, what are some of the other factors? So 5 one of the things I hear all of the time is how rates are 6 very low so it makes it easier for companies to make 7 their interest payments. We have argued that is 8 something that is risky to rely on for various reasons, 9 you know, one being that the Fed was able to cut rates by 10 500 basis points or more in the last 2 cycles. That 11 tailwind on the way down may not be there to the same 12 extent this time. 13 Second, we hear about, you know, high leverage 14 but very strong business profiles for a lot of these 15 companies, but I would ask the question, how often do 16 business profiles not look strong, at least on average, 17 near the top of a cycle when the economy is very strong, 18 as it is today? 19 And then, lastly, you know, these companies, 20 despite high leverage, often will pledge to de lever over 21 time or at least commit to an IG rating, but, in fact, we 22 looked at a lot of large M&A deals in this cycle. And, 23 in fact, many of them have missed on their de leveraging 24 targets or at least taken longer than they said to get 25 there. And, again, remember, this isn't a healthy 0149 1 economy when monetary policy has generally been very 2 accommodating. It only is going to get harder from here 3 to meet those de leveraging pledges over the next couple 4 of years, I would imagine, especially if we throw a 5 downturn or a recession into the mix. 6 What are the implications of all of this? One 7 is that, you know, downgrade activity could be heavy in 8 the next cycle. And we do a lot of work on this in our 9 BBB note. We looked at downgrade rates in the past three 10 cycles, not really taking a view on whether this one is 11 going to be better or worse but applying those downgrade 12 rates to the size of credit markets today. And we get to 13 significant downgrade volumes again when the credit cycle 14 ultimately turns. 15 And we can, you know, debate back and forth 16 around the exact numbers, and I am sure we all have 17 different views on that. But I do think this will be a 18 theme regardless. And what I would also note is that 19 historically those big downgrade rates have come after 20 credit markets have already priced a lot of it in. 21 Second implication and related is that, you 22 know, this activity could exacerbate some of the 23 liquidity risks on the way down in the form of credit 24 markets that are very big in size and a lot of bonds. At 25 some point, that may have to change hands into what is 0150 1 often an illiquid market on the way down. BBB PAR 2 outstanding today is two and a half times as large as the 3 size of the entire high-yield market. 4 Just the last thing I would end with, I think a 5 good case study to take a look at for maybe how it could 6 play out is the energy bust that we went through just two 7 to three years ago. So, for example, mid 2014, you know, 8 we pointed to energy as a risk along the theme I 9 mentioned up front: follow the debt growth. We got some 10 of the same kind of pushback. Well, these companies have 11 strong business profiles. They have manageable leverage 12 levels. Rates are low. Maturity profiles are backdated. 13 We then got a macro catalyst, right? It was oil prices 14 cracking. It was unexpected, as they always are. Credit 15 spreads widen significantly in 2014 through early '16, 16 kind of anticipating the defaults and downgrades to come. 17 The spread widening was exacerbated by weak liquidity on 18 the way down. And then the defaults and downgrades 19 ultimately followed. 20 So to me, this is by no means a perfect 21 analogy. The macro catalyst will almost certainly be 22 different next time. That is always how it is. But I 23 think this is at least a hint of what it could look like 24 in the next go-around. 25 I will stop there. 0151 1 MR. HEANEY: Adam, thank you. 2 Eric? 3 MR. BEINSTEIN: Just continuing and trying not 4 to throw too many numbers at you, having gone through 5 some of the views, but since '06, the high-grade bond 6 market has grown by 3.7 times, the high-yield by 1.9 7 times. And so the growth in debt really has been focused 8 more in the high-grade markets. And I think that is 9 partly due to the banking regulations that changed around 10 the world post-crisis and bank credit became less 11 available at the same time as QE forced investors to look 12 at other assets besides government bonds. 13 And so you had the catalyst for more issuance 14 in the bond market from the bank-regulation changes, and 15 you had the catalyst for more interest in buying the 16 bonds from QE. The combination led to very rapid growth 17 in the market. And, of course, slow yields made 18 companies eager to take advantage of that confluence of 19 events and grow. 20 And just another take on a number that Adam had 21 said, right now the BBB market is 4.3 times bigger than 22 the BB market. That was 2X 10 years ago. And so a lot 23 of the discussion when we think about downgrade risk, we 24 think about not only will there be more downgrades. What 25 will be the impact of those downgrades? Well, the scale, 0152 1 the difference in scale, between the two markets suggests 2 that the high-yield market will have more trouble 3 absorbing if there was a large number of downgrades than 4 it has in the past, just because the quantums are 5 different there. 6 In terms of M&A, you know, M&A has been a key 7 piece of this story. Since 2015, M&A bond issuance in 8 the high-grade bond market represents 18 percent of all 9 bond issuance to find transactions. In the research from 10 myself that you have in the pre materials, we have a note 11 out on M&A. And we looked at a lot of the big 12 transactions since 2015: 32 deals. And we had the list 13 in there. Leverage on average in those transactions went 14 from 2.6X to 4 times. And 6 quarters later, that 4 times 15 leverage has come down to 3.6. So on average, leverage 16 went up 1.6 times and only unwound one-quarter of that 6 17 quarters later. And these are transactions since 2015. 18 So generally this dynamic has played out in an 19 environment of solid or even strong economic growth. So 20 post these transactions, companies have not actively de 21 levered. That is a mix across 32 deals, some more than 22 others, of course. 23 One of the things I think that is interesting 24 on this is if you look at the largest issuers of high- 25 grade debt and you look at their interest expense and you 0153 1 look at their dividends and zero buybacks, in most cases, 2 there is much more money going to shareholders than is 3 being used to pay coupons. And so over the last few 4 years, there has been a use of the debt markets to expand 5 shareholder payouts, whether that is M&A directly or 6 through higher dividends and share buybacks. 7 I think one of the things that we are going to 8 see in the coming period if we have a slower economic 9 growth is we are going to watch how flexible is that 10 split. You know, we had one of these 32 companies just 11 last week announce cutting their dividend in half in 12 response, really, to a rating agency questioning the de 13 leveraging path. I think that is a theme we are going to 14 be focusing on. But as we saw in the energy selloff, 15 often that was not enough or not quick enough to save 16 ratings from going down. So companies may say they have 17 that flexibility, but whether it is enough at that time 18 will be interesting to watch. 19 In terms of leverage trends, you know, we look 20 at a subset of high-grade issuers. So it is not an 21 expansive, but it is about 200 of the largest companies. 22 And we estimate that right now, about a quarter of high- 23 grade debt is over four times leverage. That is not 24 evenly spread across industries. And in the research 25 that you have, we have some more detail there, but it is 0154 1 in commodities. It is in utilities. And if we look at 2 the ratings migration in the last few years, energy and 3 M&A account for a lot of the moves into the BBB, BBB- 4 space. 5 So there are specific drivers. Obviously 6 energy in '16 caused some I think changes in ratings 7 thinking. And M&A is a discrete event. But, for 8 whatever the reasons, we are finding ourselves, you know, 9 perhaps late-cycled with high leverage with a very large 10 market. 11 And I would also say that, you know, the 12 traditional buyers of high-grade bonds, U.S. insurance 13 companies is probably the biggest piece. You know, their 14 assets have not grown by 3.7 times since '06. The market 15 has been able to grow because other people have come into 16 the market. And that includes a significant share of 17 overseas investors. It is hard to pin down because the 18 data is not really available. One of the topics that we 19 are going to talk about is different investor types, but, 20 you know, I had the opportunity to visit some of these 21 investors around the world. And certainly the size of 22 credit teams is not the same as you have domestically, 23 especially, as you know, the people we have here with 24 very strong credit skills, people analyzing U.S. credit 25 in a time zone 12 hours away. It is more difficult. And 0155 1 so I think we have yet to see what the ability of some of 2 these investors overseas to hold on to bonds if we do 3 have a downgrade cycle is going to be. That is one of 4 the I think interesting questions. 5 MR. HEANEY: Thank you, Eric and Adam. And 6 both Adam and Eric have published, and we have it in the 7 pre-meeting reading materials but with exhibits about 8 this ratings upgrade/downgrade ratios and as you do the 9 joiners and the levers. I will get back to that after we 10 go through other panelists' comments. 11 Let me turn now to Tom Murphy, who based on the 12 panel call may have some differing thoughts on where we 13 are today. 14 MR. MURPHY: Thanks, Michael. And thank you 15 for the invitation. 16 I think Adam and Eric and their colleagues on 17 the sell side have done a tremendous job of raising this 18 issue, but it has been in a kind of a top-down fashion. 19 I will give some more top-down statistics as well, but 20 just to put it in the context of an institutional 21 investor, we really think of things from a bottoms-up 22 perspective. So we have 175 equity and fixed-income 23 analysts that analyze credit and companies. So I am not 24 going to disagree with any of the comments or the 25 statistics that Adam or Eric made. I just want to push 0156 1 it forward a little bit so we can continue the 2 conversation. 3 First of all, BBBs as a percentage and I am 4 actually going to use Bloomberg, Barclay's indices. BBBs 5 as a percentage of the index actually peaked in 2015. 6 They have been coming down since then. So three years of 7 reduction, a percentage outstanding of the index are 8 BBBs. 9 Forty-five percent of that growth in BBBs has 10 come from financials. And 90 percent of that has come 11 from banks. And two-thirds of that has come from five 12 institutions. So there was a re rating of the global 13 financial system X post the financial crisis and another 14 re-rating X the issues in the euro zone. So, again, when 15 we are talking about companies that we are worried about 16 getting downgraded to high yield and potentially 17 overwhelming the high yield from a scale perspective, I 18 don't think we are talking about institutions that are 19 BBB+ that used to be A or AA. I don't think that is the 20 issue we are talking about when the cycle turns. 21 An additional 33 percent of that growth has 22 come from two U.S.-based telecommunication companies that 23 have gone from A to BBB or BBB+. Some of it is due to 24 M&A transactions. Some of it is due to the shareholder- 25 enhancing activities that Eric had talked about. So, 0157 1 again, I don't think those are institutions that we are 2 worried about from a downgrade perspective should the 3 cycle turn. And the last piece would be another 15 4 percent came from 2 auto companies that, frankly, were 5 upgraded from high yield. So between those 3 sets, you 6 have come up with 90 percent of the change in the amount 7 of BBBs. Again, none of those statistics are incorrect. 8 It is just from a bottoms-up investor perspective, I 9 just wanted to peel back the onion a little bit 10 differently. 11 Some other street research, not the work done 12 by Adam and Eric, have shown that BBB leverage has 13 actually come down over the last couple of years. And it 14 is A leverage that has actually gone up. So, again, I 15 know we are talking about companies that could be at risk 16 of the downgrade to high yield, but the leverage from BBB 17 companies has been a little bit more measured over the 18 last few years. It has actually been the leverage of A 19 companies that have actually gone up a little bit. 20 Two more points. Not all BBBs are created 21 equal. Again, a BBB-rated utility holding company that's 22 cash flow has come from A-rated, regulated, different 23 state utility operating companies I think has a different 24 risk profile than a BBB- commodity cyclical producer that 25 has a global footprint that has to worry about commodity 0158 1 price changes and currency changes and things like that. 2 So, again, not all BBBs are created equal. I always 3 find it interesting when I read things in the press when 4 it talks about a company that is BBB-rated. That is two 5 notches from junk. It is also two notches from A rating. 6 So, again, the directionality is probably down, but the 7 directionality is not down for all of these companies. 8 Then the last thing I would say is if you look 9 at BBB- that have downgrade watches or reviews, BBB- that 10 are on watch for downgrade, that percentage is actually 11 quite low as a percentage of the amount to BBBs 12 historically. And I am sure my colleagues from the 13 rating agencies will talk about that. 14 So the last point that I would make is look at 15 the breakdown of BBBs. I referenced this earlier. The 16 BBB+ component of BBBs has actually gone up by 670 basis 17 points. So, again, As going to BBB. The BBB flat and 18 the BBB- parts of the BBB index have gone down by 500 and 19 170 basis points. 20 So has there been ratings migration? There has 21 been. Primarily it has been from A to BBB+. We can 22 debate what that means from a potential economic downturn 23 perspective. 24 And then just the last point, the point that I 25 made before is just that not all BBBs are created equal. 0159 1 And we should be worrying about those that are on the 2 cusp of high-yield as opposed to those that are higher- 3 rated and that have potentially more financial 4 flexibility to offset any pressures in their business or 5 in the global economy. 6 MR. HEANEY: Thank you, Tom. 7 John Bender, would you like to add your 8 comments and thoughts on these? 9 MR. BENDER: Yes. So one of the questions, one 10 of the comments in the first part of the question was, 11 what is your first-order concern, you know, with this 12 topic and given where we are in the cycle? One thing I 13 would remind people, you know, while we all actively 14 manage corporate bonds against benchmarks, the real 15 purpose of the U.S. corporate bond market is to help 16 companies raise capital in an efficient manner. And 60- 17 something percent of the credit, a little over 60 percent 18 of the credit, created in the U.S. economy is distributed 19 in the corporate bond market. So the stability of this 20 asset class is really crucial to the stability of the 21 economy. And we need to keep that kind of forefront in 22 our mind and set up a paradigm where investors have an 23 even tradeoff between potential upgrades and potential 24 downgrades. 25 M&A has been talked about a little bit in this 0160 1 cycle. M&A always starts out at the bottom of the cycle 2 as a relatively low-risk transaction between widget maker 3 A and widget maker B. And they take a bunch of costs 4 out. And 90 percent of the synergies are realized. And 5 it works out really well for everybody. Then the 6 economic cycle continues to move on. The transactions 7 get more strategic. They get higher premiums. They get 8 more debt funding. And they become inherently riskier. 9 But in order to justify the debt financing that they need 10 to make them work, they promise to pay down debt in the 11 future. And when they come to the bond market with 12 billions of dollars of debt, they are getting credit for 13 paying that debt down in the future today in their credit 14 rating. And that sets up instability in the corporate 15 bond market, where investors have a one-sided risk 16 tradeoff. They get a bond that stays the same rating if 17 they pay off the debt or maybe moves up a little bit in 18 three years forwards. But if 12 or 24 months later, the 19 company is running behind, that investor is in an 20 adversarial position. 21 We manage money for institutional pension 22 funds. They are trying to diffuse liabilities and create 23 asset liability programs that are prudent to keep 24 promises for pensioners. We need an even tradeoff in the 25 marketplace for investors to have a chance to come up 0161 1 with a reasonable outcome on their plans. And so that 2 statistical bias to always overrate M&A-related issuance 3 and it is a big issuance, right? AT&T did a big deal. 4 Recently Tom Hess did a big deal. There was a big deal 5 announced today in the paper that is going to be debt- 6 funded. And if these companies are always giving credit 7 at the time of issuance for a debt that is not going to 8 be paid back three or four years, you have a long time 9 before you have a more stable company. And I think that 10 is something that we should really think about. 11 I have to believe the rating agencies have the 12 more conservative bias in their rating so that there is a 13 more even tradeoff for investors over time. 14 MR. HEANEY: Brian Kennedy, your thoughts? 15 MR. KENNEDY: So we actually did a little bit 16 of work ourselves, just in response to what Adam and Eric 17 had written. I guess the first thing I would say is this 18 is likely not today's or tomorrow's problem. We are not 19 calling for recession at any time in the next 12 to 15 20 months. There are certainly quite a few I would say 21 strong clouds gathering in here. And we all see what the 22 issues really are. But we have seen this before in some 23 respects. You have seen it in the prior cycle on the 24 high-yield side. And so I think the big issue today, 25 really, is just, who are the buyers if you do see this 0162 1 kind of cascading effect given the size of the high-yield 2 market? 3 Actually, if you think, look about that. Just 4 look at the high-yield market itself. You have seen 5 negative net issuance in the high-yield market for four 6 straight years. So you have GM and Ford and other 7 companies that, you know, maybe are looking at being 8 downgraded again. And the magnitude of size that they 9 would be bringing into the high-yield market is quite 10 interesting. 11 So the other thing I would say is from the 12 liquidity standpoint, today the liquidity function in the 13 bond markets is much different today than it was during 14 the last cycle. And I think that is very important. I 15 traded at Loomis for 15 years. And when I traded, I used 16 to be able to on the desk have a portfolio manager come 17 to me and bring me a list of bonds they wanted to buy 18 during a downturn. Today is a PM. When I go to the 19 desk, I really have to ask them what is trading because 20 so many things are quoted today. And the breadth and the 21 depth of the markets is much less because of what we have 22 been through. So I think from a liquidity standpoint, I 23 think that is something that needs to be thought about. 24 You know, when we looked at the statistics on 25 this, we came up with a smaller subset of names that were 0163 1 either medium or high risk of downgrade. And, you know, 2 if Morgan Stanley is talking 550 to a trillion, we are 3 talking, you know, 350 to 5. And that is again in a 4 recessionary environment, but you are looking at fairly 5 some of them are smaller issuers. So we are not 6 looking at all of the big mega cap issuers that I was 7 just talking about that would really deluge the high- 8 yield market moving down into high-yield. So you are 9 looking at some of the same industries we have talked 10 about, whether it is communications or consumer cyclical 11 or energy. 12 You know, the energy example, I think one of 13 the reasons we had such a bad experience there is that I 14 don't know that we saw more than a handful of people who 15 were thinking oil was going from 100 to mid 20s. And 16 that at the time I think was a big deal because you had 17 companies that were cash flow-positive or cash flow- 18 neutral assuming a certain level of the price of oil. 19 And that didn't hold up. And so I think what we found 20 from that today is that oil companies, a lot of them in 21 general right now are spending within cash flow and so 22 have learned their lesson a little bit. 23 And so when we look out at industries right 24 now, it is difficult to see that one industry that we are 25 looking at, which is typically what is the first kind of 0164 1 domino to fall in something like this. And we are 2 looking for that, but that first industry out there is 3 really not evident at this point in time. 4 So I would just say that, you know, in general, 5 we looked at this and said a little bit of a smaller 6 subset, but certainly the storm clouds are gathering in 7 here and this is, you know, obviously something we are 8 paying fairly close attention to. 9 MR. HEANEY: Thank you, Brian. 10 Let me turn to the participants from the rating 11 agencies now for their comments. Craig, would you like 12 to kick off. 13 MR. PARMELEE: Thank you, Michael. We 14 appreciate you allowing S&P Global to participate in this 15 event. 16 Once again, I am Craig Parmelee. And my role 17 at S&P is I am the global head of the analytical 18 practices. I want to begin by taking a step back away 19 from the BBB discussion for a moment and just talk about 20 rating since this panel is entitled "The Role of Credit 21 Rating Agencies in a Higher Leverage Environment." I 22 want to talk a little bit about the foundation, what is 23 in our ratings, and some of the changes that we have made 24 since the financial crisis because I think it is 25 important to state that here. 0165 1 First and foremost, credit ratings are forward- 2 looking opinions about creditworthiness. Ratings are a 3 tool that help financial markets to function more 4 smoothly. They are a common language. We rate debt in 5 125 countries. And when we make the statement that a 6 U.S. corporate is rated BBB and a Japanese bank is rated 7 BBB, we are offering a common language that the credit 8 quality is relatively similar for those two entities or 9 the debt that they issue. 10 In order to be much more transparent around our 11 ratings, after the financial crisis back in 2009, we put 12 out an article entitled "Understanding S&P's Ratings 13 Definitions." This article basically laid out economic 14 stress scenarios and attributed rating levels to those 15 economic stress scenarios. And the underlying premise of 16 this paper was that if you are a AAA-rated entity and you 17 face a AAA level of economic risk, you ought to be able 18 to start without defaulting. It doesn't meant you won't 19 be downgraded during that period but without defaulting. 20 And we explain clearly what we mean by a AAA or by 21 the way, a AAA level of economic stress in that article 22 is equivalent of the Great Depression. But we lay out 23 what we mean by BBB levels of stress. We point to 24 historical experiences of those stress periods so that 25 the market has a good understanding what is baked into 0166 1 the level of ratings that we have. 2 About a year later, we felt like it was 3 important to put out another paper, and we called it our 4 credit stability criteria. This is a paper that defined 5 the tolerance that we have for rating movements during 6 moderate periods of stress over certain time horizons. 7 So, for instance, this paper says that we wouldn't rate a 8 security AA if we thought that given a moderate amount of 9 stress within a year's timeframe, it would likely move to 10 A. So our criteria is built to take into account these 11 tolerances. 12 There were many other lessons learned from the 13 financial crisis. One was the need to eliminate silos 14 and to track risk across geographies and across sectors 15 in a much more formal way. To this end, we created what 16 we called credit conditions committees. These are 17 regional groups of our most senior economists, our senior 18 researchers, and our senior analysts, who get together on 19 a quarterly basis to discuss top risks. These 20 individuals lay out those risks, discuss the impact, 21 share that with the market in a commentary, perhaps in a 22 webcast. And that information is passed on to our 23 analytical teams to help them to have that systemic 24 understanding and thought process as they deliberate in 25 their rating committees. We have sector-level committees 0167 1 that are similar to this that meet much more frequently 2 that we call analytical oversight and consistency 3 councils, who do the same sort of thing on a sector level 4 within their region. 5 Since the financial crisis, credit rating 6 agencies have become highly regulated. We heard from 7 Jessica some of the rules that we operate under. And we 8 take our obligations under those regulations extremely 9 seriously. And it is not just regulations imposed by the 10 SEC but by many other regulators around the world. We 11 emerged from the financial crisis as a much stronger 12 organization. We have invested substantially in our 13 processes and our procedures to protect the integrity of 14 our ratings, our people, our methodologies. We have 15 brought in new leadership, and we have enhanced our risk 16 management. 17 I want to touch a little bit about ratings 18 performance. We certainly acknowledge that the ratings 19 that we issued on U.S. housing-related structured finance 20 securities in the mid 2000s did not perform to our 21 historical standards, but if you look at the other 22 ratings that we issued in that same timeframe, the 23 outcomes were very similar to that of other stress 24 periods. And that includes the ratings that we issued on 25 U.S. corporate securities. Ratings are not built to 0168 1 survive the financial cycle. Ratings do move given 2 certain levels of stress, although that movement tends to 3 be somewhat modest. For us to build ratings that don't 4 move across economic cycles, number one, we would have to 5 have an understanding at the beginning of the cycle what 6 the timeframe and severity of that cycle would be and 7 also we would have to put in place such Draconian 8 assumptions into our ratings during expansionary periods 9 to the extent they would likely not be accepted by the 10 marketplace. 11 I want to touch briefly about BBB-rated 12 obligations. And, again, there are a lot of numbers that 13 are going around. And, generally speaking, I am in full 14 agreement with the premise that is being said, but there 15 are a couple of points that I want to add. Very large 16 increase in BBB-related debt. Tom talked about the 17 banks. And that is an important component. In 2015, S&P 18 downgraded the holding companies of Bank of America, 19 Morgan Stanley, Goldman Sachs, and who is the one I 20 missed? And Citigroup. And that added $600 billion 21 worth of debt to the BBB category. That downgrade was 22 based on a structural analysis. That downgrade had to do 23 with our belief that an effective resolution regime had 24 been implemented in the U.S., that the FDIC's work 25 reached the point that OLA could be implemented. And 0169 1 that created a distinction between the position of 2 holding company debt holders versus operating company 3 debt holders, which led to that downgrade. 4 The telecommunications downgrades of Verizon 5 and AT&T in 2013 and '15 had at 300 billion. So, again, 6 that is about a trillion dollars in additional debt into 7 the BBB category. The risk of having eight panelists is 8 you are going to get some people saying the same things. 9 Let's talk about BBB risk for one second, 10 default risk. We know it is very low. Most years, BBBs 11 don't default. If you go back to 2002, when we had the 12 bursting of the telecom and technology bubble, the 13 default rate for BBBs rose to 1.3 percent. But what we 14 are really talking about is rating movement. And I agree 15 very much with the idea that we really ought to look at 16 this from a bottom-up perspective because it is easy to 17 extrapolate. And from a bottom-up perspective, I think 18 we have got to start with the most vulnerable BBB-rated 19 credits. And those are the ones that are rated BBB- and 20 that either have a negative outlook or are on current 21 review for potential downgrade. This data goes back to 22 July, but in mid July, there were 28 companies that met 23 that definition. They had in total about 90 billion 24 worth of debt outstanding. 25 On average in a given year, 4 and a half 0170 1 percent of U.S. companies become fallen angels. They 2 move from the BBB category to the BB category. That 3 figure rises during times of stress. So during 2009, 4 that figure rose to 7.6 percent. If you go all the way 5 back to 2002, that number was 9.2 percent. There are a 6 lot of ways to look at this. If you isolate U.S. 7 corporate, take financials out of the mix, and you look 8 at the 2 year default rate over '08 and '09, you are 9 talking about 9 and a half percent. So I think that is 10 the way we ought to begin to look at things. And no 11 cycle is the same. There are always differences that 12 need to be taken account of. 13 If you look at the entire BBB- debt universe 14 today, we are talking about $640 billion. It would be 15 unprecedented for all BBB- to be downgraded during a 16 downturn because of the diversity of the sectors that we 17 are talking about there. 18 We think that if you use that sort of 9 and a 19 half percent 2-year default rate or if you look at it 20 rating level by rating level, BBB-, B, and BBB+, and 21 study historical performance during downturns, you 22 probably end up with potential downgrade exposure in the 23 2 to 3 hundred million range. 24 I know my time is up. I have got more to say 25 on M&A and other points, but let me defer and let my 0171 1 colleagues speak and 2 MR. HEANEY: Rest assured I will come back to 3 you on the M&A once. 4 MR. PARMELEE: Yes. Thank you. 5 MR. HEANEY: Dan, please, your thoughts as it 6 relates to what you have heard or prepared remarks from 7 Moody's Investors? 8 MR. GATES: Sure. I will comment. Coming 9 late, I will comment on the remarks that others have 10 made. First of all, I am going to speak of BAA and not 11 BBB given that I am from Moody's. So certainly JP Morgan 12 and Morgan Stanley in their publications, they treat the 13 two as similar. I may also speak of B2, which they trade 14 as similar to B. 15 We have observed the growth of the BAA rating 16 category. If you look at ratings now versus 10 years 17 ago, corporate, both nonfinancial and financial, there 18 has been a downward shift within investment-grade. That 19 is one of the reasons that we have heard participants 20 talk about the large size of debt that is rated BAA. A 21 lot of that is issuers that were downgraded from the A 22 category. And there has been a similar trend within 23 speculative grade. So at this point, a majority of the 24 U.S. nonfinancial corporations have a rating in the BAA 25 category, BAA1, BAA2, or BAA3, which is the lowest 0172 1 investment-grade rating. Those three categories 2 constitute a majority of companies by number. And in 3 spec. grade, we have the lowest rating distribution in 4 both spec. grade and investment grade that we have had. 5 In spec. grade, 60 percent of U.S. spec. grade, 6 nonfinancial corporations, have a corporate family rating 7 that is our benchmark rating. And then debt is notched 8 against that based on seniority of claim. That 60 9 percent are rated B2 or lower. So it is a very low 10 rating distribution. 11 And historically we like others that have 12 spoken, you know, we have done many, many decades of 13 studying of how ratings perform, how companies perform. 14 The vast majority of your defaults come from that B2 and 15 lower category, which is a very large category. We have 16 published that the conditions are right. In the next 17 stress, probably a recession -- that is the normal stress 18 -- and the typical profile for defaults is that defaults 19 are below their long-run average for a long period of 20 time, which has been the case in the U.S. since 2009. 21 And defaults go through a period of spike: two years, 22 three years, sometimes four years. And that is really 23 where stress forces those weak companies that we have 24 very low ratings on to default. So we have noticed this 25 increase in leverage. We do have ratings across the 0173 1 scale that are lower today than they were 10 years ago. 2 There has been some talk about mergers and 3 acquisitions. We have never rated to the peak leverage 4 at closing, not where an investment-grade company with 5 both the ability in a publicly stated plan to de-lever 6 we form our own view. We don't just accept management's 7 plan. In doing that, we have many cases where we 8 downgrade, but, whether we downgrade or we don't, we are 9 balancing healthy skepticism against an experience with 10 companies that have the ability and a public commitment 11 to de-level typically compete with a substantial portion 12 of that plan. They don't necessarily achieve the whole 13 plan, although we have had cases recently where we have 14 downgraded companies that have outperformed their plan. 15 And we have then upgraded them several years later. So 16 it can go both ways. 17 We had some comments about market perception of 18 leverage and people are talking about debt-to-EBITDA. 19 You can't reduce credit to one number. Ratings are 20 looking -- they are a forward-looking opinion of the 21 relative risk of default and loss on debt instruments. 22 And we are using methodologies in rating committees. 23 Ratings are not determined by one analyst. They are 24 determined by rating committees through a robust process. 25 And in doing that, we are looking forward at what is the 0174 1 likely outcome when there is a merger and acquisition 2 transaction. 3 There was a recent Bloomberg article, cited 50 4 companies where there had been merges an acquisitions, 5 made this allegation that Moody's is rating too high 6 because the debt-to-EBITDA of these companies, many of 7 them, is consistent with BA, the highest category of a 8 spec. grade, rather than BAA. You can look at a bond 9 that trades and these are mostly large companies, 10 where their bonds trades -- and you can derive an option- 11 adjusted spread for that company, which is effectively 12 the market's vote on the credit risk of that bond. 13 So I had someone do that for those 50 14 companies. And, indeed, the bulk of those companies did 15 not align with Moody's rating, but more than twice as 16 many, the bond-implied rating was higher than Moody's 17 rating, not lower. So the market view, at least the 18 consensus view and certainly not if one has the consensus 19 view, the market view on at least that sample of 20 corporates, is that Moody's ratings are actually 21 conservative. 22 MR. HEANEY: Thank you very much, Dan. And we 23 will come back to the implied ratings with that in a 24 second. 25 But I would like to hear comments from Van and 0175 1 your thought process from Kroll. And are there different 2 approaches in how you think about the same topics? 3 MR. HESSER: Thank you, Michael. Thank you for 4 having us. Thank you to the panel for the invitation to 5 speak. 6 Unlike probably everybody here, everybody in 7 the room may not know who we are. Kroll Bond Rating 8 Agency was started in 2010 by Jules Kroll with the idea 9 that post the financial crisis, having learned everything 10 we did about how the rating process worked through the 11 crisis, that the markets could use a better rating 12 agency. And that has been sort of our guiding force over 13 the last seven or eight years to produce timely, 14 transparent research to test conventional thinking. 15 Right? We think competition in the ratings business, 16 competition for credit opinions is a good thing. Right? 17 The more, the merrier. Again, we are happy to do 18 research-based work that will, again, sometimes 19 oftentimes agree with conventional thinking or the 20 consensus. The other times, we think it is very much in 21 our interest and in the market's interest to hear where 22 our opinions might differ from that conventional 23 thinking. 24 You know, sort of Tom mentioned earlier our 25 research is fairly straightforward. Nobody is splitting 0176 1 the atom here. Credit research is a bottoms-up company- 2 by-company endeavor for us. We don't go into it with 3 preconceived biases, right? We don't think about coming 4 out on the conservative side or an aggressive side. Our 5 goal is to get the right rating that is going to be 6 durable through the cycle. 7 As we get on through the discussion today, I am 8 sure we will talk about M&A, but using M&A as an example, 9 companies are dynamic entities, right? We are not rigid 10 in that. We don't have sort of prescribed bans for how 11 we would measure financial or operational risk. Rather, 12 you know, a big part of what we do, an essential and 13 critical part of what we do, is to spend a lot of time 14 with management trying to think through what they are 15 doing strategically. Right? So our ratings, by 16 definition, we think, always try and be forward-looking, 17 trying to take into consideration what management's plan 18 is. We use the past as a guide. We are not held hostage 19 to it, however. We have been steadfastly focused on the 20 future. 21 We believe in evolution. Right? So with each 22 crisis, if we take it in sort of the credit markets, we 23 don't assume that it is necessarily going to repeat. 24 Credit cycles exist. So we are not naive to believe that 25 cycles don't matter. But we do believe that managements, 0177 1 boards, investors, regulators where applicable, policy- 2 makers all learn from every cycle that we go through. 3 And, again, that thinking we take into account with how 4 we think about how a company is going to perform, again, 5 going forward. 6 We have been very successful in our seven or 7 eight years. I think our message has sort of resonated 8 with markets. Let me talk a little bit about one other 9 thing that I think; I am not going to dwell on some of 10 the things that were already laid out, but I do want to 11 talk a little bit about, you know, in the event of a 12 significant market disruption, do we think there is going 13 to be a significant market disruption in the next 14 downturn from a wave of downgrades on BBBs? The 15 prospects are very real, right? So you hear about the 16 numbers that are out there. There are lots of mitigating 17 circumstances that I think you really need to take into 18 consideration. 19 A couple of thoughts that we would add to that 20 discussion. One, we believe that markets are not as 21 rigid as some would believe. Right? If a wave of BBBs 22 is downgraded, it might overwhelm the high-yield buying 23 capacity as is defined today, day one. We all know that 24 these things are slow-moving train wrecks. Right? They 25 can take years to play out. So to assume that all of 0178 1 this is going to result in a market shock is a hard thing 2 for us to imagine. But assuming that there is true 3 foreselling -- that is a big assumption -- a bid will 4 emerge we think when the price is sufficiently cheap. It 5 will then come from unconstrained buyers who see relative 6 value in an attractive way. So the true size of the 7 buyer base will not be what we think it is today. 8 This is not to say that the redistribution will 9 be seamless because there likely will be significant 10 volatility and dealer quotes as market participants go 11 through price discovery, but it will get done. We saw a 12 similar threat in and around 2005, when the autos -- 13 right? Ford and GM at the time, were among the largest 14 corporate issuers at the time, downgraded or were 15 threatened to be downgraded at an investment-grade and 16 the market thinking, along the same lines as today, is 17 there enough high-yield buying power to absorb that? As 18 it turns out, the redistribution happened fairly quietly 19 and efficiently. It wasn't seamless. But that stood as 20 contrary to many investors' expectations. 21 So, again, from our perspective, you know, 22 again, the credit market, BBBs have emerged as sort of 23 the sweet spot for reasons that lots of people have 24 talked about. We think it is still going to be a big 25 part of the investment-grade market, but we think it is 0179 1 going to be a relatively durable one as we go through 2 right? The next downturn. 3 MR. HEANEY: Thank you, Van. 4 It has been brought up a couple of times. So I 5 just want to follow up with the M&A question. Eric 6 brought it up. Both have published on it, both Eric and 7 Adam. And I completely agree credit analysis is a three- 8 dimensional tool. So, just to use leverage, just to use 9 any one factor is a little crude, but just for the sake 10 of the conversation, I think, Eric, the number you had 11 talked about was leverage going from 2.6 to 4 and then 12 down to 3.6 since 2015 in this M&A-type wave. Adam, you 13 have published on it as well. 14 I guess it raises the question to the three 15 participants from the rating agencies. And we can 16 certainly open it up afterwards. Is the rating process 17 in an M&A cycle different or the same from a "Here comes 18 a brand-new company coming to market, $5 billion market 19 cap. It is getting rated for the first time, happens to 20 be a BBB-rated credit"? Is the analysis any different? 21 And, Dan, you touched on it a bit, but is it any 22 different in an environment like this for M&A? And how 23 do you think about that? 24 And we will open it up eventually to the buy 25 side. Excuse me. Because John kind of kicked this a 0180 1 little bit when he talked about, you know, some of the 2 assumptions made in the ratings following M&A. 3 MR. GATES: So the rating process is not 4 different. It is the same process, whether it is a new 5 issuer coming in for a rating, an existing issuer where 6 we are reassessing the rating. In M&A, what you are 7 dealing with is you are dealing with an event so you have 8 a change. 9 So you need to make a reassessment. And 10 certainly when we do that, we go through the same 11 process. We use the same rating methodologies. We use 12 the same standards. 13 I need to explain that because JP Morgan and 14 Morgan Stanley have noted that debt-to-EBITDA for a BAA 15 company is after -- at the time of mergers and even a 16 year after has been higher than it was before for the 17 same rating level. A couple of things there. One is 18 that our timeframe isn't necessarily 12 months after the 19 acquisition. 20 We will publish a press release when we 21 have an acquisition event, whether we downgrade, whether 22 we have a negative outlook, whether we affirm. And in 23 that press release, we will give insights as to what we 24 are looking for and what our view is on what could change 25 the rating. 0181 1 Frequently, these BAA transactions that have 2 been talked about are transformative events. So you have 3 seen some recent cases, for example, in the media sector, 4 in the healthcare sector, where an issuer is buying 5 another issuer that is a third of its size or half its 6 size or sometimes close to its size, really transforming 7 the business, so even at the end of de-leveraging. And 8 usually in those cases, there are plans to sell assets. 9 There is a mix of two companies that are in a similar 10 industry or adjacent. And the firm that is making the 11 acquisition doesn't want to keep all those assets or they 12 have been required by regulators to promise they will 13 divest some of those assets. So they have assets that 14 are earmarked for sale. 15 They also typically, being investment-grade 16 companies, have a great deal of financial flexibility to 17 pay down debt if that is what they choose to do. And 18 that really is the key question, is are they credible or 19 aren't they? Sometimes you find that the business 20 disappoints and that is why they don't de lever as much 21 as they promised. It is more likely that they don't de 22 lever as much as they promised because they changed their 23 mind. 24 They find something else they want to buy. They 25 want to put their cash to use somewhere else. But, on 0182 1 average, they do achieve, you know, a substantial portion 2 of their de-leveraging. And we take that into account. 3 We take into account that you afterwards have a larger 4 business that has got a more diverse source of assets and 5 cash flow, typically has an improved competitive 6 position, which gives you reason to believe over the long 7 run that that company's cash flow will be more stable 8 than in the past. 9 So all of that is factored into the analysis. 10 As I said, we frequently downgrade. We just are not 11 typically downgrading based upon the debt-to-EBITDA 12 metric at the time of the merger or even expected six 13 months or a year after. We are looking further out than 14 that. 15 MR. HEANEY: I will open it up broadly. Craig? 16 MR. PARMELEE: Sure. I will just add a 17 comment. I mean, I think generally we agree with the 18 comments that were made there: you know, assessing the 19 business impact, the strategic fit, the degree of 20 integration risk. 21 But I have got a little data because we asked 22 our corporate team to go back and look at M&A 23 transactions over the past couple of decades. And so 24 they looked at every investment-grade U.S. M&A 25 transaction that we rated between 2004 and 2017. 0183 1 So let me acknowledge the weaknesses of the 2 data. We are only talking about 13 years. And, as was 3 stated earlier, the timing when the M&A transaction 4 actually takes place is important here. But take the 5 data for what it is. 6 Roughly a third of the time, we downgraded the 7 company at the close of the transaction. So, then, 8 moving forward beyond that point, 2 years later, about 85 9 percent of ratings remained unchanged. And of the ones 10 that did change, it was almost equal between upgrades and 11 downgrades. I think it was 8 percent downgrades and 7 12 percent upgrades. 13 So, again, perhaps some weaknesses in that 14 data. So take it for what it is. But that is I think 15 helpful here. 16 MR. HEANEY: Van, anything to add? 17 MR. HESSER: The only thing I would add I think 18 is that we think about, again, the idea that companies 19 are dynamic and are going to change. And so when we 20 issue a rating, built into it is a certain amount of flex 21 -- right? That gives, you know, a management team the 22 ability to acquire. 23 And then we will go back in and reassess, have 24 a discussion with management, reassess to see if there is 25 a material change in the probability of default. If that 0184 1 is the case, then we would downgrade. If that is not the 2 case, then we will again try and get a sense as to what 3 the plan is over the near- to mid-term. 4 One point we would make is that we don't think 5 volatile ratings serve markets well. And so we are not 6 looking to change our ratings with every zig and zag in 7 news flow. 8 MR. HESSER: We try and see through that and 9 give a rating that we think is going to be durable 10 through the cycle. 11 MR. HEANEY: Other comments, reaction? Tom? 12 MR. MURPHY: Yeah. I think it's interesting. 13 We need to spend more time talking about managements and 14 most of these companies are going to -- we're talking 15 about these large capitalization companies and Dan 16 mentioned, you know, they're much bigger. They've got 17 much more scale. 18 Most of these companies have the wherewithal to 19 reduce leverage, if pushed or if they wanted to. It 20 really comes down to the willingness of the management 21 team and, you know, we don't buy companies if we haven't 22 talked to the management team or met the management team 23 and gone through Ks and Qs and comments that they've made 24 at industry conferences or at analyst days, and for a lot 25 of these BBB companies, it comes down to is the 0185 1 management team willing to pay the price to pull the 2 levers to either keep a rating or keep investment grade 3 metrics, and I think that's something that gets lost in 4 there's going to be a trillion dollars of companies that 5 are going to be downgraded to high yield. 6 So, you know, it comes back to what Van talked 7 about before, bottoms-up analysis of the companies, their 8 strategic positioning, and their operational and 9 financial leverage, but, more importantly, has the 10 management teams been good stewards of the balance sheet, 11 and will they pull the levers when pushed? 12 MR. HEANEY: John? 13 MR. BENDER: Yeah. So I agree it comes down to 14 an assessment of management quality. I'm assuming you 15 lend money to somebody, there's an element of trust 16 involved, that you have to make a judgment, and we all 17 have to make judgments. 18 The thing I would stress is, as the market gets 19 bigger and becomes more important to the stability of the 20 economy, most of the risk is transferred at the time of 21 new issuance of the bond right from the company to the 22 new investors and at that time, it's very -- I think 23 there's an obligation for that rating to be more balanced 24 than it's been. 25 If you look at the 30 years I've been in the 0186 1 corporate bond market and through the different cycles, 2 there's been a qualitative misjudgment of an industry by 3 a rating agency in every one of those cycles. It's not 4 the root cause of it but it's exacerbating the difficulty 5 in the market every time. 6 Different industry each cycle that never repeat 7 and won't be the cause of the next one because they get 8 disciplined. We're repeating some sectors. We've had 9 banks in 1990. We had them again '07. But the same 10 thing happens qualitative overly-positive biased 11 judgments set the market up for disappointment and it's 12 happened time and time again over 30 years and it's for 13 us happening again now. 14 MR. HEANEY: Adam? 15 MR. RICHMOND: Yeah. Just a couple points I 16 would make. 17 You know, I agree with the ability to having 18 levers to pull, the ability to pull levers is important, 19 but, you know, the bottom line is that companies usually 20 in aggregate don't pull those levers at the right time, 21 right. So, you know, confidence is probably the biggest 22 driver of debt growth in a cycle and typically when 23 confidence is booming late in a cycle and animal spirits 24 are off to the races that's when companies are levering 25 up their balance sheets and, you know, probably when they 0187 1 should be pulling the levers and then they de-lever in 2 aggregate usually when they're forced to through a credit 3 cycle. 4 You know, we've talked a lot about some of our 5 stats, you know, the stats we quoted in terms of 6 downgrade volumes and, you know, some of the different 7 ways we can slice it and dice it. 8 Again, I would just reiterate that all we're 9 doing there is taking downgrade rates from past cycles 10 and applying it to the market today, right. If we take 11 out financials, which makes sense, the ranges we get move 12 from something like 550 billion to 1.1 trillion down to, 13 let's say, you know, 450 billion to 750 to 850 billion. 14 So it's less, but these are still big numbers that we're 15 talking about and I think it starts to get a little bit 16 dangerous when we start, you know, slicing and dicing the 17 market to make it look better, you know, if we start 18 taking out, for example, the healthy sector that's not 19 going to get downgraded in this cycle. It's not longer 20 an apples to apples comparison, right, if you take out 21 the healthy sector that wasn't going to get downgraded in 22 past cycles. 23 If we're talking about numbers of, you know, 24 200 to 300 billion in downgrades, that may be the case, 25 but then we have to understand what we're saying. What 0188 1 we're saying is this is going to be a much more benign 2 downgrade cycle than we've seen in the past and we have 3 to be able to explain why that's the case. 4 I would argue there's reasons to suggest the 5 other end of the spectrum, not a much more benign cycle. 6 You know, in terms of leverage levels, you know, again, 7 we all slice it differently. You can look at the 8 broadest measures, non-financial, corporate debt to GDP, 9 that is fed data. That's as high as it's ever been. 10 I agree with the point that was made earlier 11 that this is not necessarily a problem for today in terms 12 of the big wave of downgrades, right, but what matters 13 for us is when markets price it in. So markets are going 14 to start pricing this stuff in well before the downgrades 15 happen, well before the economy starts to weaken and just 16 the last point I would make is that, you know, this is 17 not the key issue for us in credit markets. This is one 18 of many issues. 19 It's something that could sort of exacerbate 20 challenges on the way down but to me this is just one 21 stress point in credit markets, not the only one, by any 22 means. 23 MR. HEANEY: Please, Craig. 24 MR. PARMELEE: Just one comment I had there. 25 So the structure of the BBB indebtedness today, at least 0189 1 by S&P ratings, are the 22 percent are BBB-, 46 percent 2 are BBB, and 32 percent are BBB+, and so 78 percent are 3 in the second or highest category, and I think if you 4 look back at cycles, the risk of downgrade from the BBB 5 level tends to be the ones that are one notch away. 6 Some that are two notches away get downgraded 7 and then some that are three notches away get downgraded 8 but it's a much smaller percentage. Now every cycle's 9 different. 10 MR. HEANEY: Yes, Eric? 11 MR. BEINSTEIN: Maybe one of the reasons that 12 we're here today is, it was interesting the comments that 13 we heard about since '04 to '17 about that M&A, but, you 14 know, this year downgrades of investment grade are 15 running 2:1, downgrades or upgrades, and nominal GDP is 16 growing more than five percent. Last year was net, small 17 net downgrade. The prior two years were big net 18 downgrade years. 19 So this is the fourth year in a row of, you 20 know, downgrades exceeding upgrades within IG space by 21 quite a margin with a strong economy and so longer-term 22 trends are what they are, but this is not -- you know, as 23 John said, if we were thinking that ratings are sort of 24 the best estimate, so that bias would be up and down, you 25 would think in a strong economy, maybe the bias would be 0190 1 a little more up and we're not seeing that. Energy is 2 part of it. M&A is part of it, but there's always going 3 to be something that's part of it, but I think that's 4 notable. 5 MR. HEANEY: Okay. I'm going to switch gears 6 slightly and address these questions to the three Buy 7 Side participants because there's a couple of different - 8 - when I referenced earlier potential change in models, 9 one of them is when you think about investment guidelines 10 and when you go out to solicit through the process of 11 trying to gather more assets and you're making investment 12 pitches, the question is are you required to use ratings 13 and where are you required to use ratings? Which 14 agencies do you use? Are there agencies or companies 15 you'd like to use but your firms prevent you from doing 16 so? 17 So I'd just love to switch the tone of the 18 conversation to that and happy for any one of the three 19 to jump in, please. 20 MR. KENNEDY: I would say that when we are out 21 talking to a prospect, we have a template, a guideline 22 template that we provide them with. If they are opening 23 up a separately-managed account where it's just their 24 money that we're managing, it's not a comingled vehicle, 25 they have the ability to tailor the guidelines in 0191 1 whatever way they would like. 2 Obviously we would, you know, have a 3 conversation around the ability to meet our performance 4 targets, depending on what the guideline changes would 5 be. 6 We typically in the guideline template use 7 Moody's, S&P, and Fitch. That's what we have. We do 8 sometimes traffic in securities that don't have ratings 9 from any of those rating agencies, and I will say that as 10 a portfolio manager, I actually cannot buy anything into 11 a portfolio without having a Loomis Sayles rating on it. 12 So we do all of our own research ratings and 13 come up with a rating independent of what the rating 14 agencies are saying and so if there is no rating agency 15 rating on a particular security that we want to buy, in 16 our template, we typically write that the Loomis Sayles 17 rating would substitute for the rating agency's rating. 18 MR. HEANEY: Can I ask just one follow-up? Do 19 you use any rating agencies not registered with the SEC? 20 MR. KENNEDY: No, no. 21 MR. MURPHY: I'll follow up on Brian's point. 22 You know, we'd prefer to spend time talking to prospects 23 about our internal risk rating system and that's great in 24 the process that we have in the 20+ years of history and 25 the robustness with which it's used across all of our 0192 1 fixed-income sectors, but the facts are institutional 2 client guidelines reference ratings. 3 '40 Act mutual funds reference ratings. Both 4 of those types of portfolios reference third party 5 indices providers, which reference ratings in them. When 6 we're managing money for our insurance accounts and for 7 third party insurance accounts, capital and regulatory 8 requirements require ratings, and I would just echo 9 Brian's comment that it's Moody's, S&P, and Fitch are the 10 ones that are really embedded into that process. 11 MR. HEANEY: John? 12 MR. BENDER: Yeah. I won't spend a lot of time 13 because it's very similar. 14 We have 17 credit analysts globally splitting 15 between London and Chicago. They independently assess 16 credit quality but we don't put internal ratings on 17 bonds. We put relative value scores, improving, 18 declining, credit quality, kind of things like that on 19 scale of minus three to plus three to give portfolio 20 managers to more exposure or less exposure than the 21 benchmark. 22 Because of the benchmark comment and because 23 our clients are largely institutional pension funds and 24 insurance companies trying to hedge liabilities in a 25 risk-based capital or tracking an accounting benchmark, 0193 1 ratings are very important to them and Moody's, S&P, and 2 Fitch are our ratings and we're primarily an investment 3 grade player. 4 So I would say about half of our book is 5 usually a seller on the downgrade to high yield. We do 6 have some of our book that has flexibility and to work 7 with us. So the insurance companies will work with you, 8 depending on book yield trade-offs compared to loss, 9 things like that, but I would say roughly half of the 10 book would sell on a downgrade to high yield. 11 MR. HEANEY: John just crossed into my next 12 question. So I'll ask it of the other two. 13 Just to describe for the FIMSAC Committee the 14 process by which you handle credits that become fallen 15 angels, that become high-yield bonds out of the 16 investment grade portfolios, how much are you forced to 17 sell? What percentage do you have discretion over? Walk 18 us through that process. Brian? 19 MR. KENNEDY: I'll just say in summary first 20 that probably the last thing I want to do in a bond is 21 aggregate it from an investment grade to high yield is to 22 sell it. It's probably the worst time to do that. 23 For my products, we have a bucket of high yield 24 built into the guidelines and so it's anywhere from 15 25 percent for our more conservative accounts up to 50 0194 1 percent for our more aggressive accounts and, you know, 2 as part of our strategy, we look to deploy a part of that 3 bucket or all of it, depending on what we feel like the 4 valuations are like in the marketplace. 5 We will at times buy bonds that are R-rated 6 investment grade, knowing they're going to high yield or 7 that we feel are already being valued as high-yield 8 securities. 9 I would say there's not a lot of times where we 10 have to sell immediately. I would say that if our bucket 11 is full, if we have a period of time where our bucket has 12 been full of downgrades or maybe we have bought ahead of 13 it and more companies were downgraded than we thought at 14 the time, then we could also ask the clients for a waiver 15 to hold, and also in the guideline template the clients 16 will give us a period of time to sell beyond the initial 17 downgrade. So we may get 30, 60, we may get six months 18 where we have time just to sell. 19 So in most cases, I would say no, we don't have 20 to sell right away and again it's probably the last thing 21 I want to do. 22 MR. MURPHY: I'd agree with Brian. I like the 23 way he put it. You know, in the guideline template that 24 we show prospects, we ask for a five-to-10-percent 25 bucket, nominal bucket, to own below investment grade 0195 1 securities. 2 My team managers, primarily investment grade 3 managers, investment grade portfolios and mandates. So 4 that's really just five to 10 percent is to be able to 5 catch a rising star or to own a credit that does get 6 downgraded by the agencies where we disagree or just 7 fundamentally we have a disagreement with the agencies as 8 to whether or not this company is an investment grade 9 company. 10 So again to Brian's point, I mean, the last 11 thing we want to do is be foresellers. So if our clients 12 can't give us that five-to-10-percent mandate or can't 13 give us that flexibility because of regulatory 14 requirements that they have on their side, you know, then 15 we are going to be very sensitive to owning things that 16 are BBB- downgrade and outlook-negative. So it's nowhere 17 near the 50 percent that John talked about. It's in the 18 order of low single digits and my hope would be it would 19 be zero. 20 MR. BENDER: One thing to clarify, Michael. We 21 do have a 120 to 180 days usually to liquidate those 22 securities, so it's not day one. 23 MR. HEANEY: If I could, let me go back to a 24 point that was made earlier and it can be open to any of 25 the panelists. 0196 1 But how do ratings differ and movement of 2 ratings differ, depending on the type of investor or 3 geographically, where they're domiciled, difference 4 between insurance company, asset manager, hedge fund, as 5 you have experienced it, maybe for the Buy Side 6 participants, and then for the three Buy Side 7 participants, the experience you have as they're 8 domiciled in different places outside the U.S., and is 9 that vastly different than the U.S. mandates that you 10 have in terms of some of these? 11 MR. MURPHY: It's a great question. I think 12 the answer is to be determined as far as what the effect 13 is going to be. I would just say that we manage money 14 around the globe and the reliance on the NRSROs outside 15 of the U.S. is higher than it is here in the United 16 States. 17 I just think that folks like Brian and John and 18 myself, we have much deeper credit research staffs that 19 have, you know, 15 to 20 years' history and our clients 20 here in the United States have much more history with 21 going through that process and going through cycles. 22 So I do think that's one of the big risks to 23 the things that I'm going to talk about, is if we do get 24 this downgrade cycle in conjunction with the economy 25 turning down, I do think the risk of assets moving and 0197 1 being foresellers is much higher overseas than it is 2 going to be here. So I do think that's one of the 3 biggest risks we have. 4 MR. KENNEDY: I would disagree. I think if you 5 look at the statistics of foreign holders of corporate 6 bonds over the last few years, they have absolutely 7 skyrocketed. 8 I think what you've seen recently in the last 9 year maybe is a little bit of a pullback because of the 10 strong dollar. The hedging costs have made it 11 uneconomical in some respects for them to hold as much as 12 they have, but if you go back a year or two, the primary 13 -- I mean just a huge amount of buying from foreign 14 issuers -- I'm sorry -- from foreign entities of 15 corporate issuers. 16 So they hold a lot of bonds right now and 17 volatility and the lack of credit experience would, I 18 would think, lead them to be sellers initially before 19 waiting. 20 MR. BENDER: So in our book, we haven't raised 21 a lot of money in Asia but we've raised a lot of money in 22 Europe. I'd say it's more mandate and outcome-dependent. 23 If the client's more return-seeking, they're a little 24 more open. If they're using bonds to hedge liabilities, 25 it's more similar and the guidelines are more similar and 0198 1 the selling is more similar. 2 MR. HEANEY: Let me switch gears just one more 3 time for two questions to the Rating Agency participants 4 and then we'll open it up to the FIMSAC for questions. 5 But for the committee to understand and some 6 people will have experience around the table with this 7 and many will not but just the ratings process, the 8 relationship between issuer and rating agency, as an 9 issuer comes in for a rating, engaging in the services 10 with the credit rating agency and then if you can 11 incorporate into that unsolicited credit ratings, how you 12 approach that and how you think about that, as well. 13 Craig? 14 MR. PARMELEE: Sure. I'm happy to take it. So 15 I represent the analytical side of S&P. So I'm not at 16 all involved in any of the commercial aspects of the 17 relationship. 18 So the way it works with my team is that, you 19 know, if it's a new issuer, an initial meeting would be 20 set up. There'd be typically a presentation on the part 21 of the company or the issuer to the analytical team. 22 There'd be a lot of questions, a lot of back and forth. 23 There would be a process where over the next two or three 24 weeks, there's quite a lot of dialogue, follow-up 25 questions. It all factors into the sort of information 0199 1 that's used in the rating committee. 2 After a rating is issued, typically there would 3 be regular discussion with the management team. It would 4 often be in the U.S. quarterly in conjunction with 5 earnings being announced. Usually that would take place 6 over the phone but generally there would be a face-to- 7 face meeting once a year. 8 So that's sort of the extent of how the 9 analytical relationship works. 10 In terms of unsolicited ratings, you asked the 11 question are there times that we do assign ratings on an 12 unsolicited basis and the answer is pretty rarely and 13 what would drive us to do it? If we thought that there 14 was a particular market interest about our opinion on a 15 particular issuer or issue and there was sufficient 16 information in the market for us to rate it without 17 engaging with that entity, then we might do it. 18 Certainly in the U.S. market or the Canadian 19 market, we wouldn't do it if it was less than a billion 20 in outstanding issuance, but it's pretty rare that we do 21 it at all and when we do do it, we very clearly in our 22 write-ups and on our website indicate in fact that this 23 is an unsolicited rating so that those in the market know 24 that. 25 MR. HESSER: Very similar process, I think, 0200 1 that Craig described for what we do at KBRA. 2 Thorough due diligence and again you mentioned 3 earlier that management discussion is a critical and 4 essential element of what we do and when we think about 5 the idea of unsolicited ratings, they've never made sense 6 to us. We can't do surveillance because we don't have 7 enough information. When we don't have enough 8 information, we don't feel like we are able to sort of 9 give the quality of work that is the standard in our 10 outcomes. 11 We've only done unsolicited ratings in the odd 12 instance where a structured transaction relies to some 13 degree on the credit support typically of a bank, but as 14 a general matter of policy we don't do unsolicited 15 ratings because of that lack of management in our action. 16 MR. HEANEY: Dan, anything to add? 17 MR. GATES: Yes, I'm also in the analytical 18 side and we maintain a strict separation between the 19 analytical side and sales and marketing. So I'm blind as 20 to the process of an issuer deciding to choose Moody's 21 for a rating. 22 But once they've done that and we go through 23 the rating process, the relationship with the issuer 24 revolves around understanding the issuer's story and 25 plans to help us analyze sources of information that are 0201 1 independent of the issuer, such as audit and financial 2 statements, information from competitors, things that are 3 independent of the issuer and its agent. 4 And as for unsolicited ratings, we do reserve 5 the right to assign an unsolicited rating when we 6 perceive a strong market interest in a Moody's opinion 7 and we also have sufficient information to form that 8 opinion. 9 So in most cases, if there's an unsolicited 10 rating, it's an issuer that is in the public markets, has 11 got published audited financial statements that are 12 available and other information, such that you could make 13 an assessment without the assistance from management in 14 terms of understanding their story. 15 MR. HEANEY: Let me pause and open this up to 16 the FIMSAC Committee for questions and comments and we've 17 already got a few lined up. Gilbert? 18 MR. GARCIA: Thank you. This is for all of you 19 or any of you. You've all talked a little bit about sort 20 of the credit trends and potential market disruption. 21 Is there anything that we should be looking at 22 here at FIMSAC for the next year to help minimize that 23 when the next crisis does happen? Tom? 24 MR. MURPHY: That's a great question. I think 25 just the knowledge that the folks from the sale side have 0202 1 put out there as far as this is a potential issue. 2 Again, my only caveat would be as opposed to the top down 3 let's look at the bottoms up. 4 Not to be cavalier but my job is to not own any 5 of those companies, so our clients can outperform, but I 6 totally understand and respect the issues they could have 7 from the marketplace perspective. So I just think, you 8 know, having knowledge and understanding that this is an 9 issue and having it out there. 10 Though I will tell you, you know, if I had a 11 dollar for every time somebody asked me about the BBB 12 issue, I'd have a lot of dollars and, you know, it's 13 clearly an issue. The numbers are big just because the 14 market's really, really big, right? I mean, we've done a 15 trillion and a half dollars of issuance in the investment 16 grade market each of the last five years and that's to 17 help grow the economy and a lot of that's been taken away 18 from bank issuance. 19 I mean, you wonder why bank lending is down so 20 much, a lot of it's because it's been done in our 21 marketplace and the U.S. capital markets is the best 22 place to raise money and finance these entities. So, you 23 know, I really think just having an understanding of 24 what's going on and, more importantly, just tracking the 25 economic trends is key. 0203 1 MR. KENNEDY: Can I just add one thing? I 2 think the horse is out of the barn a little bit already 3 in this in that you have had the issuance and the build- 4 up are here. 5 I think the issue becomes the how do the 6 markets react post the downgrades and the selling and how 7 big of an effect is that, and I go back to this earlier 8 but one of the things we're really concerned about is 9 just the liquidity function in the marketplace and so 10 you've had the issuance, the debt's out there, the buyers 11 own bonds. 12 The question is are you going to be able to 13 sell it in an orderly fashion with the advent of ETFs 14 that may be selling them, as well, and some other 15 entities, as well, so foreign buyers that we talked 16 about. So the liquidity issue, I think, is to me is the 17 next, I think, conversation to have at some point. 18 MR. HEANEY: And I think, Tom, you raised the 19 point, and it's a good one, and John alluded to it and, 20 Brian, now, you as well. The burden of risk has shifted 21 from the banks to the investors, down to the retail 22 investor, depending on what actual specific investment 23 product we're talking about and so the issuance has 24 happened because the banks have not lent as much and so 25 the capital markets have grown. 0204 1 This is kind of just Eric and Adam started us 2 with the concerns then of what happens should things slow 3 or there's a surprise shock and it is completely centered 4 around that this risk burden has shifted into that space. 5 Yes, John? 6 MR. BENDER: One thing I would add to the 7 warnings, keep an eye on the quantitative tightening 8 process at the Fed, the raising of interest rates. 9 Credit cycles tend to get more volatile when the Fed goes 10 from being easy monetary policy to tighter than normal 11 monetary policy, and we're approaching that level 12 sometime in the next 12 months. So you could end up with 13 a fine-tuned inverted curve which could exacerbate market 14 volatility. 15 MR. RICHMOND: And I would just sort of echo 16 the points that have been made around liquidity risk, but 17 I would really try and drive home that we've talked a lot 18 about BBBs but to me that's just a small slice of it. 19 You know, we haven't talked about the risks and 20 the leverage loan market or the risks and the high-yield 21 market. You know, to me, if you think about -- like I 22 always like to think about on the way down in every cycle 23 there's some sort of accelerator which causes valuations 24 to deviate from fundamentals for periods of time when 25 things are selling off. 0205 1 So the last time it was the deleveraging of the 2 financial system and that risk seems like it's much lower 3 today. Today, it seems like it really is these liquidity 4 risks in the form of just a massive retreat over so long 5 in the cycle and I would argue it's almost been a 6 desperate retreat where credit has been treated almost 7 like a rates product for some investors, especially 8 abroad, which has led to massive growth in credit 9 markets, plus a lot of bonds at some point potentially 10 have to change hands, and so to me that is, I think, the 11 key risk on the way down. 12 I again like to look to early 2016 when high- 13 yield X-Energy was trading at 750 basis points. 14 Fundamentally, that made no sense but why was that? I 15 mean think about the liquidity challenges at the time. 16 MR. HEANEY: Eric, please. 17 MR. BEINSTEIN: There is a flip side to some of 18 this growth, though, is that the market is much more 19 diversified. 20 I mean, if you go back to 2010, the biggest 21 seven issuers of high-grade bond market were the big 6 22 banks and GE, which was a financial at that point pretty 23 much. 24 You know, now you have a much more diverse base of 25 issuers. There used to be 400 issuers in the high-grade 0206 1 market, now there's over a thousand, and so from the 2 market overall, I think that makes us more stable, even 3 though some of those names, of course, will suffer in a 4 downturn. 5 MR. HEANEY: Larry? 6 MR. HARRIS: We've talked a lot about market 7 conditions and potential risks looking forward in the 8 future. 9 Is there anything that the SEC presently is 10 doing that it shouldn't be doing that might mitigate 11 those risks or, alternatively, are there problems that 12 the SEC could solve through regulatory intervention that 13 would also mitigate those risks? 14 So I've got eight authorities here. I'm just 15 hoping that perhaps you guys can tell us what we should 16 or should not be doing. 17 MR. BEINSTEIN: I mean, this is start with what 18 we're doing right. There's another discussion on the 19 TRACE System and the benefits and the negatives on 20 liquidity that that might bring about and the pilot study 21 on that. 22 We think that's important and so, you know, 23 this point on liquidity, to the extent that for large 24 trades, it has been impacted by TRACE, studies there I 25 think may be helpful. 0207 1 MR. HEANEY: Gilbert? 2 MR. GARCIA: Again, this is for everybody. In 3 reality, you know, we've been talking about liquidity for 4 awhile here and trying to figure out is it worse, is it 5 better, etcetera. 6 What do you all think? I mean, given 7 everything that you just said, is it worse now with post 8 Lehman, with some of the big players gone, or is that 9 just a perception? 10 MR. MURPHY: I think it's different. I think 11 the right word is different, and I've known John for 30 12 years, and we used to manage portfolios versus one 13 another, and I kind of always knew what they were 14 thinking. 15 We might have a different view of a credit. He 16 might be a seller, our team might be a buyer, but I kind 17 of knew what his criteria were for buying or selling 18 securities, and it's been alluded to a couple times up 19 here, and Brian talked about it, as well, this vast new 20 trove of foreign investors. You know, is it about cross- 21 currency hedging changes? Is it about the movement of 22 their currency versus the U.S. dollar? Is it about 23 regulatory changes in the Taiwanese life insurance 24 industry? 25 Is it QT that's been crowding people out of 0208 1 their own market and into the U.S. dollar-denominated 2 corporate bond market? 3 So if I think about that from a sell side 4 firm's perspective, from someone that's sitting at JP 5 Morgan or Morgan Stanley, it's a little bit more 6 difficult to think about what the investment criteria, 7 the buy or sell criteria is for these new investors as 8 opposed to guys like John and Brian and I sitting some 9 place in the United States managing pension funds and 10 mutual funds in insurance company portfolios. 11 So I think the TRACE issue notwithstanding, I 12 think the job of a sell side trader is a lot more 13 difficult than it was 15 years ago because, yes, there's 14 this new set of investors but we don't really know 15 exactly what their buy and sell decisions are necessarily 16 based on. 17 A lot of times it's not about I think this 18 credit's going to do better, I think this credit's going 19 to do worse. I think this credit's over-valued, I think 20 this credit's undervalued. 21 There's a lot more things that are outside of 22 what's the relative value of this company vis a vis other 23 opportunities that are driving their investment decision. 24 MR. BENDER: The only thing I would add to that 25 is liquidity is cyclical and market directional. It's a 0209 1 lot better when spreads are doing well than it is when 2 spreads are doing poorly and it's a lot worse when 3 something's going below investment grade. You know, so 4 it's a very dynamic concept. 5 Single-name liquidity in the market today is 6 challenging because of the size outstanding for some of 7 these companies compared to the size of capital allowed 8 to be allocated and trading is -- there's a big 9 difference there than what there used to be. There's 10 some innovation coming on the sell side with things like 11 portfolio trading and things to help move more risk for 12 people using more names and more diverse portfolios that 13 are supposedly easier to place lower bid offer spreads. 14 So they are projects that people are working on to try to 15 improve it. 16 MR. HEANEY: Chairman Clayton. 17 CHAIRMAN CLAYTON: Thank you. Thank you all for 18 being here, and I think I hear broad agreement that it's 19 different. It's different, because we wanted to foster 20 growth, but we also wanted to, you know, add resiliency and 21 de-risk our banking sector. So the growth had to go 22 somewhere. The growth could -- into the corporate bond 23 issuance, and a lot of that's triple B -- I want to thank 24 Larry for trying to help me out. Let me -- let me put on a 25 little -- a little narrower slice on that. What I'm hearing 0210 1 is, let me put it this way. What would worry me is forced 2 selling, selling -- or for no reason other than I have to, 3 not because it makes sense from a credit exposure point of 4 view. And we talked about ratings triggers. I heard 5 something around, you know, exchange rates and foreign 6 buyers. Are there other areas, pockets of forced selling, 7 that we should be aware of or be thinking about? 8 MR. KENNEDY: So the liquidity requirements around 9 the 40 Act funds that go into effect next month are going to 10 leave you an illiquid bucket of 15 percent. As we see 11 volatility in the marketplace and the depth and the breadth 12 of the market narrows, those -- those buckets are likely to 13 be more filled up. And if I get over 15 percent, I'm not 14 forced to sell to get back in compliance there. And so that, 15 to me, is a -- you're -- it's almost, you're being forced to 16 sell at a point in time when -- 17 CHAIRMAN CLAYTON: Now, we're always -- we're 18 always looking for places where -- 19 MR. KENNEDY: That would be one -- that would be 20 one that I've thought of, though. 21 CHAIRMAN CLAYTON: -- something in the market -- 22 MR. KENNEDY: Yeah. 23 CHAIRMAN CLAYTON: -- is -- is creating behavior 24 that's -- 25 MR. KENNEDY: Yeah. It's almost, it's -- as Tommy 0211 1 said, it is self -- it's self-fulfilling in a way. 2 CHAIRMAN CLAYTON: Yeah. 3 MR. HEANEY: Other thoughts? 4 MR. RICHMOND: And I -- would just say from a top 5 down perspective, if you look at Fed data, the two big 6 marginal buyers of credit in this cycle, or at least the 7 increase in ownership of corporate bonds, was mutual funds 8 and ETFs. And that was 2008 through 2013. And so you could 9 call that kind of the QE trade. And then it was non-U.S. 10 investors, from 2015 through 2017, and maybe that's sort of 11 the global QE trade. And so to the extent, you know, there 12 are risks on the way down when sentiment really turns into an 13 illiquid market, I think, you know, we could ask what's going 14 to happen to, you know, what would happen to those buyer 15 bases, or those, you know, percentages that have gone up so 16 much in this cycle. To me, you know, that's one way we could 17 look at it. 18 CHAIRMAN CLAYTON: Let me just, I'm going to ask 19 one more. But I'll call -- I've got a sell question, which 20 is, we talked a little earlier today about leverage ETFs and 21 some of that. Are there passive investment strategies, 22 besides ratings-based passive investment strategies, where 23 people have to sell that trouble any of you? 24 MR. BEINSTEIN: ETFs don't get a lot of focus. But 25 you know, right now, less than one percent of high-grade 0212 1 bonds are in ETFs, and about three percent of high-yield 2 bonds are in ETFs. You know, there's a lot more percentage 3 in mutual funds. And so, I don't know that -- I think ETFs 4 get more headlines than they deserve, at least at this point 5 in their, in their size. 6 CHAIRMAN CLAYTON: I was kind of asking the 7 opposite question, which is are there -- 8 MR. BEINSTEIN: Whether we missed -- 9 CHAIRMAN CLAYTON: -- are there non-ETF strategies 10 that are passive that would have the same impact? 11 MR. BEINSTEIN: I mean, I'm sure my colleagues will 12 have things to say about the passive investment side. It's 13 certainly growing rapidly in both high-grade and high-yield 14 in terms of the percentages, but still relatively small. 15 There's questions on, you know, what does that do in terms of 16 forced demand for new issue, and therefore maybe overstating 17 demand or not giving enough differentiating between pricing. 18 Everybody has to buy the bond. The pricing is different. 19 But I think it's still too small to really be affecting new 20 issue pricing. But I'll leave it to my colleagues. 21 MR. BENDER: Yeah, so in the U.S., we're strictly 22 an active manager of credit. It's a philosophical thing. 23 Globally we do manage some passive mandates out of our London 24 office. They haven't been problematic managing liquidity so 25 far. But I do think it raises the -- you know, if passive 0213 1 funds become more prevalent in the market in credit, and they 2 just take the rating, the published ratings and become 3 reliant on them, and become passive followers of that, it 4 makes the area that I'm making to make the rating more 5 statistically unbiased. Through time, more important. 6 MR. HEANEY: Mahir, did you want to add to that? I 7 wasn't sure if -- 8 MR.WORAH: I think -- I think Jay was just, he was 9 just referring to the corporate bond market. That's -- 10 that's what they're talking about. But then there's these 11 others, semi-passive smart bidder strategies. We saw them, 12 saw them in February, whether it's disparity, whether it's 13 managed volt. A lot of the insurance companies have these 14 managed volt strategies. And they manage, they lever up the 15 equity exposure to 10 percent volt. As volt goes up, they're 16 forced to sell equities. So for example, we manage a number 17 of those strategies for insurance companies. So there's 18 these other strategies out there, but they're not related to 19 corporate bonds. 20 MR. HEANEY: Larry. 21 MR.HARRIS: So the SEC exercises power in our 22 markets, or influence, both through its regulation, but also 23 through its ability to deliver messages to investors and to 24 issuers. So my question is, are there any messages that you 25 all would be delighted if the Chairman or the other 0214 1 Commissioners would express to the markets, and particularly, 2 it might -- so some things like the potential benefit of re- 3 opening bonds so that we have fewer bonds out there. Many 4 people have noted that some of our liquidity problems have to 5 do with the fact that we have so damn many bonds. Or perhaps 6 messages to people about in the event that the market moves 7 very quickly, you know, something like that, and some 8 opinion. So any messages that you would like to see the 9 Commission express? Of course, only in their own name. But 10 with the benefit of your advice. 11 MR. HEANEY: Sonali. 12 MS. THEISEN: All right. Was anyone going to 13 answer? Okay. So, I will come back to -- 14 MR. BENDER: I did say something about it. 15 MS. THEISEN: I wanted to just piggyback on 16 something that John has come back to. And I'm just thinking 17 back on projects that I have been involved with in former 18 careers, such as credit grades, model and expected 19 probabilities of default. As -- as a complement -- not as a 20 substitute but as a complement to the traditional ratings, 21 are there analytical products or quantitative models that are 22 offered in the marketplace, that the industry should 23 potentially embrace? Particularly going back to the comment 24 around passive strategies. Are there models, or what are the 25 kind of limitations of doing so? Again, not necessarily as a 0215 1 replacement to ratings, but something that removed the 2 subjective measures that John has touched on, about such as 3 pledges to de-lever by -- by issuers. Is there -- are there 4 any kind of models or metrics that the industry should be 5 looking at more? 6 MR. BENDER: So our analysts do model out the 7 individual credits and make their own forecasts and 8 recommendations. We spend a lot of time following equity 9 research as well, attending equity conferences, understanding 10 where CEOs are taking companies. Because one of the important 11 things to think about isn't just sources of cash but uses of 12 cash. What are they going to spend money on that could hurt 13 me or help me? But I would say one of the real outside 14 influences that's important is the equity cushion underneath 15 the credit, right? How is the -- how is the PE multiple 16 expanding or contracting, the cash flow multiple, compared to 17 the amount of leverage they have on the balance sheet? 18 That's telling me how much cushion I have, and how confident 19 or not I should be in their ability to service. 20 MR. MURPHY: Yeah, I'm with John. It really comes 21 down to your, your individual credit analysis that you're 22 doing within your own firm. Your question is more, are there 23 models that are out there. And I know both Eric's firm and 24 Adam's firm both have things like that. But if your question 25 is, are any of those going to get enough traction that 0216 1 they're going to not supplant the rating agencies but kind of 2 get up to that level, I just don't see it. I kind of ask 3 Adam and Eric what -- what they think. Because they do have 4 tools that we will look at, and other firms on the sell side, 5 of head tools that we've used historically, to John's point, 6 look at equity volatility, things like that. Look at spread 7 moves, and them, how spreads translate across ratings and 8 spread volatility. 9 But I just don't see anything getting into the -- 10 getting into the level that's going to supplant for -- you 11 know, for some of the investors that aren't as deep from a 12 credit experience standpoint. You know, anything that's 13 going to supplant what the agencies do. 14 MR. RICHMOND: Yeah, I mean, I think that -- I 15 would agree with that. I mean, there's going to be -- 16 there's plenty of models you can build or look at that can 17 assess probability of default and risk across credits. But 18 part of the issue is, you know, you -- and part of the 19 challenge is, you know, that I have been trying to talk about 20 is, you almost need sort of a cycle model on top of it, 21 right? Because ultimately we're trying to forecast -- well, 22 many of us are trying to forecast where spreads are going, 23 and a lot of these fundamental factors, you know, they can 24 matter in relative terms, but they're not going to matter a 25 lot in absolute terms until all of the sudden they matter a 0217 1 lot, right? 2 So, I mean, that's the challenge with, you know, 3 building any one, you know, great model that's going to, you 4 know, that's going to be the end all, be all, is -- part of 5 it is bottoms up and part of it is, comes down to your view 6 on where we are in this cycle. 7 MR. HEANEY: I just want to take the opportunity 8 now to thank the panelists, and albeit this was a big panel, 9 but this was great participation by all, and helped to frame 10 a topic that I'm sure we will have much more discussion on. 11 But for us, as a FIMSAC committee, this has been a great, a 12 great panel. Great to hear the differing views, and a lot 13 for us to think about. So thank you very much. 14 We are going to quickly transition, and stay on 15 time, to open this up as I spoke about this morning, as a 16 last discussion point, which is topics for us to engage in, 17 as we think about FIMSAC 2019. As we just heard from this 18 panel, there is a lot to think about, and perhaps this is 19 worthy of a subcommittee. But let's get on the table either 20 people's thoughts about the relevance of this for future 21 subcommittees. We heard Elisse talk about 144A and 22 Information Dissemination as a potential topic. It was 23 brought up on either side, both from Chairman Clayton and 24 from Brad, about LIBOR and SOFR. So let's -- I'd open it up 25 to people who have thoughts about topics going forward, in 0218 1 '19. Larry. 2 MR. HARRIS: I have several topics; a few that 3 we've already addressed, and some that we haven't. Okay. So 4 with respect to what we've not addressed at all, what we 5 don't have -- we don't necessarily have regulatory authority, 6 but we have influence. I think that it would be useful for 7 us to explore the question of the establishment and growth of 8 state municipal bond banks. So a state municipal bond bank is 9 a bank that is chartered by a state for the purpose of 10 helping its local jurisdictions with their municipal finance 11 problems. 12 And in particular, in its most extreme form, it 13 actually issues bonds on behalf of the entities, so that the 14 Riverside Mosquito Embankment District doesn't have to go to 15 the market. Instead the, say the California State Municipal 16 Bond Bank does it on behalf -- and perhaps not by issuing a 17 specific bond, but a bond that's guaranteed by the state. 18 The benefit of doing this is that it can vastly decrease the 19 number of bonds that are coming to the market in the 20 municipal area. And in that way, create the potential for 21 having active markets, and thereby increasing liquidity and 22 lowering the financing costs for the municipalities and other 23 entities. We don't have -- I don't think we have much 24 regulatory oversight here, but we may have influence. We may 25 be able to explore issues that prevent those markets or those 0219 1 banks from forming. So it's something to think about. 2 Another issue I'd like us to spend more time on is 3 the question of riskless principal trades. We've talked 4 about it a bit in subcommittees. Riskless principal trades 5 are essentially brokered trades. But in the fixed income 6 markets, these broker trades are marked up, instead of having 7 commissions attached to them. The markets are now being 8 displayed on an ex, post basis. But it makes it very 9 difficult still for people to shop for brokers, as they would 10 in any other market, by looking at commission schedules and 11 asking what's the price, on a going forward basis. 12 In subcommittee discussions, we've been told that 13 the large markups that we're seeing on these riskless 14 principal trades are often compensation for financial 15 advisory services. So I'm a little bit troubled by that, 16 because when you mix compensation for two services, things 17 become hard to control and you often create difficult 18 incentives. So the two services of course are execution 19 services, which is the arrangement of a brokered transaction; 20 and the other service is financial advice. The problem here 21 is that when financial advice is paid through transactions, 22 there's an incentive to generate a lot of transactions. And 23 we've seen this problem in other places before, and the SEC 24 has intervened in other places when these problems have 25 arisen. So I think it would be sensible for us to think more 0220 1 about -- about understanding brokers or broker dealers as 2 brokers when they are essentially engaged in a brokerage 3 business. And if they're providing financial advice, they 4 should do it in wrap accounts, or some other structure like 5 that. And then finally I'd like us to consider further how 6 information is provided to clients, retail and institutional 7 but primarily retail, on a pre-trade basis. The extent to 8 which TRACE data is made available to them, and also the 9 extent to which consolidated quotes are made available, where 10 those quotes exist. 11 We've heard a fair amount about the adequacy of 12 that data, but I note that, as we've spoken before or 13 certainly I have, that these bond are just securities, and to 14 the extent that similar data would normally be available for 15 an equity trade, we have to ask whether they should be 16 available for bond trades, when they are available. And -- 17 and for which types of bonds. Maybe only for the -- the top 18 300 bonds maybe. Maybe only during the first month of their 19 existence, or something like that. But there's data proven to 20 be extremely valuable to investors in equity markets. And 21 it's hard to imagine they wouldn't be similarly valuable in 22 the fixed income markets. 23 MR. HEANEY: Thank you, Larry. Sonali. 24 MS. THEISEN: Sure. So I wanted to come back to, I 25 think, a point that Chairman Clayton raised earlier this 0221 1 morning, around pricing. So data, as we know, has become a 2 topic of ever-increasing importance and focus across all 3 markets. Even just, you know, we know from the SEC equities 4 roundtables from last week, I think within FIMSAC we've 5 already tackled data extensively in some of our 6 recommendations, including today's new issue, on ETF, 7 referenced data proposals. 8 But earlier today, in Chairman Clayton's opening 9 remarks, he requested that we consider pricing inputs to 10 models and how those inputs are formed and used. And I would 11 note that I think one related topic, which FIMSAC has not yet 12 explored, is derived data, such as third party composites. 13 And the fixed income markets are increasingly becoming more 14 reliant on these composites, particularly with the growth of 15 ETFs and portfolio trading, systematic strategies, and just 16 more generally, more broadly, factor-based risk transfer. So 17 I think as we look ahead to 2019, I would suggest we explore 18 further the role of derived data in fixed income markets to 19 better understand the current practices related to 20 construction methodology, contributions, and access. And I 21 don't necessarily have a view, I just think that this is an 22 important -- a growing area of importance for fixed income 23 markets, and it would be wise for this group to proactively 24 delve into this area. 25 MR. HEANEY: Thank you, Sonali. Kumar. 0222 1 MR. VENKATARAMAN: Based on what I have read, it 2 appears that the turnover of bond indexes tends to be very 3 high, relative to the equity market. And because institutions 4 have to be -- institutions benchmark against the index, to 5 the extent that the index construction itself is suboptimal 6 or has some inefficiencies, it leads to additional trading 7 activity on the institutional front. And so I was wondering 8 if it would be worthwhile for us to better understand how 9 index providers construct their indexes and whether it is -- 10 whether it's optimal. 11 MR. HEANEY: Thank you, Kumar. Yes, please, Brad. 12 And then Mihir. 13 MR. WINGES: Yeah, the last couple years we've seen 14 on the fixed income side, specifically in the retail space in 15 the industry, a large increase in the structured corporate 16 note market. And some of the nuanced types of taxable 17 instruments that are being created. I think it would be a 18 good idea for the Committee to look into the liquidity of 19 those markets, and the lack of possible or potential 20 understanding of the buyer of, you know, the size of the 21 issuance, the impact of having a CUSIP on a bond that only 22 has 100,000 total for the entire issuance. The number of 23 underwriters involved in that transaction; and therefore if 24 they do need to go to the market for a bid, that underwriter 25 might not even be in business anymore, or let alone, no one 0223 1 else will have any of the documents around it. And then the 2 lack of the ATS quotes. 3 I think that entire market has grown considerably 4 over the last few years, and it's being issued by corporate 5 issuers that are very well-known names. So there might be a 6 false sense of security by the retail investor on what they 7 actually own when they maybe go for a quote. 8 MR. HEANEY: Thank you, Brad. Mihir? 9 MR. WORAH: A couple of things. One is just 10 addressing Kumar's point. It's -- I think it's kind of 11 tough, because a lot of the turnover is just aging, aging and 12 coupon payments. And I don't know how you -- how we get 13 around that, bonds age. So, that -- obviously there's other 14 inefficiencies, but that's the biggest chunk of the turnover 15 in bond indexes. 16 But I think if the SEC is the right body, I mean, 17 we should spend some time thinking about the LIBOR/SOFR 18 transition. You mentioned it. There's a number of bodies, 19 the ISDA is the looking at it. And you know, the -- most 20 market participants are involved. But I think with the SEC, 21 you know, there is looking at it from a swap and derivative 22 point of view. But if terms of cash bonds, to the extent 23 that new issuers are regulated by this institution, what we 24 are seeing is new issues that mature after 2021, when LIBOR 25 is supposed to go away; issuers are still not -- they still 0224 1 don't have any fallback language in their prospectuses. 2 Sometimes it's as simple as "best efforts." You know, if -- 3 if LIBOR goes away, then it'll be "best efforts." I think 4 people are counting on what ISDA does, and then the cash bond 5 market will default to ISDA's choice for fallback language, 6 but we're just not seeing it explicitly, and there should be 7 some effort to either mandate issuers or have it -- have the 8 fallback language explicitly in there, or some efforts to 9 converge on a solution. 10 MR. HEANEY: Thank you. Tom. 11 MR. GIRA: This is the trade reporting issue. It 12 sort of relates to some of the things Larry raised. And you 13 know, when TRACE first started, back in 2001, there was a 14 decision made to have the trades be reported, inclusive of 15 compensation, so that the -- the commission or the markup. 16 And, you know, I think what we're struggling with sometimes 17 is that can -- we're sort of cluttering and mixing things 18 that aren't mixed in other markets. So I just wonder whether 19 it might make sense to explore, given the evolution of the 20 market, given now that there's some markup disclosure, 21 whether -- whether that construct that was developed 17 years 22 ago still makes sense, and whether we should still get that 23 information, but not have it be included in the reported 24 trade. 25 MR. HEANEY: Thanks, Tom. Lynn. 0225 1 MS. MARTIN: Yeah, just echoing what Mihir and 2 Sonali were both mentioning on the derived data side, one 3 caveat or one knock on the effect of the LIBOR/SOFR 4 transition could be on the derived data models that actually 5 go into pricing of all these fixed income securities, so it's 6 not just about the transition from the swaps, from the rates 7 that are mentioned in the actual prospectuses, but it also 8 has evaluated data; providers are potentially using 9 LIBOR/SOFR depending on the rate to extrapolate additional 10 derived data products on the markets. So I actually do think 11 it's a wide-stretching issue. 12 MR. HEANEY: We're well into 2020 now on issues. 13 So we're going to have to get Chairman Clayton to extend this 14 another year. Larry, did you have one more? 15 MR. HARRIS: So moved. Thanks. 16 MR. HEANEY: Thank you, everyone. One more. 17 MR. HARRIS: In market structures where strangers 18 can trade with strangers, you tend to have the potential for 19 a lot of liquidity, because anybody can supply a demand for 20 liquidity or express it. Those market structures are only 21 possible -- these are all to all structures -- when 22 creditworthiness is somehow dealt with through a 23 clearinghouse. And so I think it would be sensible for us to 24 explore the problems that are associated with counterparty 25 risk and of creating an environment where it's much easier 0226 1 for anybody to trade with anybody, subject to guarantees made 2 presumably by a credit, credit-clearing agencies, and then 3 ultimately from brokers or the exchanges that serve the 4 brokers. 5 MR. HEANEY: Thank you, Larry. Horace. 6 MR. CARTER: So, I'm following onto Sonali and 7 Lynn's -- excuse me -- Lynn's comments. An idea for that. 8 We -- we learned earlier today that 25 percent of the bonds 9 that are going across ATS are trading. And these are -- 10 there are an enormous number of responses coming back for 11 these RFQs in terms of price information. That's a rich field 12 to mine or rich field to harvest price information that's not 13 currently disseminated. So if we're looking for more price 14 inputs, I don't know what -- I don't know whose data that is. 15 I don't know who's supposed to have access to it, and who is 16 not, and so on. But that would be -- in my -- it would be 17 worthy of a discussion anyway. 18 MR. HEANEY: Agreed. Thank you. Anyone else? 19 Chairman Clayton. 20 CHAIRMAN CLAYTON: So -- so this is -- I don't want 21 this to be work for the members of the FIMSAC, who, I want to 22 again thank you all for, for giving us your time and your 23 input. 24 But one thing that I think we would consider doing 25 as a -- as a staff, but then for your review is summarize the 0227 1 work of this year and the issues you've identified going 2 forward, as many of you have said the sunlight is sometimes 3 the best solution to market issues. This Committee has done a 4 great deal of work, and I think it makes sense to summarize 5 that work over the course of the year, including identifying 6 the issues you've listed today for future consideration. I -- 7 our markets should have the benefit of that. So a project 8 that, you know, we will take on is to do a short summary for 9 your review. Because I think you've done a terrific job. 10 MR. HEANEY: And Chairman Clayton, that would come 11 to us prior to the January meeting; is that the thought 12 process? 13 CHAIRMAN CLAYTON: Right. 14 MR. HEANEY: Okay. There's a wide variety of 15 topics, a page and a half full. So I -- we thank you for the 16 thoughts and the ideas. I just want to reiterate how I 17 opened up the meeting. My comments was thanking this group 18 for an incredible amount of participation and energy and 19 incredible progress that was made. I don't think any one of 20 us could have thought in January we'd be where we are so far 21 to this point. So I thank you. And, but as you know, we're 22 not done yet. I'm sure there'll be subcommittee calls, 23 plentiful subcommittee calls between now and the end of the 24 year. 25 So I just want to do one last shout out, which is 0228 1 to the people behind the scenes. Although many would know 2 them as just David Dimitrious, Tom Eady and Ben Bernstein, 3 who do so much of the heavy lifting to make this Committee 4 what it is. And so our gratitude for those as well. As you 5 know, it's January 28th which is the next meeting. And 6 again, I'm sure we'll be back to you with not only the five 7 page paper, but some of the thoughts about where we go 8 forward for FIMSAC in 2019. So at this point, I will 9 entertain a motion to adjourn. All in favor? Safe travels. 10 Thanks very much. 11 (Whereupon, at 4:05 p.m., the meeting was 12 adjourned.) 13 * * * * * 14 15 16 17 18 19 20 21 22 23 24 25 0229 1 PROOFREADER'S CERTIFICATE 2 3 In the Matter of: SEC FIXED INCOME MARKET STRUCTURE 4 ADVISORY COMMITTEE 5 6 File Number: OS-1029 7 Date: Monday, October 29, 2018 8 Location: Washington, D.C. 9 10 This is to certify that I, Christine Boyce 11 (the undersigned), do hereby certify that the foregoing 12 transcript is a complete, true and accurate transcription 13 of all matters contained on the recorded proceedings of 14 the investigative testimony. 15 16 _______________________ _______________________ 17 Proofreader's Name) (Date) 18 19 20 21 22 23 24 25 0230 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 __10/29/2018___________, at Washington, D.C., 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: 10/29/2018 15 Official Reporter: Kevin Carr 16 17 18 19 20 21 22 23 24 25