0001 1 UNITED STATES SECURITIES AND EXCHANGE COMMISSION 2 3 4 5 6 ROUNDTABLE ON FIXED INCOME MARKETS 7 8 9 10 TUESDAY, APRIL 16, 2013 11 Amended: June 14, 2013 12 13 14 15 Multipurpose Room L-006 16 100 F Street, N.E. 17 Washington, D.C. 18 19 20 21 22 23 24 Diversified Reporting Services, Inc. 25 (202) 467-9200 0002 1 SEC COMMISSIONERS PRESENT: 2 CHAIR MARY JO WHITE 3 COMMISSIONER ELISSE WALTER 4 COMMISSIONER LUIS AGUILAR 5 COMMISSIONER TROY PAREDES (telephonic) 6 COMMISSIONER DANIEL GALLAGHER 7 8 SEC STAFF PRESENT: 9 RAMSAY, John, Acting Director, 10 Division of Trading and Markets 11 BURNS, James, Deputy Director, 12 Division of Trading and Markets 13 EADY, Thomas, Division of Trading and Markets 14 CROSS, John, Office of Municipal Securities 15 EDWARDS, Amy, Assistant Director, Risk, 16 Strategy and Financial Innovation 17 STARR, Amy, Office of the Chief Counsel, 18 Division of Corporation Finance 19 20 21 22 23 24 25 0003 1 C O N T E N T S 2 PAGE 3 INTRODUCTION 7 4 PANEL 1: CURRENT MARKET STRUCTURE FOR 22 5 MUNICIPAL SECURITIES 6 PANELISTS: 7 Robert F. Auwaerter, Principal, Head, Fixed 8 Income Group, The Vanguard Group 9 John Bonow, CEO, The PFM Group 10 Larry Bowden, EVP/Director, Fixed Income 11 Sales and Trading, Stephens Inc. 12 Ric Edelman, Chairman/CEO, Edelman Financial 13 Services 14 Joseph A. Hemphill, III, CEO, Regional 15 Brokers Inc. 16 Burton Hollifield, Professor of Financial 17 Economics, Carnegie Mellon University 18 Craig A. Noble, Managing Director, Head, 19 Retail Fixed Income, Wells Fargo Advisors 20 Benjamin S. Thompson, CEO/Managing Principal, 21 Samson Capital Advisors LLC 22 Thomas S. Vales, Chairman/CEO, TMC Bonds LLC 23 24 25 0004 1 C O N T E N T S (Continued) 2 PAGE 3 PANEL 2: CURRENT MARKET STRUCTURE FOR 101 4 CORPORATE BONDS AND ASSET-BACKED 5 SECURITIES 6 PANELISTS: 7 Steven C. Genyk, Managing Director, Head, 8 Fixed Income Capital Markets, Janney 9 Montgomery Scott LLC 10 Michael A. Goldstein, Professor of Finance, 11 Donald P. Babson Chair in Applied Investments, 12 Chair, Finance Department, Babson College 13 Nancy Mueller Handal, Managing Director, 14 Head, Structured Finance, MetLife 15 Colin Heffron, CEO, GFI Group Inc. 16 Jonathan Horne, EVP/Portfolio Manager, 17 Pacific Investment Management Company 18 Richard M. McVey, Chairman/CEO, MarketAxess 19 Kevin Molloy, Managing Director of 20 Fixed Income, NYSE Bonds 21 Eric J. Pitt, Managing Director, J.P. 22 Morgan Securities 23 Neil M. Schloss, Treasurer/VP, Ford Motor 24 Robert G. Smith, President/CIO/Principal, 25 Sage Advisory Services 0005 1 C O N T E N T S (Continued) 2 PAGE 3 PANEL 3: POTENTIAL IMPROVEMENTS TO THE 178 4 MARKET STRUCTURE FOR MUNICIPAL 5 SECURITIES 6 PANELISTS: 7 Robert F. Auwaerter, Principal, Head of 8 Fixed Income Group, The Vanguard Group 9 Burton Hollifield, Professor of Financial 10 Economics, Carnegie Mellon University 11 Lynnette Kelly, Executive Director, 12 Municipal Securities Rulemaking Board 13 Jason Lehman, Co-CEO/Managing Member, 14 Headlands Technologies LLC 15 Marshall Nicholson, President, Knight BondPoint 16 Craig A. Noble, Managing Director, Head, 17 Retail Fixed Income, Wells Fargo Advisors 18 Benjamin S. Thompson, CEO/Managing Principal, 19 Samson Capital Advisors LLC 20 J. Ben Watkins, Director of Bond Finance, 21 State of Florida, Chairman of the Government 22 Finance Officers Association Debt Committee 23 Brad Winges, Head, Fixed Income Sales, 24 Trading and Underwriting, Piper Jaffray, 25 President, Piper Jaffray Investment Management 0006 1 C O N T E N T S (Continued) 2 PAGE 3 PANEL 4: POTENTIAL IMPROVEMENTS TO THE MARKET 232 4 STRUCTURE FOR CORPORATE BONDS AND 5 ASSET-BACKED SECURITIES 6 PANELISTS: 7 Steven C. Genyk, Managing Director, Head, Fixed 8 Income Capital Markets, Janney Montgomery Scott LLC 9 Nancy Mueller Handal, Managing Director, 10 Head, Structured Finance, MetLife 11 Colin Heffron, CEO, GFI Group Inc. 12 Richard G. Ketchum, Chairman/CEO, FINRA 13 Richard M. McVey, Chairman/CEO, MarketAxess 14 Kevin Molloy, Managing Director, Fixed 15 Income, NYSE Bonds 16 Neil M. Schloss, Treasurer/VP, Ford Motor 17 Dexter Senft, Managing Director, Morgan Stanley 18 Robert G. Smith, President/CIO/Principal, 19 Sage Advisory Services 20 Kumar Venkataraman, Chairman, Finance 21 Department and Fabacher Endowed Professor 22 of Alternative Asset Management, Southern 23 Methodist University Cox School of Business 24 Christopher J Vogel, Managing Director, Global Head 25 Fixed Income and Currency Trading, BlackRock Inc. 0007 1 P R O C E E D I N G S 2 INTRODUCTION 3 MR. RAMSAY: Good morning, everyone. I'm John 4 Ramsay, Acting Director of the Division of Trading and 5 Markets at the SEC, and it's my pleasure to welcome you 6 to the Commission's Roundtable on Fixed Income Markets. 7 We are pleased you are able to join us for what 8 we anticipate will be an informative and productive 9 series of discussions. 10 The staff of the Division of Trading and 11 Markets, Risk, Strategy and Financial Innovation, 12 Corporation Finance, and the Office of Municipal 13 Securities -- John Cross, head of the office is to my 14 right here -- have spent a great deal of time and effort 15 in planning this event, and I'd like to thank all of them 16 for their efforts in putting the Roundtable together. 17 I'd like to extend a special thank you to all 18 of the panelists who are here and on the panels that 19 follow for the time and expertise that they are lending 20 to the Commission and to the staff in addressing a topic 21 of great importance for the markets. 22 This Roundtable will consist of four different 23 panels. Participants on the one first panel will talk 24 about the characteristics of the municipal securities 25 market today and how that market has evolved in recent 0008 1 years. 2 Participants in the second panel will focus on 3 the characteristics of the corporate bond and 4 asset-backed securities markets and the evolution of 5 those markets, also how they compare to the municipal 6 market. 7 After lunch, the participants in the third 8 panel will discuss whether there are any steps that might 9 be taken to improve aspects of the structure of the 10 municipal securities market, and finally, at the end of 11 the day, on the fourth panel, we will similarly discuss 12 whether there are steps to take to improve particular 13 aspects of the functioning of the corporate bond and 14 asset-backed securities markets. 15 A member of the Commission staff will moderate 16 each of the four panels. I'll moderate the first one. 17 At the start of each panel, the moderator for the panel 18 will introduce the panelists. The panelists will not be 19 asked to provide opening remarks, although we welcome any 20 prepared remarks, which will be placed in the public 21 comment file. 22 At the start of panels number one and two, Drs. 23 Hollifield and Goldstein will provide opening 24 presentations summarizing the existing academic 25 literature concerning the municipal and corporate 0009 1 markets. 2 The moderator will then kick off the discussion 3 by posing one of a series of broad topic questions for 4 consideration by the panelists. Our panelists are 5 encouraged to ask questions or respond to comments by 6 other panelists. Staff may also ask follow up questions. 7 We hope this to be interactive as much as 8 possible, but we will prod the discussion if we need to. 9 I want to encourage our panelists and 10 interested persons to submit comments, empirical research 11 and observations on the topics we are discussing. We 12 hope and expect the Commission will benefit in the coming 13 months from any information that is submitted. 14 Again, I want to thank you for attending this 15 important event. 16 At this time, I'd like to acknowledge 17 Commissioners Walter and Gallagher, who are present here 18 with us. Commissioner Paredes is not able to attend in 19 person but I understand he is listening by phone and 20 following us by webcast. Commissioner Aguilar is 21 presently representing the Commission at a North American 22 Securities Administrators Association conference this 23 morning, and he will be joining us a bit later today. 24 Finally, I'd like to turn it over to our new 25 Chair, Mary Jo White, who I think has now just about a 0010 1 week under her belt. We are excited to have her at the 2 Commission, and especially pleased to have her here at 3 this Roundtable. 4 With that, the floor is yours. 5 CHAIR WHITE: Thank you very much. Almost a 6 week. Thank you very much, John. Let me say good 7 morning to everyone, obviously, and repeat John's welcome 8 to the Commission's Roundtable on Fixed Income Markets, 9 and also repeat a very special welcome and thanks to our 10 very distinguished panelists. We appreciate very much 11 your taking the time to share with us your ideas and 12 thoughts. 13 Let me just pause for just a moment before I go 14 on to acknowledge the events in Boston of yesterday. I 15 know all of our thoughts and concerns are with those who 16 were hurt or lost their loved ones in Boston yesterday. 17 I want to convey mine and I'm sure everyone's 18 condolences to all the runners and families affected by 19 yesterday. 20 I am very pleased we all made it here safely 21 today for this Roundtable, and I'm very glad to be able 22 to open the Roundtable, and I do hope to be back later 23 today to hear live from the panelists about the current 24 state of the fixed income markets, and about potential 25 improvements that can be made to promote a more robust 0011 1 market for all market participants, and to the extent I 2 am not here physically, you can rest assured that I will 3 see it all. I'm really very excited about it. 4 Needless to say, the fixed income markets are 5 critical components of our national economy. At the end 6 of 2012, there was $3.7 trillion in aggregate principal 7 amount outstanding of municipal securities, and $15 8 trillion of corporate bonds. In recent years, as most of 9 you know, the SEC staff has been dedicating increasing 10 resources to the public evaluation of the fixed income 11 markets. 12 Most recently, this effort was launched with 13 the Commission's July 2012 report on the municipal 14 securities market. In that report, the Commission 15 focused on the trading of municipal securities and made a 16 number of recommendations to consider for improving 17 secondary market trading for all market participants, 18 including retail investors. 19 Those recommendations generally included 20 consideration of proposals to improve pre and post-trade 21 transparency, promote the use of transparent and open 22 trading venues, and buttressed existing dealer 23 obligations by among other things requiring dealers to 24 disclose their mark-ups in riskless principal 25 transactions, requiring dealers to seek best execution 0012 1 for customers in municipal securities transactions, and 2 encouraging or requiring dealers to provide customers 3 with relevant pricing information in connection with 4 their transactions. 5 The municipal securities report, however, was 6 only the first step in what needs to be a deeper dive 7 into and across the fixed income markets. Today, the 8 Commission is taking an additional step in its continued 9 study of the market structure for fixed income 10 securities, focusing on the trading of not only municipal 11 securities but also corporate bonds and asset-backed 12 securities. We will also be discussing the various 13 recommendations contained in the report. 14 As everyone can see from your materials, we 15 have assembled a very impressive list of panelists from 16 the academic community, market intermediaries, issuers, 17 investors, and regulators to discuss both the current 18 state of the market structure for fixed income securities 19 and potential improvements. 20 We are very excited about this Roundtable. We 21 all look forward to a very robust conversation today, and 22 many more discussions going into the future. 23 Let me stop there to introduce Commissioner 24 Elisse Walter, who has been a tireless and effective 25 leader of the Commission's efforts in this area. 0013 1 Commissioner Walter, it is my privilege to turn over the 2 microphone to you, and to thank you for your dedication 3 and continued resolve to improve the fixed income 4 markets. 5 Thank you. 6 COMMISSIONER WALTER: Good morning and thank 7 you for that kind introduction, Chair White. 8 I, too, would like to thank our distinguished 9 panelists for taking time out of your busy schedules to 10 be here with us today. 11 As you know, this is an area in which I have 12 long been extremely interested, and I must say I am 13 extraordinarily excited that this event is taking place 14 today and we will be moving forward to take a deep dive 15 into the fixed income markets. 16 I should tell you that I will not be able to be 17 here this afternoon because I, too, am going to appear at 18 the NASAA meeting, but I will follow everything that 19 happens this afternoon with great interest. 20 In October of last year, I gave a speech at the 21 SIFMA Bond Summit, which was entitled "Bringing Municipal 22 Bond Trading into the Light." I don't generally mention 23 titles, but I think that one is particularly telling in 24 light of our subject matter today. 25 And in that speech, I focused on a central question 0014 1 concerning the municipal securities markets. Why have 2 the benefits of technological advancements in the trading 3 of municipal securities not flowed to retail investors? 4 This is particularly concerning given the key 5 role retail investors play in this market. Today, 45 6 percent of the outstanding principal amount of municipal 7 securities is held by individual investors directly. 8 It is clear that the municipal securities 9 market has experienced important technological 10 advancements, electronic trading has grown significantly, 11 as demonstrated by the rise of alternative trading 12 systems. 13 The Commission's July 2012 report estimated 14 that ATSs could account for as much as 50 percent of the 15 number of trades of municipal securities. 16 I'm quite concerned that the benefits of 17 technological advancements flow disproportionately to 18 certain market participants. 19 Municipal bond dealers appear to have leveraged 20 technology to reduce their distribution costs, for 21 example. However, trading of municipal securities remains 22 decentralized and opaque. Investors have preferred 23 dealers and dealers trade selectively with 24 counterparties. 25 The GAO issued a report discussing the market 0015 1 structure for municipal securities early last year. In 2 that report, it made a number of statements concerning 3 the relative informational advantages of the different 4 market constituencies in the municipal securities market. 5 Two key findings were that institutional 6 investors traded at more favorable prices than retail 7 investors, and price dispersion, in other words, the 8 range of prices, was wider for smaller trades than for 9 larger trades. 10 The GAO recognized a variety of factors that 11 could contribute to these findings, including a lack of 12 price transparency, which allows better informed 13 investors to obtain more favorable prices. 14 The issue of pricing and transaction costs is 15 of great interest to me, not only in the municipal 16 securities market, but also in the corporate bond 17 markets. I'm particularly interested in the extent to 18 which pricing and costs are related to greater or lesser 19 transparency in the fixed income markets. 20 The historical justification for the current 21 fixed income market structure has been the difficulty for 22 dealers to aggregate and disseminate interest broadly in 23 a wide variety of securities. 24 I believe we should explore whether technology 25 could largely solve that difficulty by enhancing 0016 1 transparency for the benefit of all market participants. 2 I look forward to a robust discussion today 3 concerning the role of technology and how technology can 4 be leveraged to improve transparency, efficiency, 5 liquidity, and market quality. 6 In addition, I look forward to our dialogue on 7 the corporate bond and asset-backed securities markets, 8 and how those markets compare to the municipal securities 9 market, particularly in light of technological 10 advantages. 11 Additionally, while I suspect there are many 12 similarities among the market structures for corporate 13 bonds, asset-backed securities and municipal securities, 14 I'm eager to learn how the differences in these markets 15 may drive certain characteristics of the market 16 structures and certain considerations for potential 17 reforms. 18 For example, we know that the corporate bond 19 market has a larger institutional investor presence than 20 does the municipal securities market. According to the 21 Federal Reserve's recent Flow of Funds Report, about 35 22 percent of corporate bonds are held either directly or 23 indirectly by retail investors, as compared to 24 approximately 75 percent in the municipal securities 25 market. 0017 1 In addition, we know there are significantly 2 fewer corporate bond issues outstanding than there are 3 municipal issues outstanding. How do and how should 4 these differences impact the market structures for 5 trading these securities? 6 Given the over-the-counter nature of the fixed 7 income markets, hearing more about the existing inventory 8 practices of market intermediaries will be extremely 9 helpful. 10 Also, I understand there have been recent 11 efforts by some market participants to create trading 12 platforms that would enable buyers and sellers to meet 13 directly, potentially altering the role of 14 intermediaries. 15 Learning more about the advantages of and 16 challenges involved in creating these types of trading 17 platforms will be useful to the Commission. 18 As technology allows for improved market 19 transparency, we must ensure that those benefits are 20 realized by all investors. The Commission has 21 recommended consideration of a number of initiatives 22 along these lines, and I'm interested in engaging in a 23 discussion about all of these, either in the course of 24 this Roundtable or in subsequent dialogues, specifically 25 whether panelists believe investors in fixed income 0018 1 markets generally would be served well if existing dealer 2 obligations were buttressed. 3 For example, the Commission recommended that 4 consideration be given to requiring dealers to disclose 5 their mark-ups in riskless principal transactions, a 6 measure that has come up repeatedly, but we have not yet 7 been able to implement despite several attempts over the 8 last three decades. 9 In addition, while best execution has been a 10 common law duty for many years, the Commission also 11 recommended that consideration be given for a specific 12 best execution rule for the municipal securities market. 13 I believe the conditions for a robust best 14 execution rule may now be present given technological 15 advances and dealers' increasing access to price 16 information and liquidity. Indeed, dealers trading 17 corporate bonds have been subject to a best execution 18 rule promulgated by FINRA for years. 19 The Commission also has recommended 20 consideration be given to requiring dealers to provide 21 their customers with relevant pricing information at the 22 time of transactions. 23 And finally, I look forward to a discussion 24 concerning the information loop between the primary 25 offering process and the secondary trading markets. The 0019 1 fixed income markets are critical components of our 2 economy, and the ability of issuers to access the capital 3 markets sufficiently is vital. 4 I'm concerned that the failure of the secondary 5 municipal securities markets to realize fully the 6 efficiencies of available technology negatively affects 7 investors who may obtain worse prices for securities than 8 they would in a more modern market structure. 9 I'm also concerned this impacts municipal 10 issuers who may not receive as advantageous prices for 11 their securities as they would if there were a more 12 efficient and liquid secondary market. 13 As you can see, we, and I guess particularly I 14 have many questions and look forward to a robust 15 discussion that includes as many perspectives as 16 possible. 17 I would now like to introduce my colleague, 18 Commissioner Gallagher, who I know along with other 19 colleagues on the Commission takes a great interest in 20 understanding and tackling these issues. 21 Thank you. 22 COMMISSIONER GALLAGHER: Thank you very much, 23 Commissioner Walter. In a rare display of Government 24 efficiency, I'm going to dispense with formal remarks. 25 Actually, it's more in recognition of the fact that I'm 0020 1 been sick as a dog for the last five days and haven't 2 finished them, but I'm going to post them up on the web 3 after this. 4 I'm very excited to be here. I really want to 5 thank Commissioner Walter for all the work she has done 6 particularly over the years in the municipal securities 7 space. I think as simple in many ways as our report was 8 that we issued last Summer is, it actually was such a 9 tremendous leap forward. 10 You know, sitting here looking at John Cross 11 and recognizing that we now again have a formal Office of 12 Municipal Securities, another major leap forward, even 13 though kind of simple on its face. 14 These are tremendous steps for this Agency and 15 we have to keep going in that direction, expanding our 16 expertise outside of muni's, too, and focusing truly for 17 the first time on corporate's and asset-backed's in a 18 sophisticated way, trying to stay ahead of the curve is 19 really important to me. I know it is to Commissioner 20 Walter, Chairman White, and the rest of the Commission. 21 This is a tremendous event. I'm thrilled that 22 we are here. I thank former Chairman Walter for keeping 23 this Roundtable highest on our list of priorities and 24 these issues high on our list of priorities, and I thank 25 Chairman White for keeping it there as she came in, too, 0021 1 because I've been talking a lot in speeches and elsewhere 2 about the basic blocking and tackling of this Agency. 3 Nothing is more basic to our oversight 4 responsibilities than the fixed income markets, and quite 5 frankly, nothing has been given shorter shrift because 6 nine times out of ten, they don't pose the problems to 7 us, the obvious problems, the May 6 like problems that 8 cause us to spend a lot of attention on an issue, and it 9 shouldn't be just because they don't cause a problem that 10 we're not paying attention to them. 11 So I think this is terrific, happy to be here 12 and I want to thank all the panelists and the staff for 13 all the hard work that went into this presentation today. 14 Thanks very much. 15 MR. RAMSAY: Thank you very much. Again, a 16 special thanks to both Commissioners Walter and Gallagher 17 for their constructive engagement today. I know they 18 both have been thinking about and providing constructive 19 thoughts on these issues for many years. This is not 20 something they have come to recently. Having them so 21 actively involved in this Roundtable is a huge benefit to 22 us. Thanks to them for their time. 23 We are uncommonly efficient in having gotten 24 through all the opening remarks before 9:00. Maybe we 25 will be able to actually finish the first segment a tiny 0022 1 bit early or perhaps not. 2 In any event, we will go ahead and start with Panel 1, 3 which as I mentioned before, will discuss the 4 characteristics generally of the municipal securities 5 market and evolution of that market in recent years. 6 PANEL 1: CURRENT MARKET STRUCTURE 7 FOR MUNICIPAL SECURITIES 8 MR. RAMSAY: I'd like to briefly introduce each 9 of our panelists. First of all, Robert Auwaerter. 10 Hopefully, I didn't mangle that too badly. President and 11 Head of Fixed Income Group at The Vanguard Group Inc. 12 John Bonow, CEO of The PFM Group. Larry Bowden, who is 13 Executive Vice President and Director of Fixed Income 14 Sales and Trading at Stephens Inc. 15 Ric Edelman, Chairman and CEO at Edelman 16 Financial Services. Joseph Hemphill, III, CEO at 17 Regional Brokers Inc. Burton Hollifield, Professor of 18 Financial Economics at Carnegie Mellon University. Craig 19 Noble, Managing Director and Head of Retail Fixed Income 20 at Wells Fargo Advisors. 21 Benjamin Thompson, CEO and Managing Principal 22 at Samson Capital Advisors, and finally, Thomas Vales, 23 Chairman and CEO at TMC Bonds LLC. 24 We also have seated around the table staff from 25 the Divisions of Trading and Markets, Risk, Strategy and 0023 1 Financial Innovation, and Corporation Finance. Also, as 2 I mentioned, John Cross from the Office of Municipal 3 Securities, and my Deputy, Jim Burns, will be joining a 4 bit later. 5 Before starting the panel discussion, I'd like 6 to welcome Dr. Hollifield. As I mentioned, he's a 7 Professor of Financial Economics at Carnegie Mellon, and 8 has also written extensively on the municipal securities 9 market. 10 Dr. Hollifield has kindly agreed to help kick 11 us off on this first panel by making a presentation 12 summarizing some of the academic literature concerning 13 the municipal securities market. 14 Dr. Hollifield, thank you very much. If you 15 would like to take it away. 16 DR. HOLLIFIELD: Thank you for asking me to 17 talk here. As was mentioned, I'm a Finance Professor at 18 Carnegie Mellon, and my current research is trying to 19 understand trading and transaction costs in over-the- 20 counter markets. I've done some work in how the 21 municipal market works. 22 My comments today will focus on what academics 23 have studied on the municipal bond market in general, on 24 what we know about secondary market activity in the 25 municipal bond market, and what we know about primary 0024 1 market activity in that market. 2 Maybe to set the stage, as has already been 3 mentioned, the municipal bond market is a large and 4 important market. There are some numbers on my slide, 5 but it's really quite a fragmented, illiquid and opaque 6 market. I think mainly due to the nature of the 7 securities themselves, which are fragmented. 8 Just some basic facts. Most bonds are small, 9 and after they have been issued, they trade quite 10 infrequently. For example, just some recent data I've 11 seen. Bonds, after they have been issued on average can 12 trade three or four times per year, and many, many bonds 13 do not trade after issuance. 14 Because of the fragmentation, illiquidity and 15 opacity in the market, most transactions through 16 customers are routed through bond dealers. There are 17 roughly 700 dealers who are active in every month, so 18 this is some figures from a recent working paper by Dan 19 Li and Norman Schurhoff, who have documented some of 20 these facts. 21 This is a very fragmented and small market, not 22 small, but small amount of trade in many of these 23 securities. Historically, the bond market has had little 24 pre or post-trade transparency, meaning that it was 25 difficult for retail investors to know current prices, 0025 1 but this has changed in stages. 2 For example, in January 2005, trades were 3 required to be reported to MSRB within 15 minutes, and 4 those trades are now available to investors on the 5 Internet. 6 I'd like to just talk a little bit about what 7 we know about secondary market trading. This is a figure 8 that I've taken from a recent working paper by Dan Li and 9 Norman Schurhoff. It's a plot of two things. The solid 10 line that slopes down is a plot of the average mark-up by 11 transaction size, the mark-up being basically the 12 difference between buy and sell prices. 13 I think there are two striking things to me 14 from this figure. One is simply the magnitudes for the 15 small transactions. We're looking at numbers around two 16 percent. These are on bonds where two percent seems like 17 a reasonable fraction of a year's interest payment. So 18 these bonds are quite expensive to trade for retail 19 customers. 20 The second striking fact we see here is unlike 21 in the equity markets, the larger the volume, the lower 22 the mark-up is. So retail flow seems much more 23 expensive. 24 Another fact that we see is there is much more 25 price dispersion in retail sized trades than in 0026 1 institutional sized trades. We will see many different 2 prices, if you can find them, for similar bonds over 3 similar periods of time. 4 A recent working paper by Richard Green and co- 5 authors actually documents some similar evidence in the 6 BABs market, which itself was designed to maybe deal with 7 some of the fragmentation issues. I believe there is 8 some evidence from BABs that we see the same kind of 9 issues. 10 I would also like to talk a little bit about 11 the primary market. Of course, when we look at the 12 secondary market, and especially look at small trades, we 13 might be worried that the bonds are expensive to trade, 14 which means there is no trade in the bonds as investors 15 decide they do not want to trade the bonds. 16 Alternatively, it could be expensive to deal 17 with the bonds on the dealer side because there's no 18 flow so there is kind of a circular problem there in some 19 sense. 20 So one place we could look at understanding that 21 would be in the primary market where we know the bonds 22 need to be distributed and sold. 23 One of the striking findings that has been 24 documented by some work by Rick Green and I was that if 25 you look at small sized transactions in the primary 0027 1 market, you see many, many different prices for similar 2 bonds. You see small retail bonds having wide price 3 dispersion. You also see small sized transactions having 4 higher average prices -- there tends to be a lot of price 5 dispersion in the primary market. That is evidence from 6 before 2005. 7 I would like to talk a little bit about what 8 happened in 2005. One argument for the price dispersion 9 that we saw in these bonds and the average prices is the 10 fact that it was hard for retail investors to get prices, 11 but of course, the increased availability of transactions 12 made that information available. 13 One natural question to see if we can 14 understand the importance of the opaqueness is to 15 understand what happened in the 2005 event. 16 These are just some figures of average mark-ups 17 to show you the kind of prices we were seeing. On the 18 horizontal axis, we are looking at the day after the 19 issue. On the vertical axis, we are looking at the 20 mark-up measured in basis points. The top curve are the 21 prices at which bonds were sold to customers at. The 22 middle curve are the prices that bonds were purchased 23 from customers at. 24 You see two patterns there. You see the 25 average price for sales to small customers is increasing 0028 1 over time, the inter-dealer transaction prices do not 2 seem to be increasing over time at quite the same kind of 3 rate. Those are averages. 4 Here is some information about the price 5 dispersion. These are the day by day distributions of 6 prices for these bonds conditional on not being sold 7 at the re-offering price. There are two things to see 8 here. One is simply the significant spread in these 9 prices. A zero in this plot means a price at the 10 re-offer price. Five means five percent higher. It's 11 quite a lot of price dispersion. 12 So prior to 2005, due to the increasing 13 transparency, we saw quite high prices, quite high price 14 dispersion. There was an event in 2005 where information 15 became available, and I think what we learned from that 16 event is informative to some of the issues here. 17 After the prices became available, the 18 dispersion in prices dropped dramatically, which is 19 telling us the transparency did in fact increase the 20 ability for small retail traders to understand what the 21 securities were worth and for them to understand what 22 they should be paying. 23 But you don't actually see the average price drop, 24 and there was an interesting effect here. We saw for 25 small trades, the average price indeed dropped. For 0029 1 large trades, in fact, there was a slight increase in 2 prices. 3 My interpretation of this is the transparency 4 certainly changed the information that people in the 5 market had but it also changed the distribution of prices 6 and changed somehow the relative bargaining power 7 information that different traders had in the market. 8 The reason I'd like to mention this is this 9 means transparency in and of itself does not necessarily 10 have to drop all prices. It can have some sort of more 11 subtle effects. 12 In a recent study written by Richard Green and 13 co-authors, they looked at this in the BABs market versus 14 the tax exempt market, which is more recent data. We see 15 very similar patterns in the BABs or Build America Bonds 16 market. 17 I think the last point I'd like to make is 18 prior to the release of data from the MSRB in the early 19 2000s, not much was actually known about these markets 20 publicly. There was no data for anybody to study. 21 I would like to take this opportunity to say I 22 think for whatever change has happened, it's quite 23 important that data or information become available for 24 people to look at, to understand how these things work. 25 I think it is also important as different changes or 0030 1 rules get phased in to be cognizant of what we could 2 learn from these changes in the structure, and to make 3 that information available for people to understand. 4 That is what I mean by the phrasing of the 5 changes. I think it is quite important to be able to 6 allow people to study how these things work. 7 I think that is where I'd like to stop at this 8 point. 9 MR. RAMSAY: Terrific. Thank you very much, 10 Dr. Hollifield, for a really interesting and striking 11 presentation in a number of respects, and that should 12 provide some good crisp for the mill on some of the 13 topics we would like to talk about for the first panel. 14 With those remarks as a back drop, I'd like to 15 kick off the discussion. Incidentally, I understand I am 16 to call on Commissioners if they signal, I've been told, 17 by placing a hot pink card on the table. I don't know if 18 they in fact have such a card. 19 COMMISSIONER GALLAGHER: Just as long as you 20 don't call us if not signaled. Be careful, John. 21 (Laughter.) 22 COMMISSIONER WALTER: It's very nice to know 23 what the hot pink card is for. I wondered what sort of 24 a signal it was. 25 (Laughter.) 0031 1 MR. RAMSAY: There we go. Why hot pink, I 2 don't know. That will be conspicuous, I guess, in any 3 event. Please don't be shy. 4 In any event, first question, Commissioner 5 Walter had talked about the role of technology in the 6 municipal securities market and that was a significant 7 theme as well in the municipal report that was issued in 8 July. We wanted for our first segment today to talk a 9 little bit about the role of ATSs, alternative trading 10 systems, and also broker's brokers in the trading of 11 municipal securities. 12 I'd like to start perhaps by beginning either 13 with Tom Vales or Joseph Hemphill, whichever might want 14 to start off, by offering some general thoughts about the 15 role of those mechanisms in the municipal market, in 16 general, your thoughts about how the use of technology 17 has or has not impacted pre-trade transparency, liquidity 18 and efficiency over time. So, either of you? 19 MR. VALES: I’ve got a pink card up, you want me 20 to start off? From our side of it, the role of 21 technology I'm going to say that's played the most 22 significant role over the past decade has been clearly 23 the ability for market participants to post live 24 inventory. 25 I have always been vocal in that I'm a big 0032 1 believer that the RTRS being able to have real time 2 pricing enabled dealers to post live content, not so much 3 because it was used for pricing, but because it changed 4 the dynamics in the market. 5 By that I mean that in the old days when you 6 purchased a security, it was very easy to take that bond 7 in and not post any one where that trade took place and 8 then mark it up by whatever amount you wanted to mark it 9 up. 10 That lack of information gave dealers a lot of 11 control in terms of the pricing that they wanted to present 12 to the market. 13 With the advent of real time trade reporting, 14 what happened was everyone basically knew the cost basis 15 of a trade so when a dealer bought a position, there used 16 to be a thing called the "QT," which was a quiet 17 situation, those ceased to exist. 18 Now every time a trade took place in the 19 market, almost immediately the market was aware of where 20 that bond had just traded, and therefore, knew what the 21 price of that security was. 22 We clearly saw on our part then that the 23 ability for dealers then -- because they recognized they 24 were long this inventory and they didn't have any 25 secrets, we saw the posting of live content become much 0033 1 more prevalent. We saw many more dealers start to 2 participate on the electronic Exchanges. 3 One of the peculiarities I would say of 4 transparency as well is it almost caused this asymmetric 5 pricing, if you will. That is that because traders knew the 6 cost basis, they very rarely were going to pay 7 significantly more than the last trade and I think that 8 helped, and I think the study shows that that helped bring 9 spreads down a lot. 10 In addition to that, when the market started to 11 run, they knew that a trader would have profits in a 12 position, so it was much easier still to say hey, I know 13 the guy's got money in it, I'm going to bid tighter. 14 On the flip side of it, when the market started 15 to come off, they knew that they had losses in it, and you had 16 to mark down your bonds significantly then for dealers to 17 step in. It was easier where you had to mark up less on 18 the up side, more on the down side. 19 The other thing is I think the advent of 20 technology or the electronic trading made bonds more 21 accessible. We said that when we started the company, we 22 really were just trying to make the process of buying 23 securities more efficient, and we never really wanted to 24 say it would expand the client base. 25 But clearly, what happened was that a lot of advisors 0034 1 who used to use mutual funds or started to use ETFs, the 2 ability to trade a customized portfolio became so much 3 more efficient. That process that used to take an hour to 4 build the ladder portfolio can now literally be done in 5 minutes. We started to see greater participation by the 6 advisor community as well. 7 The final aspect of that is a lot of the on 8 line brokerage community, once the ATSs became more 9 established, started taking connections or taking direct 10 line feeds of our content and making municipal securities 11 directly available to retail, and that is actually a 12 fairly significant component of the overall market now. 13 MR. HEMPHILL: From our standpoint at Regional 14 Brokers, we are traditionally a voice broker like the 15 ATSs, so over the past ten years, we have evolved from 16 being a traditional voice broker where we went out to the 17 marketplace to solicit bids on bids wanted or show our 18 offerings in the marketplace, to where now through a 19 third party vendor, Fab.com, we get 70 percent of our 20 bids coming in on bid wanted’s on a daily basis 21 internally, which is huge for two reasons because it has 22 openedup our exposure on our bid wanted’s to firms that we 23 had no idea ten years ago even existed, and it's freed us 24 up on the phone. 25 The one thing it hasn't changed in muni's in my 0035 1 opinion is we deal with at Regional Brokers bonds of 100 2 or less, and the problem with odd lots are the retail 3 traders are very busy internally, and being that they are 4 very busy internally, they don't have time to bid the 5 Street all the time. Their priority is to bid internally 6 and worry about the Street secondary. 7 That poses a challenge for an inter-dealer 8 broker when we get a list of ten items that are all five 9 and ten bond lots because the retail trader doesn't 10 necessarily have time or the priority to bid our list. 11 So what we need to do as opposed to including 12 advertising bonds on our website and Bloomberg, we are on 13 the phone continuously reaching out to these traders with 14 known interest to bid these bonds. 15 It can be very challenging in the marketplace 16 due to time constraints, and the contraction on those trading 17 desks. It still in my opinion is a big challenge, and one 18 of the biggest challenges still out there are five bond lots 19 because a lot of firms will not bid five bond lots on the 20 Street any more due to compliance reasons, so they just do 21 not bid them. 22 If you had five years ago ten firms bidding 23 five bond lots, they won't touch them now. So now, those 24 prices are being depressed and we have to work twice as 25 hard to go out and get as close as you can to the right 0036 1 bid, say 50 bonds traded in the morning at 98 and we have 2 five out in the afternoon, quite frankly, that bid could 3 be 95. 4 MR. RAMSAY: Could I ask you to stop there just 5 a second because you said for compliance reasons, and I'm 6 kind of interested to understand that a little better. 7 MR. HEMPHILL: Of course, I'm not on the 8 trading desk but the feedback I get from the traders in 9 the secondary market is that in order to bring in five 10 bonds, there's a transaction cost and if they want to bring 11 them in at 98 and Compliance comes in in the afternoon 12 and says well, you know, bonds just traded at par, you 13 need to mark those bonds back down. It's just a 14 headache so they just chose to stop bidding them. 15 The net result of that is when we do have these 16 five bond lots out, they are trading two or three points 17 cheaper than they should be because of the lack of liquidity 18 in the odd lots. 19 As it gets higher, if you can find somebody who 20 has matcher's to those five bonds, you're probably okay, 21 but it's just a challenge as an inter-dealer broker to go 22 out and solicit bids on the odd lots, because it's a game of 23 getting the trader's attention and that's the biggest thing 24 we do, you know it's great that we use Fab.com to advertise 25 our bid wanted’s, and they do a very good job bringing the 0037 1 bids in but we spend on one item we spend sometimes 15 to 16 2 phone calls on a five bond lot just trying to find the right 3 bidders to compare against the MSRB trades as well as our 4 internal bid history. 5 If I can say one thing, the secondary market 6 for odd lots is still an ongoing challenge in the muni 7 market. It's every day. We do get it done obviously, 8 but it is a challenge getting retail traders to bid on a 9 lot of these odd lots. 10 MR. RAMSAY: Craig, were you interested in 11 adding to that? 12 MR. NOBLE: As an user of both broker's brokers 13 and ATSs, just a couple of comments. Both are very 14 usable and for different reasons we use almost every ATS 15 and almost every broker's broker because as both these fine 16 gentlemen will tell you, the marketplace is such that if 17 you don't use them, you may miss a market and the problem 18 we have is just the technology to try to bring in -- it's 19 one thing to do voice communication with a broker's broker, 20 that's pretty easy but to bring in the different ATSs onto 21 a desk becomes time consuming and we have been working on 22 it now probably for three or four years and finally have 23 gotten full connection with, I think, most ATSs in the 24 marketplace. It has not been easy. 25 I want to make a couple of comments about the 0038 1 marketplace as a whole, going back to Dr. Hollifield's 2 presentation. 3 We kind of lose track of the fact that the 4 fixed income markets are not the equity markets. The 5 equity markets historically have been a trading 6 marketplace. People buy and sell equities at a fairly 7 rapid pace. 8 I'm not saying there are not buyers and 9 holders, but when you look at the fixed income markets, 10 we are trading securities that sometimes are 30 years 11 out, sometimes are anywhere between a couple of months 12 and ten years out, but the average maturity we sell on my 13 desk is anywhere between 10 and 15 years. 14 And with that, as much as the transaction cost may 15 be higher on odd lots, if you amortize that transaction 16 cost over the length of the time that an average customer 17 owns those bonds, it's de minimis and we do lose that factor, 18 the fact that when we buy and sell, there are higher 19 transaction costs for a five bond piece than a 100 or a 20 million bond piece, and we know that. 21 It has been extremely tough now. I have a 22 number of traders. We do 1,500 plus bid-wanteds a day 23 internally alone on my desk and then if we get a phone call 24 on a five bond piece, we have to stop, look to see if we want 25 to bid that five bond piece, if it comes in on an ATS, we 0039 1 have to look on the ATS to see if we want to bid on that 2 piece. It's a hard job and I'm not sayint that the MSRB 3 has done an excellent job posting prices, and I think prices 4 have come down just because of real time pricing, but there 5 is a significant difference in our marketplace. I don't want 6 to lose sight of that fact. 7 MR. RAMSAY: Appreciate that. Yes? 8 MR. BOWDEN: I'd just like to add to that as 9 well, along with what Craig is saying, the discussion 10 about the difference between institutional pricing and 11 retail pricing. Sometimes I worry that it's not fully 12 understood what goes into that retail offering. 13 An institutional offering, when you complete 14 the transaction, you send out a confirm, maybe one 15 monthly statement, and the security is moved off your 16 books and it goes to that institutional buyer. 17 Along the lines of the 15 to 30 year bond that 18 an individual buys and you look at that mark-up and what you 19 have over the entire life of that bond, we not only send 20 out a confirm, we have to send out monthly statements, we 21 have to send out monthly pricing, we have to provide 22 portfolio analysis capabilities to those brokers. You 23 have to track interest payments. You have to look for 24 any extraordinary effects to that bond over the life of 25 the bond. 0040 1 So there is an added expense to that retail 2 marketplace that you don't typically experience in the 3 institutional marketplace. 4 MR. RAMSAY: Thank you. Appreciate it. 5 Perhaps I could drill down a little on the point that was 6 made about difference between equity and fixed income 7 markets. Certainly, we appreciate those. 8 In the equity markets, the use of alternative 9 trading systems as we know has been used in general terms 10 to aggregate trading interest and also to promote 11 competition among quotes from various kinds of dealers. 12 In the sense that I have, and I think this is discussed 13 some in our municipal report, in the municipal market, the 14 use of ATSs, and I guess broker's brokers, both, has been 15 much more targeted and limited, that there are a large 16 percentage, a significant percentage, of trades that are 17 done, relatively small retail trades that are done through 18 the ATSs, but a very small percentage of the total volume, 19 and very limited availability of price information that 20 may be available through the ATSs to anybody other than 21 other dealers, perhaps in some cases other institutional 22 investors. 23 From that perspective, according to some, the 24 ATSs have been used in a much more limited way to help 25 sell dealer inventory but not as a mechanism to kind of 0041 1 provide more broadly price information or price 2 competition among participants in the market. 3 Is any of that a fair characterization? Even 4 if it is, are there reasons why ATSs ought to be 5 performing a very different kind of role in this market? 6 Yes? 7 MR. AUWAERTER: Hi. I work for Vanguard, a 8 large mutual fund organization. We certainly look at 9 ATSs. The problem with them when you're dealing with 10 larger size is you leave a footprint in the market that 11 disadvantages you. 12 You don't want to be in a situation where you 13 are putting a large block size out for a bid on an ATS, 14 because frankly, every dealer and institutional investor 15 will know it, and your price will drop. That is sort of 16 the irony of the fixed income market, people try to set 17 up against you. 18 It's better to try quietly to go around to 19 dealers where you believe they have an axe to take on 20 those securities if you're selling or if you're out there 21 looking to buy ones that you think are from a sectoral 22 perspective more experts on that sector in terms of 23 bringing out the securities because if you advertise it 24 to the world, you just destroy your good price execution. 25 MR. THOMPSON: Ben Thompson. I work with 0042 1 Samson Capital Advisors. We are a money manager working 2 with individual clients in separate account structures so 3 a little bit different from the Vanguard model. 4 Interestingly, I think the large block trading 5 on ATSs may be part a function of what Robert said, but 6 honestly, it has more to do with anonymity, so you may 7 not want anonymity when you're transacting in a large block 8 size because there is market demand for it, so you know 9 where the market appetite is, and you want to basically 10 maximize the value of that transaction for your client, 11 buying or selling. 12 So you may not want the anonymity whereas with 13 smaller trades, you may have more ability to transact 14 over a longer period of time, so the ATSs allow you to 15 have –- would allow a dealer to have an inventory up for 16 weeks because it could take a long period of time to sell 17 numerous small lots. 18 One of the things we haven't touched on is the 19 fragmentation. Dr. Hollifield's presentation spoke to 20 this, but when Craig talks about 1,500 trades or 1,500 21 bid-wanteds, it's not 100 issuers 15 times in those 22 bid-wanteds, it's 1,500 different issuers, 1,500 23 different obligors so you're talking about very 24 dissimilar instruments in incredibly large numbers 25 and the difficulty of the market to digest the 0043 1 information, assess the risk, make credit decisions, and 2 then achieve some measure of pre-trade price 3 transparency, even on the institutional trading desk, is 4 actually pretty difficult. 5 The challenge with the ATSs and the broker's 6 brokers even in the institutional world, we take in as 7 much information as possible to accelerate the credit 8 decision and the price discovery, but the sheer number of 9 dissimilar items is one of the most challenging pieces. 10 MR. VALES: I think also you have to look at 11 the structure of the market again, and that when you look 12 at the institutional market, it's really in terms of the 13 volume of trades, it's small. There is approximately 1,000 14 inter- dealer trades a day that are taking place of a 15 million bonds or more. That's it, in an overall market 16 where there is approximately 35,000 trades. 17 The reason ATSs do not have as much business on 18 the block side is the benefits of having an aggregated 19 marketplace and bringing thousands of users together just 20 isn't necessary in the block market. That market is 21 small. You don't need a sales group of hundreds of people 22 to cover the institutional accounts. That market is 23 still largely voiced because of the dynamics of the 24 marketplace. 25 There is always this comparison, why aren't 0044 1 muni's more like equities. I think it's pretty clear 2 from the panel thus far that clearly the number of 3 credits in the municipal market is drastically larger. 4 There are a million different CUSIPs versus the 6,000 5 listed stocks. 6 For an ATS, one of the things we love about 7 muni's from a technology perspective is they are very 8 easy to maintain. Apart from having to get the terms and 9 conditions, which is a whole other set of issues, the 10 pricing of muni's is very simple to maintain because on 11 average, they only update about three times a day. 12 That's it. 13 You compare that to BABs, taxable muni's, which 14 are priced against Treasuries, they are updating hundreds 15 of times a day. 16 So the characteristics in the market are just 17 incredibly different. The other thing is we have about 18 50,000 line items of inventory. Again, compare that. We 19 are a small ATS. Compare that against the 6,000 listed 20 stocks. We have 50,000 line items with over 275 firms 21 trading on a daily basis, that is a lot of 22 diversification. 23 MR. RAMSAY: I think I have noticed the first 24 flashing of the pink card. Commissioner Walter? 25 COMMISSIONER WALTER: Thank you, Director 0045 1 Ramsay. As I look back over my tenure as a Commissioner, 2 and I know Commissioner Gallagher feels the same, I feel 3 that in this arena, I keep having to reassure people that 4 we get it, that the fixed income markets and the 5 municipal securities markets in particular, but fixed 6 income markets are not like equities and 7 we are not trying to impose an equity market 8 structure on fixed income. Our questions are given the 9 characteristics that you describe, and we are well 10 familiar with them, what can we do to have this market be 11 both more efficient and more effective, and to me, most 12 important, fairer and I think there are some low hanging 13 fruit that at least we should consider. Some of that is 14 represented in the muni report. 15 But the ultimate question is we need to understand 16 all these characteristics and try to figure out how 17 either under the circumstances that exist today or with 18 changes in business practices, we can get this market to 19 function better, not like the equity markets, but to 20 function better because the more I look at it, the more 21 concerns I have about fairness in this marketplace and as we 22 certainly have seen very clearly across the board since the 23 financial crisis once investors lose their confidence, we 24 have a real problem, and there are real characteristics here 25 that lead me to worry about investors being in exactly 0046 1 that situation, in particular the fact that I think 2 there is somewhat of an evolution. 3 I know there is not as much trading, but 4 clearly there is more of an evolution for people to not 5 just buy and hold but to sell at various times and for 6 various reasons. 7 I think as difficult as it is for investors as 8 they purchase these types of securities, it's even more 9 difficult when they sell. 10 I think what we need to figure out with great 11 creativity and I think we can assure you we don't have 12 the answers, that's why we are here, is how do we wrestle 13 these factors together and come up with some constructive 14 ideas. 15 We are also, I think, not looking for nirvana. 16 We are looking for better than what we have today. Maybe 17 after that, we can get better than that. 18 So, more an editorial than a question. 19 MR. RAMSAY: If anybody would like to respond 20 or riff off of that. Yes? 21 MR. HEMPHILL: I have a couple of ideas around 22 that. One of the things I see every day, if we have 100 bid 23 wanted's out on a daily basis, just take a number, say we 24 have 100 out, we only trade 27 percent of our bid-wanteds 25 so that means there are 73 high bids on 73 CUSIPs that 0047 1 the information goes nowhere. 2 You know the high bid, maybe the cover bid, 3 maybe the third bid will get posted on that information, 4 where they stood, and use that information. The rest of 5 the marketplace doesn't use that information. 6 And as a muni's broker's broker, we are natural 7 aggregators of this information so what if we could 8 collectively use that information for the betterment of 9 pre-trade price transparency. For instance, what if we 10 could repack that information, you could send it down 11 to the MSRB so if I'm a private investor and I own ten 12 Pennsylvania GOs, I can see the MSRB trades, but that's 13 really it, away from me going to my broker, to their 14 trading desk, to the whole process. 15 What if I were to pull up either right on my 16 statement or go to EMMA and put that CUSIP in, not only see 17 where it's recently traded, but it's been out for the bid six 18 times over the past three months and maybe hasn't traded. 19 But if it hasn't traded, it at least gives me a 20 real expectation of where that bid is going to come in as 21 opposed to my broker coming back and saying well, I can 22 pay you 98 for your ten bonds, but my statement for the 23 past four months has said they are worth 105. That's 24 when retail gets upset and the SROs hear about it. 25 So, what if using our bid information, I think 0048 1 that will give retail much more of an education as far as 2 what their bonds are worth and more knowledge, and lower 3 the expectations or understand the expectations when those 4 bids come back because a lot of times, we'll have a CUSIP 5 out six times over six months, and I track it, and they 6 haven't traded at all. 7 If you are an investor and you own that CUSIP, 8 you are just getting that one high bid but you don't 9 understand it's been out for bid several times over, and 10 that is the market, but it just hasn't traded. There is 11 really nothing to compare to away from getting the bid 12 from your broker in a trade several months ago. 13 COMMISSIONER WALTER: I think that is a 14 suggestion that is well worth pursuing. I clearly don't 15 have the expertise to know how difficult it would be to 16 accomplish. 17 You raised another question which is valuation. 18 Do we need to take another look at how valuation appears 19 on the statement, is that a false sense of confidence 20 when the investor is looking at an 105 when all they can 21 get is a 98, or how do you tackle that issue? 22 MR. RAMSAY: Yes, Craig? 23 MR. NOBLE: We try to tackle it. We have the 24 same problems. The difference is when I started in the 25 business back 32 years ago, you priced your bonds once a 0049 1 month and it was pretty easy. It wasn't any better, but it 2 wasn't as ongoing a problem as it was before because you 3 only had to price your statement. There was no on line 4 systems where you could see your price every day. 5 A large part of our problems come back from the 6 fact that everybody has on line statements now so you have 7 customers who are tracking the prices of their bonds like 8 their equities on line day after day, and seeing either 9 some movement, no movement, or to your point, a full 10 sense of my bonds are worth 105 but I get a 98 bid. 11 We have had the two pricing services, and I see 12 Larry looking down. We have had these conversations 13 before at different meetings. It's tremendously hard for 14 them to price a million CUSIPs in a municipal world 15 daily, and that is what they do, without using some kind 16 of matrix that will have significant price variations 17 versus what a real bid is for a five bond piece or a ten 18 bond, a hundred bond or a million bond piece. 19 I think the idea about bid wanted's is one we 20 have talked about. Just to give you an idea, it isn't 21 any different in my world than your world. We do about 22 1,600 bid wanted's a day. Our average is about 35 23 percent of those getting hit. 24 It isn't like we send out everything out to the 25 broker's brokers or the ATSs, just to see what pricing 0050 1 is. We have our own customers doing it, our own bid 2 wanted system also. 3 It is something for us to look at because he 4 does make a very good point, because if my 1,500 bid-wanteds 5 of which let's say 1,000 are going unsold, those 1,000 6 prices are not getting into the MSRB system, and maybe 7 they would help with price dissemination. 8 It also goes back to -- we have talked about 9 this a number of different times -- client and broker 10 education becuase prices are available. EMMA has done a 11 pretty good job of putting out where the prices are. I think 12 a large part of the public just doesn't know they are out 13 there. We make that pricing available to our brokers 14 when they buy a bond, it's up on the screen. 15 We have looked at the hit rate. It is not as 16 high as we would like it to be, because you would figure 17 that a broker would want to see exactly where something 18 had traded with before he purchased it. 19 We are in the process of doing a broker 20 education program on that, but it's basically trying to get 21 the customers and the brokers to know where the prices 22 are becuase they are available. 23 MR. EDELMAN: I'm glad you mentioned the notion 24 of broker and particularly consumer education. That's from 25 my perspective as an advisor to the retail marketplace, the 0051 1 real key element that you have also touched on, Commissioner 2 Walter, is that from the consumer's perspective, I 3 appreciate all the dilemma's you have got and the 4 challenges dealing with the complexities of the 5 marketplace and the incredible distinctions between the 6 bond market and the equities market, but at the end of 7 the day, consumers are being lied to. 8 They do not understand two fundamental facts. 9 First, that the yields on bonds are negotiable at the 10 retail level, that when they are told by their broker, 11 oh, here's a muni I've got for you and here's what the 12 yield is, they think that yield is in stone. 13 They don't realize the yield is a result of the 14 pricing that the brokerage firm is creating and it is a 15 negotiable security and can be negotiated and in fact can 16 be shopped by going to another firm where they might get 17 a better price, resulting in a better yield. That is the 18 first element. 19 The second element is the monthly statement. 20 It is grossly inaccurate. You are right, it will say on 21 the statement 105, they go to trade it for 98, and we 22 understand why that happens, there is no deceit intended 23 by the industry, but from a consumer's perspective, they 24 are shocked to discover that what they thought was 105 is 25 in fact 98, and they don't know that until after they 0052 1 have sold it and have gotten the confirmation. 2 We have huge education and disclosure problems 3 within the industry, and we are, I think, are forgetting that 4 at the end of the day, these million issues and the huge 5 number of trading you are doing in the round lots, et 6 cetera, they are down at the retail level where somebody 7 is buying a $5,000 bond or a $10,000 bond, they are not 8 investing tens of millions at the institutional side. 9 These are ordinary, routine investors. 10 As you pointed out, Dr. Hollifield, 70 percent 11 of the market are from individuals. These are retirees, 12 these are low risk investors, these are income oriented 13 investors, and they have no idea what they are doing and 14 how the game works and we need to not lose focus of those 15 two facts. 16 MR. BOWDEN: Can I add something? There were 17 a couple of general statements made there, lied to and 18 grossly inaccurate. 19 MR. RAMSAY: I thought that might provoke some 20 interest. 21 MR. BOWDEN: Our firm has been in the municipal 22 bond business for over 80 years this year and we have been 23 dealing with municipal bonds for that entire period of 24 time. I've been there for 33 years and we have both 25 issuers and investors that have been clients of our firm 0053 1 for longer than I've been there. 2 It's not totally accurate that that exists at every 3 firm and I think we need to make a distinction of that. We 4 go to a lot of detail, especially on our statements. We 5 will get five point variances from the pricing services 6 where we have to go in and manually try to look and see 7 what is a fair price to go on those statements. 8 We trade a lot of non-rated paper in Arkansas and 9 it doesn't mean it's a lower credit. A lot of our school 10 bonds, which are the highest rated credits in the state, 11 are small, but they are non-rated, so they carry a much 12 higher valuation than what the pricing services will come 13 back with, because they are primarily looking at just 14 maturity and non-rated. 15 But we do go to a lot of extra effort to try to 16 get the statements as accurate as we can. It has to be 17 understood that that is just a snapshot in time, at the 18 time the statement is printed. 19 And we do an awful lot of work with our high net 20 worth brokers to make sure that the information they are 21 providing to their clients -- they have the most accurate 22 and available information in the marketplace, because we 23 do take a lot of pride in those offerings. 24 MR. AUWAERTER: Just to that point about education 25 and information. I think it's important for the retail 0054 1 investor to realize it's not the bond's price that is 2 most important, it is the yield, because you could have a 3 situation where they may have bought the bond two/three/four 4 months ago and there had been a substantial change in interest 5 rates, and that could be at least part of the reason for the 6 discrepancy in price from when they originally bought it. 7 I think what will be helpful, and again, this 8 comes down to education and disclosure, is to maybe 9 provide a benchmark at the time when the investor bought 10 the security, say an AAA benchmark, and show the yield 11 differential between that and the yield at purchase, and 12 then they can try to look and see where that bond is 13 trading, if they are thinking about selling it 14 two/three/four/six months later. 15 That said, pricing bonds in a municipal market 16 is definitely an art, not a science because remember, there 17 is one big difference between muni bonds and corporate 18 bonds, and that is the call feature. 19 Years ago, when I got in the business in the 20 late 1970s, corporations issued a lot of bonds with call 21 features. They don't issue them now. The reason being 22 because it's an institutional investor market. There were 23 tools such as calculators, option adjusted spread calculators, 24 that made institutional investors realize that the call 25 option they were effectively selling, they weren't 0055 1 getting paid enough for. 2 Now every corporate bond comes as a bullet 3 maturity, that is not the case on the municipal side, and 4 that is because of the fragmented nature of the market. 5 That is why call options still exist and they are tough to 6 figure out. 7 So even with my suggestion I just talked about, it 8 is really hard to price a bond with a tough call 9 structure. My traders spend a lot of time on it, and it 10 is definitely an art, not a science. 11 MR. RAMSAY: Thank you. Yes? 12 MR. BONOW: A couple of other quick comments. 13 We represent the issuer side of this grand equation and 14 see about 900 or so transactions a year. I think we have 15 some evidence that the fluidity of the market and the 16 availability of price information, really the RTRS made the 17 biggest single move in recent times, because we have some 18 evidence that when more monoline bond insurers were 19 active, it was a much more commoditized marketplace. 20 You could look to that period of time and then 21 look at 2005 when there were eight, I think, AAA rated 22 insurers back in 2005, all quite active each in their own 23 way, but we had this great price dispersion and then we had 24 a change in how prices are reported and things changed 25 rather dramatically. 0056 1 I think that is a good piece of evidence that 2 we can look to. In the suggestion bucket, I would place 3 perhaps the following, roughly about half the 4 transactions that we help our issuer clients bring to 5 market as their advisor are priced through a negotiation 6 and that negotiator, that pricing wire, is resident in a 7 service called Dalcom that is not widely available to the 8 retail marketplace certainly, and it may not be widely 9 available to even some state specific brokers. 10 So that could make another column, if you will, of 11 price information very live, very real price information 12 available to folks before those bonds are actually free 13 to trade or as they are free to trade. That is perhaps 14 something to consider. 15 I think at the end of the day, the advent of 16 ATS or what have you, it will help liquidity. That is 17 good for the cost of capital side of the equation, namely 18 the issuer side of the equation. As we try to marry what's 19 going on in the secondary market with what's happening in 20 the primary market, I think the way that retail investors 21 access information and the fact that we really have an 22 amalgamation of state specific markets in the muni world, 23 it's a lot of state markets that come together in many 24 ways. That needs to be recognized as well because you 25 have some brokers that are active only in a particular 0057 1 state. You have some Funds that are state specific. 2 You have some investors that access information just 3 historically within their state, because they can't 4 get an exemption, at least at the state tax level, from 5 offerings from other states. 6 And so we have to think somewhat myopically about 7 what's happening in some states as well when we try to 8 bundle the market together. 9 MR. RAMSAY: Commissioner Walter? 10 COMMISSIONER WALTER: A follow up question to 11 that. In my opening statement, I raised the issue, and I 12 would love to get your views on it, how do you feel about 13 the impact of the way the secondary market works on the 14 primary market offering process, and in particular, 15 pricing in the primary market? Is there an impact? Is 16 there a direct correlation? 17 If we were able to get greater transparency and 18 a better and more robust price discovery mechanism in the 19 secondary market, would it help in the primary market as 20 well? 21 MR. BONOW: I certainly think it has to help. 22 The question is how much would it help. Again, from our 23 clients' perspectives, anything helps, that gets bonds 24 that are eventually going to be 70/75 percent in a retail 25 either direct or indirect holding, a long term buy and 0058 1 hold mentality, the quicker they can get there and the 2 quicker that eventual yield, that cost of capital can be 3 translated back and benefit the issuer, would certainly 4 be an improvement that the issuer community would 5 welcome. 6 I think part of the challenge is that the way 7 muni bonds also are sold, whether it's through a 8 competitive sale, when people don't have 100 brokerage 9 relationships as retail buyers, they have probably one, 10 and if that retail buyer isn't active in the syndicate 11 for that competitive sale through pre-sale offerings, or 12 isn't in a negotiated underwriting syndicate, they won't 13 have immediate access to bonds so there will be some 14 interchange two or three turns before they actually 15 receive the bonds. 16 The relationship between the individual 17 investor and the broker as opposed the 28 or 38 largest 18 institutional investors that are on everyone's speed dial 19 and they have access to all these systems, it's a 20 different dynamic in terms of how they will eventually 21 see price and how and when they will see it, and as that 22 translates back into the cost of capital for the issuer, 23 it's kind of cloudy. We were talking about opaqueness in 24 this, but it's going to be some time before that price is 25 actually realized by the retail buyer because they have 0059 1 to wait for it to hit their broker's desk or their 2 broker's system before they in turn say yes, I'm 3 interested. 4 I think we have pre-sale retail offerings are 5 the best way to get as many brokers as possible engaged, 6 and competitive sales that extend for multiple days, so 7 you can actually access this very, very large but diffuse 8 marketplace, are ways to possibly address that. 9 That is where I think the value would come from 10 the secondary market upstream to the primary market. 11 MR. RAMSAY: If we could perhaps to pivot back 12 a little bit and incidentally, again, encourage any of my 13 colleagues on the staff to feel free, even if you don't have 14 a pink card, feel free to raise your hand and I will call 15 on you. 16 To return a little bit to this theme of the kind 17 of bargaining power leverage that retail investors do or 18 don't have, it struck me, even assuming retail investors 19 don't know they can negotiate for price, then the 20 question does become if they did know, do they have an 21 adequate basis or what could be done in order to give 22 them an adequate basis to actually use that authority. 23 I thought there were some helpful suggestions 24 in terms of bid wanted information. Presumably, they can 25 go to the MSRB system to find out if there have been any 0060 1 kind of recent transactions on a particular bond, but 2 given the characteristics, it is probably unlikely that's 3 the case. 4 Is there other information that either is 5 available now or could be made readily available in order 6 for retail investors actually to use the ability to try 7 to negotiate for a better price? Yes, Ben? 8 MR. THOMPSON: Thanks. That is exactly the 9 right question to ask. I don't know that providing more 10 information -- providing more information is certainly 11 good. It will inform the primary and secondary processes 12 and will enrich the market. 13 The challenge is the volume of information is 14 so high right now, interpreting it is the difficulty. 15 You commented that taking the information from RTRS and 16 seeing a specific bond trade is unlikely, but the key is 17 you have to be able to look at groups of bonds that have 18 similar characteristics and understand what the relative 19 risk premia is for that category. 20 That is, of course, what an institutional 21 trader does when he or she assesses the value of a 22 security. Being able to quickly check disclosure 23 patterns and use things that are in the SEC report that 24 have been discussed obviously here, and being able to 25 look at groups of similar securities that have traded 0061 1 recently, and understand where the security you're 2 looking at measures relative to those. 3 Yield spread is another component. Having 4 indexes and benchmark curves available on them. EMMA, 5 obviously, has accelerated the information available to 6 investors exponentially, so RTRS started it. 7 EMMA delivers infinitely more information now 8 to individual investors, but it's subject to 9 interpretation. There has to be better guidance as to 10 how to understand the information that's coming in and to 11 distill it into an usable information set. 12 MR. VALES: If I could add to that as well. 13 One of the characteristics again that we see in our 14 marketplace, in terms of negotiation, we do have 15 negotiation, we do support negotiation at both the odd 16 lot and the block level. 17 Approximately 30 percent of the trades that 18 take place in our application are due to negotiation and 19 interestingly, the percentage is tied to the size of the 20 trade, meaning that larger trades tend to be more 21 negotiated than smaller trades. 22 Similar to the comments that I think a number 23 of participants have echoed here, is that the issue that 24 you will always end up having in the odd lot market is 25 because retail traders are so overloaded with bid wanted 0062 1 flow, with trying to post their offerings, with managing 2 their books, that it's very difficult for them to constantly 3 be managing an negotiation on an odd lot situation. 4 So what we see, especially for a lot of the middle 5 to larger size firms, is that they have algorithms that 6 are in place and so if a bond is priced near market levels and 7 a bid would come in, it's going to be a system that's 8 responding to the quote, not a trader. 9 And unless it gets kicked out, which is generally 10 because it's a larger block size, that will never touch a 11 trader's hands. That process can be completely 12 automated. 13 Having said that, a lot of firms will say if 14 it's an odd lot position, they don't want to negotiate on 15 price that they don't have the time to manage that. I 16 completely agree with Ben. 17 I think the issue you have, and I always think 18 about this with bid-wanteds, is that a lot of times you get 19 participants' information, but it's difficult to 20 interpret that information, so I think about this all the 21 time. 22 We run approximately 3,000 bid wanted's on a 23 daily basis, approximately 20 percent of those trade. 24 For the bonds that don't trade, a lot of times we are 25 wondering is it because the bids weren't sufficient or 0063 1 was the seller looking for too high of a price and 2 really, that is where a market professional comes 3 in who can digest that information and make a 4 decision looking at comparable bonds, looking at the 5 overall market, looking at credit trends, to decide the 6 value of that security. 7 MR. BOWDEN: And one other thing,just to add 8 about whether the price is negotiable, again, I only have 9 the experience of our own private client brokers and our own 10 investment advisors, but I know those brokers know that 11 price is negotiable, and they are in fighting for their 12 customers, if they have to have a specific yield to 13 accomplish a certain goal. 14 So, it's not like if we own bonds in inventory, 15 it's posted, and that's it, that there is no room to talk 16 or discuss. We put a lot of emphasis, I guess, on that 17 broker's relationship with that client, and knowing what 18 is an appropriate yield and what is a needed yield that 19 they are trying to accomplish for that suitability. It 20 doesn't mean a trader can agree to that price, but it 21 certainly is negotiable. 22 MR. RAMSAY: I think I saw another pink card. 23 Commissioner Gallagher? 24 COMMISSIONER GALLAGHER: I was busy blaming 25 Commissioner Walter for the selection of hot pink. It 0064 1 seems like something she would have picked. 2 I'm going to ask a real basic question. It's 3 one I've been asking for years. I've never really gotten 4 a good answer to it, but it seems to me, we talk about the 5 differences between the equity markets and fixed income, 6 whether it be municipal or corporate, and we understand 7 that, as Commissioner Walter said, we get that. 8 In this quest though, to Commissioner Walter's 9 question, of how can we make it better, it seems like one 10 of the key differences, the amount of data, the amount of 11 information in your markets, is something we could actually 12 use to benefit retail investors. 13 One thing that I learned along the way here 14 when I was back on the staff is the incredible amount of 15 information that's in the CUSIP number itself that you guys 16 understand, maybe your proprietary systems understand, 17 but the average investor thinks it's just an identifier 18 number. 19 It's kind of like DNA. Is there a way for us 20 to uncoil this, unleash it, in a way, on EMMA, on the 21 trade systems, somewhere publicly available that could 22 help retail investors manipulate that data in such a way 23 to help them make better investment choices? 24 MR. HEMPHILL: Well, I think there possibly is. 25 We’re in talks with a group of inter-dealer brokers and 0065 1 ATSs if this could work, to pool our information 2 together, aggregate this information, not only for the 3 retail investor on their statement, as we discussed, but 4 it may help -- if you can pool the information that we 5 get from the other 73 bid-wanteds that don't trade today, 6 if I have 100 out in a given day, pick a number, that's 7 hard information as opposed to, I think, the frustration 8 out there with the third party eval services. 9 I'm not by any means talking them down, but 10 there is some frustration out there. I think it's a 11 fact. 12 But if we could aggregate this muni information, 13 that's hard information that either the retail investor 14 can use, the MSRB can use for EMMA, and you also could 15 use that for back office pricing of muni's, and also 16 could probably also help out the PCAOB with auditing 17 firms when they are auditing the broker-dealers or the 18 mutual funds that they audit because then they will have 19 hard information as opposed to the pricing services 20 which seem to be off, which is the feedback I’m getting. 21 We don't use pricing services, but I can tell you one thing 22 that has helped in the muni market has been Rule G-43. 23 G-43 has helped us out at Regional Brokers. We deal again 24 with 100 bonds or less. We deal with retail, regional retail, 25 up and down the Mid- Atlantic, the Midwest, and the Southeast. 0066 1 We deal in smaller lots and not so much in -- 2 We will deal on the institutional side, but 3 it's mainly taking an institutional block and selling it 4 to the retail traders for their retail. 5 So we're not involved with the ten million bond 6 lot trades a day. However, going back to G-43 and how it 7 has helped us, G-43 has brought a greater awareness 8 between myself and the traders or a liaison, as far as the 9 prices I'm giving them on their bid wanted's. 10 Fifteen years ago, the bids would come in. You 11 would send them to a liaison, to a trader, and it would 12 just be passed along. Now, it's early, I know, it's only 13 been about four months since this has been implemented, but 14 it is now if I get a bid wanted and I have these 15 pre-determined parameters, according to G-43, that are an 16 internal use tool for us to use, but it's based on our 17 recent bid history and MSRB trade history. Non-eval 18 situation, it's hard facts. 19 So if that bid is below that, there probably is 20 a reason that that's probably not the right bid, but then 21 I have a conversation with the trader or the liaison, stating 22 that this is below my internal parameters, which I just need 23 to let them know, according to the rule, but it's putting 24 them on notice that they need to look into it further and 25 most firms are really taken with that. 0067 1 So there is a greater conversation between the 2 two since G-43, and I think it has benefitted the market. 3 I think if we look back a year from now, I think we will 4 definitely see the net effects to that. 5 MR. RAMSAY: Craig? I think you were 6 interested in getting a point in. 7 MR. NOBLE: You make a very good point. I 8 think a large percentage of the investing public doesn't 9 know how much information is available if you have a 10 CUSIP number. 11 It goes back to, number one, broker education, 12 client education. It also goes back to what you make 13 available to your brokers, back to Larry's point. An 14 informed broker, I think, will help the investing public 15 more than anybody else because they are the front line. 16 They are talking directly to the client where the trading 17 desk is not. 18 I put my inventory in competition with the 19 Street. I put the full Street inventory up on our 20 brokers' screens along with my inventory. My inventory 21 is highlighted so they know whose inventory it is, but 22 they get to see the Street inventory right next to us, 23 giving them a good comparison of my prices are right 24 versus where the Street's are. 25 It also allows the broker to have direct 0068 1 information supplied from both EMMA and outside news 2 services we provide because continuing disclosure is a big 3 factor of the sale of municipal bonds, too, so we want to 4 make sure they have current, accurate disclosure 5 information to make that decision. 6 It's really making sure that the broker is 7 knowledgeable, knowing what he has available and what's 8 available in the marketplace and it goes back to what Ben 9 says. We also can do, I think, a better job of 10 aggregating information that you would have to be able to 11 look at like and kind bonds because again, we have all 12 talked about the diversity in the marketplace and it's 13 really hard to see if that five bond piece that may not 14 have traded for six months is at the right yield, but if 15 you knew that a state GO in a similar rated state, if you 16 had to look at an index, let's say, and that's been trading 17 at a four percent yield, you know you are within a band, 18 that you're going to get the right yield. 19 It isn't all about the price. We have talked 20 about this. It is about the yield. The yield is what we 21 need to look at because pricing concerns can be varied by 22 callability, what's out there in the marketplace and 23 what's happened over the last couple of days. 24 So I think it's a very good point. 25 MR. CROSS: Let me just follow up with one 0069 1 quick question, on exactly the point you were raising, 2 sort of the converse of getting more information from 3 CUSIPs. 4 In the monith (?) department, a theme we have 5 heard, I think, in the huge advances with EMMA in the last 6 few years, is to look more at improving the sort of user 7 friendly ability to not just go on EMMA and find a CUSIP 8 number and look at a bond, but sort of to improve the 9 comparability features, to be able to search batches of 10 comparable credit bonds on EMMA. 11 Do you think that is a good thing, to channel 12 improvements in that system? 13 MR. NOBLE: I think it's a very good idea. The 14 question, of course, is technology. The ATS we use, if 15 you have a broker looking, you can search by CUSIP on our 16 ATS or a system, if you want to look for a bond, or you 17 can search by security features, if you want an AAA rated 18 -- it doesn't exist too much any more – an AAA rated four 19 percent yield going out in 2035, you can then search by those 20 same functionalities. 21 I think you make a very good point. To your 22 point, we talk about CUSIPs but we really need to talk 23 about like and kind bonds and the ability to search in like 24 and kind bonds, and I think it something that exists in 25 EMMA, it just needs to be refined. 0070 1 MR. CROSS: I think one of the big concerns I 2 see in general in the market is with the total demise of 3 the bond insurer companies, there is much more emphasis 4 on comparability of comparing direct credit analysis and I 5 think we need to look at ways to make that easier, recognizing 6 the diversity of the market. 7 MR. RAMSAY: Yes, Ben? 8 MR. NOBLE: Very true. 9 MR. THOMPSON: That’s exactly the question I 10 had. I would add to it - it's predicated on the 11 education process taking place ahead of time so when you 12 use an enhanced search feature to group like bonds, you 13 understand that the credit rating, the maturity, the call 14 feature, sinking funds, and geography all play a role, so 15 the investor is educated about which toggles they have to 16 hit to group those similar bonds. 17 That is also an important part of EMMA but I 18 think you are getting to exactly the right place -- the 19 enhancements get an investor much closer to what a market 20 professional would use to make a decision, but it is 21 predicated on them investing the time and actually 22 understanding how these components affect valuation. 23 And that's important. 24 MR. BOWDEN: We also would support trying to 25 make it more user friendly on the EMMA system. You would 0071 1 know more than we would, but our experience is the system is 2 not used nearly to the extent that you would think it is and 3 I'm not sure why that is. I don't know if it's the lack 4 of usable information or if it's how easy it is to 5 actually get in and use the system. 6 Craig, I don't know if your experience is any 7 different. 8 MR. NOBLE: I think part of the problem is just 9 people knowing about it. I think the system is very well 10 designed and very easy to use. I mean, we have a direct 11 connect for our brokers to use it. 12 I think part of the problem is if you were to 13 do outside this room, which is all investor 14 professionals, how many people actually know EMMA exists. 15 That's part of the problem. I think EMMA is a great 16 system. It's the fact that a lot of people just don't 17 use it. 18 MR. VALES: We love EMMA. Kind of an 19 interesting aspect of EMMA is, what's really valuable 20 about it is, if we pull up EMMA, and we have links for EMMA, 21 it doesn't get clicked as much as when you have something 22 that's happened recently. 23 One of the things we do because of the time 24 stamping that EMMA provides is we highlight when there is 25 a recent material event out that heightened awareness is 0072 1 then what promotes EMMA. That is where we are seeing 2 much more use. On kind of a day to day basis for 3 somebody coming in and looking and doing an analysis, it 4 doesn't get nearly as much attention. 5 I think that is an important distinction in 6 terms of how EMMA could move going forward. 7 MR. RAMSAY: I’ll take one more question on this 8 line and then maybe talk a little bit about dealer 9 inventory. 10 MR. BONOW: I just had a couple of quick 11 suggestions with regard to EMMA. One is that what we don't 12 see in the municipal marketplace very often is reverse 13 inquiry. We don't have retail buyers certainly saying 14 hey, this issue is coming, I want a piece, and can I get a 15 piece. Yet, they are very interested in supporting local 16 issuers, they are very interested in knowing what's 17 coming from their state. 18 EMMA doesn't have much in the way of a forward 19 calendar, so they can say, oh, this bond election I voted 20 for last October is now coming next week so they don't 21 have a sense of kind of the follow through for issues. 22 I think it would help drive down costs for the 23 issuer because you would have more demand, up front 24 demand, and you can drive down some of the price 25 disparity or price dispersion because I think issuers 0073 1 will be closer, excuse me, retail investors would be closer 2 to the initial offering price. 3 MR. RAMSAY: Commissioner Walter? 4 COMMISSIONER WALTER: Just a quick comment 5 before you move on, because I, too, would like to hear 6 about inventory. 7 I think we completely agree about how much of a 8 step or steps forward EMMA has been and I do think it is 9 probably one of the few things that I have seen in the 10 last five years for which there is unanimity, and I would 11 compliment our colleagues at the MSRB who are also very 12 flexible in terms of a forward calendar for improving 13 EMMA. 14 I would suggest that one of the things that we 15 think about after this Roundtable is over is how we can 16 put the most effective educational campaign together. We 17 have heard that several times already. I suspect we will 18 hear it more. 19 To my mind, the most effective way to do that 20 is to have a public/private partnership so that we are all 21 aiming to do this together with perhaps the content 22 really sponsored by the neutral parties, the SEC and the 23 MSRB, but using the dissemination channels that are so 24 effective within the industry perhaps to get the word 25 out. 0074 1 There are probably lots of other ways to do it, 2 but I do think it is something we should put on the "to 3 do list." I think EMMA can become an addiction as soon 4 as you use it, and not a bad one but you have to get people 5 to use it in the first instance. 6 MR. RAMSAY: Yes, Commissioner Gallagher? 7 COMMISSIONER GALLAGHER: Just a real quick 8 follow up on CUSIP before we let you go on CUSIP. Is there 9 anything else we need to be focusing on here at the 10 Commission with respect to CUSIPs, their uses, their 11 constraints, the fees, or other things that you guys see 12 as a particular issue within the current market structure 13 that we need to focus on? 14 MR. VALES: Love for the fees to come down. I 15 also think it would be helpful for the Investing in Bonds 16 site to promote the use of root CUSIPs, so when an 17 individual is looking at a particular bond or likes a 18 particular security, that he can also see how similar 19 maturities or structures have traded. 20 MR. BONOW: Similarly, to Larry's point about 21 high grade but un-rated school district issuers in 22 Arkansas or something, which is emblematic of the 23 municipal marketplace, these small one off, once every 24 ten years, if you will, kind of issuer, something that 25 can tie through a search process the CUSIP number to like 0075 1 credits or like types of issues, as opposed to just the 2 issuer root and then it's all resident with that issuer, 3 and you may have five or six hits and not five or six 4 hundred hits. That to me would be helpful in broadening 5 the sense of like credits in the marketplace. 6 MR. RAMSAY: I would note that I'm in the realm 7 of kind of aggregating information, providing more user 8 friendly ways of bringing the information together. 9 On our afternoon panel, MSRB will be 10 represented. I believe they will be able to provide their 11 thoughts on how EMMA in particular might be enhanced or 12 improved in that regard. 13 We started to get a little bit into our 14 questions about dealer inventory, trends, dealers' 15 willingness to take on inventory. I want to make sure we 16 hit that. 17 Tom Eady on our staff, I think you were going 18 to maybe kick us off in that direction. 19 MR. EADY: I was curious to hear from Craig and 20 Larry as market makers, as dealers in this market, and 21 also from the investors, Ben, Ric and Bob, as to what 22 your experience has been recently in the dealers that you 23 traditionally do business with, their willingness to hold 24 inventory, the impact that may have on the liquidity that 25 you receive from them, in particular, and from the market 0076 1 in general, and whether you see a rise in the prevalence 2 of riskless principal trading. 3 To the investors on the panel, whether you are 4 aware if a bond is offered or is being bid, whether it is 5 on a true principal at risk basis from the dealer or on a 6 riskless principal basis. 7 MR. BOWDEN: Well, I guess the first question, 8 I would say in general, dealer inventories are less than 9 they have been historically. That may not be the case at 10 every firm but I would think in general they probably are. 11 A lot of that, I think, is attributable to the 12 use of the ATSs and broker's brokers. We, along with 13 Craig, we advertise our inventory along with others' 14 inventories. We try to give our brokers as much access 15 to whatever is available as we possibly can. 16 The term "riskless principal," I know we 17 discuss that a lot when we are in these meetings, but 18 when I have those discussions with my management at my 19 firm, they don't understand what "riskless principal" 20 means. They believe there is risk in every trade, which 21 we do as well. 22 I still think that is a difficult situation to 23 describe. The one thing that has changed over the last 24 ten years -- when I first got in the business, a 25 commitment to buy or a commitment to sell a bond was a 0077 1 given, whether it was an institutional, and in most 2 cases, individuals. If there was a commitment to pay for 3 that bond on settlement date or sell you that bond on 4 settlement date, that was going to happen. 5 Counterparty risk has changed the entire risk 6 management structure of all our firms. That is something 7 we have to monitor. We are being asked to pay more 8 attention to it by the regulators as well. 9 But a riskless principal trade can, the thing that 10 is maybe not known is there is still a lot of research that 11 can go in, maybe you know where a block of bond is, you 12 know where it's located, you know who is holding that 13 bond, you think you have a potential buyer for that, but 14 you still put a lot of research into that even though you 15 may go effect that trade in a single day at a given price 16 on both buy and sell. 17 I don't think you can just come up with a 18 blanket definition of "riskless principal." I think 19 there is a different type of transaction involved and 20 certain transactions, based on the amount of research 21 that had to go in to effect that trade. 22 MR. EADY: Do you think it would be helpful if 23 there were more definition around what a "riskless 24 principal trade" is? It sounds like you are struggling a 25 little with the terminology, what really is a riskless 0078 1 principal trade. 2 MR. BOWDEN: Absolutely. I'm not sure that a 3 true riskless -- maybe we need another word instead of 4 "riskless principal." 5 MR. NOBLE: I agree with Larry. The key about 6 a riskless principal transaction is yes, you might effect 7 it because you didn't own it in inventory, but once it's 8 on my books and I effect the trade, I've taken risk. 9 A client, retail client, has a certain amount 10 of comfort that if I sell them a bond, I'm going to stand 11 behind it. It goes back to everything else that goes 12 along with it, for the next 10 or 15 years, they want me 13 to make sure I maintain research on it, I've looked into 14 the bond, the continuing disclosures that happen, and 15 everything else that happens in that transaction. 16 As soon as I effect a principal transaction, 17 riskless or coming from my inventory, I'm on the hook 18 because it's going to go into a customer's account and 19 then I'm going to have to maintain that relationship going 20 forward. 21 Let's change it back to inventories. 22 Inventories are down. We have had a couple of different 23 events, I don't want to use the Meredith Whitney event, 24 you know, Puerto Rico happening right now, the fact that 25 the marketplace has changed dramatically because of what 0079 1 happened in 2008, the fact that Bear, Lehman, we have 2 caused a lot of it, by Wells, it’s taken a number of 3 names out of the marketplace. 4 They have been substituted by a whole bunch of 5 smaller broker dealers that have left those firms and 6 probably started, but there is a lot of large market 7 participants that used to keep the marketplace kind of 8 firmed up that are gone. It has affected the marketplace 9 as a whole. 10 My inventories are down mainly because my risk 11 managers, they used to see very little risk in muni's 12 until -- I keep going back to the Meredith Whitney 13 debacle, and when my inventories went down significantly 14 over that period of time and I took sizable losses, they 15 told me to cut back my inventory so I'm holding less 16 inventory. 17 I'm using the ATSs more for liquidity and for 18 inventory. My inventories are down about 20 to 25 19 percent versus probably the high's I would have but I'm 20 still owning like anybody else here a fair amount of 21 municipal securities on my books to offer to retail 22 clients. 23 MR. RAMSAY: To what extent is that affected by 24 the lack of or trend away from credit enhancement? Does 25 that make a difference in terms of your willingness to 0080 1 hold bonds in inventory, or is it affected mostly by 2 other factors? 3 MR. NOBLE: It has in some ways because as soon 4 as you have a credit event, going back to Puerto Rico, 5 used to be able to look at the insurers and watch them 6 stand behind the credits. Now you are looking at the 7 credit to stand behind the credit so we have to do more 8 credit work, back to Larry's point. We have to be more 9 considerate of what we own. 10 It takes us more time to actually -- if there 11 is a large bid wanted out there, we do a little more 12 research than we used to do because as we said, there 13 were eight insurers, the market was basically an AAA 14 marketplace, you knew what you were going to buy, you 15 were buying the underlying insurance company. Now you 16 are buying the underlying credit. It does take us longer 17 to make those decisions. 18 MR. BOWDEN: The spreads have narrowed, that is 19 obvious in the marketplace. The municipal bond market 20 inventory is almost impossible to hedge. You really 21 cannot protect the market risk. There is not a vehicle 22 out there that is a good proxy for municipal hedge. 23 MR. RAMSAY: So if inventories are down, I guess 24 one question is with the greater availability of ATSs and 25 other ways to source funds for customers, does the 0081 1 reduction in inventory make a difference, have an impact 2 in terms of liquidity for customers or not? 3 MR. AUWAERTER: From an institutional investor 4 perspective, it certainly does, with that reduced amount. 5 To add to the points about inventories going down, some 6 of it is due to regulation, with some of the banks being 7 forced to carry more capital. 8 I also hear a lot of discussion about what is 9 going to be the potential impact when the Volcker Rule is 10 passed by the Commission or other members of FSOC. There 11 is a lot of uncertainty there. That is also creating 12 concerns. 13 One thing that I worry about as a mutual fund 14 investor is how are investors going to react to a rise in 15 rates. It's going to come some day, maybe even in my 16 lifetime. At that point, when investors get either their 17 mutual fund statements or get their brokerage statements 18 and see at least an unrealized negative rate of return, 19 how are they going to adjust to that because we have a 20 lot more people who are in the market that years ago 21 would have been in a Money Market Fund or in bank CDs. 22 I know my organization and others do a lot of 23 education to try to explain that to people but I do worry 24 from an institutional investor perspective, we may all be 25 potentially trying to get through the door at the same 0082 1 time to sell, if there is a sharp spike in rates for some 2 unknown reason. 3 Liquidity, I would say, is the number one 4 concern at my shop. 5 COMMISSIONER GALLAGHER: More on the editorial 6 side than a question, as you saw Elisse and I here 7 shaking our heads in vehement agreeance with you. As 8 they would say to me, it's one of the really important 9 reasons why we are here today. 10 In general, the muni side is obviously hugely 11 important on the retail investors. We are seeing with 12 the $2 trillion in issuance on the corporate side last 13 year and increased retail participation, as folks seek 14 yield, this is a massive issue for the Commission. 15 We can debate around the edges what role the 16 Commission plays with respect to systemic risk, right, 17 that has sort of been the hot debate since the crisis, but 18 for things that are just so squarely within our 19 jurisdiction, this issue, the exit risk here to retail 20 investors when interest rates go up is probably the 21 biggest thing I worry about in my job. 22 I worry a little less when folks are invested 23 in The Vanguard Fund than they are directly, because I 24 think your institutional power, trading power, can help, 25 despite the fees to be paid. 0083 1 This is one of the key reasons I think we are 2 sitting here today. Any further thoughts that folks have 3 on that more macro issue along the way, I'd love to hear. 4 MR. RAMSAY: Ben? 5 MR. THOMPSON: Commissioner Gallagher, you're 6 echoing what Bob said, is a concern in the market. If you 7 look at the evolution over the last five years, we have gone 8 from a market that had bond insurance, unlimited amounts of 9 capital available in an arbitrage community, in 10 municipal bonds, that doesn't really exist any more. 11 If you look at those periods of systemic 12 weakness in municipals, there is an arbitrage community 13 that absorbed the excess supply. That doesn't exist now. 14 We have moved towards an end user market more 15 like what we were 25 to 30 years ago, where we probably 16 as a firm deal with more different counterparties than we 17 have ever dealt with. The top counterparties are still 18 the top, but they are a less significant portion of our 19 business, and the depth of their balance sheet certainly 20 feels substantially less, I mean just our models require them to 21 hold less because municipals have been more volatile, so 22 the ability to absorb a significant amount of mutual -- 23 fund flows which we haven’t seen in a significant period 24 of time, is completely uncertain. 25 COMMISSIONER GALLAGHER: Yeah, and again, this 0084 1 was my fear, exit strategy, interest rate environment, 2 hopefully it will happen at some point they go back up or 3 else we are in real deep trouble but you add onto it some 4 of the California bankruptcies, and this notion that the 5 bond holders are going to be wiped out so payments can still 6 be made into pension funds and things like that, boy, tack 7 that onto rising interest rates, and we have a real Armageddon 8 on our hands here. 9 I think it is something the Commission needs to 10 pay attention to, both of these issues, what role does 11 the protector of investors have with respect to bond 12 holders getting wiped out who thought they were buying 13 inherently risk free products. 14 MR. RAMSAY: Craig? Ben? 15 MR. THOMPSON: There are back stops. The tax 16 exempt client world, international investors, pension 17 funds, have historically been the buyers of last resort 18 for municipals so I don't know that we're concerned about 19 a systemic breakdown, that is really a market collapse, but 20 I think the significance of the volatility could be much 21 greater, the depths of the draw down could be larger, which 22 will be more punitive for longer duration investors obviously 23 coming out of a low yield environment. 24 I think the market absorbed the credit events 25 in California much better than we all anticipated. This 0085 1 is actually a credit to the individual investor from the 2 education process. 3 They came through the Meredith Whitney period 4 and over the following year, became much more educated about 5 fundamental municipal credit and much more sophisticated 6 in their view of the real risk they are taking, so that 7 the efforts on the part of improving continued disclosure 8 are timed well, because investors are more receptive. 9 Something like Puerto Rico is a different order 10 of magnitude. That's going to be a different event, and 11 the sustained rise in interest rates as we saw in the 12 fourth quarter of 2010, when the Treasury market began to 13 rise, municipal bond yields began to rise, any of these 14 declining outflows, mid-term elections, Meredith Whitney, 15 the discussion of state bankruptcy, we had all those 16 things compounding the draw down. It wasn't that bad 17 actually. It wasn't the end of the world. 18 MR. RAMSAY: I want to acknowledge that 19 Commissioner Aguilar has joined us now, and thank you for 20 that. Commissioner, I know you have some remarks to 21 give. We can wait until this segment is towards the end 22 or you can do them now, whichever you prefer. 23 COMMISSIONER AGUILAR: Thank you, John. I 24 assume we have already broken the train of thoughts. 25 Instead of doing it twice, I'll just give some remarks 0086 1 now. Let me apologize for being late. Under the law, 2 Section 19(d) of the Securities and Exchange Act of 1934, 3 the SEC and the North American Securities Administrators 4 Association have to meet once a year. I am the SEC's 5 NASAA liaison, so I was there this morning. I'm sorry I 6 am late, but I am very pleased to be at the Roundtable on 7 Fixed Income Markets. I strongly support the 8 Commission's effort to evaluate ways to improve the 9 transparency and efficiency of the fixed income markets. 10 Before I continue, I assume we have given the 11 disclaimer. Anything I say you can use against me but 12 you can't use it against anyone else. 13 (Laughter.) 14 COMMISSIONER AGUILAR: As an SEC Commissioner, 15 I consider the protection of investors, particularly 16 retail investors, to be my primary obligation. This is a 17 specific concern in both the municipal securities and 18 corporate bond markets, where retail investors play an 19 important role. 20 For example, retail investors are the largest 21 holders of securities in the municipal securities market. 22 According to recent statistics, retail investors hold 50 23 percent of municipal bonds directly, and another 25 24 percent indirectly through mutual funds, closed end 25 funds, and Exchange-traded funds. 0087 1 Today, retail investors hold approximately 75 2 percent, up to $3.7 trillion, of municipal debt issued. 3 In addition, as of March 2013, retail investors hold 4 approximately 28 percent of the total outstanding 5 principal value of the corporate bond markets. 6 The numbers make it clear, retail investors are 7 significantly invested in these markets. However, 8 although retail investors are significant participants in 9 both municipal securities and corporate bond markets, 10 they do not receive the same level of protection in their 11 respective markets. 12 From a regulatory perspective, the biggest 13 difference is that corporate bonds that are sold to the 14 U.S. investing public are generally required to be 15 registered with the SEC. While in contrast, there is no 16 expressed statutory authority that requires the 17 registration and reporting of municipal securities to the 18 SEC. 19 Simply stated, the municipal securities market, 20 despite its size and importance, has not been subject to 21 the same level of regulation as other sectors of our 22 capital markets. 23 The Commission's investor protection efforts in 24 the municipal securities market has been primarily 25 through the regulation of broker-dealers and municipal 0088 1 security dealers. In contrast with most other public 2 offerings, the Commission does not receive or examine any 3 offering documents related to municipal securities prior 4 to the issuance of the offering documents to the 5 investing public. 6 In fact, provisions commonly known as the Tower 7 Amendments expressly limits the Commission's authority to 8 require municipal securities issuers either directly or 9 indirectly to file any application, report or document 10 with the Commission prior to any sale of municipal 11 securities by the municipal issuer. 12 As stated in the Commission's July 2012 report 13 on the municipal securities market, and I quote, 14 "Investors in municipal securities are often not afforded 15 access to the types of timely and accurate information 16 available to investors in other securities." 17 Just last year, the Commission brought 17 18 enforcement actions related to misconduct in the 19 municipal securities market. This is more than doubled 20 the number filed in 2011. 21 As the 2012 report noted, many of the 22 Commission's enforcement actions related to the municipal 23 securities market "Involved materially misleading 24 statements and omissions and disclosures relating to 25 municipal securities." 0089 1 It is clear that a greater focus on this market 2 is needed in order to protect investors. However, for 3 both the municipal securities and corporate bond markets, 4 today's Roundtable is an important step. 5 As the primary regulator of the U.S. capital 6 markets, the SEC must continually seek information to 7 understand the current structure of our markets, and be 8 able to objectively assess how the markets are 9 functioning. 10 By facilitating a dialogue between regulators, 11 market participants, and other stakeholders, today's 12 Roundtable will be a step forward in the process. 13 In this regard, I note that the majority of the 14 panelists for today's Roundtable are representatives from 15 the financial services industry. In fact, out of the 27 16 panelists, I believe we only have four panelists who are 17 representing retail investors. 18 As I said at the Roundtable on Market Structure 19 in June 2012, it is critically important to have a 20 Roundtable that allows for robust discussion of many 21 views, particularly the views of retail investors. These 22 are the investors that are directly or indirectly 23 providing the bulk of all capital invested in the 24 securities. 25 I also believe it is important to have diverse 0090 1 Roundtables, not only diversity in thought, but also 2 diversity in gender and ethnicity. I am particularly 3 struck by the fact that today's Roundtable only has two 4 female panelists and it appears there are no panelists of 5 color. 6 Studies have shown that diversity helps bring 7 new perspectives to challenging problems. I believe 8 Commission sponsored Roundtables should reflect the 9 diversity in our nation, and I hope and expect that the 10 staff will seek out diverse panelists for future 11 Roundtables. 12 As to the panelists who are with us today, 13 however, I do with heartfelt gratitude thank you for 14 being with us today to share your views. All of you have 15 important information to share about the fixed income 16 markets, and I very much appreciate that you have taken 17 time from your very busy schedules to be here with us 18 today and to share your views. I know there are a 19 million other things you could be doing. We are thankful 20 that you are here. 21 I hope as the discussion today unfolds that you 22 keep in mind the needs of the Main Street investors. 23 These are the investors who deserve a market environment 24 structure that is fair, transparent, and orderly, and I 25 know you will do so. 0091 1 Thank you. I apologize for the interruption in 2 your train of thought, but maybe I've given you some food 3 for thought. 4 MR. RAMSAY: Thank you very much, Commissioner, 5 for the very thoughtful comments. We appreciate that. 6 With regard to retail investors, in particular, perhaps 7 if we could talk a little bit more about pricing, which 8 we touched on, but basically the basis for pricing retail 9 trades and how dealers do that, how they determine what 10 the prevailing market price is for purposes of 11 determining how much one can mark up a retail trade. 12 Does anybody want to offer some thoughts on is 13 there adequate guidance out there, do people have a 14 general understanding of what the basis is for 15 determining the prevailing price or is there more that 16 can be done to give guidance on that score? 17 MR. NOBLE: I guess I'll start because Larry 18 looked at me, so it was kind of like okay, we're the two 19 retail guys, pricing inventory for retail customers. One 20 follow up to Commissioner Gallagher. FINRA had an investor 21 alert on duration, which I think was very well written, and 22 we are using it as part of an education project to make 23 sure that our retail investors know what duration actually 24 means to them and what happens. 25 I don't want to lose the fact that they started 0092 1 the conversation, and I think it was a good conversation 2 for them to start. 3 COMMISSIONER GALLAGHER: I thought it was 4 particularly well done and incredibly well timed. I give 5 FINRA tons of credit for that. 6 MR. NOBLE: It was very well timed. Back to 7 retail pricing. Pricing of inventory, retail or institutional, 8 in my shop, is the same. We look at a number of factors, 9 mostly looking at comparable securities, EMMA information 10 if available, available information from the ATSs, and we 11 take in the feeds from all ATSs to see what's out in the 12 marketplace at any one time, and then use that along with 13 credit information, information on maturity, and coupon 14 because coupon does have a function of what the pricing will 15 end up to be. 16 That is something we may need to talk about 17 because I think Larry and I have had discussions about 18 the proliferation of premium coupons in the new issue 19 marketplace, which retail customers truly haven't grasped 20 the need to buy premiums in the marketplace, especially 21 if we are going to have an interest rate change or shift. 22 But we use all that information. I think the 23 information that we have at least at my shop is fairly 24 comprehensive and available as much as I need. 25 Larry? 0093 1 MR. BOWDEN: We do not have separate retail and 2 institutional inventories either. We have single 3 inventories. Still, though, I think you have to look at the 4 size of that trade and what the total dollar amount is in that 5 trade to try to transact it. There is a built in expense 6 in every trade that we do. I don't know what it would be 7 in Craig's shop or whatever. 8 But we certainly appreciate all the regulatory 9 emphasis, but it does build a cost into each trade. We 10 have a lot invested in legal and compliance and 11 regulatory maintenance. It could be $20 to $30 a trade 12 that's built in just based on regulatory compliance. 13 Depending on the block size of that trade, the 14 overall percentage is certainly going to vary just to 15 cover the cost of doing that trade. 16 I was surprised at the two percent, 1.98 17 percent, or whatever was published in the study. That 18 certainly seemed high to me. But anyway, we certainly 19 do not offer different pricing for retail or for 20 institutional. 21 MR. RAMSAY: David? 22 MR. DIMITRIOUS: Further to this point John was 23 bringing up about the prevailing market value and how you 24 guys determined prices, regarding the mark-ups 25 themselves, do you distinguish between mark-ups on 0094 1 riskless principal versus trading out of inventory when 2 calculating your mark- ups? 3 MR. BOWDEN: I think the overall price and 4 yield determines the mark-up. The bonds are trading on a 5 yield basis. That's what you're trying to compare. You 6 are trying to compare yield to existing other securities 7 out in the marketplace and what is favorable. 8 We don't buy a bond, put a mark in it and just 9 see what the yield is. You have to price based on the 10 end resulting yield to the customer. 11 MR. NOBLE: Similar, riskless principal or 12 principal. Since we are in a marketplace where especially 13 with price dissemination, when you look at EMMA, you can't 14 tell if it's riskless principal or principal. It's what 15 the yield was. 16 So if a bond is supposed to be yielding a five 17 percent, to get to a customer, you may have a small mark, 18 you may have a little larger mark, it's trying to get 19 to a yield, and that's the end result, if I have riskless 20 principal or principal. 21 MR. EADY: To what extent do you currently 22 disclose to the customer what the mark-up is on the 23 trade, whether it's out of inventory or on an agency 24 basis or riskless principal basis, if you want to define 25 a non-inventory trade? 0095 1 MR. NOBLE: Agencies would be disclosed because 2 as such it would be disclosed on the confirm. Any principal 3 transaction, the mark-up is not disclosed. 4 MR. BOWDEN: That is the same at our firm as 5 well. On our managed accounts, it's all done on an 6 agency basis. Our principal transactions, again, it's 7 traded on a yield basis and it's not disclosed. 8 MR. EDELMAN: This contributes, in my opinion, 9 to the lack of consumer understanding that these numbers 10 are negotiable because they don't realize there is a 11 decision being made by the seller or by the broker as to 12 what those prices are in a competitive environment. 13 MR. EADY: On the agency trades, where the 14 amount of the mark-up or commission is disclosed, what 15 percentage of your business is done on an agency basis 16 currently? 17 MR. NOBLE: Well, agency with commission disclosed, 18 probably hardly anything. Agency transactions as a 19 whole, probably about 15 percent of my business because 20 that's all my managed business. Any managed transactions 21 similar to yours has to be done away from my inventory. 22 That managed account would then be charged a fee on an 23 annual basis for managing that account, similar to a 24 mutual fund. 25 MR. BOWDEN: The discussion about the managed 0096 1 accounts and the agency business, it is similar within 2 our firm as well, again, this is another demonstration of 3 the fragmentation of our marketplace or whatever. 4 Again, if we are trading in a lesser known 5 credit, a more geographically local credit that is very 6 unknown, we can put it out with the broker's brokers, we 7 can put it out with the ATSs, and in a lot of cases, we 8 will get zero bids. 9 We do have a letter that we have to then get 10 signed by the customer to authorize us to then trade that 11 on a principal basis because our bid -- a lot of time we 12 will put bonds out. 13 We just had a block of 250 bonds for an 14 individual where our internal bid was $7 more than the 15 bids we got from the ATSs. We are doing everything we 16 possibly can to make sure the customer is getting a fair 17 and equitable price for that security, but just putting 18 it out with ATSs and broker's brokers is not going to 19 guarantee you that you are going to get a bid. 20 MR. NOBLE: To follow up, we have seen a move 21 in the industry to more managed accounts because the 22 external managers are selling the professional management 23 of that portfolio, so these would be separately managed 24 accounts, not so much a mutual fund or ETF. 25 But the risk that we have in some of that, to 0097 1 Larry's point, is we have asset managers buying million 2 bond blocks at favorable prices, no question about it, 3 but then they are taking those million bond blocks and 4 dividing it between 20 or 50 accounts and they end up 5 with 25 bonds or 20 bonds or 5 bonds in each account, and 6 when that account makes the decision to liquidate, that's 7 when they get hurt. 8 And a managed account has an annual fee on that 9 account every year where as much as you may say that a $100 10 fee on a $5,000 bond, if that's a 15 year security, 11 that's the end of that commission. That's done. So 12 there’s a differential that’s happening in there also. 13 MR. BOWDEN: And if you do break that down, not 14 to keep dwelling on this, but the postage alone on 15 an account on an annual basis will eat up - now, 16 certainly, there will be other securities in that 17 account, hopefully, but there certainly is an ongoing 18 expense with holding a longer term municipal bond. 19 MR. EDELMAN: While we are mentioning the 20 notion of pricing versus yield and such, and these matter 21 an awful lot, I want to go full circle back to what 22 Commissioner Aguilar said and what Commissioner Gallagher 23 pointed out. You didn't know, Commissioner, but your 24 comments were not a deviation from our conversation 25 because that was exactly on point of where we were. 0098 1 There is a Tsunami coming. You referred to it, 2 Commissioner, as Armageddon. Vanguard highlighted it very 3 clearly because we all know in this room what's coming when, 4 not if, but when the interest rates rise, there is going to 5 be a crushing impact on the value of these securities. 6 and since 70 percent of the market are individual 7 retail investors who are also older, tend to be retired, 8 income oriented, and safety conscious, these are the 9 folks who fled the equity markets in 2008 if they were 10 there in the first place to what they perceived to be the 11 safety of the municipal bond market. 12 There is an Armageddon coming, not because of 13 the markets, the markets will survive it, the way we 14 survived 2008, but the individuals may not survive it, so 15 while we are thrilled at FINRA's investor alert, which I 16 highlighted prominently on my radio show, we need to see 17 more and more of this. 18 We could argue the pricing and the inventory 19 and principal trades versus broker trades and yields, et 20 cetera, at the end of the day, it's the Tsunami yield 21 curve that's going to be the crushing element in the 22 marketplace, and that's where I think our primary 23 motivation and focus really needs to be for the next 24 several years. 25 MR. BOWDEN: We also need to not forget there 0099 1 are some ways you can protect a little bit of that risk, 2 hopefully, there is more laddering of maturities now 3 where the individual is not just buying the 30 year bond, 4 that they are actually spreading their assets out along a 5 maturity ladder so they have current dollars coming 6 mature each year to reinvest in a rising interest rate 7 environment. 8 Also, we don't need to forget that as long as 9 that credit is worthy, it will continue to pay interest. 10 It's not that the customer is going to lose everything. The 11 market value of their portfolio will certainly go down, 12 but most of these people have invested in these 13 securities for a flow of income. 14 Yes, it will be affected by rising interest 15 rates, not to flow of income, but yes, the market value 16 will, but as long as they are holding those bonds and the 17 creditworthiness of that bond is there, it's still going 18 to pay off at maturity. 19 MR. EDELMAN: That's psychologically not what 20 my experience has been at the retail level with these 21 investors. They tend to be unhappy people no matter what 22 happens. If they keep the bond, their yield is a lot 23 lower than their friend who just bought a new bond at a 24 higher yield, and if they sell the bond, they just lost 25 money in the sale. 0100 1 So they are unhappy if they keep it. They are 2 unhappy if they sell it. That's part of the frustrating 3 element of this marketplace. 4 COMMISSIONER GALLAGHER: One thing that I think 5 we can all agree on in this space is no one should lose 6 money because of a dearth of available information, 7 either from the regulators or from their financial 8 services professionals. 9 MR. BOWDEN: There's another unknown out there, 10 I don't want to turn this into a political discussion, 11 but the 28 percent cap on tax exempt income, I don't 12 think any of us really know if that comes to fruition 13 what impact that could have as well. 14 MR. RAMSAY: Yes, Commissioner Walter? 15 COMMISSIONER WALTER: Let me just take a second 16 and push back a little bit on Ric. I found that when we 17 went around and did our field hearings, putting together 18 the municipal securities market report, and absolutely 19 insisted apropos Commissioner Aguilar's comments, that we 20 have actual retail investors testify, which was a 21 tremendous challenge to find them, and we weren't looking 22 for people who thought they had been defrauded. We were 23 looking for people to talk about their every day 24 experiences. 25 I was surprised that what we got was not a 0101 1 litany of complaints but a combination of suggestions for 2 reform and positive comments about the people they dealt 3 with in the marketplace. 4 I actually do think if we can reach the retail 5 investor community, which does present very significant 6 challenges, that they are a constructive force to try to 7 move forward in this area, although I do recognize there 8 is among us, as a perennially unhappy person, there is 9 that element in the constituency. 10 MR. RAMSAY: Thank you for that. That may be 11 an appropriate note to close on so we stay on target more 12 or less. I want to thank everybody for their 13 participation. This has been very helpful and enjoyable 14 for me, so thanks. 15 (Recess.) 16 PANEL 2: CURRENT MARKET STRUCTURE FOR CORPORATE 17 BONDS AND ASSET-BACKED SECURITIES 18 MS. EDWARDS: Welcome back to Panel 2. Panel 2 19 will focus on the characteristics of the corporate bond 20 and asset-backed securities markets and how they compare 21 to the municipal securities markets. 22 First, like in the last panel, we will 23 introduce the panelists. We have starting on the far 24 side, Steven Genyk, who is Managing Director and Head of 25 Fixed Income Capital Markets at Janney Montgomery Scott. 0102 1 We have Nancy Mueller Handal, who is Managing Director and 2 Head of Structured Finance, MetLife. 3 We have Colin Heffron, CEO, GFI Group Inc. 4 Jonathan Horne, Executive Vice President and Portfolio 5 Manager at Pacific Investment Management Company. Richard 6 M. McVey, Chairman and CEO, MarketAxess. We have Michael 7 A. Goldstein, Professor of Finance and the Donald P. Babson, 8 Chair in Applied Investments and Chair of the Finance 9 Department, Babson College. 10 We have Eric J. Pitt, Managing Director, J.P. 11 Morgan Securities. Neil M. Schloss, Treasurer and Vice 12 President at Ford Motor Company. Robert G. Smith, 13 President, Chief Investment Officer and Principal at Sage 14 Advisory Services. Kevin Molloy, Managing Director of 15 Fixed Income, NYSE Bonds. 16 Before starting our panel discussions, I'd like 17 to welcome Dr. Michael Goldstein. Dr. Goldstein has 18 kindly agreed to make a presentation summarizing the 19 academic literature concerning the corporate bond 20 markets. 21 Dr. Goldstein is Chair of the Finance 22 Department and Professor of Finance at Babson College 23 where he holds the Donald P. Babson Chair in Applied 24 Investments. 25 He is also currently an Honorary Professor at 0103 1 the Queens University at Belfast, and was Visiting 2 Professor at Trinity College in Ireland in 2009. He has 3 also taught at Babson College and at the University of 4 Colorado at Boulder, where he received the Procter & 5 Gamble Teaching Excellence Award. 6 Dr. Goldstein received a B.S., an MBA, an MA 7 and a Ph.D. in finance from the Wharton School at the 8 University of Pennsylvania. 9 Dr. Goldstein. 10 DR. GOLDSTEIN: Thank you. First, I want to 11 thank the Commissioners and the staff and all the people 12 who put this together for inviting me and putting 13 together what looks to be just absolutely fascinating 14 panels. I greatly appreciate being here. 15 We have a very long of questions to try to 16 cover in a presentation, so I'm going to do my best 17 quickly, and certainly if later on someone wants to see 18 something in more detail or see again, let me know. 19 So one of the questions was general market info. I 20 didn't know how much was known about what the markets 21 look like right now in terms of just general size, so 22 very quick review. I always put this up to remind people 23 that the amount of outstanding in the corporate bond market 24 exceeds the combination of Federal and state debt, so 25 collectively, this is a very important market for the issuers. 0104 1 In terms of amount that is traded, it's about 2 2 to $3 trillion per year, in terms of the total dollar amount 3 that is traded on TRACE, all in. It turns out if you 4 kind of go back up one slide, you will see if there is 5 $20 trillion trading and about $3 (trillion) or so trade per 6 year, maybe $4 (trillion), it is only about 20 percent that 7 gets turned over per year. 8 So one of the questions was what does the market 9 look like based on trade size. I interpreted that as 10 institutional size versus retail size, and in academics, 11 we tend to use if someone is trading more than 100 bonds, 12 we will call them an "institution," and if less, we will 13 call them "retail," and in terms of volume, as you can 14 see, there are an awful lot of bonds that are traded, but 15 in terms of dollar volume or regular volume, it's almost 16 all institutional. 17 But in terms of the number of trades, the 18 observation is it's mostly not institutional so the red is 19 the institutional while the blue is the number of trades. 20 A lot of trades, about a third of them are institutional, 21 maybe a little less, but in terms of volume, they count 22 for between 94 and 97 percent of everything that is 23 traded. 24 In terms of per month, just to put this up 25 quickly, you can actually see a couple of interesting 0105 1 things, again, in terms of number of trades, there is a 2 big difference, and the other part I just thought was 3 very interesting, somewhere around the crisis time, the 4 actual number of trades per month shot up. In fact, you 5 can actually see the percentage of institutional trades 6 dropped and has now basically recovered so that is part of 7 the effect of the transition. 8 There are dealers in this market and the 9 biggest ones trade a lot of stuff, and they tend to 10 account for the bulk of it. Using a subset that we kind 11 of came up with, about a third -- there are about five 12 firms that seem to account for about a third of all the 13 trading, and then it goes on from there. 14 This will come as no shock probably to anybody, 15 but corporate bonds are highly illiquid. One sub-sample 16 I took of 47,000 TRACE eligible bonds, during the 17 entire time period, I saw no trades for 18 percent of 18 them, so here's a couple more slides. 19 The median number of trades in the past 30 days 20 was about 14, which means your middle kind of bond 21 trades, once every two days, if you think of it that way, 22 and ten percent of the bonds had one trade basically in 23 the past month. That is a very sharp drop off in 24 trading. Again, it comes as no great shock to people. 25 Last bit on infrequent trading, the blue bar one, 0106 1 there was one trade that year. That bond had at least one 2 trade that year and you can kind of see when I dropped to 3 125 trades, which is approximately, you know, one every 4 two days, you are dropping by a sizable amount, about 30 5 percent, and for the number of bonds that traded basically 6 once a day, once every 250 days. You had 250 trades or more. 7 I'm not saying they traded every day. I'm just 8 saying the cumulative amount, if you averaged it out, 9 would be about equivalent to one per day. Only 20 10 percent of the bonds hit that. 11 And this is conditional on having a trade at all 12 that year. Similarly, you can see it by month. Of 13 course, the numbers are smaller because a bond that only 14 trades once a year, there are 11 other months where it 15 didn't trade at all. 16 So there was a long list of questions, and 17 academic papers have written a variety of things. I'm 18 going to try to summarize just a few of them. I want to 19 apologize to all the authors who I didn't cite, and I 20 also want to just let you all know that Exchange has 21 a wonderful staff of economists here who have written a 22 lot of great papers, and I'm probably not going to cover 23 all their papers, but there are quite a lot of excellent 24 people that could do probably better than I. 25 For example, Amy, or Dr. Edwards, I should say, 0107 1 along with Larry Harris and Mike Piwowar, had an 2 excellent paper looking at price transparency when TRACE 3 came on and they found there was a reduction in transaction 4 costs due to that. 5 There are a lot of numbers in their paper, so I 6 kind of said, yeah, it looks like it's about five to 25 7 basis points in reduction in costs. 8 There was a paper that looked at insurance 9 company data and they also found not only were there 10 reductions in costs when TRACE came on, but also even 11 other bonds got some benefit, even the ones that weren't 12 disseminated immediately. 13 I had a paper with Edie Hotchkiss and Eric 14 Sirri where we actually looked at triple B bonds. I guess 15 I can now say I was one of the people who decided which one 16 of 120 bonds got turned on way back, I guess that was 17 April 2002, and there was a matched pair, and we also 18 found there was a reasonably large effect, with no big 19 effect on volume. 20 Another question that was in the list of 21 questions was where does most of the trading occur. Most 22 of the data, again, Dr. Edwards as well as other people 23 that have worked here, have indicated that most of the 24 trades tend to happen on the over-the-counter market. 25 There is some trading on the New York Stock 0108 1 Exchange. I'm sure we will hear more about how much 2 trading there is on the New York Stock Exchange, but the 3 academic literature tends to say it's smaller and most of 4 it is not happening there. 5 Liquidity. There were a lot of questions on 6 liquidity. I tried to summarize a variety of papers. 7 Liquidity can be defined a variety of ways. It could be 8 on bid-ask spread, could be on volume, could be on what 9 portion of the yield spread over Treasuries, after you take 10 away the credit portion, what portion is due to liquidity 11 effect, and I tried to summarize it in one slide. It's a 12 little bit hard. 13 But there is lower liquidity or larger costs if 14 your bond is older, if it's a longer maturity, and 15 certainly if it's not on the run, on the run's tend to 16 have five year, seven year, ten year, those tend to have 17 lower transaction costs, but as soon as a seven year 18 becomes a six year, the transaction costs tend to go up. 19 If you are a smaller issue, it tends to have a 20 little less liquidity. Certainly, if you are a small 21 trade, it's notably more costly for you to trade. Tends 22 to be true in the academic literature, though there is some 23 differentiating, that if you have a lower rating, it 24 tends to cost a little more to trade. If you are not 25 144(a), costs a little more to trade. 0109 1 Over time, trading costs have varied, so time 2 frame does matter. There is an interesting paper that 3 talks about if frequent traders -- if the people who hold 4 your bonds are people -- if your bond is primarily held 5 by people who tend to trade, it turns out that the 6 liquidity costs in your bond are less. It’s a latent 7 liquidity paper. I think it is the Mahatia, et al paper. 8 It turns out some bonds are top bonds, we have 9 an issuer, Ford Motor Company issues a lot of bonds, I 10 think 250, I might have heard, so not all their bonds 11 have the same liquidity and it tends to be that the 12 market focuses on a top bond, and that changes over time. 13 There is some evidence that in fact those bonds respond 14 very fast to information, so liquidity tends to get 15 pooled in a bond. 16 Just a little aside, as an academic, sometimes 17 I feel like we are being unfair. Ford Motor Company has 18 one equity and 250 bonds. If we make 250 equity slices, 19 I'm not sure how often all of them would trade. All that 20 is concentrated in one security while when you are a 21 corporation, if you have many, many bonds, a lot of the 22 interest might be dispersed. 23 I also have a paper I didn't cite up here. One 24 of the interesting questions is for issuers, does the 25 expected liquidity of your bond affect the coupon you 0110 1 have to pay and the yield you have to pay because it's all 2 nice and good in the after market, trading costs are lower, 3 but how does that actually lower the costs for Ford, and 4 it turns out that it is true, if people expect your bond to 5 have lower liquidity, you will actually pay lower, and 6 that lowers your cost of capital. That is something that 7 probably deserves more research, but it's useful. 8 I happen to have a paper on how liquidity 9 varies for illiquid and highly liquid bonds. I just 10 thought I'd throw up what I calculated here along with 11 Edie Hotchkiss on dealer round trip costs so this is our 12 estimate of spread, which is a dealer takes in 100 bonds 13 or more, and then parcels them out to customers. What is 14 the spreads they pay. 15 What I wanted to highlight is you might think 16 it would have varied with liquidity, but it turns out not 17 so much, that actually the stuff that can trade what I 18 calculated to be equivalent to around 30 times a day, 19 it's not really 30 times a day, that's an average of that 20 bucket, they may actually have higher spreads than some 21 of the stuff that trades very, very infrequently. 22 The three to ten days is the bucket it had in 23 the past 30 days, between one and ten trades and I’ll come 24 back to that, and it's even more true when you get to 25 high yield. 0111 1 I personally find it fascinating that the 2 calculation for something that is a C rated bond that 3 trades between one and ten times in a month has a dealer 4 round trip cost of approximately equal to a BB bond that 5 trades all the time. 6 The point being it's not linear, which is kind 7 of an interesting result. 8 What happened during the financial crisis? 9 There is a nice paper that is about the financial crisis 10 by Dick Nielsen and Feldhutter and Lando. They looked at 11 what happened to the portion of spreads over Treasuries 12 and what portion of it was due to liquidity and how did 13 it get affected. 14 The end result really is as you might expect, 15 yes, during the financial crisis, it turned out that the 16 amount that the yield spread increased due to illiquidity 17 increased, the more illiquid's had to pay the more yield 18 spread, and also people's sensitivity to that increased, 19 so that how much people cared about the illiquidity 20 changed. 21 Because it was a double effect. It mattered and 22 people cared so it mattered even more. That is the easiest 23 way to say it. 24 Also, they have a very nice couple of graphs, 25 which I didn't copy, but looking at what happened to the 0112 1 Bear Stearns underwritten bonds right around Bear Stearns 2 falling and the Lehman Brothers' bonds, right around when 3 Lehman fell. 4 And as you might expect, those particular bonds, 5 the liquidity costs shot up dramatically. 6 So this is a table from their paper. Well, I 7 re-typed it in. To kind of give you an idea of what the 8 liquidity component is, how much does liquidity matter in 9 actual basis points as part of your yield. 10 And for very high rated bonds, it actually wasn't 11 such a big deal, but as you go down the credit rating 12 spectrum, it tends to matter and you might notice that 13 during the crisis, this stuff jumped quite dramatically, 14 that the liquidity component became one to two percent 15 for BBB or high yield bonds. 16 The portion of the spread, what I have at the 17 bottom of that Table 5 is the portion of the spread -- 18 yield spreads went up, the portion that mattered due to 19 liquidity went up, and the percentage of the spread that 20 was related to liquidity went up. Generally speaking, 21 financial crises are not good for the trading of bonds or 22 their costs, and should try to be avoided. 23 Bond new issues, issuers tend to care about 24 this, how do things trade, how do things trade over time. 25 That was one of the other questions that was asked. 0113 1 Literally last week on Thursday, I saw a paper presented 2 by Dr. Karen Craig from the University of Tennessee. 3 She had this graph, and I said can I take this 4 graph and present it, and it was a graph that she has of 5 what is the dollar value of IPOs, again, in academics and 6 elsewhere, we tend to focus on the equity IPOs, which is 7 the yellow bars. Sometimes if you had any SEOs, which is 8 the green bars, but the red and blue bars, which are the 9 issuance of corporate debt, you can see, just swamps 10 that. 11 Again, it is an important question. Sometimes 12 it doesn't get focused on enough. 13 I have a paper so I'm going to talk about my 14 paper for a second on this which talks about new issues and 15 says there is a wide price dispersion, so at the first few 16 days of trading, there is quite a lot of different prices. 17 I'll show you a graph in a second. 18 Corporate bonds when they are issued are under 19 priced, just like equities, and that has been found also 20 in muni's, it has been found kind of across the board. 21 There was a reduction in under pricing and 22 price dispersion as time went on and as transparency 23 increased and the non-syndicate members account for a 24 larger portion of trading over time. Initially, the 25 syndicate members are quite important. 0114 1 And then we looked at dealer inventory for bonds 2 that went below their offering price and bonds that did not. 3 I'll show you one or two data, because sometimes data is 4 helpful. 5 So our biggest volume findings, because there 6 was a question about how does volume change over time, and 7 I'll put up data in a second, trading volume is highest 8 in the first couple of days and drops off, that retail 9 investors seem to come in or retail size investors seem 10 to come in not on the first or second day but a little 11 bit later, which you will see. The book runners or 12 manager’s market share fall. 13 An incredibly small graph which you can't read 14 but I couldn't figure out how to get it all another way. 15 The quick thing to see is all bonds on the first day, 16 there are 31 million average trading volume and by the 17 tenth day, it falls to seven. That's a pretty big drop 18 when you think about it. That is very, very sizable. 19 And the underwriter share of volume drops from 20 the first day of about 81 percent down to about 54 percent. 21 If you look at a monthly basis, daily trading in the 22 first month is like 11 million in terms of trading 23 volume, but by the sixth month, it drops down to two. 24 So there is a big drop off. 25 The underwriter share of volume drops from 0115 1 about 60 percent down to under 50 percent, at 47 percent. 2 Just a graphic to kind of show you this is 3 also true if you volume, but in terms of number of trades, 4 institutions come in big in the beginning, particularly 5 by the second day, and I have kind of highlighted that 6 the individuals, smaller sized trades, tend to come in 7 a little later. 8 I'll let the market participants explain how 9 that happens, but I have to say, if I were an investment 10 banker having to place stuff, I'd probably want to place 11 it with big institutions first and then have them roll it 12 off. 13 This is a quick graph of a disseminated bond, a 14 bond that was disseminated, issuance, and the variance in 15 prices on a given day. The little blue dots are trades 16 less than 100 bonds, the harder to read but yellow 17 triangles, which there are fewer, are trades bigger than 18 100 bonds, and as you can see, there is quite a lot of 19 dispersion, particularly in the buy's, on a given day. 20 I'm just going to argue that can't be interest rate 21 changes. It moved up from the offer price. There are a 22 lot of people buying, an awfully lot of different prices 23 in these bonds. 24 Last thing about inventory and then we will 25 move to the next bit, there is some question about what 0116 1 does your inventory look like when the bond trades 2 above its offer price versus what does the dealer inventory 3 look if it trades below, and very quickly, I don't see this 4 as being a huge difference. I suppose some of the underwriters 5 might disagree. These are not massive numbers. We have 6 other numbers about percentage and positive territory of 7 numbers of outliers. 8 So a good way to think of this, I think the 9 underwriters are doing their job in either direction and 10 they don't get stuck holding a boatload of inventory if 11 the bond starts trading below its original offer price. 12 To close up on this, because I don’t want to 13 take up a lot of time, inventory and holding period. So 14 it got kind of interesting in this due to the spread 15 thing. It turns out that people -- dealers tend not to 16 hold stuff for very long, they oddly enough, tend to hold 17 higher rated securities that are more liquid longer than 18 lower rated securities that are hardly ever traded. I'll 19 show you that and why, I guess now. 20 Here is a holding period for investment grade, 21 and for investment grade, especially if you don't focus 22 on the AAAs, it does vary a bit with level of frequency 23 of trading, although probably not as much as you might 24 think. 25 I just want to highlight the investment grade, 0117 1 somewhere around eight or nine days worth of holding 2 period, and even the AAA, most traded stuff, trades, it 3 takes about four days on average for people to get rid of 4 it. 5 Now if you go to high yield bonds where you think 6 you might get stuck with it longer, I just want to 7 highlight that a C rated bond that trades between one and 8 ten times over the previous 30 days somehow has a holding 9 period less than an AAA bond which in the previous 30 10 days is trading an awful lot. 11 So this to me looks like magic. Somehow a bond 12 that hardly ever traded, the dealers were able to not 13 have to hang on to it very far. How do they do that? I 14 don't know. I don't have any tapes and I don't want to 15 say I know. 16 But I did check this question, which is what 17 percentage of the units that I saw and were able to 18 follow, they were able to get rid of the whole thing 19 within a day, so there was a zero holding period. If you 20 got rid of it within the day, I counted that as zero 21 holding period. 22 You might notice for investment grade bonds, it 23 doesn't vary that much. They are somewhere in the 30s, 24 let's go that way. But the percentage of days for zero 25 holding period for high yield bonds clearly drops as it 0118 1 becomes less frequently traded and as the credit rating 2 goes down so they actually hold it, again, for the least 3 amount of time, those CCC or C rated bonds, they are going 4 to be able to get rid of it right within a day. 5 That was a lot of stuff in a short period of 6 time. Here's the gist of it. It's a large illiquid 7 market generally with a few liquid bonds. Transparency 8 helped with liquidity as measured by a variety of ways. 9 What your bond looks like matters, the issue 10 size, the credit rating, its age, and a whole bunch of 11 factors about bonds do matter for liquidity, and some 12 research indicates if you pick your bonds right, so it's 13 expected to have more liquidity, you might actually pay a 14 little less. 15 Financial crises are never good for liquidity. 16 That was a strong finding by a paper but probably not a 17 great surprise, but glad to see that actually was true. 18 IPOs, new bond issuances are under priced. They 19 have a sharp drop off in trading, not wildly dissimilar 20 to equities in that way. Big price dispersion on the 21 first day and continued for quite a while, and that price 22 dispersion is probably much larger than what you see in 23 equities, and the holding by dealers and by the 24 underwriters don't seem to vary that much, depending on 25 whether the bond traded up or down. 0119 1 The last bit is holding periods are notably 2 shorter for the less liquid bonds with a lower credit 3 rating. 4 That would be my 15 minute summary of three 5 pages of questions. 6 MS. EDWARDS: Thank you, Dr. Goldstein. At 7 this point, we will open it up to the rest of the panel. 8 Before we do that, I just wanted to remind everybody to 9 try to turn off their mics when they are done speaking. 10 When we don't turn off the mike's, sometimes we get 11 feedback. I just wanted to remind everybody of that. 12 I guess the first question for the panel would 13 be how do the corporate bond and asset-backed securities 14 markets differ from municipal markets that we talked about 15 in panel one. 16 MR. SMITH: One of the things I think is 17 incredibly obvious is that the ability to hedge risk in 18 the taxable bond arena far exceeds what you can do in the 19 municipal arena, from a derivative standpoint. 20 Credit default swaps, various different 21 interest rate mechanisms, whether they are 22 Exchange-traded or otherwise, make it a decidedly 23 different market in terms of being able to manage risk as 24 a risk manager. 25 Those are things that I think are important. 0120 1 The other thing is how securities are valued. 2 Increasingly, the corporate bond market is being priced 3 off the swap market, not necessarily the cash market. 4 It's the tail wagging the dog. I think that is an 5 incredibly different environment than what you will find 6 on the tax exempt market. 7 MS. HANDAL: The asset-backed securities market 8 is also very different from the muni market for a variety 9 of reasons. One is also the institutional investor 10 presence in the market as compared to retail. In 11 general, it is an institutional investor market. 12 I heard very little even anecdotal evidence of 13 retail players, I believe, sort of private wealth, and 14 maybe some of the smaller issues get placed into retail 15 hands, but for the most part, it is an institutional 16 investor market. 17 It is also highly diversified, so you have 18 everything from very liquid ABS, such as credit card and 19 auto backed receivables, to esoteric, such as time share 20 receivables and franchise loans and things like that. 21 Additionally, it's based on the thought process 22 that these are bankruptcy remote, special purpose 23 vehicles, so essentially what you are doing when you look 24 at an asset-backed security is you are not only looking 25 at the sector, the issuer, the rating, but you are doing 0121 1 as much loan level detail/collateral analysis as you can. 2 So you are truly beholden to the collateral 3 performance in the bond rather than an issuer, which is a 4 little different. 5 In addition, the market size outstanding right 6 now is about $1.2 trillion, and again, I'm just talking 7 about asset-backed securities, not RMBS or CMBS, which is 8 a different story. 9 MR. SCHLOSS: I guess I would only add to that 10 because as an issuer, not only Ford, but Ford Credit is 11 really the bigger issuer, so we do both unsecured and 12 ABS. 13 I can just emphasize what was just said about 14 the ABS market, the complexity of not only the individual 15 asset classes but the structures themselves, so the level 16 of transparency that exists on TRACE or in other ATSs for 17 the unsecured market really doesn't exist today for 18 asset-backed, and there are several reasons why it 19 doesn't, and maybe shouldn't, from the standpoint of the 20 investors who are buying it needing to really understand 21 the underlying collateral, the underlying structure, is 22 it subordinated, not subordinated. 23 It is less about the issuer, per se, as the 24 value of the collateral. That becomes a really difficult 25 disclosure item from the standpoint of a publicly traded 0122 1 system. 2 Obviously, the prospectuses and the level of 3 work that's done at the investor level is much more 4 significant, and as a result of that, where we will get 5 200 to 250 different investors in an unsecured deal, we 6 will get 25 to 50 in a comparable sized secured 7 transaction. 8 MR. HEFFRON: I would just add on the previous 9 panel, I think there was a figure quoted of 50 percent of 10 direct holdings by retail in the municipal bond market. I 11 run a company that has only institutional clients. I 12 would expect that the corporate bond market has nowhere 13 near 50 percent of direct holdings. I'm not sure what 14 the percentage is. 15 But we run institutional business, it sits in 16 between many of the banks, but also there are a multitude 17 of online retail bond electronic platforms popping up. 18 So while the prices are becoming more and more 19 closely related, I don't think there is a direct holding 20 from the retail sector equivalent in the corporate market 21 versus the municipal bond market. 22 MR. GENYK: Amy, I would just add a couple of 23 things. Clearly, the corporate market is much larger in 24 terms of size, but it has a smaller number of CUSIPs. I 25 think that is a factor that we all heard this morning. 0123 1 The other issue, too, is TRACE has been in 2 place since 2002. EMMA has been terrific for the 3 municipal industry in terms of transparency and price 4 disclosure, but it's relatively new, particularly when 5 compared to TRACE. 6 MS. EDWARDS: All right. Thank you. We are 7 going to cover a lot of the same territory in the 8 corporate panel today that we covered in the municipal 9 panel earlier today. If you remember, we started out by 10 talking a bit about ATSs and their role in the municipal 11 bond market. 12 Right now, we will move on and talk about the 13 role of Exchanges and ATSs and dealer-brokers in the 14 corporate and asset-backed securities, how they compare 15 to one another and how their use of technology has 16 impacted pre-trade transparency, liquidity and 17 efficiency. 18 I think we can start out with our corporate 19 bond Exchange, Kevin Molloy. 20 MR. MOLLOY: Thank you. Just to touch on the 21 history of the Exchange, the Exchange has always traded 22 corporate bonds in one way, shape or form, and has had 23 varied degrees of success. I think it’s a little known 24 fact but as recently as 1991, they were doing 25 approximately 2,400 trades a day across about 1,700 0124 1 corporate bond CUSIPs, which I think puts the Exchange in 2 a pretty fair light compared to where the ECNs are 3 trading today. 4 Our platform is basically a regulated Exchange. 5 It has live prices, executed, you send an order in, you 6 get an execution back, so there is no subject markets or 7 last look. You have to be a member or sponsored by a 8 member to get to the Exchange. 9 We provide pre-trade transparency both through 10 a data feed that's available to anyone who wants to buy 11 it as well as on our own website, through BondWatch. We 12 have currently in the marketplace today quotes across 13 about 1,600 CUSIPs. That continues to grow as we add 14 more liquidity. 15 We do think the Exchange, to your point on your 16 presentation, is probably not doing a tremendous amount 17 of volume but we do think we could be doing more volume 18 in the space. 19 One of the things we do look at is we compare 20 our markets just to understand quality of markets versus 21 the TRACE inter-dealer market and we do see anywhere from 22 300 to 500 trades a day that could be filled on the 23 Exchange given the print. We do think we have some 24 viable liquidity. That will continue to grow. We can be 25 an efficient mechanism to get better liquidity into the 0125 1 marketplace, differentiating liquidity, and I think 2 better prices for investors. 3 MS. EDWARDS: Thank you. Richard? 4 MR. McVEY: Sure. Happy to chime in. We run 5 an institutional electronic marketplace primarily in the 6 corporate bond high yield and emerging markets. Most of 7 our trading is done through a multi-dealer request for 8 quote or auction trading protocol. 9 But just speaking to the question about 10 transparency, there are multiple ways that we think we 11 assist investors in the price discovery process. First, 12 we do aggregate dealer inventory on the market access 13 systems so that institutional investors have an easier 14 time finding bonds that might fit their portfolio needs 15 through the various search capabilities on the Market 16 Axess system. 17 We have also fully integrated TRACE data into 18 the MarketAxess platforms so that any time an investor 19 and a dealer are engaged in an electronic negotiation on 20 the system, they each have access to the last five prints 21 from TRACE at the bottom of the electronic trade tickets. 22 And then the auction process is open to investor 23 choice with respect to how many dealers they inquire, which 24 greatly depends on the liquidity of the bond and the size 25 of the trade they may be transacting. There is a highly 0126 1 competitive and transparent auction process to help 2 institutional investors get to the best price on each 3 trade. 4 The other interesting point, and we will fall 5 out on the positive side of TRACE, there might be some 6 different views around the market and even on the panel 7 today, but we are proponents of TRACE, and we think it 8 has been especially important post-crisis as investors 9 have needed alternative sources of liquidity and market 10 transparency, and it has helped new participants come in 11 with new capital that is going to work in the secondary 12 markets. 13 The other important feature of our platform is 14 that a completed trade triggers an electronic post-trade 15 message to the dealer's trade capture system, which in 16 turn triggers the price report to FINRA. We then consume 17 the TRACE tape from FINRA back onto the MarketAxess 18 system. That complete loop on an electronic trade is 19 typically done within 60 seconds. 20 So I think there have been significant steps 21 forward in terms of the transparency that's available 22 today in the corporate bond market. 23 MS. EDWARDS: Commissioner Gallagher? 24 COMMISSIONER GALLAGHER: Thanks. Sorry, I 25 violated both rules. I didn't use my pink card and I 0127 1 left my microphone on so I'll try to behave going forward. 2 Kevin, I'll look at you and ask the question, 3 but anybody can answer. You told us a little bit about 4 the platform at the New York. What happened? Right, I 5 mean, 1991 is a good point in time. If I recall, if you 6 go back into the 1940s, you are talking much, much higher 7 numbers on a percentage basis being handled on the Exchange. 8 And as Robert pointed out earlier here, the pricing 9 in the fixed income markets is being driven by the swap 10 markets. Congress has now told us in Dodd-Frank that the 11 swap markets should be essentially Exchange-traded 12 markets as a policy decision for the country. 13 We are regressing on one side here, in 14 corporate debt markets, and we are then progressing 15 potentially on the other side towards Exchange-trading 16 for swaps. 17 What happened and what should we be thinking 18 about with respect to Exchange-trading? 19 MR. MOLLOY: I can't really speak to the 20 specifics of certainly the markets from the 1940s and 21 1950s as far as the transformation that took place then. 22 I'm not that old. 23 (Laughter.) 24 MR. MOLLOY: But I think when you look at what 25 happened to the Exchange, one of the key things was in 0128 1 order to trade a bond on the Exchange, you actually had 2 to specifically list each CUSIP. Over time, if you start 3 looking at 2000 forward, when you look at the Exchanges 4 and data on their website, listings just dropped off 5 completely. 6 So you went from I mentioned something around 7 1,700 securities to somewhere in the low 200 to 300s. A 8 lot of these listings actually are listed for regulatory 9 reasons, for Blue Sky for various tax purposes so a lot of 10 the issues wouldn't really be considered tradeable securities, 11 it can be very small. 12 The Exchange eventually got a rule passed in 13 2007 that allowed them to trade the corporate debt of any 14 company that was listed on the NYSE. I really just think 15 anyone here can attest to when you lose volume on a 16 platform, it is extremely difficult to get it back. 17 I think that was probably the driving cause of 18 the platforms' decline from that so-called 1991 up to the 19 mid-2000s. Certainly, other competitive platforms came 20 on and offered value to investors and the volume moved 21 elsewhere. 22 MR. McVEY: Just to be on record, I wouldn't 23 say that the credit markets have moved backwards in terms 24 of electronic trading in any respect. 25 If you look at both the retail and the 0129 1 institutional markets, open competitive electronic 2 marketplaces are growing. MarketAxess is fortunate to be 3 one of those, but if you look back at 2009, we 4 represented just over six percent of all TRACE volume. In 5 the last six months, we are about 13 percent of all TRACE 6 volume. 7 If you look at smaller trade sizes where the 8 benefit of a competitive auction outweighs the cost of 9 information leakage, our market share is significantly 10 higher, and we are pleased now to count 1,000 11 institutional investors as our clients and 85 market 12 making dealers, so the electronic marketplace is moving 13 forward. 14 What we have found, and coincidentally, we, 15 too, were attracted to the open live market environment 16 all the way back in 2001, and acquired a company that 17 Kevin used to work for at that time, and what the 18 institutional market has found, and it deals with the 19 fragmentation that we talked about earlier, and the 20 number of CUSIPs, and how infrequently they trade, but 21 consistently in the institutional market, an auction 22 based RFQ approach is delivering a better price of 23 execution than the live markets that are available. 24 It is very difficult for dealer market makers 25 to keep thousands and thousands of live markets in 0130 1 quality bid offer spreads on a screen, whether that is an 2 Exchange or otherwise throughout the trading day. What 3 they do really well is compete for orders as they come in 4 electronically. 5 Today, and who knows exactly what the future 6 will hold, but the experience to date has been the 7 auction based models have grown much more quickly than 8 the live markets. 9 MR. HEFFRON: Can I just reiterate the 10 electronic trading in corporate bonds has certainly not 11 gone backwards. In 2011, firms like mine did almost 12 nothing electronically. In 2012, we did over $20 billion 13 in notional, and that continues to grow. 14 Throughout that time, we started doing more and 15 more sectors and more and more CUSIPs, to the point now 16 where we do 700 to 1,000 CUSIPs a week. Some of that is 17 repeatable. Some might be 500 originals. But as Rick 18 said, it's not been esentially an autobook that has 19 taken off. 20 Part of what might have happened between the 21 1940s and 1950s and has gone on is the complications of 22 leaving live markets out. When you have 50,000 CUSIPs, 23 it became not really possible, you didn't have enough 24 technology to handle it and change of prices on any given 25 credit event. 0131 1 So other technologies more dynamic have come 2 into the fold like RFQ, like matching, like auction 3 technology, like open matching, sort of like a liquid 4 net, if you will. That has been our fastest grower in 5 the institutional market. 6 MR. GENYK: Commissioner, I would say from a 7 retail perspective, we have also seen advances in 8 electronic trading. We conduct approximately between 90 9 and 95 percent of our retail corporate bond trades 10 through an ATS. 11 MR. MOLLOY: Just to echo some of the comments 12 there. I think one of the things we are seeing, and you 13 are certainly right, putting your markets out in a live 14 executable environment certainly carries inherent risk, but 15 we are seeing from a technology perspective a lot of the 16 moves and the resources that other markets have used start to 17 move to the fixed income market. 18 And we are seeing participants begin to feel price, 19 you know, certainly not thousands and thousands of 20 securities, but it could be 500, it could be a thousand, 21 you know, in a live executable environment, and feel 22 comfortable with the risk they’re taking. We think that 23 is a trend that is going to continue, and certainly is 24 more friendly to an Exchange type environment. 25 MR. PITT: I want to chime in on that. J.P. 0132 1 Morgan's business is predominately or exclusively dealing 2 with institutional investors, and we do a 3 disproportionate amount of our business with the bigger 4 institutional investors and in the bigger trade sizes. 5 So on my desk, which is an institutional market 6 making desk, we don't tend to deal very often in very 7 small sized trades, more typically, some of the 8 institutions who deal with individuals would come into us 9 for an offering and then be divvying that up into smaller 10 sized trades. 11 We look at the plethora of different options 12 that are emerging for trading bonds as a very positive 13 development for the marketplace. It appears to us that 14 in very small sized transactions, something more Exchange 15 like, may work, may make more sense. There is some 16 traction that’s being gained. In the sort of middle 17 territory, the RFQ model, that my colleague over here was 18 just talking about, has gained a lot of traction. 19 But it is still the case that 50 percent of the 20 volume that trades in the marketplace for high grade 21 bonds is in block trades of over five million in size. 22 Obviously, by trade count, that is a much smaller number. 23 But there is still a lot of dealing that goes on, 24 most of which is not being transacted over these 25 electronic platforms in very large size, and it's not 0133 1 that the platforms wouldn't allow you to trade that size, 2 but it's that the investors feel based on their knowledge 3 of the market and their experience they can get better 4 execution dealing over-the-counter. 5 The question is would it work to electronify 6 that part of the market, is there a problem that needs to 7 be solved in terms of dealing with the block size market 8 and could these platforms effectively solve that. 9 I think the presentation that we started with 10 told us a lot about the range of liquidity that's 11 available in different bonds, and there was a slide that 12 said the corporate bond market is very illiquid, but I 13 think that paints what may be a slightly misleading 14 picture of the actual liquidity in the bond market. 15 In the equities market, you can always get a 16 very, very tight market. It always works on at least the 17 typical retail trade size. Funny things can happen when 18 large size needs to be transacted. 19 In the bond market, it may be true that there 20 are certain bonds that nobody is quoting, not only in a 21 day but in a month or even in a year. There may just not 22 be a quote, there may not be a trade, there is really no 23 transparency, but if an investor shows up with the need 24 to raise, for example, say several billion dollars in an 25 afternoon, that can typically be done using the large 0134 1 institutional dealers as an intermediary, in the context 2 of the market, without moving the market. 3 Maybe you are going to pay a one point 4 liquidity premium on a typical day to move several 5 billions of dollars, and what we have in the equities 6 market is those block transactions are more and more 7 difficult to get done, and things like the flash crash 8 showed kind of what happens when that can go awry. 9 The key is that the market structure needs to 10 work for business as ususal and it also needs to work in 11 times of stress. We saw the over-the-counter bond market 12 functioning in size and in orderly fashion in the height 13 of the stress in 2008 and 2009. 14 My assertion is that if you tried to force that 15 onto an Exchange like model, you might meet with more 16 uneven results. We do think that all the variety of platforms 17 that are out there is very positive for the market, and 18 there probably are different needs represented at the 19 retail end, the small institutional, and then at the sort 20 of big block end of the market so different structures for 21 those different markets is probably a good outcome. 22 MS. EDWARDS: You answered my question that I 23 didn't even ask. So how much pre-trade transparency does 24 -- has technology brought to the market in the last few 25 years? 0135 1 MR. SCHLOSS: I'll give you a issuer's 2 perspective, and I think TRACE has developed a lot over 3 the last -- you know, since 2002, that gives us a little 4 bit more price transparency before issue. Although it's 5 a post record, it does give us a sense of where secondary 6 markets trade although it's small size. 7 So I think from a standpoint of the online 8 systems, the pricing, TRACE, it's a data point. But if 9 you're out trying to raise a billion, a billion and a 10 half, you're still relying heavily on your day-to-day 11 conversations with investors, your day-to-day 12 conversations with the banks from the standpoint of where 13 your large blocks are trading. 14 And again, TRACE, at 5 million or more doesn't 15 give you a whole lot of clarity around is it 5 million, 16 is it 10, is it 50. It doesn't give you a lot of 17 transparency although it gives you a data point that, you 18 know, 10 years ago we didn't have. 19 We have a pretty liquid yield curve, so we 20 have a pretty good sense of where our bonds trade, and so 21 it really helps to have a market data point in addition 22 to, you know, seeing what we see on a normal basis. 23 MR. HORNE: Yeah. I would just echo that. 24 And, you know, I think for most sets of bonds, you know, 25 TRACE is most useful to give you a sense of how 0136 1 frequently it trades. But for -- from the perspective of 2 actual price transparency itself, you know, given that 3 most corporate bonds either don't trade every day or 4 certainly don't trade in material size every day, like 5 TRACE is not going to tell you all that much. 6 And so you're still getting, you know, the 7 preponderance of your market color from a pricing 8 perspective from over-the-counter methods. Although, you 9 know, with -- since the Bloomberg interface has become 10 ubiquitous on Wall Street, that usually is your first 11 point of interface. 12 So at least for the -- from the perspective of 13 the institutional investor, you know, the first thing you 14 look at is all the Bloomberg runs that you have from 15 dealers that are an indication of where they would make 16 markets and bonds, and that's usually where you first 17 have a sense of general pricing, and then conversations 18 with dealers after that is where you're able to really 19 sharpen -- get a sharper idea of where real risk would 20 transfer in a given bond. 21 MR. SMITH: On that, with regard to the price 22 discovery in terms of TRACE and so forth, one of the 23 things that we, as a middle markets kind of smaller guy 24 -- you know, we're probably the David amongst the 25 Goliaths here. But you know, you look at a TRACE system, 0137 1 the vast majority of my clients face-to-face don't even 2 know what I'm talking about, and yet they're dealing with 3 dealers all day long, traders, brokers, that are honest 4 brokers, every day -- can't check them because they don't 5 know. 6 So I think a lot has to be done with regard to 7 education. But let's say they leave the money with us. 8 One of the things that would be helpful with TRACE is 9 what kind of price is it. Is it a dealer-to-customer 10 price? Is it a dealer-to-dealer price? 11 Give me some color because I know that there 12 are dealers that just trade amongst themselves. And all 13 of a sudden, bam, you know, I come in. I think I've got 14 a strong, firm market. Boomo, a whole satchel of stuff 15 just dropped on me, and all of a sudden I'm down a point. 16 Why? Well, I don't know who's actually posting that 17 price and why. How? Is it a sucker bet? Sometimes that 18 can happen. 19 And for the little guy who's not a frequent 20 trader, that's a problem because he can't see that 21 because he's not in the flow. So an infrequent guy 22 probably is getting -- you know, having a problem with 23 that kind of a system. 24 Is it from an ATS, or is it from a closed 25 trading club, you know? Where did that price come from? 0138 1 You know, was it a G session or was it something coming 2 off, you know, a trade web or rather a market access 3 trade? 4 Something along those lines to give me some 5 sense of how deep the market may be and who's -- where is 6 the price action coming from -- that would be extremely 7 helpful. And I don't think that that that's going to 8 expose anybody's interest. But I think it's important 9 for the guys who are the less frequent visitors. 10 And, oh, by the way, 35 percent of a 9 11 trillion/$10 trillion market is not a small population of 12 people. That's significant. That's the tail that will 13 wag the dog. And I think in today's world more 14 information, not less, on how paper is moving and from 15 which hands and why is a very important feature that 16 could be added to TRACE. 17 MS. HANDAL: I'd like to second that just in 18 general from the ABS perspective. While it is an 19 institutional investor market and is a large size market, 20 we would really like to see trades posted on TRACE. So 21 it's not -- it's not there right now. And a lot of it 22 has to do with our experience during the financial crisis 23 when there was virtually no liquidity in the market. 24 And as a buy-and-hold investor in particular, 25 you know, sort of the dearth of information in trying to 0139 1 figure out where things should trade versus where they're 2 being quoted was, you know, highly stressful to say the 3 least. And I would say that more information is better 4 than less. 5 And, in particular, if you limit it to certain 6 asset classes, even as a starter -- so Master Trust and 7 the credit card structure, which are frequently traded, 8 highly liquid, you know, AAA autos -- just to see how the 9 information flowed, I think, would be helpful. 10 We do the same thing as Jonathan here as far 11 as using Bloomberg. You can put in a CUSIP. You can put 12 in a different asset class and figure out where dealers 13 are quoting things. And you can look at both two-way and 14 one-way markets on it and figure out bid ask, but that's 15 now. 16 And in times of a stressed market when, as our 17 colleague from Vanguard talked before, when interest 18 rates rise and everybody's running for the door, I think 19 we're going to want more information rather than less as 20 a buy-and-hold investor. 21 MR. MCVEY: Just, you know, adding on the 22 corporate bond side, here again, I think technology has 23 played a significant role in improving pre- and post- 24 trade transparency over the last seven or eight years. 25 And you have on the retail systems an aggregation of 0140 1 dealer pricing and dealer inventory. You have the same 2 in the institutional markets. 3 There's a lot of detail available in TRACE. 4 The inter-dealer trades are separated from the client 5 trades. The trade sizes are very clear. There's a 6 sensible stop point at 5 million for a high-grade bond 7 and 1 million for a high-yield so you can see how pricing 8 works in different size trades. 9 And, you know, the Bloomberg runs are available 10 for market participants to aggregate in a convenient way 11 through advancements in technology as well. 12 So there are a lot of pricing tools that are 13 out there today that really were not there 10 or 11 years 14 ago. What I would caution is I think pre-trade 15 transparency in a less liquid market can actually go too 16 far, that there's a point where a client and a dealer 17 engaged in a trade negotiation on a less frequently 18 traded bond I think works better if that negotiation is 19 allowed to go to completion before it is reported 20 publicly because I think in the middle of that 21 negotiation, if you take away the private negotiations 22 that are going on, dealers will have less incentive to 23 commit capital to that trade, and investors are at risk 24 of seeing worse pricing. 25 So I do think the negotiation that takes place 0141 1 in a less liquid market is really important to 2 transaction cost. 3 MS. EDWARDS: All right. Thank you. Let's 4 talk a little bit about dealer inventory and inter-dealer 5 brokers for a moment. Dr. Goldstein presented some 6 somewhat surprising results on the inventory holding 7 periods for -- in the corporate bond markets. I guess 8 are there similar holding periods in the ABS market, and 9 is it surprising to the panelists? 10 MR. SCHLOSS: I guess -- and I don't have the 11 data, but I can tell you that from a standpoint of the 12 ABS market it is very much a buy-and-hold. And so 13 dealers aren't necessarily required to hold anywhere near 14 the amount of inventory that they would otherwise hold on 15 the unsecured side. 16 And it goes back to the type of security it is, 17 the type of structure, the work that's done by the 18 institution that's buying it, and so there just isn't the 19 same secondary market flow that would require large 20 dealer inventories. 21 MS. HANDAL: Actually, within ABS we do see, 22 you know, pretty decent-sized dealer inventories now. I 23 would say that prior to the crisis they were larger, and 24 then during the crisis they dropped -- you know, call it 25 50 to 100 percent depending on the dealer and the asset 0142 1 class. 2 But if you're talking in general across asset- 3 backed securities, you do see inventory, particularly in 4 assets where people are trying to make a market. For 5 example, if somebody is trying to bid up a lot of 6 subordinate student loan transactions, they'll have an 7 inventory if they're making a position. 8 I think the nature of the market is such that 9 in good times such as now and with, you know, sort of the 10 search for spread assets there are inventories. There 11 is, you know, support to the market. 12 During the crisis, I would argue that 13 inventories shrunk rapidly, went to zero, and most of the 14 trading that was done was crossing bonds. So riskless 15 trades between -- so it was the broker-dealer acting as 16 an intermediary between two end investors within asset- 17 backed securities. So it's a tenuous market as far as 18 inventory, and it's really going to be driven by 19 liquidity and the state of the financial markets. 20 MR. HORNE: And in the corporate market, you 21 know, that continuum that Professor Goldstein mentioned 22 about, you know, the more liquid bonds tending to have 23 longer dealer holding periods versus the least liquid 24 bonds -- actually, that is not at all surprising. It 25 makes perfect sense. 0143 1 And I think what it reflects is, you know, for 2 the more liquid bonds, dealers are willing to take risks 3 on those bonds because they have many more degrees of 4 freedom in terms of how to hedge them, whereas once you 5 go all the way down to CCC or C land, you're talking 6 about distressed, highly idiosyncratic situations where 7 there's really no ready way to hedge so that the only 8 time a trade does get done is when you're directly 9 matching a buyer and seller. 10 MR. GENYK: I would say as a non-bank dealer, 11 our inventories across asset classes have gone up over 12 the last five years. And you know, we're -- for us it 13 becomes a business decision and a risk management 14 decision. 15 I would say that, you know, corporates, in 16 contrast to the previous panel, the municipal panel, 17 municipals are very hard to hedge, as we heard, and 18 corporates are easier to hedge, and there's many 19 different ways to do it. But I think for the non-bank 20 broker-dealers, you're going to see it vary firm-to-firm. 21 MR. HEFFRON: Just to reiterate, I wasn't 22 surprised that the holding period for a high-yield bond 23 was shorter than an investment-grade bond as well because 24 I think it's priced to get off the dealer's balance sheet 25 as soon as possible, and it becomes a risk management 0144 1 type exercise. 2 In terms of inventories, you know, we as -- we 3 don't know what our customer's inventory is. We read 4 about them after the fact. But we definitely think that 5 some of the lower inventory months or quarters have led 6 to more electronic trading, particularly when you can 7 trade in the middle of the spread on a matching session 8 or an auction session or, you know, some sort of 9 Exchange-type session. 10 So that if anything it's led to -- between the 11 capital restraints and the Volcker Rule and some of the 12 things that the banks are having to contend with, it's 13 led to more of a agency style of business for some of 14 them. 15 MR. PITT: I would totally agree with the 16 comments that were made about the holding period. What 17 first seems kind of funny about the lower rated stuff, a 18 much higher percentage of that business is done on a 19 riskless principal basis. 20 I would agree with the panelists from the last 21 -- the initial panel that riskless may not be entirely 22 riskless but certainly that you have a buyer and a seller 23 at the same time. 24 Looking at our business in investment grade, a 25 very low percentage of trades that we do are riskless in 0145 1 that fashion. We are typically providing liquidity with 2 our own balance sheet on almost every trade that we do. 3 And, you know, on some of the public figures 4 that are out there about inventory, one of the 5 distinctions that I'd like to make is just between net 6 inventory or net balance sheet and gross inventory or 7 gross balance sheet. So if I look at J.P. Morgan's 8 figures pre-crisis to now, our net inventory is 9 definitely lower. In other words, we're just not running 10 a big long book. 11 I think yields are very low, and it's not very 12 interesting to me to run a big -- I'm not an asset 13 manager. I'm a liquidity provider. However, if I look 14 at my gross balance sheet, which would be the sum of all 15 the positions, be they long or short, it is down by a 16 much, much lower proportion. 17 So you know, there is a different risk 18 management culture for sure that pervades the industry, 19 and net balance sheet is down. But it's down in a way 20 that is not at all proportionate to the decrease in net 21 balance sheet that's being provided. 22 So I have a big, complex book of longs and 23 shorts that are being used to offset each other. When an 24 account wants to buy a bond that I don't have -- well, if 25 that's a bond that doesn't trade at all in a year, I 0146 1 probably wouldn't make a short offering in that because 2 I'll never find it again. 3 But it if it something that I think there's a 4 reasonable chance of finding in the market, I can be 5 short that and I can buy something else against it. And 6 when an account wants to sell me a bond, I mean there's 7 really almost no bond that we wouldn't buy because there 8 are not enough bonds around, and everybody wants bonds. 9 So liquidity is very easy when people ask me to 10 buy things, and it's sometimes trickier when people ask 11 me to offer things, but the gross balance sheet around is 12 definitely not down as much as the net. 13 MR. HORNE: I think it's also important, and 14 there's been some recent -- have been some recent papers 15 written on this. But I think a lot of people will look 16 at the Fed data on dealer inventory and, you know, look 17 at just the cataclysmic drop post-crisis in that data, 18 but that data includes both non-agency mortgages and 19 corporate's. 20 And so, you know, a lot of that is the decline 21 in the price of non-agency. Is it isolated to the 22 corporate market? Yes. Dealer inventories are certainly 23 materially lower than post-crisis, but it's not nearly as 24 much of the "C change" as the Fed data suggests. 25 MR. SMITH: One of the things I would add, 0147 1 going back to the ABS and CMO and other structured paper 2 with the madam down at the end, I mean it's important, 3 you know, when it's important. When it's not, nobody 4 cares. 5 The fact of the matter is that pricing 6 requirements are a function of where the market is 7 moving. And the more excited a market becomes, the 8 tendency is that people want more information about it. 9 So ABS markets are largely defensive in 10 nature, front-end paper, you know. That's not where 11 yield is right now, and some people aren't really that 12 concerned about where is price discovery, what was the 13 history and so forth. 14 But we, again, I think we need to prepare for the 15 future. People start rolling down the curve, rates go 16 up; I want to know more about the ABS market. I want to 17 know more about short-dated CMOs. Give me prices. Give 18 me prices. 19 The fact is that we're in a void. You see -- 20 you get air pockets of information. Price discovery is a 21 matter of hit and toggling through your Bloomberg screen 22 everyday. Who's got a run? Who's got a run? Who's got 23 a run? There's no central system, not a good enough one. 24 And that's time-consuming, if not costly, in terms of the 25 moving market. 0148 1 So we could do a lot better in anticipation of 2 where because we're all sitting here worried about the 3 great rotation, rates going up. Well, there are some 4 things, some informational voids that we can fill to help 5 people with that adjustment. 6 MS. EDWARDS: And how important are inter- 7 dealer brokers today in managing dealer inventory? 8 MR. HEFFRON: Incredibly. I'll let somebody 9 else start with that. Okay. So we play an incredible 10 role in managing, you know, risk and exchanging risk 11 between institutions. The role of the inter-dealer 12 broker in corporate bonds has changed pretty 13 dramatically. 14 We were traditionally a CDS house, and we got 15 into cash corporate bonds I'd say couple of years ago, 16 maybe three years ago. 17 And as I said earlier, over the course of 18 time, our business is now nearly 50 percent electronic. 19 And of that 50 percent, 100 percent of the 50 percent, a 20 very high percentage of the 50 percent is in matching 21 trades where there's a curve that we put out to all our 22 customers, and they can choose to buy or sell at the 23 point. If there is a buyer or a seller opposite them, a 24 trade is executed immediately and sent to TRACE. It's 25 completely electronic. 0149 1 So that's about half of our corporate bond 2 business. The other half is in, you know, deep, illiquid 3 distressed bonds that can be anywhere from a million to, 4 you know, 30, 40, 50 million. And that's a negotiated 5 process between buyer and seller. It's anonymous. We 6 clear the bonds through GFI. 7 So, you know, that's been a hybrid that's 8 slowly growing more and more electronic. I'd stress that 9 it's not relied on in a central limit order book because 10 that's just not been the style of how people want to 11 trade. It's much more of a session based trading 12 environment. 13 A lot of that has to do with, you know, a 14 credit event taking place and a guy not wanting -- a man 15 or a woman or a trader not wanting to leave out a price 16 or a multitude of prices across, you know, your screen. 17 Sort of the opposite side of the coin is our 18 index business, which is because it's less sensitive to 19 any single piece of news, is much more sort of 80 percent 20 central limit order book and then 20 percent matching. 21 So those are the roles. It's the business has 22 become incredibly more sophisticated, particularly for 23 those who have invested in technology and are prepared 24 to, you know, report the data to TRACE. I think it needs 25 to be there in 15 minutes or probably within a couple of 0150 1 minutes. So that's my pitch. 2 MS. EDWARDS: Okay. Well, let's move on to 3 talking about dealing with customers. So how -- how do 4 brokers deal with retail and institutional orders 5 including the process for identifying an appropriate 6 security to buy and finding potential sellers and buyers? 7 And how do brokers provide customers with best execution? 8 MR. GENYK: I'll speak to both the retail and 9 institutional because our firm has both. From a retail 10 perspective, we use ATSs extensively, you know, 90 to 95 11 percent of our volume. We use a request for quote 12 structure, and then we make all of the line items on the 13 ATS inventory available to our clients for purchase, and 14 that's all on the retail side. 15 From a best execution standpoint on retail, 16 it's pretty straightforward with TRACE. You know, we 17 make sure that we have a commission grid, and we make 18 sure that every trade is done in the context of the 19 market after inclusion of the retail broker commission. 20 And we use TRACE data to do that. 21 Institutionally it's a far different market. 22 We're not like my colleague from J.P. Morgan, and so we 23 operate in a far different business where we run smaller 24 business. Our average trade size institutionally is 25 about a million dollars, and that's largely done in the 0151 1 traditional voice over-the-counter methodology. 2 MR. PITT: It's difficult for me to speak to 3 the retail part of the business, but you know, one thing 4 I wanted to mention, as the number of small trades in the 5 corporate bond market go up and the percentage of volume 6 of the bond market that's done in small size trades goes 7 up, it actually -- it seems like it jumped up all at once 8 and it isn't really going up any more, but it did go up. 9 We are investing pretty significantly in 10 technology that allows -- you know, because a typical 11 trader on my desk might look at 50 bonds most of the time 12 that they're trading, and they're doing larger size 13 trades in those bonds, and it's difficult for them to 14 deal with a high number of smaller inquiries over the 15 course of the day. 16 So we're investing heavily in technology that 17 would allow us to sort of automate more of our responses 18 to inquiries and allow us to integrate more effectively 19 with all these new venues that are gaining traction in 20 the marketplace. 21 So, you know, while we don't deal directly 22 with the individual investor, we are building out our 23 ability to provide liquidity through all these sources 24 that eventually get to the individual investor. And 25 frankly I think we're intending, as we build out that 0152 1 technology, to deploy more, not less, balance sheet in an 2 effort to provide two-sided prices across a wider array 3 of bonds. 4 You know, dealing with the institutional 5 market is obviously a very different -- you know, a 6 different process, and the investors typically have done 7 a lot of homework and are quite sophisticated. So they 8 often know more than we do about the landscape of, you 9 know, pricing available to them in the marketplace. 10 MR. MOLLOY: I'd just like to add that, you 11 know, given -- as the Exchange, we do provide a lot of 12 executable quotes and pre-trade transparency. We do 13 think that, you know, that's something that the community 14 should consider when evaluating best execution. 15 MS. STARR: Hi, I have a question. With 16 respect to both the pricing as well as the liquidity, 17 what are the differences between the registered market 18 and the 144A market? 19 MR. SCHLOSS: Again, I'll -- from an issuer’s 20 perspective, we don't do a whole lot of 144A because of 21 our size. But the little experience we have is that they 22 are a lot less liquid, and they price at -- on a primary 23 basis they're going to be anywhere between 20 and 50 24 basis points wide of where a registered would be. And 25 we're even finding that the 144As that we've done -- the 0153 1 demand for reg rights and the demand for getting them 2 registered has a meaningful benefit to the investor from 3 a standpoint of overall liquidity. 4 MS. HANDAL: I would actually -- from the 5 standpoint of an investor in asset-backed securities, 6 it's a little bit different because it is an 7 institutional market, so predominately many of the 8 players -- most -- can invest in 144A and public 9 securities. 10 And in general, the difference in the market is 11 the type of securities that are traded off of the 144A 12 shelf. They tend to be a little bit more esoteric, and 13 CLO is more highly negotiated transactions. 14 So from the standpoint of liquidity, there is 15 some element of it where pricing is a little bit wider 16 due to liquidity, but a lot of it is really just due to 17 the nature of the asset classes that trade that way. And 18 we've actually seen pricing in the 144A market sort of 19 more in line with the public market now than it has been 20 ever before in the past. 21 MR. RAMSAY: Yeah. I was just going to ask, 22 circling back to the municipal market, again, for -- if 23 any of you have a perspective on it, to the extent that 24 transaction costs in municipals tend to be substantially 25 higher than they are in corporates from the data that I 0154 1 have seen, do those differences make sense to you? Do 2 they -- are they justified? Do you understand them? Are 3 there -- does anybody have any thoughts about that? 4 MR. SMITH: As a risk manager, I can 5 understand some of that because, you know, clearly it's 6 hard to find matchers. It's a -- you want to talk about 7 fragmentation. I mean they're the poster child. And so 8 obviously that's going to enter into the willingness to, 9 you know, make a market. 10 I think though that at this point in time what 11 is starting to happen is that because of the reduction in 12 rates that we've seen, you know, everything is elastic 13 except for the spread, and then the spread starts 14 gobbling up more and more of the net return. What do you 15 think happens to the market? It stops because nobody can 16 make money on a after transaction basis. There's nothing 17 left on the table, and that just promotes more 18 illiquidity. 19 So you could say some of it is the highly 20 fragmented nature of the market, but it's also the 21 extremeness that we find ourselves in in terms of what 22 the return is on the capital that the investor is putting 23 forward and what the transaction costs are merely, I 24 think, is a reflection of how difficult it is to deal 25 with the fragmentation. 0155 1 And I can understand some of it. I would hope 2 that there would be greater efficiencies. I listened to 3 all the technology gurus that we have at the table. I'm 4 sure something could be done better than there is in 5 muni’s because we trade muni’s as well as corporates. 6 But we are talking light-years of difference in terms of 7 the realities of how to get through a day. 8 MR. HORNE: I mean I think you can within the 9 corporate market itself, you can see a continuum of a 10 certain subsection of it. It does look a lot like 11 municipal bond markets, so, less elements, less liquid 12 elements of high yield, you know, smaller issuers that 13 will only issue once every three years, or something like 14 that, and are under 500 million total issue size. 15 That looks a lot like the municipal market in a 16 lot of ways, and, you know, I think the difference 17 between those sorts of securities and say the 30 percent 18 market value of the high yield market that trades reasonably 19 frequently, or the more liquid parts of the investment 20 rate market, is that in the former, your holders of bonds 21 are fairly concentrated in terms of their horizon type, 22 in terms of what motivates them to buy or sell. 23 They're more, you know. It's going to be more 24 buy and hold; don't move things around that much. You 25 don't have any speculative involvement, generally, in 0156 1 those sorts of securities, they are hard to short, et 2 cetera, et cetera. 3 Whereas, the more liquid parts of the market, 4 BB and single B high yield, financials and investment 5 grade, there are a lot of different kinds of market 6 participants in there with all sorts of different time 7 horizons, you have a lot greater speculative involvement. 8 So it just makes it more liquid. There's a much greater 9 likelihood that there will be someone else that wants to 10 trade on the same day that you do on those kinds of areas 11 of the marketplace. 12 MR. SMITH: I think just an aside, just to 13 think about what drives the market sometimes, even in the 14 muni's, you look at the BABS market. There's a distinct 15 difference between what's included in the BABS Index and 16 what's not in terms of liquidity and markets. And so one 17 of the things I would think that the Commission should 18 look at is we are all performance oriented in the risk 19 management world, you know. 20 And the dealer community is here to kind of 21 help us get through that, but this whole thing known as 22 indexes really is a driving force in terms of where is 23 liquidity. Because if you're outside of what's inside 24 the Index, then from a relative performance standpoint, 25 you're taking a big risk. 0157 1 Now that may pay off in terms of alpha, but the 2 fact of the matter is that most indexes that are put out 3 there for performance bases are understood by the buying 4 public to be really liquid reflections of reality. And 5 they're so far from it, it's not even funny. 6 But at the very least, going back to the 7 Exchanges, why can't we get people looking at these 8 indexes, at least the corporate indexes, of across the 9 ETF complex of corporate bonds? You've got a thousand 10 securities across all of them, 1,100 securities across 11 the top six that control $50 billion. 12 Why not have markets, at least in those top 13 1,000 securities, or some form of a price that people can 14 sit down and say, okay, I know where they got the value 15 of this thing. There are things that are there that we 16 should be thinking about that are forces and influences 17 in the fixed income markets that go way beyond what a 18 dealer does. 19 But we should be appreciative of the fact that 20 as a risk manager, I have to live in that world. Nothing 21 could be more apparent than in the BABS world, in muni’s, 22 in terms of the difference in prices, because it's in the 23 Index or out of an Index. 24 MS. EDWARDS: In the earlier panel today we 25 talked a lot about the role of negotiation in getting to 0158 1 the price. I guess what's the role of negotiation in the 2 corporate markets by institutions or by retail investors, 3 and how much do the investors ultimately know about the 4 compensation that the dealer is getting in transacting 5 the bond? 6 MS. HANDAL: I can speak from the perspective 7 of the asset-backed securities market, if that helps. So 8 to answer your question on a new issue, ABS, it's not 9 highly negotiated, because it's essentially the issuer 10 will come out and say here's what we want to offer, and 11 they'll do price talk. 12 And dealers will go about and say are you 13 interested here or not, so you pretty much know where 14 spreads are going to come, and you can push back a little 15 bit by saying no, I'd be interested in X amount of debt 16 spread and Y at the next spread. 17 So there's a little bit of negotiation there, 18 but given the great rotation, as mentioned before, and 19 the way that deals are flying off the shelves right now, 20 there's not a lot of negotiation. It's pretty much here's 21 the price and that's where you're buying it. 22 When it comes to secondary trading there is 23 more negotiation and it just depends on whether it's a 24 bond in inventory or if it's a bond that the dealer is 25 sourcing through another client, and you can go back and 0159 1 forth, and there is negotiation there. When it comes to 2 transparency, the dealer compensation, that's really 3 opaque right now. 4 And I would say that's one of the issues coming 5 out of the crisis we really were made aware of, you know, 6 the inability to know how much broker-dealers are making 7 on any given trade. 8 And particularly if it's a trade that's in 9 their inventory and they've taken the risk, that's fine 10 and that's pretty much held for a certain period of time. 11 It makes sense to us that it's a little bit more opaque 12 when it's crossing bonds between two end investors. We 13 feel that it should be disclosed. 14 MR. PITT: As we've seen, the vast majority of 15 the traded volumes, particularly in the institutional 16 market, are concentrated in a small number of issues. 17 And, typically, even without TRACE, the institutional 18 investor base that I mentioned before has a very clear 19 picture of where those liquid bonds are, because they're 20 getting multiple runs from large dealers that are all 21 going to line up very closely. 22 So generally speaking, there's pretty good pre- 23 trade transparency on the stuff that actually trades, and 24 then the things that don't trade as much, there's less 25 pre-trade transparency, because no one really knows 0160 1 exactly where that bond should be until there's a buyer 2 or a seller of that bond, in which case the negotiations 3 kick in. Typically, when an institutional trade goes 4 through as riskless principal, in other words, you have a 5 seller and you take an order to go try to find a buyer, you 6 know. 7 Much of the trading is in TRACE bonds, so if 8 it's an infrequently traded bond, then everybody's going 9 to know where both sides of the trade happened, and then, 10 of course, the liquid stuff, it all happens in a very 11 sort of well-illuminated type market anyway. 12 But you know, a large number of the trades we 13 do are principal trades, so we're just saying this is 14 where I'll buy or this is where I'll sell, at which point 15 there's no longer really a conversation to be had about 16 mark-up or bid offer spread, because you're taking risk. 17 Once you've taken risk, things can start to 18 happen. There could be news. There could be macro news, 19 micro news. You might have -- actually, we maybe are not 20 doing a good job managing our inventory, but my inventory 21 turnover for specific positions is significantly longer 22 than the slide that we saw earlier this morning. We tend 23 to have most positions on our books for a couple, three 24 weeks. 25 That's an average, you know, which is probably 0161 1 skewed by the fact that there is some amount of riskless 2 principal trading that goes on. So for things that we're 3 actually principaling the average holding period is 4 probably higher than that, at which point, you're buying 5 something that you sold a couple weeks ago or selling 6 something that you bought a month ago. There's no longer 7 really a mark-up or bid offer spread that's really being 8 crossed. 9 MR. MCVEY: Yeah. Just to add to that, I agree 10 with all of the points that Eric made, but just to put a 11 little bit of math behind some of the TRACE data, we've 12 looked very closely at the high grade market, and if you 13 had the advantage of having all of the customer orders in 14 one place at one time during the day, you could match 15 about 15 percent of daily trade volume, and nobody has 16 access to that book at one time during one day. So the 17 vast majority of trades are done with dealers as 18 principal. 19 And so to your question about what do the 20 customers know about the dealer compensation, the dealers 21 don't even know. They're taking on that risk and they're 22 going to manage that over a period of time until they can 23 find the other side of that trade. The actively traded 24 bonds, there's plenty of information for everyone. 25 If they're trading frequently, they're 0162 1 available on TRACE. People do have an understanding of 2 where they trade. TRACE is marking client buys versus 3 client sells. It's very easy to determine bid offer 4 spreads in those instruments for all kinds of different 5 trade sizes and give investors confidence that they're 6 getting a fair price. 7 The second part of this is in the institutional 8 market. Institutional investors do take great comfort in 9 the competitiveness of the auction or the RFQ process. If 10 they're going out to 20, 30, 40, 50 dealers in some cases 11 to get competitive pricing back on either a single 12 security or a list of bonds that they're trying to buy or 13 sell, they take a lot of confidence that they have gotten 14 the best execution and they found a competitive price for 15 the bond that they're trying to trade. 16 MS. HANDAL: Just to reiterate that, I'd agree 17 with you on a BWIC situation where you're putting bonds 18 out to a bunch of dealers. But, I wanted to also 19 reiterate that talking about ABS where we don't have 20 TRACE information on bonds that have traded, so from the 21 perspective of if you're a buyer, not a seller, and 22 you're not putting out a large BWIC, we don't have that 23 trade information. 24 And also during -- and I agree wholeheartedly, 25 if a dealer is taking a bond into inventory, then their 0163 1 mark-up is their business. What I was referring to 2 earlier was strictly on riskless trades where they're 3 crossing bonds between a buyer and a seller, and that's 4 more in a highly illiquid period, so, during the crisis 5 when people weren't taking inventory, or even now today 6 with some of the smaller deals that go through. 7 MR. GOLDSTEIN: A couple quick things. My data 8 did not include riskless principal. Those were all 9 principal trades. And I next want to tell a story. I 10 want to highlight -- I don't have data behind this story, 11 because it's just a story I'm kind of making up to match 12 the data, but here's my guess as to one of the reasons 13 why we see this. It's a very logical thing. 14 Something that trades 30 times a day, right, 15 and the dealer, if they're going to buy it, they should 16 take it into inventory. They have a reasonable chance 17 someone's going to call them in the next couple of days, 18 and they're acting as a dealer. But here's my story, 19 which may or may not be true for things that hardly ever 20 trade, like say once in the last month or zero times last 21 month. 22 Somebody calls you up and says I'd like you to 23 buy this or I'd like to sell this to you. Well, if 24 somebody asked me -- I'm just pretending I'm a dealer now 25 -- if somebody asked me for a quote right now, I'd give 0164 1 you a ridiculous quote, but I'd probably say here's my 2 ridiculous quote, give me 40 minutes. 3 Because one of the other things dealers have is 4 they know customers. They know who might have it or they 5 know who they might be able to place it to. I'll give 6 you a quote. No one has to worry about the market 7 running away. It hasn't traded in 30 days. It's not 8 going to trade in the next five minutes. And I'll call 9 you back. You know, if you haven't executed yet, I'll 10 maybe give you a better price. 11 They call around. They don't have the order, 12 but they call around and get an idea of where they might 13 be able to place it. They're doing what book building 14 would be, you know. They're getting some information and 15 they give somebody a much better price. 16 And then as a result, they're able to turn it 17 around within the day, and they don't charge everybody 18 that much money for it, because they have trouble taking 19 the risk. They don't have a firm order on either side, 20 but they are using the fact that these aren't anonymous 21 customers, is I guess the way I'd want to say it, that 22 they actually have an idea. 23 If you call me, I can always call you back, and 24 in the meantime I'll call PIMCO or I'll call MetLife and 25 say to Metlife look, I notice you bought this. I know 0165 1 you have this in inventory, because I sold it to you 2 eight months ago, you know. It's probably hedging 3 something else. I have a different bond in inventory 4 that has the same duration, a lot of the same 5 characteristics, you know. Can I kind of swap this out 6 with you or something like that. 7 Now, that's just a story. I have no proof that 8 that happens, but it's consistent with the data, and 9 what's nice to see is they seem to charge you less when 10 they go through the search function. 11 And it is still a risky trade. I want to 12 highlight I'm not saying it's doing anything illegal. 13 It's still a risky trade. There's not a guarantee on 14 either side, but to be perfectly honest, if you called me 15 up about something traded once in the past year, I would 16 probably say I know neither one of us have a big rush on 17 this. Let me take a moment to look at it. 18 And it's consistent with one of the academic 19 papers I cited, which said look, it kind of turns out 20 that the bonds that tend to get held by people who trade 21 more have lower trading costs, because their theory is 22 dealers know where to go to find it or offset it. It's a 23 little easier for those bonds and the bonds that hardly 24 ever trade. I just want to highlight I've no data. That's 25 just a guess. It's a consistent story, and I think it's 0166 1 economically consistent with what people do. 2 MR. SCHLOSS: I guess I'll go back to the 3 original comments that were made from the standpoint of 4 negotiations between issuer and investor. And really, 5 you can almost draw yourself a grid, depending on sort of 6 investment grade, high yield, good market, bad market. 7 And the leverage point that exists from a pricer or a 8 price taker really depends upon the market environment in 9 which you're in. 10 When we come to market, we will issue unsecured 11 or ABS for that matter. We have a pretty good idea of 12 where the market is, not necessarily of where it is in 13 size, and that's what we use, sort of the book building 14 process, and in today's market where the market is very 15 strong, there's less negotiations, but in a bad market or 16 in a difficult market, investors can somehow drive the 17 pricing a little bit more than they would otherwise. 18 I think the other thing we're finding is as 19 several of the investors are getting extremely large, 20 they tend to have more voice in the process, and so as 21 you're building the book, the more you can narrow the 22 discussion down and build your book around the largest 23 investors. 24 You're taking a little bit of leverage away 25 from them, but as investors are getting bigger and bigger 0167 1 and more clout, and we heard here earlier this morning, 2 you know, how they're buying a large amount and then 3 divvying it up among accounts, and as they get bigger and 4 bigger and bigger, they get more and more leverage in the 5 process. 6 MS. STARR: I have a question. In terms of the 7 types of deals that are done say in the initial offering, 8 are they done where you have more of a book building, you 9 know, investor identification process before the deal, or 10 are the dealers doing them as bought deals and then 11 finding investors? 12 MR. SCHLOSS: Yeah. I think in this 13 environment, if you go back 10 or 15 years, there were a 14 lot more bought deals done. In today's world where the 15 balance sheets of the institutions aren't being used, it 16 is all book building, and our deals tend to be very 17 transparent, or they're very transparent with us as the 18 books get built with who the investors are, and then we 19 actually get involved in the allocation process. 20 That's a lot easier now that we're investment 21 grade than it was when we weren't from the standpoint of 22 the choice of investors we have, but we are actively 23 involved, because we want the distribution, from an 24 institutional perspective. 25 We also run a retail program where we price 0168 1 weekly, comparable to institutional levels that allow 2 retail investors to buy directly at fixed fees from a 3 standpoint of what the brokers get, and then costs for us 4 comparable to institutional. 5 MR. SMITH: I'd like to add to that, again, 6 book building is great, but I think we've gotten into a 7 little bit of a bizarre situation where you get over 8 subscriptions of two, three, four and five times for 9 relatively hot deals, because everybody's piled into the 10 same canoe. We're at the lower end of the spectrum in 11 the sense of we've got $11 billion under management. 12 With all due respect to some of the large guys, 13 you know, they come in and they all take a book, and 14 we've gotten to a point now where I may have an interest 15 for 5 to $10 million, but I actually have to go in and 16 ask for a hundred million and end up with a million on a 17 good day -- on a good day. 18 That is a huge risk to my business, because one 19 day it may be you've got a hundred, and I don't know how 20 to get around that, but it's scary. But the other side 21 of that -- I read in CIO Magazine recently that there's 22 concern on the part of the middle markets people, the 23 lower, the smaller cap guys, how do I raise capital. 24 Because if you can't raise $250 million or more, don't 25 even think about going to that bond market. You are shut 0169 1 out. Nobody wants to trade your bonds. 2 So think about everything else, but don't think 3 that you're getting anywhere at the corporate bond 4 market. That's now how capital formation is supposed to 5 occur in our economy. So again, the liquidity concerns 6 are now backing up the system in my view, and I think 7 that has -- if you want this economy to have legs, you 8 want growth and prosperity, it's got to filter down 9 beyond the big guys. 10 How you get beyond that in terms of capital 11 creation, you know, has to be addressed at the market 12 level because if they're shut out because they can't get 13 efficient pricing because it's not big enough, because 14 ah, I can't make a market, who cares? Then we have a 15 problem in the long term. 16 MR. PITT: A couple of comments on that. The 17 first one is, you know, the small bond issues in the 18 capital formation process that you were referring to may 19 be seen in the municipal market. If you have very small 20 issues, it's very difficult to get liquidity in those 21 issues. 22 So I would argue that securitization of debt 23 may not be the best lending market for small issuers. 24 You know, there is a very important role that bank 25 lending is supposed to play in financing small size 0170 1 companies, and it's just very difficult to educate 2 investors on a very small issue, and as the municipal 3 market, I think, bears out. 4 And the only other comment that I would make is 5 please don't feel singled out at all in the frustration 6 that you have in getting allocations on deals. I think 7 that our larger investors would have a very similar 8 comment about their inability to get the size that they 9 want, and it's no different from the middle size 10 investors. This has been a market where there has been a 11 structural bond shortage. 12 Net issuance of spread product, when you factor 13 in coupons and central bank buying, is sharply negative; 14 so it's a contracting market in an environment where 15 investor desire to have access to fixed income, whether 16 it's at an individual level or whether it's based on 17 regulatory changes and pension fund accounting, there has 18 just been an increasing appetite for -- you know, there's 19 more and more demand and there's less and less product. 20 So it's a pretty structural issue, and good for the 21 issuers, by the way. 22 MR. SMITH: It's great for the issuers. I'll 23 take your problem any day compared to mine, but I think 24 the issue here is distribution. You know, we want distribution 25 or we don't. Do you want a narrowly defined market 0171 1 controlled by a few, or do you want a widely distributed 2 market so everybody can enjoy at least some degree of 3 liquidity, and failing to give people something on the 4 break on a new issue, as the Professor and others have 5 said, look, most of the juice is in the front end, okay, 6 of a deal. As it comes to rest, pricing action quiets 7 down. People tend to say okay, fine. The float's going 8 to be less. I'm going to hold them. 9 I think that in the broad interest of the 10 marketplace, to enhance liquidity, you have to have more 11 players, not fewer players. To enhance liquidity, you 12 have to have more price discovery, not less price 13 discovery. 14 Right now, we're talking about an ecosystem, a 15 collection of vendors, that looks like a Darwinian 16 experience for the small to average guy, and we may 17 eventually evolve into a more collective system, but the 18 buy side in my view has to push for a more collective 19 system. Give me a mall where I can go to specialists 20 within the mall, but I have a central trading system. 21 We are very much in favor of a more collective 22 system than a disaggregated system, because I feel like I 23 have to be a fly to watch all the screens every day on my 24 trading desk. They're very unconnected in many different 25 ways. 0172 1 So you've got closed networks. You've got 2 networks like MarketAxess. You've got guys that are on 3 line. Who knows, you know, Craigslist may get into the 4 business. 5 But the fact of the matter is -- I'm not trying 6 to be glib about it, but it's getting harder and harder 7 for risk managers to find where they can get in and out 8 of markets efficiently on every single day. And not to 9 get on the back of Mr. Edelman's comments, but we are in 10 a very extended risk posture in the investment management 11 community. 12 Realize that the average intermediate bond fund 13 out there in corporate land has now something close to 40 14 percent in BBBs or less securities. Five years ago, it 15 was 22 percent. Okay? 16 Realize also that the average maturity on 17 trading corporate bonds has risen by 83 percent in the 18 last decade to 14.5/15 years from 8. So from both a 19 duration and from a credit quality standpoint, we're 20 super extended, and 80 percent of the bond Funds out 21 performed their indexes last year. 22 But the first quarter of this year, for the 23 first time in five years, we had a negative print on the 24 aggregate bond index. Everybody's feelers are up, and 25 I'm not sure that the market mechanism here needs less 0173 1 distribution, it needs more. It needs more players at 2 this point, not fewer. 3 MR. HORNE: To comment again on the liquidity 4 point and access to market by smaller issuers, yes, 5 liquidity is certainly worse than it was pre-crisis. I 6 don't think that's necessarily -- I don't think you can 7 blame that necessarily on just market concentration. 8 I think it has more to do with (a) the rise in 9 fundamental, economic uncertainty in the marketplace, 10 which then on behalf of the ultimate investor, because 11 PIMCO's not the ultimate investor, our investor is the 12 ultimate investor. They want products that have shorter 13 term liquidity terms. They want to be able to exchange 14 or redeem MetaFunds very quickly, and you even see this 15 in hedge fund space. 16 You've had a huge push towards more liquid 17 versions of things, so there's very little ultimate 18 appetite on the ultimate investor's perspective to invest 19 in something that they have to hold for five years if 20 there's not liquidity there. 21 So once that's the case, you immediately start 22 to push out access to securitized debt for very small 23 issuers, because it's not going to be liquid, and there 24 aren't enough pockets of investor preference that are 25 willing to hold things for a long time, if they have to. 0174 1 So I think it's a product of the investor behavior 2 changes that you've seen post crisis. 3 MS. EDWARDS: Okay. One last question before 4 we break for lunch, and I'm sure everybody's hungry at 5 this point. But we've talked a bit and moved on to 6 talking a bit about the primary markets and a bit about 7 the secondary market. 8 What effect does the market structure of the 9 secondary market have on the efficiency of the primary 10 market? And when we think about the primary market, we 11 can think about both registered and unregistered issues. 12 MR. GOLDSTEIN: All right. I'll take a shot 13 very quickly. So let's imagine you changed the market 14 structure so nobody could ever trade after a bond was 15 placed. So you issue the bond. It goes somewhere. You 16 can't touch it and never sell it. 17 That would cause issuance costs, yields that 18 you have to go to pay off. It would just be more 19 expensive, so a very short version of what the research 20 indicates, my own and others', is when market structure 21 changes in a way that enhances post-issuance liquidity 22 that turns out to lower the cost to capital for the 23 issuing body in a way that one would hope would 24 ultimately increase. 25 The research. I haven't gone on to see does 0175 1 that increase employments and other things, but general 2 finance theory would suggest if you can lower the cost of 3 capital, it would be good. So the very short version -- 4 I'll let everyone else chime in -- is whatever change the 5 market structure increases liquidity in a way, it will 6 hopefully lower the initial cost and in effect cost to 7 capital, and hopefully our cars will get cheaper. 8 MR. HORNE: Yeah. I mean just think about the 9 process that we would take when we look at a new company 10 that might be coming to market. You know, we're going to 11 do our fundamental credit analysis. We're going to think 12 about what is a fair compensation for credit risk for 13 this company, and then we're going to apply a liquidity 14 premium, if there needs to be a liquidity premium, and 15 something that's like a U.S. bank that's extremely 16 liquid, that trades all the time. 17 You know, our reservation price is going to be 18 driven almost entirely by fundamentals. If it's a 19 company that we've never seen before, even if we really 20 like it, we're going to charge some liquidity premium 21 that we think reflects the cost to be able to get out of 22 it if we need to. 23 MR. SCHLOSS: I think the other piece that fits 24 into this is the size of the transaction itself on 25 primary from the standpoint of enough liquidity to where 0176 1 it freely trades for some period of time after issuance. 2 Now obviously over time, as the data -- and I think our 3 securities would support the data -- it will trade. It 4 gets too big, it gets expensive. So there is a size that 5 you need to be, but don't go too big without having to 6 pay a penalty, and there's a size too small that you will 7 pay a premium for as well. 8 And that's evolved. You know, for us it used 9 to be half a billion dollars was about the right size. 10 Now if it's not a billion, it's going to have trouble 11 from an overall pricing perspective. But in some cases, 12 a billion and a half might be too big, and so there is 13 that, and it will vary within the market. 14 ABS is smaller and it's also because there are 15 different tranches. You have different types of 16 investors who play into that, and as I said earlier, our 17 ABS is a short duration security, so the secondary 18 trading in it is less than you might see in other ABS 19 markets. 20 MS. HANDAL: And I'd follow that up with just 21 the fact that if you look at the issuance in the ABS 22 market pre-crisis, sort of during crisis and post- 23 crisis, a lot of it has to do with the perceived 24 liquidity within the secondary market. So if you look at 25 peak issuance was in 2007 at about $375 billion, and then 0177 1 we saw a precipitous drop-off to about $150 billion in 2 2008 and 2009, settling in at about 100 in 2010. 3 What happened there is you started to see more 4 secondary trading. People got more comfortable with 5 where markets were. There was more inventory buildup and 6 more trading along the way, so that the market has healed 7 and recovered to the point where last year we did see 8 about $240 billion of issuance. 9 So ability to trade on the secondary market, 10 even for shorter securities, does drive the ABS primary 11 issue mark, and I'd reiterate, when we look at 12 auto-backed ABSs or very short duration, we use them as 13 cash substitutes, but it's also very helpful for us to 14 know that we can trade out of them if we need to. 15 MS. EDWARDS: I guess we'll end just a couple 16 minutes early if none of the other panelists have 17 something to add. We'll now have an hour for a lunch 18 break before the start of our third panel. As you leave 19 the building, please leave your visitor pass at the front 20 desk and allow sufficient time to go through security 21 screening upon re-entry. 22 The third panel will start at approximately 23 1:45, and thank you very much to the panelists for coming 24 here today and sharing your thoughts. 25 (A lunch recess was taken.) 0178 1 A F T E R N O O N S E S S I O N (1:48 p.m.) 2 PANEL 3: POTENTIAL IMPROVEMENTS TO THE MARKET 3 STRUCTURE FOR MUNICIPAL SECURITIES 4 MR. CROSS: Why don't we go ahead and get 5 started? I'm John Cross from the Office of Municipal 6 Securities, and this is the welcome you back from lunch, 7 and this is the third panel, in which we are going to 8 discuss -- this is the answer panel. We are going to 9 discuss ideas to improve the structure of the municipal 10 market and transparency, liquidity, and other structural 11 aspects. 12 Let me quickly introduce all of our very 13 distinguished panelists. We had a great discussion this 14 morning. 15 Starting at this end and going down, Robert 16 Auwaerter, Principal and Head of the Fixed Income Group 17 at Vanguard. Joseph Hemphill, CEO at Regional Brokers, a 18 broker's broker. Burton Hollifield, Professor of 19 Financial Economics at Carnegie Mellon University, who 20 gave us a nice presentation this morning as background. 21 Lynnette Kelly, Executive Director of the 22 Municipal Securities Rulemaking Board. We spent the 23 morning praising her organization. Jason Lehman, Co-CEO 24 and Managing Member at Headlands Technologies, 25 representing a retail investor. 0179 1 Marshall Nicholson, President at Knight 2 BondPoint, an alternative trading system. Craig Noble, 3 Managing Director and Head of Retail Fixed Income at 4 Wells Fargo Advisors. Ben Thompson, CEO and Managing 5 Principal at Samson Capital Advisors. 6 Ben Watkins, Director of Bond Finance for the 7 State of Florida, and Chairman of the Government Finance 8 Officers Association's Debt Committee. Brad Winges, Head 9 of Fixed Income Sales, Trading and Underwriting for Piper 10 Jaffray, and President for Piper Jaffray Investment 11 Management. 12 Let me kick this off with topics, and I'm 13 sympathetic to one of the panelists this morning with the 14 observation that we had a million questions in our 15 materials, but just in general, let's start with pre- 16 trade transparency. 17 I think we heard a lot this morning, a few 18 little broad-brush observations, institutional investors 19 seem to do just fine and get all the information they 20 want, retail investors seem to need all the alternative 21 trading platforms, big institutional investors and big 22 underwriting firms at the same time may like to use the 23 various kinds of technology to reduce their need for 24 inventories. 25 We heard some sympathy for greater disclosure 0180 1 of bids wanted. We had a lot of discussion on sort of a 2 common consensus point, that more data and information in 3 this industry would be helpful. 4 Discussion prompted by Commissioner Gallagher 5 on building on information from CUSIPs and how we may 6 deploy that to better use. 7 To start this off, let's talk about whether or 8 not, just the general question of do people think that 9 making the bid and ask prices or bids wanted or other 10 trading interest information on electronic platforms 11 operated by ATSs or broker's brokers more broadly 12 available to the public will help the structure of the 13 market, and what kind of considerations should we take 14 into account there. 15 Just to begin, let me turn to Marshall 16 Nicholson and ask you to start that off maybe. 17 MR. NICHOLSON: For some reason, I knew you 18 were going to do that. Actually, over lunch we decided 19 that probably the easiest way to fix what ails the 20 municipal bond market is just to eliminate all income 21 taxes. Kidding, of course. 22 MR. CROSS: Maybe go in the opposite direction 23 there. 24 (Laughter.) 25 MR. NICHOLSON: There was some discussion that 0181 1 came out of the SEC's study on the municipal bond market 2 and the thought that ATSs could perhaps be a solution to 3 providing more transparency to the marketplace. We did a 4 little bit of analysis around that. 5 Actually last night over dinner, Tom Vales, who 6 was on the earlier panel, who is a friend and very good 7 competitor of ours, and I were talking about what would 8 that really mean for the marketplace, right. 9 I think everyone here has kind of thrown around 10 there is something like a little over a million 11 outstanding CUSIPs. The last time I looked, I think in 12 our security master database, we had 1.2 million CUSIPs. 13 To me, if you took all the ATSs that are involved in 14 posting some sort of municipal offerings, you would 15 probably come down to maybe something like 60,000 in a 16 given day, CUSIPs, that might be offered in the 17 marketplace. 18 Of course, unfortunately, those are going to 19 trade, some are not. I would kind of turn it back over 20 to everyone else and say you are roughly talking about 21 six or seven percent of the market that is outstanding 22 that may be offered on these ATSs on a given day if you 23 aggregated all the various ATSs. 24 To me, that is a step in the right direction. 25 It's more transparency than was there previously. The 0182 1 data is readily available. The technology to provide 2 that data into the marketplace is readily available. In 3 fact, we have actually worked on a couple of arrangements 4 with some of the financial portals. I won't name them, but 5 they are some of the more heavily trafficked financial 6 portals that retail investors typically look for for 7 stock quotes and research on a variety of items. 8 For one of those, we actually provide the bond 9 offerings off the BondPoint platform to that financial 10 portal, and they turn around and then they make that 11 available to retail investors. I'm sure some of you guys 12 might be aware of it. 13 I have been very surprised at how much activity 14 those financial portals get. One of them in fact has 15 over 20,000 searches a month that get run on that data 16 platform. Certainly, there is some value in that data. 17 It is somewhat limited, you know, as everyone says here, 18 to kind of the broker-dealer community, but in fact, 19 there are steps being made to make that available more 20 readily to the public via different vehicles than I think 21 normally most people would have considered. 22 MR. CROSS: So do you think this is a good 23 idea? 24 MR. WINGES: I would jump in there and say that 25 price transparency and information flow obviously is a 0183 1 good idea in any market, and to the extent we can 2 increase that, the better. 3 I think a lot of the consternation comes around 4 the pricing, whether you are talking about monthly 5 pricing or third party evaluator pricing, relative to, in 6 the example earlier today, somebody has a monthly statement 7 where they are pricing 50 bonds or 100 bonds on their 8 retail statement at 105, and they put the bonds out for 9 bid and they get a 98. 10 Another data point out there is roughly 25 to 11 30 percent of all the bid wanted's trade in the open 12 market, the other 70/75 percent that don't trade, there 13 is no transparency on what those high bids were. 14 To the extent that transparency was available 15 and the third party pricing services could use those 16 bids as a better indication of what the true market value 17 was and in this case, the evaluator would change the 18 evaluation on the monthly statement for the retail 19 accounts from an 105 to a 98. 20 It's more than just that one retail client that 21 had the 50 bonds out for the bid. It is all the other 22 people that own a like in kind security or a matching 23 security that wouldn't be privy to that information, 24 unless they went back to that same broker-dealer or got 25 the information from their registered rep, because it was 0184 1 their broker-dealer that put the bonds out for the bid. 2 From that perspective, I think that would 3 certainly be a large help to narrow the bandwidth, if you 4 will, between this 105 and 98 type number that got thrown 5 around today. 6 MR. THOMPSON: I agree, I think more data 7 certainly helps because it will fill in some of those 8 gaps that don't already exist in the real time trade 9 reporting. But the challenge is there is already a large 10 volume of data coming at investors, professional investors 11 and retail investors, and the difference between an 12 institutional trader and the individual investor is the 13 institutional trader can digest that information quickly 14 and take the pieces out that are relevant to pricing the 15 next trade. 16 I think it really comes down to investor 17 education, about how to use the information. It starts 18 with a background of what are the inputs to valuing a 19 municipal security relative to a risk free curve because 20 it has to start with a benchmark curve, a smooth 30 year 21 maturity curve, because that's how the market issues 22 bonds. 23 And then each security has a risk premium built 24 in, and this is what the trader at Citigroup does or the 25 trader at Piper Jaffray does when he or she bids a bond 0185 1 in the open market or offers one. 2 So the individual investor needs to understand in 3 relatively simplified terms how you would price a bond in 4 a certain region, with a certain maturity, with a certain 5 coupon, with a certain call feature. 6 EMMA has made incredible advancements in the 7 ability to distill the trade data down by those criteria. 8 So you can find trades that would be relevant. The only 9 weak link is the investor's understanding of how to actually 10 recognize why five trades that occurred within the last 11 five hours are relevant to the one bond that hasn't 12 traded in six months, but they are very similar. 13 So enriching the data is great, but I think 14 the challenge is getting the investor to invest the time 15 in understanding how to value the security and making 16 sure they are aware of the tools that actually give them 17 all that data that is for the most part in the market 18 already. 19 MR. AUWAERTER: I think from transparency, 20 putting bids and offers that aren't executed is generally 21 good, just be aware there is one risk to that, and you 22 see it in institutional markets, it's called "painting the 23 screen," where somebody will put up a bid or offer which 24 they know is not going to get traded out there, but they 25 are trying to set something up in the marketplace. 0186 1 That would be the only potential risk that if 2 somebody is trying to play games with a group of retail 3 customers and they might be using that way to put a price 4 up there, 98, when they know the real bid is par, but 5 they are the only one showing the bid up there when they 6 are showing a high offer when they know the offer should 7 be lower. 8 There is always a little bit of risk of that 9 going on when you do that because remember, the trade is 10 not executed. Anybody can put a number out there if they 11 are not held to it by the marketplace. 12 MR. CROSS: Just to follow up on that, in terms 13 of kind of the candidates for what information to be 14 encouraging to put up there, should there be limitations 15 or should we be looking at bid and offering prices or go 16 beyond that to request for quotes in auction type 17 transactions? What kind of limits or what should we aim 18 for in the particular information we are trying to 19 encourage? 20 MR. AUWAERTER: At the end of the day, the most 21 valuable information is where somebody is willing to put 22 their money where their mouth is, to buy or sell something. 23 As you move away from that, you do have the 24 risk I just referred to before. Hopefully, most people 25 are doing the right thing but there is some opportunity for 0187 1 somebody to try to game the system. 2 MR. LEHMAN: I would agree with many of the 3 points raised, but I think in general, the more 4 information that's available, the better off the investor 5 is in terms of being able to evaluate a larger set of 6 information. 7 I think there are a lot of brokerage firms out 8 there who do a very good job right now of providing 9 filtering tools to the individual investor, to sort 10 through inventory offerings in the marketplace, and to 11 refer to historical price data. 12 And just offering more avenues for these firms 13 to help the individual investor and giving them more 14 information, I don't see how that could be a bad thing. 15 MR. NICHOLSON: Just to add one item to that, 16 in addition to -- Mr. Cross, I think you mentioned 17 limitations. I think it also has to come with an 18 expectation. 19 Assuming there was some sort of movement where 20 you are going to post ATS prices on say EMMA or some other, 21 or ATSs themselves, are required to post them out to some 22 sort of publicly accessible website, what has to be kept in 23 mind is those are prices at which dealers can trade with 24 each other over these ATSs and does not include mark-ups, 25 mark downs, commissions, whatever might go along with that. 0188 1 What concerns me is creating an atmosphere that 2 the retail investor thinks these are prices that are 3 available to them. They are, but after commissions, 4 mark-ups, mark downs, et cetera, are added to them. 5 MS. STARR: I have a related question. To what 6 extent would the prices on the ATS or even from the third 7 party reflect issuer information that is either in or not 8 in the marketplace? 9 You were talking about the retail investor 10 being able to look at the actual prices coming off, but 11 do those prices reflect what may have happened with an 12 issuer, whether or not they have provided their financial 13 information or lack of financial information, et cetera? 14 MR. WINGES: I'll just jump in there. You tend 15 to find in the case of California issuers that we are 16 hitting the tape with bad news. Within an hour, a fair 17 amount of bid wanted's of that matching name come out 18 for the bid. 19 So relatively quickly, you start to see what the 20 clearing price is on a particular security if it has bad 21 news out there about it or good news for that matter, if 22 it's on upgrade watch or a host of some kind. 23 MR. NOBLE: But I think your point was more to 24 what information is available about the security along 25 with the price. If you go to EMMA, you will be able to 0189 1 see most of the current information available. 2 Like I said, we subscribe to another data 3 provider to scrub -- there has been talk about having 4 real time news, and if you just do a Google search on an 5 issuer, you might get a whole bunch of other news. We 6 subscribe to a service that actually narrows that down, 7 just the financial news and news on the individual 8 security. We provide that to our brokers. Our brokers 9 have that information at their desktop. 10 To your point, I don't know if every individual 11 investor would have it. If you are not using a full 12 service brokerage firm and you might be using a firm that 13 doesn't have all the information, you are right, that 14 client, unless they go to EMMA and do a CUSIP search, and 15 the disclosure information would be there or not be 16 there, if they didn't file a disclosure on it. 17 MS. STARR: But it is also a question of how much 18 of the information is actually taken into account in the 19 bids, and what role that plays in terms of the bid, and 20 when you have historical pricing, obviously, if something 21 happens in the interim, it would not have taken it into 22 account, so from a timing standpoint on the pre-trade 23 pricing information, I mean are there thoughts? 24 What are people's thoughts about how you get 25 closer in time to bids relating to when you want to 0190 1 trade? 2 MR. NOBLE: You make a very good point. The 3 problem with pre-trade information or even post-trade 4 information is you don't know the full market out there. 5 We had a discussion before this. The municipal 6 market flows with liquidity. As liquidity dries up, your 7 bids will suffer. In the case of let's say a couple of 8 California issues, as we get more bidders out there, more 9 demand for liquidity, and liquidity not being there, your 10 bids will go down but that would not show you in your pre- 11 price information, so yesterday's bid wanted information 12 may have nothing to do with today's bid wanted 13 information. 14 To your point, the client could either be 15 seeing good or bad news that makes no difference in the 16 price they actually got so it is having a client that 17 understands what is happening in the marketplace and 18 having the pricing information. 19 So it is having all the information available, 20 but it is tougher, I think, for a retail client to put 21 everything together. 22 MR. WINGES: That information flow can be 23 daunting as well. One thing that this morning's panel and 24 I haven't heard it talked about yet is the market is 25 moving. The data and the price data, pre-trade price 0191 1 transparency, may be from a trade that occurred 2 yesterday. The market is a point lower today or a point 3 higher. In most cases, it will be a trade that occurred 4 two or three days ago. The market has moved pretty 5 dramatically from there. 6 That dove tails into disclosing mark-ups. You 7 may be working for a point and a half on a trade and 8 selling it to a retail client, but you may have lost two 9 points on your hedge, and you are actually losing money on 10 the trade. 11 So the transparency right to the client looks like 12 you are actually making a lot of money on the trade when 13 actually you are losing it because the market moved. 14 Without that full spectrum of understanding, 15 which I think the institutional marketplace has, you can 16 talk to an institution in the morning and if the market 17 is rallying, they won't be shocked that you have marked 18 your bonds up. Vice versa, they are expecting that you 19 will discount it if in fact the market is selling off. 20 But if you are just going off pre-trade price 21 transparency from former trades in the matching CUSIP, 22 you are not going to have any of that reflection. That 23 is where you need to rely on the guidance of your 24 registered rep or your advisor to say here's the 25 spectrum, here's what I'm seeing out there. Here's the 0192 1 information flow. Here's what is on the tape, and here's 2 why we are higher than that or here's why we are lower 3 than that, and give them, I think, that perspective, that 4 probably an individual investor without some guide along 5 the way won't be able to process all that information. 6 MR. HEMPHILL: This morning I touched on, take 7 a number, if we have 100 bid wanted's on a day and 8 Regional Brokers had 27 percent trades, so again we have 73 9 bid wanted's that do not trade, 73 high bids that can be 10 used. 11 To Brad's point, if that information can be 12 provided on a pre-trade basis, moving forward to a retail 13 investor, they may be able to better assess their 14 holding. They put something out for the bid and we 15 discuss the expectations because of the third party 16 eval, but at least they have that information to see 17 that they may not have an MSRB trade report because 18 they may not have traded in six months but as I said this 19 morning, perhaps they've been out for the bid a dozen times 20 over the past six months, and all the bids are maybe 21 within two or three points of each other, because that's 22 the market. They just haven't traded. 23 The ATSs and the broker's brokers are sitting 24 on that information now, and it's not being used and if we 25 could aggregate that information, provide it to the MSRB, 0193 1 and so if I'm a private investor and I plug that CUSIP into 2 EMMA, not only can I see where the recent trade history 3 is, maybe there is no trade history, but look, it's been 4 out for bid 12 times over the past six months. 5 That gives me a better idea of what my bonds 6 are worth, and my broker will know that as well, and it's 7 a better conversation to have as opposed to the eval says 8 I should be getting six or seven points higher. 9 One other point as far as education, I think 10 the retail investor needs to understand the process by 11 which if they want to buy or sell their muni's, just 12 understand the process, that it goes from their broker to 13 their trading desk. They know that but they think it is 14 sort of black hole from there. 15 They need to understand that it goes to an ATS 16 or inter-dealer broker who then goes out and solicits bids 17 interest on their item and then reports back to the 18 liaison or the trading desk. 19 I think if they understood that entire process 20 and how it works, I think they would be perhaps a little 21 bit more comfortable in understanding their muni 22 holdings, especially when they are going to sell, and 23 they are disappointed in the net results. 24 I think it will help out the entire process 25 down the road. 0194 1 MS. KELLY: Sp what you are hearing is very 2 consistent with the MSRB's vision of what we call EMMA 3 2.0, the next iteration of the EMMA website, which really 4 culminates in what we call a "central transparency 5 platform," where you have all of the disclosures 6 available for each individual bond. 7 You have all the post-trade transparency, and 8 over time, you start adding incrementally different kinds 9 of pre-trade transparency, whether they are indices, 10 whether they are firm executable bids, whether they are 11 requests for quotes. 12 Obviously, through a very extensive comment 13 period, we want to really understand the costs and 14 benefits. We are asking dealers to reprogram their 15 systems, and this is very expensive. We really want to 16 make sure that the information that we are gathering is 17 really helpful to retail investors. 18 But I think the key is exactly what Ben said, which 19 is education. You really have to couple all of these new 20 enhancements with very significant education for retail 21 investors so they understand. 22 This central transparency platform, in our 23 view, would not be an execution site, but it would just 24 again provide one central place in the market for all of 25 that pre- trade and post-trade and disclosure 0195 1 information. 2 MR. RAMSAY: Lynnette, what is your vision 3 around the timing of doing that, assuming everything else 4 fell in place? 5 MS. KELLY: Yesterday. No, the first step is, and 6 we are actually having conversations with the Commission 7 about adding yield curves to EMMA, one or more yield 8 curves, different kinds of yield curves, again, with 9 corresponding educational information, not only for 10 retail investors, but remember, the EMMA website is also 11 an incredible tool for municipal issuers. 12 Under Dodd-Frank, we have an additional mandate 13 to protect state and local government entities and the EMMA 14 website has a whole other set of functionality for the 15 municipal issuer community. We want to make sure that 16 they are equally accessible and available and easy to 17 understand for the issuer community. 18 And then -- so again, incrementally over time, 19 we've just come out with a concept release. We'll do a 20 series of concept releases on how to build to that 21 central transparency platform. 22 The first one talked about trade reporting and 23 whether 15-minute trade reporting requirements is still 24 the right timing or if that should be accelerated. We 25 also asked questions about the architecture and whether 0196 1 or not the current sort of trade reporting through DTCC 2 still made sense or is a TRACE like reporting model 3 better? 4 So there's a lot of infrastructure and nuts 5 and bolts that has to be dealt with at the same time the 6 policy issues are being considered. 7 MR. CROSS: So one thing it sounds like is, if 8 bid and ask information is a good idea, the big caveat 9 with it is a lot of investor education as to what it 10 means and what it is and is not accomplishing. 11 How best to move in that direction? Do we do 12 that as regulators or does, you know, the market go off 13 on, you know, its sort of self-help advances in this 14 technology? You know, what's the best course there? 15 Should we be mandating something or -- 16 MR. LEHMAN: Well, education -- 17 MR. CROSS: -- will the market ever do it by 18 itself? 19 MR. LEHMAN: Sorry to cut you off. 20 But education really has to start with 21 transparency. And there's nothing that confuses an 22 investor more than opacity. And I think Joseph started 23 to get to that point which is it's very confusing for a 24 retail investor to try to sell a bond and he simply gets 25 a bid, one price back from his broker, and it says this 0197 1 is the price, take it or leave it. 2 And that level of opacity just creates 3 confusion about, well, how do you get to that price? And 4 it creates suspicion and mistrust. And what is going to 5 solve that is just providing -- lifting the covers and 6 providing transparency about every part of the process. 7 Providing the bids and offers, providing indications of 8 interest, and the education will come from that because 9 that will enable the retail investor to start asking all 10 the questions, and enable the brokers and the 11 institutions who are supporting the investors to provide 12 them with tools to properly evaluate all this data. 13 And so that's really where it has to start, is 14 with the full transparency. 15 MR. CROSS: Let me digress back a little bit 16 to one kind of negative concern about the idea of more 17 transparency on bid and offer type information. 18 And a point that was made this morning sort of 19 with the big institutional investors is kind of the 20 concern that posting information about big block trades, 21 you know, would have a negative effect. People would 22 just be resistant to that. 23 You know, is that a case of, you know, 24 defining big block trades which maybe don't want to use 25 the ATSs anyway or, you know, how should we deal with 0198 1 that issue? 2 MR. AUWAERTER: Well, right now trades up to 3 five million are disclosed right away, and any trades 4 above five million, say five-million plus. So if you're 5 concerned about transparency to the retail investor, 6 you're getting plenty of information within 15 minutes of 7 the trade, if I'm correct. 8 And then after a week's time, the total size 9 is listed. I mean, if the trade turned out to be 30 10 million, it doesn't take a detective from CSI to figure 11 out that it's probably two or three potential holders of 12 that security, because you just go on Bloomberg, look up 13 the CUSIP and you see who the holders are. So it does 14 make it difficult to move them out in the market, so it 15 will actually have the impact of reducing liquidity. 16 Just think about it, if you're a dealer and I 17 do a trade with you, and I -- and you're willing to take 18 down 30 million, which is a big trade in the muni market, 19 if everybody knows it right away, they're going to stack 20 up against you. If it's five million-plus, it could be 21 six, it could be seven, it could be more, but they don't 22 know. 23 MS. KELLY: And we've certainly looked at 24 that. We're actually looking at whether or not to have 25 total par unmasking and actually make the trade size 0199 1 immediately available, but we have an academic study 2 commissioned, and expect that some of the findings will 3 help us decide whether or not there will be a significant 4 impact on liquidity, and it's really the mutual fund 5 group that, you know, are very concerned about any par 6 unmasking. 7 MR. AUWAERTER: I think if you talk to traders 8 in the street, you don't have to be an academic to figure 9 that out. It's just common sense if it's disclosed right 10 away. 11 MR. WINGES: And I think you have -- 12 MR. AUWAERTER: And also, I will note that on 13 TRACE, there was a discussion by FINRA on increasing it 14 above five million, and they pulled back from that after 15 hearing from a lot of institutional investors, their 16 retirement systems, life insurance companies, asset 17 managers, mutual fund companies. 18 MR. WINGES: I think you have to start the 19 conversation with "what's the intent?" Are you just 20 adding rules for the sake of adding rules and more 21 transparency? 22 MR. CROSS: We're not just adding rules for 23 the sake of adding rules. Let me go on the record with 24 that. 25 MR. WINGES: Right. But what I mean by that 0200 1 is, I view the issue -- it's like the muni market and 2 liquidity is like a river, and you've got the issuer at 3 one end and you've got the retail investor at the other. 4 And oftentimes, adding rules and regulations are like 5 putting boulders in the river, and at some point, if you 6 put too many rules and regulations in there, the 7 liquidity flow is going to stop, and it becomes more 8 pronounced in volatile markets. 9 And from that standpoint, giving more 10 information flow around whether it was 10 million or 20 11 million or 30 million bonds, I really don't think in the 12 broader sense it's helping anyone. 13 And the large institutions, from what I 14 understand, and have heard and the traders aren't asking 15 for that, and I don't -- to the best of my knowledge, we 16 don't have a regulatory issue in the institutional 17 market. 18 There's concern, from what I understand from 19 the Commission, around the retail market, so I'd rather 20 see the Commission focus on those sets of rules versus 21 maybe adding other boulders in the river, if you will, to 22 use my analogy, that may not help the market and could 23 hurt it in terms of liquidity and a more volatile market. 24 MR. CROSS: I mean that actually is part of 25 the reason for the question, which is that most of the 0201 1 theme about more information for retail being disclosed 2 through ATSs and the like, or through EMMA, aims at 3 retail and aims at smaller transaction. And the biggest, 4 you know, concern that I've heard expressed about it is 5 the big block trades with institutions that don't 6 particularly want to use this anyway. 7 MR. NOBLE: Well, to that point, you've got to 8 remember that most of the large institutional traders 9 don't use the ATSs. They use either voice communication, 10 Bloomberg -- if you don't discount that as an ATS. So we 11 wouldn't really be tapping that market in the first 12 place. 13 So I mean, the question I have is really, what 14 are we trying to gather here? If we're trying to feed 15 information to EMMA, which is what I think we're trying 16 to do, then how are we going to do it? Are we going to 17 use it as a stand-alone central point of contact outside 18 the ATSs, or are the ATSs the means to get that 19 information to EMMA? And that's really the question I 20 have. 21 Because we've been talking a lot about ATSs in 22 this room, and there are quite a few of ATS participants. 23 And I'm just trying to get -- are we trying to drive -- I 24 use, like I say, every ATS. But there are a lot of firms 25 that only may have one connection or no connection to an 0202 1 ATS. And are we trying to drive everything to an ATS? 2 That's just a question I have for the Commission, because 3 I'm trying to figure out where we're going. 4 Because EMMA is -- would be my point of 5 contact. You want bid, offer or any other information, 6 I'm fine with sending my information to EMMA. 7 COMMISSIONER GALLAGHER: Somebody took my pink 8 card during the lunch break. 9 (Laughter.) 10 COMMISSIONER GALLAGHER: Burns, if I see you 11 wave a pink card, I'm going to -- 12 MR. CROSS: The whole organization comes to a 13 halt here. 14 COMMISSIONER GALLAGHER: Those things have 15 been marked. 16 Well, I had the same question sitting here 17 thinking, because I agree with Jason's point on 18 education, information is power, as they say. So how do 19 we do it? I mean is it order handling rules? I know 20 we're not -- we've been promising you we're not going to 21 lay equity markets over fixed income markets. 22 But let's -- just to use a common terminology, 23 is it order handling rules on fixed income side? Is it 24 souping up EMMA even further so that you have bid 25 information there? Is it through the ATSs? I mean what 0203 1 would work if that's the ultimate way to get education? 2 MR. LEHMAN: Well, Commissioner, I think the 3 most impactful thing that you can do for the retail 4 investor is to facilitate other options for them when 5 they need to sell a bond. Right now the only viable 6 option for a retail investor who needs to sell a bond is 7 a principal bid from a dealer. And often they're held 8 captive to the dealer, the brokerage firm in which they 9 hold the position. 10 And many brokers will go out and they will 11 accumulate bids from a number of dealers in an effort to 12 get them the best price, but not every broker does that. 13 And so some sort of a reasonable best execution standard 14 would be helpful there. 15 But as an alternative to that, the retail 16 investor should have the option and the ability to place 17 a resting limit order that's displayed somewhere in the 18 marketplace so they have an alternative to the principal 19 bid. A principal bid can be expensive, particularly for 20 an illiquid instrument and small size. And what we see 21 is the smaller the size of the transaction, the larger 22 the mark-up because it's a nuisance for a dealer to deal 23 with a five-lot transaction. 24 And so that's the perfect opportunity for an 25 investor who doesn't need to sell it immediately but 0204 1 maybe is willing to wait even hours or days to get an 2 execution done. They should be able to rest a limit on 3 an ATS or somewhere else where the offer can get 4 displayed to the marketplace and that the market cannot 5 trade through that level. And so that if somebody's 6 interested in that price, they -- anybody can go and 7 execute. 8 MR. NOBLE: And that's all well and good in 9 the equity market, but as we know by just in this room, 10 there are at least five ATSs in this room that deal with 11 individual firms around this room that -- so if I put a 12 resting order on Marshall's, it wouldn't hit Tom's or it 13 wouldn't hit Bonnie's. I can just go through the names 14 of the people in the room. You do have that problem. 15 And then you also have the technology issue of 16 how many resting orders do you want to have sitting in a 17 system? I don't disagree with you that it would make 18 sense, but at some point it's going to move down the 19 line. It's kind of like being on Craigslist, if you 20 don't put it on the first day, by the third day it's 21 already down to the 150th item and nobody's calling you. 22 And that's what's going to happen with that. 23 MR. LEHMAN: Yeah, but we went through this in 24 the options market, and the exact same problems existed 25 in the options market. We had fragmented liquidity, many 0205 1 Exchanges, they weren't interconnected. A limit order on 2 one Exchange had nothing to do with an order on another 3 Exchange. The Commission mandated a linkage system, 4 required them to go and talk to each other, simply talk 5 to each other. So if there's a resting limit on one 6 Exchange and somebody wants to trade on another Exchange 7 at the same price, or a trade through that price, they 8 have to respect the limit order on another platform. 9 And we heard all the same criticism at the 10 time, everyone said, you know, it's never going to work, 11 it's too messy, and it eventually got solved, and this 12 was ten plus years ago. It's -- technology's come a long 13 way from there. 14 MR. NICHOLSON: I think the technology certainly 15 exists from doing that from the operator of an ATS 16 perspective. What concerns me is I think everyone in here 17 agrees, and I think someone said earlier that pricing muni 18 bonds is more of art than sciences. You know, I like the 19 idea of empowering the retail investor, but what concerns 20 me is who determines the price of their limit order? 21 You know, so you have a limit order that's out 22 there, and who determines whether or not that's within 23 market context. And then you have another retail 24 customer that comes in and perhaps removes that limit 25 order. You know, ultimately the broker/dealer is going 0206 1 to be responsible for that execution that takes place by 2 their client. 3 I just -- you know, if we talk about central 4 limit order book trading, I certainly think it's applicable 5 in certain bond types, but I would say that any of them that 6 is not is certainly muni bonds. And you know, I could 7 see a whole host of issues arising from allowing retail 8 investors to directly post limit orders for municipal 9 bonds. 10 Not saying that it can't be done 11 technologically, but there needs to be, you know, that 12 follow-through on education, and people really have to 13 have an understanding of what they're dealing with to 14 make that determination. 15 MR. EADY: Just to clarify, Jason, is that 16 what you were suggesting, that the retail investor need 17 to have direct access to place their own limit order or 18 working through their broker who can advise them on 19 perhaps what a good price is? 20 MR. LEHMAN: Yeah, I think working through 21 their broker is fine. To clarify, I think having the 22 broker represent the retail interest as a principal, as 23 agent, simply I'm -- if Craig is my broker and I go to 24 him and I say I want to sell this bond, and I'm not happy 25 with the principal bid I get back, and I tell him, well, 0207 1 I want to rest this order at a price, and he can advise 2 me to whether that price is reasonable, what he thinks 3 about that. And then he can represent that order on my 4 behalf on a platform. I'm not actually putting it there, 5 he's doing it on my behalf. 6 MR. HEMPHILL: If I can add, I understand 7 ATSs do a very good job in the muni market, but I don't 8 think -- and this may be a little self interested, but 9 inter-dealer voice brokers who have -- 10 MR. CROSS: We actually haven't heard that 11 much about the broker's broker role. 12 MR. HEMPHILL: Okay. 13 MR. CROSS: So feel free to -- 14 MR. HEMPHILL: I will. 15 So Regional Brokers is an inter-dealer voice 16 broker, and we're a bit of a hybrid because we have a 17 second party firm that runs our website. And as I said 18 this morning, we get -- 70 percent of our bids come 19 through that website and it's very efficient. 20 But the interesting thing about -- I think 21 what gets lost in the conversation is that, you know, 22 again, the ATSs do a very good job, and most of the firms 23 that I feel that get very good execution for their customers 24 put their bid wanteds out on an ATS and give it to a firm 25 such as Regional Brokers, and we advertise them on our 0208 1 website and we make phone calls. 2 And putting up a ten bond line on an ATS 3 doesn't exactly guarantee you best execution because 4 there could be a CUSIP where I know there's three bidders 5 in the country on that CUSIP, and if two of those three 6 individuals are out that day -- and this sounds 7 borderline ridiculous -- or in meetings, that bid's not 8 going to be there. 9 Now I can go back to the trader or their 10 liaison and relay that information to them, communicate 11 that, that I've got three bidders and I've got one bid 12 back. He's two points cheaper than he was yesterday when 13 we had him out for the bid, and the other two bidders are 14 not in today. So you may not get the bid that you saw 15 yesterday. 16 Where, if it's on the ATS, there's no 17 communication and you get no feedback, and I think the 18 communication between the trader and the secondary 19 market, the inter-dealer broker, is essential. I just 20 didn't want that to get lost in this conversation. 21 MR. CROSS: So since you bring that up, why 22 don't we switch gears slightly to that topic. Should 23 there be a best execution rule for the municipal 24 securities market that would require brokers to seek to 25 obtain the most favorable terms available to customers? 0209 1 MS. KELLY: I think I have to start on that 2 one. 3 We don't have a rule that is entitled "best 4 execution." Clearly we have pricing rules, G-18, 5 G-30, that require fair and reasonable pricing. We 6 have Rule G-17, you must be fair to your customers. So 7 we have what we believe is the equivalent of the 8 reasonable diligence standard that's in the FINRA rule. 9 Where we differ from FINRA is that we don't 10 have the -- we don't have the -- you know, the back end, 11 where you have to review on a quarterly basis the best 12 execution, the facilities, the -- not on a trade-by-trade 13 basis, but just the facilities, et cetera. And that's 14 tough to do in a thinly traded market. 15 We are and have been studying a best execution 16 rule for the municipal market. We're starting to engage 17 the Board in those conversations, but FINRA's clearly the 18 most appropriate model for our market, but it is not -- 19 it is not as simple as it might sound. 20 MR. CROSS: So what is it about the FINRA 21 model that presents the most challenges for muni's? 22 MS. KELLY: So there's -- you know, you've got 23 to use reasonable diligence to ascertain the best market, 24 number one. And then you have to use -- you have to 25 periodically conduct -- I've got the language, 0210 1 periodically conduct a regular and rigorous review at 2 least quarterly, you know, retroactively to make sure 3 that, you know, your policies and procedures and your 4 choices of execution venues are, in fact, you know, the 5 best. Or you know, that they are satisfying the best 6 execution. I think FINRA just had an enforcement action 7 announced today on the best execution case. 8 So the first part is fine, that's not a 9 problem at all. But it is that, you know, how do you 10 know, in again a thinly traded municipal market, how can 11 you go back retroactively to ascertain that? And again, 12 it's just a challenge. It's something we've been 13 studying and looking at. 14 MR. RAMSAY: I was going to say, it strikes 15 me, Lynnette -- I fully appreciate the fact that the 16 extent and nature of the obligations may be different, we 17 might not be able to just graft the FINRA rule on top of 18 it. It occurs to me if you -- unless you have some means 19 to -- or requirement to survey alternative methods of 20 execution or results from doing that, then how do you 21 know if the obligations are ever violated? You know, 22 what does it mean at the end of the day? It seems like 23 the tail piece is, you know, the key part. 24 MS. KELLY: Right. So maybe the first option 25 is, you know, requiring brokers or on a voluntary basis 0211 1 to talk about their execution venues, to make those 2 public. And then -- or requiring certain execution 3 venues. So there's different steps that can be put into 4 place. 5 MR. CROSS: I mean, that's -- you allude to 6 sort of a combination of better education of the 7 available ways to trade securities and maybe considering 8 requiring people to make that information known. 9 MS. KELLY: Very consistent with the SEC 10 report and the recommendations therein. 11 MR. LEHMAN: I think the biggest issue that 12 would give rise to best execution problems in this -- in 13 a retail space, again, is potential conflicts of 14 interest. And this would revolve around a broker 15 transacting with their own dealing desk, and I think the 16 vast majority of brokers do practice reasonable best 17 execution, but there are certainly some cases where a 18 broker is going to preference or to really look to siphon 19 retail flow to its dealing desk, and that's the piece of 20 the -- the conflict that we want to highlight here. 21 And a reasonable standard should require the 22 broker to go out and seek arm's length interest from 23 other parties that are unrelated. 24 MR. HEMPHILL: Well, one suggestion is if you 25 put it out, you can have the desks bid. But what if you 0212 1 also had to put it out on an ATS or an inter-dealer broker, 2 so you would have the competitive bid with the desk, and 3 if the Street was the better bid, then that would be the 4 -- you know, that's your price. But at least you're 5 seeing what the trading desk bid was, and then what the 6 interest is in the secondary market as well. 7 And if it's combined with the two, then you 8 could prevent the broker just going to his trading desk 9 and getting the number and instead of it going back to the 10 customer. This way it's going out for competitive bid, 11 and at least you know where the secondary market is at 12 that time compared to where the desk bid is. 13 MR. WINGES: I think it would also help if -- 14 excuse me -- if the trading desk of the firm which has 15 the bond out for the bid bids blind. In a competing bid 16 with another ATS what tends to happen is the bid comes in 17 from the ATS, and then the desk decides whether or not 18 they want them at a slightly higher price. And then the 19 trade execution occurs through that desk and not out on 20 the ATS. 21 So what ends up happening on the ATSs is, 22 you have these ridiculously low trading volumes on the 23 number of bid wanteds that are being executed when, in 24 fact, a lot of those bonds are trading through the desks 25 that put them out for the bid at just a slightly higher 0213 1 price. Where if they were required to bid blind, it 2 would help the client on both sides, the seller. 3 But in addition to that what it would raise is 4 the amount of bonds that would trade electronically over 5 the ATSs because the other traders wouldn't feel like 6 they would get plussed at the last minute by a trading 7 desk who put it out for the bid. 8 So your confidence in your ability to execute 9 and actually spend the time, you know, I would know -- I 10 know I would rather have our traders spending time where 11 there's 50 percent execution or 60 percent execution than 12 on ATSs where there's 25 percent. 13 And so looking at nuances like that that would 14 help improve the liquidity in the market, I think, and 15 those are the types of rules that can help guide that 16 liquidity a little bit better. That will cause the 17 marketplace to behave more in an open free market 18 capability. 19 And back to a comment earlier about the 20 options market and you referenced the equity market -- 21 and I appreciate you saying we're not trying to become 22 the equity market, I think everybody here appreciates 23 that, but with that regard, the big major component that 24 a lot of people continue to miss is municipal bonds are 25 the only thing you can't short. 0214 1 So your ability to hedge it or your ability to 2 give an offering on a bond where you have a demand is 3 non-existent until you already own the bond. And that is 4 a major difference for the muni market over all the other 5 asset classes. 6 And you can't short the muni because you can't 7 short -- or you can't create phantom tax-free interests. 8 And again, a lot of people don't understand that. So it 9 makes it extremely difficult to hedge your position or 10 create a two-way liquid market. 11 MR. CROSS: So take just a second and compare 12 that with the way people hedge in the corporate market. 13 MR. WINGES: So corporate's you could -- let's 14 say you're long a ten-year corporate, Ford Motor Credit. 15 You can be long one coupon and short another coupon, so 16 you've hedged out all your credit risk on Ford. You 17 maybe have slight interest rate exposure differential 18 with your coupon interest, depending on what maturity 19 you're short, but you've taken out all your credit risk 20 because it's Ford for Ford. 21 You could choose to short Treasuries against 22 the Ford but then you've taken out your interest rate 23 exposure, but you haven't taken out your credit exposure 24 of Ford. Now take a ten-year muni, what do you short? 25 Well, you short Treasuries, but muni's and Treasuries, 0215 1 the correlation is inconsistent. So you haven't taken 2 out most of your interest rate risk because you still 3 have embedded interest rate risk by being short 4 treasuries. 5 And you have all the basis risk of the muni 6 exposure relative to tax rhetoric or different things 7 that are going on with the muni credit. And that will 8 tail around on you and your Treasury may not be moving. 9 Or there might be a flight to quality because of some 10 news event and the Treasury will run against you and the 11 muni bond won't have changed. 12 So it's extremely difficult to hedge out not 13 only your interest rate exposure but also your credit 14 exposure in the muni market, because you can't do like 15 you do in the corporate market where you short Ford for 16 Ford or Ford for treasuries and have a strong 17 correlation. 18 So as you build a large portfolio of thousands 19 of different CUSIPs, whether you're a mutual fund trying 20 to hedge out some of their interest rate exposure, or 21 you're a trading desk like at Piper Jaffray, we spend 22 more time trying to hedge out our interest rate and our 23 basis risk, or our credit exposure of municipal debt, than 24 all of our other desks combined, because hedging interest 25 rate and basis risk on those other desks is relatively 0216 1 easy. 2 It is by far the hardest thing to do, is hedge 3 out your exposure. And what ends up happening is you 4 have far less inventories being carried by the Street in 5 municipal securities because you've got all the nuances 6 of the individual credits, you've got the nuances of 7 hedging and your exposure to -- again, tax changes are 8 pretty dramatic. 9 No different than you'd have exposure of a news 10 announcement on a Ford credit, well, you can imagine, the 11 muni market is one of the only markets where they can 12 change some tax policy or rhetoric in Congress and affect 13 your entire inventory and portfolio in one day. 14 And this latest proposal of 28 percent, we saw 15 that. Or Meredith Whitney, look at what we saw happen 16 there. It was a massive amount of money lost because 17 people couldn't hedge out that risk. And all that really 18 happened was that the basis or the credit value of 19 municipal bonds went down. It wasn't that interest rates 20 moved. 21 MR. LEHMAN: So I just want to add a quick 22 comment to that. I agree with everything that was just 23 said. It's -- it is very hard to hedge municipal risk, 24 therefore it's very hard for dealers to make competitive 25 prices. And so principal bids are expensive for 0217 1 customers to exit positions. 2 And so what we -- the best way to again to help 3 the retail investor is to facilitate movement to an 4 agency solution for selling a bond, which means going out 5 there and trying to find the other side of the trade 6 rather than having a dealer sitting in the middle of 7 every transaction for small size. 8 And that's how we deal with the problem of the 9 risk because he's right. The dealer who has to make a 10 price for this very illiquid instrument he can't hedge, 11 he has to take a spread for that, and that's expensive. 12 And it's expensive to the end investor. 13 MR. CROSS: Great point, yeah. 14 MR. THOMPSON: I agree you have to give them 15 more education about their options in selling. I think 16 taking away the intermediary -- I mean, I'm on the 17 institutional buy side, so I benefit from, you know, best 18 execution price transparency. But I also appreciate the 19 difficulty in managing the volume of information. 20 I think when you ask Marshall or you ask you 21 ask Tom about the number of items they have in their 22 systems at any one point in time, and the age of some of 23 those items, it's very obvious that if you had a central 24 repository of bonds being sold by individual investors 25 directly without a financial intermediary, it would be 0218 1 the Craigslist problem times a hundred. 2 So you need a financial intermediary as the 3 advocate for that client. But you can't dictate 4 necessarily the mechanism they'll use because the market 5 is not efficient enough to let one mechanism be the de 6 facto best solution. So if we try to dictate which 7 mechanism is the best way for a client to move away from 8 their dealer, you could end up with lowest common 9 denominator, something that actually passes tests but 10 actually isn't the optimal execution for individual 11 trades. 12 So if someone's selling a bond in California, 13 those set of brokers, they should be contacting 14 California. If it's New York, it's someone in New York. 15 If it's the Southeast, it's different. So there are 16 vagaries to the way transactions occur, particularly in 17 the sell side, that are very situation-specific. 18 So the financial intermediary is the one who 19 has to be enlisted as the advocate for this. The 20 situation where they're pulling transactions towards 21 their own trading desk, I mean there are lines where 22 there's, you know, fraudulent activity or there's -- they 23 are essentially, you know, conducting business in a way 24 that violates the rules. 25 You have to basically create a rule set that 0219 1 requires them to disclose the options that are available, 2 have the conversations with the client about the 3 mechanisms that they can utilize. I think a client 4 can already turn down a bid if they're presented by a bid 5 from a dealer. They can say, no, I want to sell it at a 6 higher price. Craig's clients can all say that. They may 7 not be fully informed that they can say that, so they 8 have to be informed that they can turn down a bid. 9 And they should also -- there should also be a 10 requirement that they're informed about other alternative 11 mechanisms. How tightly you define that becomes kind of, 12 I think, dangerous territory because you could end up 13 mandating something that is sub-optimal. 14 MR. EADY: Again, just to clarify, Jason, were 15 you suggesting that the agency model completely replace 16 the personal bid? 17 MR. LEHMAN: No. Again, all of this would be 18 done through a financial intermediary. It's just simply 19 requiring that the financial intermediary offer the 20 option of representing a limit order in the marketplace 21 somewhere that gets displayed to the marketplace. That 22 can be an ATS, that can be anything that displays that 23 order to a marketplace where that limit order is 24 respected in some way. 25 But again, we're not -- I'm not suggesting in 0220 1 any way that you remove the financial intermediary. 2 MR. CROSS: Why don't we move on and talk a 3 little bit about the primary offering process, and 4 whether or not improvements in price transparency in the 5 secondary market, their impact and how they could enhance 6 the primary offering process for municipal securities, 7 and turn to Ben Watkins from the State of Florida who has 8 remarkably been quiet for almost the entire panel. 9 MR. WATKINS: John knows me well, it's hard 10 for me to sit still and not open my mouth when I have an 11 opinion, because I will. 12 But so as -- you know, before we move on to 13 the primary market offering process, just an observation. 14 So when I hear the discussion relative to the secondary 15 market execution and the different ways that bonds can be 16 traded in the secondary market, it's not something that 17 is a direct interest to the issuer that we follow on a 18 day-to-day basis. but it's extremely important in the 19 sense that it provides liquidity to the marketplace for 20 an extraordinarily diverse group of issuers who can 21 access credit across a broad array of different credits 22 and the market operates fairly efficiently in that 23 regard. 24 But having options to execute, from my 25 standpoint, makes sense. As a general proposition, the 0221 1 notion of providing more information to the marketplace 2 as a theoretical construct is something that I think 3 everybody can agree on and is generally supportive of. 4 But the way that rules are designed or mandates are 5 promulgated can significantly affect and could 6 potentially impair the liquidity that we enjoy in the 7 market today. 8 And so word of caution, so when I hear my 9 largest bond holder say that, you know, the $5 million 10 filter is an important thing for him to execute trades, 11 if he can't execute trades efficiently and effectively 12 for his purpose, I ultimately am going to bear that cost 13 of that. 14 There's no question about that in my mind, it 15 will come back to the issuer community in the form of 16 higher interest rates if liquidity is impaired 17 inadvertently through mandates or rules promulgated 18 relative to that. 19 So secondary market liquidity is important, I 20 guess is what I'm trying to say. It's an important 21 consideration. It's hard to promulgate a uniform rule 22 for a bifurcated market. Because there's no doubt that 23 the institutional market is capable and aware and the 24 retail market is in a different mindset entirely. 25 You need to be careful about the kind of 0222 1 information that's provided and the value of that 2 information. Is that information going to be useful 3 information to a retail investor, that they can 4 meaningfully act on and substantively affect the outcome 5 of their trade? 6 So that's my cautionary note. 7 Relative to secondary market trading in 8 forming primary market levels, so there's two different 9 ways an issuer can use the secondary market trading 10 information. One is to assist in pricing bonds in a 11 primary market offering, something akin to pre-trade 12 transparency. What is the right price for -- what is the 13 right level in terms of interest rate in a primary market 14 offering? 15 But also secondarily is a post-sale evaluation 16 of how those bonds were priced. Through the secondary 17 market trading information, you can tell how close you 18 came to the mark. So when they're free to trade, if they 19 trade up or down and by how much, you can judge how well 20 those bonds were placed in the primary market offering. 21 MR. CROSS: Ben, I'm just curious. On your -- 22 on the first part of your point, you know, it's kind of 23 the, I think, see - change that EMMA introduced in the Spring 24 of 2008 when they first started putting all this 25 information out there on prices. 0223 1 I'm just curious, as a big issuer for a state, 2 how has that affected -- have you all used that 3 information to your benefit in pricing bonds? Or how has 4 that affected your all's views of the market? 5 MR. WATKINS: Okay. So we do use the 6 information and it is a -- it is generally a helpful 7 tool, but only in a very general way. And so we're using 8 the secondary market trading information to try to 9 discern levels, but the precision with which that 10 information can be used is limited. 11 So there -- we view it as one additional 12 dataset. And any dataset is welcome, so we're looking at 13 benchmark scales, we're looking at comparable deals that 14 are executed in the primary market, in and around the 15 time of our offering. And we're having conversations 16 with dealers relative to spread to an index and where we 17 might trade relative to that. 18 MR. CROSS: To follow up on that slightly and 19 on something that Lynnette mentioned earlier, do you 20 think it would be constructive to have more information 21 about basically yield curves and different kinds of 22 benchmarks made more available and explained more through 23 EMMA and the MSRB? Or is that something that you just -- 24 I mean, you all are sophisticated issuers, so maybe you 25 already know all that, but do you think that would be a 0224 1 helpful thing? 2 MR. WATKINS: I think it's generally a tool 3 that's used in the marketplace currently. It is helpful 4 in a general way, but I would have to say that the 5 indexes can be misleading. In other words, if the 6 indexes were absolutely accurate in terms of a reflection 7 of the market, then your spread should stay relatively 8 constant over time for a given security. That's not the 9 case. 10 The spreads move around based on market 11 conditions, liquidity in the market and sometimes the 12 indexes themselves may not be an entirely accurate 13 reflection of the market going forward. It's more of a 14 look back than it is a look forward. But I think it's 15 generally helpful, but again, as a general guide and a 16 useful tool, and sort of not the be-all end-all. 17 MR. CROSS: Commissioner Gallagher, did you 18 find your pink card? 19 COMMISSIONER GALLAGHER: Someone got me a 20 replacement, so I actually did not have a question, I was 21 just illustrating that I got it back, in case I do have a 22 question. 23 (Laughter.) 24 MR. CROSS: Well, I'll tell you what, I'll ask 25 your question for you then, which is to follow up on -- 0225 1 there's only a few minutes left in this session, and I'd 2 like to follow up on a point that is a little bit less 3 price transparency and a little bit more of this huge 4 coming issue relating to how to address and manage the 5 ultimate unwind of the Fed's zero interest rates for as 6 far as the eye can see policy. And you know, how are you 7 all addressing that? How do you think regulators should 8 address that? 9 COMMISSIONER GALLAGHER: John, can I add to 10 that without my pink card? 11 Factor into it what Jason told us, what we 12 heard earlier about exiting, right? If the big problem 13 is the sell on the retail side, I mean, it's going to go 14 hand in hand with the problem that John poses. So I'll 15 be interested to hear. 16 MR. WINGES: I think I'll let Craig speak as 17 it relates to retail specific. But I think -- and I hate 18 to keep repeating information as key, but one of the -- 19 probably the least understood things in the muni market 20 as it relates to coupons and valuation is de minimis, and 21 that is going to be a word that people are going to learn 22 very, very quickly when we're a couple hundred basis 23 points higher in the muni market. 24 Prices are going to logarithmically start 25 dropping as de minimis starts coming into play, and 0226 1 there's been some recent proposals by the Administration 2 about how to treat de minimis. We have to very careful 3 on how that's being treated, how it's being reported, how 4 it affects prices. 5 And I would encourage you to pay close 6 attention to that. Because for those of you that don't 7 understand de minimis very well, it essentially takes 8 into account the discount on the bond and treats it from 9 a tax perspective differently, depending on how many 10 years you are from maturity and how many years you are 11 from the original issued discount price. 12 And as a lot of debt, to your point or question, 13 the last few years has been issued with extremely low 14 coupons. That will come into play. And from what we've 15 seen, a fair amount of pricing services still struggle 16 with that, a fair amount of ATSs still struggle with 17 that. 18 And that's going to probably be one of the 19 bigger points that you're going to see a wide disparity 20 in evaluations on customers' portfolios where they may 21 have a portfolio evaluation of, let's say 90, and they 22 put the bond out for the bid and they can't figure out 23 why they got 80. And they got 80 because after taking 24 into account de minimis, they really are getting about 25 88.5. 0227 1 But that education of explaining that not only 2 to the individual investors but actually to the industry 3 in general, because that has not come into play 4 significantly because we've historically been in a bull 5 market for a very, very long time. 6 MR. NOBLE: Well, I mentioned it before and I 7 think it goes back to investor education, it goes back to 8 the FINRA, you know, investor alert on duration. It's 9 making sure that our clients understand that what has 10 gone up for the last 20 years can go down. I mean that's 11 the biggest problem. 12 I mean I started in the market when you could 13 buy a 16.75 Treasury non-callable for 30 years. I doubt 14 many clients of ours held that to maturity, if they did, 15 they did very well. 16 But it's a point of concern of do the clients 17 know that a lot of their bonds that are trading at 110 18 premiums are in fact worth par. And if you don't sell 19 them now, they will mature at par which, if you're okay 20 with, and they may trade at 80, but they will mature at 21 par. 22 As long as the clients understand that and are 23 willing to take a below at the time investment rate of 24 return, our clients, most, 99.9 percent of our clients 25 will get out of the bonds at par, which is a tough 0228 1 lesson, because they haven't seen that for the last 20 2 years. 3 And I think it's admirable that FINRA and the 4 SEC are very concerned about it. We are concerned as a 5 firm about it. We have actually just answered a meeting 6 invite that we are sitting down and having a special 7 meeting with our analysts to talk about how we're going 8 to put together a report and actually have a confirm 9 stuffer on duration, and I think it's just making sure 10 all our customers are educated on what can happen. 11 And I don't know if what, away from the FINRA 12 investor alert, whatever we can do about it, because it's 13 not just muni's, it's going to be every product. 14 COMMISSIONER GALLAGHER: As one of your 15 clients, I'm pretty happy to hear that, Craig. 16 MR. NOBLE: The problem is I seem to have more 17 clients that are either here or the MSRB or someplace 18 else. I am very worried about my confirm's and 19 everything else that are coming. 20 COMMISSIONER GALLAGHER: I don't read 21 anything. That's the problem. 22 MR. NOBLE: Oh, good. 23 COMMISSIONER GALLAGHER: Yeah. I mean, I am 24 the average investor. The staff does all the testing on 25 me. Yeah. 0229 1 MR. NOBLE: Good. 2 (Laughter.) 3 MR. RAMSAY: We'll have that stricken from the 4 transcript. 5 (Laughter.) 6 MR. CROSS: Tom, I know you had a question to 7 follow up on something we skipped over, but wanted to 8 make sure we touched. 9 MR. EADY: Yeah, we'll circle back there while 10 we still have this esteemed panel in front of us to 11 provide some guidance on this. 12 The municipal securities market report that 13 we've referred to several times today that was published 14 last year, one of the recommendations in the market 15 structure section was that the MSRB should consider 16 requiring the disclosure of mark-ups on risk or mark 17 downs on riskless principal trades to their customers, 18 and I would like to hear from the panelists what their 19 views are on that recommendation. 20 MR. NICHOLSON: Well, the data is there, 21 right? I think Craig mentioned earlier that with regards 22 to EMMA, you know, the issue is that a lot of retail 23 investors just don't know of its existence or what 24 information may be contained in there. 25 But with regards to the trades that are 0230 1 reported, you know, you could certainly -- any retail 2 investor could go in and see that there was an inter- 3 dealer trade and then there was a buy from a customer or 4 a sell from a customer at a different price and back 5 into, you know, what the markup or markdown is. 6 So short of actually disclosing on a confirm, 7 right, which is what you could do, you know, it goes back 8 again to education. If that seems to be an issue that 9 retail investors are concerned with that, you know, in fact 10 the data is there. It can be used and, if someone is 11 curious of what is the compensation that was paid to 12 their broker-dealer, they could certainly do so. 13 MR. NOBLE: And we have customers that do 14 that. We have brokers that do that. We have brokers and 15 customers that go to EMMA, look at the price of where I 16 paid a bond if I bought it at 99 and for some reason it's 17 being sold at par, they know exactly what part went to 18 the broker, what part went to the firm. Everything's 19 available. 20 I don't know if it makes it any clearer, I 21 don't know what we're trying to accomplish by disclosing 22 the markup. Are we trying to limit mark-ups or just disclose 23 mark-ups? Because that would be my question back to you. 24 Because you can easily, especially on your 25 riskless principal trades since they're happening the 0231 1 same day. If I buy a bond from Knight BondPoint at 99 and 2 I sell it to Jason at par, somebody's going to know what 3 I made, and that information is readily available, real 4 time. 5 MR. THOMPSON: I think if you just be very 6 careful how you define it. I think a riskless principal 7 trade is one where there's an order pre-trade. If it 8 takes place -- if the order is given after the trade's 9 been consummated and someone is at principal risk, if you 10 say five minutes or an hour or three hours is riskless, I 11 think that's a difficult decision -- that's a difficult 12 argument to make. 13 But if there's a pre-trade order given, it's 14 truly riskless. So I think it's more -- it's valuable 15 information and how you present it is up for discussion. 16 You have to be very careful how you define it. 17 MR. CROSS: So I think we're going to close 18 out on this panel. But before doing so, I just wanted to 19 thank everyone for the very helpful perspective and 20 emphasize that, you know, I think the really constructive 21 SEC report on the market structure ideas from last Summer 22 involves an area in which we want to vet the ideas. 23 We want -- you know, in order to build 24 incrementally on those for the benefit of the markets and 25 liquidity, it will require, you know, a combination of 0232 1 market efforts and certainly consensus among market 2 participants, and as well as, you know, the role of the 3 SEC. 4 So thank you all very much, and I'll now turn 5 it over to Jim Burns for the final corporate panel. 6 (Recess.) 7 MR. BURNS: If we could try to get started, 8 that would be great. 9 Okay. I think we will try -- I know a few 10 people are out there calling their brokers trying to sell 11 their muni portfolios, but we better get going. 12 And that was for purposes of waking people up, 13 I hope. Actually this has been a terrific Roundtable, 14 and I think this last panel will be just as exciting. 15 We're very pleased you could join us. 16 We're turning now -- panel four is going to 17 discuss whether there may be potential steps to improve 18 transparency, liquidity, efficiency or other structural 19 aspects of the corporate bond and asset-backed securities 20 markets. 21 PANEL 4: POTENTIAL IMPROVEMENTS TO THE MARKET 22 STRUCTURE FOR CORPORATE BONDS AND ASSET-BACKED SECURITIES 23 MR. BURNS: To begin, I'd like to introduce our 24 panelists. Many of them are returning from our second 25 panel, but for those who haven't been here before, let me 0233 1 just start first with Nancy Mueller-Handal, who has come 2 to us from MetLife where she is head of Structured 3 Finance. Collin Heffron, who's the CEO of GFI Group. 4 Steve Genyk from Janney Montgomery. 5 A newcomer to the panel but known to many of 6 you, Richard Ketchum from FINRA. Richard McVey who is 7 with us -- forgive me, Richard, it's been one of those 8 days -- from MarketAxess. Kevin Molloy who has come to 9 us from the New York Stock Exchange. Neil Schloss who 10 comes to us from Ford Motor Company. 11 Dexter Senft who comes to us from Morgan 12 Stanley, and I want to talk with you afterwards about 13 crossword puzzles. For those of you who haven't read the 14 bio's, he constructs them, so he will, I hope, be able to 15 help us puzzle through some of the things here. Sorry, I 16 couldn't resist. 17 Robert Smith, who was speaking with us 18 earlier, who is CIO and President at Sage Advisory 19 Services. Kumar Venkataraman, who is a Professor at SMU 20 who has joined us, our academic here. 21 For your benefit, we had a wonderful 22 presentation from Michael Goldstein earlier in the day. 23 We're going to ask you to be more of a repeteur than to 24 give a presentation in the beginning. 25 Finally, Christopher Vogel who comes to us 0234 1 from BlackRock. 2 I thought where we might start this afternoon 3 is returning to the subject of transparency. 4 Commissioner Gallagher, who left just a moment ago, sort 5 of helped us kick things off there asking about whether 6 and how things have progressed or regressed in the area 7 of transparency. We had a very lively discussion in the 8 first panel ranging from remarks Richard McVey had about 9 certain concerns about whether pre-trade transparency 10 could have impacts on pricing. Robert Smith spoke a bit 11 about trying to force further price discovery and finding 12 ways to help with that. 13 We heard about block trading, how much 14 activity 50 percent of trading seems to be done in 15 blocks. Kevin Molloy, of course, gave us a sense of 16 what's been going on in at least one -- in an Exchange 17 venue to try to foster Exchange-based trading. 18 And so with that, I thought what we might do 19 is launch off and try to bring into the conversation 20 people like Chris, Richard and Dexter who haven't been 21 involved yet, on whether the panelists believe that 22 making trading interests displayed on electronic 23 platforms operated by ATSs or inter-dealer brokers 24 broadly available to the public could enhance pre-trade 25 transparency and improve the liquidity and efficiency of 0235 1 the corporate bond and asset-backed securities markets. 2 So maybe starting at the end of the row with 3 Chris. 4 MR. VOGEL: Good afternoon everyone, and 5 thanks for inviting myself and BlackRock to the panel. 6 With regards to transparency, I think we at 7 BlackRock feel there is ample transparency out there in 8 investment grade corporate bond trading between the 9 trace prints, the closing services, the information 10 aggregation services, to bring all your runs in together 11 at one place, some of the modeling's. 12 You can certainly get an academic price of a 13 security. But really what's most important for us is 14 where can we transact. So you can spend as much time as 15 you want looking at a bond thinking where it should 16 trade, what its value is. but what's most important is 17 where will it trade, where are the bids, where are the 18 offers? 19 And for us at BlackRock, we like to have a 20 portfolio of mediums that trade. We rely on the dealers 21 for our RFQ. We're a huge user of the ECN, we trade 22 north of 60 percent in block count on the machine. We 23 believe in list trading, we believe in displaying axes 24 and we certainly believe in different types of pools, 25 both dark and dimly lit. 0236 1 MR. BURNS: I'm looking down the panel just to 2 see -- Dexter, I don't know whether you have anything to 3 add? 4 MR. SENFT: Sure. The point was made in two 5 earlier panels, once by Bob Auwaerter and then by Richard 6 McVey. I think actually Bob might have made it twice -- 7 that liquidity and transparency are different things. 8 Over long periods of time they may move up together, but 9 in the short run they can be inversely related. This is 10 particularly true for a large trade size for the reasons 11 that those two outlined. 12 It could also be true in smaller size for some 13 sort of weird reasons in the category of unintended 14 consequences. So I guess the way we think about 15 liquidity is, you know, it's -- well, first of all, 16 because they can move inversely, it's important to 17 understand which one is your objective, because you can't 18 maximize both in the short run. 19 Our perspective at Morgan Stanley is that 20 we're trying to get the best execution for our customers 21 and therefore will side with better liquidity every time, 22 and if that happens at the cost of reduced transparency, 23 then so be it. We're always in favor of liquidity. 24 Liquidity, if you think of it as the ability 25 to get the best price for your transaction in a short 0237 1 period of time, then as you work through sort of 2 practically what that means, it can be different for 3 small trade sizes and large trade sizes, and therefore 4 the impact of greater transparency can be different for 5 those, too. 6 So rather than rehash some of the arguments 7 that were made in the earlier panels, I'll just let it go 8 at that. But I do think it's dangerous to ask the 9 question, you know, what can we as regulators do to 10 improve liquidity and transparency, because often times 11 you just are going to be hard pressed to do both at the 12 same time. 13 MR. VENKATARAMAN: I would like to present a 14 slightly different perspective on the benefits of 15 transparency, which I think has not come up in our 16 discussions today, and that relates to just looking at 17 the players in the market. 18 So recently I've looked at some data provided 19 by FINRA on structured products, and when you look at the 20 transactions in these markets, one thing you observe is 21 that there are a certain number of dealers who dominate 22 the market. So if you look -- if you take the top five 23 or top ten dealers and you look at their market share in 24 terms of number of trades executed, it's a very large 25 number. 0238 1 And I think one of the lessons from the 2 financial crisis has been that when you have 3 intermediaries who play a crucial role in the market, if 4 there are certain kinds of funding shocks that hit some 5 players in certain markets, then those shocks get 6 transmitted to other markets through these 7 intermediaries. 8 And so one of the things that one needs to 9 keep in mind when you're looking at a dealer market is 10 whether there are some dealers who are very large with 11 respect to the overall market. 12 And so one thing we learned from the 13 transparency experiment for corporate bonds is that when 14 transparency was introduced in corporate bonds, when last 15 prices were reported, smaller dealers were able to 16 compete better because in the world before transparency, 17 they were just seeing a small slice of the order flow, 18 and so they could not get a sense of where the prices 19 were, and so they did not have the same amount of 20 information that larger dealers had. 21 But in the world of transparency, I think that 22 landscape becomes more balanced, and so we see in the 23 data that smaller dealers are able to compete more 24 effectively in a transparent market. 25 And so I think that transparency has these 0239 1 other spill-over effects that may not be very obvious, 2 just looking at transactions costs, including the fact 3 that it may reduce the kind of systemic risk that the 4 regulators may be worried about. 5 MR. BURNS: If one were trying to think of 6 ways to foster public dissemination, if one were looking 7 for ways to right-size that -- not saying that we are or 8 we aren't -- how would you go -- how would you tackle 9 that? How would you size it? What would be the 10 characteristics you would look for? 11 MR. VENKATARAMAN: I think if you have a few 12 players who are executing 30 percent of the trades that 13 you observe in the market, then you're just worried that 14 if there's a shock that that particular intermediary 15 faces, it could significantly affect the liquidity in the 16 market. 17 So I would suggest that by looking at, you 18 know, some of the metrics like the Herfindahl Index and 19 such, and examining the relationship between the number 20 of dealers who are active in a particular market and the 21 market quality that we observe, including volatility 22 and liquidity, we would get a better sense of what would 23 be the impact of having a few large players who are 24 dominant in certain markets. 25 MR. McVEY: Just to follow on that, I agree 0240 1 with the comments about the benefits of transparency and 2 increasing market participation. I actually want to 3 share a little bit of good news on that front, post- 4 crisis, in that the combination of transparency and 5 access to electronic order flow has enabled a lot of new 6 dealers to get into the corporate bond markets. Some of 7 them previously were equity market makers and alternative 8 market makers, others were regional dealers serving 9 either smaller institutions or retail clients. 10 But if you look at the market access 11 experience post-TRACE and post-crisis, we had 35 dealer 12 market makers on the system pre-crisis. We have 85 13 today, and those new dealers that came on since 2009 are 14 now responsible for 30 percent of our trades and about 25 15 percent of our volume. 16 So there's been a very important contribution 17 to liquidity at the margin from the new dealers that have 18 come in to help offset the drain in liquidity from some 19 of the dealers that did not make it through the crisis, 20 and others that through the increase in capital 21 requirements are holding less inventory for market making 22 than they did in the past. 23 The important thing is that if you look at the 24 large institutional clients that we serve, the biggest 25 concern is coming around the block trading market and the 0241 1 changes that are taking place there, and we talked a 2 little bit about it in the earlier panel, but you know, 3 the trade data would suggest that block trading was about 4 55 percent of all of TRACE four or five years ago, and 5 it's in the low 40s now. It's around 42 percent of 6 TRACE. 7 And this is happening at a time when the 8 overall corporate bond market has gotten significantly 9 larger. The outstanding debt in the U.S. corporate bond 10 market is up about 40 percent over the last four years, 11 so the institutional money managers are managing more and 12 more corporate debt and the block trading market is 13 getting smaller. 14 I would worry if there were any proposed 15 changes to the way that FINRA reports TRACE data today 16 with respect to its impact on the block trading market. I 17 think it's a very sensible compromise to have block 18 trades reported as five million plus, and exposing more 19 with respect to the exact size of those trades, in my 20 opinion, would reduce the incentives that dealers have to 21 make markets in block sizes at a time when block 22 liquidity is already shrinking. 23 And dealers do need time to manage their risk 24 in block trades. And to me, the way that FINRA has 25 compromised with the industry for a five million 0242 1 threshold point in high-grade corporates and one million 2 threshold point in high-yield, is serving both sides of 3 the market. It's getting the transparency that's needed 4 out to the retail market on where trades are taking 5 place, and it's allowing dealers to make markets in block 6 sizes for large institutions and manage their risk 7 appropriately as they work down that position. 8 So I think that huge strides have been made on 9 transparency and I think retail to large institution is 10 well served by the TRACE tape. We're excited that 11 there's an expansion in the TRACE tape into the asset 12 backed markets and 144As, but I quite like the way that 13 it's handled now so that large risk positions are not 14 exposed too soon in the block trading market. 15 MR. KETCHUM: Far be it for me to step in when 16 people are saying nice things about FINRA. I'm not used 17 to being part of that conversation but all right, thank 18 you. 19 (Laughter.) 20 MR. KETCHUM: I think you -- I think this is a 21 terribly important question, and obviously I think the 22 panel is doing what they properly should do, which is 23 first distinguishing between post-trade transparency and 24 pre-trade transparency, and then some of the issues 25 around that. 0243 1 From the standpoint of post-trade 2 transparency, I won't repeat Kumar's stuff, but I do 3 think the experience, including the effort to balance 4 liquidity, competition and access to information for a 5 range of investors have been pretty darn impressive. 6 Which is not to say it's the same for every 7 security, just as we now sit in the more complex world of 8 mortgage-backed securities and asset-backed securities 9 and recognize with the huge number of CUSIPs and huge 10 different designs of the security, and the care that has 11 to be taken in thinking about those issues. 12 But I do think that the assumption with 13 respect to post-trade transparency of a strong 14 presumption is the right way to be. We're -- FINRA's 15 about to take, I think, the next logical step from the 16 standpoint of at least handing it over to the SEC and 17 considering it with the Board in the next two days, which 18 is to take a look at the steadily and substantially 19 growing 144A market that from a pricing standpoint really 20 does inform the entire marketplace, plus the value of 21 having that information there. 22 I think -- I personally think -- we'll see 23 what the Board thinks and what the Commission thinks -- 24 it is an important step. 25 I think that the points made with respect to 0244 1 concern about block liquidity is always a valid and 2 important point to the issue. I would like to clarify, 3 FINRA actually has not walked away and determined that 4 there should never be any adjustment to the existing caps 5 that are in there today. We do agree with the points 6 made by the industry that given the many changes that are 7 occurring off of a variety of regulatory issues and 8 market issues and the corporate bond market today, that 9 slow is better than fast there. 10 And I can't imagine a world where there ever 11 would not be caps and would not focus carefully on the 12 impact of blocks. I can't imagine a world where the cap 13 for high-yield securities was not a million, and I think 14 it's not at all clear that sits as a very justifiable 15 number as the activities occurred there. 16 On the pre-trade transparency, obviously we 17 don't bring the same expertise, other than in that vast 18 world. We're not looking to have the same thing from an 19 equity standpoint. But I would say that it seems to me 20 that there are some things to feel very heartened for 21 from a pre-trade transparency standpoint, and some things 22 to at least watch carefully, although I'd agree with them 23 that care here and a careful examination is the right way 24 to go. 25 I think the growth from an ATS standpoint and 0245 1 the increased flexibility, the requests for quotes in the 2 dealer marketplaces have been terrific, have provided 3 additional reference points from the standpoint of the 4 quality of execution, additional liquidity, additional 5 competition, and I think really great stuff. 6 It is always a worry as those type of systems 7 grow, prosper and become more significant and start to 8 define best price. When that information is not readily 9 available across the marketplace, it always creates risk 10 from the standpoint of intermediary pricing that can be 11 addressed from a best execution and mark-up standpoint, 12 but not -- but somewhat awkwardly and not terribly 13 effectively. 14 And I think the statement in the last panel, 15 I'd underline again, which is there's always a lot of 16 education issues when you start providing retail 17 investors with more information. but fundamentally, 18 education starts with transparency. And without -- and 19 certainly in the present Internet environment, we see 20 explosions from the standpoint of firms' education and 21 tools that are provided to retail investors once the 22 transparency is there. 23 So a lot of issues, not simple questions, not 24 simple answers that are the same for all range of complex 25 securities on the corporate and mortgage-backed side, but 0246 1 I do think this is an appropriate issue for the 2 Commission to continue to look hard at. 3 MR. RAMSAY: Richard, I think you sort of 4 anticipated my next question and I don't know, maybe 5 answered it, too. I'm not sure. But around block sizes 6 and acknowledging that there is always -- defining what 7 you mean by a block and creating some exception or carve- 8 out or different treatment for block size, there's always 9 a tradeoff in terms of the value of greater transparency 10 versus other kinds of values. 11 But there's -- the assumption is there's 12 nothing magical about any one particular number, and to 13 the extent the market has changed over time, so that the 14 way that the market actually thinks of what a block size 15 might be changes over time, then perhaps the systems of 16 people who are providing transparency ought to adjust as 17 well. I think I heard you say that. 18 MR. KETCHUM: Yeah, I absolutely agree with 19 that, John. I -- you know, I think this should be driven 20 by statistical analysis not just by instincts. I think 21 some valid concerns were raised from the standpoint of 22 when we put this out for comment, which is why we're not 23 determining to go back and look at those caps at the 24 present time. 25 But I do think, you know, caps have the value 0247 1 from the standpoint of encouraging block positioning, 2 encouraging holding positions, and that's a good thing. 3 They have a cost from information, I get that's not with 4 respect to the largest institutions that sit with no 5 meaningful market disadvantage with respect to the 6 dealers. but a cost of it being able to value difference 7 for prices and whether that has something to do with 8 respect to the size of the trade or not. 9 So you always give up something one way or 10 another, and those things, I think it makes sense to have 11 care from the standpoint of approaching it, but also to 12 treat each as a subset of the market. You heard my 13 prejudice with that, I have a little trouble in my mind 14 defining a million as necessarily an inviolable place 15 with respect to a steadily increasing unrated marketplace 16 on the corporate side. 17 So I think there's things to continue to look 18 at, but I take Richard's points, and I think that that 19 really needs to be done from the standpoint of pretty 20 careful statistical analysis. 21 MR. EADY: So if we could maybe circle back 22 and talk a little bit more about pre-trade transparency, 23 again Rick, you alluded to it just a second ago. But in 24 the earlier panel, I remember Richard McVey and Kevin 25 Molloy said a couple things about displayed pre-trade 0248 1 pricing, and I think, Richard, you and Richard McVey were 2 making the point that there -- it could have detrimental 3 effects during a trade negotiation if there's too much 4 public awareness of trading prices, the prices that are 5 being negotiated. 6 Is that true for all sizes of trades? I 7 wanted to get a sense from you because you made a lot of 8 comments about block trades and the information that 9 those convey. Is there some level, without trying to 10 pick a threshold today, where that information is less 11 sensitive and more beneficial to the market generally? 12 MR. McVEY: I think that there is. And you 13 know, we're excited about the developments we see in 14 automated dealer market making and smaller trade sizes. 15 And that's a place where there could be a lot of value in 16 the aggregation of those prices being made broadly 17 available. 18 I think in the institutional market and 19 infrequently traded securities, which we all agreed 20 corporate bonds and municipal bonds are, what I was 21 cautioning against is the requirement that somehow during 22 the middle of an auction process or a negotiation with 23 one or more dealers, that an institution would be 24 required to expose those prices before a transaction 25 takes place. 0249 1 I think, to me, that would serve as a powerful 2 detriment from dealers providing competitive levels to 3 begin with, and in all likelihood is going to end up with 4 worse transaction prices rather than better. 5 But we are -- you know, the evolution of the 6 market has been really interesting. It -- nothing was 7 really trading back electronically at all in the 8 corporate bond markets back in 2000 when we started and 9 others started to emerge on the market. TRACE started in 10 2002. The market has moved forward. There's 11 significantly more transparency, there are many 12 electronic venues in the retail and the institutional 13 market. The electronic share is growing. 14 And I do see the market making evolution 15 taking place, too, and the number of dealers who are 16 starting to be able to auto quote more bonds in smaller 17 trade sizes to form the beginnings of a live market that 18 would help with pre-trade transparency and in retail 19 trade sizes, I think is very encouraging. 20 MR. EADY: That's helpful. And Kevin, just to 21 circle back to something you said in the earlier panel, 22 that there are live bid offers, two-sided markets in many 23 securities -- 1,600 securities, I think you mentioned -- 24 CUSIPs; is that right? 25 MR. MOLLOY: Yes. 0250 1 MR. EADY: But you said -- you made a 2 reference that you were seeing trades away on TRACE in 3 securities where you had live bids and offers displayed. 4 Now presumably when you're looking at these, are these 5 trading away at better prices? 6 MR. MOLLOY: No, we're looking at the markets 7 to really evaluate how competitive our markets are. So 8 what I was referring to there, in order to determine the 9 viability of the live prices we have is right, you know, 10 are these good markets, are they bad markets? 11 One of the things we do see is, when you look 12 at our markets versus the inter-dealer trades, right, as 13 on any given day we will see, you know, 300 to maybe 500 14 trades. It varies depending on the liquidity of the 15 trades that take place, right, that we could have filled 16 on the Exchange, that you know, if we have a bid for 100 17 bonds and 50 trade at a worse price, right, and we look 18 at that and say hey, let's trade, we could have filled. 19 So I think it kind of speaks to the 20 fragmentation of liquidity in the market, the OTC nature 21 of the bilateral relationships between customers, and 22 that no one really has an aggregated view of the 23 liquidity in the marketplace. 24 And I think the panel before us kind of spoke 25 about that, around the ideas of pre-trade transparency 0251 1 and you know, would it be beneficial to have some type of 2 centralized independent system that displays this 3 liquidity or this pricing for the market. 4 And to Rick's point, I think that, you know, 5 there is some level where the benefits of transparency 6 would outweigh any type of slippage around, you know, 7 making that transparent in the marketplace. So you know, 8 whatever that number is, I think the market would 9 naturally determine that. But we do think that there's 10 certainly, you know, a large subset of trades that take 11 place on TRACE. 12 And I think we went over the numbers earlier 13 as far as trades that are less than a million, trades 14 that are less than 100,000, I think they are somewhere 15 around 66 percent of the market. That, you know, if 16 there was some type of pre-trade transparency around a 17 centralized quote, that that would be beneficial, I 18 think, to investors, it would be beneficial to our -- you 19 know, to our platform to get other people to quote 20 knowing that their price will be seen by the marketplace. 21 I think those are the kind of things that can 22 going forward enhance liquidity in the marketplace, and 23 as another panelist said earlier, broaden that base of 24 participants that want to play in the market. 25 And I think that, you know, Jason mentioned as 0252 1 well that when you start to create that kind of 2 ecosystem, it does create opportunities for retail to be 3 able to post with some comfort around knowing where the 4 market is and I think streamlines the ideas around best 5 execution, and really, broadens sort of that, both the 6 investor base which we should certainly see volumes grow, 7 and I think also just liquidity in the marketplace. 8 MR. HEFFRON: Just to add to some of the stuff 9 that people have said, not to be missed is we're still 10 talking about the corporate bond market. It's still 11 completely illiquid, it's got 50,000 issues. 12 So throwing it on a screen and hoping that 13 people show up and make a price is not exactly a simple 14 thing. So I just want to remind people that there is a 15 huge evolution going on now. There weren't electronic 16 bond trades five/ten years ago. My company's nearly half 17 electronic on its bond trades. It's post-transparency, 18 everybody knows where it's traded. It can be a dark 19 pool, it can be a central limit auto book, it can be an 20 RFQ trade. 21 So I would just be careful on trying to get in 22 the way of that evolution because the technologies keep 23 coming and keep coming, and the competition is fierce. 24 And the broadening of the market makers has continued 25 since the 2008 crisis. 0253 1 You know, we speak to so many more 2 counterparties that are streaming liquidity to us or 3 point and clicking liquidity to us, and I think when 4 we're sitting here in a year, you're going to see a 5 continuing diversification of how liquidity is formed and 6 how liquidity is accessed, and it will be mostly 7 electronic. 8 MR. SMITH: I'd like to raise a point on that 9 as well. 10 One of the things -- I'm just kind of 11 listening and I'm trying to understand the logic. We've 12 got a market that on the new issue is getting larger and 13 larger. We keep bringing elephant sized deals into the 14 marketplace, but we're still talking about price 15 discovery at the retail level and that's good enough for 16 everybody. 17 I think the caps, particularly on post-trade, 18 have to be adjusted to reflect the reality. The fact is 19 we're not bringing $200 million deals any more, these are 20 billion dollars and up. Ford Motor is a perfect example. 21 And a $5 million block or $10 million block on a billion 22 and a half deal, quite frankly, doesn't mean a lot, but 23 it does tell you a little bit more about where the 24 institutional market is, or could be, above a $150,000 25 trade. 0254 1 So I think we have to think of the logic in 2 the sense of I'm starting with jumbo elephants and 3 offering them on the break, and then all of a sudden, my 4 price discovery is based on retail odd lots in terms of 5 where this thing is valued. There's a lot -- there's a 6 distance between the lip and the cup there that quite 7 frankly, you know, I'm not comfortable with. I'd like to 8 know more. 9 MR. SENFT: Tom, I was -- I just had a comment 10 with regard to your last question about pre-trade 11 transparency, and particularly I think -- I don't want to 12 put words in your mouth, but could it backfire for the 13 small trade? Was that basically the idea? 14 MR. EADY: (Off microphone.) 15 MR. SENFT: Right. So lest anybody at the 16 Commission get the wrong idea, based on what I'm about to 17 say, my job description is that I work in electronic 18 markets, and I'm having a good time, because market and 19 regulatory forces are making things more electronic and 20 you know, that's what I do for a living. So I can give 21 the speech about the benefits of electronic markets and 22 the transparency that they create. 23 And I can tell you how the market has 24 benefited from the presence of MarketAxcess and NYSE 25 Bonds and Bonds.com and the Bloomberg and everything 0255 1 else. Unquestionably, you know, the market is better off 2 for having these venues in places where we can show what 3 we are looking to buy or sell on a given day, and I think 4 every market participant connected to those systems would 5 agree. 6 But you do have these weird edge cases where 7 things can sort of backfire on you. So one of the 8 popular notions being explored today is the idea of, you 9 know, sort of enhanced auction systems and letting buy 10 side holders cross with one another, and some of them 11 operate continuously, some of them operate discreetly at 12 points in time. 13 And imagine, if you will, that there was a 14 system where everybody in the world participated in an 15 auction that took place at 2:00 o'clock in the afternoon, 16 and we showed for a particular CUSIP, you know, everybody 17 wanted to buy or sell, put your indications in. And 18 you've got a pricing service or a bunch of dealers who 19 have agreed, you know, that the mid market for this ought 20 to be somewhere around 99. Let's form our indications 21 around that. 22 And you wind up getting $5 million worth of 23 sell interest and -- well, actually let's do it the other 24 way around, $5 million of buy interest and $50 million of 25 sell interest. Well, that's a lot of information that we 0256 1 didn't have before. And you know, there's this 2 presumption that it's a good thing to have that 3 information. 4 Well, I guess it is a good thing to have that 5 information, but I can assure you, once that information 6 is in the marketplace, 99 is not the right number any 7 more, right? Clearly you've got, you know, an imbalance 8 in terms of the amount that would be sold versus bought 9 at that price. 10 You could cross $5 million at 99, but then the 11 bid would suddenly plummet for the other guys who wanted 12 to sell the 45 that was left. So you know, there can be 13 people made happy by that disclosure that afternoon. 14 There's $5 million worth of happiness in the marketplace, 15 and there's $50 million worth of unhappiness. 16 You can get, you know, the same sort of weird 17 unintended consequences perhaps, you know, from a 18 suggestion made in the last panel, that maybe a retail 19 broker should be required to give their customers the 20 option of putting bids or offers out on a limit order 21 book. 22 Well, okay. You know, it's really hard to 23 argue with giving somebody an option to do something, but 24 it could backfire, right? You put a limit order out, in 25 that case a relatively illiquid muni, your limit order 0257 1 stays there for three consecutive days. We've adjusted 2 it so it's market appropriate all the way along, and 3 there's just no takers on the other side. 4 Well, meanwhile, everybody in the market is 5 seeing that that bond just isn't trading for the last 6 three days. You have not helped your customer versus 7 just, you know, the fact that he could have taken a bid 8 or an offer from his broker on day one. 9 So you do have to be careful. 10 MR. BURNS: I want to change, if we can, this 11 is a very rich vein of conversation, but to make sure we 12 touch on various other things we want to touch base 13 about. 14 MR. SCHLOSS: Just one final -- 15 MR. BURNS: Oh, sure. 16 MR. SCHLOSS: On that last question, I think 17 what we're struggling with is you've got primary and 18 secondary. You've got the previous discussions on the 19 muni market and the other extreme would say the U.S. 20 Treasury market, which we can argue has sort of full 21 transparency both pre and post. It's got the auction, 22 it's a bid -- it's a supply/demand that's going to 23 dictate price. 24 I know a lot more about muni's today than I 25 did when I walked in this morning. That's the opposite 0258 1 effect. And the corporate bond market is sort of in 2 between. And what we're struggling with is trying to 3 figure out, you know, how much post, how much pre, what 4 size degree of transparency do we want, and try to find 5 that right balance of pre and post. 6 And it really -- you know, there's a lot of 7 the corporate bond market that can mirror what's going on 8 in the Treasury market. Very liquid, large size auctions 9 are in development, there's platforms out there being 10 developed. They will work to some extent. You've got 11 the other extreme on smaller size bonds that really do 12 benefit from the five million cap. 13 And so trying to find one sort of model that 14 fits all is going to be really hard, if not impossible, 15 to do. And the market and the technology is going to 16 evolve on either side of those, and hopefully the munis 17 with all the discussion and the comments this morning, 18 and what you guys are doing to try to fix some of this 19 mirrors -- pushes that further up. 20 But we're all trying to find sort of a 21 platform that works for everything, and we're not going 22 to find that. There's going to be evolution along the 23 way and there's going to be technology that supports 24 that. And more is better; more transparency is good to 25 the extent that it doesn't bog down the market. 0259 1 And that's the one worry bead that I would 2 leave with, is just, you know, we've got to make sure 3 that the process stays, the issuers can issue, secondary 4 bonds move, and we don't compound it with more and more 5 data that just slows down the process. 6 MR. BURNS: On the subject -- so the segue now 7 to better investor information. In the second panel, 8 Nancy, we heard you talking a bit about how you would 9 like to see more data posted to TRACE on the ABS side. 10 And Robert, I think you talked also about wanting to see 11 more information about where prices were coming from, 12 dealer to dealer and the like. I'd be very interested in 13 exploring that vein. 14 But the question I've been told we absolutely 15 need to make sure we get to is whether investors should 16 be provided more information about the compensation of 17 broker-dealers who are trading in a riskless principal 18 capacity. What would the benefits and burdens be that 19 might accompany requiring better disclosure about dealer 20 markups. 21 MS. HANDAL: So I truly believe that if you 22 are not taking the risk and you're simply crossing bonds 23 -- so this is a different story than even having it in 24 inventory. So if you've pre-negotiated the trade on both 25 sides, the information should be disclosed. 0260 1 And what that will do is provide more 2 liquidity and better price execution for the investors at 3 the end of the day, right? Because if I know Broker A is 4 going to come out and charge me two ticks, and Broker B 5 is going to charge me a point, I'm going to go with 6 Broker A obviously. 7 And that -- considering that they're riskless 8 trades and the sort of way that the market works over 9 time, it's just going to benefit the market as a whole to 10 have that information. And the example that I would use 11 would be, you know, obviously during the crisis, trying 12 to trade bonds and find price transparency was muddied by 13 brokers crossing bonds without a clear indication of 14 where one side was selling and where one side was buying. 15 So it was even harder to get an indication for the true 16 liquidity in the marketplace, the true price discovery in 17 the marketplace. 18 So to the extent that that's disclosed, I 19 believe that it adds to investor confidence, ability to 20 buy and hold bonds and ability to feel comfortable that 21 they can go out and trade bonds in the future. 22 And it also takes a bit away from the need to 23 -- through the past few years, we've seen some of these 24 buy-side to buy-side electronic platforms come through, 25 and they're pretty impossible within the ABS market right 0261 1 now, just because of the variety of CUSIPs, the 2 structures, the deals, some of the things that Neil was 3 talking about earlier. 4 But the broker benefit is that they do know, 5 okay, you're Investor A and you hold this type of paper. 6 I know Investor B likes this type of paper, so yes, they 7 should be paid something for that knowledge. But it's 8 the matter of, you know, how much that should be 9 disclosed. 10 MR. GENYK: I agree with a lot of what you're 11 saying. Our firm is in the retail business and in the 12 institutional business. On the retail side, for a purely 13 riskless transaction, we do not mark it up and we simply 14 charge a commission, and the retail financial advisor 15 works -- negotiates that commission with their client and 16 discloses that commission. 17 I think in the institutional marketplace it's 18 more complicated for the reasons that Nancy mentioned. I 19 mean if you're finding a security that you know somebody 20 wants to buy, or the reverse, and there's a lot of 21 digging and a lot of research involved in that, there is 22 the possibility for the dealer to receive compensation, 23 as long as that compensation is fair and fairly 24 disclosed. 25 So I think -- you know, again I think back to 0262 1 what Collin said earlier and Rick and others have said, 2 you know, it's not a one-size-fits-all solution, 3 particularly for this kind of issue. 4 MR. VOGEL: We just had one real example 5 with regards to a fallen angel that happened in 2011, and 6 we had some forced selling, and due to some of the 7 challenges in the broker community, the broker-dealer 8 community, there were no bids for a certain security that 9 had fallen out of the Index. 10 And as we were forced sellers, the agency 11 model rose to the top of the heap and they did a 12 tremendous job of finding the bids. We needed to move a 13 pretty good slug of paper. And the question really, 14 should we have been bogged down with what are they 15 charging for that? They found bids, we had no bids in 16 the market at the time, we needed to sell these bonds. 17 And we managed to move in excess of 100 million with an 18 agency model. And they were not taking any direct risk. 19 I knew they had found the other side, whether it was a 20 distress buyer, a ten percent yield buyer or just a 21 legitimate fallen angel buyer who has a portfolio of 22 these securities. 23 So in that context, I think it's tough to say 24 it should be, you know, a quarter tick, a half a tick. I 25 think it's -- I think it depends on the market 0263 1 volatility and the situation. 2 MS. HANDAL: I don't disagree with that. And 3 that's why I said as long as it's fairly disclosed and 4 you know what you're in for when you're selling the 5 bonds. 6 MR. VENKATARAMAN: I'd like to come back to 7 the question of what is the most effective way to provide 8 markup disclosure to customers? Ideally we would for 9 each transaction like to say what was the mark-up, but in 10 reality, bonds are not frequently traded. We don't have 11 pre-trade information to come up with a benchmark nor 12 what the mark-up would be. Those are other challenges we 13 work with. 14 But it is also the case that although we can't 15 estimate the mark-up for each transaction, we can -- we 16 do have approaches to estimate average transactions cost 17 for a type of bond based on, you know, pulling together 18 data say over the last three months for bonds with 19 certain characteristics, because we observe the buys and 20 the sells and we can come up with statistical techniques. 21 I know Dr. Edwards has done that at the SEC. 22 These are well-known models used by institutional 23 investors. 24 So one idea might be to put together a grid 25 say based on issue size, credit quality, trade size, et 0264 1 cetera, and report say by -- FINRA can do it on their 2 website or the Commission can put together a website 3 using the transaction data, which we already collect, 4 where we report what would be the expectation on 5 transactions cost of a certain size for this kind of a 6 bond. 7 And I think that would be helpful because if I 8 am a municipal bond investor or a corporate bond investor 9 and I'm looking at a certain yield of say five percent or 10 six percent, and I'm looking at the transactions cost of 11 one percent or, you know, point seven five percent one 12 way, then at least I can as a retail investor make an 13 informed decision on whether this is the kind of 14 transaction that I want to get into. I think it's just 15 more information for an uninformed investor when they 16 make a decision. I think that could be one type of 17 information that may be useful. 18 MR. SENFT: I guess my gut reaction to that 19 proposal is that it sounds imminently reasonable to make 20 more information available based on information that's 21 already out there. 22 But the problem is that a lot of information 23 gets destroyed in the production of the averages that 24 would be in your table. So you know, the average does 25 include some riskless principal trades. And by the way, 0265 1 not all riskless principal trades are created equal, some 2 truly are buy-side crossing with itself. We run a system 3 that allows our customers to do that. It's no work on 4 our part, or very little, to run it. We have a small 5 mark-up and we disclose it a priori. Everybody's happy. 6 There are others such as the situation Chris 7 outlined where, you know, it looks like a riskless 8 principal trade after the fact, but it took a lot of work 9 on people's parts to make that trade happen. People 10 might have worked for hours or days or who knows how long 11 to pull off what was a riskless principal trade. After 12 which, by the way, TRACE does a heck of a job of 13 reporting what the mark-up was. 14 But the problem is when you're combining the 15 riskless with the non-riskless in the same table, you're 16 going to have some problems there. 17 MR. HEMPHILL: Yeah, just to -- go ahead, 18 Collin. 19 MR. HEFFRON: I was just going to say we're 20 talking about stuff that's not on TRACE, right? That's 21 why you can't calculate the mark-up. So you know, 22 everything that's on TRACE, as Dexter said, they do a 23 nice job so you can calculate it, riskless principal or 24 -- because you can see it trade. It trades once, it 25 trades twice. 0266 1 MR. McVEY: Yeah. As long as it's TRACE 2 reported and FINRA has dealer buys and sells, you can 3 observe that. 4 But you know, to Dexter's point, it doesn't 5 really account for the majority of the market. There are 6 70,000 high-grade CUSIPs available on MarketAxcess today 7 and TRACE has about 4,000 unique bonds that print on an 8 average day. 9 But when we're trying to analyze transaction 10 costs, and I think we've done as much of this as anyone 11 else, and created a bid/ask index, the real problem is 12 most of the bonds are only trading one or two times a day 13 and they're not riskless trades. So we actually have 14 looked at the number of bonds out of that universe of 15 70,000 that actually have ten or more prints on both 16 sides of the market in an average day, in any trade size, 17 and you get down to about 25 corporate bonds. 18 If you want to look at only institutional size 19 trades where TRACE reporting has ten or more buys or 20 sells in the same bond and over 100,000 in trade size, 21 you're down to about ten bonds. 22 So the data is not there because the trades 23 are not there. So it's very, very difficult to have a 24 far-reaching average bid/ask spread when so many of these 25 bonds trade so infrequently, and they're carried by the 0267 1 dealers on their balance sheets for longer periods of 2 time until they can market the bond and find the other 3 side of the trade. 4 So my concern would be for that tiny slice of 5 the market where we have the data, there is reasonable 6 evidence to talk about bid/ask spreads and costs and 7 dealer mark-ups, and it's very clear on those kind of 8 bonds what the institutional mark-ups are and what the 9 retail markets are, and retail mark-ups are larger for 10 all the reasons that have been discussed through the day 11 today. 12 The problem with it is it can be highly 13 misleading with all the other bonds that trade so 14 infrequently where the dealer capital and the dealer 15 market making is what's providing the liquidity to make 16 the market operate and where those bonds may not print 17 again for a couple weeks. And that, to me, trying to 18 have a dealer then communicate something about mark-up 19 two weeks after they bought the bond is a very difficult, 20 if not impossible exercise. 21 MR. BURNS: Tacking once again, we wanted to 22 touch base briefly on best execution. And Rick, I'll 23 look to you, Rick Ketchum, and perhaps Dexter, you. 24 How are brokers complying with FINRA's best 25 execution rule which today applies to corporate bond 0268 1 transactions? Are there areas where additional guidance 2 on best execution might be helpful? 3 MR. KETCHUM: Well, I'm sure there is 4 additional guidance because any time that the market's 5 in flux, there are always concerns and issues from the 6 standpoint of how each dealer may interpret it. 7 I think dealers have built a reasonable amount 8 of controls and documentation around the basic process 9 they do to go through their reasonable diligence process. 10 And so when we get out, it's not perfect, and as was 11 noted in the last panel we do see instances in which 12 there are failures and will bring an occasional 13 enforcement; yes. 14 But the basic documentation -- this is 15 something they focused on -- I do think there's a range 16 of concerns right now that perhaps argues for more 17 guidance. It certainly argues for us to continue to look 18 at the area. You've got an evolving ATS market, as noted 19 before, not all corporate bonds, much less the range of 20 mortgage-backed and asset-backed bonds are equal. 21 There is -- we have seen instances of some 22 dealers tending to use a single ATS, not validating that 23 with a range of other sources, and we've had concerns, 24 and have concerns, about potential misleading pricing 25 that comes into some of the efforts to do that. 0269 1 So I think there's -- you know, and of course, 2 that's going to evolve over time as the ATSs grow larger 3 and choices between those products, et cetera. It could 4 be helped a little bit by the range of things noted here, 5 everything from riskless principal reporting requirements 6 to greater transparency there. 7 So I think it's a good time for us to look at 8 and provide some greater guidance about some of the 9 things we see on the whole. I wouldn't -- I would say, I 10 think, on the whole we see reasonable diligence around 11 the dealers and care in how they build it, but we do see 12 differences as we see the market changing, and we see 13 some that are operating in a way that I'd say is pretty 14 close to shortcuts and may not always give an accurate 15 and thorough analysis of really what best execution is. 16 So from all those standpoints, I think this is 17 an area that's going to be in flux, and probably an area 18 that justifies further guidance over time. I'd be 19 interested in what people feel from the industry side. 20 MR. SENFT: Well, best execution is a -- it's 21 a concept that once again, you know, seems like it ought 22 to be easy enough to define. But it gets tough, 23 depending upon circumstances. So in the markets where I 24 work, the electronic markets which are dominated by, you 25 know, for the most part, lots of relatively small trade 0270 1 sizes, where bids and offers are posted perhaps 2 indicatively, perhaps firm, where RFQs are the bread and 3 butter of the corporate bond market and you can get some 4 subset of Rick McVey's 85 guys to respond with a price, 5 you've got a pretty good idea that you got a fair 6 execution after that process, so I'm not worried about 7 the electronic markets. 8 But the markets that I get more concerned with 9 are the institutional trades, particularly around the 10 block sizes. And I guess I would just take a moment to 11 point out that in addition to this concentration that 12 exists, that several people have now pointed out on the 13 sell side of the corporate bond market. How if you go to 14 Professor Goldstein's slides and just add up the share of 15 the top ten, it came to 56 percent. 16 Well, if you look at the top ten holders of 17 corporate bonds in the United States, they own 47 percent 18 of the market. And by the way, that's up from 27 percent 19 ten years ago. So you know, I'm not saying that's a good 20 thing or a bad thing. I'm pretty sure you think it's a 21 good thing. But it's a thing, it's a real thing. 22 And that means that when those ten guys or the 23 next ten guys are looking to change asset allocations in 24 their portfolio, we're talking about some large size 25 requirements. And I think Chris would probably agree 0271 1 that if he's in a hurry to move a fair chunk of bonds, 2 that it might not be about, you know, could I have gotten 3 a better price if I waited a day or spread it out over 4 time, but really, what can my, you know, group of dealers 5 do for me right now. 6 So how you measure best execution against that 7 complex equation I think is tough to say. 8 MR. VOGEL: I'll just jump in with some stats 9 for everyone here. 10 So in Q-1, we traded 80 percent of our 11 investment-grade trades and block count electronically, 12 25 percent in notional terms. So our teams are split, we 13 have a credit team that handles predominantly the smaller 14 lot trades, and they'll use ECNs exclusively. And we 15 always check on any day where the institutional traders 16 who are trading the block-size versus the small lots 17 traders who are trading the small size. 18 Whenever they're in the same CUSIP trades in a 19 day, you'll look at a report and it's hard to fathom that 20 they are usually on top of one another. There's one or 21 two basis points difference, but it's not always the 22 institutional block size trader on the phone who's 23 getting a better price. So we do monitor that. 24 We do feel that, you know, having the right 25 mix of -- as we stated earlier, the ECNs are incredibly 0272 1 important and efficient and we may not use all of the 2 85 dealers Rick said, but we're certainly using 60 or so, 3 so we're kind of pushing some of our risk out and being 4 very mindful of our concentration risk. With the same 5 group of dealers, we have started to use some list 6 trading where we're in -- when markets are quieter and 7 you can move things or try to move some things closer to 8 mid. We're certainly not afraid to put some lists out 9 there. 10 We've experimented with some of the crossing 11 engines that Dexter had mentioned. They can be 12 incredibly helpful. The problem there is just the sheer 13 number of CUSIPs that are outstanding versus what you can 14 do in a day or a week is sometimes a little frustrating. 15 But it's certainly another way to be moving risk. 16 And again, what's always front and center at 17 BlackRock is just liquidity, and we just have, you know, 18 an insatiable appetite for liquidity, and anything that 19 comes online or is out there, we'll certainly look at it. 20 MR. McVEY: Now just one twist on the best 21 execution questions that have been brought up today is, I 22 think in most of the panels they've been focused on the 23 dealer side, and your regulated investment managers have 24 best execution requirements as well. 25 And from my perspective, it's something that 0273 1 could use some revision and updating. I think many 2 investment managers have worked with a relatively murky 3 best execution guideline as a regulated investment 4 manager. And the world has changed, and many of them 5 still have policies that suggest that getting at least 6 three dealer prices puts them in some form of a safe 7 haven for best execution. 8 And for some subset of securities, investment 9 managers are still getting three dealer prices and for 10 others they are not. But the world is also opening up 11 electronically, as has been pointed out in this panel, 12 where investors are starting to interact directly with 13 each other. And in that case, at least one of them has 14 to lead with a price, and it does not fit with best 15 execution guidelines that many of them are working with 16 today. 17 So I think in terms of utilizing technology to 18 open up the markets and find alternative sources of 19 liquidity, to promote the capital markets and what is a 20 rapidly growing U.S. credit market, it would be helpful 21 to revisit the best execution guidelines on the 22 investment management side and work with the industry as 23 the market is evolving away from just purely a client to 24 dealer model. 25 MR. BURNS: I think with the time left us, 0274 1 we're going to turn for a moment to the offering process. 2 But before we do, I just want to look down the aisle both 3 ways and make sure there aren't any other questions kind 4 of in the -- I think in the trading and markets hub 5 before we move on. Or Commissioner Gallagher has 6 anything else? 7 (No response.) 8 MR. BURNS: Great, thank you. 9 So with the little bit of time that's left, we 10 heard a little bit also in the second panel, we kind of 11 bit around the edges about difficulty some firms 12 encounter in being shut out of the market if they're not 13 there in any size, even middle and larger firms are 14 having troubles as well. 15 And I just wonder if any of the conversation 16 that we've had this afternoon related to price 17 transparency could turn around and enhance the primary 18 offering process itself for corporate bonds or other 19 fixed income securities. Are there other ways to improve 20 pricing of corporate bonds or asset backed securities? 21 And for that, Neil, I might turn to you first. 22 MR. SCHLOSS: I think, as I said this morning, 23 the process for us is working pretty good, with access to 24 investors and we play an active role in the dissemination 25 of how the bonds get distributed. Obviously, certain 0275 1 deals we have less say on it because of the investor base 2 and who's in it and where their price is, but to the 3 extent we can, we play sort of an active role. 4 So we are very concerned with the larger 5 investors who are dominating the market, because we do 6 like -- we like them in the deal, but we also like a 7 broad distribution of investors, and so we are helping to 8 facilitate that. 9 And at least from what we see today, and the 10 real test to this market will be how it reacts when the 11 market isn't so good -- at least today, with broad 12 distribution, broad interest, good transparency, both pre 13 and post, given the amount of bonds we have outstanding 14 -- we're pretty comfortable with the level of 15 transparency there is today. 16 Again, I think the real test will be when the 17 markets turn and go opposite direction, you know, will 18 some of these trading platforms, will technology actually 19 help on the offering side. 20 And I know they're being developed, we've seen 21 several. And there's a balance of how much to use those 22 versus what might be available through direct and through 23 brokers and dealers and banks. 24 So I think it's a balance, and again, we're 25 going to try to use -- given the amount we do every year, 0276 1 both in unsecured and ABS, we really have to try to 2 balance our distribution the best we can and not overly 3 concentrate that in any one channel, both private and 4 public. 5 MS. STARR: I have a question. In terms of 6 -- you were talking about the ten sort of largest players 7 in the market, what is the characteristic of those 8 investors? Are they Fund families? What are they 9 generally? 10 MR. SCHLOSS: On the investor side, it's the 11 PIMCOs, it's the BlackRocks, it's the Vanguards, it's 12 those type of large both mutual funds for the broader 13 distribution, but also, you know, large, whether they're 14 private or public, managed money for private or public 15 pension funds, and, you know, there's just a large pool 16 of money. 17 And so when they come to us and want, you 18 know, $200 million of a $1 billion transaction, that 19 causes -- you know, obviously I don't like the 20 concentration in a single investor, but it gets split up 21 pretty quickly into individual Funds within their 22 structure. 23 And so from the standpoint of once it gets 24 placed, then it becomes the individual owner of that 25 account who's actually dictating, you know, whether it's 0277 1 active or not, how the bonds get traded out of there. 2 MS. STARR: Right. But then we had heard this 3 morning that because of the needed liquidity on the Fund 4 families, in the event of redemptions, et cetera, that 5 that played a role in whether or not the middle market 6 companies could actually access the corporate debt 7 market. Is that really the framework on, you know, what 8 the risks are or why the middle market companies are 9 having a harder time accessing the corporate debt market, 10 you know, at least on a -- I would suppose on a 11 registered basis? 12 MR. SMITH: Let me comment to that. I mean, 13 again, I think this kind of -- from a logical standpoint, 14 while I agree with the diffusion of a particular takedown 15 by one of the oligarchs of the bond industry, it's fine, 16 and that's just reality. 17 The fact is it's still one investment committee 18 and it's still one risk decision set by and large, and it 19 disseminates through the group. So if they decide they 20 don't like something, boom, it's not going into the 21 Funds, it's not going into the separate accounts, it's 22 not going into the ETFs, whatever we control. 23 So there is an uniformity of thought there, 24 because it's one firm, it doesn't mean you have six 25 different ways of looking at the market, otherwise nobody 0278 1 would give money to them. 2 So logically you have to sit there and think, 3 well, geez, aren't these guys really controlling the 4 action, and that's the hundred-pound gorilla in the room 5 here. It's like the buy side is taking control, that's 6 the fact, and the sell side pretty much is doing what the 7 buy side wants, my view, our view. That's not a terrible 8 thing, but it's a thing nonetheless, to use Dexter's way 9 of looking at it. 10 I think that there are inefficiencies that 11 could be inherent in that, particularly when you're in 12 fast-moving markets. I think they are inherent 13 inefficiencies in terms of why are bonds moving and in 14 what hands are they, and in what size are they moving. Am 15 I aware of the fact that 100 million of my bonds have 16 just traded underneath me into the bad side of my price? 17 I don't know, because it's not disclosed. 18 When you get to the trading, does it go to 19 Rick? You know, here again, we've got what in essence is 20 becoming a bond trading club. I hate to put it in the 21 vernacular, but to allow only 50 or 60 guys in and say 22 this is the way it's going to be and we're going to tell 23 you which bonds are going to be good for everybody, and 24 we'll only get a 15-minute look before we actually get 25 underway here to figure out what's going on, and we're 0279 1 only going to charge you a couple of bips and everybody 2 else, sorry, you're not -- you don't have the flow to be 3 in this game. 4 Just inherently, I have a problem with that in 5 terms of defining that as an efficient market. It's an 6 efficient market for 50. It's not an efficient market 7 for 500. 8 MS. STARR: So the retail then really comes in 9 in the after-market rather than at say initial issuance 10 as you would say in an equity deal? 11 MR. SMITH: I would tell you in a hot deal, 12 Mr. Retail, unless he's getting through one of these guys 13 and it's disseminated down through the mutual fund, 14 there's not a snowball's chance that he's going to get 15 that. Very few times in a hot deal, and that's kind of 16 the world we live in. 17 MR. SCHLOSS: Although there are unique 18 programs intentioned to go to retail directly, and I know 19 we have one, and it's about somewhere between five and 20 ten percent of our unsecured issuance. 21 MR. CROSS: I have a question, slightly off 22 topic but related to the difference between sort of the 23 primary market price and the impact on the secondary 24 market. 25 There's been a big controversy the last few 0280 1 years as to whether or not it is an outlier or a common 2 structure of the market that bonds trade up quickly in 3 the secondary market in the week or two after a primary 4 offering, and whether or not that's just the way the 5 market is supposed to work to foster liquidity, or 6 whether that is a case that should raise concerns about 7 basically whether or not issuers are getting the best 8 price in the primary market when they turn around and 9 watch the bonds trade up shortly after that. 10 Anyone care to comment? 11 MR. SCHLOSS: Since the issuers are the guys 12 at least that make the decision, I'll start. 13 I think by and large, we would like bonds to 14 break slightly positive on a trade, and slightly 15 positive, if the price at par is 100.5, not 105. 16 Now the art in the execution is a part of 17 that, and I think from our perspective, you know, we're 18 in the market a lot so, you know, keeping the investor 19 base engaged in the primary is important. Every once in 20 a while we don't have one that breaks positively, which 21 is okay, and so it's important that you do have a 22 balance, but you don't want to mis-price it. We have a 23 requirement for our shareholders to get the best price. 24 The issue you've got and the uncertainty you 25 have is you don't exactly know how much demand there 0281 1 truly is after the market breaks, and how much demand you 2 fill in. 3 That's what -- I'll go back to what I said I 4 think earlier, from the standpoint of how we allocate the 5 bonds. You know, somebody puts in an order for 200 6 million. I think Bob, you said it earlier, you put in 7 for 100 and hope to get five. You know, that inflates 8 the overall size of the book and so you really have to 9 understand what the real size of the book is when you're 10 setting price and where investors fall out on that scale. 11 But like, you know, if you're a frequent 12 issuer, you'd like to have a bond trade slightly positive 13 on the break. 14 MR. CROSS: Do you have any sense of whether 15 or not there's sort of a common practice through typical 16 trends on the up tick shortly after the secondary market? 17 MR. SCHLOSS: No. I mean, it's pretty -- I 18 mean, we try to get as fine-tuned as you possibly can. 19 ABS, I think, works the same way, and Nancy, I don't know 20 if you've got a flavor for that. But I think ABS, 21 smaller investor base, you know pretty well where it's 22 trading, the asset types, but I think a similar faction 23 there, too, is you want to have the investor base. 24 MR. CROSS: I mean the obvious slight tension 25 there is between more liquidity in the market and best 0282 1 price for issuers in the primary market. 2 MS. HANDAL: I just mentioned on the asset- 3 backed securities market, essentially, you're exactly 4 right. And what happens -- I talked about this a little 5 bit before -- is the brokers will kind of test the market 6 to say, okay, are you in at -- call it swaps plus 50. 7 Yes, I'm in for 100 million. Are you in at swaps with 8 48? No, I drop. 9 So they're sizing the book based on that, and 10 that's the best transparency that they can get. And it's 11 going to be distorted by what we call people, you know, 12 padding the orders. And it's part of what happens when 13 there's a food fight for bonds, which is what's going on 14 right now. 15 And it -- you know, at various points in time, 16 it's pretty natural for the break to be positive. And 17 the reason that I'd say that is that if you have a 18 limited number of bonds and you have a lot of buy and 19 hold investors buying them after the break, so when 20 you're issuing, it's sort of supply and demand. 21 You issue and you put out 500 million of bonds, 22 there is, you know, a limited amount of demand, and you 23 figure out where that is on the secondary when most of 24 those bonds are put away by a buy and hold investor, or a 25 portion of them are. There's just naturally a smaller 0283 1 amount of supply so it should break positive. I think 2 that's healthy and pretty natural. 3 MR. SCHLOSS: And I think getting back to 4 something that was said earlier today, you know, when you 5 think about dealer inventories, and the -- you know, the 6 normal underwriter who would have taken -- you know, 7 depending on how the deal was short or long, depending on 8 the strength of the transaction, those are becoming fewer 9 and farther between, too. 10 So they're not putting their balance sheets up 11 anywhere near what they once did from the standpoint of 12 initial transactions, which again causes more distortion 13 on the ability to get that price sort of exactly right on 14 issuance. 15 MR. CROSS: Right. Very helpful perspective. 16 Thanks. 17 MR. VOGEL: Just real quick, I'll just add I 18 think this is being exacerbated because the inflows in 19 investment grade credit funds is up 40 billion net 20 inflows this year. You've got QE all over the globe, 21 people fighting for mortgage product and spread products. 22 So I mean I understand you're saying on the break they're 23 trading up, it's just net issuance is not where it needs 24 to be. 25 The question is if this rotation that I've 0284 1 heard mentioned a few times today ever happens, you know, 2 will this be the case in a bear market for bonds? And 3 for those that didn't sell their duration yet, their ten- 4 year notes trading at 172, which is a mere five basis 5 points off the low yield of the year, you've got plenty 6 of time to get in there and make some good sales. 7 (Laughter.) 8 MR. BURNS: And here we try to get people to 9 avoid talking book. 10 We had a -- this is terrific, and as usual in 11 circumstances like this, just a rich conversation. 12 Fortunately, our comment file will be going on and 13 staying open long after we've all left for the day. 14 I'm going to turn it back over to John Ramsay 15 or John Cross, if they have any final remarks, but thank 16 you to this panel very much. 17 MR. RAMSAY: I will just echo my thanks. 18 We've talked about a lot of important things today, 19 identified important things, and I found it terribly 20 interesting. Certainly appreciate all of the time that 21 people have devoted. 22 This is kind of the best kind of public/private 23 partnership to be able to talk about these kinds of 24 things in as open a way as we can. I know this will be 25 very helpful as we go forward in refining our thinking 0285 1 about next steps in the fixed income space with 2 Commissioner Gallagher and others. And so again, thank 3 you very much. 4 (Whereupon, at 4:15 p.m., the Roundtable was 5 adjourned.) 6 * * * * * 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0286 1 PROOFREADER'S CERTIFICATE 2 3 In the Matter of: ROUNDTABLE ON FIXED INCOME MARKETS 4 File Number: 4-460 5 Date: Tuesday, April 16, 2013 6 Location: Washington, D.C. 7 8 9 This is to certify that I, Susan Watkins, 10 (the undersigned), do hereby swear and affirm 11 that the attached proceedings before the U.S. 12 Securities and Exchange Commission were held 13 according to the record and that this is the 14 original, complete, true and accurate transcript 15 that has been compared to the reporting or recording 16 accomplished at the hearing. 17 18 19 20 ____________________ ____________________ 21 (Proofreader's Name) (Date) 22 23 24 25 0287 1 REPORTER'S CERTIFICATE 2 3 4 I, Jon Hundley, reporter, hereby certify that the 5 foregoing transcript of 285 pages is a complete, true and 6 accurate transcript of the testimony indicated, held on 7 April 16, 2013, at Washington, D.C. in the matter of: 8 ROUNDTABLE ON FIXED INCOME MARKETS. 9 10 I further certify that this proceeding was recorded by me, 11 and that the foregoing transcript has been prepared under my 12 direction. 13 14 15 16 Date: __________________________ 17 Official Reporter: __________________________ 18 Diversified Reporting Services, Inc. 19 20 21 22 23 24 25