0001 1 UNITED STATES SECURITIES AND EXCHANGE COMMISSION 2 3 4 5 6 ROUNDTABLE ON DECIMALIZATION 7 8 9 10 11 12 TUESDAY, FEBRUARY 5, 2013 13 (Amended: March 13, 2013) 14 15 16 17 18 19 20 21 22 23 U.S. Securities and Exchange Commission 24 100 F Street, N.E. 25 Washington, D.C. 0002 1 SEC COMMISSIONERS PRESENT: 2 Elisse Walter, Chairman 3 Daniel Gallagher, Commissioner 4 5 PARTICIPANTS PRESENT: 6 James Burns 7 Colin Clark 8 Chris Concannon 9 Brian B. Conroy 10 Kevin Cronin 11 R. Cromwell Coulson 12 Joshua L. Green 13 Frank M. Hatheway 14 Chris Isaacson 15 Paul Jiganti 16 Scott Kupor 17 Craig Lewis 18 Maureen McCarthy 19 Lona Nallengara 20 Maureen O'Hara 21 Adam V. Reed 22 Stephen Sachs 23 Jeffrey M. Solomon 24 David Weild 25 Kent Womack 0003 1 C O N T E N T S 2 PAGE 3 Opening Statement 4 James Burns 4 5 Daniel Gallagher, Commissioner, SEC 5 6 7 Panel 1: Evaluating Concerns Relating to Tick 8 Size for Small and Middle Capitalization 9 Companies 31 10 11 Panel 2: Evaluating Concerns Relating to Tick 12 Size for the Securities Market 13 Generally 87 14 15 Panel 3: Studying the Effects of Alternative 16 Tick Sizes 146 17 18 19 20 21 22 23 24 25 0004 1 P R O C E E D I N G S (9:33 a.m.) 2 MR. BURNS: I think we will try to get started. 3 Good morning, everyone. I'm Jim Burns, Deputy Director of 4 the Division of Trading and Markets at the SEC. 5 It's my pleasure to welcome you to the Commission's 6 Roundtable on Decimalization. We are pleased you have been 7 able to join us for what we anticipate will be an informative 8 and productive series of discussions. 9 We had hoped at this point to introduce our 10 Chairman, Elisse Walter, and our Acting Director, John 11 Ramsay, who are in Brussels. They will be joining us by live 12 feed in a little bit. We will turn to them when I get the 13 cue they are in situ. 14 While we are waiting for them, I would like to 15 convey the regards of some of the Commissioners who are not 16 able to join us today. Commissioners Paredes and Aguilar are 17 not able to attend in person but they send their regards and 18 Commissioner Aguilar in particular sends his regrets that 19 illness prevents him from being here in person. I understand 20 he is following us by webcast. 21 We are very pleased to welcome Commissioner 22 Gallagher who joins us here on the podium. Commissioner, 23 thank you for being with us. We would appreciate any 24 comments or observations you might wish to share here at the 25 outset or as the morning progresses. 0005 1 COMMISSIONER GALLAGHER: It's good to be here. I 2 was going to make a little joke about the Commissioners who 3 aren't here, but I know Luis is watching, and I know he'll 4 call me in his hoarse voice and yell at me if I make a bad 5 joke, so I won't do it. 6 It's great to see everybody here. I know there are 7 a lot of people watching on the web, too. 8 I think this Roundtable is just a great indication 9 of how important what might seem like small issues are to the 10 work of the Commission and to investors. 11 Something like decimalization, given our agenda 12 these days with the Dodd-Frank rulemaking's, with the JOBS 13 Act, and with our day to day work, could just be overlooked 14 as something unimportant. 15 To the contrary, it's a very important issue as 16 Congress told us when they mandated the study that the staff 17 carried out last year in the JOBS Act. 18 I am very agnostic as to where this goes, open 19 minded. This is truly a Roundtable where I think we have the 20 right people here. We have the right learning and expertise. 21 The Commission as well as the Commission staff can learn and 22 make a decision as to how to proceed. 23 As we saw last Friday with our Small Business 24 Committee and their recommendation that we move down the path 25 towards a pilot in this space. This is a real issue that 0006 1 impacts capital formation at the very important, critical 2 early stages for companies, I think, considering IPOs. 3 I think because of that, we should take it 4 incredibly seriously. I'm looking forward to the 5 interactions today. I will, as Jim pointed out, jump in 6 where I see fit. 7 Thanks everybody for being here. Thanks to the 8 staff for putting it together. Great job. 9 MR. BURNS: Thank you, Commissioner Gallagher. 10 Thank you, too, for your leadership and dedication to the 11 staff's efforts on this and other important initiatives. 12 For starters, the staff has had a hard time 13 wheedling down the list of invitee's to the esteemed 14 panelists you have here today. 15 I want to remind folks at the outset, other 16 interested parties, please offer your comments or 17 observations. We would like to include those in the comment 18 file. Already, we have a very robust comment file. We do 19 appreciate your comments as they may come in. 20 A special thanks -- we will do this periodically 21 throughout -- to the panelists before us and on the panels 22 that follow for the time and expertise they are lending to 23 the Commission and the staff as we assess a topic of great 24 importance. 25 Quickly, in particular, today's over-arching topic 0007 1 is whether the current tick size regime in the U.S. equities 2 markets can be enhanced to more fully promote vital public 3 policy objectives, such as fair and efficient trading, the 4 protection of investors, and vigorous capital formation. 5 We are tackling this in three parts today. The 6 first panel is focused on tick size as they affect small and 7 mid-cap companies. The second panel is focused on tick sizes 8 in the context of the equities markets more generally. 9 After lunch, the third panel will focus on ways to 10 generate useful data to enable the Commission to evaluate 11 tick size issues in as disciplined and a responsible manner 12 as possible. 13 Among other things, the third panel will assess 14 whether a pilot program might be implemented to assign 15 varying tick sizes to a control group of stocks of different 16 types of companies. 17 The objective of such a pilot as well as other 18 potential means of generating data would be to establish a 19 solid empirical basis for any action the Commission might 20 determine to take or not take in the future. 21 We are optimistic that a data driven approach can 22 be developed that will prove fruitful for addressing tick 23 sizes, and more broadly, I think this type of data driven 24 approach -- we all think, I believe -- can provide an useful 25 template for addressing many of the other complex and 0008 1 pressing market structure issues that currently face the 2 Commission. 3 I can run down a number of important steps we have 4 taken in recent years, market-wide circuit breakers, market 5 access rules, updated SRO rules relating to breaking of 6 trades, sub-quotes, or other enhancements on the horizon or 7 with us now, audit trail, large trader, the new MIDAS system 8 that will collect and facilitate data analysis within the 9 Division and among the offices of the Agency, and the quantum 10 leap in hiring of personnel with specialized expertise. 11 These initiatives have begun and should continue to 12 strengthen our approach to the U.S. equity market structure. 13 There is plenty to do. 14 In addition to the tick size issues, there are many 15 other issues relating to market structure that we should 16 further examine and we will endeavor to do so. 17 As many of you know, our concept release asked 18 questions about tick size a number of years ago. There are 19 many other areas of endeavor we want to pursue under that 20 concept release. 21 Whatever the Commission may do, the most promising 22 path forward is to adopt a data driven approach to these 23 issues, to provide a solid empirical basis for the 24 Commission, to enable it to address open issues and determine 25 what if any regulatory actions are appropriate and how they 0009 1 might best be calibrated. 2 In sum, we hope and expect the Commission will 3 benefit very much in the coming months from various sources 4 of empirical research. 5 Today's discussion on tick sizes is an excellent 6 example of how we can pursue an approach that enables 7 progress on a wide range of other issues. 8 I am not sure if we have the Chairman yet. If not, 9 I'd like to turn it to a key leader in our efforts, Craig 10 Lewis, the Commission's Chief Economist, and Director of 11 Risk, Division of Risk, Strategy and Financial Innovation, 12 for his introductory remarks. 13 Craig? 14 MR. LEWIS: Thank you, Jim. I'd like also to begin 15 by thanking everyone for joining us today for this important 16 Roundtable to discuss the impact of tick size on small and 17 mid-sized companies, market professionals, investors and U.S. 18 security markets. 19 As always, I must make clear that the views I 20 express today are my own and do not necessarily reflect the 21 views of the Commission or my colleagues on the Commission 22 staff. 23 The genesis of today's Roundtable is a study 24 conducted by staff in my Division, the Division of Risk, 25 Strategy and Financial Innovation, in response to a 0010 1 congressional mandate included in the JOBS Act. 2 Specifically, the JOBS Act required the Commission 3 to examine the effects of decimalization on initial public 4 offering's and small and middle capitalization companies. 5 Staff in RSFI worked diligently with staff from the 6 Division of Trading and Markets and the Division of 7 Corporation Finance to complete this study within the 90 time 8 frame mandated by Congress. 9 We took a three pronged approach to conducting the 10 study within this time frame. Specifically, the study 11 entailed (1) reviewing empirical studies regarding tick size 12 and decimalization; (2) participation in a meeting of the SEC 13 Advisory Committee on Small and Emerging Companies on the 14 impact of market structure on small and middle capitalization 15 companies, and (3) a survey of the tick size conventions in 16 non-U.S. markets. 17 A report of our study was delivered to Congress in 18 July of last year. The report, which is available on the 19 Commission's website, explained that the impact of mandating 20 an increase in the minimum tick size for small capitalization 21 companies on the structure of our markets and on the 22 willingness of small companies to undertake initial public 23 offering's, is at best uncertain. 24 In light of this conclusion, the report ultimately 25 recommended that the Commission not proceed with specific 0011 1 rulemaking to increase tick sizes but should consider 2 additional steps that may be needed to determine whether 3 rulemaking should be undertaken in the future. 4 One of the additional steps recommended in the 5 report was for the Commission to solicit the views of 6 investors, companies, market professionals, academic's and 7 other interested parties on the topic of tick sizes, how to 8 best study its effects on IPOs, trading and liquidity for 9 small and middle capitalization companies, and what if any 10 changes should be considered. 11 Indeed, this recommendation is what brings us here 12 today. 13 Before we start soliciting your views, I thought it 14 may be helpful to discuss examples of what the Commission 15 staff's findings have revealed thus far. 16 Although the staff ultimately concluded that the 17 impact of mandating an increase in the minimum tick size is 18 still uncertain, the staff did report findings that should 19 help illuminate our discussions today. 20 First, the staff undertook a literature review of a 21 number of academic studies that discussed the effects of the 22 change to decimal pricing on overall market quality and where 23 possible, the impact on small and middle capitalization 24 stocks. 25 As summarized in this report, the academic 0012 1 literature documents not only lower spreads after 2 decimalization but also smaller trade sizes, decreased long 3 run volatility, and no apparent reduction in market maker 4 profitability. 5 On net, these findings suggest that the initial 6 effect of decimalization was improved market quality. Most 7 notably, for large capitalization securities. 8 For those papers that examined smaller and middle 9 capitalization securities, the effect, particularly for 10 spreads, appears to be relatively minor, but for reasons 11 outlined in the report, it is difficult to draw strong 12 conclusions from a review of these available studies with 13 respect to potential adverse effects on the change to 14 decimals on small and middle capitalization companies today. 15 The staff also considered the discussion and 16 material presented at a meeting of the SEC Advisory Committee 17 on Small and Emerging Companies on market structure issues. 18 As the staff explained, at that time, members of 19 the Advisory Committee commented that it may be hard to 20 disentangle the impact of decimalization on small company 21 IPOs for several concurrent reasons, including the enactment 22 of the Sarbanes-Oxley Act in 2002, and the more recent 23 emergence of high frequency trading in dark pools. 24 I would note, however, that just last week, the 25 Advisory Committee, after further review, recommended that 0013 1 the Commission adopt rules to increase tick size for smaller 2 Exchange listed companies in the United States that will 3 allow such companies to choose their own tick size within a 4 range designated by the Commission. 5 The staff study also surveyed tick sizes in a 6 variety of countries that have significant smaller company 7 IPO activity. This review revealed that other countries use 8 multiple tick sizes in contract to the one size fits all 9 approach in the United States, which suggests there may be 10 viable and perhaps preferable alternatives to uniform tick 11 size rules. 12 Although it also is worth noting that the study 13 explains that regulatory decimalization in the market lowered 14 the minimum allowable tick size, it did not mandate market 15 participants quote narrower spreads and quoting narrower 16 spreads appears to have been the result of continued market 17 forces. 18 Thus, although we are left with many questions, our 19 study has established a foundation for analyzing issues 20 related to tick sizes. 21 This Roundtable represents another step in 22 collecting valuable information on how best to proceed in 23 this complex area of regulation, and will hopefully give us a 24 better understanding of the costs and benefits of potentially 25 changing tick sizes. 0014 1 Finally, I would be remiss if I failed to echo 2 Jim's sentiment that the study also provides an example of 3 how data driven empirical analysis can illuminate policy 4 discussions at the Commission. 5 I wholeheartedly believe that this approach if done 6 appropriately strengthens the Commission's policy 7 considerations. 8 My Division hopes to continue to work as partners 9 with our colleagues in the other divisions to develop similar 10 approaches to addressing the numerous regulatory issues that 11 face the Commission, including the market structure projects 12 that Jim mentioned. 13 I want to thank you again for attending this 14 important Roundtable. I look forward to listening to your 15 views and the helpful and productive discussions that we will 16 have today. 17 Now, I will turn the floor back over to Jim for 18 further comments. 19 MR. BURNS: Thank you very much, Craig. I thank 20 you on behalf of Trading and Markets for all the efforts you 21 and your staff have put into this and some of the other 22 important equity market structure initiatives. Indeed, many 23 other initiatives we take on. 24 I am not sure whether we have the Chairman with us 25 yet. 0015 1 I am pleased to introduce Lona Nallengara, the 2 Acting Director in the Division of Corporation Finance, who 3 along with his Corporation Finance colleagues, leads many 4 staff efforts relating to issues that affect small and mid- 5 cap companies. 6 Lona, we will give you a signal when the Chairman 7 and John are on line. 8 MR. NALLENGARA: Okay. We are going to start the 9 first panel? 10 MR. BURNS: If you folks can indulge us, we will 11 let you know if we need to pause for a moment. We will tack 12 on some time at the end. 13 MR. NALLENGARA: Thank you, Jim. Our first panel 14 is titled "Evaluating Concerns Relating to Tick Size for 15 Small and Middle Capitalization Companies." 16 Before we get started, I thought I'd introduce some 17 of the staff members we have today that are going to help 18 facilitate the discussion. 19 You have already met Jim and Craig. To my right, 20 Kathleen Hanley, is our Deputy Director in the Division of 21 Risk, Strategy and Financial Innovation, and also our Deputy 22 Chief Economist. 23 Down at the other side working from the end, Daniel 24 Gray, Heather Seidel, David Shillman, and Gregg Berman. They 25 are in the Trading and Markets Division, and Amy Edwards, 0016 1 also in Risk, Strategy and Financial Innovation. 2 Together, this group and their teams are leading 3 our efforts in thinking and re-thinking market structure 4 issues. We are glad they are here today to help us with the 5 discussion. 6 I would also like to welcome our panelists today. 7 We have an impressive group here today. We appreciate the 8 time they have taken to join us. 9 I would like to ask them to each introduce him or 10 herself to everyone here. We are hoping to have as much time 11 as possible for both the discussion as well as the Chairman's 12 remarks, so if I can ask each of the panelists to keep their 13 introductions brief. There is a lot of information about 14 them in the materials and on our website. 15 I will start with -- as a product of the Canadian 16 Educational System -- Professor Womack, to start with the 17 introductions. 18 MR. WOMACK: Thank you very much. I have a career 19 that started in investment banking. After that, I went back 20 to get a Ph.D. at Cornell. My research has been about the 21 equity markets, specifically IPO markets, conflicts of 22 interest and research on IPO firms. 23 When we get to the topic today of how this might 24 impact research, the research environment for securities, 25 hopefully I'll be able to add a couple of things. 0017 1 Right now, I'm at the University of Toronto in 2 Canada, and expect to stay there for my career. 3 MR. BURNS: I'm going to break in right here. Do 4 we have the Chairman and John Ramsay? 5 CHAIRMAN WALTER: You do. 6 MR. BURNS: Terrific. We are delighted you could 7 join us. We are in the middle of introductions, so we will 8 turn the floor over to you and then to John, and then we will 9 continue with introductions. Thanks for joining us. 10 CHAIRMAN WALTER: Thank you very much, Jim, and 11 after the drama it has taken to get us here and actually on 12 camera, it would be nice if we had something very dramatic to 13 say, but I can assure that, that's not the case. 14 MR. RAMSAY: We don't. 15 CHAIRMAN WALTER: Good morning to everyone. We are 16 sorry to interrupt the flow but we did want to be able to say 17 a few words to you. Welcome, all of you, and to say that 18 John and I both regret that we are unable to join you. 19 John and I are both out of the country today 20 discussing over-the-counter derivatives with our counterparts 21 from Europe, Asia, and the Americas, and thus, not able to 22 join you for this important event. 23 We do, however, want to let you know that we are 24 very interested in the issues concerning decimalization, and 25 we are very eager to hear from all of you and to get the 0018 1 results of your discussions today. 2 Promoting capital formation and economic growth 3 are, as you know, central objectives of the SEC, and the 4 equity markets we regulate, and we have a special concern for 5 the needs of smaller companies that need access to capital to 6 grow, expand, and flourish. 7 I know that many feel that today's market 8 structure, including the current tick size regime, does not 9 recognize the particular needs of smaller companies, and that 10 a detailed examination of the current structure is important 11 and appropriate. 12 I am pleased that this issue is getting needed 13 attention today, and that today's panels include a broad 14 range of participants who will be able to speak to the needs 15 of smaller companies and give the current structure the 16 scrutiny that it deserves. 17 We want to thank all of you for your participation 18 and for lending your expertise to this effort. 19 I am also very interested in discussion today about 20 the potential for a pilot program and how it might be formed 21 to put to the test some of the ideas being discussed today. 22 Given the complex nature of our modern markets, we 23 have increasingly called upon the ability to test potentially 24 significant structural changes in a well thought out and 25 controlled fashion that can provide us all with the types of 0019 1 data and information we need to make informed decisions. 2 It will be interesting to hear from all of you, if 3 you think that would be a worthwhile endeavor here, and if 4 so, how you would structure it. 5 As I said, I'm grateful to all of you for your 6 willingness to share your time and expertise, and I'm also 7 grateful that Craig, Lona, Jim, and their staffs, as well as 8 Commissioner Gallagher, will be participating in our absence. 9 You have a dedicated and insightful staff team, and 10 they will be briefing me fully on today's discussions and how 11 we can best move forward in the future. 12 Thank you, and I'll turn the floor over to the 13 Director of our Division of Trading and Markets, John Ramsay. 14 MR. RAMSAY: Thank you, Chairman Walter. I know 15 the staff of the Division of Trading and Markets and other 16 Divisions have spent a great deal of time and effort in 17 planning this event. 18 From my Division in particular, let me especially 19 thank Deputy Director Jim Burns, Gregg Berman, the new 20 Associate Director for our new Office of Analytics and 21 Research. Heather Seidel, Dave Shillman, Dan Gray, Ilya 22 Fradkin, Steve Kuan, and Kelly Riley. 23 Thanks also to Commissioner Gallagher and to the 24 panelists who have given their valuable time to contribute to 25 the discussion. 0020 1 I hope you will take this effort as an indication 2 of how seriously we are considering the funding needs of 3 smaller companies and the topic of decimalization in 4 particular. 5 I know this event will do much to advance the 6 discussion around this multi-faceted topic, but beyond 7 advancing the dialogue, I also hope that the insight gleaned 8 today will help the staff to develop a concrete course of 9 action, pilot program or otherwise, going forward, and beyond 10 this issue, I hope that this effort may serve as a model as 11 we review other aspects of our equity market structure. 12 What that means for me is being willing to 13 challenge old assumptions and think creatively, and where 14 possible, gather hard data that can inform our policy 15 choices. 16 This has to be a group effort, the staff working 17 with every constituency that is affected by a particular 18 issue, and if done in that way, is sure to pay off for 19 investors and also the companies that rely on them to prosper. 20 Thanks again for all of you being there. Again, to 21 the panelists, for taking their very valuable time. 22 (Live feed connection lost.) 23 MR. BURNS: I think he was going to say thank you. 24 (Laughter.) 25 MR. BURNS: Just like the Super Bowl power going 0021 1 out. 2 David almost doesn't need an introduction here, but 3 David? 4 MR. WEILD: Thank you. First of all, I want to say 5 thank you for including me, and I want to just acknowledge 6 the fact that I think you have some terrific public servants 7 here. I have enjoyed the opportunity to get to know people 8 at the SEC. I think it's a tough spot to be, but I think you 9 are doing a terrific job, and with a little bit of luck, will 10 make some progress. 11 I was former Vice Chairman of the NASDAQ Stock 12 Market. I oversaw it at the time, more than 4,000 listed 13 companies. I followed my mentor, Hardwick Simmons, over 14 there. 15 Prior to that, I had run Corporate Finance and I 16 ran Equity Capital Markets, pricing over 1,000 equity 17 transactions, about 500 IPOs, during the course of my career. 18 For a while, I was the co-Chair of Strategic 19 Planning for the Investment Bank, Equity Research Sales and 20 Trading, obviously prior to the Global Research Analyst 21 settlement back when you could do it that way. 22 At Grant Thornton, who has been absolutely terrific 23 supporting our work, we have written a number of sort of 24 thought leadership pieces, which I think informed much of the 25 JOBS Act, including the first one in 2008, IPOs in the ICU. 0022 1 That looked at the very material drop off in the 2 number of IPOs, particularly small IPOs, that started to 3 occur in the late 1990s, coincident with what I call part of 4 the trinity here, which is the order handling rules in 1997, 5 followed by Regulation ATS, alternative trading systems, 6 culminating with decimalization in 2001. 7 Then we followed that up with a paper called "The 8 Wake Up Call for America," which actually started to document 9 the systemic decline in the number of listed companies in the 10 United States, the punch line of which is we have lost 44 11 percent of all listed companies since the peak in 1997. 12 Think about that number. It's sort of staggering. This 13 is operating companies. We take out the Funds. We are down 14 44 percent. 15 For me, what I'm passionate about, is how do we 16 reverse that trend because we think it's killing jobs in this 17 country, we think we have lost about 10 million jobs in terms 18 of opportunity costs. We can debate about that. 19 We think it's a big number particularly when you 20 look at Enrico Moretti's work, Professor at the University at 21 Berkeley, that says there are five service sector jobs that 22 are created for every one technology job. 23 Thank you. 24 MR. SOLOMON: Hi. My name is Jeff Solomon. I'm 25 actually the CEO of a growth investment bank called Cowen and 0023 1 Company. We do focus on small company capital formation in 2 both equity and the debt markets. 3 I'm also a member of the Committee on Capital 4 Markets Regulation, and really today, an unofficial 5 Ambassador of a number of equity market participants, buy 6 side/sell side clients and non-clients alike, with whom I've 7 engaged on this topic. 8 I also was a huge proponent of the decimalization 9 move at the end of the last decade or the beginning of the 10 last decade, and I feel like I've come almost full circle on 11 it, particularly as it relates to small companies. 12 Widening tick increments for small cap stocks I 13 definitely think will help to foster a better ecosystem, and 14 it brings back fundamental investors who are otherwise 15 paralyzed by the speed of the market and the lack of real 16 fundamental buyers and sellers at finite gathering points. 17 My view is that widening increments will actually 18 enable firms like mine to begin initiating more research on 19 small cap companies, once again, which in turn spawns the 20 necessary liquidity in the marketplace for new issuers to 21 consider doing IPOs, IPOs that will raise capital to hire 22 more people, as David mentioned. 23 I read David's piece. It certainly motivated me to 24 try to do something more than just talk about it. It 25 inspired me to get involved, to see if we could make some 0024 1 positive change. 2 While I have a definitive view that decimalization 3 is one of the root causes of the current equity market 4 dysfunction for small caps, I'm also willing to allow that 5 there are other factors contributing to the lack of 6 liquidity. 7 I'm not polemic about that. What I would say is 8 I'm here primarily to urge you to consider trying a pilot 9 program. At the end of the day, we will know whether it 10 works or doesn't work only after we have seen it. 11 There will be a number of people on this panel and 12 maybe subsequent panels who will give you tons of views and 13 tons of studies that will support or rebut or claim, and I 14 think the answer is we just don't know until we try. 15 That's the purpose of the pilot program, which is 16 to really to try to let us see if it does get better, and 17 let's give it enough time to develop. 18 As we have our discussions here today, I found an 19 interesting quote from Professor John Kay who actually did a 20 similar kind of study in the U.K., talking about their equity 21 markets. 22 He did testimony last year at the National 23 Association of Pension Funds in December and he said the 24 following: "If equity markets provide the link between the 25 activities of publicly traded companies and returns to 0025 1 beneficiaries, it follows that the functions of equity 2 markets are to enhance the capabilities of publicly traded 3 companies on the one hand and to achieve the best possible 4 returns to saver's on the other." 5 This, he said, "Is the central criteria by which 6 the performance of equity markets is to be judged." 7 Hopefully, that provides some context, and we are 8 here to help in any way we can. 9 MR. NALLENGARA: Thank you. Ms. O'Hara? 10 MS. O'HARA: Good morning. I'm Maureen O'Hara. 11 I'm the Robert Purcell Professor of Finance at Cornell. I've 12 been there forever. I do research on a wide range of issues 13 in market micro-structure. I've worked on a wide range of 14 topics including the effects of fragmentation on market 15 quality, the effects of high frequency trading on how market 16 making works. 17 Lately, I've been working on issues relating to 18 liquidity or illiquidity of stocks, particularly those with 19 low prices. 20 I've also in the past worked on the liquidity of 21 IPOs, and in particular, what happens to IPOs as they enter 22 into the market with respect to the trading of IPOs, the 23 roles of the underwriter market maker and the roles of 24 others. 25 Certainly, a wide range of issues that come to the 0026 1 floor today. I think a lot of these issues that we are going 2 to be speaking about cover a wide range of issues, including 3 sort of the liquidity costs that individual investors face. 4 I'm looking forward to hearing from my colleagues. 5 I also, in addition to my day job, am Chairman of the Board 6 of Investment Technology Group, ITG. We do a lot of the 7 transaction costs analysis in the market. 8 Through our studies we have a feel for how 9 transaction costs have changed in these markets, and I am 10 looking forward to hearing the comments from my colleagues 11 today. 12 MR. KUPOR: Good morning. My name is Scott Kupor. 13 I'm the Managing Partner at Andreessen Horowitz. We are a 14 three and a half year old venture capital firm focused 15 principally on investing in IT companies. We do invest 16 multi-stage, so anywhere from very early stage companies all 17 the way to growth capital opportunities. 18 We also have the privilege of all of us having come 19 from the operating world, so we have had some experience both 20 financing new companies as well as actually being part of 21 companies that have gone through the IPO process. 22 I'm very excited today to be a part of the panel 23 and hopefully lend some perspectives based on that 24 experience. 25 Thank you. 0027 1 MR. GREEN: My name is Josh Green. Thanks very 2 much for inviting me to be here today. I began in Silicon 3 Valley in the 1980s as a corporate and securities lawyer. I 4 had the privilege of working with entrepreneurs on taking 5 over 100 of their companies public and saw the rise of the 6 four horsemen, the demise of the four horsemen, and have been 7 doing this for quite a long time. 8 In 2006, I decided to leave the legal profession 9 and join Mohr Davidow as a general partner investing in clean 10 technology investments. 11 I am today the Chairman Elect of the National 12 Venture Capital Association. We represent about 90 percent 13 of the overall venture capital industry. 14 As I think most of you know, the NVCA was a leader 15 in the adoption of the JOBS Act, specifically under Title I 16 of the JOBS Act. We believe it is very effective -- although 17 it's quite preliminary at this point -- on ramping of 18 companies to the public markets. 19 The issue now before us is the public markets are 20 the highway and they have potholes. One of those potholes 21 has to do with decimalization. 22 I look forward to sharing my thoughts on how we can 23 improve the system through pilot programs to derive empirical 24 data that will ultimately encourage more capital formation by 25 small and emerging companies. 0028 1 MR. COULSON: Hi. I'm Cromwell Coulson. I am the 2 CEO of OTC Markets Group. We operate financial marketplaces 3 for 10,000 U.S. and global securities. Our trading system is 4 an SEC registered alternative trading system. 5 We have a broad range of securities, 1,500 ADRs, 6 1,500 ordinary shares, 600 small community banks across the 7 country. We have 3,000 SEC reporting companies. We have 500 8 or 600 non-SEC reporting companies that use our disclosure 9 services to publish information into the market. 10 We have a mix, all from the very high range global 11 500s down to the smallest penny stocks. We have very strong 12 less active, less liquid securities. 13 In the search for an efficient market, we may have 14 made our markets too slippery for investors. I think the 15 most important thing is how do we make the investor 16 experience when they are buying or selling us small 17 securities that they get liquidity, that they see liquidity, 18 that they don't see a tiny amount of liquidity that then when 19 they take it, yo-yo's away before they can buy the amount 20 they want to own. 21 That's what we have done. We have created a market 22 where it's very slippery as we have gone to remove friction. 23 We have experience with price increments because 24 back in 2004, we had Professor Christy take a look at our 25 trading data through that year. 0029 1 He came up with the idea that marketplaces should 2 organize quoting activity so that liquidity would cluster. 3 He found natural clusters in price increments in our 4 marketplace. 5 We implemented those. Those actually worked really 6 well because our goal was to increase displayed liquidity by 7 broker-dealers. 8 However, as manning and limit order protection came 9 along, FINRA did not share our view that orders should also 10 have increments, and you can't really have "increments" 11 without order increments because there is a lot of reverse 12 manning problems, which is technical speak for a broker 13 buying at one price and having to sell at a worse price 14 because it's not far enough away from the bid. 15 What we did learn was you need to have both quotes 16 and orders. Two, we recently had the tier sizes of our 17 marketplaces lowered by FINRA rules. 18 We have seen a lot of interesting data there, too. 19 We lost about 15 percent of our broker-dealer's price quotes. 20 They didn't leave the market. They went un-priced. 21 We actually didn't see a contraction of spreads. 22 They stayed the same in community banks, everything else 23 widened a little bit. We saw more inter-day volatility. We 24 saw significantly less, over half of our displayed liquidity 25 went away because the tier sizes were lowered. 0030 1 This is what's really great about actually making 2 changes and watching it. As we were going through the debate 3 with FINRA, their reviews were saying broker-dealers aren't 4 going to lower their tier sizes to the minimum if we lowered 5 the minimum. You can't be sure of that. Well, we can be 6 now. 7 There was also a viewpoint that if the minimum tier 8 size was lowered, there would be more broker-dealers wanting 9 to be dealers in the space. We actually haven't seen that. 10 What we have seen in our marketplace is the 11 transactional providers, the highly automated firms, those 12 are our largest customers now, and they are really good at 13 providing efficient transactions, but we are losing out on 14 the pricing specialists. 15 We are losing out on the regional investment banks, 16 the retail banks, that provide the price setting 17 functionality, and we need to bring those back. 18 We are big supporters of bringing in increments, 19 but we also think if price increments come in, market makers 20 should be displaying larger sizes. 21 That is because we should go back to the investor 22 experience. The investor should start seeing more liquidity 23 offered by our markets at levels at which they want to 24 transact. 25 We run an e-Bay of markets, and right now, the 0031 1 shopkeeper shelves are too thin for investors. We need to do 2 some experiments to try to fix that. 3 MR. CONROY: Good morning, everyone. I'm Brian 4 Conroy. I'm President of Fidelity Capital Markets. 5 Fidelity, as many of you know, is one of the largest 6 providers of financial services in the country, with just 7 under $4 trillion in assets under administration and 8 management. 9 We are a firm that is focused on the small 10 investor. We service tens of millions of retail investors 11 and shareholders in our mutual funds. 12 We greatly appreciate the opportunity to work with 13 the Commission and the staff today, coming from the 14 perspective of the firm, which does not have an economic 15 interest in market structure as we do not provide research 16 and we do not provide investment banking services or collect 17 underwriting fees. 18 Thank you. 19 PANEL 1: EVALUATING CONCERNS RELATING TO TICK SIZE FOR 20 SMALL AND MIDDLE CAPITALIZATION COMPANIES 21 MR. NALLENGARA: Great. I think we are ready to 22 start. We have a list of questions here that we were hoping 23 to get through. I am hoping that once we get started, the 24 discussion can flow naturally and we may not necessarily need 25 to turn to each of the questions. We are hoping we can cover 0032 1 the topics in the questions by the end of our session. 2 Taking the questions out of order, to throw 3 everyone off, I thought it may be helpful to get a sense of 4 the IPO market today, what impact if any the JOBS Act has had 5 on the IPO market. 6 David, you have written extensively on this. Jeff, 7 you have a lot of sort of real hand experience. I am 8 wondering if you could give us a sense of where you think the 9 IPO market is now, what impact the JOBS Act has had or could 10 have on that. Let's start there. 11 MR. SOLOMON: To start with, certainly doing 12 confidential filing's has helped. You can point to one area 13 where issuers are more willing to go down the path of being 14 able to file confidentially. 15 That was a big concern for a lot of people, 16 especially in the tech and health care areas where there was 17 some concern about disclosing company strategies any time 18 before you absolutely need to do that. 19 We have seen a number of companies tap the market 20 using the confidential filing mechanism. 21 It's hard to say that there are more companies that 22 are doing it than would have otherwise done it. I think if 23 that's the measure for whether or not the JOBS Act is 24 happening, it's still too early to know. 25 The IPO market is still very challenged. I still 0033 1 see and speak to a number of people in the venture community 2 who weigh every day why they would think about trying to 3 generate liquidity for their investment positions when they 4 could just go on and sell those companies. 5 If you think about the strategy that has been 6 inbred into the source for these new companies, which would 7 be the venture capital community largely, I think it's still 8 very much at a minimum, run a dual track to see if I can get 9 these companies sold. 10 In a worse case scenario, I'm not even really 11 thinking about the public markets because I'm actually from 12 the outset not preparing my companies to go public. I'm just 13 preparing them to go to a certain size and then sell to a 14 larger aggregator. 15 It will take some time for that to work through. I 16 don't think the JOBS Act in and of itself will actually 17 change market liquidity or will change a number of issuers, 18 but it certainly is a step in the right direction, and I was 19 a huge proponent of that. 20 MR. WEILD: I would just add a couple of things. 21 First of all, 90 percent of the companies qualify as an 22 emerging growth company under Title I of the Act, and 90 23 percent of those seem to be filing as confidential filer's 24 from what we have seen. To echo Jeff's comments, that's 25 really getting used. 0034 1 I think the other provision that we find some 2 excitement with when talking to people on the Street and down 3 in Silicon Valley is the testing the waters provisions, 4 specifically for de-risking highly technologically 5 complicated companies. 6 In the biotech sector, I've heard some folks that 7 run equity capital markets, some of the firms like Jeff's, 8 saying the process at some point going forward won't take a 9 biotech company public without first investing a lot of time 10 meeting with institutional investors and understanding 11 whether or not the market is there. 12 I think that will cut failure rates over the long 13 run. It will take a while for that to take hold. I think 14 that was very beneficial. 15 The rest of it I think is going to take longer to 16 kind of play out. I think it was Josh that mentioned you 17 need the highways to be attractive because ultimately, once 18 companies go public, they have to sustain issue price. 19 If they break issue price because of the lack of 20 after market support, economics in the IPO window shut very 21 quickly. Predictability is not there. The risk of failure 22 is high. People start to avoid the IPO market right at the 23 get-go. 24 MR. NALLENGARA: Much of the current discussion 25 that we are having today -- this is an issue that we have 0035 1 been talking about for a long time -- much of the issue has 2 been focused on the IPO market and searching for causes for 3 the huge decline in the number of IPOs. 4 Maybe I could ask someone to talk about the impact 5 of decimalization on the IPO market but then contrast it or 6 correlate it to other sort of regulatory events or other 7 events over the last 15 years. 8 One of the questions that we have had, one of the 9 questions the Advisory Committee has had, is sort of 10 isolating what impact decimalization or what impact other 11 events have had. 12 MR. KUPOR: I agree with the general proposition 13 which is there are a number of things that have obviously 14 created this issue, but they all kind of fall, at least from 15 my perspective, under the general rubric of how do we create 16 enough liquidity and enough kind of financial incentive to 17 actually create markets in small cap stocks. 18 Decimalization certainly is a component of that. I 19 think Jeff mentioned it as did David, global research 20 settlements, very significant in that respect. 21 You really had two kind of very fundamental 22 components that provided most of the revenue and most of the 23 profit making opportunities to support research at the end of 24 the day. 25 Decimalization certainly created profits on the 0036 1 trading side of things, often used to cross subsidize 2 research in most of these institutions, and at the same time, 3 pre-global research settlement, you had a significant portion 4 of investment banking revenues obviously going also to create 5 a profit pool that could subsidize research. 6 I think you have certainly a combination of those 7 things along with as David mentioned, ATS and others. 8 When you kind of bubble all these things together, 9 at the end of the day what you have done is you have taken a 10 lot of the financial lubricant out of the system for small 11 cap stocks, and therefore, things like research, which 12 obviously helps create and enable trading liquidity, things 13 like profitability at the trading level which incent market 14 makers to actually take risk positions in these companies, 15 the sum total of those things certainly has attributed to 16 drying up a lot of the trading activity and interest in small 17 cap stocks. 18 MR. WOMACK: I think this issue has focused on 19 decimalization as the problem. I think we are missing 20 something in terms of the very big picture - the economics of 21 capitalization has changed in the last few years in a pretty 22 big way. 23 I brought a graph which I guess I can't show, but 24 essentially it shows as IPOs have collapsed over this last 25 decade, private equity has taken off. 0037 1 If you're looking for something that's more causal 2 than decimalization, I think it's really more about this 3 issue of maybe the perfect market that we all wish and hope 4 was still there, the IPO market is vibrant, maybe it just 5 isn't the driving force any more. 6 I have a suspicion that we are trying to make 7 something happen that isn't there. 8 With respect to decimalization per se, if I look at 9 the five or six years before decimalization finally kicked 10 in, the IPO markets were going down precipitously during that 11 period of time. 12 One can't argue that decimalization actually caused 13 that. It didn't even kick in until four or five years later. 14 I have a healthy suspicion up front before we even 15 start thinking about this that many other things are going on 16 and decimalization is possibly a small part of the fix but 17 not a very big one. 18 MR. GREEN: I want to go back to a comment that 19 Jeff made earlier, the M&A activity is a source of liquidity. 20 However, this is all a very careful balance. We must 21 understand that many of these initiatives that were taken 22 over time, whether they be regulatory, SOX, or otherwise, or 23 were exogenous to that, 2000 .com bust, just market forces 24 themselves, 2008 financial crisis, had an exaggerated effect 25 on the most fragile of all public markets, and that is the 0038 1 IPO market and the like. 2 As a result of that, the absence of a viable IPO 3 avenue has had a significant impact in my view on the 4 robustness of the mergers and acquisitions market. 5 It's a check and balance to keep those prices 6 appropriate that ultimately provide the return to the 7 investors that make all that work. 8 My own feeling is that decimalization has had an 9 impact. It's one of those factors that is actually quite 10 important to retaining robust public markets, so that becomes 11 an essential part of the ecosystem as this very fragile 12 ecosystem has existed in the past. 13 MR. NALLENGARA: Professor Womack raised a comment 14 on what may be the real impact of decimalization. Brian? 15 MR. CONROY: I think he made some great points 16 about the fact that there are a variety of events that have 17 taken place over the last dozen or so years which have 18 contributed to the decline of the IPO market. 19 We at Fidelity do not feel that decimalization is 20 one of the major contributors to that. We have not seen the 21 evidence that would indicate in any way it has had any effect 22 on the IPO process itself relative to the others. 23 What we find interesting is most of the discussions 24 are around the fact that the increase in tick size would 25 support the economics of firms to create research for these 0039 1 small and mid-sized companies at a time when these banks, 2 these firms, have the ability to raise commissions to their 3 clients. This is no longer a net market, this is a 4 commission based market. 5 The question we have is if the economic lever 6 exists for firms to raise commissions to subsidize their 7 research, to create the opportunity for a more robust small 8 and mid-cap research marketplace, therefore, IPOs, why isn't 9 that taking place? 10 Why are we talking about a pilot program to 11 artificially increase the tick size to create a marketplace 12 which helps market makers at a time when many of these firms 13 aren't even electronic market makers. 14 MR. NALLENGARA: Professor O'Hara? 15 MS. O'HARA: Let me follow on Brian's comments, 16 which I think really focus on the fact that as we look at 17 what has happened with decimalization, we have to recognize 18 it changed the picture for everyone. 19 It changed it for everyone in many ways in a good 20 direction. As we look at transactions costs, no matter how 21 you measure them, transactions costs are a lot lower now than 22 they ever were. That used to be considered a good thing. We 23 used to think that was a great idea. 24 Now it's cheaper for people to trade in these 25 markets. We generally think that induces people to be more 0040 1 willing to trade. Now we are hearing the other side of the 2 coin, maybe traders, per se, are better off, at least for 3 listed stocks, and I'll defer to Mr. Coulson about his OTC 4 stocks, but at the same time, we are arguing that we need to 5 provide more returns to the market makers, not necessarily 6 because we want to make market makers better off, but because 7 we want them in turn to subsidize research. 8 I think that linkage is really tenuous. 9 MR. SOLOMON: Let me just explain a little bit 10 about some of the ways the market economics work. I would 11 love to go into Fidelity and ask them to pay me six cents a 12 share and not have to worry about it. The likelihood of them 13 actually doing that is zero. 14 I would argue just like Professor O'Hara says, 15 that's not really the way the world is going. 16 Fidelity should be trying to reduce its transaction 17 costs as much as it possibly can; it should. The idea of my 18 going into Fidelity and saying hey, why don't you just pay me 19 six cents, I can write more research. That doesn't work. 20 It's not practical. 21 What I would say is decimalization, again, one part 22 of it, but a key part to driving research coverage in the 23 after market. That's what people, I think, don't quite 24 understand. 25 If you can't afford to write research in the after 0041 1 market, then it's really hard to convince a company they 2 ought to go public because of the fear there might be a lack 3 of liquidity. 4 From our standpoint, I'll speak as a small cap 5 issuer, Cowen and Company is a public company. Just this 6 past week we did an acquisition. I have two research 7 analysts that cover us, two research analysts from growth 8 banks. 9 I can't get the bulge banks to pay any attention to 10 me because I don't pay them any fees. They can't make any 11 money trading my 350,000 shares a day at a penny increment. 12 If you total up the sum total and the spread in my stock over 13 the course of the year, if you had 100 percent of the market 14 share in Cowen stock, you make $700,000. It just doesn't 15 work. 16 The only reason why people cover us is because they 17 think we might pay them an investment banking fee at some 18 point. 19 Now, one of the two firms that cover us, KBW, just 20 got bought by Steifel. I can't get the KBW analyst on the 21 phone to talk about the acquisition that we just did that 22 transforms our company. The only person that showed up was 23 the Sandler analyst. 24 I think there is something more to this than just 25 whether or not we are reducing costs. I'm all for that. We 0042 1 do it. We have done it on our own. This is about being able 2 to provide an ecosystem for an after market, that enables 3 firms like ours and other firms to cover companies like ours 4 without having to worry about when is the next investment 5 banking fee coming to come to pay for that. 6 MR. COULSON: As we are talking about having 7 increments, I don't think the debate should be about having 8 increments that really artificially widen where things trade. 9 They should really be about organizing where trades 10 are clustering now. When you go watch an auction at 11 Sotherby's for a Warhol, they are not running the auction at 12 penny increments for the Warhol. eBay has increments. 13 Marketplaces use increments as an organizational 14 skill, and that creates meaningful price impact for people to 15 be trading. 16 We have had a push in our market structure which 17 comes from when Congress brought out the national market 18 system. Congress said there should be the opportunity for 19 investors' orders to interact without the intervention of a 20 dealer. They said "opportunity." 21 We have had a regulatory view that there is 22 something icky when a broker-dealer supplies liquidity to 23 their investors. There is something bad when you are 24 providing a service that your customer wants. 25 The regulatory process has been trying to turn 0043 1 broker-dealers into agency brokers going to an order book. 2 That is a viewpoint. 3 We have been slowly -- except for the fastest, most 4 high tech, whatever increments we want to do, we can go to 5 .00001 and send it to Knight, UBS, that will be fine. They 6 will wiggle around in slippery markets at high speed and will 7 be incredible at it. 8 We are blowing away the price setting world. We 9 are blowing away the participants that either through their 10 investment banking, through their institutional sales and 11 trading, through their retail offices -- we used to have 12 Merrill Lynch or A.G. Edwards who would trade a community 13 bank where they had a retail office. 14 Those were great price setting firms. They would 15 be there with real liquidity. The transactional firms will 16 be fine. We need to bring in those price setting firms 17 again. 18 MR. WEILD: I wanted to circle back to something 19 that Kent said a little bit about, the problem starting 20 before decimalization in 2001. 21 There were a series of things, if you look at the 22 history, the first thing that put a spotlight on this was the 23 Christy and Schultz study which came out in 1994. Then we 24 had the Manning rule, which basically said don't trade in 25 front of your customer in 1996. In 1997 was the order 0044 1 handling rules, but there was no way to reflect the customer 2 limit orders into the market until you did it electronically 3 with Reg ATS. 4 The seismic shift in the market was really in 1998, 5 Reg ATS. Everything was leading up to that. We went from a 6 quote driven market where traders could manage their risk to 7 an electronic order driven market, and what we call the 8 bankable spread or the maximum opportunity on the desk went 9 from the maximum quarter point quoted spread to the minimum 10 tick size overnight, which was a 32nd for these smaller 11 stocks or 3.125 cents. 12 As a practitioner running strategy for the firm, we 13 went through a massive realignment internally. You can 14 remember the SOES, the small order execution system. 15 These things were actually gutting the economics 16 and people were moving out of these markets and trying to 17 adjust. 18 I had a conversation with Dan Case, who was an old 19 friend, who told me when they sold Hambrecht & Quist, he was 20 the President of Hambrecht & Quist, and Dan unfortunately 21 passed away in 2002, but he said one of the reasons that they 22 were selling to J.P. Morgan was they thought in a bad banking 23 market the cash equities business was no longer going to be 24 profitable enough to sustain the infrastructure. 25 SPEAKER: He was right. 0045 1 MR. WEILD: He was right. I talked to Dan Sturgis, 2 who was the CEO of Josephthal that did small IPOs. He's an 3 investor in the second market. Dan told me that with the 4 order handling rules coming in, they knew that was the end 5 right there of the small IPOs. 6 Kent said private equity is the place to look. We 7 have looked at private equity. The problem is that private 8 equity are cash flowing businesses. If you look at the whole 9 portfolio of other kinds of businesses, the technology 10 businesses, small IPOs have been sort of eviscerated from all 11 sectors. 12 You could see uptake in the private equity cash 13 flowing businesses and you probably have, but then the 14 question you have to ask is why haven't you seen a sustained 15 small IPO market in other non-private equity sectors, and the 16 answer is you haven't. 17 All small IPOs have dropped like a stone over that 18 period really from sort of 1996 to the culmination in 2001. 19 There was no pilot program done when the order 20 handling rules and Reg ATS were implemented back in 1997 and 21 1998. This was a wholesale change for our stock markets. 22 It's amazing to me, and I have to ask the question 23 why it didn't occur, short term measurements, which is what 24 most micro-market economists do, are not helpful when you 25 understand that all the people changes that go on are long 0046 1 term. 2 We don't react over night, firing research 3 analysts, cutting compensation, because it puts us at a 4 competitive disadvantage. This stuff goes on over years and 5 years. 6 Just like a buddy of mine who is Professor Robert 7 Schwartz from Baruch said, stock markets are still adjusting 8 to changes that occurred in the late 1990s even today. 9 Take away is just don't look at decimalization. 10 You really have to look at the order handling rules and Reg 11 ATS. That is where we saw a very quick drop off in the small 12 IPO. Ask why we didn't do pilot programs back then, and also 13 recognize that it's not just simply private equity companies 14 that are depressed in small IPOs, it's everything across the 15 board. 16 MR. NALLENGARA: Professor Womack? 17 MR. WOMACK: Good points, David. Before we get too 18 far away from the topic that's come up already, Professor 19 O'Hara and Mr. Solomon mentioned the idea of research coming 20 back into the market through some kind of liquidity provision 21 and profitability, I think that is a really big question. 22 I guess I'm skeptical as to whether we would see 23 small investment banks provide more research coverage if a 24 higher tick size created a more liquid and profitable 25 environment for them. 0047 1 I've talked to a couple of research directors 2 recently after I knew I was going to be on this panel to ask 3 them do you think that's what would happen, if the market was 4 more vibrant, you would spend more money in research on 5 smaller companies. They were quite dubious. 6 MR. SOLOMON: I'm a CEO, and what I would tell you 7 if asked the question how many small cap -- how many non- 8 large cap stocks do you cover at your firm, which is focused 9 on financing growth companies? How many, as a percentage of 10 our research universe? What do you think the percentage of 11 small cap stocks are that we cover today? 12 MR. WOMACK: Not very many. 13 MR. SOLOMON: Not very many. The reason is because 14 I can't get paid for covering small cap stocks. 15 MR. WOMACK: No amount of extra -- 16 MR. SOLOMON: However -- 17 MR. WOMACK: From five cents or ten cents is going 18 to change that. 19 MR. SOLOMON: However, if I saw the opportunity to 20 make liquid markets and reasonable spreads -- we are hiring 21 our first small cap trader. I'm dedicated to figuring out 22 ways to bring back liquidity, and I'm going to put my own 23 capital at risk in the marketplace to create liquidity, not 24 asking for guarantees, I'm going to put my own capital at 25 risk to try to drive liquidity to names we think are 0048 1 important to our coverage universe. 2 As a part of that, we will cover those companies. 3 One will follow the other. It will take some time. As David 4 said, if we go through a pilot program, don't make it like a 5 year because I would assure you that yes, Kent, you will be 6 exactly right, no one will rush out to re-hire a bunch of 7 analysts to publish on small cap stocks because it's 8 expensive to do that. 9 You need to let that ecosystem regenerate. When 10 the coverage begins to take hold, you will see liquidity come 11 back to the market because there's no place today for 12 individual investors or small cap institutional investors to 13 find the information they need to make intelligent investment 14 decisions. It just doesn't exist, for the most part. 15 MR. NALLENGARA: Brian? 16 MR. CONROY: I guess I would just shift the 17 conversation a little bit and just remind everyone that in my 18 opinion there has been a paradigm shift in research in this 19 industry. 20 Part of that paradigm shift is because of the 21 decrease of publishing analysts on the Street for a variety 22 of economic factors, many buy side firms, including Fidelity, 23 have greatly grown their research in-house staff. We have 24 over 100 analysts now on the asset management side focused on 25 following companies. 0049 1 While we are not publishing, Jeff and his firm will 2 come in, I'm sure, four times a year and meet with the 3 analysts and the portfolio managers who are going to make the 4 investment decisions on behalf of their Funds. 5 That paradigm shift has occurred. That is one of 6 the reasons Jeff suggested firms like ours aren't willing to 7 pay the six cent commission that we were 20 years ago, and 8 that we are concerned that if there are changes in 9 decimalization, increasing the tick size, our transaction 10 costs go up to no discernable benefit for either a retail 11 client or end investor. 12 As far as market makers in this space, most of the 13 traders on the asset management side I'm sure would tell you 14 they don't want partners when buying and selling positions 15 for the firm in these small and mid-cap stocks. 16 We don't need small liquidity events in order to 17 build a position. That just creates more obfuscation in the 18 marketplace. We would rather it be an agency market, and we 19 recognize that the main reasons there is less liquidity in 20 these are the obvious ones, the float is smaller, position 21 concentration is higher. 22 I think it is important to separate decimalization 23 and the economics of research from this because the world has 24 begun to move on. 25 MR. LEWIS: I'd like to revisit a point that Kent 0050 1 made earlier in the presentation. Before I joined the SEC, I 2 was a Finance Professor at Vanderbilt University, and I 3 routinely teach classes on corporate finance. 4 Part of every one of these classes is a discussion 5 about the equity offering process. We go through IPO issues. 6 We go through seasoned equity issues. 7 One of the stylized facts that emerges from the 8 empirical evidence around IPO issues is the cost to an issuer 9 of actually going public. 10 Professor Womack noted the rise of private equity 11 as a channel for raising equity capital. I'd like to hear 12 your comments on the change from using public equity as a 13 source of capital to possibly a lower cost source of capital, 14 i.e., going through the private placement market and the 15 exceptions to the Reg D offering process, if you have any 16 thoughts about that. 17 MR. KUPOR: I have a couple of comments. We 18 certainly have seen both in the form of secondary markets, 19 private and secondary markets, as well as new entrants and 20 sources of capital. 21 Even in the businesses that we are in, which is 22 principally financing IT companies, we even see companies 23 like Fidelity and T. Rowe Price and others actually 24 participating in late stage private financing's now, which is 25 something at least we hadn't seen in quite a while. 0051 1 We have also seen other alternative sources of 2 capital like secondary markets. 3 I think one interesting policy question that it 4 does raise for the SEC consistent with protecting investors 5 and ensuring fair markets is are you at risk of potentially 6 having more of a two tier market system over time, which is 7 you have private equity investors like ourselves, you have 8 kind of secondary equity investors who are accredited and 9 therefore obviously can participate in those markets, and yet 10 the public markets where you have individual investors being 11 able to actually trade those stocks continue to be much less 12 accessible to higher growth companies. 13 One of the things certainly that we see and you see 14 this in a lot of the numbers is an elongation of the time to 15 going public for these companies which in general also means 16 that the growth trajectory and the kind of appreciation 17 opportunities in the public markets are probably less so than 18 they were when we had a more robust small cap market. 19 I agree with the general statement which is I think 20 there are secondary sources of capital coming into this 21 market. I think there is a bigger policy question 22 potentially for the SEC and other organizations as to what 23 the implications of that are if we don't enable some of these 24 companies to go public at a time which is early in their life 25 cycle where there is more growth trajectory and therefore 0052 1 hopefully appreciation for public market investors. 2 MR. GREEN: I would add a couple of things to that. 3 One is what I'll call the growth equity sector, not the pure 4 private equity sector, but the growth equity sector is a 5 reasonably small amount of the total capital markets in terms 6 of the availability of capital. 7 I don't see that changing exponentially over the 8 near term because the opportunities still needs to 9 understand, the companies themselves. It requires a 10 significant amount of due diligence, relationship building 11 and all the rest of it. It's much more of an individualized 12 dance. 13 The public markets by contrast, notwithstanding the 14 cost to go public, provide a far lower cost of capital than 15 growth equity will in terms of raising that money. 16 Not only that, but enables the fundamental issue of 17 liquidity, and that promise of liquidity is actually one of 18 the major motivator's that a private equity or growth equity 19 investment cannot assert. 20 We actually have now a subcommittee in the National 21 Venture Capital Association on growth equity, to help 22 stimulate that and encourage it. We do think it's a sector 23 that needs more encouragement. 24 I don't think it will be able to become -- just in 25 terms of dollars alone in the foreseeable future -- anything 0053 1 close to what the reasonable potential is of the public 2 markets to provide capital. 3 MR. NALLENGARA: Gregg? 4 MR. BERMAN: Can I go back to a couple of comments 5 about the implication of changing tick sizes to something 6 that is larger? 7 One of the common refrains that we hear is I can't 8 buy stock, it's illiquid because the spread is too high 9 already. For many small and mid-cap stocks, they don't trade 10 the penny. 11 The actual physical quoted spread might be seven 12 cents, eight cents, 15 cents, 20 cents if it's an extremely 13 liquid stock. 14 What aspect of changing the tick size would either 15 stabilize that size or can you talk a little bit about why 16 the existing wide spreads for illiquid stocks don't seem to 17 provide the motivation that's needed to gain the profits and 18 produce that research. 19 MR. COULSON: I think the real issue is if you 20 don't organize it into increments, there's not as big an 21 incentive to show any size because someone is just pennies 22 ahead of you for 100 shares, and marketplaces, there is a 23 slight view that marketplace prices are set by people who 24 make offers. They make a bid. They make an offer. That's 25 not really true. 0054 1 Market prices are set by people who take offers and 2 hit bids. There are lots of people who have houses for sale 3 and get an offer that's ludicrous. 4 What you want to do is incentivize the display of 5 liquidity and you want to make for that security the two -- 6 for someone to move ahead, that they have to move to the next 7 price point. It shouldn't be an artificial price point. 8 That shouldn't be the idea of increments. It should just be 9 the next price point. 10 There are community banks that are actively traded 11 on our market, trade 30 times a day. There is almost 30,000 12 seconds in a trading deck. 13 There is a lot of time in between those price 14 points. It's pretty easy to go through and say what are the 15 price points of these classes of securities. 16 There is a great paper by AMF, which is probably 17 the smartest thing I've seen on this and explains it better 18 than I ever could. It comes from the French, so it's a 19 little complicated. We probably use a little brutal 20 simplicity in America. 21 The idea is we should have price points based on 22 price and trading velocity and number of trades in a deck. 23 That organizes it. It brings together the structure, and 24 that way, the next price is something that has meaning. It 25 has economic difference. 0055 1 For something that trades 10,000 times a day, that 2 next price is very tiny, that next increment, but we stopped 3 organizing it for smaller companies. 4 MS. O'HARA: I want to pick up on comments Gregg 5 led us to and Cromwell has focused on as well. It is 6 certainly true there are stocks that have wide spreads, but 7 we also have to be really careful, there are lots of small 8 stocks now that trade with very, very small spreads. 9 For example, my colleague, Gideon Saar at Cornell, 10 we were chatting, and he gave me some numbers. I think they 11 are interesting. These are from NASDAQ. We are talking 12 about stocks that are now listed. 13 If you look at stocks that trade less than $250,000 14 a day in value, which are fairly illiquid traded, 469 of 15 those, but 211 of them have spreads that are less than a 16 nickel. 17 If you now start putting nickel spreads on these 18 stocks, you are essentially by definition just raising the 19 transactions costs for everybody else relative to where they 20 are now. 21 If you say let's restrict it to stocks that trade 22 $100,000 a day, which really is not a whole lot, there are 23 257 of them and 108 of them have spreads less than a nickel. 24 Even going down and saying what if they have less 25 than 100 in market cap, you can have 95 of them that have 0056 1 spreads less than a nickel. 2 I think what we are hearing is a Tale of Two Cities 3 almost here. It is true there are stocks that have high 4 spreads, but I think one of the things that is difficult and 5 certainly things that we as researchers find is it's not as 6 straightforward as it looks. 7 You can't just say look at all low priced stocks, 8 they have terrible spreads because a lot of them don't, even 9 stocks that aren't trading as much as you might expect, some 10 of them have very low spreads. 11 As we begin to change these tick sizes, I think we 12 have to recognize the complexity of what we are doing here. 13 You are not just painting the hole in the wall. 14 You are painting the whole wall. There are all kinds of 15 problems that I think arise when we do that. 16 MR. COULSON: There are 600 community banks on our 17 platform which I think is a great example. They have 18 averaged -- the last I looked, the spreads were about 20 19 cents. 20 If your kind of average spread is 20 cents, you are 21 not really incentivized to show much size if someone can pop 22 ahead of you for 100 shares and make the spread 19 cents. 23 It should be a stock by stock basis. I think the 24 greatest thing is -- I'd love to have a pilot program run 25 across our market, and let companies opt out if they want, 0057 1 and that is I think we would see demand, and there would be a 2 great test on markets. 3 Markets should set what is the increment of quotes 4 and orders. Trades should take place any place. Actually, 5 for institutions, they should never trade on the bid and 6 offer. They should try to trade as close to the middle as 7 possible so they are not giving away information which way 8 the blocks are going. 9 We shouldn't be forcing institutions to have to 10 trade on bids and offers, but we should be setting those 11 increments and having the markets stack up. 12 We have had a trend. It used to be investors 13 traded with marketable orders, and because there was on the 14 offer that price. Now as markets have gotten slippery, you 15 have to have a limit order. The investor comes in and the 16 idea that a retail investor has market impact is slightly 17 ludicrous now. 18 MR. SOLOMON: If you are a retail investor, if you 19 ever asked your retail firm to give you a volume weighted 20 average price execution over the course of the day, they will 21 look at you funny. 22 I did that last week actually. I did something I 23 haven't done in a while, I bought a stock. It's hard for me 24 to buy a stock because of a lot of the compliance rules, but 25 I did it just to see what would happen. 0058 1 I put in an offer to buy a stock that was trading 2 at $13.50 by $13.51, and before the open, I put in to buy 3 4,000 shares of a stock that I knew had averaged around 4 200,000 to 250,000 shares a day. 5 Not inconsequential as a percentage of the daily 6 average trading volume, and I asked them to give me a volume 7 weighted average price between now and the end of the day and 8 then allocate it to three different accounts, which is 9 basically trader speak for how I used to trade my 10 portfolio's. 11 I want to buy over a period of time and I want to 12 allocate it down. 13 By the time the market opened, the stock was 14 trading at $13.66/$13.67. I called them back and I said do 15 me a favor, I was looking for the volume weighted average 16 price over the course of the day, I hope they're not walking 17 up the bid because they see a 4,000 share order on the desk. 18 They assured me that wasn't the case and they 19 called me back at 9:28 and said to me oh, we can't do volume 20 weighted average price. We need a limit order. I'm like 21 really. A limit order, what's that? That's code for I want 22 to rip your face off. 23 Fine. I gave them a limit order at $13.68. Don't 24 you know, the first print, $13.68. Three minutes later, 25 $13.47. Really. 0059 1 I called them on the phone. That $13.68 print 2 better not be my print, because if it is, I'll start an 3 arbitration just for laughs. It doesn't matter to me, it's 4 4,000 shares, it's 20 cents. It's not going to make or break 5 me. 6 When we are talking about issues, about protecting 7 retail investors, and there is a lot of discussion in here 8 around how decimalization impacts costs, and yes, it's true, 9 but there is a lot more things that go on in the market that 10 disadvantage retail investors than penny increments. 11 There is a lot that the SEC could and should be 12 doing to actually open up order types for retail so they can 13 trade like everybody else does. 14 I think that is a key thought, if there is a lot of 15 push back around running a pilot program, let's think about 16 other things. I think Professor O'Hara is right, we should 17 be thinking about other things like addressing maker/taker 18 rules, that just level the playing field for the average 19 investor so that you can begin to have a little bit more 20 price discovery, fundamental price discovery, rather than 21 casino type transactions. 22 Those are things I would say are far more 23 fundamental to investor protection than whether or not we are 24 trading in penny increments. 25 MR. CONROY: I would agree. I think it is a 0060 1 marketplace that has become more sophisticated and not all 2 participants in the marketplace have kept pace. 3 The typical retail investor is not buying 4,000 4 shares in a small or mid-cap stock. They are buying much 5 smaller share sizes and can take advantage or be advantaged 6 by the smaller tick increments. 7 I think the issue is making sure that the investors 8 who are participating in the marketplace, if they are not as 9 sophisticated as Jeff is, should become some, and tools being 10 provided to them. 11 MR. WOMACK: Gregg, I wanted to give you an example 12 about pre-Reg ATS, how we traded stocks and moved them 13 through. 14 Just say a stock was trading 10 by 10.25 with a 15 quarter point and it was a quote driven market. We can buy a 16 block of 100,000 shares at 10, mark it up with a quarter, 17 sell it in 1,000 share increments at the retail system or 18 call institutional accounts. 19 We could make as much as 2.5 percent. We could 20 turn around and print it in the middle up an eighth and make 21 an eighth of a point to Fidelity. 22 As soon as we went into the Reg ATS post-world 23 market, and when people progressively started to be able to 24 step in front of you for a penny, all of a sudden, we 25 couldn't count on an economic opportunity, so we couldn't 0061 1 commit that capital. 2 All the capital came off the desk for these small 3 cap stocks. 4 Brian, you made some points earlier. We do 5 analyses on the institutional marketplace. I will tell you 6 about 80 percent of the commission opportunity rests with 7 about 100 institutional investors in the United States. 8 Fidelity is at the top end of the curve, very close to the 9 top end of the curve. 10 The difference in terms of assets under management 11 top to bottom, the 50th account is about 1/20th the size of 12 the largest account, $50 billion versus $1 trillion, up at 13 the top end of the curve, Fidelity, BlackRock, T. Rowe Price. 14 If you look historically, going back 15 years ago, 15 you will find that all the largest accounts have cut as a 16 percentage of their assets under management their allocations 17 to small cap, micro-cap and nano-cap stocks. 18 If you look down the long tail of the distribution 19 curve to the 2,000th largest investor, you will find they 20 have increased their allocation to small cap, micro-cap. 21 That's the self directed market. 22 What has happened is you have gutted the economic 23 incentives for people to distribute securities so the breadth 24 and the quality of the distribution of Wall Street on IPOs 25 has declined. 0062 1 The after market distribution has declined because 2 there is no longer any economic incentives, and Wall Street 3 has reacted by concentrating its resources on a smaller 4 number of institutional accounts, it has closed down every 5 middle market institutional sales desk. 6 Every firm had them. We had one at Prudential 7 Securities. Gone. 8 Retail brokers used to actually cover smaller 9 institutional investors that were not so into the 10 institutional marketplace and they would be supported with 11 capital from the NASDAQ or the upstairs New York Stock 12 Exchange trading desk. 13 This has been a massive deterioration in 14 distribution fundamentally. What we see is IPO sizes have 15 gotten larger, but IPO success rates have declined steadily 16 over 15 years. 17 That's very scary to me. It's gone from about 75 18 percent the way we define them, 70 percent of all IPOs were a 19 success, going back into the early 1990s, to about 30 percent 20 today. 21 As stocks are breaking issue price because the 22 distribution support isn't there, they are more mature so 23 they should be better companies, why do you think we are not 24 getting more IPOs done? Because investors aren't making 25 money because the deals aren't being supported. 0063 1 MR. CONROY: I would say while that was all 2 occurring, marketplace has shifted from a net market to a 3 commission based market. In this market, brokers do have the 4 ability to charge more commissions per trade. 5 The economic forces of the buy side are preventing 6 them from doing so, as Jeff pointed out earlier. 7 I'm not quite sure why the industry should be 8 supportive of increasing tick sizes and going back to try to 9 relive a marketplace that is a spread market which no longer 10 exists. 11 MR. SOLOMON: I think it is becoming increasingly 12 difficult for the buy side to out perform indices, and I 13 think we hear that every day from clients, like Fidelity, 14 that it is really super hard now to provide incremental alpha 15 above the beta benchmarks. 16 The easiest way actually to out perform benchmarks 17 is to participate in new issue. 18 I'm not surprised to find retail investors and 19 investors at large choosing to invest in ETFs and lower cost 20 investment vehicles because they are seeing interestingly 21 enough that the money that they are paying to act as managers 22 isn't really worth the return they are getting over and above 23 the benchmark. 24 I would say we are actually on the same side here 25 in that regard. A healthy IPO market provides long term 0064 1 active managers with positive returns over the long haul. 2 I'm not interested in going back to Fidelity and 3 asking them for six cents. That's not the point. The point 4 here is we have a responsibility to protect the ecosystem 5 that allows companies to become public, and that means 6 driving liquidity back to smaller issue where the bulk of 7 these companies come, and decimalization is one very 8 important piece to that, and widening tick increments 9 incentivizes people to actually do fundamental price 10 discovery rather than electronic price discovery. They are 11 two very different things. 12 What has also gone on here is I think there is the 13 maker/taker dynamic that goes on in the market and that 14 provides for a significant amount of volume that has nothing 15 to do with price discovery. 16 It is just how fast can you get that execution done 17 so you can be the maker and get the rebate, and somebody else 18 is the taker. 19 It is a zero sum game, by the way, between makers 20 and takers, and the people taking the spread out of the 21 middle are the people that are running alternative trading 22 venues. 23 They are not really interested -- they are 24 interested in order matching at this point. There is really 25 no price discovery in order matching. It is just I want to 0065 1 be the biggest. I want to have the most flow. I want to 2 have the most connectivity because I can make the penny 3 increment between the maker and the taker. 4 That kind of behavior, I'm not saying it's the root 5 of all evil, but it does need to be tailored and dialed back 6 a little bit when you're talking about small companies where 7 fundamental price discovery is so important to creating 8 functioning after markets. 9 COMMISSIONER GALLAGHER: I have a question for 10 David. This conversation is fantastic, by the way. Self 11 sustaining, easy for the moderators here. 12 I think David hit on something that I've been 13 thinking a lot about and talking about publicly, which is the 14 fear at least I have on the Commission, that we are servicing 15 a symptom and not the illness. 16 Decimalization could be simply a symptom, which is 17 the point I was getting at in my opening remarks. 18 I'm fascinated by this discussion. The order 19 handling rules. People don't talk about the order handling 20 rules. I've been talking about -- to Cromwell's point -- the 21 1975 Act amendments forward. 22 It's about time to question these things. If not 23 change them, at least severely question them before we start 24 treating more and more symptoms and making things more and 25 more complicated. 0066 1 We have a pretty complicated carpet we have woven 2 here on market structure. 3 Specifically as to your example about pre-order 4 handling rules, how a big block would have been treated, do 5 you have a recommendation in that space? 6 Should there be a carve out within the order 7 handling rules for certain sized companies? Is there 8 something else we should be thinking about in that vein? 9 MR. WEILD: The genesis of the tick size idea was 10 at a dinner a year and a half ago with a bunch of market 11 professionals, former Stock Exchange executives, and sell 12 side and buy side folks. 13 People just thought it would be easy to implement. 14 Part of it is if you were given a clean piece of paper and 15 you wanted to redesign the entire market, what would you do. 16 I'm a big believer that you shouldn't get in front 17 of your customer. The flip side is when you have to reflect 18 every single order into the market and then you start to 19 proliferate the number of price points, the real economic 20 incentives and the ability to manage risk that traders used 21 on trading desks was a function of the reliability of the 22 spreads. 23 To Maureen's point, even if something is at a 24 nickel or 20 cents, the fact is it's illusory. I can't bank 25 that spread. I'm not going to commit capital. 0067 1 I think if you can create price increments which 2 are larger -- as Cromwell has pointed out, they sort of 3 naturally -- remember, the penny tick is not natural. We 4 ordained it. We made a determination. 5 If you go back, this goes back to the discussion 6 around the order handling rules and Reg ATS. When those were 7 being implemented, if you go back to the record, the SEC 8 decided to do it anyway, they had received quite a bit of 9 feedback from the Street that it was going to harm capital 10 formation. They believed otherwise. 11 All we are talking about is adequate economic 12 incentives for small cap companies. You don't need it for 13 large cap which trade, as the academic's would say, these are 14 symmetrical order book marks, with lots of buyers offsetting 15 lots of sellers, naturally visible, on everybody's radar 16 screen, they benefit from the network effect, but small 17 company stocks need to be sold. They need to be marketed. 18 If there is no economic incentive to do that, a 19 $250 billion market value company is 2,500 times larger than 20 a $100 million market value company. 21 Yet, the basic economic model that we have created 22 is the same for the one that is 2,500 times larger than the 23 little one. On its face, that just doesn't make any sense, 24 if you're trying to optimize markets and create balance in 25 the system. 0068 1 I think we have to be very careful of ideologies 2 that favor low cost investment when in point of fact, as 3 Warren Buffett has said, you know, it doesn't bother him to 4 pay more on a transaction because he's a long term investor. 5 What we are catering to is trading. As a result, 6 we have seen more and more coverage of the Street's shift 7 from investors to the traders, because there is where the 8 money is, which is exactly the people from a balance 9 standpoint, the investors, with some exceptions, that issuers 10 don't need to reach. 11 One of the unintended consequences is if you look 12 at the client access business, non-deal road shows, Wall 13 Street is wasting the time of issuers, running them around in 14 front of institutional investors who are not natural buyers. 15 Many times I see this with our clients. They are 16 smaller cap, they are in front of somebody who can't possibly 17 own a $500 million equity market value name. 18 Why? Because they're an expert in a vertical, 19 providing information about their competitors in that 20 industry to an institutional investor that is the client of 21 Wall Street. 22 There are a lot of things that have effectively 23 shifted all sorts of work onto the backs of issuers. Issuers 24 are woefully unprepared, and as a consequence, they are 25 floundering as public company executives because they don't 0069 1 come to it with the knowledge base that the panelists do 2 here. They are ill equipped. 3 I think we just need to get adequate economic 4 incentives, not to gouge people. Time and price priority, 5 very important. As an investor, in Jeff's example, if he's 6 first at a price point, if he's paying a nickel or dime, 7 whatever the case might be, he should be able to get up, if 8 he's first in line, he should get the execution, and that 9 instills enormous amounts of confidence to retail investors 10 because it creates a sense of fairness. 11 Right now, this cat and mouse stuff that goes on, 12 you know, where people are putting in trades for retail 13 investors, undermines investor confidence, including my 14 friend, Bob Aber, who used to be general counsel of NASDAQ, 15 who is an active trader. 16 He always complains about the fact that as an 17 active investor, when he puts his little orders in, somebody 18 always automatically pulls him out, and then the order in 19 front of him disappears. Bang. Bang. Bang. 20 I think we can do a lot by just bringing it back to 21 the center a little bit and making sure there is an effective 22 seat at the table for the issuers that need to have a 23 reasonable experience as a public company and the large 24 investors, but also to understand that distribution, there is 25 an enormous amount of small cap distribution in this country 0070 1 that is not being accessed because nobody can afford to cover 2 it any longer. 3 MR. COULSON: In some ways, we're not a very big 4 IPO market. Actually, when David was Vice Chairman of 5 NASDAQ, they did become a public trading OTC first, and then 6 they did a secondary on NASDAQ. 7 It's really about creating the quality of market 8 long term. We have two separate subjects. If we have a 9 quality of market, and our markets, we are a significant 10 source of Exchange upgrades, about 40 last year when to 11 NASDAQ and NYSE, and bringing companies in, and getting that 12 conveyor belt. 13 The JOBS Act, what has the JOBS Act done? I ask 14 when is the JOBS Act going to be done because in the smaller 15 company financing space, it's all Reg D. We're not done on 16 that. We have not even started on Reg A, 144(a). 17 Those are the things that are going to go create 18 equity that can then go be tradeable, and equity is worth 19 more if it's tradeable. If it's only a tiny amount and the 20 investor experience is not good -- investors want to be able 21 to trade. 22 Just to address, I do have a disagreement about the 23 idea that the person who makes the offer should be first in 24 line. 25 When I go buy an iPhone and I show up at the store 0071 1 with an ad of an iPhone offering $25 cheaper, I don't go to 2 the store guy and he says okay, I'll match that price, I 3 don't say oh, no, I want the phone in the ad. Can you go to 4 California and get that phone and that guy who made the 5 offer. 6 This is how competition -- the taker, the consumer 7 should be the one with the choice, they should get the best 8 price but they should choose based on the quality of service, 9 based on where they go, and that incentivizes more aggressive 10 off bids and offers. 11 That should stay in our markets. We really have a 12 dynamic market eco-structure where it's not one size fits 13 all. We have different types of brokers that fit large 14 institutions, liquid, and that is incredible. 15 Sadly, they don't cross every trade Fidelity has. 16 In fact, the majority of institutional trading flow is now 17 being broken up into tiny retail sized bits and filled by a 18 liquidity provider. That's a little bit of a dark secret 19 that liquidity providers are still very important in this 20 agency market. 21 It is not to create a disadvantage for dealers but 22 for us to go back and have fair competition between brokers 23 and dealers. The Doctrine of the Church has been read that 24 we should prefer agency brokerage. Being a dealer is icky. 25 Supplying liquidity puts you in a questionable position, even 0072 1 if you are supplying liquidity at the best price. 2 I think we should bring it back and just have it be 3 fair and organize the market so it works for agency brokers, 4 it works for dealers. 5 Our marketplace, we have agency brokers, we have 6 NYSE's ArcaEdge ECN, who does about eight percent of the 7 share volume, and we have market makers. They all interact 8 and compete and investors get to choose. 9 That's what we want. We want equal competition, 10 not an one sided push, you should become an agency broker, 11 and then the institutions will pound you on per share 12 trading. 13 COMMISSIONER GALLAGHER: Thanks, Cromwell. I just 14 want to pick up before we end this panel here on a last point 15 David made on research, that we aren't going to have 16 liquidity. 17 I want to just challenge that point because this 18 came up in our Small Business Committee meeting last week, 19 why are they recommending to us to do this pilot on tick 20 sizes. 21 There was actually some debate. Was it research or 22 was it simply just liquidity, was it the ability to make the 23 spread. 24 I'd like to just throw that back. I don't know, 25 Kent, if you have a thought on that. 0073 1 MR. WOMACK: I was actually thinking about that 2 just this second. It strikes me that these things are very 3 separable, and what we have been talking about for the last 4 10 or 15 minutes here is improving the liquidity environment. 5 I just don't -- I'm much more skeptical that we can 6 put a gun to anybody's head and say produce more research. I 7 don't really see what the mechanism is for driving a larger 8 amount of research, will in fact the small brokerage firms 9 with that liquidity provision throw off some of it for 10 research. 11 How are we going to know? I don't know how we are 12 going to test that very easily in any reasonable length of 13 time. 14 MR. SOLOMON: I would just say from my standpoint 15 as somebody who can make that decision every day, we are 16 taking companies, small companies, on the road and doing non- 17 deal road shows all the time. 18 We do take them in to see large buyers who we know 19 can necessarily accumulate a position because there is a lot 20 of transmission of knowledge about what is happening in that 21 particular industry, but we are also taking them out there 22 because we do need to cultivate a very personal relationship 23 between those management teams and the ultimate end buyers, 24 and there are a number of them out there. 25 What I would say is a quintessential part of being 0074 1 able to get in the door to have those meetings is published 2 research. 3 We have turned our business model around and said 4 corporate access is not just an administrative back water 5 where you are filling meeting calendars between companies and 6 potential portfolio managers. It is actually a driver of 7 value and connectivity between the end buyer, the investor, 8 and the management team. 9 You want that conversation to happen as many times 10 as you possibly can so the investor is better informed and 11 the management team actually appreciates what the owners in 12 the company want him or her to do in that position. 13 Writing research gets you the ability to open the 14 door at the buy side. If you don't write research, it's just 15 hard to get in the door because they are like why are we 16 talking about this. You don't write on it and I can't buy 17 it, why are we talking about it. 18 I start with my research report. Here's our view 19 on this company. We are bringing management into Boston and 20 we would like them to see you and here's the reasons why, in 21 three bullet points on the front page. Can we see you for 15 22 minutes. 23 That's why research is important and it drives the 24 back end of the business which ultimately leads to capital 25 raising and ultimately leads to IPOs and job creation. 0075 1 That's really the connectivity here. 2 COMMISSIONER GALLAGHER: Brian, on that point, on 3 the buy side, the utility of the research, because that is an 4 interesting point. 5 MR. CONROY: As I stated earlier, we, too, are 6 skeptical that somehow decimalization is linked to the 7 economics of research, especially in a time when we are no 8 longer a net market, we are a commission based market. 9 Brokers do now have the flexibility to raise commissions. 10 We think the real issue for looking at IPOs is 20 11 years ago the average underwriting spread for a small or mid- 12 cap company was 500 to 700 basis points. Today, it is 13 exactly the same. 14 If you look at how market structure has improved 15 liquidity, quote size, reduced volatility across the board, 16 what stands out unique in our market is the fact that the 17 underwriting spreads and the lack of transparency in equity 18 issuance has remained where it has been over the last 20 19 years. 20 I think one of the reasons the buy side is 21 skeptical about raising spreads to subsidize research would 22 be where is that 500 to 700 basis points going now. 23 MR. KUPOR: I agree with Jeff, and David, you may 24 agree or not, but I think even if you didn't get research, if 25 you got better trading liquidity in these stocks, to David's 0076 1 point, making sure it didn't just take a penny increment to 2 step ahead of price and you incented people to actually have 3 larger trading volumes, I think that would be a major push 4 forward here in terms of helping the trading liquidity of 5 these stocks and then enabling them to go public but also to 6 have a robust trading environment that allows them to do 7 things like secondary offering's to raise capital in the 8 market. 9 I take Jeffrey at his word that providing financial 10 incentive would translate into research. I think even in the 11 absence of that, if you solved the liquidity problem 12 fundamentally through the tick sizes, that still has a 13 tremendous impact on the attractiveness of the capital 14 markets for small cap issuers. 15 MR. GREEN: Yes, stated just slightly differently, 16 liquidity is the major point. Research, in my view, is a 17 tool to increase liquidity. Therefore, in stimulating that 18 and all the rest of it is -- I don't want to say it's a side 19 show, but it is not the main event. The main event is 20 liquidity itself. 21 I just can't help but say since I have the mike for 22 maybe the last time here, just a couple of thoughts as you 23 design a pilot program moving forward, there is a tendency in 24 the law and regulation to design these like engineering is 25 designed, and this is relating back to where I live in 0077 1 Silicon Valley. 2 Therefore, very discernable, a pragmatic type of 3 approach and the rest. 4 I would consider these pilot programs to be more 5 like biology. We don't really know what the results are 6 going to be. We are going to try certain things. We are 7 going to try to get end points. 8 What biologists do is try to get end points from 9 their experiments and figure out what that is. For me, it is 10 liquidity. That is the bottom line. 11 What does liquidity really mean? I divide it into 12 three different areas. One is what does it do to pricing, 13 what does it do to volume, and then finally, is it 14 encouraging investors as opposed to trading. Does it mean 15 trading is unimportant, in fact, it's absolutely critical. 16 Is it emphasizing more investment in small cap stocks. 17 MR. WOMACK: At the same time, I think factoring in 18 the incremental costs to the investor in moving this tick 19 size, we are completely in a cost/benefit tradeoff situation 20 here. 21 MS. O'HARA: I still think we need focus, as you 22 pointed out, if the focus is on liquidity as opposed to 23 research, then I think we need to use the data that we have 24 to help us here. 25 The data that we have actually suggests that the 0078 1 transactions costs, the trading costs of trading small stocks 2 has actually been falling dramatically. 3 ITT Research, Jackman King, put out a very nice 4 piece showing, for example, that trading costs for small 5 stocks decreased at a faster rate than it did for large cap 6 stocks from 2008 to 2012. 7 I really like the way you focused our attention, 8 that there are two issues here. There is a liquidity issue 9 and there is a research issue. 10 I think one of the interesting challenges that the 11 SEC will have to think about is what is going to be the point 12 of your pilot. I think there can be different goals and 13 different directions that this can go. 14 COMMISSIONER GALLAGHER: Just to be absolutely 15 clear, despite the nodding of heads over here, I don't think 16 we should assume a pilot. I think the Commission has a lot 17 of hard decisions to make here as to whether this makes sense 18 to proceed, to the discussion we had with David, are there 19 other things we should be thinking about. 20 MR. WEILD: Personally, I'd just do it. That's me. 21 (Laughter.) 22 MR. WEILD: They did it back with the order 23 handling rules and Reg ATS. I think it's about visibility 24 and creating and sustaining visibility and driving down cost 25 of capital for small cap stocks at the end of the day. They 0079 1 have to have competitive costs of capital. 2 I think research -- there is a great paper that was 3 just put out by Tim Keating at Keating Capital that just 4 charts the lack of equity research coverage. That's helpful 5 to the process, but salesmen, sales traders, they make phone 6 calls. They create the visibility and the liquidity. The 7 capital committer's facilitate that as much as the research 8 analyst. 9 A lot of that research is even worse. Most of that 10 small cap research isn't being actively marketed anyway 11 because there is no sales incentive to do it. It's a place 12 holder to keep the client happy because they may do an 13 investment banking transaction at a later date. 14 It's very misleading. I would say you just have to 15 get people focusing on these stocks. You have to get it to 16 the point where Fidelity will reverse the trend they have had 17 for the last 15 years, which is cutting their allocations as 18 a percentage of their portfolio to small, micro and nano-cap 19 stocks, and actually starting to raise it because they can 20 get a bid. There is liquidity for them to put size to work. 21 If we start doing that, we will get America back 22 into business. We will bring back $25 million IPOs again. 23 We will start to create jobs and reverse this 15 year 24 decline. 25 MR. COULSON: The electronic market makers, the 0080 1 highly automated ones, I have not heard concerns of 2 profitability in trading the small cap stocks on NASDAQ 3 because they are great at wiggling around. 4 Every study is a series of statistics, what's the 5 spread, but what is the effective spread, what is the wiggle. 6 Where is the satisfaction. Are we designing -- in our debate 7 with FINRA over new tier sizes in our marketplace, the real 8 debate was who was more important, the 500 share investor of 9 a $6 stock or the 500 or the 100 share investor. 10 I think we need to put it in a little bit of the 11 middle, that there is enough liquidity that there is not 12 these 100 share spreads, that it starts to stack up. 13 On the other point, we have been building an 14 ecosystem at the top level of our marketplace with OTC, you 15 actually need to be sponsored. We have used attorneys which 16 do a passive quality check, but we also have investment 17 banks. 18 What we have seen is research is one side, but the 19 real value is having an institutional firm take you on the 20 road, take you to conferences, be there as a call to trade 21 the stock when a block comes up. It's that sponsorship. 22 In Europe, we have the nomad corporate broker 23 bringing in that institutional connection to the markets for 24 our company, because we are creating these orphans. 25 If you see a lot of research, it can be good and 0081 1 bad, but you really need corporate coverage, you need someone 2 on the institutional desk who knows the company, knows the 3 story. 4 In our market space and in the bottom half of 5 NASDAQ, people don't do secondaries and capital raises. They 6 are doing private placements. They are in the pipes market. 7 They are going to a few sophisticated investors. 8 What you really need if you are in the capital 9 raising mode as a public company is to be out there talking 10 to these investors before, because when you do a capital 11 raise as a pipe, you get such a better capital raise if you 12 have knowledgeable investors. If they are fundamental 13 investors rather than someone who is trying to just arbitrage 14 it into the market. 15 That part of connecting, we have a problem if it 16 just becomes all these structural arbitragers for the small 17 company capital raises doing a pipe, rather than we have 18 institutions that are fundamental investors, no companies 19 that corporate brokers have taken them out to. That is part 20 of the connection of getting these firms back trading these 21 stocks. 22 The high speed firms, they are fine. Make it 23 whatever you want. Set the rules. They have great 24 technology bases. The big five/ten of those will just move 25 to trade whatever type of security. 0082 1 We want to bring back the ecosystem to the price 2 setting, the corporate financing, and it's about quality of 3 market experience. If we have quality markets, we will get 4 more publicly traded companies. 5 MR. WEILD: Chris Jacobs, who was the co-Chairman 6 along with Steven Graham of the SEC Advisory Committee on 7 Small and Emerging Companies that made the recommendation on 8 Friday to increase tick sizes, she took her stock chart back 9 when the order handling rules and Reg ATS were implemented 10 and just overlaid it, and found that even though she was 11 growing at 30 percent, compound annual growth rates dropped 12 off the cliff. 13 It might be helpful for people to take a look at -- 14 that's just one data point -- what happened around that 15 particular period. That was a shift in stock markets. 16 All the focus of the academic papers that I have 17 read have been around decimalization. 18 MR. BERMAN: Could I ask just a quick question? We 19 have heard a lot of numbers in terms of what is a small cap. 20 I know that from a Fund standpoint, from an investor, if I 21 invest in a small cap fund, that's generally $1 billion/$2 22 billion/$3 billion, but I've also heard $10 million and $25 23 million. 24 David, I think you made a good point there could be 25 a huge difference, factors of 100, in the way we treat 0083 1 companies. 2 Are folks thinking about small cap and emerging 3 markets the same way? Are we looking at $10 million 4 companies, $25 million companies, $100 million, or are we 5 still in the $1 billion and $2 billion companies? 6 Should all those be treated the same or do we need 7 to have different regimes for each of those? 8 MR. SOLOMON: I have to say from our standpoint, I 9 view everything $1 billion and south is small, small or 10 micro. I don't really draw a distinction between small and 11 micro. I feel like they are all the same. 12 I'm not really for having all these different 13 regimes for different sizes. I just think fewer choices is 14 better in this particular case. 15 Once you get to $1 billion, there are enough people 16 to pay attention and that's not really an issue. 17 If you look at where research coverage has as a 18 proxy for where liquidity is, you will see if you're $1 19 billion and above, there are plenty of people that cover you, 20 and you don't really have issues in terms of getting access 21 to information in the market, and the liquidity improves 22 significantly at $1 billion and north. 23 I think from my standpoint, I would say $1 billion 24 and south is really where we are focused. 25 Just to kind of give you guys an idea, biotech has 0084 1 been around for two decades or so, a little more than two 2 decades maybe. I just want to point out that there are still 3 only about 20 companies that have market caps north of $1 4 billion. 5 All of the financing that has gone on, all the drug 6 discovery and all the creation of -- what I would say is in a 7 very robust capital raising part of the economy, we still 8 have the vast majority of those companies sub-$1 billion 9 period. 10 The expense associated with researching those 11 companies is not cheap. It's expensive. They're 12 complicated. 13 We can boil it down, rather than just look at the 14 numbers, look at specific industry types. 15 Just in terms of the micro cap, Microsoft only 16 raised like $56 million when it went public, something like 17 that. There were 114 investment banks and the spreads were 18 still at five or six percent. 19 When you actually think about that and you think 20 about the fact that last year Linkedin went public and they 21 raised $350 million and there were four banks on the cover. 22 Ask yourself the question, how does that happen. 23 If spreads stayed constant, why did Linkedin -- how could 24 they only have four banks, and the answer is those banks need 25 that economics from the IPO to justify research coverage. 0085 1 Ask any large organization, you look at the way the 2 covers read today, the large banks are commanding 70 and 80 3 percent of the economics in the IPO because they know they 4 can't make any money in the after market. 5 I don't particularly think that's healthy for a lot 6 of companies that don't raise to the level of the attention 7 of a Goldman Sachs or J.P. Morgan. It's not healthy. It's 8 not a healthy environment for them. 9 We should have a number of different investment 10 banks. The way to create that is to create economics that 11 are balanced and reasonable in the after market so people can 12 make money either on a net basis or a commission basis. We 13 should have that. It should be a stated objective of this 14 body to try to foster that. 15 MR. WEILD: Gregg, I can send it to you. We use $2 16 million as small cap and small micro cap, sub-$500 million 17 nano-cap, sub-$100 million. 18 What troubles me,and I would agree with everything 19 that Jeff said, except I'd take it even higher because we 20 have a sample set that we look at, IPO performance trends 21 over time for the large IPO, defined as larger than $500 22 million. This is initial float size. 23 That has been declining and converging on success 24 rates for smaller IPOs as time has gone on. 25 I had conversations with -- he's now over at 0086 1 General Atlantic -- David Topp, who used to be the head of 2 Global Equity Capital Markets at J.P. Morgan. His view was 3 you need at least half a billion dollars in float to be able 4 to sustain organically after market interest in the current 5 market structure. 6 MR. NALLENGARA: We are short on time. Cromwell? 7 MR. COULSON: $500 million. Below $500 million, 8 your shareholder base is going to be all stock pickers for 9 the most part. 10 MR. NALLENGARA: Josh? 11 MR. GREEN: Just very quickly from the venture 12 capital perspective. We really look at 300 to $400 million 13 thresholds because of the way the venture capital investment 14 model works to be the logical floor for a market cap to go 15 public. That is what our industry, our companies are focused 16 on. Anything below that would be a micro cap. 17 We also have to account for the fact that after 18 market trading, it may take those numbers lower than those 19 market cap numbers, and we still need to support those 20 companies because they can dip and then come right back up as 21 well. 22 MR. NALLENGARA: I hate to cut this conversation 23 off because it has been fascinating and extremely helpful for 24 us. 25 As I scanned through the questions, I think we have 0087 1 covered the topics. 2 I want to thank the panelists for participating. 3 Hopefully, they will stay around, if anyone out there wants 4 to corner them and pepper them with more questions, I am sure 5 they won't like that I'm going to say they would welcome 6 that. 7 We are going to take ten minutes and then we will 8 be right back for our second panel. Thank you. 9 (Brief recess.) 10 PANEL 2: EVALUATING CONCERNS RELATING TO TICK SIZE FOR 11 SECURITIES MARKET GENERALLY 12 MR. BURNS: We are going to try to get started 13 again. Thanks, everyone. This is a group that doesn't get 14 to see each other very often, I take it. 15 MR. NALLENGARA: We are going to start now so if I 16 could ask everybody to take a seat. If you want to chat, 17 there is a beautiful atrium there, you can do it there as 18 well. 19 MR. BURNS: Suitable for capital formation outside. 20 Thank you. Thanks, Lona. He speaks with that authoritative 21 Canadian voice that demands respect. 22 We are delighted to be starting on our second 23 panel. This group has a rather hard act to follow. That was 24 a very brisk dialogue that opened up a number of lines of 25 inquiry. We know we have brought together a really good 0088 1 thoughtful group of individuals who will carry the dialogue 2 forward. 3 To set sort of an initial theme, we can begin to 4 delve more deeply into the question of fostering the 5 ecosystem for the secondary market. 6 It could have been noted at the outset, this entire 7 subject has taken up a notable amount of interest since last 8 year. I think I counted Senate and House hearings, there 9 were three in July, September, December, where 10 representatives from a number of the institutions who are 11 here today talked about re-thinking the one size fits all 12 approach that comes with decimalization. 13 I am sure we are going to invite the panelists to 14 introduce themselves. We will pose a question or two. We do 15 have a number of fruitful areas that we can follow up on, 16 including the premise of a tick size as the issue relative to 17 things like other rules that govern our markets, Reg ATS and 18 order handling, which came up. 19 Thinking about how we benchmark and measure the 20 impact if we were to pursue a tick size pilot, the standards 21 of increased liquidity, the fostering of research. A number 22 of fruitful areas we can explore as the panel goes on. 23 I will turn to this group and maybe moving from the 24 left side of the table to the right, I'll invite you to 25 introduce yourselves. 0089 1 MR. REED: Thank you for that. My name is Adam 2 Reed. I am Associate Professor of Finance at UNC, the Julian 3 Price Scholar there. 4 My research has been on a wide range of trading 5 issues. I've done some work on mutual funds, some work on 6 market participation. Sort of a focus of my work has been on 7 short selling and in particular the impact of the regulatory 8 environment and the information environment with respect to 9 short selling. 10 Also, I've done some work on IPOs, both from the 11 perspective of short selling, its involvement in IPOs, and 12 also from the perspective of what makes a firm choose to go 13 public in the first place. 14 MR. JIGANTI: Good morning. I first would like to 15 thank the SEC Commissioners and staff for their continued 16 hard work creating and maintaining the most robust financial 17 markets in the world. 18 My name is Paul Jiganti and I'm a Managing Director 19 at TD Ameritrade in charge of market structure and customer 20 advocacy. 21 I started trading in 1985 on the Floor of the CBOE 22 as an independent trader. In 1992, I moved to Susquehanna 23 and traded for 16 years, doing just about every type of 24 trading that Susquehanna did during that period of time. 25 I was involved at the CBOE as a member or the chair 0090 1 of many committees. In 2007/2008, I was on the Board of 2 Directors. 3 Since then, I created a dispersion trading model at 4 Geneva Partners, left to work as a consultant at NASDAQ, 5 introducing a new index product line. 6 In July 2012, I joined TD Ameritrade in my current 7 position. 8 I applaud the SEC for continuing to explore ways to 9 make the equity markets better for all investors. Retail 10 customers have never had a better experience in the markets. 11 They have the lowest commissions ever, narrower spreads, 12 firms with an array of routing destinations to choose from 13 for best execution, and protections, all in under a second in 14 most instances. 15 TD Ameritrade clients are filled at the quoted 16 price or better 96 percent of the time for the entire route 17 of the order. My fear is any change to this fragile 18 ecosystem could be to the detriment of the small investor. 19 Thank you. 20 MR. ISAACSON: Good morning. My name is Chris 21 Isaacson. I'm the Chief Operating Officer of BATS Global 22 Markets, the third largest Stock Exchange here in the U.S. 23 and the number one equity market center in Europe. I'm also 24 a founding employee of BATS back in 2005. 25 Good morning and thank you for the opportunity to 0091 1 participate in today's Roundtable. I think it is a very, 2 very important topic that is being discussed. 3 Decimalization has led to tighter bid/offer spreads 4 resulting in reducing costs to investors that Paul just 5 talked about, but as we heard in the first panel, one size 6 fits all approach to tick sizes may have unintended 7 consequences of impeding the capital raising ability of small 8 and mid-cap companies. 9 Tick size regimes can also increase the cost of 10 trading in large, mid and small cap companies alike, 11 depending upon a company's liquidity profile and share price. 12 As a result, in Europe, for instance, the industry 13 developed a dynamic tick size regime designed to find an 14 optimal balance between incenting liquidity at the inside 15 while minimizing the maximum spread investors would pay to 16 access that liquidity. 17 We are proud to have played a leading role in that 18 initiative, which has been successfully adopted across 19 Europe. 20 We are encouraged that in hosting this Roundtable, 21 the Commission is examining not only concerns regarding 22 potentially larger minimum tick sizes for small and mid-cap 23 companies, but also broader questions about the current tick 24 size regime. 25 We believe that a wider tick size may make sense 0092 1 for new issuers or illiquid securities, small and mid-cap 2 companies, as well as certain higher priced but still very 3 liquid securities, where an one cent price change may be 4 immaterial to the notional cost of execution, but where 5 investors are forced to access multiple pools of liquidity to 6 complete a single trade. 7 We also believe that a narrower tick size is 8 appropriate for certain highly liquid lower priced securities 9 where evidence suggests that an one cent minimum tick size 10 may be constraining price formation. 11 We also believe that as part of any pilot program 12 analyzing effects of changes in the minimum tick size the 13 Commission should consider allowing trading centers to 14 display locked markets in certain highly liquid securities. 15 Because of the ban on locked markets, most 16 Exchanges have implemented their systems such that better 17 priced orders are frequently displayed to reflect the 18 compliant yet less aggressive price level. 19 This is known in the industry as "price sliding" 20 and is necessary under Reg NMS to prevent locking the 21 markets. Even though these orders would have improved the 22 national next bid or offer. 23 This in turn may be inhibiting the public price 24 discovery process and depriving investors of the benefits of 25 trading in a zero spread market. 0093 1 BATS believes the Commission should implement a 12 2 month pilot program with minimum tick sizes based upon 3 considerations of the securities' quoted effective spread 4 leeway, which is the number of ticks inside the average 5 quoted spread, share price, and liquidity. 6 Details of our proposal can be found in our written 7 submission to the Commission, but in short, we are proposing 8 a pilot program that would allow a five cent minimum tick 9 size and an one cent minimum trading increment where either 10 the average daily volume in a security is less than 100,000 11 shares or the security is priced greater than $100 per share, 12 and the average effective spread that investors are realizing 13 as greater than five cents. 14 Secondly, a half cent minimum tick size in certain 15 very liquid securities with an effective spread of less than 16 one cent, and an average quoted size at the national best bid 17 or offer of 10,000 shares. 18 Finally, the display of locked markets in very 19 liquid securities with narrow effective spreads. 20 We believe this would be the most effective manner 21 for the Commission and the industry to determine the 22 potential benefits of tick sizes to companies and the 23 investing public. 24 Thanks again for inviting me to participate. 25 MS. O'HARA: I am Maureen O'Hara. I am the Finance 0094 1 Professor at Cornell University. I do research on market 2 micro structure. I also am Chairman of the Board of ITG. We 3 are an agency/broker-dealer firm that does a lot of work on 4 transaction cost analysis among other things. 5 MR. HEALY: Hi. My name is Pat Healy. I have an 6 MBA and CPA, former CFO in the banking industry, and for the 7 last 17 years or so, I have run a company called Issuer 8 Advisory Group. As a result of that, I bring to you today an 9 unique perspective, and that is of issuers. 10 I can tell you without getting into a long 11 description here that issuers generally like decimalization a 12 lot. However, there have been some unintended consequences, 13 and the general consensus of the issuer community is their 14 shares, which are traded on these markets, largely become 15 chips in a casino. 16 They would like to see some changes, and I think 17 they would wholeheartedly favor some move toward the topic of 18 discussion here today. Thank you. 19 MR. HATHEWAY: Thank you. I'm Frank Hatheway. I'm 20 the Chief Economist to NASDAQ Stock Market. 21 I would like to thank the staff for including me in 22 today's proceedings, and we will talk about erasing the tape 23 in a minute. 24 Prior to joining NASDAQ, I was a finance professor 25 at Penn State University. Prior to that, I was an options 0095 1 trader in the old days. 2 In my role at NASDAQ, I talk to regulators, members 3 of the trading community, listed companies, large and small, 4 about issues of concern to them, market structure is a 5 particular focus of what I spend my time on and my group. 6 I arrived at NASDAQ shortly after decimalization as 7 we were finishing our report on decimalization. To summarize 8 some of the less well covered parts of that report, small 9 stocks did not particularly benefit, a quarter of the stocks 10 listed on NASDAQ, most of them inactive, actually experienced 11 an increase in spreads with decimalization. 12 Since then, spreads have tightened for all 13 securities, as Maureen said in her review of the trading cost 14 analysis work ITG has done, particularly for larger stocks, 15 many of which are tick constrained, trading at a penny 16 increment. 17 For the less active stocks, the benefits have been 18 less obvious and probably less dramatic. 19 When you think about liquidity in smaller 20 companies, a pilot to see if perhaps wider quote increment 21 might improve the environment for these companies, I think 22 it's a good idea. 23 More generally, tick size is just one element of 24 market structure that is overdue for assessment. NASDAQ has 25 reported multiple Commission proposals in improving equity 0096 1 market structure and moving us away from an one size fits all 2 approach in many, many aspects of how we run the markets. 3 The markets are continuing to evolve. Regulation 4 must be an iterative and incremental process. 5 I encourage more dialogue such as this one to 6 identify and evaluate alternatives which improve the market 7 for investors, listed companies, and the American public. 8 I look forward to your questions, and thank you 9 again. 10 MR. CONCANNON: I'm Chris Concannon, Partner and 11 Executive Vice President of Virtu Financial. We are an 12 electronic market maker. We make markets in several thousand 13 companies around the globe, here in the U.S., Canada, Europe 14 and Asia. 15 We also run a designated market maker on the Floor 16 of the NYSE where we employ actually human traders, and we 17 make markets in about 250 names which are by definition small 18 and micro cap stocks. 19 I think the most important point, some of us have 20 touched on the topic, decimalization of all the market 21 structure changes that we have witnessed over the last 20 22 years, decimalization has delivered more profits to investors 23 than any other market structure change. 24 In the context of this Roundtable and talking about 25 decimalization, it was an unbelievable success. Where we 0097 1 have failed our market is where we have treated it in the one 2 size fits all, and we will hear that phrase over and over 3 again. 4 I do think we need to at a Federal level customize 5 our market to the nature and characteristics of the 6 instruments that we are trading. That means trade increments 7 need to be adjusted. That means market structure features 8 need to be thought about. 9 We come ironically from a market structure that did 10 in fact have different market structures for different 11 symbols. If you think about the 1990s, despite all the 12 flaws, there was a dealer market in the OTC, and many 13 companies started on the dealer market and then they 14 graduated to the auction market, either on the American Stock 15 Exchange, and if they were very successful, they would 16 graduate to the NYSE. 17 We did in fact accidentally have a migrating market 18 structure that was dealing carefully with the symbols that 19 were listed on those markets, and today, listing standards 20 are currently across our markets which are very low. We have 21 some of the lowest listing standards we have ever had. 22 All companies just fall into the same market 23 structure that a Microsoft and Cisco exist. I think that is 24 ultimately the flaw that we are identifying on this panel, 25 not questioning the success of decimalization. 0098 1 Thank you for having me. 2 MR. CRONIN: Thanks for having me. I appreciate 3 the opportunity to speak here today. My name is Kevin 4 Cronin. I am from Invesco. Invesco is a global independent 5 asset management firm with assets under management of $687 6 billion. 7 Essentially, I am here to represent the interests 8 of long term investors, which I think all would agree are the 9 cornerstone of efficient and effective market structure and 10 capital formation. 11 Within the context of that, I will tell you that I 12 think generally we do believe there have been advances made 13 with respect to our investors' ultimate outcome as a result 14 of the changes in market structure. 15 I think some will take probably more credit than 16 they are due on this panel. I think actually if you were to 17 dis-aggregate our interest in things, in terms of transaction 18 costs, as we will get into invariably, spreads is not one of 19 them. 20 Narrowing spreads is an interesting academic 21 concept but in the construct of an institutional trade, when 22 you have 150 or 200 percent of the daily volume, spread 23 doesn't really factor into the equation. 24 We worry instead about things like posted 25 liquidity, not transient posted liquidity, it's really there. 0099 1 We think while there have been tremendous strides 2 made in market structure, there is clearly room for 3 improvement. We certainly applaud the Commission for 4 examining decimals. 5 I think as others have described, it certainly 6 sounds reasonable to take a look at how decimals have 7 interacted in the marketplace and the impact, intended and 8 other, that has been had as a consequence. 9 Our own theory is that certainly from a large cap 10 perspective, things are better, but it's far less clear to us 11 as others in the mid-cap and small cap space that the 12 benefits of electronic trading, algorithmic trading and other 13 things that we now take for granted have really brought the 14 level of sophistication, equality, efficiency to the mid and 15 small cap world that I think most of us would have hoped. 16 I thank you again. I would like to again take the 17 opportunity, I suppose I would be remiss, as the only 18 institutional investor on the panel, to say while we 19 certainly appreciate the investigation of decimals, there are 20 other issues that we would hope regulators would take a look 21 at in the not too distant future, certainly things like 22 maker/taker economics, internalization, the proliferation of 23 order types and the lack of a consistent regime around the 24 disclosure of those order types, and lastly, order 25 cancellation, something we probably all see a lot as we see 0100 1 flickers and that sort of thing. 2 As far as I can see, it doesn't add to investor 3 confidence and quality of markets. Certainly, to confusion 4 and at some point probably deleterious to investor 5 confidence. 6 Thank you again and I look forward to answering 7 your questions. 8 MR. BURNS: Thanks to all the panelists for that 9 introduction. Thanks to Frank for that moment in levity. 10 Another merger. We were all kind of wondering. 11 Before I turn it over to my colleagues in Trading 12 Markets and Risk Fin and Corp Fin for questions they might 13 ask, I think maybe just as a table setting exercise, why 14 don't we jump right into measures, liquidity, fostering 15 research. 16 The European models are going to be an interesting 17 topic. 18 Is there anything -- we began to talk about the 19 impact of decimalization in the last panel, is it a cause, is 20 it a symptom, is it a consequence or something coincident or 21 additive. 22 Anything you would like to augment, touch upon, 23 anything you think we didn't get into? Anything else? 24 MR. CONCANNON: I would just add -- first of all, 25 we are supportive of a pilot. I think we are not going to 0101 1 know -- there are no studies that will really give us a 2 conclusion around trade increments in the U.S. There are so 3 many different market structure features of our market versus 4 other markets. 5 I think they are wonderful guides. The AMS study 6 was actually very impressive. I think there are a number of 7 things that suggest we should do something. The one thing I 8 would suggest is in the context of a pilot program, I would 9 take a hard look at market maker obligations. 10 If we are going to change the trade increment, we 11 should take a hard look at market maker obligations, with the 12 interest of increasing those obligations, in exchange for an 13 increase in the potential profit as a result of the trade 14 increment. 15 MR. BURNS: I'm going to turn to my colleague, 16 Gregg Berman, from our new Office of Analytics and Research. 17 MR. BERMAN: I would like to ask all the panelists 18 their opinion on why the market hasn't solved this aspect or 19 this problem for itself. 20 Certainly, Commission rules are consistent with a 21 marketplace aim we would like to have, the two cent or five 22 cent tick, and competitive pressures may say if I do five 23 cent but someone else has an one cent, that's a problem. 24 We are hearing from everybody saying yes, this is a 25 problem. What's preventing the marketplace from saying 0102 1 actually we can have our own pilot, so the need for a 2 specific Commission action on this. 3 I'm wondering if folks can comment on that. 4 MR. HATHEWAY: I think you can get some insight on 5 that on what's happened even where we have a tick and it's 6 appropriate with a penny. We don't have uniformity because 7 there are very, very strong economic incentives to defect 8 from what everyone else has to do. 9 Everybody follows the rule. There are exceptions 10 to the rule. If you can exploit those exceptions, you do so. 11 It gives you an advantage. 12 In the European context or here in the United 13 States, if you have a voluntary approach or you have an 14 approach that is taken unilaterally, and we have tried some 15 with different market structures, it isn't necessarily going 16 to produce the outcome you desire, either experimentally, 17 speaking as an economist, or practically speaking as a 18 representative of NASDAQ, because there are always 19 opportunities for people to do what again with the academic 20 hat on we would call an epsilon better, just a tiny bit 21 inside what you are trying to do. 22 I think that is the main reason you haven't seen a 23 market solution, that this is a situation where you probably 24 have a legitimate market failure around coordination. 25 MR. CONCANNON: I would agree. I think the 0103 1 challenge -- we chose a path which is a competitive market 2 structure, and that means those Exchanges and those 3 alternatives are going to compete with one another. 4 The outcome of that is actually very helpful for 5 the economics of the end user, but it also delivers this 6 issue where sometimes the Federal Government has to say hey, 7 we all need to agree on these certain dynamics, so maybe it's 8 a national market system. 9 I think at the national market system level, they 10 could do it, but they do need a push, a fairly strong push to 11 decide on trade economics. 12 You can't have one Exchange have different quote 13 increments than other Exchanges. It just won't work. That 14 is what you had in Europe. They tried to have an industry 15 solution. It's been helpful. There clearly needs to be a 16 broader solution in Europe as well. 17 MR. ISAACSON: I will speak to the push in Europe 18 that we were part of. There was somewhat of a race to the 19 bottom regarding tick sizes, and we came together and said 20 this is not in the interest of anyone, we should agree upon 21 tick size tables, and while not completely uniform across 22 Europe, I think where we landed is a rational tick size 23 regime. 24 It always can be improved, but there is an ongoing 25 evaluation of those tick sizes and what securities fall into 0104 1 the tick sizes, and whether or not they should be in or out. 2 I think that is a healthy evaluation if given enough time to 3 gather data about how the security is or is not trading. 4 I would also say that I did sign a letter a couple 5 of years ago, our timing was impeccably poor, before the 6 flash crash on May 6, with NYSE and NASDAQ saying let's do a 7 tick size pilot. In that case, it was regarding tighter tick 8 sizes for very liquid securities, which we have yet to hear 9 from the Commission on. 10 I would say we have come together in that regard. 11 We would still stand behind that but the reason we are here 12 today is about wider tick sizes for small and mid-cap 13 companies. I think for the same reasons we could come 14 together, but I don't think one of us doing it is going to be 15 good for the industry. 16 Similarly, lessons learned from the flash crash, 17 one thing was volatility mechanisms implemented by a single 18 Exchange in a fragmented market doesn't make much sense. 19 Tick size experiments in a fragmented market by a 20 single market doesn't make much sense. 21 MR. HEALY: To broaden your question just a little 22 bit. You talk about market solutions, and typically when you 23 talk about market solutions, it's the trading community. 24 I think the issuer community would come at that 25 from a different angle. They would say why don't you let me 0105 1 pick the tick size. I'm the guy who is sitting there with no 2 liquidity in my stock. I'm the guy that can't get anybody to 3 cover me. Why don't you give me the option to go to a 4 nickel, and if I think that's good for my shareholders, I'll 5 do it. 6 You have to make that available to get the market 7 solution that you are talking about. That is not currently 8 available. 9 MR. REED: I would just add echoing Chris' thought 10 a little bit. The extent to which this is still seen as a 11 problem really depends on the point of view and who you are 12 asking. 13 I think there has been tremendous benefit to 14 decimalization, and I think this issue is inherently 15 controversial in the sense that tick sizes benefit some 16 market participants at the expense of others. 17 I think when you are sort of thinking about this, I 18 absolutely applaud the Commission for taking a careful view 19 on this and think about how the data has been used in the 20 past, how the data reflects the views of those people making 21 the points. 22 MR. CRONIN: My own sense has been that the 23 conflicted economic models that exist today certainly don't 24 perpetuate a solution in the industry, so Exchanges becoming 25 for profit, for example. 0106 1 We have seen a number of different competitive 2 efforts but am I the only one that seems to sense that most 3 of that is around pricing. 4 We hope that the competition that would be spurred 5 by NMS would really be about innovation and ways that we 6 could make trading costs less impactful. 7 My sense is unless and until there is either a 8 mandate that it goes to a specific level or some greater 9 change in the economic incentives of all, that it is going to 10 be very, very difficult to get to a solution that I think is 11 amenable to all. 12 The other thing I would say is while I recognize 13 that there are going to be those with different thoughts 14 about how this whole thing has kind of played out from their 15 own perspectives, I would take a step back and say as we try 16 to prioritize, what's the important voices in this whole 17 thing. 18 I do hope that we don't lose sight of the fact that 19 retail investors by and large get the majority of their 20 exposure from institutional investors. As we try to solve 21 this complicated problem, let's not just address success by a 22 self directed retail investor. 23 Again, if we are solving it for the greater of 24 market participants, we would just suggest that institutional 25 interests are obviously very important as well. 0107 1 Unfortunately, there is just a natural tension that 2 exists between what might be the best outcome for retail 3 oriented investors, the self directed variety, and bigger 4 institutions. 5 In recognition of that, when we talk about things 6 like dark pools and other things, there are real reasons to 7 kind of keep things different, that this one size fits all 8 paradigm just doesn't work in the U.S. market structure nor 9 should it. 10 I think that applies not only to retail and 11 institutional phenomenon but clearly to capitalization of 12 companies and stocks and how they trade. 13 MR. JIGANTI: You can't have the retail investor -- 14 you have to have them present at the table when these things 15 are thought of. 16 I believe that as retail investors look at the 17 marketplace, that's the advertisement. That's the part of 18 the market that is at the inside quote. The institutional 19 investors have every tool at their advantage to use all the 20 different market mechanisms to achieve their desired 21 position. 22 MR. CRONIN: Of course, you recognize that if you 23 are buying 200 shares, it's a far different proposition than 24 buying 300 percent of the daily volume. 25 We have to have additional tools, obviously, 0108 1 because of the impact that our orders could have in a 2 marketplace. 3 I would say that these kinds of tools should be 4 available to all kinds of investors, and by my best 5 recollection, there are a number of different ways that 6 retail investors get exposure today and access to the markets 7 that they clearly didn't have. 8 As we talk about advances in market structure, it's 9 clear that there have been real advances made. I think the 10 retail investor in particular has been a tremendous 11 benefactor. That is great. 12 Let's make sure -- we are sorry for this 13 complicated problem -- that it is not one dimension we are 14 trying to solve for. That's my only point. 15 MR. JIGANTI: I guess I'm looking at the market as 16 a whole, in the lower cap stocks. When you are talking 300 17 percent of the average daily volume, that is completely 18 different. 19 In our comment letter about this, if we are going 20 to go forward with a pilot program, it does make a lot of 21 sense in some of these areas to widen the tick size. 22 My concern clearly is if you are solving for 23 liquidity, it's one thing. If you are solving to get more 24 research in that group, it's completely separate. 25 MR. CRONIN: I think as we have looked at this, 0109 1 this isn't really about necessarily solving the research 2 problem. It is about solving the liquidity problem. 3 In many respects, that becomes a virtuous cycle. 4 Why don't we post bids and offers today in small and mid-cap 5 stocks? Because the price increment that somebody can jump 6 in front of those posted bids or offers is a penny. It's a 7 very minimal amount of risk that is required to take, to jump 8 in front of an order. 9 What we are saying is well, in large cap, that's a 10 problem, so we just hide more things, we break it up over the 11 course of the day, we adapt and we adjust. 12 In the small and mid-cap space, you just don't 13 have that secondary liquidity part that happens with natural 14 market makers or whatever, because not only is it then the 15 erosion of an incentive, frankly, from an investor's 16 standpoint, but so, too, from a market maker's standpoint, if 17 you can't really make any reasonable economic return on the 18 investment, you won't bother making markets. 19 I think the proposition that we have put forth, and 20 again, this is not necessarily supportive or not of research, 21 is this is a liquidity matter. If you can solve the 22 liquidity matter, nothing draws a crowd like a crowd. 23 I suspect that if you were to find there were more 24 incentives to post bids and offers for institutions, as 25 institutions view the stocks that typically wouldn't fall 0110 1 within their screens because of liquidity concerns, trading 2 just doesn't support what's necessary for a mutual fund to 3 buy a security, if you change that paradigm, all of a sudden 4 you have changed the attraction of that security, which by 5 its own right brings more trading activity to it, and I 6 think, again virtuous things can happen. 7 The question isn't would it be reasonable to take a 8 look at this because decimals have been in place for 10 or 12 9 years. The question is is there a way to actually make this 10 much more effective for a group of stocks that don't seem to 11 have participated at the level of benefit that some of the 12 others have in the current market structure. 13 MS. EDWARDS: I have a quick follow up question for 14 Patrick. There is a stream of economic literature that says 15 issuers do have control over their tick size, and that is 16 relative tick size relative to their price. They can split 17 their stock and basically affect their tick size. 18 Why doesn't that work? 19 MR. HEALY: I don't really think we are talking 20 about the right part of the market that is in a position to 21 split their stock. This is the low end, the low cap. 22 I think if you look at those statistics, I'm not 23 sure those are even splittable stocks. 24 What we are talking about here is thinly traded 25 stocks that can't get coverage. Split the stock, put more 0111 1 shares out there as a percentage of volume, it will increase 2 the liquidity in the stock. 3 I'll go back to something that Maureen said in 4 the prior panel, and that is you should be focused on the 5 dollar value traded. You can split the stock all you want. 6 It is not going to increase the dollar value traded at all. 7 It's going to be more shares at a lower price. 8 I don't think that is getting to the issue. The 9 issue is the economics have been drained out of the research 10 side. That CFO is sitting there and he would love to get 11 some research. He would be happy to pay for it. If he were 12 to pay for that research, it would be tainted by the payment. 13 The only way for him to get the exposure he's 14 talking about here is for the economics to return to that 15 part of the research business. 16 Splitting the stock gets you a couple of headlines, 17 gets you a little talk, gets you more shares outstanding. It 18 doesn't do anything with the dollar value traded and it 19 doesn't do anything in terms of getting your research 20 coverage. That, I think, is the problem. 21 MR. HATHEWAY: I've had these conversations with 22 companies large and small. The answers are several. It is 23 costly, one of the big ones. The main one for smaller 24 companies, say $20 stock that wants to get the equivalent of 25 a nickel, it's a $4 company, once it does the split. $4, can 0112 1 you guys hold stocks much below $4? 2 MR. CRONIN: We hope they go a lot higher. 3 MR. HATHEWAY: Yes. Firms begin to actually become 4 concerned about meeting listing standards/thresholds if they 5 get their stock price too low. I think for the small 6 companies, those are issues. 7 Another thing that often gets ignored, if you are 8 talking about a high multiple split for a higher priced 9 stock, you have to buy back fractional shares typically. We 10 are talking about companies that are not big and are not 11 flush with cash and not in an easy position to buy back 12 fractional shares or fractions of round lots. 13 There are a number of quite practical economic 14 reasons that people don't manage their stock price to produce 15 desired tick size. 16 MR. BERMAN: Is there a reason why they don't start 17 off with the tick size? I had an opportunity to take a 18 company public. The number you are looking for is 20, give 19 or take 2. There is something about a $20 stock. Doesn't 20 matter how big or how small. That seems to appeal to 21 investors, both in retail and institutional. 22 If I wanted to have a tick size that was ten times 23 larger, I would just go out at $2 to start with, really 24 doesn't change any of the economics of the company at all, 25 but I have created a tick size by myself without any action 0113 1 on anybody's part, that is effectively ten times larger. 2 In Europe, stocks trade in the single digits all 3 the time. It seems to be very natural that stocks trade at 4 50 pence, 80 pence, 1.5 Euro. 5 What is it about the magic numbers that it has to 6 be in the mid-two digits, apart from listing standards, but 7 presumably all those were geared -- based on the same 8 results. 9 MR. HEALY: I think I can help you with that. Put 10 yourself in the seat of a CFO doing an IPO. An investment 11 banker comes over and he says look, $20. You're thinking 12 maybe $10. All of a sudden, you're a Groupon or another tech 13 stock that has seen a 75 percent decline in your market value 14 because you came out of the gate very strong and things 15 didn't go well. 16 All of a sudden, you're a penny stock. You have 17 listing issues. Nobody is going to take that risk. It's 18 really a matter of risk management. $20 is a very prudent 19 level to price the stock to give yourself some room. 20 If you look over the last couple of years, 21 particularly in the tech sector, room has been necessary. 22 That's why they do it. 23 MS. O'HARA: I'd like to just add a couple of 24 comments about these issues of quoting increments and 25 liquidity. 0114 1 Again, it's kind of an interesting question about 2 we do know that stocks with low prices have higher relative 3 ticks. It isn't true that in some sense, we have the same 4 tick size because it differs. 5 An interesting way to think about these issues is 6 to think about matching stocks on industry and market cap 7 that differ with respect to the price levels they trade at. 8 I've been working on some research along these 9 lines, looking at that kind of issue. I'll point out we are 10 still not finished with this research, so I can't quite share 11 it all yet. 12 One thing that is very intriguing is that stocks 13 that have higher relative ticks, professional market makers 14 spend more time at the tick and provide more orders than do 15 institutions. 16 It's kind of an interesting issue if you compare it 17 relative to your control group. It incentivizes the 18 professional market makers and they have changed their 19 behavior more than the institutional traders. 20 Again, as we think about how changing ticks affects 21 the big picture, we have to recognize that for many stocks, 22 in fact, maybe for most, the bulk of the liquidity now is 23 provided by sort of the professional market makers. 24 This can affect the liquidity. Whether it affects 25 the research that you are looking for, I'm not so sure. 0115 1 MR. HEALY: If I could respond to that, I think 2 what you are saying is that the economics of that relative 3 tick size gives rise to professional traders who add 4 liquidity. I think that proves my point. 5 If you put the economics back into the equation, 6 you are going to get more research coverage. It's going to 7 happen. That is why I think issuers are craving coverage 8 because they want liquidity in their stocks. 9 MR. CONCANNON: Just two points. We heard it on 10 the last panel and we are hearing it now. When I think about 11 tick size and even a pilot, I think when it comes to 12 research, Wall Street has a research issue, regardless of 13 tick size. It has a research issue. If you talk to the 14 large banks writing research, they don't even think about the 15 small caps in this research issue. 16 They are having a hard time paying for research in 17 the large cap activity they have. 18 Wall Street has this research issue, tick size, we 19 could go to $1, we are not going to solve the research issue. 20 That's how expensive research is. 21 I also agree that companies benefit from research 22 and research coverage, but I think that is one of those 23 things the market has to solve. 24 Playing with the market structure, micro market 25 structure to solve that dynamic, I think will ultimately be 0116 1 too costly to the end investor, that we have to fund and 2 create inefficiencies in our market to somehow create this 3 research demand when it is a problem no matter what. 4 I want to make sure it's clear that this is more 5 about liquidity. This pilot is really studying the posted 6 liquidity on our market and whether or not we can adjust 7 that, and if we do adjust it and we are successful, does it 8 benefit the buy side and does it benefit the retail. 9 That should be our narrow kind of goal of analysis. 10 Whether or not we create research, we trade 250 names of 11 small caps, we don't make enough money to put people on the 12 Floor let alone write research. 13 I can e-mail you a page from Yahoo. That is about 14 all we can do. 15 MR. BURNS: Let me pick up on a point, if I may, 16 going back to something Pat said, talking about issuers 17 determining tick sizes. 18 Two things. I'm taking up something from the last 19 panel. Is there a number we ought to be looking at on size 20 of company, above or below which, for purposes of the pilot, 21 we ought to be looking at? I think a number came out at 22 about $1 billion, maybe someone said $500 million. 23 Secondly, how would an opt in work? I'm thinking 24 of where we sit as regulators, thinking about a potential 25 pilot. Can you really have an opt in regime, should you have 0117 1 an opt out regime. That is to say you try to come up with 2 kind of a cross section. 3 I'm thinking of your CFO as well, who has to go 4 into the CEO and say guess what, boss, we're in this pilot, 5 and you don't really know what the outcome is going to be. 6 Sort of a compound question but throwing it out for 7 you guys to talk about. 8 MR. HEALY: Let me take the first crack at that. 9 First of all, I think and I hope you will do the pilot. As 10 you embark on that, I think we need to be careful in a couple 11 of ways. 12 One, we need to isolate the variables to the 13 problem at hand. The quality of the company is every bit as 14 important as the tick size. In fact, more important. You 15 get to the question -- I think the essence of your question 16 is or one of the parts of the question is who do you include. 17 I would say and I am sure the fellows from the OTC 18 market would not support this, but I think first of all the 19 stock has to be listed on either NASDAQ or the New York Stock 20 Exchange because of the higher listing standards. 21 Secondly, I think it has to be profitable. The 22 reason it has to be profitable is we are trying to remove 23 variables from this question and we want to focus on what is 24 the impact of tick size to liquidity. 25 It's very important that we get a control group 0118 1 that doesn't have other factors that influence liquidity. 2 I do think in response to your other question the 3 opening bid on this thing ought to be a nickel, a nickel 4 spread. To add to that, I believe it should be issuer 5 choice. 6 I think you will have a group of profitable NASDAQ 7 and New York listed stocks who will line up. You will have 8 to have a Lotto to determine who is in this pilot. 9 The CFO is going to say look, I'm frustrated. I 10 can't get our story on Wall Street, I don't have any research 11 coverage, we don't have any investment bank deals pending, 12 where else am I going to turn. I'm going to volunteer for 13 the pilot so that maybe I can draw some attention to us. 14 I have a great story, nobody is listening. 15 MS. O'HARA: Can I just raise an issue? This is 16 the academic in me. Academic's hate things where people get 17 to sign up, because if we are actually going to be able to -- 18 if we want to actually be able to ascertain the impact of 19 this experiment, we need to be able to control for everything 20 else. 21 When you allow people to sign up, there will be a 22 certain type of company that may find this more attractive 23 than others. Now as we try to put together a control group, 24 you really run into problems. 25 Should the Commission do this, perhaps one way to 0119 1 address your concerns is the initial pilot does not involve 2 people volunteering, but having seen the outcome of the 3 pilot, perhaps the long run down the road might be allowing 4 that to happen. 5 MR. HEALY: I respect your opinion. I just want to 6 raise the point here that you are not going to have any 7 shortage of volunteers. You are going to have a volunteer 8 Army that is more than you need. You won't have to go to a 9 draft here. You will get issuers that participate. 10 MR. REED: I would like to say I absolutely agree 11 that when academic's try to study these things, the word they 12 use is "endogenicity," that the type of issuer that would 13 choose a new tick size is somehow different from the type of 14 issuer who didn't choose a new tick size. 15 There might be sort of a hybrid approach where you 16 could allow an opt in but then somehow randomly assign which 17 firms actually got to participate in the pilot. 18 I think the Commission has shown excellent ideas 19 about this, the Reg SHO pilot. There might be a hybrid 20 approach where you have randomization. The question that the 21 academic's could answer would be slightly different in this 22 case, and it would be comparing opt in firms who made it into 23 the sample versus those who didn't make it into the sample, 24 and not able to compare those that didn't opt in, but 25 nevertheless, it would give the academic's a little something 0120 1 in terms of an experiment that they could trust. 2 MR. HEALY: I would just tell you, the companies 3 would defer to your good judgment on that. That would not 4 become an issue. 5 MR. JIGANTI: You would have to compare them 6 against like companies in the market that we are looking at 7 today. If you're going to say it's an opt in program but 8 we're going to compare it against the companies that wouldn't 9 make it into your pilot, I think that is where the problem 10 is, and I think you came up with a great solution, Adam. 11 MR. HATHEWAY: I just want to support what Pat is 12 saying. I don't think you will have trouble with companies 13 opting in, and we have probably a quarter of the list, small 14 stocks to very big stocks, that trade at a dime or more. 15 That is ten ticks. In the old days, nobody traded at ten 16 ticks if there was any interest in the security at all. 17 I also don't want to lose sight of the fact that 18 with our petition with Chris' firm and the New York Stock 19 Exchange, my former or future employer, we have probably 300 20 or 400 stocks that are trading at a penny or less. 21 I take partial issue with Maureen's point. As the 22 price comes down and the spread approach is a penny, that's a 23 displayed spread. The traded spread actually comes well 24 within a penny because you have increasing levels of price 25 improvement, increasing levels of mid-point trading, and 0121 1 other types of activities that essentially show a constrained 2 spread on the stock. 3 That raises costs for investors just as much as the 4 spread that is sort of too wide and too unstable. 5 We think a little bit about sort of the broader 6 context here. 7 If you are doing a pilot, you want to do a pilot 8 that gives a little and takes a little. Parts of the pilot 9 are lower costs to investors for narrowing the spread has a 10 certain appeal in some constituencies that may take a dim 11 view of widening the spread. 12 MR. ISAACSON: I want to follow up on what others 13 have said regarding the opt in. As I mentioned our comment 14 letter in my opening comments, I think there should be some 15 gross criteria set in order to be able to opt in, your 16 security has to be of a certain security profile or liquidity 17 profile today, trading less than 100,000 shares ADV or 18 trading less than a certain notational value every day or 19 lower than a certain market cap. 20 I think there is maybe asymmetry potentially in 21 potential conflicts of interest with an issuer saying I want 22 to trade a very, very liquid security, some CFO saying I want 23 my security to trade at a nickel or dime or whatever, and 24 that would be born entirely by the retail investor, 25 especially as Frank said, currently that security is trading 0122 1 at a spread of less than a penny or at a penny today. 2 I think there is just gross criteria, 100,000 3 shares a day or $250,000 a day, some sort of opt in, and then 4 randomization about who is in the control group as part of 5 the pilot would be in order here. 6 MR. HEALY: I'm a big fan of common sense. There 7 were two parameters I threw out there, has to be a listed 8 stock and they have to be profitable. If there are other 9 criteria that folks think make sense, I'm all for it. 10 Again, we're trying to isolate for variables here 11 that give us the right answer with respect to liquidity. To 12 the extent that you have criteria that eliminates those 13 variables, I agree. 14 MR. ISAACSON: One of the criteria, certainly not 15 as it relates to mid and small cap companies, but you have 16 outlier securities that have been incredibly successful 17 companies, Google and Apple, that trade at effective spreads 18 that are 5/10/20 cents per share. 19 I'm not convinced that a penny tick is the 20 appropriate tick size for those securities. 21 Securities that have prices that are greater than 22 100 shares, I think, should be part of the pilot even if they 23 happen to trade more than a certain de minimis amount of 24 trading every day. 25 MR. CRONIN: I think we would tend to agree with 0123 1 that, but I think it's probably important as we approach this 2 pilot program to do it very incrementally. 3 I think the Hippocratic Oath probably applies well 4 here, "above all else, do no harm." 5 In the large cap space, I think generally we all 6 agree that the experience is much better than it has been. I 7 don't think any of us disagree that there is probably room 8 for improvement in some dimensions as well. 9 The point is as you begin this study, it is both 10 important to sort of identify a group specifically as opposed 11 to a broad swath of the market, and really see if there is 12 value in this enterprise. 13 I think probably it does extrapolate beyond if 14 truly there is value. 15 If I could just quickly opine on whether the 16 company should be the determinant of that. I suspect there 17 is probably an element of investors and Exchanges who should 18 probably also be involved in that, because there are a number 19 of variables that kind of matter when you look at liquidity. 20 Whether it is included in an Index, whether it 21 actually has a top ten holder's list of real investors, these 22 kinds of things ultimately have a pronounced impact on 23 liquidity profiles. 24 Just saying that a company is $1 billion and is in 25 the pool and they have opted in or opted out, it doesn't seem 0124 1 like it is setting ourselves up for real scientific data. 2 By the way, we are all for data in these kinds of 3 enterprises. I don't know about you but I don't really have 4 a day in my recent past where I haven't seen a study that 5 says this is true or this is false, et cetera. 6 The best way to try to figure out if these kinds of 7 initiatives are valuable is to get the data. I agree with 8 the academic side on this, which is look, if you have a bunch 9 of guys just opting in because they think it is going to be a 10 better outcome for their security, it's kind of comical from 11 my perspective to think somebody thinks that if their stock 12 is now at a nickel as opposed to a penny, that somehow there 13 is more intrinsic value in that. 14 The other side of it is I think if we were just 15 trying to get to a place where we are bringing value, trying 16 to least complicate this, this seems to be the best approach 17 and having people with the appropriate amount of perspective 18 to help make those decisions and set reasonable criteria with 19 no opt out's. That seems like the right way to do it. 20 MR. JIGANTI: I think a volatility component in 21 this would make a big difference. If you measured volatility 22 along with the notional changes every day, the overall 23 volume, along with the other criteria, and really had a think 24 tank or group of probably similar to this Roundtable to aid 25 the SEC in finding the right mix would come up with a pilot 0125 1 program that would probably be most effective. 2 MS. O'HARA: Can I raise a couple of issues? If 3 you are thinking of designing a pilot, ideally you want to 4 make sure your control group is sufficiently a control group, 5 so that you can actually draw inferences from what you get. 6 There are a couple of things that make sense. 7 Market cap is exogenous to a firm in a sense. Ideally, what 8 you want to be able to select on are things that are kind of 9 exogenous, things like market cap makes sense. Industry. 10 There can be differences in the liquidity across 11 different stocks because they have different appeal to 12 different sorts of investors. 13 You may want to think about controls based on size. 14 I think you would be hard pressed to argue that once you get 15 above $1 billion in market cap, you are really talking about 16 small companies. 17 I think there are really two stories here. One is 18 the really big guys and one is the little guys. If you are 19 really talking about the little guys, certainly in the 20 studies I'm familiar with, stocks that are really below about 21 $250 million in market cap are really very different even in 22 stocks between 250,000 and 500,000. 23 Once you get above $1 billion, they are really not 24 so little any more. 25 As I look at these issues, things like size as 0126 1 measured by market cap are important. 2 The challenge when you begin to put in trading 3 volume is trading volume in turn is actually part of the 4 liquidity of the stocks. The reason you have the volume is 5 in part because of the liquidity you have, so that becomes 6 exogenous and that gets kind of messy. 7 There are interesting things. A few years ago we 8 might have said if you're going to have a control sample, you 9 should be matching New York with New York and NASDAQ with 10 NASDAQ. 11 Nowadays, because of the nature of trading and 12 where things go, those issues are less important but it still 13 becomes the case that as you begin to form your control 14 samples, it may not be that easy to find stocks that are kind 15 of in the same industry but are just bigger. Some of these 16 stocks are really so tiny, finding controls is challenging. 17 I just raise the issue, any kind of strategy that 18 you look at, and volatility fits into this, because 19 volatility is also kind of an endogenous variable that 20 depends on liquidity as well, so selecting on that is a 21 difficult thing to do and I wouldn't advise it. 22 However, looking at things like trading volume and 23 volatility as the outcome of the change is surely something 24 you would want to do if you were doing this pilot. 25 MR. HEALY: If I could just add to that. I do a 0127 1 lot of comparisons between companies looking to either list 2 on the New York Stock Exchange or NASDAQ. Sometimes I've had 3 companies looking to switch markets. Sometimes I had 4 companies in an IPO, and they want to see where others have 5 been. In other instances, you have spin off's and mergers 6 and other things. 7 Invariably, we come back to exactly what Maureen 8 just said, and that is how do you get an apples to apples 9 comparison. We use multiple variables to come up with those 10 apples to apples comparisons. 11 For example, if you are in the same industry and 12 you are Kraft, there is only a handful of companies that you 13 can compare yourself to. You can have trading peers. You 14 can have companies of similar volatility, similar market cap, 15 similar stock price. 16 What you do is you find those common denominators 17 and you put them side by side and you say okay, is this 18 giving me the right answer. 19 I really think that is what you are talking about 20 here as you tailor your list of companies to include in the 21 pilot, there are multiple factors that you want to take into 22 consideration. 23 Just to reiterate a point I made before, the 24 listing standards between NASDAQ and the New York Stock 25 Exchange are pretty darn close. They do set a floor, at 0128 1 least in my mind, that you want to be north of. Thank you. 2 MS. HANLEY: This has been very interesting. Just 3 to confirm, we don't know if we're going to do a pilot yet. 4 MR. HEALY: I just like to think positively. 5 MS. HANLEY: We were hoping to get some information 6 from the third panel on whether that is the appropriate 7 mechanism to study this. 8 I thought some interesting things came up in this 9 discussion. In particular, the notion that perhaps we should 10 not at the Commission focus simply on small and mid-cap 11 stocks, that perhaps large cap stocks could benefit with a 12 closer look at tick sizes there, perhaps even going in the 13 opposite direction. 14 I wanted to hear what the panel had to say about 15 that. 16 MR. CONCANNON: I do think if we are going to 17 choose a pilot, and when I look at the liquidity challenges 18 that we have in our market, it's at the lower end of the 19 market. We have let a lot of listed companies come in, so 20 there are standards, but they are fairly low historically. 21 There are some companies that probably shouldn't be listed. 22 There are some companies that are going to trade poorly no 23 matter what we do. 24 There are thousands of companies that have not 25 traded well for retail or for the buy side. More 0129 1 importantly, the buy side has had the biggest challenge 2 trading these mid-cap and less liquid stocks. 3 I do think when we think about our market 4 structure, these stocks have been treated the same way our 5 most liquid largest caps have been treated. 6 When I think about who deserves the pilot and where 7 we will get the best economic studies from, I do think 8 focusing on a simple group of less liquid names as opposed to 9 sticking in some large cap names that people aren't 10 complaining about, there is plenty of liquidity, we might be 11 putting those names in for a different reason, and we can 12 impact other things by putting those names in. 13 I think keeping the pilot simple, keeping it fairly 14 targeted will actually be a successful pilot and we will 15 actually get it done. If we complicate it with all these 16 other market structure potential's, then we may not have a 17 pilot. 18 MS. HANLEY: The pilot is one issue. I think you 19 mentioned before that you thought there were some issues in 20 the larger cap, particularly even in tick sizes, that could 21 be addressed by the Commission. 22 I am interested to hear whether the one penny tick 23 is fine at the large end and we should just leave it there or 24 should we go to sub-pennies. What do you think? 25 Are there any issues there, not necessarily as part 0130 1 of looking at small and mid-cap stocks, that may be a 2 different examination, but since we have the opportunity here 3 to talk about tick sizes in general, whether any focus should 4 be paid to the larger end of the market. 5 MR. SHILLMAN: In connection with that, there has 6 been a lot of talk about maker/taker and if we were to go to 7 smaller tick sizes for the most liquid stocks and perhaps 8 couple that with a re-evaluation of maker/taker, would that 9 be something worthwhile doing. 10 MR. CONCANNON: I guess the first question is what 11 are we trying to solve. Is it a liquidity issue in the large 12 cap stocks or is it an economic issue. 13 I start from the premise that we have to look at 14 every stock individually or the characteristics of all our 15 stocks, both price, liquidity, float, and how they trade, and 16 determine the right market structure which includes trade 17 increment. 18 There is this organic thing called a "market," and 19 there are different instruments trading in it, and we have to 20 constantly fine tune that. We haven't touched it in over ten 21 years. 22 I think looking at large cap, small cap, I agree, 23 we have to look at it. I think the effective spread in some 24 of the large cap names suggest the market could deal with a 25 smaller tick size. 0131 1 I do think we start bumping into the original 2 issues that we dealt with in Reg NMS, which was does everyone 3 have access to that lower tick size, and is it fair access, 4 did they see it, can they trade it. 5 I think with the sub-penny stocks, maybe we have 6 found a way to do that. 7 It does raise other issues around how to get to the 8 lower tick size in the larger cap stocks. 9 I guess what I'm saying is a pilot on the small end 10 will reveal a lot of information whether there are benefits 11 to liquidity or not, and that may result in changing things 12 across all stocks, large and small. 13 MR. ISAACSON: I think we would recommend from 14 BATS' perspective that we have probably a separate pilot for 15 very liquid or lower priced securities that appear to be tick 16 size constrained, but we would do it as part of this holistic 17 review of tick sizes that we haven't really looked at in over 18 a decade. 19 I refer back to the AMS study, this notion of 20 spread leeway, which really drove the discussion in Europe 21 when we talked about putting in their tick size regimes, what 22 is the optimal spread leeway or the number of ticks in 23 between the average quoted spread. It's a very academic 24 thing to think about. 25 We think the number should be at least one. You 0132 1 can argue about whether it's one to two or one to five. We 2 think it ought to be at least one. In most highly liquid 3 securities in this country, it's zero, because it's a penny 4 wide every day, and that is good for investors who are only 5 paying a penny wide spread. 6 Paul is offering them great executions and Kevin, 7 through an institution, is getting the advantage of that as 8 well. Potentially, that is where most of the trading occurs 9 as well. 10 I think if you do a pilot that is separate from 11 trying to solve the small or mid-cap problem, you can 12 actually give decreased costs to investors even further with 13 the large amount of trading that's going on in the U.S. 14 MR. JIGANTI: We are talking about three issues 15 here. Within the liquidity portion of this, we are talking 16 about the low end, the lower micro cap end. We are talking 17 about sub-pennies in the super liquid, let's call it the Bank 18 of America's of the world, and then we will talk about the 19 Apple/Google world. 20 Three completely different panels, three sets of 21 data. Not to be confused with sprinkling any market 22 structure changes like maker/taker changes. 23 My concern is really the retail investor, of 24 course, what are they going to do when they look at the 25 screen and they see a half tick wide spread. I don't know 0133 1 that gets them to where they feel more confident in the 2 marketplace. Is it going to create a little more 3 "flickering?" Hopefully, not. Presumably, in those super 4 liquid's, you probably wouldn't have that. 5 Then challenging the very high priced stocks when 6 notionally, you are talking about a much smaller spread. 7 Apple being 40 times larger than Bank of America, presumably 8 that is a 40 cents spread. 9 I would argue it has completely different trading 10 characteristics and the retail customer has the ability 11 through a firm like TD Ameritrade use every one of the venues 12 that is out there to get the best execution that is 13 available. 14 MR. HATHEWAY: As I have said a number of times, as 15 have other people, I don't think one size fits all works 16 particularly well. The European experience shows we can 17 manage as an industry multiple tick sizes. We can manage 18 maker/ticker fees with structures other than 30 cents per 100 19 to take. We do that in the States below $1. 20 Reconsidering a lot of things that we put in place, 21 as Chris said, a decade ago, certainly makes sense. 22 When you have a highly retail focused security, big 23 institutional following as well, 100,000 price changes a day, 24 100 shares on either side of the market, similarly situated 25 company, 80 price changes a day, 100,000 shares on each side 0134 1 of the market, the stock price difference isn't a factor of 2 1,000, it's more like a factor of 100. It is essentially ten 3 times excess price changes in high priced stocks and ten 4 times cued up liquidity in the low priced stocks. 5 It just doesn't seem that a penny works for 6 everything. If re-evaluating pennies also means it's time to 7 re-evaluate some of the other things that were put in place, 8 in large part, in the name of simplicity, we were manual when 9 we did decimals. We were manual when we did NMS. 10 Poor Chris had to keep all those price schedules in 11 his head. Now, it's all programmed. You can sit there and 12 talk to panels like this and not worry about actually trying 13 to make money. 14 MR. CRONIN: He's not poor old Chris any more. 15 He's just old Chris now. 16 (Laughter.) 17 MR. CRONIN: I think -- 18 MR. CONCANNON: Can we focus on decimalization, 19 please? 20 (Laughter.) 21 MR. CRONIN: I've seen the numbers of the high 22 frequency guys. Come on. 23 I think our thoughts on sub-penny pricing is at 24 best we are dubious, as you might guess. With a move to sub- 25 pennies, and I can understand why the Exchanges would 0135 1 propagate such a move because it will likely result in more 2 activity, and at the end of the day, that's how they get paid 3 -- this activity is entirely, at least in my mind, very, very 4 difficult to determine whether or not it is valuable or not. 5 At the end of the day, trades can just be trades. 6 It doesn't necessarily equate to liquidity. I think that is 7 what you get. 8 From our perspective, giving somebody price 9 improvement of a tenth of a cent on 100 shares is not 10 material. Certainly, relative to the potential impact that 11 we can have in the market structure by allowing that kind of 12 behavior, so specifically this whole notion of crowding out. 13 We want to post a bid or an offer, so now we have 14 taken the minimum risk increment to do so from a penny to a 15 tenth of a cent, or essentially the increment that you can 16 make back by the rebate. 17 If anybody doesn't think that's going to ramp up 18 activity, you are crazy. 19 What does that mean? You could certainly have so 20 much more quote volume that you really have issues with the 21 transmitability of data and by extension could get a lot more 22 of these technology driven problems that we have intra-day. 23 I suspect while it may be a good idea for some and 24 it might, at this juncture, it seems like an incredibly bad 25 idea to move the increment down to a sub-penny, certainly 0136 1 from our perspective. 2 As you described today, people are already getting 3 the benefit of trading less than a penny. Maybe that is in 4 crossing networks, the trade at mid-point. That seems like a 5 reasonable outcome. It still is only 50 cents on 100 shares 6 traded. Probably meaningful at some point if you do enough 7 trades. 8 MR. CONCANNON: One thing I would add to that, and 9 that is something that has not come up, trade increment does 10 impact order to trade ratio. It does impact quote data. 11 BATS has it in this study in Europe. If you touch 12 trade increments of liquid stocks, you are playing with 13 market data, and the consumption of market data and fair 14 access to market data. Touching trade increments, decreasing 15 trade increments, will increase order to trade ratio's, will 16 increase quote traffic. 17 We need to be mindful of that. Touching trade 18 increments at the less liquid, you just don't have the number 19 of quote updates in less liquid names, so there is no 20 structural impact of that change in terms of a small and mid- 21 cap pilot. 22 MR. BURNS: Can I throw a quick question in? 23 MR. HEALY: If I could just add something before 24 that. I just want to echo Kevin's comments here from a 25 little different perspective. 0137 1 You heard my opening comments that one of the 2 concerns of issuers is the markets will become a casino. I 3 will tell you the perspective of most issuers I speak with is 4 if you go to sub-penny pricing, which very few of them 5 understand, the casino just got larger. I don't think they 6 have any interest in the casino getting larger. 7 MR. HATHEWAY: It's sub-penny quoting essentially 8 we are talking about. 9 MR. HEALY: It's the wrong direction. 10 MR. BURNS: We got down the rabbit hole of the 11 pilot a little bit here, but recognizing we are going to lose 12 some of the panelists who have had some experiences in 13 Europe, I wonder if anyone can offer observations about 14 lessons learned, pitfalls, experiences. 15 You have multiple jurisdictions sometimes doing 16 slightly different things. 17 Anything you would offer? 18 MR. HATHEWAY: I think one of the key take away's 19 from us in Europe is the importance of uniformity. That is 20 something I said in my other remarks we don't particularly 21 have here, even though we have the penny increment we have 22 been talking about. 23 The schedules worked. We don't have a problem, we 24 are not aware of problems with member firms, and managing 25 schedules -- there are certainly behavioral changes that 0138 1 occur when stocks pass a price increment that moves them from 2 one tick to another. Changes in the professional community, 3 changes in the sort of investor community, which is one of 4 the things the European regulators need to consider. 5 Also, what was applied in Europe originally with 6 the one exception of the U.K., was simply price based. I 7 think that was probably a little simplistic. 8 Generally, we found the European experience to be 9 very instructive for us in terms of a good learn about how 10 things could be done differently. 11 It's probably not a model we could use for a study, 12 Maureen, because there are too many other differences. I 13 think it is a very positive approach to dealing with the 14 price increment and trying to take some of the factors out of 15 the equation that are really not central to what it should 16 cost or what the appropriate cost of trading ought to be. 17 MR. ISAACSON: I would just add to Frank's comments 18 that I think in the U.K., where there is some liquidity 19 component and not just price, I think that has been 20 instructive for us, and we would favor -- it is also to 21 decide what is a high liquidity security in the U.K., there 22 is a periodic evaluation of whether or not the security is 23 high liquidity. 24 I think making sure that evaluation period is long 25 enough is important. You don't want securities going in and 0139 1 out of a certain liquidity profile group. 2 All and all, I think it has been a successful 3 venture, and the uniformity is really paramount, you can't be 4 competing on tick size. I don't think as Exchanges that is 5 really something we should be competing on. 6 I want to make one other point. While there is 7 potentially some concern about sub-penny quoting, I 8 appreciate Kevin's comments about is price improvement of a 9 tenth of a penny really material, I would actually agree. 10 What was troubling to us is if you look at off 11 Exchange venues today, the amount of price improvement that 12 is actually occurring on those off Exchange venues is less 13 than that in most cases. 14 It begs the question of if you're going to change 15 quoting increment, you probably need to change trading 16 increment. That is what we proposed for a pilot for small 17 and mid-cap for a nickel quoting increment that you would 18 have a trading increment of a full penny to force people to 19 not be offering de minimis amounts of price improvement, 20 immaterial amounts of price improvement. 21 I think that is really what is going on in our 22 market in the high liquid securities today where you have 23 people offering de minimis amounts of price improvement and 24 an artificially wide quote. 25 MR. CRONIN: I think we would agree with the 0140 1 concept that there is today a lot of internalization that 2 takes place within the quote which is immaterial, so that 3 price improvement that is being provided to orders that never 4 see the light of day in the marketplace that are being price 5 improved for again increments that I don't think many people 6 would come to the conclusion is relevant. 7 That gets to bigger internalization issues. Again, 8 I'm trying not to complicate any potential pilot program that 9 would be done, but I do believe as I mentioned in my opening 10 statement that decimals, while an important concept here, is 11 not the only thing that probably should be looked at by 12 regulators. Internalization is an issue. 13 Could you get to a point where we agreed that if 14 there is price improvement on a security that takes place 15 mid-point, is that good enough? Does it have to be an 16 increment better than that? 17 I think that would be a reasonable discussion to 18 have, but what has happened is these things continue to 19 happen in a vacuum with no aligned economic self interest of 20 everyone, so you don't get a common answer that would 21 probably be the one for the common good. 22 MR. ISAACSON: I think that the $15 million that my 23 clients have achieved in price improvement back to their 24 pockets, back into the economy, is significant. 25 MR. GRAY: I guess I would note we have already had 0141 1 some really interesting and useful submissions in the comment 2 file for quantitative ways to think about tick sizes and what 3 they should be and perhaps how to do a pilot. 4 It seems like there are two really differing 5 approaches. One is the spread leeway approach that has been 6 discussed here. Another one, which was probably focused more 7 on at the earlier panel, is the clumping, a liquidity 8 clumping approach. 9 I would assume that any pilot would try to evaluate 10 those two ways of thinking about tick sizes. 11 My question for the panel is how would we know? We 12 talked about liquidity or not. If we had this pilot, 13 presuming we had it and you tested the clumping versus spread 14 leeway, what is the prime metrics you would look at to 15 determine whether liquidity got better or worse, and would 16 they differ depending on whether it was a large cap versus 17 small cap company? 18 MR. CRONIN: I think defining success in this thing 19 is kind of an important concept. I think unfortunately you 20 are likely to find that as you ask each one of us, we might 21 define that either a little bit different or radically 22 different. 23 If you start with the premise that institutions are 24 the most important, we would certainly acquiesce and tell you 25 that we would be okay with that. I'm kidding. Thanks for 0142 1 laughing. 2 (Laughter.) 3 MR. CRONIN: We have very good measures of what 4 costs in trading are, as I hopefully gave some insight into 5 earlier. Spread isn't one of them. When we look at our 6 costs, as we kind of aggregate the different parts of 7 transaction costs that we incur, the smallest part of it is 8 commission. 9 Commissions might be seven/eight basis points, but 10 market impact and the information leakage that results from 11 large institutional trades entering the marketplace delay all 12 the different things that can happen, front running orders, 13 most certainly. 14 Those kinds of transaction costs can accrue to 15 50/60/70/80/100/150 basis points in certain securities. 16 From our perspective, we have a reasonable baseline 17 to draw upon. By the way, that is independent of volatility. 18 I think one of the dirty secrets is people talk about 19 transaction costs being lowered dramatically, and they have 20 been in many respects over the last ten years. 21 Volatility is the number one determinant of 22 transaction costs for institutions. As you lower it, and 23 let's be clear, over the last ten years, if you were to look 24 at nominal volatility, it was clearly lower now than then, we 25 would expect costs to go down. 0143 1 Is that necessarily a market structure thing or is 2 that something that was more related to just the volatility 3 environment more broadly. 4 Separating out volatility from the other parts of 5 our transaction costs is actually kind of an important 6 concept, but to my point, I think there are reasonable 7 metrics certainly that firms like Maureen's provide, ITG, 8 that can give us very good insight into the efficacy of this 9 effort. 10 As I said earlier, if it is that there is much less 11 price impact as we go in because the increment of risk goes 12 from say a penny to five cents, and as such, we have a lot 13 more incentive or maybe it's less disincentive to post our 14 bids and offers because if somebody wants to jump in front, 15 it's a more meaningful economic increment that they have to 16 do or risk they take to do so. 17 We might find that as a consequence, there are more 18 posted bids and offers of much higher quality and duration 19 and size. 20 Those are the variables that matter to us. If you 21 have 100 up market that's a penny wide, and that's just 22 flickering back and forth, what is the value of that. Is it 23 really an effective spread market of a penny? 24 I submit to you that when you go to reach over and 25 take the offering and it disappears before you can do that, 0144 1 and by the way, it disappears on all 13 Exchanges at the same 2 time, that's not really something that we could all say is a 3 great market structure. That is not something that really we 4 could say is indicative of transaction costs. 5 From an institutional perspective, I think there 6 are very specific things that could be brought to the 7 discussion to help academicians and others who would be the 8 third parties to help opine on whether or not this worked, or 9 in fact, whether it did work for institutions. 10 The other thing is I would hope you would see 11 obvious things like there are more companies coming into the 12 market, there is much more liquidity. 13 If you look at the ownership and the list of people 14 who bought these companies as opposed to what it was 15 historically, you have some pretty good circumstantial 16 evidence as well. 17 MS. O'HARA: Can I follow up? I think again it 18 comes back to what is the goal of this thing. I agree with 19 Kevin, the metrics for evaluating things like what are 20 transaction costs and trading costs. 21 I completely agree that you looking at quoted 22 spreads as only a minor issue here. It has to do with the 23 effective spreads, the amount of effective spreads. 24 There are some widely agreed upon metrics that we 25 can look at. However, I think the interesting thing about 0145 1 this study may actually not be captured by any of that. 2 If you listened to the first panel, the first panel 3 was convinced that the main thing this would do would bring 4 lots of people back into putting orders into the book and 5 institutions who wouldn't have done it before are now going 6 to populate the book, and the books are going to be really 7 big, and everyone is going to be great. 8 That's really an interesting question. Who is 9 going to be putting orders into the book if you change this, 10 and you cannot ascertain that by simply looking at the prices 11 and the quotes. 12 You actually need order data. Again, let me put my 13 academic hat on here. In order to be able to kind of 14 understand the ecology of liquidity, you actually have to 15 understand sort of how long orders stay on the book, who put 16 them there, where did they come from, who is more likely to 17 be cancelling. 18 That sort of data is available. I'll put in a 19 little plug here for the NYSE who has provided data for their 20 market, trying to sort that out. Hopefully, Frank will feel 21 similarly, willing to do it. 22 I would encourage if you are going to do a study, 23 recognize that it's not like the old days where orders are on 24 the book, they are all the same. Orders on the book are not 25 all the same. 0146 1 It would be very, very helpful if you are going to 2 gauge success to understand if the greater liquidity you got 3 came from a broadening of your participation in the book or 4 it came from an increased willingness of Chris and his market 5 makers to be willing to step up and just throw more and more 6 orders on. 7 I'm not saying necessarily one is better than the 8 other. When you think about the goals that you have been 9 hearing from a lot of divergent group of people, there are 10 people who would not view having more high frequency market 11 makers just providing more liquidity as the desired goal. 12 I just raise that. I think you need more data, and 13 I think you will need a much, much deeper analysis than just 14 effective spreads. 15 MR. BURNS: Professor O'Hara, you have given us a 16 great note to end on. We can continue talking. I hope 17 people will over lunch. We will be back at 2:30 this 18 afternoon when we will talk a little bit more about a 19 potential pilot and what that might look like. 20 Thank you very much to this panel. I am very loath 21 to cut us off here. This has been very helpful. We look 22 forward to continuing these conversations with you. 23 (Whereupon, at 1:02 p.m., a luncheon recess was 24 taken.) 25 A F T E R N O O N S E S S I O N (2:33 p.m.) 0147 1 PANEL 3: STUDYING THE EFFECTS OF ALTERNATIVE TICK SIZES 2 MR. BURNS: I think we're going to try and start 3 the third panel. Oh, we're missing the third panel. No, 4 here we are. It's hard to see down this row here. 5 After a terrific morning, two very lively, 6 thoughtful panels, we are turning to the afternoon session, 7 and where we study the effects of the alternative tick sizes, 8 and for that I'm going to turn the floor over to our Chief 9 Economist and the Director of RiskFin, as we affectionately 10 call his division, Craig Lewis. 11 MR. LEWIS: Thank you, Jim. 12 I thought what I would do, what we're going to try 13 to do today is try to solicit views on what we are trying to 14 accomplish with the development of a possible pilot. I 15 thought I would begin my remarks by discussing what I think a 16 good pilot program involves, and after that I will then turn 17 it over to the panelists. Many of you have actually 18 introduced yourselves already, but I'll give you a chance to 19 re-introduce yourselves to those who are just tuning into the 20 Webcast now, and I will then begin by a lead-off question, 21 and we'll go from there. 22 So the way I wanted to think about pilots is one of 23 the things that we do in the Division of Risk Strategy and 24 Financial Innovation is we help work and develop the economic 25 analysis that supports policy decisions and rule writing. 0148 1 Part of good economic analysis has four elements. 2 The first element is to identify a market failure, 3 understand what it is, what the problem is that you're trying 4 to solve. 5 The next is once you have identified that market 6 failure, your next step is to identify a baseline from which 7 you can begin to design a regulatory solution or, in this 8 case, to design an appropriately scoped pilot program. 9 Having identified the baseline, one can then 10 discuss what the economic effects a pilot program is designed 11 to address a market failure so we can better understand what 12 the market failure is, but we are interested in knowing what 13 the benefits, potential benefits, from this are as well as 14 the costs to market participants, as well as those 15 registrants that are actually required to comply with a pilot 16 program. 17 And then finally, the last piece of a good economic 18 analysis is to consider the alternatives and what the actual 19 economic implications are of the choice that you're 20 recommending or the study that you're conducting, as well as 21 what other possible policy choices could be made. 22 So this morning, we heard a lot of discussion about 23 what the problem is, and I felt that coming away from this 24 morning's two panels, there were a number of issues that 25 appeared to be on the table. The first session focused on 0149 1 the role of capital formation, the public market, 2 particularly that formations of equity, the role of research, 3 equity research, and how market structure issues might impact 4 the provision of equity research and how that might 5 ultimately affect the security issue process. 6 The second panel we had this morning focused more 7 on what I felt were market microstructure issues: liquidity, 8 the provision of liquidity, how more liquid markets might be 9 able to support additional components like equity research. 10 So when you are, from my perspective, trying to 11 design a pilot, I'm looking for feedback and input on your 12 ideas of what might be the effective way to put one of these 13 together. We want to understand what exactly it is that 14 we're trying to accomplish with a pilot program. So that 15 would be the first objective of having a pilot study. 16 So understand what we're trying to solve and then 17 some issues came up. Once we have a clearly articulated 18 scope for the project, the next came down to more 19 methodological issues, sort of econometric issues related to 20 how we would go about constructing a pilot that could 21 actually answer the questions we have in mind. 22 And so we had some concerns about how would you 23 identify a treatment group; what would be the right control 24 group; what are the factors one would consider when designing 25 an appropriate control. 0150 1 Maureen raised a point about voluntary 2 participation in a pilot program and some of the problems 3 associated with allowing a pilot to be constructed by looking 4 for volunteers versus a more scientific approach would be to 5 go out and actually not only identify the control group, but 6 to identify the treatment group as well. 7 So there are a lot of issues on the table that 8 we're interested in hearing your feedback about, and then it 9 comes down to once you have these samples, keep in mind that 10 one of the things I think is important when you are designing 11 and trying to construct a control group, is that the control 12 group should be calibrated to actually answering the 13 questions that are on the table. 14 So if the questions that you're seeking to address 15 in a pilot program are around liquidity, the control group 16 should be focused around liquidity. If it's designed to be 17 addressing equity issues, it should be looking at equity 18 research potentially as one of the potential controls. 19 So with that in mind, I think I'd just like to turn 20 it over and let everyone introduce themselves. 21 Maureen. 22 MS. O'HARA: Well, thank you. I'm Maureen O'Hara. 23 I'm the Robert Purcell Professor of Finance at Cornell. I've 24 spent a large part of my career studying issues in market 25 microstructure, and I've worked on a wide range of topics, 0151 1 including the effects of fragmentation on market quality, the 2 role of different market mechanism in impounding information, 3 the nature of trading platforms. Most recently I've been 4 working on issues relating to high frequency trading. 5 I also am Chairman of the Board of ITG, which is a 6 global agency brokerage firm. 7 MR. JIGANTI: Good afternoon. I thank you for 8 having me. 9 I am Paul Jiganti. I'm Managing Director at TD 10 Ameritrade in charge of micro market structure and customer 11 advocacy. I have been in the trading business for 25 years 12 before coming over to TD Ameritrade and working in my current 13 role. Most of my time, 16 years was with Susquehanna and a 14 year at NASDAQ. 15 I look at this potential pilot program as a great 16 data source and the ability to go forward and make some real 17 judgments of how the market could look going forward. 18 MR. WOMACK: My name is Kent Womack. I am the 19 Manulife Chair in Financial Services at the University of 20 Toronto, after 16 years at the Tuck School at Dartmouth. I 21 was previously at Goldman Sachs where I did a lot of selling 22 of the IPOs and small companies that we're talking about 23 here. 24 My research has been in the fields of IPOs and 25 investment research, specifically, the value of analyst 0152 1 recommendations, and so I will likely talk about that in a 2 little bit. 3 MR. REED: I am Adam Reed. I'm the Julian Price 4 Scholar and Associate Professor of Finance at UNC 5 Kenan-Flagler Business School. I have a Ph.D. in finance 6 from Wharton, and I've worked at the Federal Reserve for a 7 couple of years before the Ph.D. 8 I've studied a large range of trading issues, 9 including issues on mutual funds. I've had a special focus 10 on short selling and equity lending, and in that capacity 11 I've looked at a number of regulatory changes and the effects 12 of those changes on short selling, and benefitted from the 13 Commission's SHO pilot on short selling. 14 I've also done a lot of work on IPOs from both the 15 perspective of short selling and also from the perspective of 16 innovations that go into IPOs. 17 MR. SACHS: Hi. I'm Steve Sachs. I'm head of 18 capital markets for ProShare Advisors. I joined ProShares 19 about 18 months ago after a career in equity and derivatives 20 trading that spanned institutional buy side firms from 21 fundamental stock pickers to quantitative shops, you know, 22 small, mid, large cap across, you know, really the sort of 23 full size spectrum of an institutional money manager. 24 And I certainly want to thank the agency for 25 holding this roundtable today on obviously an extremely 0153 1 important topic, not just this individual topic, but a lot of 2 the things that came out in the first couple of panel, in 3 particular, about just the broad sort of state of the market 4 structure in this country and certainly taking a look at this 5 particular issue is very timely. 6 MR. HATHEWAY: Thank you. I'm Frank Hatheway and 7 the Chief Economist of NASDAQ OMX. It's a role I've had for 8 about the last 12 years. Prior to that I was a finance 9 professor at Penn State University, and prior to that I was a 10 floor trader in the options industry at the Philadelphia 11 Stock Exchange. 12 I had a long set of experiences with pilots of 13 various types. A short sale pilot which Adam referred to 14 under Reg. SHO was, of course, one. Penny pilots in options; 15 my group will be performing a great deal of analysis on the 16 upcoming limit up/limit down pilot as one of the follow-on 17 initiatives of the Commission in response to the flash crash 18 of May 6th. 19 I think pilots are a terrific way for the 20 Commission and market participants to gather information 21 about alternative mechanisms that might be used in the U.S. 22 securities markets, and I look forward to a more detailed 23 discussion of the proposal under consideration here. 24 Thank you. 25 MS. McCARTHY: I'm Maureen McCarthy. I'm the 0154 1 Director of Sales and Trading at JMP Securities. JMP 2 Securities is a San Francisco headquartered, equity focused 3 investment bank, focused on research. We have 350 names 4 under coverage approximately, and 87 percent of those are in 5 the micro to mid-capitalization. 6 We worked on 64 follow-ons and eight IPOs in 2012, 7 and have raised 75 billion for our clients at an average deal 8 size of 142 million. 9 Most of my career has been in the mid to micro cap 10 space. I have made markets for Kidder Peabody in the late 11 '80s and then for Montgomery Securities in the mid-'90s. I 12 was the head of trading and eventually sales trading for 13 Roberts & Stevens from 1996 to 2002, and I've been at JMP 14 Securities for nine years. 15 And, incidentally, JMP Securities is itself a micro 16 cap publicly traded New York Stock Exchange stock that is 17 majority held by JMP Group, JMP. 18 MR. CLARK: Good afternoon. My name is Collin 19 Clark, and I am Senior Vice President of the Strategic 20 Analysis and Market Data at NYSE Euronext, where I lead the 21 data analysis team in the areas such as market quality, 22 client behavior and pricing. 23 I think I'd like to, you know, first of all 24 congratulate Frank for joining the company. 25 (Laughter.) 0155 1 MR. CLARK: I was not aware of that. 2 I've been at the exchange for seven years. Prior 3 to that, I was an equity research analyst at Merrill Lynch 4 and Citigroup for six years covering brokers, asset managers, 5 exchanges, trading companies. Prior to that I worked at 6 Fidelity for five years where I had various roles, including 7 worked as a mutual fund analyst. 8 I'd just like to spend a couple minutes just making 9 the panel aware that NYSE Euronext provided a written 10 statement in support for the pilot that increases tick sizes 11 in less liquid securities, and we believe this could have a 12 positive effect on depth and liquidity. From an issuer and 13 investor perspective, there is a lack of liquidity in 14 thousands of stocks in our markets. This leads to more 15 volatility and makes it more difficult to attract 16 institutional investors and analyst coverage. 17 From a market maker perspective, the current market 18 structure is not properly incentivizing brokers to display 19 liquidity, and we believe changing the tick size could help 20 to consolidate liquidity at wider price points thus reducing 21 undercutting strategies. 22 I want to thank the Commission for inviting me to 23 participate in today's panel and the staff for organizing a 24 review of the topic that we believe is deserving of 25 attention. 0156 1 Thank you. 2 MR. LEWIS: Okay. I'd like to thank all the 3 panelists for agreeing to come down here and spend some time 4 with us today. 5 I thought before we move on since we're coming back 6 after lunch, I will also introduce the rest of my colleagues 7 that are from the SEC staff. To my immediate left or your 8 far right would be Lona Nallengara, who is the Director of 9 the Division of Corporation Finance here at the Commission. 10 To my right is Amy Edwards, who is an Assistant 11 Director in Division of Risk Strategy and Financial 12 Innovation, and she is responsible for the Office of Markets. 13 We have Kathleen Hanley, who is my Deputy Director 14 in RiskFin, as well as Deputy Chief Economist. 15 At the far end of the table we have Gregg Berman, 16 who is leading the Office of Analytics and Research. 17 Oh, Jim, you're hiding back there. My apologies. 18 That was the visual contact I was looking for. 19 So to my far -- is Jim Burns, Deputy Director of 20 Trading and Markets; David Shillman, Heather Seidel, and Dan 21 Gray. 22 So thank everybody and look forward to your 23 comments. 24 So I guess I will start out by throwing the first 25 question over the transom, and what we're doing is 0157 1 considering the first two panels. What information would 2 help us better understand the effect of tick sizes on all or 3 a subset of stocks? What alternatives to a pilot should the 4 Commission consider in gathering additional information on 5 the effect of tick sizes? 6 One of the things I'd be interested in hearing from 7 academic colleagues would be do you feel that there are any 8 gaps in the existing academic research that the pilot could 9 be used to fill. 10 MS. O'HARA: I'll start us off. 11 There are huge gaps in what we know in academia 12 about markets because so many of the studies that we have 13 done in the past that have been referenced by you and others 14 are necessarily dated, and particularly a lot of the studies 15 that were done about decimalization. There's a lot of 16 excellent studies, but they were done, many of them, in, you 17 know, the early 2000s, which is really in the period before 18 the high frequency world began. 19 So academics love pilots, right? Because we would 20 love to know more about what's going on in markets, and if 21 you would design a careful experiment that we could analyze, 22 we're always happy to do that. 23 There is, however, sort of an interesting set of 24 questions in the sense that I don't know that we actually 25 know that much about the nature of the liquidity provision 0158 1 process and markets right now as it is, and you know, the 2 ability to do research these days in markets and to 3 understand what you're looking at is non-trivial because it 4 requires data that most academics just cannot get. Tack data 5 on a monthly basis is not sufficient to address these issues, 6 and so we've been very fortunate in academia when the markets 7 have been willing to support this. 8 And, you know, Frank Hatheway and NASDAQ provided 9 an HF database which is very handy. The New York Stock 10 Exchange has been helping me with some work on trying to 11 understand the nature of liquidity at the NYSE. So I think a 12 pilot can be very, very useful, but it may also be helpful to 13 be able to have data available that allows people to just 14 evaluate the current structures of market because within 15 those markets, for example, the tick size actually, the 16 relative tick size, does differ across stocks. If you can 17 match stocks that have the same characteristic but one is a 18 $4 stock and one is a $40 stock, then you actually have a 19 situation where there's a relative tick size that's ten times 20 as large as the other one. 21 So should you decide not to do a pilot, I would 22 like to hope that maybe you could encourage organizations to 23 step up as the NYSE and NASDAQ have done to try and provide 24 data that could help, and obviously if you do do a pilot, 25 then the issue is, you know, to be very, very careful how 0159 1 it's structured so that you don't get caught up in the 2 plethora of issues that really surround liquidity provision 3 in markets now. 4 MR. LEWIS: Adam. 5 MR. REED: Yes. I would say that a pilot is 6 probably one of the best, if not the best, ways to study this 7 issue. I think in my mind the key defining feature of a 8 pilot is that there's a built in randomization that you 9 basically can't get except for the luckiest of natural 10 experiments, and I think, you know, the Commission's 11 willingness to design a pilot, I think, definitely goes in 12 the right direction. 13 From my experience with the Reg. SHO pilot, I think 14 it provided a good test bench for lots of issues around short 15 selling, and I would say it even went past that. The people 16 sort of think of it as a nice instrument to use in 17 identifying liquidity overall, and they'll think about some 18 of the work that came out of that pilot study as a way of 19 testing various hypotheses even beyond the direct hypotheses 20 about short selling, in particular. 21 MR. CLARK: If I may just highlight, actually, an 22 academic paper done during -- it was a 1999 tick change that 23 happened on the Paris Bourse where some securities had lower 24 ticks and some had higher ticks, so kind of a unique event, 25 and the conclusions were interesting, and that might be a 0160 1 worthwhile review if you haven't read it, but you know, the 2 summary, at least as far as stocks that had higher tick 3 increases, the outcome was limited to no impact on spreads 4 and increased displayed liquidity in that market, so just one 5 study to note. 6 MR. LEWIS: Let me ask a related question. If you 7 were to identify the market failure that a pilot study is 8 going to be designed to answer, in your view, what is that 9 market failure? What is the pilot designed to look at? 10 MR. SACHS: I can certainly jump in from the 11 practitioner's perspective. 12 I think that there's actually really multiple 13 things, but if I had to prioritize them and rank them, 14 displayed liquidity, from the institutional investor's 15 perspective and the access -- in the ability to access that 16 displayed liquidity is probably at the top of the list. 17 One of the things that, you know, we see already in 18 the data, you know, from the academic studies, and if you, 19 you know, talk to practitioners and traders of, particularly, 20 small and mid-cap securities, is that -- and it was alluded 21 to in one of the -- in the first panel this morning -- while 22 the tick size is a penny, the actual, you know, spread that 23 you're trading at in most of these securities is typically 24 far greater than a penny; it's something in the, you know, 25 probably five-to-eight-cent range, if I had to, you know, 0161 1 anecdotally take a guess. 2 So, the market, in and of itself, is already 3 aggregating liquidity, you know, at the points where supply 4 and demand, you know, meet 5 One of the things that, you know, the institutional 6 community keeps, you know, falling back on with the issues 7 is, again, displayed liquidity, access to the liquidity or 8 the -- the flickering issue, things of that nature. 9 So, from a pilot program perspective, which I do 10 support, and I agree wholeheartedly that I really don't think 11 that there is a better way to gather actual data that we then 12 could use for a qualitative analysis, an empirical analysis, 13 that doing a pilot program. 14 There are multiple models around the globe that you 15 could use from a multi-tier, a multi-tick tiered system, but 16 ultimately, looking at a number of issues, but I would argue 17 the displayed liquidity issue and the -- the aggregation of 18 that displayed liquidity issue is certainly going to be first 19 and foremost from an institutional investor's perspective. 20 MR. HATHEWAY: If I were -- I think I'm going to 21 rank some possibilities for market failure for things that 22 may be harder to investigate to easier. 23 In the panels earlier today, part of the issue, and 24 I think a key part of the issue that brings us here is the 25 change in the capital formation process for small companies. 0162 1 Essentially that comes down to cost and availability of 2 capital and public equity markets cease to be as attractive 3 to small companies as it once was. 4 Essentially, there is an increase in the 5 illiquidity discount that small companies face. So, the 6 first market failure is an illiquidity discount inadvertently 7 imposed by decimalization. 8 I think that's very hard to observe and very hard 9 to measure. 10 So, we're then left looking at proxies, of which 11 liquidity, as Steve was just saying -- and now I'm not 12 talking about an illiquidity discount that investors have to 13 take into account that raises the cost of capital of firms 14 but simply liquidity getting a trade done in the market is a 15 proxy for what might have happened to illiquidity discount. 16 That can be investigated through a pilot. 17 The other potential market failure, as we were, 18 again, discussing this morning, is the coordination failure, 19 lack of uniformity around trading increments, and what impact 20 that may well have on incentives to display liquidity, on the 21 ability to access liquidity; heard very different points of 22 view this morning from different constituencies as to whether 23 the ability to trade on one set of increments in one type of 24 markets and a different set of increments in another type of 25 markets is beneficial or harmful or not. 0163 1 The final failure is whether there is insufficient 2 revenue associated with bringing small companies to market or 3 having small companies represented as part of your sales 4 portfolio for your sales force to fund research. 5 As, again, some of the points made this morning I 6 agree with, that's probably the most -- the least directly 7 related to tick size and the most difficulty to explore with 8 a pilot, cause there may well be a great many factors at 9 work. 10 That's my list. 11 MR. BERMAN: Could I ask a follow-up question? 12 Steve, you mentioned displayed liquidity. If 13 there's 10,000 shares out to 10 cents today and then we 14 engage in the pilot, could you comment on the three different 15 aspects I think that -- that we could see in change, though. 16 One is the 10,000 stays. Nothing changes. The 17 second is that the 10,000 becomes 20,000, so it increases, 18 but the 20,000, I think you mentioned flickering or it's less 19 accessible. So, how would we actually measure whether or not 20 that 20,000 is valuable compared to the original 10? 21 Or the 10,000 over 10 ticks becomes 5,000 at one 22 tick and then 5,000 at the next tick, so it's the same 10,000 23 but it's now aggregated, cause I think you mentioned 24 aggregation. 25 Do you think about each of those three differently, 0164 1 and should there be different criteria? 2 MR. SACHS: I don't think that there should be 3 different criteria, because what I'm discussing there is 4 really more the reality, as you know, of how markets work and 5 how market makers would think and, ultimately, how we've, you 6 know, these days, programmed algorithms to -- to work and to 7 think in the marketplace. 8 But the assumption is, in the concept -- and maybe 9 it's just nothing more than an assumption -- is that, you 10 know, again, given the -- the current one-penny tick 11 increment, there is an extremely low hurdle for a market 12 maker or other market participants to step into the process. 13 So, that, by its very nature, is going to encourage 14 people to step into the process, but again, given the -- the 15 market structure that we have and ultimately what we've moved 16 to post-Reg. NMS is that, you know, you don't really display 17 that liquidity, because why would you? 18 You can step into the middle of the quote for a 19 one-penny increment for as little as 100 shares or less, 20 which is exactly what your algorithm or market making, you 21 know, intention or tendency is. 22 So, if you -- if you think back to when there was a 23 -- you know, a 16th or 8th of a point spread, when the 24 economic incentive to step into the middle of that quote was 25 a much higher hurdle, the tendency of the market participants 0165 1 was to display liquidity from a priority perspective. 2 Because the incentive was not as great, the 3 incentive to hide what you were doing in the marketplace, at 4 that point, from a practitioner's perspective, was less. 5 I do fully, fully agree or acknowledge that, 6 ultimately, given the psychological change that's taking 7 place with market participants today, that you may not see 8 that net outcome, but ultimately I think it's one of the very 9 important facts that a pilot program would prove out. 10 One of the baseline assumptions from the trading 11 community, institutional trading community, is that a wider 12 spread with lower -- with higher economic incentive to step 13 in the middle would create more displayed liquidity, you 14 know, from that perspective. 15 So, I don't think of those things differently. I 16 would actually argue for keeping the variables around that 17 type of activity as few as possible to see what the possible 18 outcomes are. 19 MR. CLARK: If I could just add, I think another 20 important point to make is that, you know, a higher tick does 21 not necessarily mean a wider spread, and that obviously we're 22 talking about this tradeoff between how much additional 23 displayed liquidity we can get as well as that -- that 24 tradeoff of the spread. 25 I mean, to give you an example -- and it's more of 0166 1 a question -- you know, if you've got a stock that's got a 2 quoted spread of, say, seven and you move the tick increment 3 to five, does the -- does the stock actually start moving to 4 the five increment and actually narrow the spread where you 5 get some liquidity now aggregated at five and you get a lot 6 of additional liquidity aggregated at 10, the net result 7 being you've got a narrower spread and potentially more 8 liquidity, but I think there's this -- this tradeoff between, 9 obviously, spread and displayed liquidity that's really at 10 the core of this. 11 MR. HATHEWAY: If I could make one point, I mean 12 one of the temptations of a question like that is to say you 13 could answer it with institutional trading cost analysis, 14 such as what Maureen's -- the company Maureen's chairman of 15 does in there full-time. 16 The trouble with that approach is it assumes that 17 there is no change in the mix of what type of stocks 18 institutions invest in. It's typically transaction cost 19 analysis when you get summary statistics. It's 20 volume-weighted. 21 The fact that there is a transaction has to have 22 occurred before you can measure transactions cost. 23 If a pilot makes it less attractive to do a 24 transaction in certain types of costs, those stocks drop out 25 of the study. If it makes it more attractive to do 0167 1 transactions in stocks where there are not transactions 2 today, they come into the study. 3 It makes the analysis more complex. Folks like 4 Adam and Kent and Maureen certainly have the hardware to deal 5 with it, but it becomes now more on the -- I won't say 6 necessarily witchcraft but a more complex part of statistical 7 and economic analysis to address what sounds like a simple 8 question: Did costs go up? Yes/no. 9 If the program actually works, you're going to get 10 more liquidity, more costs to measure, and stocks that people 11 don't trade today, cause they're too costly, and that may 12 pull averages up, but in effect, the average is masking, 13 really, what we're interested in. 14 MS. O'HARA: Can I just echo a little bit what 15 Frank has brought on? I think in any study like this, there 16 are going to be winners and losers in terms of the effects on 17 stocks, right? 18 For the stock that had the seven-cent spread, there 19 may be a big winner in the sense that perhaps spreads go to 20 five. The stock that had the three-cent spread -- its spread 21 might also go to five. 22 And so, one of the challenges in this is that any 23 -- any pilot has to be able to differentiate kind of what's 24 happening both to the winners and the losers, so averages can 25 be very misleading. 0168 1 The second issue that is one to contemplate is, as 2 we look at orders that trade in markets, we already know that 3 an awful lot of liquidity now doesn't trade on the exchanges. 4 It trades elsewhere in various other settings. 5 If you widen the spread, the tick, right, then for 6 many stocks, any stock right now that's trading less than 7 that, the spread is going to be bigger, and that gives 8 incentives for things to trade off exchange. 9 So, if we define the metric just simply have we 10 lowered trading costs, we can answer that. If you define the 11 metric, have we changed the liquidity provision process 12 whereby now people who were not willing to post limit orders 13 before but are now, so books are bigger, then you need a lot 14 more data than just looking at the trades and the quote. 15 You really do need to look more carefully at the 16 nature of the liquidity on the book, and you can do that if 17 you have the data. 18 But I would suggest if you're going to do this 19 study that you really shouldn't just content yourself with 20 trying just to measure the transactions costs without using 21 the opportunity to understand more fully the liquidity 22 production process. 23 MS. EDWARDS: I think that's a good transition to 24 the next part of this panel, where we talk about a bit of 25 what the goals of a pilot should be. What do you recommend 0169 1 the goals of a pilot should be if the Commission considers 2 going in that direction? 3 And we'll transition into design from the goals, 4 but at least for right now, let's think about what goals 5 should a pilot be designed around? 6 MR. JIGANTI: I'll jump off and say do no harm. 7 You know, in a lot of these stocks, even in this end of the 8 curve, the typical retail customer gets -- 96 percent of the 9 time they get filled on their full order at the quote they 10 see. 11 So, you know, it's hard to give up the extra spread 12 for no apparent reason, for, you know, the group that I am 13 here for. 14 So, you know, do no harm. 15 MR. SACHS: Obviously the balance to that -- Kevin 16 Cronin on one of the earlier panels stated it very well. 17 Again, the vast majority of individual investors in this 18 country get their investment exposure through some sort of 19 pooled vehicle or some sort of professional managed vehicle. 20 So, from a design perspective, obviously, you know, 21 being extremely biased and close to this issue, I would 22 certainly want to see a very large dose of balance amongst 23 the constituents of what the ultimate, you know, effects 24 would be. 25 I certainly echo the effect of do no harm, but 0170 1 ultimately I think that is, one, somewhat unrealistic, but 2 two, you know, at the end of the day, you know, most 3 regulatory actions and most market structure actions have 4 some sort of reaction. Some are positive; some are negative. 5 So, ultimately, from a design perspective -- and to 6 my point, a heavy dose of balance for all of the constituents 7 and looking at it from a much bigger picture perspective, as 8 opposed to just an individual transaction or a quote. 9 One of the things that I wrote down earlier that 10 came up in an earlier panel is, all market participants, 11 myself among them, we tend to confuse quote with liquidity on 12 a regular basis, and I caution us, in the design of a pilot 13 program, to not be so narrowly focused that we make that -- 14 we make that confusion. 15 MR. GRAY: I think that's a good point, and I guess 16 I would ask Maureen, based her everyday experience with 17 trading small cap stocks, if there was a pilot that didn't 18 widen the spreads for a number of the -- widen the increment 19 for those stocks, what are the sorts of things that you would 20 focus on to determine whether the market was better and 21 liquidity improved in those stocks, beyond perhaps just 22 looking at the quoted spread? Are there other metrics that 23 you would recommend? 24 MS. McCARTHY: I think average -- average trade 25 size and not -- not -- what we would recommend, actually, is 0171 1 -- I think somebody mentioned this in another panel, but -- 2 is having a -- a study group that is the pilot, having a -- 3 almost a paired group that is the non-pilot but a study group 4 to compare it against, and then, also, of course, to compare 5 it to the statistics of before -- before the pilot. 6 In terms of -- in that regard, average trade size, 7 dollar volume traded, volume traded, and then we thought of, 8 you know, some other ways that might be outside of the box to 9 figure out whether not only market makers are accepting their 10 obligations but also if it is actually working in terms of 11 adding liquidity, and some of that can't be measured. 12 If a market maker wants to add liquidity to the 13 market, we will send out an IOI to our institutional 14 customers, and that IOI can be acted upon by the institution, 15 or it may not be acted upon by the institution, but if it 16 isn't acted upon by the institution but there is more of a 17 willingness to trade than there was before because there is a 18 perceived wider spread and a better economic incentive, that 19 might not be measured in just the stats that we were talking 20 about earlier, average trade size, and it may. 21 But one way or one additional measure that's just a 22 little out of the box is we could look at IOIs in the control 23 group prior, IOIs now, and then compared to the IOIs in the 24 -- in the study group, as well, and whether -- hopefully that 25 data is available. 0172 1 I think it is. I think it's different from 2 advertised trade volume by market makers, but -- but it is, 3 in and of itself, a metric that we could use to determine 4 whether people were more willing to -- and market makers were 5 more willing to commit capital. 6 MR. HATHEWAY: I think one of the goals we've had 7 here is capital raising for small companies, either ones that 8 would like to do an IPO or ones that are currently listed and 9 looking to raise capital through a follow on or PIPE or what 10 have you. 11 For a pilot of sort of reasonable length of time, 12 if we look at the really excellent work Grant Thornton has 13 done, it's clear that these changes are somewhat slow, 14 particularly with respect to, say, the short sale pilot or 15 the option penny pilot, where the impacts were relatively 16 quick. 17 So, it's not clear that we would have enough data 18 to answer questions like that at the end of a one-or-two-year 19 pilot. 20 A proxy that occurred to me while Maureen was 21 speaking is you can look at things like ownership mix, you 22 know, who is in these stocks that are existing in the market 23 today, who holds them, what's the institutional holding, how 24 diversified the institutional holdings are, things like that 25 that, again, are proxies for accessibility for the 0173 1 institutional community, for the retail community into some 2 of these securities. That's all data that's available. 3 MR. WOMACK: You know, when decimalization came to 4 be, it was about saving investors' bucks, right, big bucks, I 5 would say, and you know, I think one of the questions we 6 should have firmly in the front of our mind going forward on 7 something like this is what are the transaction costs that we 8 expect to be imposed on investors, if they are such, in order 9 to find out, over how many years -- three, four, five -- that 10 there's no benefit? 11 I mean, I think a healthy question is, what is the 12 likelihood, the probability that there is, in fact, some 13 benefit to tinkering with the cost of transacting? 14 I don't think -- I'm not at all convinced that I 15 have a good guess about that. I would guess that this is a 16 important question maybe this panel could opine on as to the 17 likelihood that after you've spent how many millions of 18 dollars on higher transaction costs, what do you think you're 19 getting out of it? 20 MR. HATHEWAY: The panel hasn't started arguing 21 much yet, so maybe it will start. 22 When decimalization came in, spreads went up on a 23 quarter of our stocks, mostly inactive stocks, and so, this 24 is essentially Colin's point. 25 It's not clear spreads, which is the retail cost, 0174 1 generally speaking, are going to go up or down, and then the 2 institutional costs are going to go up or down, and the nice 3 thing about a pilot -- and the short sale pilot and the 4 implementation process for decimals have this feature to 5 them, as well -- you can pull the plug. 6 If it really -- say we -- we do this pilot and DMMs 7 all quit and the market makers all quit and these stocks are 8 now just flat out not able to trade, we could change our 9 minds about the pilot, but it's not -- again, to answer your 10 rhetorical question, why do we think things may get better -- 11 well, because for the subset of stocks, things got worse, at 12 least in the NASDAQ world, very different from today, when we 13 did this -- we did a reverse 12 years ago. 14 So, I think there's some basis to think it may get 15 better. 16 MR. JIGANTI: And in your mind, is that directly 17 related to decimalization, or could it have been other 18 economic forces that have been part of that explanation, too, 19 over the last 5 to 10 years? 20 MR. HATHEWAY: Well, the change was very short-term 21 around the introduction of decimalization. So, this was the 22 pre-post data analyses. 23 I think the -- probably the biggest weakness with 24 my statistic is I'll bet for inactive stocks there's a ton of 25 noise in what the typical spread is. 0175 1 MS. O'HARA: I would just add that if, in fact, the 2 goal is to improve the liquidity for a subset of stocks, then 3 clearly what you want to be looking at are the variables that 4 would reflect that. 5 So, for example, time-rated orders on the book, 6 right? If -- if you really think that part of the problem is 7 that liquidity has been flickering, then -- and this is going 8 to incent people to put orders in the book, we should be able 9 to see that in the duration of orders on the book. 10 We should be able to see, for example, if the 11 argument is that institutions will now be incentivized to put 12 orders in the book, then when we look at -- right? -- who is 13 putting the orders in the book, we should be able to see more 14 institutions. 15 If, instead, what you're seeing is market makers 16 are now incentivized to put more orders on the book, that may 17 well lead to enhanced liquidity, but the people this morning 18 who really are hoping that these incentivized institutions 19 are going to take the extra returns they're making and use 20 that to move forward in research, if they're not actually the 21 source of any action here, that's unlikely to happen. 22 So, I think there's a variety of measures that we 23 know how to look at. We certainly know how to measure 24 transactions cost. We know how to measure relative expected 25 spreads. 0176 1 A lot of these actually turn out to be highly 2 correlated with the dollar trading volume in a stock. So, 3 you know, a very quick thing you're going to be able to see 4 is whether or not stocks in which you've raised tick sizes 5 have higher dollar trading volumes, because if you don't, 6 you're not going to be able to see a whole lot of improvement 7 in a lot of these measures. 8 So, those are things to keep in mind, I think, as 9 we go forward. 10 MR. CLARK: I think we would naturally expect that, 11 you know, market maker participation would increase 12 proportionally more than institutional posters, and that's 13 the -- that's what you're trying to -- to incent, and if it 14 also does incent institutional posting, I think that would be 15 a great thing, but typically, you know, the -- the 16 institutions are the takers of liquidity. 17 So, to Maureen's point, you would hope that that 18 would result in additional dollar value traded, and that's 19 really one of the key questions: Does displaying more result 20 in, you know, a different behavior from the institutional 21 customer and draw them in to trade more? 22 Just a couple other, I think, related metrics 23 around, you know, institutions and trading size that could be 24 interesting to look at are how quickly does it take to 25 execute a larger size order? 0177 1 Typically, in the smaller cap names, you have to be 2 very patient; it takes a very long time. So, the hope would 3 be that this increased liquidity would speed the process at 4 which someone can get, you know, a sizeable position. 5 And then the other related metric would be the 6 transaction cost analysis or what you call the slippage, you 7 know, of that trade, and what does -- you know, what's the 8 impact -- impact cost of a large trade, you know, in a market 9 where there's more depth, and hopefully those are -- those 10 are positive results. 11 MR. SACHS: I would echo that about the length of 12 time, although arguably it is a difficult metric, I think, to 13 measure and get the information on, but again, from a 14 institutional, you know, manager's perspective who is 15 operating in a market, you know, that is constrained by 16 liquidity, where you're seeking to, you know, take -- buy 17 liquidity for multiple days' worth of average volume, an 18 interesting thing that I have certainly experienced in the 19 post-decimal world, as well as has been reiterated in some 20 research, I think most recently in -- in a piece in 2004 -- 21 Bess and Binder, I believe, was the author, if I have my 22 notes correct -- institutional orders that were -- in the 23 post-decimal world -- that were less than one day in length 24 saw increased transaction costs in the decimal world while 25 multi-day orders actually saw a significant decrease. 0178 1 So, when all of us talk about decreased, you know, 2 transaction costs from decimals, ultimately my belief had 3 always been that, while we are certainly seeing a decrease in 4 cost, you know, we have simply spread out or order times, you 5 know, to account for the other effects of decimalization, 6 most namely is the, you know, lacking of resting liquidity, 7 small order size, or the decline in order size. 8 So, ultimately, arguably, it's a very, very 9 difficult metric to get your arms around and to get 10 information on, but I do think it's an important metric that, 11 if it can be measured, to look at trade length or duration in 12 the marketplace relative to these metrics. 13 MR. REED: One thing to add here is that I think if 14 we start focusing on sort of the wider effects in terms of 15 capital formation, in terms of cost of capital, one of the 16 things that starts to matter a lot is the information 17 environment that these firms are operating in, and there's 18 tremendous disparity in the information environment between 19 very large stocks and very small stocks, and so, it's 20 probably important to take that into account in some way in 21 designing any sort of pilot. 22 You'd want to be able to match carefully on 23 dimensions other than just these direct measures of trading 24 and liquidity but things like news, things like coverage 25 would sort of be at the heart of a lot of the arguments that 0179 1 were happening this morning. 2 I think those linkages -- it would be nice to make 3 sure we knew more about those linkages. 4 MS. EDWARDS: Okay. And that's a good transition 5 into actually getting into the design elements. 6 So, we have had earlier discussions today on some 7 issues related to design, on which securities should be 8 included, but I think keeping the goals in mind and what's 9 feasible to study in mind, I think I'd like to ask the group 10 to opine a bit on how to design it. 11 How should these securities be selected? I think, 12 how long should the pilot be? And what should the tick sizes 13 be for these different types of securities? Should we -- how 14 many different tick sizes do we want to have in a pilot? 15 And I guess adding onto that is what sort of data 16 would be necessary to address the goals of the pilot that 17 we've put forth today? 18 MR. CLARK: Can I take the easy part of that 19 question? In terms of the amount of time and number of 20 securities, I think it is important to, you know, have a 21 length of time that's sufficient and a number of securities 22 that's sufficient to motivate changes in behavior. 23 So, from our standpoint, you know, at least a year, 24 and you know, a pretty substantial, you know, group of stocks 25 of 300 to 500 might be needed to motivate the trading 0180 1 community to make the necessary changes to respond to the 2 change in tick size. 3 MS. McCARTHY: We agree with that. We think that 4 it should be at least a year and 500 names. Again, if there 5 could be a paired non-pilot group of 500 names, as well, that 6 would be ideal. 7 Even though it would be a year or more, even in 8 that -- even in that instance, we do feel that some behaviors 9 aren't going to change just by the nature of it being a pilot 10 and being a temporary situation. 11 That is, you may not see the level of research 12 providers improve, if you believe that liquidity and -- and 13 economics -- being incented on an economic basis leads to 14 more research providers in that name, but it takes, 15 typically, three to six months or more to research -- put a 16 research piece out on one name, and it's -- it's unlikely 17 that we're going to change sell-side behavior to incent them 18 to move into this space on a research basis if it's just a 19 temporary, by nature, pilot design. 20 And the same also might be true of liquidity 21 providers and capital commitment in that it might invite 22 those who already have those programs in place to incent them 23 to engage in those practices, but it's unlikely to, in a 24 temporary process, invite new participants. 25 So, that is one thing we would caution. One of our 0181 1 thoughts is to add non-statistical surveys to both the buy 2 side and the sell side halfway through the pilot to determine 3 whether the buy side sees an impact to what they're doing in 4 terms of liquidity, and from the sell side perspective, 5 whether such a program might invite them, once it's 6 permanent, to engage in further research, capital commitment, 7 or adding more liquidity. 8 MR. WOMACK: On the question of research coverage, 9 this is only a gut feel, but it's an informed gut feel. I 10 don't think a year would be even close to enough. We're 11 talking two or three or four before the market would 12 fundamentally change. 13 You know, there's been a lot of change in research 14 coverage and how it works since the early 2000s when, you 15 know, analyst disclosure came to be. I'm not at all 16 convinced it's going to change back very much, if ever. 17 You know, we will find a different kind of 18 provision, but I'm not exactly sure that it's going to be 19 brokerage-connected information provisions. So, I guess what 20 I'm trying to say, in a nutshell, I don't think we'll learn 21 anything in a year with respect to coverage of -- information 22 coverage. 23 MS. O'HARA: I kind of share Kent's view. I think 24 it will take quite a while before that could happen. Within 25 a year, a pilot, though, could give us a lot of information 0182 1 about changing liquidity production which is somehow related 2 to all of this. 3 I would stress, you know, 500 stocks sounds like a 4 lot, except you're going to have to control for a variety of 5 things that we know are important. 6 For example, market cap, right? You really need to 7 think about who is this pilot applying to, and you know, you 8 might -- you might design like four groups with nano-caps, 9 micro-caps, very small caps and small caps, so you'd sort of 10 have four groups, and then you also need to recognize that 11 those stocks trade at different price levels, right? And we 12 know that low-price stocks tend to have a very different type 13 of liquidity process than high-price stocks. 14 So, if you were to divine, you know, four more 15 groups on prices, then you actually now have 16 groups of 16 stocks, if you think about it, and you'd also then have to 17 design, you know, matching control groups. 18 If you pick up what I think is Adam's very good 19 point that information differences across stocks are 20 important, then you may want to think about the fact that 21 perhaps information differences are also, perhaps, not 22 totally, but correlated with industries, right? 23 Certain industries are certainly -- have different 24 information structure. 25 So, you may want to start matching on industries, 0183 1 as well, in order to really begin to say, when I get to the 2 end of this pilot, I have controlled for price, I have 3 controlled for market cap, I have controlled for information, 4 and now the differences are truly due to the change that I 5 made in tick size. 6 If you do that, you need a lot of stocks, and I 7 completely share the view that, unless it's for a year, the 8 cost to people to change their systems and keep track of all 9 this is just too large. You just won't get people, you know, 10 participating the way you want. 11 So, I mean, it's certainly doable. Doing match 12 studies like this is something that all of you are very good 13 at. 14 So, it just, though, that it's -- it's not trivial 15 to think about this carefully and plan it in front. 16 MR. LEWIS: I'd like to add just one thing for you 17 to consider, as well, from the perspective of someone who has 18 actually spent a fair amount of time thinking about how to 19 design pilot programs, and that is, when you have a -- I 20 asked everybody for what they thought the market failure was, 21 right, and we heard a number of different alternative 22 interpretations that characterize what we should be trying to 23 look at in a pilot study, and I understand that you can have 24 a very focused pilot study that takes on one of those 25 concerns; you could have a more involved, complicated, pilot 0184 1 study that's simultaneously designed to try to address a 2 number of these different issues. 3 But bear in mind that every time you bring a new 4 issue on the table, you're effectively constructing a brand 5 new sample. 6 Now, that also has to be considered against the 7 backdrop of the fact that, at the Commission, we have a 8 number of other pilot studies that are already underway, and 9 so, when you start to think about how many available 10 securities there are to use for a pilot study, you have to 11 worry about the possibility of cross contamination, that 12 what's being controlled for by using it in one pilot study 13 has to be considered how it affects an alternative pilot 14 study. 15 So, it's important to keep the focus precise, so 16 that you understand exactly the nature of the question you're 17 trying to answer. 18 MR. JIGANTI: Maureen hit it right on the head. 19 With the 16 different combinations -- and I would actually go 20 to 32 and add a couple of price points in there. You know, 21 we've avoided the number. 22 I think a two-cent pilot in some instances makes 23 sense and a nickel pilot, both, you know -- so, if we had 24 your 16, we spread them over those -- turn them into 32 -- 25 could end up with a pilot program with -- with real data, 0185 1 with price point differences, to measure against each other 2 in the same market conditions, which brings me to -- Kent is 3 absolutely correct. A year isn't long enough. 4 The liquidity providers will jump in one day one if 5 they see an opportunity that makes it work, but the -- but if 6 you want to get more research people in -- I don't know how 7 many people here have hired anyone recently, but it doesn't 8 take -- it doesn't happen in a day, and so, I imagine that 9 it's going to -- it's going to keep -- it's going to keep the 10 doors closed for a while. 11 So, if you're looking to see if the research 12 departments get to be greater, it's got to be a minimum of 13 two years, in my mind, before you start seeing any real 14 change out there. 15 MR. REED: Just to add one thing, the last panel 16 was concerned about the idea of opting in. If we had an 17 opt-in feature to the pilot, that reduces the sample size 18 significantly there and makes all these concerns even more 19 problematic. 20 One other thing I just wanted to bring up -- I 21 don't think I have a good solution to this, but if one of the 22 goals of the pilot is to think about capital formation, then 23 effectively you're trying to run a pilot for companies that 24 have not yet entered the market, and so, you'd need some way 25 of exogenously determining which firms would be allowed to 0186 1 have high tick sizes and low tick sizes, and it seems like it 2 would raise a lot of procedural issues for the SEC. 3 MR. HATHEWAY: Just to offer my two cents on this, 4 I don't think we'd necessarily need 32 partitions of 5 potential sample, taking everything outside the Russell 3000, 6 taking every nth stock is a simple enough way, and on a 7 mixing perspective, you should get a lot of the various 8 attributes you're looking to cover covered. 9 I do think Paul has a very good point, that there 10 should not be one tick for the pilot. Having multiple ticks 11 within the pilot stock allows studies to be done when 12 individual stocks cross price points. It allows us to get 13 some familiarity with what calibrations may be, and it also 14 avoids placing one-size-fits-all regime with a 15 two-possible-sizes-fits-all regime, which is not necessarily 16 all that better. 17 I think ideally, at a minimum, we need to do, as I 18 said, the stocks sort of outside the Russell 3000. There's 19 no index issue. These are essentially stocks that are not 20 the most liquid in the world. 21 Ideally, we would take a look at larger stocks, as 22 well. There are investor costs associated with a tick too 23 wide. I've been in here in the past before Citi split to 24 point out that essentially professional money was trading on 25 a half-cent spread and non-professional money was trading a 0187 1 few tenths of a cent or hundredths of a cent inside the full 2 penny spread, and that's not necessarily clear that the 3 individual investor, in cases like that, is getting the best 4 possible deal. 5 MR. LEWIS: It's a very naïve question, but I'm 6 wondering whether issuers and/or investors might put up a 7 stink about us allocating them to two cents versus five 8 versus 10. I mean, can we imagine that they would not like 9 the -- 10 MR. JIGANTI: It can't be as bad as the short sale 11 pilot. I don't think many -- you guys have to tell me, but I 12 have a feeling there were firms who were not happy about 13 losing their tick test for a study. 14 MS. EDWARDS: One quick question before we move on 15 to costs and benefits, and we'll probably address your point 16 again, Kent, in talking about costs and benefits, but there's 17 several ways that a pilot can occur. 18 Either the SEC can do a pilot such as the 19 Regulation SHO pilot or the exchanges themselves can do a 20 pilot through an NMS plan such as limit-up/limit-down pilot. 21 Which way would be the best way to go forward if a 22 tick test pilot was appropriate? 23 MR. HATHEWAY: Plans are becoming increasingly 24 popular. I suppose you may have -- it's a way to take 25 commission, share scarce resources, and leverage them in 0188 1 terms of getting things designed and implemented, I'm just 2 guessing. 3 I would guess, with a plan, there's probably a 4 little bit longer time to get it done. You've got a whole 5 bunch of SROs in the room who have to agree on something. 6 Presumably people other than the SROs would be represented on 7 advisory boards like we do with the data plans. 8 I've thought the same question, and I think, in 9 this case, given what we're already asking the SROs to do 10 with other plans, and we're getting some experience in 11 working with each other on other plans, and less 12 contentiously than recent history with the data plans, this 13 may well be a good place to try a plan. 14 If we can get limit-up/limit-down done with a plan 15 -- we haven't got there yet, but if we should, then this 16 would be something we could do with a plan, though, again, we 17 haven't discussed that internally, so I'm speaking mostly for 18 me. 19 MS. SEIDEL: Before we leave the design, I just had 20 one, further question for the panel. There's been a lot of 21 discussion about how we might decide what securities would go 22 into a pilot or have a pilot, or the goals that we might want 23 to measure during a pilot. Were we to consider a pilot, 24 should we be giving any consideration now, at this point, as 25 to sort of is there anything thought we should give to 0189 1 whether or under what circumstances securities would come out 2 of a pilot? Or, is that something we should think about when 3 we're thinking about the design of the pilot from the start; 4 or, once we're ready to have a pilot, once we actually got to 5 see some data, see what's going on? Just sort of interested 6 in views on that. 7 MS. MCCARTHY: Are you talking about an opt out or 8 are you talking about, now, the determining when this -- 9 MS. SEIDEL: Sort of whether or not we should be 10 giving thought -- 11 MS. MCCARTHY: They pass the metric. 12 MS. SEIDEL: -- right. Is there some parameters, 13 what might be the parameters, at what point consideration 14 should be given. Should there be, you know, for some 15 standards in place where securities will come out of the 16 pilot. 17 MS. MCCARTHY: Right. Our feeling is, I guess, 18 I'll answer both questions. Our feeling is for after a 19 statistically significant period of time that companies 20 should be allowed to opt out for whatever reason, if they 21 determine it's harmful to them in any way. But, separately, 22 I think you're also asking if it moves out of a market cap 23 metric into another one, what's the situation there. I think 24 it depends on the time design of the trial. If it's a year, 25 no. If it's a much longer study than a year, then I think 0190 1 one has to determine what happens once they moved 2 significantly out of the metrics. They got them into the 3 trial in the first place. 4 MR. SACHS: And, again, I would echo that. 5 Theoretically, I can't quite get my head around the opt in, 6 opt out, and how that would actually affect a study like 7 this. I think there's too many variables to control for 8 there, but, ultimately, yes. I think if you're going to move 9 stocks out of the pilot program, I think very 10 straightforward, quantitative, simplistic approach of market 11 capitalization and/or something else, as few variables as 12 possible, I think, to move something out of the program. 13 I think, depending on what the length of the time 14 of the program is -- and I do think longer in this particular 15 program is better than shorter -- I think you're going to run 16 into issues of the data being subject to too much analysis if 17 you have a large number of stocks moving out of the program. 18 And, then, what replaces those? What sort of 19 market regime are you in when those stocks come in and go 20 out, because everything we've talked about here from a design 21 perspective is extremely important, not the least of which is 22 the industry or sector the security is in. Because, again, 23 depending on the length of time -- but over a multi-year 24 period, obviously -- macro economic regimes are going to 25 change and certain stocks are going to start behaving 0191 1 differently as flows move towards those stocks and then flows 2 move out of those stocks, or, you know, the sectors. Money 3 rotates through the sectors and out of the sectors. 4 MR. JIGANTI: I think that if the stock -- one way 5 out would be to have the stock at the S&P 500. Some major 6 event would be an obvious opt out that would have enough 7 natural liquidity that would come from index funds that would 8 be trading. That becomes more of an every-day stock. That 9 would be one way out. The other is a defined end to the 10 pilot program -- you know -- if it's a three-year program 11 that is actually a three-year program and it ends with a set 12 date. 13 MR. REED: I would just add that from the 14 perspective of pure research the opt in and opt out are 15 similar issues and they have sort of androgyneity. You'd be 16 concerned about the particular type of firm choosing the opt 17 out, because of its circumstances. You wouldn't necessarily 18 be able to tell the effect of the pilot study if there was 19 this ability to opt out. 20 The other thing I'll just mention is that depending 21 on the fixed criteria that you use to allow a firm to 22 graduate or to move out of the pilot, you'd want to be 23 careful about that, and just allow for the possibility that 24 some metrics might be gained in order to effectively opt out. 25 So you'd just want to design those in ways that are difficult 0192 1 to challenge them as exogenous as possible. 2 MS. HANLEY: So I wanted to ask a question about 3 potentially a non-pilot. So does any one thing that perhaps 4 I've heard that issuers -- issuers in the last count -- 5 issuers would be happy to choose and take sides. Would it be 6 perhaps better to go to a regime in which instead of having a 7 pilot we allowed companies to choose their own tick size and 8 do it that way? 9 MS. MCCARTHY: I guess I think two things there. 10 One is I think I like the competitive, natural selection 11 there; however, I do think we're asking companies to consider 12 something about which they know less than we do on, which is 13 market structure, and the nuance is a market structure. I 14 think -- we think we would be better if we told companies 15 whether they are in or out, but as I said, I do like the 16 natural selection of that. So that's a tough one. 17 MR. HATHEWAY: Yeah. While issuer choice is very 18 appealing, I do echo Maureen's concern that it's not a lot of 19 information the firms would have for the basis of making a 20 decision. It may well be to an outcome where firm's just do 21 what everybody else does, and we don't really gain much 22 information. Issuer choice may be a possible end state, 23 particularly if the results are sort of it depends, if we 24 actually do a pilot. But I'm not sure if we would be able to 25 gain information from issuer choice, if that were the pilot 0193 1 itself. 2 MR. CLARK: The issuer choice exists today. I 3 mean, right, so all issuers can choose now to reduce their 4 stock price to effectively widen their spread and, I guess, 5 it's an option that exists today. We could try to educate 6 them better on that process; but, ultimately, what you're 7 going to end up doing is just every one of these companies 8 will be trading at below $10. And is that the outcome that 9 we're looking for? 10 MS. HANLEY: Can we move now to some costs 11 associated with the pilot? So, operationally, those of you 12 who would be involved in this, operationally, what would be 13 the cost this pilot? Would it be difficult to implement a 14 tick size change? 15 MR. CLARK: So I've talked to our technology guys, 16 because this is essentially something we do in Europe 17 managing multiple ticks, and it's not that different 18 technology between U.S. and Europe. From our perspective as 19 an exchange operator, it's not a particularly costly 20 implementation. I think the bigger costs will be probably 21 borne by the participant firms, particularly those that have 22 a large client face that's not used to seeing anything other 23 than pennies. 24 MR. JIGANTI: It's a good segue into we can make 25 the change. When options are traded in multiple different 0194 1 price levels, it was confusing going in. It's really more of 2 an educational process cost than the actual program costs; 3 you know. Every time you make those minor changes, there's 4 obviously some cost to the industry of potential hazards of 5 changing code. But since we do it every day, I think, 6 ultimately, that's not a very large cost. It's really 7 education. 8 MR. SACHS: From the investment management 9 perspective, I see little to no significant costs. You know; 10 obviously, from an operational perspective, most of that 11 would be borne by data providers, broker-dealers, exchanges. 12 Certainly, there would be costs from a time and resource 13 perspective in transaction cost analysis and reporting, and 14 the way in which we look at transactions in stocks in the 15 pilot program; but I would suspect that those costs would be 16 minimal. 17 MS. HANLEY: What about other types of costs, so 18 unintended consequences. There can be, as Maureen pointed, 19 sometimes winners and losers in this. And so how do we 20 balance that? I know we had some experience in the reg show 21 pilot but how do we balance that, particularly in stocks that 22 are what we would call, some have said here, fragile. 23 The eco system generally, we're only changing one 24 aspect of the eco system. We're not changing the many things 25 that people have talked about here, and that these stocks are 0195 1 already facing concerns about trading that the effect on them 2 may be greater, let us say, than if they did this to a very 3 large company. So if you could talk a little bit about some 4 of the potential, unintended consequences that could occur, 5 or areas which we're particularly mindful, so that we don't 6 further harm a market that others have suggested as 7 struggling. 8 MR. HATHEWAY: Well, quoting algorithms -- and 9 everybody now quotes algorithms now -- are expensive to cope. 10 There's not a lot of trading volume, and some of the stocks 11 we're talking about. And one potential outcome is you lose 12 liquidity providers, because they just choose not to recode 13 their systems. And that's a cost. It damages the pilot and 14 potentially, negatively impacts investors if you get enough 15 liquidity providers opting out. 16 There's certainly going to be some investors who 17 will pay more. I mean averages are averages. There are 18 going to be some investors who pay less. The goal is that 19 investors pay less and they have better access to small 20 growth companies. That's harder to quantify, and the welfare 21 arguments around who's paying more and who's paying less is 22 particularly difficult to quantify. 23 I mean one suggestion I made on an earlier panel 24 was that if you narrow ticks in some stocks, investors as a 25 class will be receiving the benefits from presumably higher 0196 1 volume stocks with a narrower tick and incurring some of the 2 costs associated with less active stocks with water ticks. 3 Clearly, there were members of some of the earlier panels who 4 were not receptive to narrower ticks and active securities. 5 MR. NALLENGARA: We've identified the market 6 failure, or at least we've talked about it; but this panel 7 was designed to talk about a pilot. But are there other -- 8 as we consider it and we look at costs and benefits -- we'd 9 also consider what other alternatives there would be? Are 10 there other market incentives that could be, should be 11 considered other than a pilot that any of you have thought 12 about? I'm just wondering if there's something other than 13 ticks that should be looked at in considering this market 14 failure. 15 MR. CLARK: I think this is going to be a very long 16 panel if we go there, but I think that, certainly, we're 17 cognizant that there are a lot of issues in the marketplace, 18 and this is not a miracle cure to resolve the liquidity 19 challenges of the small cap market. I think in conjunction 20 with this pilot, we need to continue to move forward with a 21 holistic review of overall market structure, various 22 dimensions of it, which also could improve the environment in 23 a less active securities as well as more active securities. 24 Obviously, there's a lot of topics out there related to 25 maker-taker models related to fragmentation, related to other 0197 1 types of, maybe, market maker incentive programs. You know. 2 We're working with the Commission now on a program in the ETF 3 space where the issuer is providing payment to the market 4 maker to meet certain performance obligations. And I think 5 that's an interesting pilot. Well, it's not a pilot yet, but 6 it has interesting potential; you know, if the results prove 7 interesting there. You know. That's something that could be 8 extended into the smaller cap space, for example. 9 MR. SACHS: Yeah. I would echo that. I mean, 10 obviously, stating the obvious here, I mean, really again it 11 depends on what problem we're trying to solve. You know. If 12 you're looking at the lack of IPO activity and/or the 13 hampering of the research efforts from the broker-dealer 14 community, I think from a market structure perspective there 15 are far less things that you could do or leverage you could 16 pull to have immediate impact. 17 If you're looking solely at liquidity and the 18 quality of market, if you will, and measuring that by some of 19 the metrics that we've spoken about here earlier, depth of 20 book, average size of trade, you know, overall liquidity 21 perspectives, then yes, I do think that there's probably a 22 fairly long list of things that have been discussed at length 23 from the maker-taker model to market maker obligations, all 24 sorts of things. 25 But, again, to echo what was just said, I do argue 0198 1 that if we have the ability to take a step back and look at 2 the entire market structure eco system as opposed to looking 3 at individual items that are on the list -- such as spreads 4 and small cap stocks and the maker-taker -- and really look 5 at it holistically, obviously, understanding the realities of 6 the world we live in, I do think we ultimately would end up 7 in a better place from a market structure perspective as 8 opposed to handling the situations in a piecemeal basis to 9 react to certain perceived disruptions or failures in the 10 marketplace. 11 MR. WOMACK: Professor Jay Ritter from Florida 12 wrote a very wise article that I think was showcased here a 13 few months ago about re-energizing the IPO markets. And I 14 think I would encourage you to go back and think about some 15 of the issues that he brought up; number one, reducing the 16 overall costs of going public. You know. This is a market 17 where despite contraction in secondary costs, the primary 18 cost of going public is still very large. You're talking 6 19 to 7% plus, typically 10 to 15% for the best deals. 20 That's a very large cost to the issuer, and I think 21 that belongs on the table too, if we are thinking about what 22 energizes investors and issuers to want to be back in this 23 IPO market again. So I wouldn't forget to go back and look 24 at some of the things that Jay twisted our arm about. 25 MR. REED: Also, and in some of Jay Ritter's slides 0199 1 related to the same paper, he mentions this idea of research 2 being paid for by the issuing firms. I think we were talking 3 here about complicated linkage between tick sizes and 4 eventually the information environment in which a firm 5 operates. I think if the firm really does value that 6 information environment, perhaps we should at least consider 7 mechanisms that allow firms to pay for that information 8 directly, instead of hoping that market maker profits will 9 generate research and that that will generate liquidity. 10 MR. BERMAN: Can I ask a follow-up question based 11 on the theme that Lona was picking up on and I think that, 12 Stephen, you and a few others were reacting to? The market 13 is very complicated. People have talked about maker-taker. 14 Both here and also in both of the first two panels, there was 15 also discussions of off-market trading. So are there things 16 that we need to do for market structure purposes to control 17 for in any sort of a pilot? 18 So where I was thinking is if we're looking for 19 improved liquidity, improved depth of book, et cetera, let's 20 assume that we find trading volumes -- average trading volume 21 -- goes up. But it's actually 70% is now off exchange, and 22 the depth of book actually has gone down. What would you 23 conclude from a pilot? Would you say, well, that's actually 24 good? Or, actually, is that bad, because we were looking for 25 depth of book? So are there linkages between off-market 0200 1 trading or over-the-counter trading that we need to control 2 for for the purpose of understanding the pilot? 3 MR. JIGANTI: If that's not real market -- if it 4 goes on the other parts of the marketplace -- if you take one 5 of those off-market trading pieces away, you will affect the 6 pilot in ways that you may not even realize. You know. So 7 many of these things we want to see how they play out in real 8 life if we're going to think about and impose non-realistic 9 limitations to it. You're not going to have the same result 10 that you would. You know; if it turns into a market 11 structure change, it's a different pilot. 12 MR. SACHS: But, again, from the practitioner's 13 perspective, I would argue that the reason that volume moves 14 elsewhere off exchange is in reaction to the market structure 15 in which we're forced to deal with today -- lack of 16 liquidity, flickering quotes, you know, noise if you will, 17 generated by the market structure that we're in. So I would 18 argue, while I'm guessing it wouldn't be at all popular, 19 controlling for all those factors and only allowing for, say, 20 true mid-point, dark crossing at the middle of whatever the 21 pilot spread is. 22 You know, it eliminated control for many of those 23 factors, so you would still allow for, again, large orders to 24 cross in a dark pool or what have you in the middle of the 25 spread, but you wouldn't eliminate all of the other sort of 0201 1 ancillary noise that you get from small price improvement -- 2 you know, tenths of a penny, that sort of thing. And, again, 3 obviously, each constituent represented here, whether it be 4 the exchange or the broker-dealer community, or the 5 investment management community, it's obviously going to have 6 a greatly different opinion of that. 7 MR. HATHEWAY: Yeah. I think, well, two responses: 8 First place, that's one of the things a control does, that 9 controls stocks, is lets you get a sense of what's happening 10 elsewhere in the market for stocks not in the pilot; the 11 other thing, the goal is depth. The goal is to improve 12 liquidity. The goal is to lower cost of capital and a more 13 attractive environment for small companies. If a larger, 14 wider display tick were to result in all these other things 15 happening, then it's probably a good income. The locus of 16 trading, per se, and competitive dynamic between NASDAQ, NYC, 17 Broker Systems, BATS, Direct Hedge, you know, their 18 exchanges, is probably not a critical point. So if a wider, 19 not particularly used displayed quote allows people to have a 20 better sense of where the price really is to get size done, 21 to get retailers done more effectively, then all in all 22 that's a good outcome. 23 And speaking now as an exchange rep, we would be 24 back here saying, hey, what these guys are doing in the dark 25 is the right way to trade with inside a displayed quote, and 0202 1 we want to replicate those mechanisms. But taking a step 2 back from that, the scenario you described could be good or 3 ill, depending on the other attributes you've already talked 4 about. 5 MS. MCCARTHY: Yeah. I agree with that. I think 6 that if you were to do what Frank said and you allowed 7 trading, just allow trading at the midpoint, you get around a 8 few things, which is the dark pools can still participate. 9 The opening cross would be critical for that, however, I do 10 think you're starting to add some elements and maybe getting 11 away from what is the end point, and maybe that's a little 12 tougher to design. 13 If you were to say that nothing could be traded in 14 the middle, I don't think that -- now I think we're into a 15 completely different study. And I'm not arguing that it 16 shouldn't be done. I'm just saying I don't know if it's in 17 -- well, without becoming a completely different trial design 18 here, and I guess that's it. 19 MR. CLARK: And I just add to that. I think the 20 study is to explore the impact of changes in tick sizes; and, 21 certainly, there should be a control group that's just 22 focusing on that change. If you didn't want to add other 23 elements to it, you could have another separate control group 24 that's looking at tick size change, plus elimination of some 25 pennies, or whatever that other variable is you want to look 0203 1 at. And that way you could look at the data to see what the 2 impact of that element is; but I think you don't want to lose 3 sight of -- just focus on -- the tick size in and of itself. 4 MR. BURNS: I guess the question I had, which is 5 quite naive, is if we were to put rules in place for a pilot, 6 is it possible that investors and/or market makers can just 7 reverse engineer around it and make it a bit null and void in 8 terms of what our true goal is? I guess Frankie or somebody 9 else would be far better at thinking about such things than 10 I. 11 MR. LEWIS: So, Frank, could I jump in here real 12 quick and sort of add something on top of that? And it goes 13 back to the point I think Greg was making. And, that is, if 14 you were going to design a pilot, and we have to consider the 15 unintended consequences of the pilot, we ought to consider 16 where order flow is likely to go. Right? So discussion now 17 is assume a static framework where there is no movement. 18 Will transactions continue to be conducted in 19 markets at the same rate, or will people be incentivized to 20 actually take their trades into dark pools? And I think that 21 goes to the arbitrage question that Kent has raised. 22 MR. HATHEWAY: In illiquid stocks, roughly, half 23 the volume is probably going to be in dark pools already, our 24 internal safety platforms of one type or another. So could 25 you move more? Yeah. Of course, it's still not 100. But to 0204 1 Kent's point about "would people try to reverse engineer it," 2 yeah. They would. This is one of the fascinating things 3 about this industry. You show people where lines are and the 4 immediately start proving. Exactly where the line in fact 5 is, through all the steps the SEC took regs show, there were 6 things that people came up with to get her on locate, to get 7 her on some of the other things that people did. 8 And I think Greg's show is many respects 9 illustrative of when the Commission has a commitment to a 10 goal and takes step by step processes to get it done, they 11 get it done. It may not be, sort of, day one; but, you guys 12 would really need it to be on making sure it's selling and 13 similar issues underlying reg show. Would people reverse 14 engineer? Yes. Would there be a change in the dynamic of 15 how off exchange trading operates? Yes. How is that going 16 to come down? These are not profitable stocks to trade. 17 That's the bottom line on these securities. So 18 when we were talking about the high frequency trading firms, 19 they're not active in these securities. It's essentially 20 various people's algorithms trying to get a trade done or 21 trying to get somebody to get a trade done at a price they 22 may not like to expost. Executors were hearing about the 23 forced asset share trader earlier this morning. Is it going 24 to change? Yeah. But I think the whole point of a pilot is 25 to find out what net-net we accomplish in terms of creating a 0205 1 more attractive ecosystem for companies whose behavior over 2 the last 12 years is clearly indicated; but, public equity is 3 just not a very attractive place for them to be. 4 And that, as just pointed out in the earlier 5 panels, has implications for entrepreneurship. It has 6 implications for the type of companies to get started. It 7 has implications for the ability of investors to participate 8 in potential, high growth companies. Risky, yes, but with 9 these companies not being a part of the companies now, you're 10 closing investors off to the opportunities to participate in 11 them. So I think it's a learn from a pilot, Maureen's point 12 about seeing more than the displayed information for the 13 markets is well taken. 14 What are people doing that doesn't necessarily get 15 displayed in our quotes is a relevant part of this today that 16 really wasn't that big a part of the story 12 years ago. The 17 world was different when we did decimalization. 18 MR. BURNS: Well, irrespective of where we go with 19 the pilot, I hope the stringers were all listening when Frank 20 said we were a "get it done" kind of agency. And, once 21 again, like the technology roundtable this past fall -- for 22 my part, parochial interest -- I think the quality of the 23 panelists and the discussion today was superb. We thank you 24 so much for the time you've taken with us; and, doubtless, 25 you and others will be coming and visiting and making 0206 1 submissions, which we welcome very much in coming weeks. 2 Thank you very much and thank you to my colleagues, 3 the staff here, the Commissioners for permitting us to hold 4 this roundtable, and for all of you. And on behalf of Greg 5 and Lana, and all of us here, thank you very much. 6 (Whereupon, at 4:03 p.m., the roundtable was 7 concluded.) 8 * * * * * 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0207 1 PROOFREADER'S CERTIFICATE 2 3 In the Matter of: ROUNDTABLE ON DECIMALIZATION 4 File Number: OS-4-657 5 Date: Tuesday, February 5, 2012 6 Location: Washington, D.C. 7 8 9 This is to certify that I, Maria E. Paulsen, 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 0208 1 REPORTER'S CERTIFICATE 2 3 4 I, Jon Hundley, reporter, hereby certify that the 5 foregoing transcript of 206 pages is a complete, true and 6 accurate transcript of the testimony indicated, held on 7 February 5, 2013, at Washington, D.C. in the matter of: 8 ROUNDTABLE ON DECIMALIZATION. 9 10 11 I further certify that this proceeding was recorded by me, 12 and that the foregoing transcript has been prepared under my 13 direction. 14 15 16 17 Date: __________________________ 18 Official Reporter: __________________________ 19 Diversified Reporting Services, Inc. 20 21 22 23 24 25 0209 1 2 3 4 Diversified Reporting Services, Inc. 5 1101 Sixteenth Street, N.W. 6 2nd Floor 7 Washington, DC 20036 8 9 10 In the Matter of: ROUNDTABLE ON DECIMALIZATION 11 File Number: OS-4-657 12 Date: Tuesday, February 5, 2012 13 Location: Washington, D.C. 14 15 This is a letter to inform you that we do not 16 release our tapes and notes. I do maintain 17 them for a period of one (1) year. 18 19 Sincerely, 20 _________________________ 21 22 23 24 25