0001
 1           U.S. SECURITIES AND EXCHANGE COMMISSION 
 2   
 3   
 4   
 5   
 6       ROUNDTABLE ON MARKET STRUCTURE FOR THINLY-TRADED 
 7                          SECURITIES 
 8               DIVISION OF TRADING AND MARKETS 
 9                                
10   
11   
12   
13   
14   
15   
16                  Monday, April 23, 2018 
17                          9:28 a.m. 
18   
19   
20   
21   
22   
23           U.S. Securities and Exchange Commission 
24            100 F Street, N.E., Washington, D.C. 
25              Station Place 1 Multipurpose Room 
0002
 1   PARTICIPANTS: 
 2   Robert Jackson, Jr. 
 3   Kara Stein  
 4   Jay Clayton(Chair) 
 5   Michael Piwowar 
 6   Hester Peirce 
 7   Jeffrey Harris 
 8   Keegan Murphy 
 9   Alex Jadin 
10   David Saltiel 
11   Brett Redfearn(Moderator) 
12   David Shillman 
13   John Roeser 
14   Dan Gray 
15   Jason Vedder 
16   Ari Rubenstein 
17   Ovi Montemayor 
18   Frank Hatheway 
19   Bryan Harkins 
20   Brian Fagen 
21   Adam Epstein 
22   Steve Cavoli 
23   Tal Cohen 
24   Chris Concannon 
25   Stacey Cunningham 
0003
 1   PARTICIPANTS(CONT.): 
 2   Brian Frambes 
 3   Brad Katsuyama 
 4   Joe Mecane 
 5   Robert A. Schwartz 
 6   Owain Self 
 7   David LaValle 
 8   Phil Mackintosh 
 9   Laura Morrison 
10   Greg Sutton 
11   Charles Thomas 
12   Kumar Vekataraman 
13   Doug Yones 
14    
15    
16    
17    
18    
19    
20    
21    
22    
23    
24    
25    
0004
 1                    C O N T E N T S 
 2                                                        PAGE 
 3   Call to Order and Opening Remarks                       5 
 4    
 5   Panel 1:  Challenges in Market Structure Performance   17
 6   Of Thinly-Traded Exchange-Listed Equity Securities 
 7    
 8   Panel 2:  Potential Improvements in Market Structure   93 
 9   For Thinly-Traded Exchange-Listed Equity Securities      
10          
11   Panel 3:  Market Structure Performance and Potential   176
12   Improvements for Thinly-Traded Exchange-Traded  
13   Products(ETPs) 
14    
15   Concluding Remarks                                     231    
16    
17    
18    
19    
20    
21    
22    
23    
24    
25    
0005
 1                     P R O C E E D I N G S 
 2        MR. REDFEARN:  Good morning.  Thank you all very 
 3   much for joining us today.  I'd like to welcome you to 
 4   the Securities and Exchange Commission's Staff Roundtable 
 5   on Market Structure for Thinly-Traded Securities.  I'm 
 6   Brett Redfearn, Director of the Division of Trading and 
 7   Markets.  And I'll be moderating our panels today, along 
 8   with my fellow staff members from the Division.   
 9             Before I begin, I'd like to provide our 
10   standard disclaimer, which is that any views I or the 
11   staff express today are my own or our own and not those 
12   of the Chairman, the other Commissioners, or the 
13   Commission. 
14             As you may know, this is the first roundtable 
15   we're going to do in a series of roundtables looking at 
16   equity market structure topics.  Today, in this 
17   particular one we want to talk about the fact that we 
18   basically have a single equity market structure for NMS 
19   Securities, large and small, liquid and illiquid.   
20             As Chairman Clayton recently emphasized, we 
21   really need to examine whether the current market 
22   structure is appropriate for all types of companies and 
23   all types of investors.  Today's roundtable focuses on 
24   whether the market structure is appropriate for thinly-
25   traded exchange listed equity securities, including both 
0006
 1   common stocks and exchange traded products. 
 2             I am convinced that the thinly traded segment 
 3   of our equity markets has not received the attention that 
 4   it deserves.  In general, securities with lower volumes 
 5   exhibit wider spreads, less displayed size than higher 
 6   volume securities, and that can lead to higher 
 7   transaction costs for investors.   
 8             In addition, institutional investors that need 
 9   to trade in large size may be deterred from participating 
10   in this segment of the market because of a small float or 
11   an inability to take sizeable positions in a name without 
12   incurring higher transaction costs. 
13             Thinly traded securities and investors deserve 
14   a market structure that fits their particular needs.  
15   This concern has been raised in a number of contexts over 
16   the past few years, and there's clear interest in ways of 
17   address this.  The recent Capital Markets Report from the 
18   Department of Treasury highlighted the challenges that 
19   face buyers and sellers when a relatively small volume of 
20   trading is spread out among many different venues, 
21   stating that quote, "The primary functions of markets is 
22   to facilitate the meeting of buyers and sellers, but with 
23   so little volume spread across so many venues, finding 
24   the other side of a trade has become harder."   
25             The Treasury Market Capital -- the Treasury's 
0007
 1   Capital Markets Report -- recommended that the SEC 
 2   considered allowing issuers of less liquid stocks to 
 3   suspend unlisted trading privileges for UTP -- or UTP, 
 4   for a stock, so that it would trade on a smaller number 
 5   of venues until liquidity reached a minimum threshold.  
 6   To maintain a basic level of competition and because of 
 7   the significance of the upstairs, negotiated and 
 8   wholesaler markets in these names, the Capital Markets 
 9   Report recommended that broker internalization should 
10   remain as a trading option for all stocks.   
11             In addition, market participants have been 
12   interested in the suspension of UTP, and we've had 
13   various forms of outreach on that specific matter.   
14             Today's roundtable is going to consist of three 
15   different panels.  The first panel will address 
16   challenges in Market Structure Performance for Thinly-
17   Traded Common Stocks.  Among other questions we will 
18   explore, what challenges do exchanges face in maintaining 
19   fair and orderly markets in thinly-traded common stocks, 
20   and why does trading in these stocks tend to occur off-
21   exchange?    
22             The second panel we'll have today will explore 
23   potential improvements in market structure for thinly-
24   traded common stocks.  And we will discuss whether 
25   limiting UTP would provide a better opportunity for 
0008
 1   exchanges to innovate to serve less liquid securities.  
 2   We will also discuss whether any solutions would need to 
 3   impose restrictions on off-exchange trading to succeed, 
 4   and how to address concerns about possible monopolistic 
 5   pricing if competition is excessively restricted. 
 6             The third and final panel we'll have today will 
 7   address the unique set of issues that are presented when 
 8   examining thinly-traded exchange-traded products.  We 
 9   will explore how market structure considerations for ETP 
10   is different from those of common stocks, and whether 
11   solutions for stocks may also be solutions for ETFs.   
12             Our panelists today represent a wide array of 
13   industry participants, and I'm looking forward to hearing 
14   what they have to say about these important issues.  In 
15   connection with this roundtable, the Office of Analytics 
16   and Research in our division has conducted an analysis of 
17   liquidity demographics and market quality for thinly-
18   traded NMS stocks.  A paper detailing this analysis is 
19   available on our website, and provides some context for 
20   our discussion today.   
21             In second, I'm going to ask Alex Jadin in our 
22   office to present the paper and the findings.  But first 
23   I want to welcome all of our panelists.  Thank you very 
24   much for your participation today.  We're very interested 
25   in hearing your views.  And in particular I'd like to 
0009
 1   thank Commissioner Stein and Commissioner Piwowar for 
 2   joining us this morning as well.  Thank you both.  So 
 3   with that in mind, let me turn it over to Alex.  And 
 4   Alex, take it away. 
 5             MR. JADIN:  Thanks, Brett.  And thank you also, 
 6   Commissioners and guests, for your participation today.  
 7   I have to supplement Brett's disclaimer.  The comments 
 8   I'm going to give today are based on a paper produced by 
 9   the staff of the Office of Analytics and Research.  And 
10   the Commission has expressed no views on the findings, 
11   analysis or conclusions in that paper. 
12             Our paper is intended to provide a basic 
13   characterization of thinly-traded equities.  It's worth 
14   noting at the outset that we don't, for example, engage 
15   in any hypothesis testing, attempt to establish any 
16   causal relationships or any historical trends.  Rather, 
17   we're attempting to sort of provide a snapshot from a 
18   recent point in time, the fourth quarter of 2017, for all 
19   NMS stocks with a focus on thinly-traded equities.   
20             The analysis included any stocks that were 
21   listed ineligible for trading for at least one day during 
22   the period. So we looked at slightly more than 8700 
23   securities, even though not quite that many tickers are 
24   eligible for trading and quoting any, on any given day.  
25   And in addition to common stocks and ETPs, we looked at 
0010
 1   the less prevalent kinds of stocks, including preferreds, 
 2   rights, warrants, ADRS, when-issued securities; 
 3   everything that's in the NMS base.  The population of 
 4   symbols of course during this period included anything in 
 5   the Tick Size Pilot, whether it's in the control group or 
 6   in any of the test groups. And while we haven't broken 
 7   anything out in the paper for the tick-size pilot, we did 
 8   run some, some basic checks for some of our key 
 9   variables.  And the -- our -- the results didn't change 
10   dramatically in any given direction.  And where they did 
11   change, they sort of moved in different directions.  And 
12   so it didn't seem necessary to break it out specifically 
13   for the tick-size pilot. 
14             For most measures, we took an average by symbol 
15   and day across the entire period.  So for example for 
16   trade count, we would take the total number of trades per 
17   day and average it over the period by symbol.  For a few 
18   of the measures we just took a single snapshot of a 
19   value, like price or market cap at that point in time, 
20   usually at the beginning of the period. 
21             Throughout the paper, we bend tickers by ADV.  
22   We use three, three bends.  One 50,000 shares or below, 
23   one between 50,000 shares and 100,000, and another above 
24   100,000 shares.  For all NMS stocks, the median ADV was 
25   just under 100,000 shares during our period, and sliding 
0011
 1   that number up to 100,000 for the sake of discussion and 
 2   exposition didn't result in any tickers changing bins.  
 3   So 100,000 seemed to work. And because we're focusing on 
 4   the thinly-traded end of the market, we drew another line 
 5   at 50,000 shares, again, another draftable point for 
 6   discussion. 
 7             As I noted, about half of all -- half of all 
 8   tickers have an ADV of less than 100,000 shares, and as 
 9   Brett and others have noted in speeches, this adds up to 
10   less than 2 percent of all NMS ADV, and the nearly 3500 
11   tickers with an ADV of less than 50,000 together have 
12   about, have less than 1 percent of all NMS ADV. 
13             And to provide a sense of proportion for the 
14   entire distribution of ADV, the top one percent, 88 
15   tickers in our data, together have slightly more than 27 
16   percent of all NMS ADV.  So -- but another way ADV is 
17   heavily tilted towards the more liquid end of the scale. 
18             So that basic framework in mind of the sort of 
19   the population, I'd like to highlight a few data points 
20   about the more thinly-traded end of the market.  If you 
21   are following along in the paper, most of these stats 
22   come from Table 2, which provides most of the stats for 
23   NMS stocks.  So the median ADV for the lowest tier, those 
24   under 50,000, is about, is just under 10,000 shares, and 
25   for the middle group it's about 71,000 shares.  And for 
0012
 1   the upper half, those about 100,000, it's over 500,000 
 2   shares per day. 
 3             At the lower end of the distribution, there's 
 4   about 40, 40 trades per day is the median number of 
 5   trades.  And in the middle, the middle group, it's about 
 6   320.  And it's over 3000 for the highest tier.  
 7             Trading for the more thinly-traded stocks is 
 8   also more concentrated than it is for the more liquid 
 9   stocks.  To measure trading fragmentation, we constructed 
10   a Herfindahl-Hirschman Index, based on the share volume 
11   of each venue.  And for our purposes, we treated anything 
12   reported to the TRF as a single venue consistent with 
13   what market-participants would see on the SIPS. 
14             And so for the values you see in the 
15   fragmentation line of the table, a smaller number 
16   indicates less fragmentation and a larger number 
17   indicates greater fragmentation.  Similarly, as you'll 
18   see in Table 7, the more thinly traded stocks have fewer 
19   exchanges quoting at the inside, as well as higher 
20   percentages of the day when a single exchange is low, at 
21   it either both or one size of the NBBO, and that's in 
22   Table 8. 
23             In addition to the higher concentration of 
24   trading in quota, the more thinly traded stocks have a 
25   greater percentage of off-exchange trading.  For example, 
0013
 1   and again going back to Table 2, for all NMS stocks with 
 2   an ADV under 50,000, a median of 52 percent is done off-
 3   exchange compared to 35 percent for the more liquid 
 4   symbols, and a much smaller percentage of the more thinly 
 5   traded names are done as block trades. Following the 
 6   definition of block size with respect to an order in Reg 
 7   NMS, we treated a block as anything of 10,000 shares or 
 8   greater or $200,000 or more.   
 9             And in the lowest tier, the median was one 
10   percent of volume executed as a block, compared to a 
11   median of 8 percent for the highest tier. 
12             Generally, similar proportions of share volume 
13   are executed in auctions across all of the volume tiers, 
14   about five to seven percent.  But these similar 
15   proportions represent drastically different absolute 
16   measures.  For the most thinly-traded tier, the median 
17   value is only about 500 shares, while it is over 15,000 
18   for the most liquid tier. 
19             Depths at the inside are lower for more thinly 
20   traded names, as Brett mentioned.  Medians for this tier 
21   are anywhere from half to one quarter the size of the 
22   medians for the more liquid range, as shown in Table 9.  
23   For example, for commons in the lowest tier, median 
24   depths at the National Best Bid were about 550 shares, 
25   while for the highest tier it was nearly 1000 shares.  
0014
 1   And for ETPs the comparable measures for 1000 shares for 
 2   the lowest and more than 3500 in the upper, in the upper 
 3   tier.   
 4             Spreads, both quoted and relative quoted are 
 5   wider for the most, for the more thinly-traded stocks, 
 6   which is seen in Table 10.  For example, the median 
 7   quoted spread for common stocks in the lowest tier was 21 
 8   cents, and only 4 cents in the upper tier.  For ETPs the 
 9   difference is 9 cents compared to 2 cents.  And the 
10   median relative quoted spread for common stocks in the 
11   lowest tier was 1.56 percent, but 0.19 percent in the 
12   highest tier.  For ETPs, the comparable numbers were .27 
13   percent and .05 percent in the highest tier.  
14             These spread statistics also highlight the fact 
15   that there are differences between the more thinly traded 
16   ETPs and common stocks.  The following data points are 
17   largely drawn from tables three and four, which break 
18   things out for commons and ETPs.  The median market cap 
19   in the smallest tier for common stocks is 124 million 
20   dollars, while for ETPs it's only 27 million dollars.  A 
21   much larger percentage of ETPs are also in the lower 
22   tiers, as evidenced from Table 1. 
23             About 66 percent of all ETPs have an ADV of 
24   less than 50,000 shares, compared to just over 19 percent 
25   of common stocks, and slightly more than 75 percent of 
0015
 1   all ETPs have an ADV below 100,000, compared to about 28 
 2   percent for all common stocks.  
 3             ETPs have a higher percentage of their share 
 4   volume done off-exchange.  For example, in the lowest 
 5   tier, ETPs have a median off-exchange percentage of about 
 6   62 percent, compared to 41 percent for commons.  And the 
 7   most thinly-traded ETPs have a lower measure of trading 
 8   fragmentation than common stocks, but their quoting 
 9   activity for the most thinly traded ETPs appears to be 
10   more fragmented than for commons, as seen in Tables 7 and 
11   8.  So for example the median number of exchanges quoting 
12   at the inside is larger for ETPs than for commons and the 
13   median percentage of the day when one exchange is alone, 
14   at either or both sides of the NBBO is lower for ETPs 
15   than for commons.  And as described earlier, ETPs have 
16   greater quoted depth at the inside than commons, as well 
17   as tighter spreads on both an absolute and relative 
18   basis. 
19             Finally, I'd like to point you towards the 
20   appendix of the paper, which provides a large set of 
21   visualizations of the basic statistics that we measured. 
22    These plots try to provide a feel of the overall space 
23   rather than provide specific details. They're really an 
24   exercise in exploratory analysis.  And together we hope 
25   that they provide a sense of the distribution of each 
0016
 1   variable, how the distribution of each variable relates 
 2   to overall ADV for NMS stocks, and then it also indicates 
 3   the relationships between each pair of variables, as well 
 4   as how those relationships play out for common stocks and 
 5   ETPs.  We hope that the basic information in this paper 
 6   can provide a useful set of benchmarks for the 
 7   conversation today and for future discussions in the 
 8   area.  We are very eager to receive feedback on the 
 9   paper. We can take feedback in the comment file to this 
10   roundtable and contact information for Keegan and myself 
11   is in the back of the paper. Again, thank you all very 
12   much for your participation today, and we really look 
13   forward to the conversation. 
14             MR. REDFEARN:  Thank you, Alex.  That was 
15   great.  A very, very helpful analysis. And you know, 
16   it's, when we think about a data-driven approach to 
17   anything we do, starting with something like that is 
18   extremely helpful.  It is clearly a different universe 
19   that we're looking at here than when people oftentimes 
20   think about the equity markets as a whole.  And you know, 
21   when I first realized it, you know, it was half of the 
22   market were securities that were trading under 100,000 
23   shares, and you look at the characteristics of this, you 
24   realize we are truly dealing with a different set of 
25   challenges than when we think about the top end of the 
0017
 1   market where the bulk of the volume is.  So thank you for 
 2   that, Alex.   
 3             And before we move on, I first just want to 
 4   introduce the staff that is here with us today, starting 
 5   with on the far left, Jeff Harris, the Director of the 
 6   Department of Economic and Risk Analysis, Keegan Murphy, 
 7   who is a Senior Statistician in OAR.  We already talked 
 8   about Alex.  David Saltiel, Associate Director for OAR, 
 9   Office of Analytics and Research.  David Shillman, 
10   Associate Director for the Office of Market Supervision. 
11    John Roeser, also an Associate Director of the Office of 
12   Market Supervision, and Dan Gray, Senior Market Structure 
13   Counsel in the Office of Market Supervision.  Thank you 
14   all very much for joining us.   
15             So with that, I want to turn to our panelists. 
16    And I think what we're going to do is we're going to let 
17   each of you introduce yourselves briefly.  We'd like you 
18   to tell us a little bit about, you know, what you do and 
19   how you interact with this segment of the market.  And 
20   just a couple minutes of, you know, if you have another 
21   high level thought on what we're doing today.  So Steve, 
22   why don't we start with you? 
23             MR. CAVOLI:  All right.  Thank you very much, 
24   Brett. 
25             MS. STEIN:  You may need to turn your mic on.  
0018
 1   You press the button on the right.  Yeah. 
 2             MR. CAVOLI:  I'm new to this.  Thank you, 
 3   Commissioner Stein.  Again, I'm Steve Cavoli.  I work at 
 4   Virtu Financial. I focus on business strategy. And of 
 5   course, we appreciate the opportunity to share our 
 6   thoughts with you all this morning.  A little, by way of 
 7   introduction, Virtu is a technology-enabled market maker 
 8   and liquidity provider to the global financial markets.  
 9   We provide liquidity in more than 25,000 instruments, 
10   including thousands of ETPs and thinly traded securities. 
11    We do all of this in more than 235 venues and 
12   marketplaces in 36 countries around the globe.  In 
13   addition to the market-making business, we also provide 
14   agency execution services to hundreds of institutional 
15   clients. 
16             As such, we support innovation and enhancements 
17   to efficiency, transparency and fairness that improve 
18   liquidity for all market participants. We strongly 
19   believe that the competition, characterized by choice and 
20   transparency, empowers investors to make better, more 
21   informed trading decisions.  This ultimately results in 
22   broad dissemination of prices and enhances the price 
23   discovery process. With respect to thinly traded 
24   securities, on any given day we trade approximately 12 
25   percent of the volume of NMS securities with an average 
0019
 1   daily volume of 100,000 shares or less.  As a result, 
 2   we're familiar with the space.  It's important to note 
 3   too, as I mentioned earlier, we trade in about 235 venues 
 4   globally, so we believe, so we have experience trading in 
 5   many different market structures, and the opinions that 
 6   we formed around this topic are the result of direct, 
 7   real world and practical experience.   
 8             As I'm sure many of you know, we submitted a 
 9   comment letter on Friday afternoon on the subject.  And 
10   I'd like to take a real quick minute to summarize two 
11   important points we made in it.  The first is that 
12   liquidity in these names is actually pretty good.  Over 
13   time we've observed that liquidity at the NBBO in thinly 
14   traded versus actively traded securities is actually 
15   higher relative to average daily volumes.  For example, 
16   in the 50,000 share or less bucket, liquidity at the 
17   touch, that is stock that one can buy or sell 
18   instantaneously, is about five percent of a day's volume, 
19   while in names which trade over 1.6 million shares, that 
20   number falls to well below half a percent. 
21             And then the second point that I'd like to make 
22   is that investor interest is really the dominant factor 
23   that drives active trading, not market structure.  Market 
24   structure changes won't alter a security's fundamentals, 
25   its float, the diversity of its ownership, its research 
0020
 1   coverage, or cause it to be included in an index.  The 
 2   securities characteristics and underlying interest in 
 3   them is just different.  It isn't wrong and it isn't 
 4   broken.  Again, thank you very much for everything, and I 
 5   look forward to talking about this this morning. 
 6             MR. REDFEARN:  Thank you, Steve. 
 7             MR. CAVOLI:  Yeah. 
 8             MR. REDFEARN:  Adam? 
 9             MR. EPSTEIN:  Good morning.  My name is Adam 
10   Epstein.  In preparation for coming here this morning 
11   from San Francisco yesterday, I spent some time speaking 
12   with a number of small cap CEOs, CFOs and board members, 
13   and they implored me to thank the Commission for hosting 
14   today's event, and I joined them in thanking everyone. It 
15   couldn't be more important to those issuers and it 
16   couldn't be frankly more important to me. I've had a 
17   front row seat to the impacts of small cap trading 
18   illiquidity for the last 15 years, during which time 
19   anecdotally, maybe not statistically, it certainly seems 
20   to be getting worse from my perspective. I'm the founder 
21   of Third Creek Advisors. Since 2010, my firm has provided 
22   a buy-side perspective to CEOs and boards of small cap 
23   companies regarding the resolutely unique challenges that 
24   those companies face day in and day out.  More 
25   specifically, capital markets, corporate finance, and 
0021
 1   corporate governance.   
 2             The buy-side perspective my firm shares with 
 3   Pre-IPO and small cap companies arise out of my seven 
 4   year tenure from 2003 to 2010, co-managing special 
 5   situation hedge funds operated by Enable Capital 
 6   Management, based in San Francisco. 
 7             Enable's hedge funds invested hundreds of 
 8   millions of dollars in more than 500 small cap growth 
 9   financings during my tenure.  From the perspective of an 
10   institutional investor, there is no doubt whatsoever that 
11   for small cap companies in the United States, the ability 
12   to access capital, number one, and the terms of those 
13   financings, is inextricably linked to trading volume.  
14   Subsequent to the financial crisis, I would point out, 
15   that link has become so pronounced that I would suggest 
16   that most of the capital in the small cap equity markets 
17   and terms for those financings are not necessarily for 
18   the best companies but they're for the most liquid 
19   companies.  And unfortunately the opposite is also true. 
20    From an issuer perspective, illiquidity is an austere, 
21   multi-faceted problem.  As a result of my firm's work 
22   with dozens of exchange listed small cap companies over 
23   the last eight years, I see what I would characterize as 
24   the insidious nature of illiquidity on a day to day 
25   basis. And I use the word insidious because small cap 
0022
 1   trading illiquidity affects considerably more than 
 2   capital formation.  Trading illiquidity gravely impacts 
 3   the ability for small cap companies to garner and retain 
 4   research coverage.  Trading illiquidity gravely impacts 
 5   mergers and acquisitions in the small cap ecosystem.  
 6   Trading illiquidity gravely impacts the ability for small 
 7   cap companies to hire and actually retain great 
 8   employees.  It gravely impacts small cap companies' 
 9   abilities to efficaciously deal with customers, vendors, 
10   potential partners.  And trading illiquidity -- and this 
11   is something that CEOs and CFOs implored me to pass along 
12   to the Commission today -- is also responsible for 
13   countless lost hours of productivity for small cap CEOs 
14   and CFOs.  So I join many others in recognizing that an 
15   enormous percentage of high quality jobs and innovation 
16   arised out of the small cap ecosystem in this country.  
17   Rampant small cap trading illiquidity is a dynamic 
18   capital markets problem, true.  But rampant small cap 
19   trading illiquidity also directly impacts how small cap 
20   companies operate. It affects how they operate.  So 
21   consequently, it's also just as impactful, not only on 
22   the small cap ecosystem but on the broader economy in the 
23   United States. And I would also submit that it 
24   jeopardizes America's global standing as the most 
25   innovative nation.  Thank you for having us today.  Thank 
0023
 1   you for conducting the forum. 
 2             MR. REDFEARN:  Thank you very much, Adam.  
 3   Brian? 
 4             MR. FAGEN:  Great, thank you.  I add my 
 5   comments in thanking the Commission for this opportunity. 
 6    It's a wonderful format.  I work at Deutsche Bank, and 
 7   in America my role is the Head of Execution Strategy for 
 8   Equities. In this role, what I largely look at is how do 
 9   we improve our execution capabilities for our 
10   institutional clients?  We are a global institutional 
11   broker; we run a global equities business across trading 
12   sales research. And what we really focus on for this, 
13   especially for this part of the marketplace is, how do we 
14   help our institutional clients' effect of limitation in 
15   this, this range of securities.  This has been a long 
16   sought after and multi-faceted issue.  Some of the 
17   comments from the prior speakers highlighted that. I 
18   won't belabor that point, but what I will say is that, 
19   you know, for us, our, our focus on improving execution 
20   quality is very, very data driven. And what we see in the 
21   data is the fact that we as a broker struggle to provide, 
22   you know, the appropriate form of liquidity, you know, in 
23   a large percentage of the names that are traded in the 
24   U.S. equity marketplace, largely because of the issues 
25   that we are here to discuss today.  Our goal is to help 
0024
 1   to solve that issue.  The institutional marketplace for 
 2   the securities is very -- it's really very, very 
 3   difficult for us to predict, and the needs of our clients 
 4   come in many, many different forms.  They come in forms 
 5   of large asset managers.  They come in the forms of index 
 6   providers, ETF providers, as well as hedge funds, family 
 7   offices and others.   
 8             That puts us in the position of having to 
 9   supply execution in all different forms. Our focus, you 
10   know, for our clients, is to help guide them to the best 
11   possible solution, you know, for all of these types of 
12   executions.  That includes our high touch cash business, 
13   our program trading business, our electronic trading 
14   business and other ways in which our clients choose to 
15   execute.   
16             The, you know, the goal, you know, for any 
17   executing broker is to connect, you know, our clients 
18   from you know, their, their investment decision to 
19   ultimately the liquidity that they seek.  This has 
20   clearly been, you know, a challenge, you know, in this 
21   range of securities, especially at times in which the 
22   markets tend to deal with stress or other portions of 
23   illiquidity that present itself around significant 
24   events.  So what we're hoping to do and what we're 
25   looking for is to create and be part of creating a market 
0025
 1   structure solution that allows all of our clients to gain 
 2   access to this liquidity and for us to help to guide them 
 3   to that in a measurable way.  Thank you. 
 4             MR. REDFEARN:  Thanks, Brian.  Bryan, with a Y? 
 5             MR. HARKINS:  Thanks, Brett.  Good morning, 
 6   everyone. I'm Bryan Harkins.  I am the co-head of our 
 7   markets division at Cboe Global Markets.  Cboe, we 
 8   operate in U.S. equities, U.S. options, futures, foreign 
 9   exchange and European equities.  So what I'd like to do 
10   this morning is to bring you perspectives of across our 
11   different asset classes, as an exchange operator, because 
12   clearly, liquidity is a topic, and illiquidity really 
13   does apply to all asset classes, and I think different 
14   markets have tried to solve this problem, this challenge, 
15   in different ways.  And so hopefully I can be helpful 
16   this morning with different perspectives.  Just to throw 
17   out some questions that I think are worth exploring 
18   around this space are things like, okay, how have the 
19   banks transitioned to really agency models and less 
20   capital commitment from banks, that I've seen over, 
21   particularly over the last decade.  How is that impacting 
22   the small cap space?  And really, in my 20 years in the 
23   exchange space, I've really seen a -- a dramatic 
24   evolution as to who are the exchange's best liquidity 
25   providers, how that has evolved really from as I said the 
0026
 1   banks to our best liquidity providers are non-banks, 
 2   people who don't have direct customers themselves.  And I 
 3   think that has an impact on how they look at supporting 
 4   illiquid securities when they don't have those direct 
 5   customers.   
 6             Profitability, you know, liquidity providers 
 7   follow the profits, so I think it's worthwhile to look at 
 8   how market-markers are incentivized to really, to stand 
 9   in there and provide liquidity in names that just don't 
10   trade very much.   
11             How has passive investing had an impact?  
12   Really, less stock-picking like the old days so to speak, 
13   how has that potentially impacted the illiquid and the 
14   small cap space.  And then lastly, just around exchange 
15   innovation, one thing just really speaking from -- really 
16   speaking on behalf of the exchanges is what my opinion 
17   is, what we don't do well is really servicing blocks, 
18   facilitating really more slower negotiated trading. We 
19   trade liquid stocks very very well. I think we all know 
20   that.  We've heard about the one size fits all model of 
21   Reg NMS, and how perhaps that should be tweaked.  So when 
22   exchanges are trying to innovate, and we've seen some 
23   rule filings over the last few years, trying to get at 
24   some of these issues, really what can exchanges do in 
25   trying to innovate within a Reg NMS market structure, 
0027
 1   that's not so easy. 
 2             MR. REDFEARN:  Thank you, Bryan.  Frank? 
 3             Mr. HATHEWAY:  Thank you, Brett.  Good morning. 
 4    My name is Frank Hatheway.  I'm the Senior Vice 
 5   President and Chief Economist for NASDAQ.  On behalf of 
 6   myself and NASDAQ, I'd like to thank the staff of the 
 7   Commission for organizing and hosting this roundtable.  I 
 8   also thank Commissioners Peirce, Piwowar, Stein, Director 
 9   Redfearn, for their leadership on this issue.  Finally, 
10   I'd also like to thank the staff of OAR for their 
11   detailed analysis of thinly traded NMS securities.  I 
12   look forward to today's panel.   
13             NASDAQ is home to not only the four largest 
14   companies in the United States, but also to many thinly 
15   traded companies.  In Europe, our Nordic Exchanges also 
16   list companies both large and small.  Small companies, 
17   through our SME market, First Norse.  We have a long 
18   history of providing markets where companies can grow and 
19   thrive in a variety of market structures.   
20             NASDAQ views the decline in the number of small 
21   companies in the U.S. as a concern.  We hold the opinion 
22   that more can and should be done to cultivate liquidity, 
23   improve the listing and trading experience in thinly 
24   traded stocks in the U.S.  Our analysis and experience 
25   have let us conclude there is no simple single solution. 
0028
 1    A year ago, we memorialized our views through the 
 2   revitalized blueprint, which sets out a comprehensive set 
 3   of recommendations, covering broad reform of the public 
 4   company; particularly reform of the obligations of public 
 5   companies, the promotion of long-termism and modernizing 
 6   market structure.  In the next few days, NASDAQ will file 
 7   an application with the Commission to permit issuer 
 8   choice as to whether to consolidate the exchange 
 9   liquidity onto a single market by suspending unlisted 
10   trading privileges for non-listing exchanges.  We believe 
11   a consolidation of exchange liquidity onto a single 
12   exchange would promote economic and pricing efficiency by 
13   focusing the listing and trading of thinly traded 
14   securities onto one exchange, where those securities have 
15   opted to suspect UTP.  Encourage national securities 
16   exchanges that are listing markets to compete and 
17   innovate more aggressively.  To list and trade securities 
18   for these smaller companies for which UTP has been 
19   suspended.  Improve market structure and market quality 
20   for securities for which UTP privileges have been 
21   suspended, incentivizing additional smaller companies to 
22   list on the public markets.  Our application envisions 
23   that over the counter trading will continue in the same 
24   form as today, in those securities where issuers have 
25   chosen to suspect UTP privileges. We believe that 
0029
 1   evidence from our European markets as well as NASDAQ's 
 2   own history supports a conclusion that dealers serve a 
 3   vital role in providing liquidity in thinly-traded 
 4   securities.  In closing, our markets are no longer able 
 5   to support small growth companies such as the four I 
 6   alluded to, which are now the largest companies in the 
 7   United States.  We are pleased with the dialogue begun 
 8   with the Jobs Act, Tick Pilot.  We look forward to 
 9   further engagement with investors, public and private 
10   companies, industry groups, policy makers, to create a 
11   market with the ability to serve the needs of small 
12   issuers as well as large.  Thank you, Brett. 
13             MR. REDFEARN:  Thank you, Frank.  Ovi? 
14             MR. MONTEMAYOR:  Good morning Commissioners and 
15   Division staff.  Thank you for the opportunity to 
16   participate in today's roundtable.  I'm Ovi Monetemayor, 
17   managing director of Financial Market Services at TD 
18   Ameritrade.  TD Ameritrade, based in Omaha, Nebraska, was 
19   founded in 1975 and was one of the first brokers to offer 
20   negotiated commissions following the passage of the 
21   Securities Act of 1975.  TD Ameritrade has long advocated 
22   for market structures that create transparency, promote 
23   competition and reduce trading costs for individual 
24   investors.  Over the last 20 years, the rise of 
25   alternative trading systems and technology advances have 
0030
 1   resulted in a market structure that provides retail 
 2   investors with an easy, low cost access to the capital 
 3   markets.  In fact, retain investors have never had it so 
 4   good.  What may be surprising to some is that this 
 5   statement applies equally to the retail experience in 
 6   trading the most actively traded securities and the less 
 7   liquid stocks.  During the last quarter of 2017, TD 
 8   Ameritrade handled over 700,000 trades per day, traded 
 9   over 440 million NMS shares per day for over 11 million 
10   clients.  For low ADV securities, we estimate that TD 
11   Ameritrade clients represented approximately 4.5 percent 
12   of the market and 9 percent of the market for NMS 
13   securities with an ADV under 1 million notional. 
14             Our experience is that retail investors enjoy 
15   the overall same execution quality regardless of the 
16   liquidity of the security. Most importantly, these 
17   clients' orders were filled in comparable price 
18   improvement percentages; they received enhanced liquidity 
19   over the amount of shares available at the inside quote. 
20    In fact, over 45 percent of TD Ameritrade client orders 
21   on symbols with less than 100,000 ADV are submitted in 
22   amounts larger than the insighted quote.  These orders 
23   are generally filled because the market makers that TD 
24   Ameritrade uses provide liquidity enhancements.  On 
25   average the liquidity enhancement averages 5.5 times the 
0031
 1   inside quote.  So for every quoted size in 100 shares, 
 2   market makers provide TD Ameritrade clients with 550 
 3   shares at the same inside quote.  The experience of our 
 4   retail investors trading thinly traded securities 
 5   demonstrates that market-center competition provides 
 6   retail investors with significant benefits.  To the 
 7   extent that some propose a single stock exchange to trade 
 8   thinly traded stocks or trade at role, TDA finds that any 
 9   such proposal would likely be detrimental to the retail 
10   client experience and contrary to one of the foundational 
11   tenants of the SEC mission to protect investors, 
12   investors such as the 11 million that TDA advocates for 
13   today.  Again, I thank you for the opportunity today and 
14   I look forward to answering all your questions. 
15             MR. REDFEARN:  Thank you, Ovi.  Ari? 
16             MR. RUBENSETIN:  Director Redfearn, your team, 
17   Commissioners, thank you for having me here today. I 
18   remember Chair Clayton's speech at the Economic Club in 
19   New York last July, and I thought he really set the bar 
20   high for technology in our industry, where he said that 
21   technology and innovation was disrupting in positive ways 
22   and that we need to ensure that today's rules governing 
23   trading reflect the realties of our capital markets.  And 
24   I think that's really relevant and germane to this 
25   roundtable and what I want to talk about and what I live 
0032
 1   and breathe every day.  My name is Ari Rubenstein; I'm 
 2   the cofounder and CEO of GTS.  GTS is an electronic 
 3   market maker.  We provide offers to buy and sell 
 4   securities across global markets and we do that 
 5   quantitatively using computers.  At the New York Stock 
 6   Exchange, we're the largest designated market maker, 
 7   responsible for over 900 public companies, amounting to a 
 8   little over 12 trillion dollars worth of market 
 9   capitalization. And we're also directly accountable to 
10   these companies to ensure that there's liquidity 
11   available for their investors when they need it. And in 
12   the category of small cap securities under 100,000 ADV, 
13   we're at 371 companies in our portfolio that we're 
14   responsible for. 
15             I'm excited for this discussion today because I 
16   think there's a big opportunity for the industry to come 
17   together and to leverage technology to try to enhance the 
18   environment for thinly traded securities and I look 
19   forward to sharing some of my ideas.  Thanks. 
20             MR. REDFEARN:  Thank you very much, Ari.  
21   Jason? 
22             MR. VEDDER:  Thanks, Brett.  Thanks to the 
23   Commission for the invitation.  My name is Jason Vedder. 
24   I am the Director of Trading and Operations at GTS 
25   Capital Management in Chicago.  We are a boutique firm, 
0033
 1   about 9 billion under management.  We focus on both fixed 
 2   income and equities, but our flagship products are in the 
 3   small cap and microcap US base as well as emerging market 
 4   space as well.  I've been at GTS for 18 years, in 
 5   different capacities.  I was, started out as a trader 
 6   there, and then became an analyst for the firm, covering 
 7   the technology, industrial energy sectors for four years, 
 8   and then decided to come back and run the trading and 
 9   operations since 2014.  Prior to my career at GTS I was 
10   also a market maker on the Chicago Stock Exchange as well 
11   as the Cincinnati Stock Exchange. 
12             And coming in in the early 90s, not to date 
13   myself, but a lot of the big liquid names were already 
14   taken to make markets in, and those are the names that 
15   you wanted to have.  The smaller cap names, the microcap 
16   names are the ones that traded 50,000 shares, those fell 
17   to my -- my pad.  And you had to be nimble in how to 
18   trade those for risk purposes as well.  So I think it's 
19   great that we're having this conversation.  I think it's 
20   long overdue. I think that, you know, the way that we've 
21   seen trading evolve over the last 10 to 15 years has been 
22   tougher in these less liquid names. I think, you know, 
23   when you look at it from at the touch perspective, my 
24   firm is at the touch multiple times during the day in a 
25   particular stock. So one at the touch datapoint does not 
0034
 1   really count for my experience in these names.  One other 
 2   comment that I would say is that passive investing has 
 3   taken a big toll in this space as well.  The inverted 
 4   bell curve that you see in these names around these 
 5   passive investments has really bifurcated this market 
 6   even more.  So I really, again, appreciate the 
 7   invitation. I look forward to a healthy conversation in 
 8   this space.  Thanks. 
 9             MR. REDFEARN:  Great.  Well, that was a very 
10   illuminating set of opening comments.  Thank you all very 
11   much for that. I think that to start up the discussion, 
12   Jason, I actually want to turn back to you for a second. 
13    Because one of the issues that has come up is when 
14   institutional investors are making decisions about where 
15   to invest or what names to get into, how does the 
16   liquidity in the name or the dynamics of a particular 
17   stock affect that decision.  So when your firm thinks 
18   about it, how do they decide, and to what extent is it 
19   ever an issue that the liquidity is so low that that's an 
20   impediment to deciding to get into the name? 
21             MR. VEDDER:  Yeah, there's -- it's something 
22   that they look at.  When we're doing bottoms up 
23   investment analysis, those names may get locked off right 
24   at the beginning of the process, so -- but in other 
25   situations, if the idea or the -- the stock, or the 
0035
 1   company that they're looking to invest in is attractive 
 2   enough, they may widen the band of what may be allowable 
 3   for going forward and investing in those names. So it's 
 4   really kind of a fluid decision process.  The PMs will 
 5   come out and talk to us on the desk to see if there's, 
 6   how to look at the volume that's been generated in the 
 7   name over the last 30 days.  But it is a deterrent that 
 8   if it's not trading enough, we won't -- we won't invest 
 9   in it, because it's just, you know, it's a two sided 
10   coin.  I mean, it may be hard to get into, but it may be 
11   really hard to get out of, and that's where you got to 
12   kind of look at both, both in and out, so. 
13             MR. REDFEARN:  Thank you. So Brian, Brian 
14   Fagen, from the perspective of you know, working and 
15   managing an institutional trading desk, when you're 
16   interacting with your clients, what do you see in terms 
17   of their interest and their traffic in these names as it 
18   compares to others? 
19             MR. FAGEN:  Sure.  This is probably one of the 
20   most difficult parts of running an institutional trading 
21   business.  You know, we -- we obviously spend a lot of 
22   time on the larger, you know, most liquid part of the 
23   market, because that's where the bulk of our investors 
24   are, but the segment of the less liquid securities 
25   becomes an issue, as we do have a good number of clients 
0036
 1   that run portfolios with those securities in them, and 
 2   their needs to trade are just as important.  But 
 3   unfortunately the markets' liquidity are not, don't 
 4   really meet that demand.  So we've had to develop 
 5   different strategies in dealing with those.  Some of 
 6   those, you know, end up representing themselves in 
 7   principal-type facilitation, where we simple take the, 
 8   you know, the position from the client in some way, shape 
 9   or form.   
10             We run a centralized risk portfolio across our 
11   equity business, and that's helped us to hold these 
12   positions for an extended period of time, to help 
13   transfer the risk from the client to ourselves, but in 
14   the agency part of our business as Bryan had, you know, 
15   mentioned earlier, that the -- you know, the ability to 
16   trade into the markets, whether it be algorithmically, 
17   using technology or finding the other side is 
18   extraordinarily difficult and probably the highest level 
19   of frustration that we find in transacting with 
20   institutional clients.  This becomes particularly 
21   protracted as it relates to the passive side, as has been 
22   mentioned a few times, when it relates to rebalances.  
23   Index rebalances tend to put the most stress on 
24   securities in the range that we're talking about, because 
25   they do tend to be accumulated over a long period of time 
0037
 1   and then rebalanced all at once with a large number of 
 2   participants all trying to do the same thing at the same 
 3   time in a very public forum, and these re-weightings are 
 4   very well publicized.  The idea behind that is sometimes 
 5   that will bring out liquidity to find the other side, but 
 6   in most cases it ends up with a small number of broker 
 7   dealers and ending up having to control a large 
 8   percentage of ADV, you know, many many days of ADV into a 
 9   very, very point in time event, which does create 
10   dislocation and stress into the marketplace.  
11              So we work with clients on the institutional 
12   side to try to find paths to work these types of 
13   positions, both the in, and Jason's point, the out is 
14   probably the more sensitive part of the equation, because 
15   the demand to acquire position is generally more patient 
16   over time.  The demand to get out is usually much quicker 
17   and shorter. We end up -- we end up on the other side of 
18   a lot of those transactions and then having to work out 
19   of those ourselves into the market or provide liquidity 
20   to other clients.  So our position in this is, I think is 
21   a little bit, it's a little bit different than many of 
22   the other participants up here.  You know, our -- you 
23   know, we don't, we don't have a direct, a direct impact 
24   on the market structure aspect of it.  You know, we're 
25   more of a, you know, a causal effect of it, and that 
0038
 1   we're trying to work within it and to try to find the 
 2   best solution, and this is clearly an aspect that for our 
 3   clients creates a lot of frustration. 
 4             MR. REDFEARN:  So to get a better sense of the 
 5   retail interest and the set of names, I wanted to reach 
 6   out to Steve and Ovi about, you know, what do you see in 
 7   terms of what is driving retail interest in these names? 
 8    Do you find that the retail investor is as equally 
 9   interested in these names or other names or what -- how 
10   can you characterize the driver behind retail interest? 
11             MR. CAVOLI:  I'll actually punt it over to Ovi, 
12   because you probably understand the underlying interest 
13   from the retail a little bit more than the actual 
14   execution.  And then I have a point about the liquidity 
15   both from the retail perspective but also an interesting 
16   difference in how -- difference in experience that we 
17   have than what Brian has at Deutsche.   
18             MR. MONTEMAYOR:  Yeah, well, from my 
19   experience, wholesalers do a great job of sourcing their 
20   liquidity for TD Ameritrade clients.  Liquidity is not 
21   something our clients have to think about, and doesn't 
22   prevent our clients from trading those securities.  We 
23   receive price improvement on 97 percent of these thinly 
24   traded securities, and seven times the enhanced 
25   liquidity.  Our experience is that retail investors enjoy 
0039
 1   the overall same execution quality regardless of the 
 2   liquidity of these securities.  In our research on market 
 3   quality and execution metrics for thinly traded 
 4   securities, we find the more relevant delineation to be 
 5   based on notional ADV instead of shared ADV.   
 6             For example, in March of this year, 814 NMS 
 7   securities traded fewer than 100,000 shares per day and 
 8   yet traded over 1 million notional per day.  With a 
 9   simple stock split these symbols would no longer be 
10   considered thinly traded by the share ADV criteria alone. 
11   In some cases it's possible that some of these securities 
12   would be more tradable if they simply split the stock.  
13   In our experience, 25 percent of our orders for symbols 
14   with less than 100,000 shares or 100,000 ADV were larger 
15   than the quoted size.  Almost exactly the same as on 
16   securities with greater than 100,000 shares ADV. On the 
17   other hand, over 40 percent of orders for less than a 
18   million ADV notional were larger than the quoted size, 
19   compared to 25 percent of those orders greater than a 
20   million per day. 
21             Therefore, the enhanced liquidity that the 
22   wholesalers provide was much greater on notional ADV 
23   symbols on the lower end of the notional ADV symbols, 
24   because of the thinner markets.  We find these numbers 
25   convincing that the notional ADV may be a better way to 
0040
 1   describe liquidity in that share ADV alone.  
 2             It's also important to highlight that the 
 3   retail beneficial relationship with these wholesalers is 
 4   preserved.  Because of the individuals' experience with 
 5   market quality and fairness is experienced by trading in 
 6   their self-directed accounts, if anything is done to 
 7   hamper these positive experiences, public perception of 
 8   the fairness of the markets could be compromised. 
 9             MR. REDFEARN:  All right.  As a DMM, in your 
10   experience trading these names and how you think you're 
11   interacting with whether it's a retail market or an 
12   institutional market, can you just share your thoughts 
13   and challenges? 
14             MR. RUBENSTEIN:  Absolutely.  On that, let me 
15   share a story with regards to when GTS became a DMM. This 
16   was a few years ago. I remember we had to become 
17   compliant with a long list of rules to be a designated 
18   market maker at the New York stock exchange.  It was 
19   different from being a market marker on any other 
20   exchange.  For starters, we had to put up our share of 
21   the hundreds of millions of dollars worth of regulatory 
22   capital that we had to segregate to back up our offers to 
23   buy and sell these securities to ensure that GTS would be 
24   there, you know, during times of stress if there was a 
25   lot of volatility.   
0041
 1             Also, we had to file, for each security, 
 2   capital commitment schedules.  So over and above our 
 3   quoting requirement, we had to make sure that we would 
 4   supply the market with capital for each security.  And 
 5   this was just a couple, a couple examples.  So I remember 
 6   being somewhat frightened managing a business that had 
 7   those types of obligations.  But what ended up happening 
 8   was, once we got going and deployed all the technology 
 9   that we built going on two decades now, we were able to 
10   hit a lot of those obligations.  And I'm very proud to 
11   say that in small cap securities at the New York Stock 
12   Exchange, GTS is number one in providing price 
13   improvement, compared to any other market maker down 
14   there, and that's the percentage of our quotes that are 
15   improving the national best bid offer. 
16             And we did this because we used technology.  
17   And I think that that could be an area that we can work 
18   to improve the markets.  What we would recommend is a 
19   modernization of market-making rules.  Some of these 
20   rules were, you know, are Depression-era rules regarding 
21   trading and market making.  They were enacted for good 
22   reason, but at a different time, when -- before the 
23   markets were electronic.  Before computers were handling 
24   trading and decision making.  And so I think we can -- we 
25   can modify some of these rules.  That will allow market 
0042
 1   makers to have more flexibility to be able to leverage 
 2   the technology that they already have.  And I think you 
 3   could marry some of these changes with an increase of 
 4   market-making obligations.  And I think that the scale 
 5   technology market marker of today has the ability to 
 6   survive with higher obligations than they have today that 
 7   were built on rules enacted a long time ago.  And I would 
 8   take that flexibility and I would also apply it to the 
 9   exchange space and allow exchanges an easier path and an 
10   easier way to innovate and come up with more solutions, 
11   especially in the area of small cap securities. 
12             MR. REDFEARN:  And I think that's one of the 
13   things we're going to talk about -- 
14             MR. RUBENSTEIN:  Okay. 
15             MR. REDFEARN:  -- further, later today, is 
16   whether or not this idea about UTP actually could foster 
17   some of that -- 
18             MR. RUBENSTEIN:  Great. 
19             MR. REDFEARN: -- type of innovation.  But I 
20   look forward to later hearing more of your, the specifics 
21   on the changes -- 
22             MR. RUBENSTEIN:  Okay. 
23             MR. REDFEARN:  -- to the antiquated rules.  But 
24   Steve, you wanted to finish a comment before. I don't 
25   know if we have -- 
0043
 1             MR. CAVOLI:  Yeah.  Thank you.  So two points. 
 2    One was, retail interest.  Ovi did a good job of 
 3   answering it. On the liquidity piece of that question, in 
 4   the opening remarks I mentioned that we trade about 12 
 5   percent of names that trade under 100,000 shares a day, 
 6   and Ovi did a good job of explaining the wholesalers' 
 7   role in providing price improvement and size improvement. 
 8    And importantly, doing it immediately.  So at the touch. 
 9   So no timing risk.  You see the screen, you go to trade 
10   stock, and you get it done.  The important aspect of our 
11   ability to do that -- and I'll talk a little bit about on 
12   the institutional side, which is opposite of Brian's 
13   problem, where they concentrated in the top, very liquid 
14   names, which quite frankly don't need a ton of help, we 
15   concentrate ourselves down on the lower end.  And I'll 
16   talk a little bit about that.   
17             But the reason we can do that with retail names 
18   and the reason we can do it with our institutional 
19   clients is because of the bilateral accountable 
20   transparent nature of the relationship.  It allows us to 
21   provide a ton of service to our clients and it allows us 
22   to understand the nature of the risk that we're taking.  
23   So we did a good job of talking about the retail side, 
24   but let's just look a little bit on the institutional end 
25   of the spectrum.  So providing liquidity to institutions, 
0044
 1   34 percent of overall executions that we provide in that 
 2   space are in smaller microcap companies, and we define 
 3   that as small is 250 million to a billion in microcap -- 
 4   up to 250 million. 
 5             28 percent of those trades are 2000 shares or 
 6   higher.  4 percent are 10,000 shares.  5000 to 10,000 are 
 7   eight percent.  And 2000 to 4999 shares are 16 percent.  
 8   So there's a significant amount of liquidity being 
 9   provided in small and mid -- small and microcap names.  
10   We can even go up to the mid-cap tier -- not just to 
11   retail clients, but also to institutional clients.  And 
12   once again, the reason we're able to do that and provide 
13   that kind of liquidity is very specifically a function of 
14   the market that has developed over the last, you can say 
15   since '75, but really since '97, '98, with the advent of 
16   order handling roles manning -- the great success of ATS, 
17   decimalization and then NMS.  It's kind of a cocktail 
18   effect. No one rule was necessary, one was necessary but 
19   insufficient to the other.  And then you had technology 
20   that came in and provided a ton of transparency around 
21   those executions. 
22             MR. REDFEARN:  So before we debate the cocktail 
23   effects of Reg NMS, I wanted to -- 
24             MR. CAVOLI:  Good cocktail effect. 
25             MR. REDFEARN:  -- go to Frank or Bryan and just 
0045
 1   sort of -- I'd like to hear your thoughts on this, but 
 2   also, you know, we are dealing with an interesting array 
 3   of market structures here, right, and there probably is 
 4   some advantage also to knowing the counter party in these 
 5   transactions in the process of being able to you know, 
 6   facilitate that liquidity provision. I guess I just 
 7   wanted to go back to sort of the exchange side for a 
 8   second and ask your thoughts about, you know, whatever 
 9   follow up you had, Frank, but your thoughts specifically 
10   about bringing liquidity together in these names, given 
11   the unique challenges there. 
12             MR. HATHEWAY:  Let me go, because I want to 
13   follow up on this point about sourcing liquidity.  One of 
14   the things we hear from listed companies, CFOs, investor 
15   relations officers, is that they can sell the story of 
16   their company's stock to an institutional investor, and 
17   then are confronted by a statement from the same investor 
18   that they're concerned about being able to get in or out 
19   of the stock, or not able at all, in their opinion, to 
20   get in and out of the stock.  That echoed some of the 
21   earlier comments that have been made that sometimes these 
22   stocks can be very difficult to trade, and we hear it 
23   from the issuer side.  In terms of gathering liquidity in 
24   a combination of a single exchange and an active over the 
25   counter market, that is essentially how Europe operates. 
0046
 1   It's how our markets in Europe operate. The details of 
 2   European securities operation are not generally well 
 3   known in the United States.  But there are three types of 
 4   trading in Europe.  We tend to think of just two.  But 
 5   there's over the counter, which it is pure over the 
 6   counter. 
 7               There is what's called on exchange, off book, 
 8   which would smell a lot like trading in the United States 
 9   that takes place over the counter, because it is bound by 
10   regulatory oversight of, in the US case FINRA, in the 
11   European case the national regulatory authority, the 
12   listing exchange.  It's also bound by the price filters, 
13   which in our world are the order protection rules.  So 
14   there's a great deal of similarity between parts of the 
15   US structure and parts of the European structure which 
16   underscore, as I said in my prepared remarks, the 
17   importance of dealers.  You cannot trade these stocks 
18   without dealers who are willing to commit capital.  As 
19   Brian said, it makes it easier for his clients in the 
20   securities where his firm is able to act as principal, 
21   and Steve said when they act as principal, and as Ovi 
22   said when he is acting or able to interact with a broker, 
23   acts as principal, it's an important part of these 
24   markets. 
25             When that activity can be spread over multiple 
0047
 1   exchanges, it becomes harder to know where to source it. 
 2   If you're looking for trying to determine where the 
 3   relevant price is, it also adds complexity to trading 
 4   these stocks.  You know, we have a number of first line 
 5   broker dealers here.  We don't have all the broker 
 6   dealers, types of broker dealers represented here who 
 7   trade less active securities.  And for many of the 
 8   smaller ones, the complexity of the market is a 
 9   challenge, and having to trade potentially 13 places is a 
10   challenge when they go to either try and address the 
11   order types that exist in a national market system, Reg 
12   NMS type of structure or simply to try and figure out, 
13   you know, where they're going to position liquidity to 
14   have the best of chance of an offsetting order 
15   interacting with them.  So we think that concentration on 
16   one market would help the markets.  And we, as I said, 
17   view our experience in Europe, even with the First North 
18   Market, which is much smaller in the United States, it's 
19   indicative of the kinds of things an exchange can do with 
20   both innovation and with the ability to focus trading 
21   within its order book, supplemented by a robust over the 
22   counter environment. 
23             MR. REDFEARN:  Bryan? 
24             MR. HARKINS:  Yeah.  I think I just want to go 
25   back to something that Steve mentioned around the value 
0048
 1   of disclosed bilateral trading.  So it allows market 
 2   makers to really control a really important risk 
 3   component, knowing your customer and you know, being able 
 4   to interact with them in a sort of confined environment. 
 5    Frankly, yeah, that's something that exchanges just 
 6   haven't done well.  We don't, you don't see exchanges 
 7   running disclosed order books or creating, as I said in 
 8   my earlier remarks, sort of slowing down the process to 
 9   allow for a more bilateral facilitation back and forth.  
10   Yeah, the notion of revoking UTP, you know, that's an 
11   idea that could work.  It's just, but I really think the 
12   bigger issue is around exchange innovation in this space. 
13    ATSs are obviously doing this and trying to do this 
14   well, but exchanges usually haven't.  And you know, as I 
15   said earlier, it's, how do you innovate within the 
16   confines of Reg NMS?  You know, we're very fixated on 
17   making sure that we don't introduce any inefficiencies 
18   into the NBBO and into a market structure that works 
19   very, very well for liquid names but just really misses 
20   the mark for the illiquid names. 
21             The other thing I'll mention too is just, I 
22   think going back to, I think it was Jason who said 
23   earlier, like he would oversee a pad of illiquid 
24   securities, and you know, we exchanges, we need to be 
25   able to better tie the illiquid names to the liquid 
0049
 1   names, and we operate a listings venue for ETPs and you 
 2   know, 90 plus, 95 percent of them just don't trade.  
 3   Different, different product set of course, but how can 
 4   exchanges better tie their rewards of liquid names to the 
 5   obligations of being able to stand in the market and 
 6   facilitate and frankly sort of babysit some of the 
 7   illiquid names.  I think there really needs to be more 
 8   explicit tying in of the economics. 
 9             MR. REDFEARN:  Interesting.  Adam, I wanted to 
10   jump back to you for a second, given that it seems to me 
11   like you are talking to the issuers in this space, and 
12   get any thoughts on what's been said so far.  But in 
13   particular, you know, I am interested when issuers have a 
14   view as to how the markets might function differently in 
15   a way that helps their securities, so any -- what are 
16   your thoughts? 
17             MR. EPSTEIN:  Yeah, well there's a -- so far 
18   we've been talking about liquidity in open market 
19   purchase; there's a whole other eco system of liquidity 
20   and that's capital formation, right?  I mean, roughly 76 
21   percent, give or take of publicly traded companies in the 
22   United States have market caps below 500 million dollars. 
23    45 percent of NASDAQ issuers actually have market caps 
24   below 300 million dollars.  They aren't even large enough 
25   to be small cap.  And the preponderance of those 
0050
 1   companies are out there, their principal concern in life, 
 2   since most of them are not cash flow positive -- and by 
 3   the way, reasonable people will differ as to whether some 
 4   of those companies should be public, but they are.  But 
 5   those companies are raising a lot of capital. When I was 
 6   a fund manager, that was a 25 to 40 billion dollar a year 
 7   industry, okay?  And I think the capital formation part 
 8   of the liquidity in the small cap ecosystem needs to be 
 9   looked at, I would respectfully suggest, in the pre-
10   financial crisis versus today.   
11             Back in the pre-financial crisis, probably a 
12   lot of people on this desk and certainly in the 
13   Commission remember that there was a really vibrant 
14   ecosystem for capital formation, for example, all the way 
15   down into the OTC BB exchanges. Subsequent to the 
16   financial crisis, the way that limited partners deal with 
17   hedge funds has changed dramatically.  Limited partners 
18   used to come to general partners and sit down -- I'm 
19   talking about the special situation ecosystem of 
20   providing capital directly to small cap companies. 
21   Limited partners would come and they would sit down with 
22   hedge fund managers, and they would say, they would point 
23   to a ticker symbol on a screen and they'd say, "Tell me a 
24   little bit about the investment thesis.  What are some of 
25   the catalyses to get out of it?  Tell me a little bit 
0051
 1   about the pricing."  Post financial crisis, that 
 2   conversation doesn't happen anymore. Post financial 
 3   crisis, the allocators from limited partners come and sit 
 4   down to the special situation fund managers that are 
 5   providing all the capital to all of these small cap 
 6   companies, and point to ticker symbols and say, "How many 
 7   days until you can liquidate that?"  There's nothing 
 8   about, there's nothing about fundamentals, there's 
 9   nothing about catalysis.  It's how many days until you 
10   can liquidate that position?  And so that absolutely 
11   infiltrates our 25 to 40 billion dollar a year industry. 
12             So when I say that -- I said earlier in the 
13   opening remarks that capital formation is inextricably 
14   linked to trading volume, today's marketplace.  And one 
15   of the anomalies that's really interesting -- you could 
16   have a whole other panel to talk about why it is that you 
17   could have a 250 million dollar company, one of them 
18   trades -- an exchange listed company, one of which trades 
19   a million dollars a day and for some reason the other one 
20   only trades $50,000 a day, there's lots of examples of 
21   that.  But when you look at those companies, what really 
22   matters, and what really happens out there to these 
23   issuers, is that if you trade a million dollars a day, 
24   you can go out and probably do a shelf takedown or do a 
25   follow on offering with no warrants and price somewhere 
0052
 1   around 15 percent off.  If you're the company that trades 
 2   $50,000 a day, what you're looking at is perhaps some 
 3   type of restricted stock deal.  You may be looking at 
 4   something that's hyperstructured with pretty austere 
 5   warrant coverage.  And to be able to dig your way out of 
 6   that operationally is really challenging.  So for those 
 7   companies, there's one whole part of the ecosystem of 
 8   liquidity, which is buying and selling stock in the open 
 9   market, but there's a big marketplace, because all of 
10   these companies require growth capital.  And what I hear 
11   from issuers over and over and over again is that gee, 
12   why is it that, that none of the special situation fund 
13   managers anymore tend to care nearly as much as they used 
14   to about the quality of the companies?  People look at 
15   ticker symbols and very quickly tell you how much 
16   percentage of your market cap you're going to be able to 
17   raise.  That's pretty challenging.  Again, we're talking 
18   about the overwhelming majority of public companies in 
19   the United States.  It's not an isolated cross-section. 
20             MR. REDFEARN:  Yep, yep, yep, yep.  So Jason, 
21   so you mentioned earlier the difficulty of getting out of 
22   names in particular can be a challenge.  I think it would 
23   be helpful for you to just talk for a minute about your 
24   trading process, right?  So you get an order from a PM 
25   and you look at it and you say, "Wow, this thing is 
0053
 1   pretty illiquid.  It might take me days," or whatever.  
 2   Tell me about that process and just how you think about 
 3   that process in terms of working that order, who do you 
 4   call;  how does it work? 
 5             MR. VEDDER:  Sure.  So if we receive an order 
 6   in a name like that, the first, the first hope is that 
 7   you're going to get something in the dark markets that 
 8   you may be able to find a natural counter to what you're 
 9   trying to do.  If that's the case, you know, maybe you 
10   can get 20 percent of your order done or whatever they're 
11   trying to accomplish.   
12             But after that it's a cat and mouse game and 
13   it's basically me spending my day, if something's not 
14   working, you're just moving on to the next venue, and for 
15   as much as cash trading has stayed relevant, and it's 
16   certainly stayed relevant in certain situations, our 
17   reliance on the cash desk has gone down quite a bit.  And 
18   it's not because we don't think that they add value, it's 
19   the fact that we're trying to control the information 
20   that's coming out of our, our firm, in terms of, and the 
21   names that we trade in.   
22             In a name like, that trades 100,000 shares, 
23   there are going to be only two to three people that are 
24   participating in the marketplace that would have a 
25   counter side to what I'm trying to do, if any.  So then 
0054
 1   you're trying to combat everybody else that's trying to 
 2   get into that market space and possibly glean any kind of 
 3   information from how you're going in, you're entering 
 4   into the marketplace.  So our first step is usually, 
 5   unless we have any kind of information from a cash desk, 
 6   that they were involved in a name in the last five to ten 
 7   days, if there's no information like that available for 
 8   the name that we're investing in, we go straight to the 
 9   algos, and the algos are our way of trying to stay as 
10   anonymous as possible.  
11             It's frustrating that I have to go to, you 
12   know, 20 different exchanges to get 10 percent of what 
13   I'm trying to do, but that's the cat and mouse game that 
14   it has become, and so you're basically just knocking off 
15   the strategies that you think will work in that 
16   particular name, and having been in those names for as 
17   long as I have, you have certain strategies that actually 
18   work well and some that don't, and so that's where you 
19   start.  
20             You start at your best performing strategies 
21   and then you work your way down to the point where you're 
22   looking at something that you probably haven't used in 
23   two months and you're like, well this is, I've got to 
24   give this a shot because I'm not getting anything else 
25   anywhere else in the marketplace.  So it is -- it has 
0055
 1   gone from being a negotiated market to more of a cat and 
 2   mouse game where you're just trying to find pockets of 
 3   liquidity that will -- you'll be able to collect together 
 4   and soon enough you'll eventually hopefully get to your 
 5   target investment.   
 6             That's, you know, the stocks that we're trading 
 7   in, I think we do them a disservice when we just 
 8   basically let them out into the wild.  With everything 
 9   else that we've created in our industry and all the 
10   regulations that we've created for everything in the NMS, 
11   that I think you know, we do it a disservice to that 
12   company to basically throw them to the wolves and 
13   basically have everybody else trade the way that they 
14   trade a stock that trades 9 million shares a day, that 
15   there should be some kind of incubator or ICU kind of 
16   environment that limits the amount of places that you can 
17   go.   
18             Because it still takes more negotiated price, 
19   more price discovery.  I think it encourages kind of a 
20   back and forth dialogue that was in the market a long 
21   time ago but has kind of gone away, that you can actually 
22   pitch yourself against another side or somebody else 
23   that's trying to get out of the name, and you can 
24   negotiate a price and you feel good about being done with 
25   that process, that you found the best, the best price for 
0056
 1   your clients on both sides, in and out. And that to me 
 2   would be something that I think would be -- it would be 
 3   something that hasn't been done before, which I really 
 4   applaud trying to do just to see what happens. 
 5             MR. REDFEARN:  Bryan?  Thank you, Jason. 
 6             MR. HARKINS:  I figured I'd make it 
 7   interactive.  I think he touched on, Jason touched on a 
 8   really important point around the, kind of the back and 
 9   forth negotiation. And I know the next panel is going to 
10   talk about revoking UTP.  Again, the idea has a lot of 
11   merit.   
12             But part of, I think, what the challenge of 
13   that is is that it presumes that the listing exchange is 
14   actually going to service that particular market the 
15   best, right?  So non-listing exchanges can sort of so 
16   called turn back the clock with different market models 
17   that address this problem.  Negotiation. You don't have 
18   to be a listing exchange to have a floor.  I know only 
19   one does, but you know, you could perhaps service the 
20   segment through more back and forth price formation 
21   outside the typical continuous electronic order book. 
22             MR. CAVOLI:  In light of the interactive nature 
23   that Bryan just established -- 
24             MR. REDFEARN:  I don't -- opening up a little 
25   here. 
0057
 1             MR. CAVOLI:  May I?  Those are very good 
 2   points, Bryan.   
 3             And to Frank's point earlier and to Jason's 
 4   also, these, it's very difficult -- like I always have 
 5   great sympathy for the regulator, I will say it, because 
 6   you are coming up, you're trying to come up with the best 
 7   set of rules.  Not the perfect set, but the best set, 
 8   because of the diverse nature of the variety of 
 9   participants, who all have different time horizons, they 
10   have different reasons that they trade, and any variety 
11   of kind of risk/reward metrics that they look at.   
12             And we have to ask ourselves when -- Frank, 
13   when you mentioned that before, are we talking about, are 
14   we confusing temporal fragmentation with geographic 
15   fragmentation? So the underlying interest in the name, 
16   the fact that there are limited or a lack of diverse 
17   holders of the name, is that why these are difficult to 
18   trade?   
19             Is that why, you know, they can't find each 
20   other, or is it because they're resting in 40 venues, you 
21   know, in the U.S. market.  So I think that we have to 
22   really make sure that we don't conflate the two.  
23   Temporal fragmentation, which is a more fundamental 
24   underlying interest, versus geographic fragmentation, 
25   which quite frankly through 611 and through technology 
0058
 1   has been stitched kind of back together.  And so it's 
 2   not, it's not like the geographic fragmentation of yore, 
 3   where you had, you know, you had the Cincinnati, you had 
 4   the Pacific, and you had really discrete pockets of 
 5   liquidity and you had different, you had different 
 6   participants who had different set of rules, that weren't 
 7   a national market system. I mean, the national market 
 8   system really achieved its intended purpose. 
 9             MR. FAGEN: I think there's a risk here, right? 
10    You know, and listening to Jason, I think it's 
11   incredibly impactful, right?  That the -- the, you know, 
12   the frustration for the manager of securities is 
13   sometimes lost when we start to talk too much into the 
14   sort of microjargon of the markets, right?  The actual, 
15   you know lifeblood of an asset manager in terms of 
16   creating return and getting access to, you know, to 
17   alpha, you know, in the marketplace.  And I think that he 
18   brought up an interesting, you know, comment that a lot 
19   of people sometimes try to stay away from because its 
20   uncomfortable, which is information, right?  You know, 
21   from a trading perspective, what I can tell you is that 
22   one of the biggest costs that we deal with in trading and 
23   implementation is information cost, right?   
24             The aspect of the marketplace knowing the 
25   existence of an order, right, and what that forces you to 
0059
 1   do.  And you know, you know Jason, you know, kindly 
 2   called it a cat and mouse game, you know, and that's a 
 3   good way to describe it. The ability to control and limit 
 4   the amount of information that you put into a marketplace 
 5   in exchange for liquidity is really what the game in 
 6   trading is about today, especially in this range of 
 7   securities.   
 8             It becomes that much more important.  It's not 
 9   nearly as important -- it has some implementation in 
10   large cap liquid stocks, but as you move further down the 
11   chain, if you think about what anyone has ever traded of 
12   securities always thought about, there's a sort of an 
13   old, you know, quantitative theory called Augment- 
14   Risk, which is basically just what people think 
15   about as a trader, which is you know, impact versus risk. 
16   I trade shorter, I create more impact.  I trade longer, 
17   you know, I take more risk.   
18             It's pretty simple.  But post -- really post 
19   NMS and post fragmentation, there's a third, what I think 
20   of as a third dimension to that equation, which is the 
21   trade off of information for liquidity.  That's kind of 
22   what we're really talking about within this range of 
23   securities.  You can go all the way down to the 
24   microlevel and talk about, you know the market making 
25   aspect and that piece of the equation, but when you bring 
0060
 1   it back up to the more macro level for the, you know, 
 2   asset manager or hedge fund or institutional investor 
 3   trying to gain you know, gain access to these securities, 
 4   it's really about when they have to implement, how much 
 5   information do they have to give up and what's the cost 
 6   of that information and finding that liquidity.   
 7             And today's marketplace and market structure, 
 8   that can be extraordinarily damaging, which then moves 
 9   the asset manager away from buying those securities 
10   because of that cost, right?  So it's not solely just the 
11   actual cost of the liquidity itself, but the information 
12   that's embedded in finding that liquidity. 
13             MR. REDFEARN:  I think that's a great point, 
14   Brian, and it does appear that the -- those dynamics are 
15   very different when you look at the smaller cap segment, 
16   right?  So on a more liquid security posting a quote you 
17   probably have a pretty high probability of fill when you 
18   post that quote. When you go down you actually have 
19   potentially a greater chance of creating impact as 
20   opposed to getting a fill, and that helps to drive some 
21   of the dynamics that we see in trading.   
22             I guess, Jason, I wanted to turn back to you 
23   just one more time in order to get a, sort of the flip 
24   side equation of the process in terms of those 
25   circumstances when you do have some idea of where to go, 
0061
 1   right?  So you've looked at Bloomberg or AUTEX or 
 2   whatever.  You've seen who the holders are; you know what 
 3   firm is trafficking in the name.  What is the process 
 4   that's typically used in that scenario in terms of sort 
 5   of the liquidity search process, and how do you think 
 6   about sort of that in the upstairs market versus when it 
 7   makes sense to sort of be engaging more on the lit 
 8   markets? 
 9             MR. VEDDER:  That's a good question.  We're 
10   creatures of habit, so when we find places to trade, 
11   we're going to stay there and we're going to -- as soon 
12   as we find an engaged contra on the trades we're trying 
13   to do, we'll definitely stay in that venue and trade it 
14   till it's exhausted or one of us is exhausted.  Yeah, I 
15   mean, that's, that's where we try to -- we're not in a 
16   space of -- we're trying to get the best prices for our 
17   clients.  We're trying to reduce the prices for our 
18   clients; that's our number one objective on our trading 
19   desk.   
20             And that may involve going to a cash desk and 
21   pay more in commission if I know that the volatility is 
22   going to be less and I know that there's somebody there 
23   that I can engage with that's going to have a meaningful 
24   amount of volume for me to execute.  And other times it 
25   may be where I'm again getting back to the cat and mouse 
0062
 1   game and then I find a mousetrap that works and I sit 
 2   there and I use it for the rest of the day.  That's kind 
 3   of the life that we life in our, in the names that we 
 4   trade.   
 5             And I mean, there's not much else other than 
 6   that, other than just kind of going to where the 
 7   liquidity is. And to that point, I mean, if there was a 
 8   way that I didn't have to go to 25 different exchanges or 
 9   25 different venues to find the one person or one 
10   contraside to invest in, it makes it a lot more efficient 
11   and the information leakage reduces considerably because 
12   there's maybe one or two venues that I can trade at that 
13   I know that the other person can't go anywhere else.  
14   I'll find them at some point.   
15             Whether they're at exchange A or exchange B or 
16   whatever, if I can limit the number of places I've got to 
17   go visit to try and find something, that makes it a lot 
18   more efficient for me. 
19             MR. REDFEARN:  But just in terms of posting a 
20   quote, displaying a quote in the market, and back to the 
21   point that was being raised earlier about sort of the 
22   risk versus reward of doing so, when you think about that 
23   or in the algorithms that you use as a mechanism of 
24   soliciting the other side of an order, you know, how are 
25   you thinking about that when you're looking at the -- the 
0063
 1   algo, algo process that you're utilizing? 
 2             MR. VEDDER:  I don't want to be posting 
 3   anything, because I know that once I post something, 
 4   somebody else is going to see it and try and get ahead of 
 5   me.  I'm just being completely honest.  I mean, we try 
 6   and employ -- we do a lot of dark pool trading, but we 
 7   also do stuff where we're looking at the inside, at the 
 8   quote, and we're looking at trying to incorporate going 
 9   into a marketplace in a stock that hasn't, nobody's 
10   jumped to the far side of the quote to execute something 
11   in the last, you know, I call it minutes, but some people 
12   in this room call it milliseconds or microseconds.   
13             It's -- but as soon as I do that in my name, 
14   I'm going to see that -- the impact is going to be 
15   instantaneous.  I'm going to see it happen on my screen, 
16   more so than any other large cap or highly liquid name. 
17   The stock is going to move on me if I decide to go and 
18   post something that I may not know that anybody's out 
19   there or not, but just kind of randomly posting 
20   something, I run the risk of having it move, and without 
21   me getting anything done. So that's a problem.  That's a 
22   problem with how I have to enter into the marketplace in 
23   -- with the current structure, in terms of saying, okay, 
24   I've got to be really quiet and I got to be very careful, 
25   and I can't really post anything, because if I do I know 
0064
 1   that someone's going to see my footprint and try to get 
 2   ahead of me. And that person may not be, you know, an 
 3   institutional investor.  Which is -- I have no problem 
 4   with that.  But again, it comes back to that game where 
 5   you're just, you're trying not to expose  your clients' 
 6   interest because you know -- you know what the outcome 
 7   could be. 
 8             MR. REDFEARN:  So we do have a challenge of 
 9   price formation in this particular context with these 
10   sorts of names, and Frank, it looked like you were eager 
11   to say something.  But I'm also curious about, you know, 
12   when we think about the important price formation process 
13   that happens in names.  And obviously, you need to price 
14   merchandise.  I'm interested in how, how you think about 
15   that, but Frank, do you have any thoughts on how we 
16   should be thinking about that in this space? 
17             MR. HATHEWAY:  Well, I think that we certainly 
18   have the evidence, you know, in our study, of the 
19   challenges, I mean the volume going from least active to 
20   the most active increasing 50 fold, the depth increases 
21   two fold.  So there's clearly more -- block size stays 
22   essentially the same.  I think we've had statistical 
23   standard errors in here that would be the same.  To 
24   Jason's point, if you post a quote in an active name, 
25   you've got a 90 percent chance that the other side is 
0065
 1   going to move across your price in 30 minutes.  You do it 
 2   in a less active name from our analysis it's one time out 
 3   of three, the other side of the market's going to move 
 4   across you or one time out of three you're going to 
 5   trade, sorry, much less than that it'll be the other side 
 6   of the market moving, moving across you.  So there are 
 7   disincentives to do that. I do think to Steve's point, 
 8   the ability to have some reputational capital in the 
 9   market matters, with, you know, we have the ability in 
10   NASDAQ to be a lot -- nobody, in other words, to display 
11   a quote in your own name.   
12             When I got into the industry, which was 35 
13   years ago, 25, that's what everybody did.  And now, even 
14   though the ability to do so is still there, people don't, 
15   in the U.S.  In our Nordic markets, everybody displayed 
16   their own name.  We took it away and made everybody 
17   anonymous.  It did not go over well with the trading 
18   community, and today only the most active stocks on the 
19   Nordic markets have the ability to be anonymous.  
20   Everyone else indicates who they are.  So that's -- there 
21   is a reputational aspect to dealer markets.  And as I 
22   keep saying, dealer markets are what these stocks are all 
23   about.  Whether you choose to be an out-loud or not is 
24   your choice in terms of sending information to the 
25   market, but if you want to establish a reputation in one 
0066
 1   of these stocks as someone you should call or otherwise 
 2   contact to trade the stock, being able to put your name 
 3   in front of the market is one way to do it, and it's one 
 4   way it was done in the old days.  As Brian has indicated, 
 5   innovation is one thing; it's difficult to do when we're 
 6   all trading the same stocks and essentially constrained 
 7   to trade in the same way.  You can be anonymous, you can 
 8   be out-loud in a NASDAQ name now, but all the other 12 
 9   exchanges are anonymous, so why be out-loud on NASDAQ?  
10   You somehow give something away. 
11             Order delivery is another thing that comes up 
12   from time to time.  It is difficult to do in a Reg NMS 
13   context, in order times less than one millisecond.  
14   You're never, in my view, in these types of securities, 
15   going to completely solve the problem of getting people 
16   to reveal the absolute best price they're going to buy 
17   and sell at, nor should they do so.  But instead you want 
18   prices where people are -- "This is the price I'm willing 
19   to reveal, if you want to trade with me right now at this 
20   size and at this price go ahead and hit me; and by the 
21   way, there might be something better if you want to get 
22   in touch and we'll see if we can work something out."  I 
23   think by concentrating liquidity in an exchange you 
24   create a context as to "Okay, this is where I'm starting 
25   from.  This is where I'm going to be displaying my best 
0067
 1   number that I choose to make public; there may be a 
 2   better number behind it."  But you're clear on what it 
 3   is. And to Steve's point about the temporal versus 
 4   geographical fragmentation, temporal we're not fixing.  
 5   You see the order counts here.  Here is temporal 
 6   fragmentation; that's always been the case in less liquid 
 7   stocks. 
 8             Geographical fragmentation, it's pretty good, 
 9   Reg NMS, binding the markets together.  It's not as good 
10   as it might be.  Cboe groups, European exchange can trade 
11   and does trade everything in Europe, where as a practical 
12   matter they trade the most active.  And the less active 
13   securities, be it ours, be it Deutsche versus Euronext, 
14   tends to be the province of the local broker, the person 
15   who has the knowledge of the company and then more 
16   importantly has the knowledge of who's been a buyer, 
17   who's been a seller, who can I call if I have a client 
18   who wants to trade.  I think that's one the things we're 
19   looking to try and reestablish. 
20             MR. REDFEARN:  Just before turning it over, I 
21   know Ari wants to jump in as well as Jason.  I think part 
22   of the idea around no UTE or centralization is that 
23   presumably there's an opportunity to solve for both 
24   temporal and geographic fragmentation.  So the ability to 
25   innovate within a particular space allows innovation to 
0068
 1   occur in different ways.  It doesn't -- not 100 percent 
 2   clear to me that it needs a vehicle -- 
 3             MR. HATHEWAY:  Yeah, the temporal could be 
 4   addressed. 
 5             MR. REDFEARN:  -- a fully continuous market. 
 6             MR. HATHEWAY:  40 trades a day, a temporal is 
 7   going to be a little tricky, but yeah. 
 8             MR. CAVOLI:  But that's the central question, 
 9   right?  That's what we're arguing is, is it in fact 
10   interest in the underlying name or is it in fact the 
11   point that interest is spread anonymously across a 
12   multitude of venues?  I mean, we all have, you know -- 
13   NMS was great because it gave us free will, right?  And 
14   we see it right now.   
15             I mean, we see that the concentration at the 
16   NBBO is 23 percent.  And either the best bid or the best 
17   offers at 42 percent.  And when you go over a 100,000 
18   shares it basically drops off to next to nothing.  So we 
19   have the ability, we have the mechanisms today to 
20   concentrate liquidity.  We have it.  The market largely 
21   has kind of done that in the liquid names, right?  We 
22   kind of, we kind of see it.  And to underscore that point 
23   about interest in an underlying name versus the structure 
24   of the market you see in, I think it's page five, or 
25   where you look at the R2 names.  What you see in the R2 
0069
 1   names, they trade completely differently.  Same bucket, 
 2   R2, in an index, trade completely differently.  So 
 3   there's fewer shares traded over the counter.  The 
 4   concentration is equal and the amount of shares that are 
 5   traded in the auctions are greater.  So I think that -- 
 6   like we were really digging into this, and I know, Brian, 
 7   getting really tedious in inside politics around it is 
 8   tough, but you kind of sort of need to, we kind of sort 
 9   of need to really understand the essence of what we're 
10   trying to solve for and what tools we have at our 
11   disposal to address those issues.   
12             And I think as a broad, as a large matter, the 
13   great competitive market that we have since NMS was 
14   introduced and the technological revolution that followed 
15   or came in conjunction with, has done an okay job of 
16   addressing all these issues. 
17             MR. REDFEARN:  Ari, I know you wanted to jump 
18   in before. 
19             MR. RUBENSTEIN:  Yeah.  Agreed, it has.  I 
20   mean, we have the most liquid efficient markets in the 
21   world, and it's not perfect, and I think, Steve, you had 
22   mentioned that you know, the Commission deserves a lot of 
23   credit for changes over the last two decades to create 
24   that.  I think institutional investors, investors that 
25   have order flow that is price impactful, price moving, 
0070
 1   because it's so large, they want to control for two 
 2   things: information leakage and cost.  And in today's 
 3   market, the way that happens is the order will initially 
 4   start, as Jason mentioned, far upstream.  Meaning, you 
 5   know, at the local bank or custodian, to see if there's 
 6   any natural internalizations that might be able to occur. 
 7    Because, you know, in a bank dark pool or a broker dark 
 8   pool, that type of transaction is really efficient, 
 9   really inexpensive, and has the least amount of 
10   information leakage.   
11             And it moves through the system all the way 
12   down to the exchange.  And the exchanges are, you know, 
13   typically much more expensive to transact on, but there's 
14   a lot more information on the exchange in the market data 
15   that they have.  But the downside to that is that there's 
16   a lot of information leakage.  And I'm not -- I think 
17   centralizing price discovery in one place, that 
18   potentially might make trading easier for the 
19   institutional community dealing with larger sizes, but it 
20   has a whole host of other potential issues.  And if we're 
21   going to centralize just on the exchange, on one 
22   exchange, but allow ATSs to continue, then, well, there 
23   can just, there can be a large amount of ATSs that can do 
24   that trading that will continue to fragment.  And you 
25   still have the information leakage issue that you'd have 
0071
 1   on the exchange.  So -- and all at the same time, 
 2   potentially limiting competition for listings.  So I'm 
 3   not sure that that really is going to solve the problem. 
 4    I think what would help is things that can promote price 
 5   makers to commit capital to interrupt the order flow as 
 6   it's getting done so that more order flow can get done at 
 7   a better price. 
 8             MR. REDFEARN:  I do want to follow up, then. 
 9   What -- in terms of -- you mentioned before some specific 
10   innovations or changes to, you know, updating the market 
11   maker model.  What are you thinking? 
12             MR. RUBENSTEIN:  Well, to start, I think there 
13   are -- and again, this is, you know, we have a fantastic 
14   system, so it's not a complaint of the current system.  I 
15   do think there can be a little bit of a renovation in 
16   some of the rules applied to market makers.  I love to 
17   use this example.  In August of 2015 when the markets in 
18   half an hour traded enormous volumes, enormous 
19   volatilities in the morning, we were making markets at 
20   GTS, and in that first half hour we traded seven percent 
21   of the turnover.   
22             And we're making markets in ETFs and the 
23   underlying baskets.  And we came across this problem, 
24   because there were a lot of underlying securities that 
25   had gone into a rule 201 uptick restriction, which 
0072
 1   essentially says you can't short the security on a 
 2   downtick, because it had breached certain thresholds.  
 3   Well, that became a problem, because we were trying to 
 4   post liquidity in ETFs and the underlying securities, but 
 5   we couldn't deterministically know how we could hedge 
 6   those ETFs, because a number of the securities in the 
 7   basket were rule 201.  So you know, our computers weren't 
 8   able to sell when they wanted to sell.  And that -- what 
 9   ended up happening was, there was just, we just applied 
10   less liquidity and that cost investors money, right at 
11   the time where there was enormous demand for liquidity.  
12   So that's just sort of one example.  If you're doing 
13   bonafide market making, should there be an exemption for 
14   market makers with that sort of rule?   
15             I think there also rules regarding borrowing of 
16   securities for market makers, that could be modernized. 
17   And the easier it is for a market maker to post two sided 
18   quotes, the more liquidity there's going to be.  The more 
19   price discovery, you know, there would be.  So you know, 
20   I'm happy to commit, you know, my time and my team's time 
21   to you know, sort of highlight some of these and work 
22   with your staff to go over some of this.  And I think, 
23   not speaking on behalf of the larger broker dealers and 
24   banks, but you know, it's hard not to look at the 
25   decrease in capital and liquidity provided upstream and 
0073
 1   some of the other rules regarding capital and trading at 
 2   that level as well. 
 3             MR. REDFEARN:  Jason, you looked like you were 
 4   ready to jump in before.  Just to go back, you know. 
 5             MR. VEDDER:  Yeah, there's a simple dynamic in 
 6   these trades coming together, right? 
 7             MR. REDFEARN:  Mm-hmm. 
 8             MR. VEDDER:  It needs to be one price at one 
 9   time at one place, right?  And it's funny, because with 
10   the tickpilot we sort of dealt with the price 
11   fragmentation a little bit, but we didn't really deal 
12   with the place or the geographic fragmentation or the 
13   temporal fragmentation which was already there. And in 
14   today's world we do have an environment where we have 
15   stocks that trade 10, 20, 30 times a day in some cases, 
16   where they are trading on 13 exchanges, 30 dark pools, in 
17   microseconds in pennies, right?  And so I think there is 
18   a question as to for these illiquid names, whether or not 
19   this is the optimal environment for that. 
20             We're not looking to go, you know, with some 
21   sort of crazy rule-heavy approach here.  I think the idea 
22   is, is there a way that we can foster the ability from 
23   our -- innovate in a way that better serves some of these 
24   names.  So with that as the backdrop again, Jason, you 
25   know, I wanted to let you jump back in.  Because before 
0074
 1   you had mentioned some thoughts about things that we 
 2   might be able to do that foster that process. 
 3             MR. VEDDER:  I think Frank made a great point 
 4   about the reputational capital that firms have, that have 
 5   kind of gone away from when we kind of went to the NMS 
 6   model.  And I think that that is something that has dis-
 7   services this part of the marketplace.  Because there 
 8   were times before that if somebody was posting a quote, 
 9   and it was a firm that I knew that our research 
10   department used, that would be the first firm that we 
11   would talk to, to get on the phone and call them to ask 
12   them what they had going on.   
13             That's gone away; that's completely gone away 
14   in this day and age, with, you know, everything trading 
15   in multiple different venues; and you don't really see a 
16   lot of the regional firms making markets at the top in 
17   these names, because I think, quite frankly, it's hard 
18   for them to make capital in a name like that that is much 
19   more volatile than an industrial car maker.  And so I 
20   think that that's gotten completely lost in this whole, 
21   in this whole process.  And I appreciate Frank for 
22   remembering that, that there is a value to that, in the 
23   sense that you would foster relationships.  And it even 
24   happened on the floor of the New York Stock Exchange.  It 
25   was, you knew who the market makers were.  You knew who 
0075
 1   were good market makers; you knew who were not so good.  
 2   That you could be a little more transparent with what you 
 3   had to do.  And so I just wanted to make that comment.  
 4   As far as how you do this, I don't know if there's any 
 5   one answer to do it.  I think what would be interesting 
 6   to see is that we strip back all the -- make it much more 
 7   pure from the standpoint of a buyer and a seller coming 
 8   together in one spot. Not any kind of access fee; not any 
 9   kind of you know, stipulation on how much it has to trade 
10   or where it has to trade or at the mid or wherever.   
11             It's just, have it be a place that can foster 
12   liquidity where someone can come up; and you only have 
13   one or two spots to go to.  There's no trying to find it, 
14   you know, me getting six different platforms before I 
15   even get a nibble on something.  I can go to a place, I 
16   can say hey, I got something to buy; is there anybody 
17   here?  And if there is, then we can negotiate price.  And 
18   maybe it's outside the context of what the NBBO is at the 
19   time, because there may be an opportunity to trade more 
20   liquidity, and if I'm there first and the other person's 
21   there, then why not?  It just seems like it would be 
22   interesting to get back to a little more pure environment 
23   to trade in these names. 
24             MR. CAVOLI:  I couldn't agree with that more.  
25   I mean, the reason we're 12 percent market share is 
0076
 1   because of that bilateral nature of the relationship.  
 2   Like we wear our nametag prominently on our uniform and 
 3   we service our clients so it's precisely -- it's a very 
 4   good point.  It's precisely why we have the market share 
 5   we do. 
 6             MR. HARKINS:  But, if I just jump in again.  
 7   The -- let me just throw out an alternate way of looking 
 8   at it.  So we're talking -- the theme here is talking 
 9   about revoking UTP or centralizing.  But perhaps -- and 
10   then we're also talking about the pain points of Reg NMS, 
11   you know, the one size does not fit all.  We're sort of, 
12   we're trying to force these smaller cap illiquid names to 
13   trade within a Reg NMS environment.   
14             Perhaps the better way to look at is is saying 
15   instead of revoking UTP should we revoke NMS for these 
16   names?  I know that's a controversial thing to say and 
17   way to look at it, but that still allows for innovation 
18   from multiple exchanges to come up with the better market 
19   model to service Jason.  But you know, maybe they should 
20   just be outside of NMS in a, you know, in a typical Auto-
21   Ex environment.  If people want more of the old-fashioned 
22   relationships.  Or, electronify that process. That can 
23   happen outside of NMS on an exchange. 
24             MR. MONTEMAYOR:  I just want to make maybe one 
25   or two points here.  You know, I think we heard a lot 
0077
 1   from Jason on -- I made a note here, Jason; hopefully I 
 2   got it right.  I think my first hope is that it's in the 
 3   dark.  I just don't see how revoking UTP would address 
 4   any of that, right?  And I think the other point I wanted 
 5   to make is, you know, one thing that the -- Alex Jadin 
 6   maybe explained in his, when he covered the report, is 
 7   that natural market forces are already concentrating.  
 8   You know, there's volume in shares under, or ADVs under 
 9   50,000 to 1.8 exchanges, if I'm right.  Right, 1.8.  So 
10   natural market forces are already doing this, so why 
11   exacerbate the problem?  Right?  That's obviously not 
12   working, right?  Revoking the UTP would only that make 
13   that worse, right?  So I just don't understand how that 
14   could be a solution to a problem that we -- for 
15   securities that already have those characteristics. 
16             MR. REDFEARN:  I mean, look, this is a very 
17   good discussion, and obviously we're getting views from a 
18   multiplicity of viewpoints, which is exactly what we 
19   wanted to try to accomplish today.  Again, the -- part of 
20   the idea here isn't simply to concentrate liquidity, it's 
21   to allow a form of innovation that might not be allowed 
22   within the existing dynamic that we have within Reg NMS, 
23   because you know, when this came into effect in 2007, 
24   right, all of the sudden it did sort of say, okay, 
25   everybody, it's all going to be, we're all going to be 
0078
 1   doing Auto-Ex and we're all going to route to each other 
 2   and we're all going to do it this way.  And you saw 
 3   certainly the traditional model that partially was seen 
 4   on the floor, pretty much changed significantly during 
 5   that period of time.  Some of what Jason is looking at 
 6   doing may have moved to more of an upstairs market, 
 7   right?  It became, you know, it became a challenge I 
 8   think for some of the markets.   
 9             Part of the idea is, is there a way in which 
10   markets might be able to repatriate some of the liquidity 
11   into the market by innovating.  And we're not even going 
12   to presume how that's done.  And part of the next panel 
13   will be indeed that exact question, which is if this 
14   happened, what would you do?  What -- how, you know, just 
15   to stimulate some creative juices here, how can we better 
16   serve this set of companies in that environment? And so 
17   we'll have an opportunity to dig a little bit, a little 
18   bit deeper on all of those things.  I want to take a 
19   minute here and just see if any of the staff have any 
20   questions or also Commission Peirce, I don't know if you 
21   wanted to throw anything out?  You're good?  Anything 
22   from any of the staff right now on these topics?  So 
23   Frank, do you think that a fully automated market is -- 
24   when we think about these stocks, is still the right 
25   thing for stocks that trade that thinly? 
0079
 1             MR. HATHEWAY:  So when I talk about a fully 
 2   automated market, I talk about essentially a 
 3   deterministic electronic limited order book where you 
 4   have put your order in, you're committed to trade at that 
 5   price with whoever comes in to interact with you.  And 
 6   no, I don't think that's an ideal market structure for 
 7   these stocks.   
 8             And I think whether you want to talk about size 
 9   negotiation or delivery or however you want to go about 
10   structuring and innovation that moves away from an 
11   automated electronic limit order book, would be helpful. 
12    We've already talked about the fact, the size in these 
13   stocks seems to be fine, but yet we're talking about 
14   illiquidity at the same time, so I -- you know, the 
15   electronic limit order book always was and still is a 
16   dominant form of trading and probably the most successful 
17   form of trading on exchanges for your active securities. 
18    For your illiquid securities, it doesn't seem to have 
19   been particularly successful, whether we talk about 
20   evidence from the early days when the market's automated, 
21   or conversation such as we're having today that reflects 
22   the world of 2018. 
23             MR. REDFEARN:  And the other sort of dynamic 
24   that I think we'll want to talk about as well, it's an 
25   interesting outgrowth was this sort of identifying the 
0080
 1   market participant, right?  So it's almost like, there's 
 2   sort of like a select net you know, -esque kind of 
 3   dynamic that's suggested there, where if -- I mean, is 
 4   that what you're suggesting, Jason, that if you saw, that 
 5   if people were out there and you knew who they were and 
 6   similar to what Frank is saying he's witnessing in the 
 7   Nordic market, that that kind of environment might 
 8   actually foster an ability for trading that might be 
 9   different? 
10             MR. VEDDER:  Yeah, I think.  I mean, to Ovi's 
11   point, when I said about, my first hope is that it's in 
12   the dark, is because I want it to be in the dark so that 
13   I don't have to go and play the cat and mouse game across 
14   all the exchanges.  That was my comment, or that's what I 
15   meant.  I've been talking about this for a few years now, 
16   is that we've gone to the point where we're trading in 
17   microseconds and quotes are changing faster than the 
18   blink of an eye or whatever.   
19             And the one thing that has not changed in this 
20   marketplace is indications of interest. Indications of 
21   interest are still the old AUTEX model.  It's static.  
22   It's most of the times, it's stale.  And a lot of times 
23   it's not real.  And so I've posed this to a number of 
24   broker dealers that we have relationships with. I'm like, 
25   why can't we come up with a way to, for the sell side to 
0081
 1   create a matrix that they can, they can decide on as far 
 2   as profitability of the account or whatever, or how 
 3   vanilla I am versus a hedge fund or versus a high 
 4   frequency trader.  That I should be able to have more 
 5   transparency to what you really have in your books, 
 6   because that to me is going to be a much better way for 
 7   me to effectively communicate, as opposed to going down 
 8   through my best trading strategy to find liquidity into 
 9   the marketplace. It just, it shocks me that even to this 
10   day that we've gotten to this point where we can have 
11   server forms next to the New York Stock Exchange or 
12   wherever, that we haven't figured out a way to 
13   effectively allowed the sell side to traffic their own, 
14   their own information to the buy side, whether that buy 
15   side -- whatever side it is that they're interacting 
16   with.  But if I'm a vanilla, you know, 40X 
17   fund adviser, I'm not going to go and take advantage of 
18   something that, you know, Brian's firm puts out there, 
19   because I can't do that.  There's no way I can do that.  
20   I'm not allowed to do that.   
21             I'm trading the stuff that my PMs are putting 
22   in, whatever.  That the relationship is known that that's 
23   the way we operate as a firm, but why wouldn't we have 
24   100 percent transparency on what they're showing?  If 
25   they show that they have a size seller in stock xyz and 
0082
 1   I'm not even trading in xyz, I'm not going to -- that 
 2   doesn't make any difference to me.  But if they're a 
 3   seller of stock abc and I'm a buyer of it and nobody else 
 4   is seeing that information unless, because I'm being 
 5   transparent with them, and they know that, if I'm going 
 6   to be the contra side to that trade, the first call 
 7   they're going to get is from me. And it just surprises me 
 8   that that hasn't really been addressed at all.   
 9             I mean that would, I think it would help a lot 
10   with the cat and mouse game that I talked about before. 
11   If I just knew where to go, I would definitely go there. 
12   I would put my stake in the sand.  I would be like okay, 
13   let's trade.  And put my own abilities up with the other 
14   person that's on the other side of that trade.  But you 
15   know, I spend most of my time trying to search the 
16   landscape to find something, you know, just putting a 
17   pole into the water and just hoping something bites.  I 
18   would just be much easier for me in these names, 
19   particularly these names, to go and say hey, how do you 
20   guys want to do this, or where do you see this, we should 
21   trade this stock, so. 
22             MR. REDFEARN:  Actually I 100 percent agree 
23   with your characterization of the IOI marketplace, and I 
24   think it can be a much, a much more robust and useful 
25   tool.  I think there's been some efforts in the last two 
0083
 1   years to start to clean that up but we're still far away 
 2   from where we need to be, especially as it relates to 
 3   this range of securities. I think being able to start to 
 4   create much more distinct segmentation in IOIs that we 
 5   have today, between, you know, what are client natural 
 6   trades, what are firm natural trades, in terms of, you 
 7   know, having real interest to trade out of a position 
 8   versus market making and the other components, is a big 
 9   step in the right direction.   
10             But I still think that there's a lot of work to 
11   do to create the filtering aspect that Jason's talking 
12   about, to really make sure that the information that 
13   these IOIs contain gets to the right place, to the -- you 
14   know, to the actual beneficial owner, you know, of a 
15   security that looks to trade, versus splashing it out 
16   into the marketplace, where there's you know, there's 
17   plenty of electronification in this marketplace as well, 
18   in terms of consuming the information and then using that 
19   to internalize into various models, much as you see on 
20   the exchanges in trading and even in dark pools today.  
21   So I think there's, there's -- when we talk about 
22   innovation, I think the IOI functionality can be 
23   something that plays well into that, regardless of -- it 
24   could be on an exchange, it could be in an ATS, but I 
25   think that's an area that needs to be explored more, 
0084
 1   especially for the less liquid component of the market. 
 2             MR. REDFEARN:  So we're -- we're getting close 
 3   to wrapping it up here. I wanted to just ask for, since 
 4   we've had this discussion -- and again, a great array of 
 5   views here, just if people want to make some just closing 
 6   comments on what your takeaways are at this point in 
 7   time.  Adam, I actually wanted to start with you, because 
 8   I know you are talking to a lot of the issuers.  So from 
 9   what you've heard -- and you see a lot of people here who 
10   are obviously committed to our markets and committed to 
11   solving some of these challenges, what are your thoughts 
12   from this discussion so far? 
13             MR. EPSTEIN:  Yeah, you know, obviously, 
14   clearly there's not one answer.  But I think when you 
15   look at, when Alex was describing some of the data, it's 
16   interesting to try and understand, when you look at the 
17   bottom ADV bracket, and then you try and transpose that 
18   to real life, companies that exist in that bottom 
19   bracket, their lives as public companies is severely 
20   handicapped.  Severely handicapped.  For example, one of 
21   the reasons you go public, access to capital. Another 
22   reason is to use your currency.   
23             Well, when your stock trades once every three 
24   days or it trades in that bottom bracket of ADV, great 
25   acquisition targets actually don't want your stock, 
0085
 1   right?  So there's no such thing actually as mixed 
 2   consideration for most small public companies. You have 
 3   to -- and by the way, if your stock is already to the 
 4   point where nobody wants it, then you actually have to go 
 5   out and raise capital.  It's hyper-dilutive.  It's a 
 6   vicious circle, right?  
 7               And so another, some other just 
 8   really quick anecdotes that CEOs and CFOs wanted me to 
 9   pass along to the Commission -- hiring people; hiring and 
10   retaining employees. When employees come, whether it's C 
11   suite or VP level employees and they're looking to go to a 
12   company and they see that the stock either is horribly 
13   trading or to your point with respect to the size of 
14   those spreads, spreads is equal to volatility.  They see 
15   a stock that goes up 25 percent one day, down 30 percent 
16   next week, and they're looking at their stock options.  
17             They think, I don't know if I want to go to 
18   that company.  So not only is it -- no only is hiring an 
19   issue for these issuers but also retention. I've seen 
20   people in the last month, I've seen senior VP level 
21   people at companies in the Valley where I'm from in San 
22   Francisco, walk away from stock options that are in the 
23   money, because they look at the stock and they realize, 
24   this stock doesn't trade enough for me to monetize it, 
25   and by the way the stock is so volatile, who knows when I 
0086
 1   go to actually exercise it, whether it's actually going 
 2   to be in the money on that day, right?  And so, not only 
 3   that, a last anecdote -- there's a company that I advised 
 4   at the beginning of last year, and was recently talking 
 5   to the CEO.  They have a single source -- they're a 
 6   technology company -- single source supplier.  Not a 
 7   great situation, but it's a very unique situation. And 
 8   their stock, the volatility in the stock has become 
 9   extreme.  
10               Trading illiquidity a huge problem for them.  
11   The credit department of that supplier actually flagged 
12   the company and said, you know, what, I don't want to go 
13   down the path with this company anymore, giving this 
14   trade.  Something's wrong with this company.  Right? And 
15   so they ran into a situation where their sole supplier is 
16   having a problem with the company, 100 percent due to 
17   illiquidity.  Right?  So when you look at the bottom 
18   bracket of that ADV there's a real life out there.   
19             Yes, there's all the different trading 
20   mechanics and all the different investor issues, but 
21   these companies are not, do not have all of the access, 
22   the access that larger cap companies have and enjoy all 
23   of the benefits of being public companies, because of the 
24   liquidity; so a pretty austere challenge for these 
25   companies. 
0087
 1             MR. REDFEARN:  Steve, final thoughts? 
 2             MR. CAVOLI:  I think this is obviously a 
 3   serious topic and we just need to really focus in on very 
 4   specifically isolating the issue if there is an issue and 
 5   then making a determination as to whether any outcome 
 6   that we, that we implement is worth the cost, you know. 
 7   So I think that that's a critically important thing; we 
 8   cannot conflate issues.  We have to isolate the issue.  
 9   Is it really an issue?  Do we have tools at our 
10   disclosure, at our disposal, currently in the market, 
11   periodic auctions, we have them; continuous trading, 
12   consolidating to primary venues; we have the ability to 
13   do all that.  The question is, do we have the tools at 
14   our disposal right now to address the issue that we 
15   identified, as is the cost worth implementing. 
16             MR. REDFEARN:  Brian? 
17             MR. FAGEN:  Well, I think, you know, we've 
18   heard a lot of different views from a lot of different 
19   parts of the marketplace, and I think that's a theme that 
20   will come through over the course of the day; obviously 
21   we're talking more about ETPs at the end of the day.  But 
22   I think that understanding the entirety of the 
23   marketplace and creating solutions that actually meet 
24   needs of all investors is incredibly important here, and 
25   I think that -- you know, I think that the first part of 
0088
 1   solving anything is recognizing what the issue is.   
 2             And I think that the data in this presentation, 
 3   which is I think is actually quite excellent is a really 
 4   good way to approach it from a data driven approach to 
 5   decide what are we really trying to solve for, right?  
 6   And I think when you hear someone like Jason, who 
 7   represents a very large component of the marketplace, and 
 8   his frustrations, I think it definitely speaks to the 
 9   fact that there is a problem that needs to be addressed. 
10    And then I think that the current structure that we have 
11   probably isn't the whole answer.  Parts of it probably 
12   are, but innovation has to be part of that, you know, 
13   largely through technology, and I think that's a very 
14   good way to start with the process. 
15             MR. REDFEARN:  Bryan? 
16             MR. HARKINS:  Yeah, I think there's -- there's 
17   pretty strong agreement around the need to do something 
18   here.  And the way I look at it, again this is the 
19   exchange perspective, is we need to provide an ability 
20   and a structure where we can innovate.  We also need to 
21   provide choice; however, there's a difference.   
22             Today's Reg NMS forces choice on people, so I 
23   think exchanges should be allowed to innovate to solve 
24   Jason's pain points there, but again, taking away that, 
25   while still allowing choice.  But Reg NMS sort of forces 
0089
 1   that choice down everyone where you know, if Jason finds 
 2   an exchange for example that is a good meeting point, 
 3   whether it's electronic, floor, and sort of so-called 
 4   turn back the clock, perhaps he should be able to ignore 
 5   100 or 200 shares on some, you know, minor exchange.  So 
 6   I think competition has got to drive innovation, and as I 
 7   said it, I'll throw it out there, let's consider being 
 8   open minded around Reg NMS and potentially even revoking 
 9   NMS for these names. 
10             MR. REDFEARN: Frank? 
11             MR. HATHEWAY:  Yeah, a couple of thoughts.  
12   First, our application, as you see, is an opt-in 
13   application firm that -- an issuer that wants to 
14   participate; and limiting UTP across the exchanges can do 
15   so but there's no requirement.  If the Commission decides 
16   to take action in this area, however, our opt-out I think 
17   would be welcome.   
18             As we've heard today, with different 
19   discussions over whether it should be turnover or market 
20   cap or ADV that's a criteria for including a company, 
21   well it's still, replacing one size fits all with two 
22   sizes fits all isn't necessarily ideal.  So offering a 
23   company that thinks their liquidity is just fine the 
24   ability to opt out would be beneficial.  Also, to Brian's 
25   point, if another exchange offers an innovative approach 
0090
 1   to trading, perhaps an issuer on -- would like to try to 
 2   experience that in addition to their primary market 
 3   listings, so they could opt out and experience that.   
 4             The second thought is that I think it would be 
 5   dangerous for the companies involved if inclusion in a 
 6   program like this was viewed as some kind of scarlet 
 7   letter, some kind of mark that there was something 
 8   somehow wrong with these companies in the negative.  So 
 9   there may be a need, either -- and  the program we've 
10   proposed certainly reflects this, or action the 
11   Commission might take, to be somewhat over-inclusive than 
12   might be strictly necessary simply to avoid putting a 
13   negative connotation on any company that was in the 
14   program.  Thank you. 
15             MR. REDFEARN:  Ovi? 
16             MR. MONTEMAYOR:  We applaud the SEC for 
17   researching market quality statistics on these thinly 
18   traded securities.  I think it revealed many important 
19   insights into the differences in trading behavior based 
20   on ADV and security types. I just wanted to highlight 
21   that it's important to us that the retail's, you know, 
22   beneficial relationship with these wholesalers is 
23   preserved, to protect the current retail experience, and 
24   we hope to be able to continue these discussions and 
25   contribute in any way possible. 
0091
 1             MR. REDFEARN:  Thank you.  Ari? 
 2             MR. RUBENSTEIN:  Thank you. One data point I 
 3   wanted to share.  Of those 371 names in our portfolio 
 4   that we're responsible for at NYSE, we speak to them 
 5   almost on a daily basis and they're starving for content, 
 6   any content that can explain price discovery in their 
 7   security. The feedback is very different amongst all of 
 8   them.  They all have very different reactions as we 
 9   explain how price discovery occurs in today's modern 
10   market.  So for sure there has to be some sort of option 
11   for these companies.  They have to have a say in whatever 
12   it is that we end up doing. 
13             MR. REDFEARN:  And Jason? 
14             MR. VEDDER:  Thanks again. This has been great. 
15    I don't want anybody to leave here thinking that I'm 
16   like the English Patient or something like that, that I 
17   need to be revived.  We're doing fine, but things could 
18   work better, right?  But we know the issue.  The issue is 
19   that one size doesn't fit all, and we've heard it so many 
20   times from everybody that's been involved in this, this 
21   part of the marketplace, that you know, I think 
22   potentially pulling back some of the stuff, like you 
23   said, with respect to NMS or in this particular spot.  
24   And Frank and Ari saying that they think there should be 
25   an opt-out option for these companies.  If they don't 
0092
 1   want to do it, that's fine, you're on your own, you know? 
 2             But also getting to the point of kind of not 
 3   looking at it from just solely a trading perspective but 
 4   even to Adam's perspective of capital formation.  Is 
 5   there ways that you could actually create a venue or a 
 6   portion of the marketplace that could also relax some of 
 7   the, the regulations behind, you know, becoming public or 
 8   having to deal with all the overhead of having to be a 
 9   public company.  That really impacts a lot of these 
10   companies; we see it even now with some of the firms that 
11   we invest in.  It's, you know, there's a big cost to 
12   being a publicly traded company, and it's not just by, 
13   just for trading purposes.  It's the whole, the whole 
14   ecosystem.  So I hope this has been a good -- beneficial 
15   for you guys in terms of where you guys see this, and 
16   think that this a great stepping -- you know, starting 
17   point for us to think about this. 
18             MR. REDFEARN:  Look this, this is great.  We're 
19   getting a lot of different views from a lot of different 
20   perspectives here. That's extremely helpful. I think we 
21   all agree that there are issues here that should be 
22   addressed in some way and that there's more work to be 
23   done in terms of getting to the solutions. Right now 
24   we're going to take about a 15 minute break and at 11:45 
25   we're going to resume here and we're going to focus much 
0093
 1   more on some of the detailed potential solutions.  Thank 
 2   you all very much for your contributions today; 
 3   appreciate it. 
 4             (Whereupon, there was a recess from 11:30 a.m. 
 5   to 11:47 a.m.) 
 6                  MR. REDFEARN:  Good morning.  Thank you 
 7   all again for being here.  We have our second panel now, 
 8   where we are going to look at potential ways to improve 
 9   liquidity or improve the market for these illiquid 
10   securities, which we talked about this morning.  Again, 
11   thank you all for being here, and thank  you, 
12   Commissioner Piwowar, for joining us, as well. 
13             Generally speaking, on this particular panel, 
14   what we're going to do is we're going to start with 
15   allowing each of the exchanges to take a few minutes and 
16   talk about this idea, which was recommended by Treasury, 
17   of, if we did eliminate UTP, what might we be able to do? 
18    What are some of the alternatives that we have in this 
19   particular space? 
20             Again, the idea here is to foster an 
21   environment for innovation, and then, after we have those 
22   conversations, then we will open it up to the rest of the 
23   panelists and continue on with our discussion. 
24             So, with that being said, I will let each of 
25   you also introduce yourselves, and maybe, since it's set 
0094
 1   up this way, Tal, we'll start with you. 
 2             MR. COHEN:  Good morning.  My name is Tal 
 3   Cohen.  I am senior vice president and head of North 
 4   American equities for Nasdaq.  I've been that since 2016. 
 5    Prior to that, I was CEO of Chi-X Global, where we ran 
 6   alternative markets in Asia, Canada, and Latin America. 
 7             On behalf of Nasdaq and for me personally, I 
 8   would like to thank the staff of the Commission for 
 9   organizing and hosting this roundtable on the important 
10   topic of improving market structure for thinly traded 
11   securities.  I would also like to thank Director Redfearn 
12   and the Commissioners for their leadership on this issue.  
13             Nasdaq is the home to some of the largest and 
14   most well-known companies in the world, but we're also 
15   the home to many smaller issuers who aspire to follow in 
16   their footsteps. While the matter of how thinly traded 
17   names fare in today's market has become a more acute 
18   issue over the past few years, Nasdaq has long held the 
19   view that more should be and could be done to cultivate 
20   liquidity and improve the trading experience in these 
21   names. 
22             As Frank alluded to earlier, a year ago we 
23   memorialized our views through the revitalized blueprint, 
24   which sets out a comprehensive package of 
25   recommendations:  one, broad reform to the public company 
0095
 1   model; two, the promotion of long-termism; and three, 
 2   modernizing the market structure. 
 3             With respect to today's discussions, it's 
 4   important that we capitalize on the opportunity to 
 5   modernize market structure for companies that are the 
 6   growth engine of our economy.  The U.S. equity markets 
 7   exist to facilitate capital, which in turn fuels job and 
 8   wealth creation, ultimately driving economic growth. 
 9             If the volume of IPOs continues to fall and 
10   more companies continue to stay private or simply stay 
11   private longer, job creation/economic growth could 
12   suffer, and average investors miss out on attractive 
13   investment opportunities. 
14             To that end, it's time to address the one-size-
15   fits-all regulatory regime.  Many of the regulations that 
16   form the foundation of today's markets, which include Reg 
17   NMS, Reg ATS, and order handling rules, were developed 
18   and implemented more than a decade ago.  
19             While the evolution of the markets under these 
20   rules unlock competition and lower trading, broadly 
21   speaking, today's markets bear little resemblance to 
22   those of just a decade ago. 
23             So, now is the time to write new rules of the 
24   road to ensure public markets are as attractive as 
25   private markets.  We can accomplish this by embracing 
0096
 1   reforms that account for the different needs and 
 2   objectives of market participants. 
 3             As part of our revitalized blueprint, Nasdaq 
 4   proposed that the boards of such companies be empowered 
 5   to choose market structure models that would consolidate 
 6   on-exchange liquidity and has the potential to improve 
 7   the trading for their stocks and better serve their 
 8   shareholders, effectively advocating for issuer choice. 
 9             This structure where displayed liquidity is 
10   concentrated on an exchange would be complemented by an 
11   over-the-counter market to preserve choice.  Listing 
12   exchanges would then be motivated to innovate/introduce 
13   tailored solutions for thinly trading securities. 
14             Exchanges would, of course, be required to 
15   treat traders and market data consumers fairly in such an 
16   environment, and we understand that needs to be part of 
17   the discussion today. 
18             Providing issues with a voice will allow for 
19   better engagement about how the markets can work for 
20   their companies. 
21             So, in closing, we hope what follows sparks 
22   dialogue and action amongst investors, public and private 
23   companies, industry groups, and policymakers.  As always, 
24   the starting point is to clearly identify the problem and 
25   then present balanced solutions, but we must be clear, 
0097
 1   modern markets can and must be flexible to serve the 
 2   needs of these issuers and investors. 
 3             Thank you. 
 4             MR. REDFEARN:  Thank you, Tal. 
 5             Chris? 
 6             MR. CONCANNON:  Thanks.  Thanks for having me 
 7   here. I appreciate it. 
 8             As I think about market structure -- and first 
 9   of all, the exclusive listing concept is an interesting 
10   concept. I'm not sure who came up with it, but they 
11   should be commended for it, and Treasury should be 
12   commended for the idea, but I do think it's -- as I think 
13   about market structure, we're in this constant balancing 
14   act, balancing the needs of certain types of investors, 
15   retail or institutional, to access public companies, 
16   providing a marketplace where public companies can raise 
17   capital, and then there's intermediaries in the middle, 
18   intermediaries that are either brokers or market makers 
19   that support the stocks. 
20             The challenge that we have had over the years 
21   is finding the right market model for all the different 
22   types of securities that we list, and obviously, we list 
23   lots of different types, from ETPs, which is the next 
24   panel, to small companies, to very large liquid 
25   companies, and we clearly haven't found the right model. 
0098
 1             Over the years, we have had different model 
 2   approaches.  If you look back in the '90s, we had IPOs 
 3   listing on a quote-driven market, like Nasdaq, and then 
 4   graduating to a auction market like New York or the 
 5   American Stock Exchange back in the day, and those were 
 6   different market models that served properly -- properly 
 7   served the different types of securities. 
 8             Reg NMS is one market model, yet we apply it to 
 9   all securities, regardless of size, and I think we all 
10   agree, as you go across the industry, that one-size-fits-
11   all is not ideal. 
12             There's a variety of different approaches, and 
13   we recognized that in the tick pilot, but that was really 
14   just a small effort to move away from one-size-fits-all. 
15             The exclusive listing concept is interesting, 
16   but it's a bit of a blunt instrument, and in that, really 
17   the over-the-counter market does solve for the liquidity 
18   challenges of small companies, as you see a lot of the 
19   volume gravitate to the over-the-counter market, and it 
20   solves the retail demand for liquidity.  However, it 
21   doesn't solve the institutional demand for liquidity. 
22             I don't think the exclusive listing concept 
23   will solve that, that institutional demand for liquidity, 
24   yet we're going to give an exclusive to a single market. 
25     
0099
 1             So, when I think about what would I do with an 
 2   exclusive listing market as an exchange operator, 
 3   obviously I would want to pay rebates to market makers.  
 4   I wouldn't want to ban those rebates; I would want to 
 5   actually have them available to market makers 
 6             I would probably look at using some of the 
 7   auctions market structure models like allocations to 
 8   market makers, certain allocation methodologies to reward 
 9   market makers, and then there would be the concept of 
10   order delivery.  I think Frank mentioned that in his 
11   talk. 
12             Auto-quote, the auto-quoting of illiquid 
13   markets, is very difficult for a market makers, so 
14   allowing some type of quote obligation but order delivery 
15   mechanism.  
16             And then, finally, the challenge of the 
17   exclusive and why I think it's a blunt instrument -- if I 
18   had the ability to have an exclusive listing, I would 
19   likely increase my market data rates, because I would 
20   know people would have to come to our market for data. 
21             So, there's some detriment to awarding 
22   exclusive listings, or allowing it, and that could be the 
23   cost that some of these markets will charge in supporting 
24   an exclusive listing. 
25             So, I'm looking forward to everyone's comments 
0100
 1   and happy to engage. 
 2             MR. REDFEARN:  Thank you, Chris. 
 3             Stacey? 
 4             MS. CUNNINGHAM:  Hi, I'm Stacey Cunningham.  
 5   I'm the chief operating officer for NYC Group, and I also 
 6   would like to thank the staff and Commissioners Piwowar 
 7   and Peirce for having us participate here today to speak 
 8   on this topic. 
 9             So, NYC operates three separate unique listing 
10   venues.  One is predominantly large corporates.  We also 
11   have a small to mid-cap growth listing venue at NYC 
12   American, and NYC Arca is the largest listing venue for 
13   ETPs, with over 1,500 products, many of them thinly 
14   traded securities.  So, when we present our thoughts 
15   today, they're really also how do we think this would 
16   impact our issuers across our various market. 
17             So, what I was hearing this morning is that 
18   centralization of liquidity would be helpful, and we 
19   agree with that, with that theme, but some of the 
20   proposals that were presented aren't really going to 
21   centralize liquidity if you're looking at the data that 
22   we heard about from Alex this morning, if we're not 
23   looking at over-the-counter trading, as well. 
24             We've done our own analysis, as well, and very 
25   aligned with what we saw in the DERA report, and even 
0101
 1   just looking at securities that trade 50,000 shares a 
 2   day,  looking at Q4, the average security on an average 
 3   day  trading across five exchanges but 18 different off-
 4   exchange venues, and in the DERA report, we did hear from 
 5   Alex this morning that was kind of looked at as one 
 6   venue, but the reality is it's not, that liquidity is 
 7   really fragmented across those venues, and if we're 
 8   looking at trying to centralize liquidity and we're only 
 9   addressing the portion that is on on-primary exchanges, 
10   we're not really addressing fragmentation, especially 
11   because the profile of liquidity  in those securities is 
12   predominantly the percentages and market share that trade 
13   off-exchange are much higher in those securities. 
14             So, if you're talking about 56 percent trading 
15   of a 50,000 share name trading off-exchange and roughly 
16   24 percent trading on the non-primary exchanges, we're 
17   not really tackling that issue. 
18             Now, I think there is value in these less 
19   liquid names to have anonymity, to have less information 
20   leakage, and the retail investors, as we heard this 
21   morning, has a very good experience. 
22             So, my point isn't just saying, hey, everything 
23   should trade on one exchange, but if we're focused on 
24   centralization and we're not addressing the over-the-
25   counter market, we're not going to have a good result, 
0102
 1   and my point to that would be why would we do it, then? 
 2             Why would we take the additional complexity to 
 3   introduce a whole other market model that really isn't 
 4   going to solve the problem that we're trying to solve? 
 5             I think when you look at being able to innovate 
 6   as an exclusive primary listing venue, what we would do 
 7   would vary on whatever proposal might be out there.   
 8             If we're talking about not having over-the-
 9   counter exchange trading, as well, and really having a 
10   centralization of liquidity that had an exemption for 
11   blocks or an exemption for price or size improvement as 
12   most retail clients are getting from their wholesalers, 
13   there I think there is an ability to do a little bit on 
14   the innovation side, but when you talk about things like 
15   periodic auctions or other mechanisms that might be 
16   implemented, if there's over-the-counter trading, you're 
17   going to run into those same problems that we heard about 
18   this morning, that, hey, there's information leakage out 
19   there.  As soon as you see something, people are reacting 
20   on the other venues, and there isn't quite as much 
21   success. 
22             So, you know, we do think that the opportunity 
23   to innovate would vary based on whether or not all 
24   liquidity is centralized or not, but I would say our most 
25   important point -- and this is one area where we are 
0103
 1   aligned with Nasdaq on this -- is that the issuer should 
 2   have a choice. 
 3             Our issuers should be able to choose whether or 
 4   not they want to be part of a program.  So, we would 
 5   oppose the idea of having a pilot where issuers are put 
 6   into certain buckets with a certain market structure and 
 7   they don't have control over being included or not 
 8   included into that bucket. 
 9             So, if the Commission were hesitant to have a 
10   permanent structure that was adopted but wanted to have 
11   some way to try it, we would suggest a two-year trial 
12   period or some kind of trial period versus a pilot 
13   program, but importantly, that the issuers have the right 
14   to choose whether or not they want to be included and the 
15   right to  exit if they feel like it's had a detrimental 
16   impact on  them. 
17             So, that's, I think, our driving force.  
18   Otherwise, we feel like if we're not really centralizing 
19   liquidity, why bother? 
20             MR. REDFEARN:  Thank you, Stacey. 
21             We're going to move on to Brad, finish the 
22   exchange part. 
23             Brad? 
24             MR. KATSUYAMA:  Thank you. 
25             My name is Brad Katsuyama.  I'm a cofounder and 
0104
 1   the CEO of IEX.  The a stock exchange that we are 
 2   building really reflects our commitment to issuers and 
 3   investors, a simpler and less conflicted market 
 4   structure. 
 5             We definitely appreciate the Commission's 
 6   recent initiatives, which prioritizes those to 
 7   constituents.  We think that, you know, they are 
 8   absolutely critical to this market, and we definitely 
 9   appreciate the opportunity to participate on this 
10   roundtable, as well. 
11             When we think about the issue of thinly traded 
12   securities, I think we've seen from the tick pilot that 
13   tweaking market structure at times is not necessarily the 
14   optimal way to serve these small cap issuers or to expand 
15   opportunities for IPOs or research. 
16             However, the U.S. Treasury report did discuss 
17   some recommendations that issuers should be given the 
18   choice to limit trading to only the primary exchange or 
19   they did say a reduced number of exchanges, and that is 
20   an interesting idea and something worth discussion. 
21             On the first recommendation, we do believe, at 
22   IEX, that limiting trading to only a single exchange 
23   would not be an ideal outcome, for a number of reasons. 
24             First, by definition, it's anticompetitive in a 
25   way that the Commission has historically rejected.  You 
0105
 1   know, we've already kind of seen the way -- and I think, 
 2   you know, Chris mentioned this a little bit -- we've seen 
 3   the way that exchanges use exclusive power to sell market 
 4   data connectivity.  It shouldn't necessarily give us 
 5   confidence that granting new ways to extend market power 
 6   is a good answer. 
 7             Secondly, we've seen how a multi-market or a 
 8   multi-exchange environment does provide stability to 
 9   issuers and investors. 
10             We had a major exchange shut down for nearly 
11   four hours a few years ago.  There was absolutely no blip 
12   on the radar for issuers or investors, and I think, 
13   without that, within a multi or a competitive exchange 
14   environment, that would not have happened. 
15             Finally, I think Stacey mentioned this.  You 
16   know, giving exclusive status to a single exchange 
17   doesn't address the issue of off-exchange fragmentation. 
18    So, I'm not necessarily sure that granting that monopoly 
19   status for a single exchange but not limiting off-
20   exchange trading would necessarily be the best outcome. 
21             On the second recommendation, limiting trading 
22   to a smaller number of venues, we do think, while we 
23   would recommend that looking to limit trading to one 
24   market per exchange family for thinly traded names could 
25   be a constructive step to improve liquidity for small cap 
0106
 1   securities, for a few reasons. 
 2             First, you know, there are 13 stock market 
 3   licenses in operation, but are 3 exchange companies that 
 4   are operating 11 and soon to be 12 of these 13 exchanges. 
 5    So, it's a market that's been excessively fragmented but 
 6   hasn't proportionally increase competition. 
 7             You know, just to say it simply, 13 stock 
 8   market licenses in operation, but there are not 13 
 9   competing stock markets. 
10             If you think about it, Nasdaq BX, I don't 
11   believe competes with Nasdaq.  If they did, I think 
12   Nasdaq BX would write research reports saying that the 
13   execution quality for trading displayed on Nasdaq BX is 
14   far superior to that of Nasdaq, but Nasdaq has longer 
15   lines to buy or sell stock, and I think the fact that you 
16   don't see that research means that we have three exchange 
17   companies, we have a lot of exchanges, but they're not 
18   necessarily competing with each other.  So, excessive 
19   fragmentation, I think, is something that I think needs 
20   to be looked at. 
21             Second, limiting trading to one market per 
22   exchange family does provide critical data to the 
23   industry, provides critical data to the regulators.  I do 
24   think giving issuers choice is really important, but what 
25   is the basis for making that choice, and I think showing 
0107
 1   them data around the benefits of reduced fragmentation -- 
 2   maybe there aren't benefits, but at least providing that 
 3   data to these issuers would be a step to help them make 
 4   an informed choice. 
 5             And I think, last, which kind of couples with 
 6   the second point there, we think a pilot program would be 
 7   an interesting step, a pilot that basically looked to 
 8   limit small cap stocks, where they traded, and again, it 
 9   would be the best way to test, study, and implement this 
10   change before giving issuers this choice. 
11             So, we thank you for your invitation to 
12   participate.  We look forward to the discussion. 
13             MR. REDFEARN:  Thank you, Brad.  Thank you very 
14   much. 
15             So, we're going to now turn to non-exchange 
16   participants on the panel to get some high-level views. 
17             So, Brian, if you want to introduce yourself 
18   and give us your thoughts. 
19             MR. FRAMBES:  Sure.  I'm Brian Frambes.  I'm 
20   the co-head of Global Cash Trading at Fidelity 
21   Investments, obviously not an exchange. 
22             I wanted to thank Director Redfearn and staff 
23   for the invite, and Commissioners Piwowar and Peirce for 
24   attention to this matter.  It's pretty important to us. 
25             Asset managers like Fidelity play an important 
0108
 1   role in the financial markets.  We provide investment 
 2   products to millions of individual investors. 
 3             Fidelity has $800 billion plus of AUM dedicated 
 4   to the equity markets.  I took a snapshot last week.  We 
 5   own 2,000 U.S. securities with market cap of less than a 
 6   billion dollars.  Seventy-five percent of those trade 
 7   less than $3 million a day. 
 8             I think that notional is a better way to look 
 9   at liquidity rather than shares.  I think that was 
10   mentioned in the panel previously. 
11             However, the fact that small cap equities don't 
12   trade often is not evidence of an inferior market 
13   structure. That's been Fidelity's position for a while 
14   now. 
15             Factors that primarily shape low liquidity are 
16   small floats, highly convicted owners of those 
17   securities, and I think you have to look at index 
18   inclusion, also, as part of that problem. 
19             That being said, we are very supportive of 
20   exploring opportunities.  We're not afraid of 
21   experimentation to improve our transaction experience 
22   around the margin, and look forward to taking part in the 
23   discussion. 
24             MR. REDFEARN:  Thank you, Brian. 
25             Joe. 
0109
 1             MR. MECANE:  Thank you again for including me 
 2   today. 
 3             I'm Joe Mecane from Citadel Securities.  We're 
 4   the largest U.S. market maker.  We trade virtually ever 
 5   U.S. equity.  We make markets on every U.S. exchange.  
 6   We're part of one out of every five shares that trade in 
 7   the U.S., or about 40 percent of the retail execution 
 8   market.  We're the largest DMM on the floor of the NYC by 
 9   a number of issues, including about a third of which have 
10   under 100,000-share ADVs. 
11             And I mention all that because it gives us 
12   unique perspective on what it means to trade these types 
13   of securities.  We interact with retail firms.  We 
14   interact with issuers.   
15             And I would say that we are broadly very open 
16   to any market structure changes that help improve 
17   liquidity, especially in this segment of the market, but 
18   would highlight two overarching points that we think are 
19   important to keep in mind with any types of changes along 
20   these lines. 
21             The first, which partly echoes the point Brian 
22   made, is that we need to be very specific about what 
23   goals we're trying to achieve with any of these types of 
24   changes.  I give a simple example of $500 stocks that 
25   trade 100,000 shares a day, of which there are several, 
0110
 1   and a $5 stock that trades 10 million shares a day being 
 2   equivalent turnover, but we're defining one as liquid, 
 3   one as illiquid, and so, that's one characteristic. 
 4             The other that I'll highlight, as Brian alluded 
 5   to, is the index component, which the fact that there are 
 6   less liquid securities that trade with different 
 7   characteristics partly implies that perhaps other things 
 8   than market structure are behind some of the changes that 
 9   we see across the liquidity spectrum. 
10             The other point I would highlight is that, as 
11   Alex alluded to at the beginning and is prevalent in the 
12   memo, we do see higher off-exchange trading in a lot of 
13   the less liquid names.   
14             A lot of that is because those stocks are 
15   dominated by retail investors, and so, retail, as was 
16   highlighted on the last panel, does receive a 
17   disproportionate benefit of the trading that happens in 
18   those names. 
19             So, any changes to that segment of the market 
20   need to be contemplated with an eye on ensuring that 
21   retail investors are not harmed through whatever changes 
22   we do decide to make. 
23             Thank you. 
24             MR. REDFEARN:  Thank you, Joe. 
25             I think I'm going to go to Owain Self first, 
0111
 1   Bob, and then we'll get back to you at the end, if that's 
 2   okay. 
 3             So, Owain, let me jump to you. 
 4             MR. SELF:  Thank you very much for having me 
 5   today. I'm Owain Self.  I'm the global head of execution 
 6   services at Millennium Management.  I would be remiss if 
 7   I wasn't to say that all the opinions I offer today are 
 8   personal and not representative, necessarily, of 
 9   Millennium Management or its investors. 
10             In my role as the global head of execution 
11   services, one of those components is to run our trading 
12   desk for our equities fundamental business, as well as 
13   working with our quantitative portfolio managers in our 
14   execution strategy. 
15             We are both a large user of the brokerage 
16   services in both high-touch and electronic services from 
17   the sell side, as well as building our own execution 
18   algorithms and using appropriate market access pipes for 
19   our brokers to the exchanges. 
20             So, we're a true firm that operates on the 
21   whole spectrum of the market microstructure. 
22             A couple of comments I would say, with what's 
23   already been said in some of the other areas, is the low 
24   liquidity or small cap or thinly traded securities in the 
25   markets have always generally been a neglected part of 
0112
 1   the market. 
 2             I think this neglection was generally helped by 
 3   people in the past with large trading desks on the sell 
 4   side and large sales desks and research, and as the sell 
 5   side has rationalized their business model and they've 
 6   looked to the profitability of individual sections and 
 7   subsections of their businesses, this area has been 
 8   somewhat neglected. 
 9             And whereas, if you count the trading desk in 
10   some blue chip liquid names, technology has had the 
11   ability to take over and replace that human capital, it's 
12   been less effective in some of the lower ends of the 
13   spectrum of the market capitalization. 
14             And you know, what we're discussing here today 
15   is how technology and how the market structure can 
16   actually accommodate some of those changes, which are 
17   far, you know, broader than we talk about here, and as 
18   mentioned already on the first panel this morning about 
19   the -- not just the market structure but the interest of 
20   participants in the space with regards to capital 
21   raising, going public, so I think, you know, it's a 
22   challenge to try and solve everything with just the 
23   market structure, but I think, you know, there are 
24   definitely things that can be done. 
25             I would echo some of the comments from other 
0113
 1   people around the allowance for more innovation and 
 2   multiple tiers of segment of market to support different 
 3   types of stock, but one thing I would say -- and the last 
 4   point -- is just to really say a word of caution, is we 
 5   don't want to create a false sense of liquidity, and when 
 6   we talk a bit about looking at how many shares trade on a 
 7   daily basis, and that may or may not be the catalyst to 
 8   perceive a stock is liquid, if we have shares just 
 9   trading hands between intermediaries and not between real 
10   investors and creating a false sense of liquidity, it may 
11   ultimately mean that, when people look to liquidate 
12   things, the cost is disproportionate to what they 
13   believed it would be. 
14             And I think we saw that -- we see that 
15   significantly in parts of the market structure today, 
16   where certain, you know, practices have meant that people 
17   believe that liquidity exists, but when it actually comes 
18   to it, shares trading on a screen doesn't mean anything. 
19             I think we have to be clear that anything we 
20   try and propose, you know, throughout your initiatives is 
21   that it tries to drive true liquidity and not just the 
22   pretense of it. 
23             MR. REDFEARN:  Thank you, Owain. 
24             Bob? 
25             MR. SCHWARTZ:  I want to thank you, Brad and 
0114
 1   Commissioners.  It's a pleasure to be here and have this 
 2   opportunity.  I am a representative, I guess, of 
 3   academia, and I want the disclaimer to be clear that I do 
 4   not represent -- my opinions are not accepted by all 
 5   academicians, and I've had some very good papers rejected 
 6   for that very reason. 
 7             Just so you know my background, I had a PhD in 
 8   economics, liberal arts economics, not business school, 
 9   from Columbia University.  It's an excellent background 
10   for dealing with the sorts of issues that I've gotten 
11   involved in, and I've been specializing in the equity 
12   market microstructure before -- I don't want to give away 
13   my age or anything, but before the field was given a 
14   name. 
15             My specific focus in microstructure is -- and I 
16   think it's the reason why maybe you thought of me here, 
17   because my focus is market structure, but I want to put 
18   in, also, it's not just market structure but also the 
19   dynamic process of price formation, how does it work. 
20             In so doing, I have paid major attention on 
21   price discovery, which I hear mentioned in settings like 
22   this rarely.  It's mentioned, but not much.  Bid/ask 
23   spreads dominate. 
24             My recent work has to do with liquidity and 
25   liquidity premiums.  It's largely empirical but with a 
0115
 1   conceptual framework, and we're finding that, for all 
 2   stocks, that there is a liquidity premium.  It could be 
 3   higher for the microcap and following this, but I would 
 4   like to study that more. 
 5             But it just underscores the importance of this 
 6   topic, because if there's a liquidity premium, it's 
 7   because share prices are trading at lower levels. 
 8             In my market structure focus for decades, I 
 9   have paid particular attention to call auction trading, 
10   so maybe we can get into that a little bit, but I will 
11   say that I was at a conference a number of years ago on a 
12   totally different subject and I was introduced by the 
13   moderator as Bob "Call Auction" Schwartz.  So, that's my 
14   full name on my passport now. 
15             But I have several publications on call 
16   auctions, and I have two empirical ones that show that 
17   price formation after the introduction of call auctions 
18   is enhanced.  One of them is from the Paris exchange.  In 
19   '96 and '98, they instituted a closing call, and then, of 
20   course, there's Nasdaq's introduction of its opening and 
21   closing calls in 2004. 
22             I also have been concerned about how additional 
23   liquidity can be attracted to a market, especially the 
24   microcaps, for today's focus. 
25             So, this is all, in my opinion, terribly 
0116
 1   important, and I opened with a thanks.  I'll close with a 
 2   thanks for having me here. 
 3             MR. REDFEARN:  Thanks, Bob.  That's great. 
 4             What I would like to do, maybe, is go back, and 
 5   Tal, in the last panel, Frank Hatheway mentioned that you 
 6   guys were going to put out -- put in something that's 
 7   sort of looking at sort of proceeding with sort of a no-
 8   UTP environment, and I think we just wanted to get a 
 9   little bit of a deeper sense about what are your 
10   thoughts, what would you do if we contemplated something 
11   like that?  What would you do in that space? 
12             MR. COHEN:  So, Frank referred to a UT 
13   application that we're going to put out imminently, and 
14   in it, we just tried to show some thought leadership.   
15             We knew there would be a dialogue and debate on 
16   the topic, and we thought it was important enough to get 
17   in front of it, and so, we used the Choice Act and the 
18   Garrett bill, at least that was going through the House, 
19   to use that as a starting point in terms of thinking 
20   about how to define the market. 
21             So, I think, in the latest iteration, it's 
22   something like 700 million to 2 billion, was the 
23   inclusion metrics. 
24             We started off, also -- and we've heard it from 
25   almost all the panelists, and so, there is real agreement 
0117
 1   on this -- it has to be about investor choice.   
 2             So, you will see us propose and advocate for 
 3   investor choice.  You will see us try to be consistent 
 4   with the legislation that's being spoken about, even 
 5   referred to in the Treasury bill, and then, when we think 
 6   about the different components -- we talked about index 
 7   inclusion.  We think that's a big part of the story 
 8   today. 
 9             So, when you think about broadening the 
10   investor base and how do you do that, index inclusion 
11   plays a big role in that, and many of the metrics we just 
12   talked about here are either market cap or ADV-related, 
13   but there's really good research around turnover as a 
14   percentage of ADV and looking at free float. 
15             And so, there are markets in the world that we 
16   can compare ourselves to that run different market models 
17   what we would consider thinly traded, and how do we fare 
18   or stack up against them when we use that as the key 
19   metric? 
20             Now, what we also heard as we went up and down 
21   here is there are some concerns around consolidating 
22   liquidity.  When we think about consolidating liquidity, 
23   we think about it simply as an enabler. 
24             We understand that there are risks with 
25   consolidating on-exchange liquidity but also preserving 
0118
 1   choice, because we understand that the ATC markets work 
 2   for a reason, whether it's retail, whether it's price 
 3   improvement or size discovery, we understand the value of 
 4   that, and we don't see the parent order.  We see child 
 5   orders come through. 
 6             So, we need to be, also, careful of our view of 
 7   the market, but it's not necessarily about monopoly and 
 8   taking that liquidity and centralizing it, because we can 
 9   address fairly easily the transaction concerns. 
10             So, we have a new transaction fee cap today.  
11   We can discuss, by the way, if that's the right number 
12   for thinly traded. 
13             On market data, in our proposal, because you've 
14   asked, what we offer up is language around not putting 
15   any market data products around -- new market data 
16   products around thinly traded securities, if this were to 
17   come about. 
18             We also wouldn't look to charge more for 
19   existing products as a result of ETP revocation 
20   increasing the value of that product. 
21             And finally, we would have to look at the SIP 
22   formula and figure out if it still made sense to have 
23   that as part of the SIP formula. 
24             So, I think they're all solvable, and it's been 
25   referred to as a monopoly or anticompetitive, but 
0119
 1   remember, you'll have off-exchange.  You want to incent 
 2   the exchanges to innovate. 
 3             Right now, if myself or the other exchanges 
 4   look to come up with a new order type, pretty quickly 
 5   we're going to copy each other, and then, we're in the 
 6   same state of affairs. 
 7             And so, it's not just about awesome innovation. 
 8    I would add it's about risk management.  And so, when 
 9   you think about a consolidated on-exchange platform, 
10   we're talking about potentially de-risking, during times 
11   of market distress, how markets work. 
12             Somebody referred to August 24, 2015.  
13   Complexity in the markets kind of led to that, and so, 
14   it's not just about innovation; it's also about helping 
15   manage risk. 
16             MR. REDFEARN:  Any specifics in terms of the 
17   market model, how the order book would work or what 
18   market makers would do or anything like that?  I just 
19   want to take it down one more notch, if possible. 
20             MR. COHEN:  So, I'll list out a few things that 
21   we've been talking about and think that would be 
22   interesting. 
23             So, Frank referred to it as reputation capital, 
24   but it's attribution, and there are markets like South 
25   Korea, China, Canada that have embraced attribution, and 
0120
 1   so, it's interesting to look at how attribution plays a 
 2   role for thinly traded securities. 
 3             So, if a regional broker steps up and wants to 
 4   own that name or own that space, as they do in many 
 5   jurisdictions and geographies, does that help?  So, 
 6   that's one. 
 7             We talked about disabusing ourselves from price 
 8   time, maybe price size, pro rata.  All of that's done -- 
 9   was trying to incent competition. 
10             So, what we'd like to do, if possible, is 
11   incent competition.  It wouldn't necessarily be about 
12   taking the ADV from 10,000 to 20,000 but making the 
13   trading experience much more positive along the way. 
14             And then the last point would be -- we're 
15   talking about auctions and we talk about intraday 
16   auctions, but also, in Europe, look at end of day 
17   auctions.   
18             You know, you have to consider the index 
19   inclusion/exclusion part of it, but we could consider 
20   looking at end of day auctions as some of the other 
21   geographies do, and looking at -- and comparing and 
22   contrasting -- what goes up in our auction for -- for 
23   thinly traded versus what goes in, if you will, South 
24   Korea, Canada, China, some of these other markets that 
25   have embraced different market models for those 
0121
 1   securities. 
 2             MR. REDFEARN:  So, Stacey, could you elaborate 
 3   on some of your thoughts in terms of what a model might 
 4   look like?  You've mentioned a few of the aspects of 
 5   that, but you know, if we were to move forward in 
 6   contemplating something, what would you be -- 
 7             MS. CUNNINGHAM:  I guess it comes down to -- 
 8   our concern would be, it would be harder to innovate if 
 9   we're just talking about centralizing the on-exchange 
10   venues, because it's already working like that today, 
11   that most of the liquidity is centralized on the primary 
12   when you're just looking at the exchange venues. 
13             So, I think some of the things that might be 
14   interesting to try for less liquid names would be hard if 
15   there's still, you know, over half of the market trading 
16   off exchanges, over the counter, and any mechanisms that 
17   you try could lead to information leakage. 
18             So, I think it really does depend on that 
19   structure.  I think we need to recognize that these 
20   securities have always been hard to trade.  It's not a 
21   new phenomenon that it's hard to trade, you know, some of 
22   the small cap names or less active names. 
23             And so, you know, I think if we were talking 
24   about purely centralizing on a single venue, we would 
25   consider things like auctions where you could, you know, 
0122
 1   attract buyers and sellers together, but I think that 
 2   would be hard to do if there are other markets actively 
 3   trading, and we would also want to be cognizant of the 
 4   impact on derivatives markets if you're suspending 
 5   trading for some period of time and just periodically 
 6   trading throughout that course. 
 7             One thing I would say, though, if we did get to 
 8   a place where we said we really do want to centralize 
 9   trading in just one venue, we agree with some of the 
10   other sentiments that we think it would be appropriate 
11   for exchange to agree, and we would agree to fee caps and 
12   things like that, if you're talking about just having one 
13   market actively trading. 
14             MR. REDFEARN:  Is the notion on the over-the-
15   counter market like just centralizing of exchange or is 
16   it trade add, block exception?  Is there something other 
17   nuance that you've thought about? 
18             MS. CUNNINGHAM:  Yeah.  I mean we think it 
19   would be appropriate to have bulk exemptions or to 
20   recognize the value of price and size improvement that, 
21   you know, predominantly goes to retail or whether it goes 
22   to institutional, as well, if we're talking about price 
23   and size improvement, but if we're just talking about 
24   run-of-the-mill trading that might be even, you know, at 
25   the NBBO, I don't think we're really solving the problem, 
0123
 1   since most of that trading is centralized within the 
 2   exchange groups already. 
 3             MR. REDFEARN:  Chris, so if something went 
 4   forward -- and clearly I've got your views from the 
 5   initial comments, but what would -- what would you think 
 6   about doing, specifically, in that market environment? 
 7             MR. CONCANNON:  Sure.  I agree with Stacey.  I 
 8   think that the challenge of an exclusive listing is -- 
 9   you still have fragmentation of liquidity in the over-
10   the-counter, and if you don't solve for the over-the-
11   counter  fragmentation, you're awarding a benefit to an 
12   exchange that -- which probably doesn't bring a lot of 
13   benefit to investors.  So, it's just a balancing act, 
14   what to do with the over-the-counter market. 
15             I think in terms of some of the innovations 
16   that you can use in the exchange world, obviously, 
17   allocations to market makers, truly how do you incent a 
18   market marker to put up size and commit capital, and do 
19   you have to do that through micro-market structure means 
20   and awarding them some benefit to being there, whether 
21   it's an allocation, a sizeable allocation, whether it's 
22   renumeration through rebates, whether it's some other 
23   exclusive benefit that you give them. 
24             You can use fast and slow markets where they 
25   have a fast response, faster than other market 
0124
 1   participants, so they can protect themselves. 
 2             And then there is just the concept of taking 
 3   away automated quotes, and we see that -- that was the 
 4   history of Nasdaq, was a nonautomated quote, and I would 
 5   argue some of the reasons why we have Reg NMS was the 
 6   damage that was caused by nonautomated quotes in the most 
 7   liquid markets. 
 8             But in a less liquid name, nonautomated quotes 
 9   is probably more reasonable.  You still need a firm quote 
10   rule, but after that, they can make decisions on how to 
11   deal with liquidity coming at them. 
12             So, I think there's a number of things that you 
13   can do.   
14             Some of those ideas apply to a non-NMS market, 
15   as well.  So, rather than giving a UTP exclusive, you can 
16   look at Reg NMS at the bottom end of our market and think 
17   of new ways to incentivize exchanges without tying us all 
18   together through the rigors of an NMS rule with automated 
19   responses and automated quotes. 
20             So, there's -- that's why I'm worried about the 
21   exclusive listing as being a blunt instrument to solve 
22   for liquidity fragmentation when there's probably other 
23   methodologies that we can use in our current market 
24   structure before we get to an exclusive listing concept. 
25             MR. REDFEARN:  Thanks. 
0125
 1             Brad, when you think specifically about your 
 2   market environment -- and I know you're looking to build 
 3   out the issuer environment, but is there something that 
 4   you would think about offering specifically there? 
 5             MR. KATSUYAMA:  I mean, I would just echo some 
 6   of the points that were discussed previously.  I mean, 
 7   you know, the elephant in the room is off-exchange 
 8   trading.  So, this was talking about consolidating 
 9   trading down to a single venue. 
10             I think that would be -- that's definitely an 
11   interesting concept, but if you're not going to address 
12   both sides of it, on- and off-exchange trading, then I 
13   think it becomes one of kind of, I guess -- it's a 
14   questionable step. It's a big step, but it's also a 
15   questionable one. 
16             I think, as we discussed it internally, part of 
17   the reason we came back and said, you know, a pilot might 
18   be interesting in terms of limiting the number of 
19   exchanges, gives each exchange family an opportunity to 
20   put their best model forward.   
21             It allows competing models and innovation to 
22   kind of, you know, try to attract an order flow, and you 
23   know, the Commission might find that the data, you know, 
24   gives very strong support to further consolidation, and 
25   then, at that point in time, down the road, you can make 
0126
 1   a more informed decision about trying to further 
 2   consolidate liquidity to a smaller number of venues, both 
 3   on and off exchange. 
 4             So, right now, you know, going from, you know, 
 5   I think Stacey's data were, you know, 5 exchanges and 18 
 6   off-exchange venues, it's going from 5 to 1 and leaving 
 7   that 18 number.  It's still massively fragmented. 
 8             So, incremental steps, we think, are really 
 9   important, and I think this is a data set that a lot of 
10   people would value, because we complain a lot about the 
11   fragmentation of our market, the complexity of our 
12   market, the costs. 
13             One thing that we haven't talked about is that, 
14   you know, small, mid-size brokers, even some of the large 
15   brokers, continue to get taxed with the cost of 
16   connecting to all these markets, and I think, you know, 
17   those costs have gotten pretty egregious. 
18             So, it's just a -- it's a market structure that 
19   isn't creating a lot of competition on the exchange 
20   front, also not a lot of competition on the broker front. 
21    You're not seeing a lot of small and mid-size brokers 
22   who are the ones that were covering some of these 
23   companies, writing research, etcetera. 
24             You're not seeing, you know, the formation and 
25   growth of a lot of those brokers like you did in the 
0127
 1   past, as well. 
 2             MR. REDFEARN:  So, I want to turn for a second 
 3   to Brian and Owain to just talk about the sort of off-
 4   exchange dynamic here and to think about it, because part 
 5   of what we're thinking about is how to create innovation, 
 6   and there's one potential path that would be create 
 7   innovation and have that be a means of attracting more 
 8   liquidity into the market and solving that way.  Another 
 9   one would be sort of more rule-based. 
10             And I guess, given that the over-the-counter 
11   market right now is -- in terms of whether it's a 
12   negotiated market -- and Joe, we'll get to the retail 
13   market -- a significant way in which liquidity is formed, 
14   you know, what are your thoughts in terms -- about how we 
15   should be thinking about the future of sort of the 
16   upstairs market as we think through possible 
17   alternatives. 
18             MR. FRAMBES:  There's a lot to unpack there. 
19             So, dark or off-exchange is a very important 
20   tool for us, and it's about managing about information 
21   leakage.  We use both the lit and dark markets when we 
22   try to accumulate or lose shares. 
23             I think I understand the argument.  If you're 
24   going to limit it to one exchange, it's a little 
25   intellectual dishonest to say, you know, you can have 40 
0128
 1   ATSs trade a security, but I think it's overstated to a 
 2   degree.  So, we ran some data for 2017 and for year-to-
 3   date. 
 4             So, 40 percent of the small cap securities are 
 5   trading in, you know, off-exchange.  Only 10 percent or, 
 6   I guess, a quarter of that is really trading in ATSs, 
 7   right?  The rest are on TRFs.   
 8             So, that's either naturals meeting upstairs and 
 9   printing to the tape through a broker-dealer or it's 
10   single-dealer platforms, or as Joe mentioned, it's 
11   retail.  Retail is pretty big in these names. 
12             So, whatever the Commission decides to do, I 
13   think dark definitely needs to be a part of the 
14   conversation.  As I mentioned, it's a way that we manage 
15   information flow or information leakage. 
16             Part of what would be -- I think would appeal 
17   to  us about limiting it to one exchange or maybe one 
18   exchange from each family is the centralization of 
19   liquidity, right?  It would become a little bit easier, 
20   and maybe you get back to where having priority on the 
21   book matters again. 
22             There are a lot of unsophisticated holders of 
23   securities like this who don't have access to large 
24   broker-dealers, who are basically just generally going to 
25   the floor through smaller broker-dealers where if you 
0129
 1   developed a book again and priority mattered, I think 
 2   we'd -- probably from the asset management universe, we'd 
 3   have to start participating a little bit more. 
 4             MR. REDFEARN:  Owain, again, from your 
 5   perspective, as we think about this issue, what are your 
 6   thoughts? 
 7             MR. SELF:  I think I would echo some of Brian's 
 8   comments.  I think dark OTC trading is very important for 
 9   us, I think both from a perspective of looking for 
10   natural liquidity and being able to tailor the types of 
11   liquidity you interact with and control information 
12   leakage with regards to who you show your order to, and I 
13   think, you know, it's -- especially in the small and mid-
14   cap space, risk trading, capital commitment from the 
15   brokers is a significant portion, as well as upstairs 
16   trading. 
17             You know, in the data as presented earlier 
18   today that echoes exactly Brian's comments of less than 
19   10 percent of liquidity being in ATSs and 30 percent 
20   being reported,  and even in the ATS space or, 
21   necessarily, in the exchange space, I think what we look 
22   for in dark is a differentiated liquidity set to what we 
23   see in lit, and I think the thing that we don't see in 
24   lit is differentiation.   
25             Brad mentioned it already today, which is we 
0130
 1   have a lot of competing exchanges which don't necessarily 
 2   deviate from the models of others or necessarily compete 
 3   with each other, so we get more of the same, and 
 4   fragmentation with more of the same doesn't really 
 5   benefit, and I think, you know, the ATSs are guilty of 
 6   that, as well, and some of the brokers, because when 
 7   they're designed in the concept of cost avoidance, as 
 8   opposed to being looked at as a way to bring 
 9   differentiated liquidity to the market, then you get some 
10   more of the same outcomes, and I think, you know -- so, 
11   it's somewhat of a circular thing. 
12             You know, cost avoidance or, you know, looking 
13   at, you know, prices being more palatable on the 
14   exchanges, not necessarily having to fund the market, you 
15   know, indirectly, the market data fees of the exchanges 
16   by paying the rebate to the market maker, who then will 
17   get the correct share up so they can receive a better 
18   portion of the market data fee, you start to, you know, 
19   deal with those situations, and suddenly you may get to 
20   the crux of the problem quicker and try and get the 
21   natural liquidity to meet each other. 
22             So, I think OTC is hugely important.  I 
23   appreciate the comments with regards to centralizing one 
24   exchange might just be dealing with one part of the 
25   problem. 
0131
 1             I think from what Bryan from CBOE said this 
 2   morning with regards to looking at innovation within the 
 3   construct of Reg NMS may allow the free market to 
 4   determine centralization of liquidity, etcetera, which I 
 5   think is somewhat fair. 
 6             So, I think, you know, just to repeat the 
 7   answer once more, I think OTC is extremely important.  
 8   Certain types of OTC are obviously less, in the same way 
 9   as lit fragmentation.  Some is important, some less so. 
10             MR. REDFEARN:  Just before we go to the retail 
11   question, Joe, on the institutional side, I mean you both 
12   are obviously very large asset managers trading in size. 
13    Given that one of the potential thoughts here was, you 
14   know, potentially a block exception or a size exception 
15   or something like that, I don't know if that -- if you've 
16   thought through or have any thoughts about -- to the 
17   extent that, you know, that was sort of the related -- a 
18   related concept, whether or not that would be okay and 
19   not impede your upstairs trading or whether or not that 
20   would potentially be -- I mean, I am aware that, you 
21   know, 5,000 shares is, you know, 10 percent of a 50,000-
22   share-a-day stock, so it's a pretty big print, but I'm 
23   just wondering your thoughts. 
24             MR. FRAMBES:  I think we could get comfortable 
25   with the block exemption, depending on what it is.  
0132
 1   Obviously, you know, liquidity is always a function of 
 2   your demand.  So, as you get larger in size and you're 
 3   demeaning more liquidity, right, we'd want to see a 
 4   smaller block exemption, because we're picking up a lot 
 5   of shares, but I generally think that's something we 
 6   could get comfortable with. 
 7             MR. REDFEARN:  Owain? 
 8             MR. SELF:  I think, you know, obviously, with 
 9   the market structure, I mean we can get comfortable with 
10   it.   
11             I think the one thing I would say about block 
12   exemption or OTC trading is, as I said before, which is 
13   OTC tends to deal with how do I want to control the 
14   information leakage of my order and the person I'm 
15   exposing my order to. 
16             It doesn't necessarily equate to I'm happy to 
17   get my entire order done right here, right now.  
18   Controlling your order and slicing your order to smaller 
19   components can often deal with decreasing of market 
20   impact and risk and overall execution strategy.  That 
21   doesn't necessarily go hand in hand. 
22             MR. REDFEARN:  Understood. 
23             So, Joe, then, turning to the retail side, I 
24   don't know if you have any stats about what portion of 
25   this market is retail? 
0133
 1             MR. MECANE:  Yeah, I can share some stats.  You 
 2   know, we would, I guess, highlight one statistic, is last 
 3   year we delivered about $450 million of price improvement 
 4   to retail investors, and if you look at the 605 stats, 
 5   that is roughly proportionate in terms of the percentage 
 6   of time inside the spread between stocks over 100,000 and 
 7   under 100,000 ADV, just to use the line that you guys 
 8   drew, but if you look at the dollar terms, the average 
 9   price improvement per market order to retail investors 
10   was about $7.47 per market order, which is greater than 
11   the average commission amount from the online brokers and 
12   about $2 higher than the average amount from over 100,000 
13   ADV securities. 
14             So, the first point I would highlight is that 
15   the amount of price improvement that is being delivered 
16   to retail investors is material, and again, to my opening 
17   point, I think something that we need to make sure 
18   doesn't get impacted in whatever changes we're 
19   advocating. 
20             The other general points I would highlight is 
21   that, you know, we're speaking as if fragmentation in 
22   general is a problem to be solved or an issue that we 
23   need to address, which to me is at least a little bit 
24   counterintuitive to a lot of the market structure 
25   conversations we've had to date across the industry, and 
0134
 1   you know, just reflecting on some  of the data in the 
 2   memo, it seems like if fragmentation was the problem, we 
 3   would see the market acting in different  ways than 
 4   concentrating more activity at the primary markets or 
 5   looking for the opening and closing auctions to be a  
 6   place where more activity would gravitate, just trying to 
 7   find a more centralized place of trading, and it doesn't 
 8    seem like we're seeing the market acting in a way where 
 9   we are seeing evidence that we should have to force more 
10   consolidation. 
11             So, one point is just perhaps better defining 
12   what we think the benefits of -- or how the benefits of 
13   less fragmentation would manifest themselves, and the 
14   other point is that we seem to be starting with the 
15   premise of taking out a rule and seeing what impact it 
16   has, as opposed to defining what we think we need to do 
17   in order to increase liquidity, what we think the 
18   quantitative outcome of that change should be, and then 
19   why do we need to pull out a rule to help accomplish that 
20   goal? 
21             It feels like we're kind of starting with the 
22   last step and then working backwards, instead of starting 
23   with the first step and moving forward, at least in terms 
24   of defining what we think the issue is that we're trying 
25   to get to. 
0135
 1             MR. REDFEARN:  I mean, we did have -- I presume 
 2   you were here for the last panel -- 
 3             MR. MECANE:  Yeah. 
 4             MR. REDFEARN:  -- and so we did have a view 
 5   that -- you know, I always point out trade needs one 
 6   place, one time, one price, right, and it's hard to solve 
 7   for all that, but to the extent that you can simplify 
 8   that process, it may actually help, not necessarily in 
 9   increasing the liquidity of a name but potentially 
10   increasing the ability to find the other side in that 
11   name, and that was echoed by Jason Vedder on the panel, 
12   which is part of why we're -- 
13             MR. MECANE:  Yeah.  I guess what I would ask is 
14   -- or what I would look for is evidence that the two 
15   sides are there but not meeting up.   
16             Like, I think even Jason's comments were I go 
17   to a venue, I try to find the other side. If I can't find 
18   the other side, then I start to interact with the public 
19   markets. There you start to interact with different kinds 
20   of liquidity.  You have different strategies for trying 
21   to aggregate that liquidity. 
22             But I would just offer that I don't think we've 
23   heard or seen evidence of two parties not meeting up.  It 
24   feels like we have an issue of more of an investment 
25   problem, where one side is there, in an illiquid name, 
0136
 1   wanting to either get out of that name or establish a 
 2   position in that name, and there isn't that natural other 
 3   side. 
 4             And so, I would just offer that we should think 
 5   about whether we're trying to solve an investment problem 
 6   with a market structure answer, and you know, if the goal 
 7   is just to try to increase average daily volume in some 
 8   of these names, I think I'll echo what I think Owain said 
 9   at the beginning, which is we need to be careful that we 
10   don't just create the illusion of liquidity of, you know, 
11   just more volume happening for the sake of happening 
12   through market makers and not traditional investors 
13   getting together. 
14             You know, one rough stat that we pulled 
15   together is, if you increased the average 100,000-share 
16   ADV name's trading volume to a million shares, it would 
17   have twice the turnover of most other stocks, which 
18   probably isn't realistic and probably isn't the goal of 
19   what we're trying to accomplish here. 
20             So, I just think separating out if we have an 
21   investment problem or a market structure problem is maybe 
22   a way to frame the question. 
23             MR. REDFEARN:  It's a good point, and I think 
24   we did hear, at least from the majority of the last 
25   panel, that they seemed like something needs to be done, 
0137
 1   and part of the question is what, specifically. 
 2             Now, the investment, finding another side, may, 
 3   in fact, be there's two sides but just not at the same 
 4   time, and in that light, it was mentioned before some 
 5   contemplation of the possibility of auctions, periodic 
 6   auctions or something like that. 
 7             Bob, if there's anything you could share with 
 8   us about what you've looked at auctions, any other 
 9   markets, or you know, what's your fascination with this 
10   issue? 
11             MR. SCHWARTZ:  You would like a short answer to 
12   that one?  Oh, my goodness.  Listening to the first panel 
13   and the discussion on this one up to this point, it 
14   strikes me -- and it fits in with my view of the public 
15   discussion anyway -- that implicitly, when we talk about 
16   these things, we're talking in the context of continuous 
17   markets.  That's not the periodic call auction. 
18             Now, I have been in favor of both, absolutely, 
19   but I think that when you address some of these issues, 
20   to think of the extent to which they can be dealt with in 
21   the context of a periodic call auction should be part of 
22   the discussion, and I think it would be terribly 
23   important -- I think there's sort of a bias away from 
24   thinking that way, and in fact, a lack of understanding 
25   of what a call auction really does, how it operates, what 
0138
 1   its dynamics are, because we live in a continuous world. 
 2             It happens to open with a call and closes with 
 3   a call.  In Europe, they have midday calls, in addition, 
 4   but most of our activity, we're not thinking in that way. 
 5             Another thing that I believe that we don't pay 
 6   enough attention to -- and I've said this -- I already 
 7   said it in my opening remarks -- is lack of attention to 
 8   price discovery, because we can always look at, you know, 
 9   a trade that people participate in and what did it cost 
10   this guy, what did it cost this other person, but if you 
11   look at in a macro sense, trades result in prices.  They 
12   result in published prices.  They result in prices that 
13   are used for a variety of reasons, and the price 
14   discovery is terribly important.   I've been spending 
15   attention to it.   
16             People can say, well, how do you measure it?  
17   How do you measure the accuracy of price discovery?  I 
18   measure it in terms of volatility, the accentuation of 
19   intraday volatility that we see, and in fact, the 
20   volatility metric is the one that I have used first and 
21   foremost in looking at the impact of Nasdaq having 
22   introduced it. 
23             Its calls are called crosses, but they are 
24   calls.  They're call auctions. 
25             And it's the volatility.  And I say why is it 
0139
 1   good? Why is it a good impact?  Because I see in the 
 2   neighborhood of the call volatility comes out in the 
 3   first moments after the market has opened with a call. 
 4             I also see that, when it closes with a call, 
 5   volatility is dampened in the minutes before the call.  
 6   It's a very useful tool for that sort of thing. 
 7             I want to say that, when we get into 
 8   discussions of fragmentation and the like, of course 
 9   fragmentation -- the nice word for it is competition, 
10   competition between different market venues, and every 
11   economist is going to be in favor of competition, 
12   especially if you have tenure, but competition comes at a 
13   cost.  It really does, because you have competition 
14   amongst different venues, but there's also competition 
15   within the order flow, and if you want to augment the 
16   competition within the order flow, it calls for 
17   consolidation. 
18             It was mentioned in the previous panel, and I 
19   didn't entirely understand it, but I'll put it in my 
20   words. There's two ways to consolidate.  One is spatial, 
21   and we've been talking about that.  The other was 
22   mentioned, temporal, but I would have gone further, 
23   because that's what a call auction is doing.  It's 
24   pulling orders together. 
25             Now, these issues, which I have -- in my own 
0140
 1   work, I haven't really focused that much on the micro 
 2   guys, but this meeting, the topic here is going to affect 
 3   the direction I take.  I think it's terribly important.  
 4   I'm not sure how unique the problems are, but they sure 
 5   as hell are accentuated and should be dealt with. 
 6             Does that answer your question, Brad? 
 7             MR. REDFEARN:  That was great.  Thanks, Bob. 
 8             Go ahead, Stacey. 
 9             MS. CUNNINGHAM:  Just one thing on auctions.  
10   We did -- we heard from institutional investors a few 
11   years  back that, hey, it would be helpful to have more 
12   auctions, and we had such success in the opening and 
13   closing auctions, and could we do that more often 
14   throughout the day?  So, we did have plans to introduce 
15   midday auctions through the day. 
16             As we got closer to launch and started to talk 
17   more about getting usage in the auctions, we started to 
18   hear, well, I'm actually not sure that I'm going to want 
19   to use the midday auction, because it's going to have an 
20   impact on all the trading that's happening in other 
21   venues, and I don't want to end up having that 
22   information leakage impact trading elsewhere, and we saw 
23   that there was concern about using them in that -- and 
24   so, we actually haven't launched them for that reason. 
25             MR. REDFEARN:  So, it's a great point, and I 
0141
 1   remember when that dialogue was pretty robust and talked 
 2   to  a lot of asset managers at the time about that, in 
 3   particular, and heard those comments, and one of the 
 4   comments that was made was that, you know, obviously, 
 5   there's a question of, do you publish an imbalance, do 
 6   you not publish an imbalance?  
 7             If you do publish an imbalance, then that's 
 8   information that then participants in the market can 
 9   essentially be trading around, and there was a 
10   conversation that, if it was exclusively listed on one 
11   exchange, that it might actually work better, because it 
12   wouldn't be a scenario where you would have a bunch of 
13   other people who go trading in front of -- or trading 
14   with that order place separately from that. 
15             The other thing I remember hearing were asset 
16   managers saying if it trades that infrequently, then 
17   there's a compelling force to participate in that, 
18   because you don't want to miss that print, right? 
19             So, depending on how often it happens, whether 
20   it's just the middle of the day or whether it's other 
21   hours, if you have this order on your book and somehow or 
22   another, all of the sudden, boom, there goes, you know, 
23   10-20 percent of the day's volume happens in that print 
24   and you're not in it, then that's going to be a problem, 
25   and so, that itself also has sort of a compelling draw 
0142
 1   for participants being in there. 
 2             And I guess the third thing about it -- and 
 3   this is interesting vis a vis the over-the-counter, the 
 4   upstairs market, is, you know, if you -- if you had a 
 5   situation like that for illiquid names and there is some 
 6   question about what is the right price and how is price 
 7   formation happening and the price formation is happening 
 8   in the auction, again it raises another sort of 
 9   centrifugal force on that. 
10             MS. CUNNINGHAM:  I just think there -- you want 
11   to be part of that print, but you can do that today on 
12   auctions, and you don't need to have exclusive listing 
13   rights to be able to introduce something like that. 
14             A primary listing exchange can say and we could 
15   say we're going to run this at 12:00 o'clock during the 
16   day, but if there's still trading that's happening 
17   elsewhere, even if it were centralized, again, it comes 
18   to the over-the-counter versus just the primary exchange 
19   centralization, because you would still see that trading 
20   happen elsewhere, and they wouldn't want to be part of 
21   it. 
22             Nasdaq also had run auctions a few years prior 
23   that were blind, and they weren't displaying in balance 
24   information.  It was just to try and meet up for names in 
25   the middle of the day and, again, had a hard time 
0143
 1   soliciting interest, because people didn't know. 
 2             So, the argument to why Nasdaq was having 
 3   trouble was, because it wasn't advertised, people were 
 4   missing each other and they weren't in the same 
 5   securities at the same time. 
 6             So, when NYC proposed it, we were going to 
 7   advertise the auction, and then, we heard, oh, no, 
 8   there's going to be information leakage. 
 9             So, I think there's just a challenge, as long 
10   as trading is occurring on other venues over the counter 
11   or elsewhere, I think it's going to be hard to solicit 
12   real interest there. 
13             MR. REDFEARN:  Joe. 
14             MR. MECANE:  I just have a question, if I can 
15   ask one, and I'm curious -- it's just an academic 
16   question that maybe Bob has an answer to. 
17             If more auctions would help or if the market 
18   wanted to gravitate towards more auctions, given that the 
19   buy side is already so comfortable with the opening and 
20   closing auctions, why wouldn't you expect to see more 
21   volume gravitating to the auctions that already exist as 
22   being evidence that maybe we should have more auctions, 
23   because we could satisfy more of the demand. 
24             MR. SCHWARTZ:  Thank you, Joe and Stacey. 
25             I don't want to defend a midday auction.  In 
0144
 1   Europe, also, where they have it, it gets less volume.  
 2   It's the open and the close is, by far, the most 
 3   important, and my understanding is that it gets 
 4   substantial volume. 
 5             But I also want to point out that a call is not 
 6   a call is not a call.  It depends on how it's structured, 
 7   and I have issues with some of the structuring of calls. 
 8             But when we turn our attention to the 
 9   microcaps, it's not a question of how often you have a 
10   call, and if you have a call too often, it starts moving 
11   over to continuous.  I mean, you want to be consolidating 
12   orders over a longer period of time. 
13             But to have the structure for a continuous 
14   market of a limit order book, let's say, for stocks that 
15   trade very little -- and by the way, I don't think, 
16   myself, that average daily trading volume -- that's not 
17   how I would define liquidity. 
18             But I would say that that's a very -- let's get 
19   another name for it, then.  It's not exactly liquidity. 
20             But it's terribly important, because if it's 
21   substantial enough, the ecology of an order-driven market 
22   can break down. 
23             An order-driven market is driven first and 
24   foremost by limit orders, and if you don't have enough 
25   activity, it makes it impossible to get sufficient depth 
0145
 1   on your book, and you break down. 
 2             So, now when you're dealing with these 
 3   microcaps, what do you do?  I wouldn't depend on the 
 4   limit order books. I would depend on calls and I would 
 5   depend on dealer input or market maker or designated 
 6   market maker. 
 7             That's the way it's done in France with the 
 8   Nouveau Marche, which was very successful, open with a 
 9   call, close with a call, combined with dealer markets, 
10   and that's fine. 
11             While I've got my red light on here -- I don't 
12   know if I should do this.  Do I have the courage? 
13             I would suggest that a major input of liquidity 
14   into call auctions for your microcap stocks should be 
15   designated market makers. 
16             Calls have to be automated.  Calls don't just 
17   happen.  You need the structure, and a designated market 
18   maker would be a fantastic addition to that, as a 
19   formalized addition. 
20             On the other hand, when you're dealing with 
21   very thin markets, very volatile markets, to have market 
22   makers go in there and impact with their orders the 
23   continuous flow, they're wasting their fire power. 
24             I mean, why should I be buying at 32.10 as a 
25   market maker if the price is really going lower?  Let it 
0146
 1   go lower. Let us get the orders together, find 
 2   simultaneously what the best price is, that's the best 
 3   price for the market maker, as well, and by the way, if 
 4   you do it that way, the market maker get price 
 5   improvement.  In a continuous market, they don't. 
 6             So, I think a lot of these things do come 
 7   together, but I'm not a proponent, necessarily, of a new 
 8   call in a continuous market. 
 9             MR. REDFEARN:  I see Chris' light flashing over 
10   there. 
11             MR. CONCANNON:  I just have a thought.  I 
12   appreciate Bob's comments around the auctions, but as Joe 
13   pointed out, we already have exclusive auctions, right?  
14   The primary markets run an open and a close, and all the 
15   other markets don't UTP the open and the close. 
16             We allow them -- now, one market did have a 
17   call auction proposal.  It's held up right now, and Bob's 
18   support would be appreciated. 
19             So, look, the exclusive auction exists for 
20   these illiquid securities, and they are supported by DMMs 
21   like Ari, and there's a lot of support. 
22             I do think we're struggling with solving 
23   economics for liquidity providers through market 
24   structure means.  I don't know that creating an exclusive 
25   place to trade solves the institutional demand for 
0147
 1   liquidity, because it's bigger than some of the symbols 
 2   that we're talking about, and we've got this pretty good 
 3   solution for retail. 
 4             When I look at the over-the-counter and what it 
 5   solves for, we don't want to damage the retail solutions 
 6   that are out there, but we want to create some type of 
 7   place where liquidity can be sourced inhouse, and I just 
 8   think using market structure is difficult. 
 9             As I think about it, I look at the corporate  
10   broker concept in the U.K., and it's a pretty interesting 
11   concept, and by rule here, I think it's a little 
12   difficult  to do, but if Ari -- I'll use Ari as an 
13   example, mostly because his pink socks keep distracting 
14   me, but if Ari could be the exclusive underwriting of one 
15   of these small companies, the exclusive market maker in 
16   an exchange like  New York, can own the 10-B-18 or the 
17   buyback program, and do employee stock purchase program, 
18   if you have this exclusive broker concept, the economics 
19   that Ari can achieve make him incented to make a solid 
20   market and be a bigger market maker in that name and have 
21   all the information flow that the buy side will want to 
22   come to Ari, because he has natural liquidity.  He is an 
23   informed liquidity provider. 
24             So, there's other dynamics, not just micro-
25   market structure and market structure, that can help 
0148
 1   here.  The issuers can help, as well.  They're just going 
 2   public and then saying, well, you're not giving me a 
 3   market that solves for my liquidity and my liquidity 
 4   demands, but an issuer can actually, in Europe, can 
 5   choose a corporate broker and help find a market maker 
 6   that will support their stock through the economics that 
 7   the issuer has to exchange, not just the exchange trying 
 8   to create economics through micro-market structure and 
 9   other means. 
10             MR. REDFEARN:  So, there's a lot here to digest 
11   as we go, but two themes that have come up that I wanted 
12   to just explore a little more deeply -- one is the issue 
13   that was raised earlier about the possibility that we're 
14   talking about a space were nonautomated quotes might make 
15   sense, and the second one is this issuer of market maker 
16   models and how the market maker model could, indeed, 
17   evolve ways that might be beneficial to these names. 
18             Tal, maybe turning to you for a second, just on 
19   either of those points, when you think about the need for 
20   an automated quote or the role of the market maker, you 
21   know, how robust that market maker model is, do you have 
22   any thoughts on how we may be able to evolve either of 
23   those threads for this discussion. 
24             MR. COHEN:  Before I do that, I'm just going to 
25   take a step back, and there's really three constituents  
0149
 1   we're trying to serve with the discussion today, and 
 2   we're talking mostly about one or two of them, but the 
 3   issuer side of the equation is a narrative, too, and so, 
 4   just -- it's important to focus on the engagement with 
 5   the issuer, the narrative around our markets, and the 
 6   public versus private debate. 
 7             So, we're having the OTC versus the on-exchange 
 8   debate, but there is also a private versus public market 
 9   debate that is being had at a higher level. 
10             In terms of some of the solutions that have 
11   been proposed and we're thinking about, we need to 
12   through and ask ourselves what is the cost-benefit. 
13             So, I think Joe talked about identifying the 
14   issue and making sure the cost-benefit is there, and when 
15   we look at today's markets, when we think about 
16   centralizing liquidity, we think about the diminishing 
17   returns of innovation and we think about the diminishing 
18   returns of NMS, as you go down the spectrum, because once 
19   again, during times of duress, is your market maker going 
20   to step up in a Reg NMS environment? 
21             Is he incented to step up because of the 
22   complexity, because of the cost of doing so?  Are 
23   institutions willing to show their hand because of how 
24   the market works today? 
25             So, the question then comes back to slow versus 
0150
 1   fast markets, and what do slow markets actually achieve? 
 2    What do they allow us to do?  They allow, potentially, 
 3   financial intermediaries to come into the market and take 
 4   on more obligation, to the extent that they feel like 
 5   there's more protection and less adverse selection. 
 6             So, in today's auto-ex world, where if you put 
 7   out your hand -- and I think Jason referred to it before 
 8   -- he would never display 100 shares today, because he 
 9   might get auto-ex'd right away and somebody might see 
10   that or somebody might just simply front-run you, but the 
11   concept, once again, with slow markets is, can you 
12   provide a market model, if you will, that would allow 
13   market makers to feel more comfortable with obligation, 
14   not just during every trading day but during times of 
15   duress. 
16             Do they understand how the markets work during 
17   times of duress?  Is there too much complexity, so they 
18   need to pull out of those markets?  Can they do that 
19   effectively? 
20             And so, today, we see a lot of rhetoric around 
21   people pulling out of markets really quickly, and that's 
22   just a function of them reacting to how the market works, 
23   because if they don't pull out of the markets really 
24   quickly today, they become fairly exposed during those 
25   times of volatility. 
0151
 1             So, that would be one. 
 2             In terms of then providing market making 
 3   incentives, Chris talked a little bit about the 
 4   incentives we could provide, but we need to be careful to 
 5   balance that with competition, so we don't want to put to 
 6   much hands in the power of one individual, even if he is 
 7   wearing pink socks, and so, we want to be mindful of 
 8   competition.  We want to be mindful of putting enough 
 9   economics out there so people can do what they need to do 
10   but at the same time not have the entire name trade 
11   through one or two broker-dealers, because I think that 
12   would probably put us in a position where there is 
13   asymmetric information in the marketplace again, and that 
14   was, you know, the yesteryear of some markets where I 
15   don't know that we want to go back to that model. 
16             So, we think slow reduces risk.  It builds a 
17   narrative.  It talks about tailoring the market 
18   structure. 
19             So, for example, once again, this LULD and 
20   market-wide circuit breakers and things of that nature -- 
21   do they work as well as they could for small cap names or 
22   thinly traded names?  Probably not.  We can tailor around 
23   that. 
24             We can talk to the constituency on the buy side 
25   around what are they trying to achieve?   
0152
 1             I don't think restricting choice -- some of the 
 2   conversation is about potentially restricting choice, but 
 3   it is a balancing act between, on the lit side of the 
 4   equation, trying to centralize how markets work and 
 5   removing complexity but preserving some choice, because 
 6   100 shares, 200 shares may actually be a meaningful order 
 7   for a thinly traded name, and that's what you were 
 8   getting to, Brad, as well, and we're not in a position, 
 9   once again, to see the parent order. 
10             So, when we think about it, there is a reason 
11   for us to kind of move away from Reg NMS.  There is a 
12   reason for us to embrace slow versus fast, and there's a 
13   reason for us to think about market makers and the 
14   obligations we'd like them to take on and the economics 
15   we provide, but we need to preserve choice along the way. 
16             MR. REDFEARN:  Owain, when you think about 
17   trading in these names, if there was a model where they 
18   were nonautomated quotes or that there was a sort of a 
19   different market maker program in place, do you have any 
20   thoughts about what you would like to see or whether 
21   there's anything that worries you about moving in that 
22   direction for your liquidity? 
23             MR. SELF:  I think, you know, if you're looking 
24   at nonautomated quotes, I think, generally, that would 
25   push more volume upstairs for us, assuming it would come 
0153
 1   with the -- it would probably reach -- necessary trading 
 2   electronically in that space, I think. 
 3             Generally, when you're dealing electronically 
 4   on an exchange, the less someone knows about my order, 
 5   the better, and that includes any obligated market maker 
 6   on a venue, you know, specifically in and around 
 7   auctions. 
 8             So, you know, I think, from that perspective, I 
 9   find, once you -- if you're going to be visible to 
10   someone, it generally tends to be more of a two-way -- I 
11   like the person to know that, you know, we're having a 
12   conversation, what we expose to them and what they do 
13   with that information, rather than doing it in a more 
14   one-way mechanism. 
15             So, I think, really, to have -- I can see the 
16   benefit for those people that want to trade on the screen 
17   and to have some stability in the stock, etcetera.  I 
18   mean, Chris mentioned the comments around some of the 
19   corporate broking and market making, which tends to be 
20   in, obviously, the much lower tier and tends to be in a 
21   much, you know, slower moving stock. 
22             So, I think, from our perspective, I can see 
23   the attraction for some market participants, but I would 
24   probably guess that the experience would move our 
25   business upstairs more than anything else. 
0154
 1             MR. REDFEARN:  Brian, do you have a similar 
 2   reaction to that? 
 3             MR. FRAMBES:  Yeah, I would echo Owain's 
 4   comments. In a sense, we already have like nonautomated 
 5   quotes.  Broker-dealers supply them for us already, 
 6   upstairs.  I think it would be -- it would make -- 
 7   nonautomated quotes would make the workflow issue for 
 8   small caps become even more intense than it is now.  We 
 9   already spend probably a disproportionate amount of time 
10   on small cap names than we do on large cap names, just on 
11   how to trade them and find liquidity. 
12             Automated quotes and the whole 
13   electronification of the market has really helped us 
14   scale our business and kept costs down. 
15             So, I would agree that, to a degree, we already 
16   have the nonautomated quote market upstairs, and if you 
17   went that way, I think it would even gravitate further. 
18             MR. REDFEARN:  Any other thoughts on that one? 
19             MR. MECANE:  I'll just add quickly, you know, 
20   we believe strongly in the idea of market maker benefits 
21   for obligations.  It's a big part of our model.  We've 
22   partnered with NYC on a lot of those markets. 
23             But I would just echo the point that we made 
24   earlier that it's definitely helpful, it's a significant 
25   improvement to market quality, but I don't think it's 
0155
 1   going to solve the goals that we're trying to lay out 
 2   here.   
 3             We'll see in a lot of the thinner names where  
 4   we'll be sometimes 30 or 40 percent of the volume that 
 5   happens in a lot of those names, and I think it helps 
 6   investors and it helps market quality in those names, but 
 7   I don't think it's enough without the fundamental 
 8   investment coming in behind it and the interest from more 
 9   buy side investors if, you know, we're trying to get to 
10   the market quality statistics that we're talking about. 
11             MR. REDFEARN:  Is there any thought that's been 
12   given to sort of enhancing the market maker obligations 
13   in the sense of, if there were -- you know, if there was 
14   a single market maker or a greater advantage that came 
15   with centralizing liquidity in some way, any thoughts 
16   about what might be done to sort of make that a little 
17   more robust than it is today? 
18             MS. CUNNINGHAM:  I would sort of add on to what 
19   Joe is saying, and I agree with him.  We do have the most 
20   stringent obligations on NYC out of any other market 
21   maker that's out there, and we do see a better result and 
22   better market quality in the last active names, 
23   significantly, and it is a selling point for our issuers 
24   when they're choosing a listing venue.  That's one of the 
25   things that they look at, and it's important to them. 
0156
 1             But it's not going to go as far as to say, hey, 
 2   somebody needs to find the natural investor on the other 
 3   side, and because this market maker is there, they're 
 4   going to suddenly be able to find that -- source that 
 5   liquidity that -- that, you know, Brian and Owain are 
 6   looking for. 
 7             What they really are doing is dampening 
 8   volatility intraday and kind of offsetting those 
 9   temporary disparities, and I think that there are things 
10   that we can explore as a listing exchange to continue to 
11   finetune those obligations, but I don't think it would be 
12   genuine to say suddenly you're going to find the other of 
13   a trade. 
14             MR. REDFEARN:  So, just to sort of continue to 
15   push this question on, you know, what would this no-UTP 
16   world look like -- so, what Tal and Nasdaq are 
17   recommending -- if that were the case, there is this 
18   question about, you know, cost, monopoly cost.  Chris 
19   mentioned market data costs would go up.  Other people 
20   might think other fees might go up, because you have this 
21   centralization. 
22             Would that kind of environment the necessitate 
23   something to be done to ensure that that doesn't take 
24   place, and if so, what? 
25             MR. CONCANNON:  I appreciate all of Tal's 
0157
 1   commitments to not charge more from Nasdaq.  We've heard 
 2   that for a while now. 
 3             It's just hard to -- it's hard for the 
 4   Commission to restrict price.  It's hard for a market to 
 5   make a commitment that we won't charge for extra ports if 
 6   I'm an exclusive listing or our market data won't go up. 
 7    I just think it's hard. 
 8             The benefits that you reward for giving an 
 9   exclusive listing to a market are quite sizeable, and 
10   it's hard to regulate away those benefits when it comes 
11   to fees and all the different means that we have as 
12   exchanges. 
13             I could charge you a higher membership fee.  I 
14   could charge port fees that are slightly higher than 
15   market. 
16             So, when I think about the exclusive listing, I 
17   like the concept.  I just worry about it's not going to 
18   deliver the benefits that we'd like, yet it brings with 
19   it a great deal of detriment over time, and maybe for the 
20   first year there's no fee hikes, but in three years, the 
21   market data fees could be something different. 
22             So, I worry about the exclusive listing.  I 
23   start from the premise that we should do something, that 
24   we have a rule called Reg NMS that, regardless of the 
25   size of your symbol and how you trade, notionally or by 
0158
 1   share, it's the same rule, and I think we were flawed 
 2   back in '05 when we thought about market structure in 
 3   that sense, that we were taking a market that did have 
 4   tiered and different structures, and we just put one 
 5   structure on it. 
 6             So, I would start with Reg NMS and looking at 
 7   ways to unwind Reg NMS in the less liquid names before I 
 8   come to the conclusion that granting an exclusive listing 
 9   is the right way to go. 
10             MR. REDFEARN:  I'm going to invite feedback on 
11   that. 
12             Joe? 
13             MR. MECANE:  This is maybe tangentially 
14   related, so I didn't want to change topics, but on the 
15   topic of fees -- and this fits into the market maker 
16   point, too, and I don't have a strong opinion on this, 
17   but I thought it was worth mentioning that obviously the 
18   Commission has the transaction fee pilot out for comment, 
19   and one of the questions that I know has come up -- I 
20   think it was in one of the Q&As -- and was discussed 
21   certainly after the MSAC proposal went in, is the idea of 
22   whether the cap on less liquid securities should, in 
23   fact, be increased or even, to Chris' point, taken away, 
24   as a way to try to improve liquidity in some of those 
25   names, and I don't have a strong opinion on that. 
0159
 1             I don't know if the exchanges do, but I would 
 2   just highlight that.  If we're talking about fees in this 
 3   type of environment and we're talking about incentivizing 
 4   market makers, that we should at least have that be a 
 5   more holistic conversation and not deal with those two 
 6   things piecemeal. 
 7             MR. REDFEARN:  Do you think that the rebates 
 8   are meaningful to market makers for their quoting in 
 9   these  names? 
10             MR. MECANE:  The point has been made that 
11   perhaps you can't pay enough to try to incentivize the 
12   type of activity that people are looking for, so it's 
13   unclear.  I think it helps. 
14             Again, I would go back to what I think the 
15   theme of some of the conversations has been, in that 
16   market structure and incentivizing market makers is 
17   helpful, but what it feels like we're really looking for 
18   is a change to the investment side of things, so I'm not 
19   sure it would dramatically change things, but I think 
20   anything along those lines marginally changes people's 
21   opportunity cost. 
22             MS. CUNNINGHAM:  I just would say that I do 
23   think the rebates and the incentives that we place in 
24   those less liquid names, where we do invest even more, an 
25   outsized amount, in our DMMS there, do lead to better 
0160
 1   prices for issuers and investors, not just in the trading 
 2   on the exchange, but in the trading off exchanges, as 
 3   well. 
 4             So, we hear -- you know, we heard from many 
 5   today that there is a lot of value in trading in the 
 6   dark, but those dark prices are being set by what the 
 7   displayed quote is, and so, if we're tightening up that 
 8   displayed quote, even if the designated market maker 
 9   doesn't become, you know, the liquidity provider, de 
10   facto, you're still leading to better investor prices for 
11   issuers. 
12             So, we would be concerned about viewing these 
13   things as two separate issues and having a transaction 
14   fee pilot that then impacts these less liquid names in an 
15   outsized way. 
16             MR. KATSUYAMA:  I think it's important to 
17   remind people the rebate is one-third of a cent, and so, 
18   at most, getting rid of the rebate would widen the spread 
19   by two-thirds of a cent. 
20             So, in terms of the notion of spreads blowing 
21   out -- I just think, mathematically, again, back to Joe's 
22   point, how much do you have to pay someone to stand in, 
23   you know, when all signals are pointing to you doing the 
24   right thing and getting out of their way? 
25             I think, back to obligations, Chris mentioned 
0161
 1   this in his point -- I think one of the pieces that is of 
 2   growing concern is kind of who are the market makers that 
 3   are left, right? 
 4             So, you have a lot of large banks who have 
 5   pulled away from those roles of making markets.  This is 
 6   a market structure that has not evolved to kind of, I 
 7   guess, give them some type of competitive edge.  It's a 
 8   speed-based market.  They get crowded out at the inside. 
 9             You know, even looking at Goldman withdrawing 
10   from being a market maker in ETFs and ETP products, you 
11   know, you do look at -- for ways to get the banks who 
12   have a relationship up and down the stack with some of 
13   these small caps, are the ones bringing them public -- 
14   what are the incentives to get them back into the market, 
15   and I think that's an interesting question.  It's one 
16   that we've talked to a number of them about. 
17             But again, the rebate -- tying the rebate, you 
18   know, to -- yes, it's fine if the market's not moving, 
19   etcetera.  Markets will widen, at most, by two-thirds of 
20   a cent, and when you're talking about already fairly wide 
21   spread in some of these small cap names, I just don't 
22   think those points correlate. 
23             MR. CONCANNON:  If I could just add -- I didn't 
24   know we were going to debate the access fee pilot here 
25   today. 
0162
 1             MR. REDFEARN:  I didn't either. 
 2             MR. CONCANNON:  I was going to actually agree 
 3   with Stacey that the rebate -- the economics that we 
 4   reward our market makers does matter.  We have these 
 5   conversations -- I know New York and Nasdaq have these 
 6   conversations, because we have market makers. 
 7             In ETP -- and I know that's the next panel, 
 8   it's material in the ETP world, and having a conversation 
 9   where market makers are calculating what their cost of 
10   having a quote in a listed ETF is and how they can get a 
11   remuneration to offset that cost, they can actually 
12   calculate the cost of trading an ETF.  Whether it's 
13   liquid or not, they can calculate that. 
14             The rebate does matter in certain symbols, and 
15   I know the rebate matters in less liquid names, as well. 
16    Is it all that material to the cost of taking on a big 
17   position that a buy side trader might leave them with in 
18   a market move?  Probably not.  It won't offset that.  But 
19   it does impact how they quote and how they think about 
20   the market. 
21             And to Stacey's point, it's a largely retail-
22   driven market, and the spread that the market maker is 
23   delivering on an exchange somewhere is impacting the 
24   effective spread that retail is getting somewhere in our 
25   market. 
0163
 1             MR. REDFEARN:  Owain. 
 2             MR. SELF:  I'm not going to turn this into an 
 3   access fee debate. 
 4             MR. REDFEARN:  Thank you. 
 5             MR. SELF:  Brad said it before I could, which 
 6   is the rebate is just a below-the-line transfer of cost 
 7   between the taker and the maker.  It's not the exchange 
 8   providing the rebate.  It's the taker providing the 
 9   rebate. 
10             So, if spreads were to naturalize to the right 
11   level, I can't see it being a significant difference in 
12   the -- to the pure level which is actually being paid.  I 
13   can't see it being a significant difference in small caps 
14   or some of the less liquid names, because I think people 
15   should want a market maker if they're incentivized by the 
16   spread capture, if it warrants the risk that they're 
17   willing to take by posting prices and not necessarily to 
18   have some secondary fee. 
19             The other thing I wanted to comment on -- a 
20   couple of other things -- you know, your original 
21   question about feedback, to your point, was -- look, I 
22   echo -- you know, I'm not trying to, you know -- I'm 
23   trying to help build solutions but not just keep 
24   reiterating problems, but you know, what we have seen 
25   with -- where exchanges have a monopolistic area on flow, 
0164
 1   they do increase fees. 
 2             Auctions are a prime example.  Where they 
 3   compete, they reduce the fee.  They subsidize it by 
 4   increasing the fee in areas where they don't have to 
 5   complete, and I can't arguably say that that won't happen 
 6   in an environment where someone has a -- you know, a 
 7   standalone place where they don't have to compete for 
 8   business. 
 9             The second thing I would say about auctions, 
10   just as a comment, is the auctions in the U.S. do not 
11   have anywhere near the amount of liquidity as probably 
12   the rest of the world in terms of the percentage of the 
13   day's volume, and we -- you know, from my personal 
14   experience in trading, it's not a great place to try to 
15   transact because of the lack of price discovery, 
16   transparency, due to the cancellation rules, and I see, 
17   you know, if there were -- and part of that reason is 
18   because they're in operation during continuous market, to 
19   echo the point that was made earlier, where if you want 
20   full transparency and price discovery and you're showing 
21   people how to trade, if the market's open at the same 
22   time, you're showing your hand. 
23             And you know, historically, my perception of 
24   why the cancellation rule exists is because people were 
25   accumulating -- people in privileged positions were 
0165
 1   accumulating liquidity to satisfy the auction, and when 
 2   people cancelled their orders, they were left, you know, 
 3   holding the shares. 
 4             So, I think there's some evolution there in 
 5   terms of the lower liquidity spectrums, where if our 
 6   auctions in these markets were better, potentially, they 
 7   could attract more liquidity for the stocks, and not 
 8   necessarily need to be intraday.  Sorry, that was on the 
 9   previous point, as well. 
10             MR. REDFEARN:  No, that's great.  We can debate 
11   auctions and we can debate -- 
12             MR. FRAMBES:  I won't bring up the access fee 
13   pilot, but just to address your original question, we'd 
14   be supportive of taking a look at a holistic review of 
15   market structure in NMS, so we should take a look at 
16   protected quotes, 611 access fee caps, all of those 
17   things. 
18             And to bring it back around to why we're here, 
19   I don't know if rebates make a difference in small cap 
20   securities.  I'm sure they do to some degree.  But we're 
21   not looking to transact with a market maker, and I can 
22   sure as hell guess they're not looking to transact with 
23   us.  It's always adverse selection for them. 
24             But the idea of centralizing liquidity in one 
25   or a couple of areas just makes some intuitive sense, 
0166
 1   right?  There are two outlets for us, dark and lit.  So, 
 2   dark, you know, we have that covered, but lit, we've kind 
 3   of conceded the quote to market makers over time. 
 4             I think if it was centralized and there was an 
 5   opportunity to get priority in the book, we would have to 
 6   contribute, at least represent a portion of an order.  
 7   Will we ever show, you know, outsized size on the bid or 
 8   offer?  I doubt it, but we will probably be there for 
 9   something. 
10             MR. REDFEARN:  I like the priority point, 
11   because with the markets as dispersed as they are, there 
12   is really no price-time priority in the books that we 
13   have today, so that's an interesting point. 
14             Joe? 
15             MR. MECANE:  Can I make one outside-the-box 
16   suggestion just around market structure? 
17             MR. REDFEARN:  Why stop?  Why stop? 
18             MR. MECANE:  So, one other thought that we had 
19   which we'll raise for consideration is -- and I don't 
20   know if Brian or Owain has run into this, but one thing 
21   to potentially look at is the Section 16 disclosure 
22   requirements for these types of names, and you know, we 
23   were talking, I think, in the earlier panel even more, 
24   about information leakage and the concern of when someone 
25   knows what your -- what you're trading or what your 
0167
 1   position is. 
 2             One thing to maybe think about is whether the 
 3   10-percent limit in the types of names that were -- or 
 4   the 10-percent disclosure requirement in the types of 
 5   names that we're talking about maybe over-signals to the 
 6   market, and perhaps if you're a passive investor, a 
 7   higher limit should apply, and again, going back to maybe 
 8   what could encourage more fundamental investment, that's 
 9   something that we think could have an impact on people's 
10   willingness to invest in the types of names we're talking 
11   about. 
12             MR. SELF:  I can say one thing which won't 
13   necessarily address that, which was the thing about the 
14   tiered structuring of all the different market segments 
15   and the rules that surround it is one thing I was just 
16   mentioning before this roundtable, is everybody talks 
17   about the liquidity problem in these less liquid 
18   instruments, by definition, and you know, what goes 
19   around the small caps, and when you look at some of the 
20   metrics, they look similar to large caps. 
21             They look similar in terms of some of them have 
22   the same percentage of turnover of the shares 
23   outstanding, etcetera. 
24             You quickly encounter the liquidity problem in 
25   a small cap than you do in a large cap, because you 
0168
 1   encounter the portfolio concentration problem in a large 
 2   cap.   
 3             You don't want to take hundreds and hundreds of 
 4   billions of dollars of position in a large cap before you 
 5   get anywhere near a percentage of the shares outstanding, 
 6   but you're very comfortable taking a $15 million position 
 7   in a small cap name, and then, suddenly, to Joe's point, 
 8   you're -- you're telling everybody what you have. 
 9             You have a massive liquidity problem, because 
10   maybe it only turns 2 percent of turnover a day in shares 
11   outstanding. 
12             So, the problems that we incur are not 
13   necessarily because there is no liquidity in these names, 
14   etcetera.  It's because the positions that people are 
15   willing to take are outsized in some of those metrics -- 
16   shares outstanding, average daily volume, and other 
17   areas. 
18             So, not just looking at this in terms of market 
19   structure, I think, is a good point that Joe brought up, 
20   is there may be other areas which can help aid in the 
21   top-down investment process around these stocks. 
22             MR. REDFEARN:  So, we have a few more minutes. 
23    I'm just going to ask and maybe just see if anybody 
24   wants any final closing salvos before we wrap up, maybe 
25   starting with you, Tal.  Any final thoughts on -- or 
0169
 1   maybe I'll ask it this way. 
 2             When we walk away from this and there's 
 3   specific things that we want to drill down on more deeply 
 4   to potentially follow up on here, what would your, you 
 5   know, top thoughts be on what we should be looking at? 
 6             MR. COHEN:  Don't forget about the issuers.  
 7   So, what we've represented here on the panel today is a 
 8   lot of discussion about micro-market structure and 
 9   potential solutions that either provide incentives in the 
10   marketplace -- and part of that is the rebate part 
11   conversation and part of that is how do we reduce 
12   complexity and risk so people want to make that 
13   investment? 
14             So, we need to think about all three 
15   constituencies in terms of the issuers, and we were 
16   talking about granting exclusivity, but what, really, we 
17   were talking about is issuer choice, and this is part of 
18   a broader recommendation where we talked a little bit 
19   about regulatory reform, but there's a lot to be 
20   discussed around disclosure and proxy and litigation that 
21   didn't get discussed on the panel today, that could 
22   probably move the needle, and so, I think that needs to 
23   be part of the discussion. 
24             We need to think about it in a combined way, 
25   and we need to think about all three constituents, which 
0170
 1   is issuers, investors, and financial intermediaries, and 
 2   you have to figure out how you're going to balance 
 3   removing complexity and providing the right level of 
 4   incentive, so market makers can come in and get paid for 
 5   coming in, but yet provide what you're looking for in 
 6   terms of quote competition and spreads, but that's got to 
 7   be meaningful in the context of institutional investors, 
 8   so when a portfolio manager says do we buy this, that the 
 9   entire alpha or idea isn't eroded by implementation 
10   shortfall. 
11             And so, I think, thinking through all three 
12   segments more holistically would be helpful, and I think 
13   -- don't let the concern of how do we manage conflicts or 
14   how do we manage certain concerns in the marketplace stop 
15   us from trying to solve this problem. 
16             MR. CONCANNON:  I think Tal used all our time. 
17             MR. REDFEARN:  The clock is 4 minutes fast. 
18             MR. CONCANNON:  I do like the -- I like the 
19   exclusive listing concept, but more because I like 
20   treating stocks of different shapes and sizes with 
21   different market structure. 
22             I do think there are things within our current 
23   market structure that we could look closely at before we 
24   use what I call a very blunt instrument called exclusive 
25   listings. 
0171
 1             MS. CUNNINGHAM:  I want to echo some of what 
 2   Tal said, just because I do think it's so important to 
 3   keep the issuer voice part of this process, and what 
 4   we're hearing from them -- we talk to them all the time, 
 5   and they don't like being put into a pilot that they 
 6   don't have control over. 
 7             So, I would urge the Commission not to do this 
 8   as a pilot, because they don't want to be structured 
 9   there, but as we are having the conversations about 
10   public versus private markets, I think it's important 
11   that we continue to give issuers the ability to choose 
12   what's right for them. 
13             We've heard so much about the retail presence 
14   in these thinly traded securities.  We don't want to push 
15   an issuer to access the private markets and leave retail 
16   out of the game because there's a model that works better 
17   for them there. 
18             So, giving them a choice to centralize 
19   liquidity, if we are truly talking about centralizing 
20   liquidity, then that applies to not just restricting UTP, 
21   which only applies to changes, but over-the-counter 
22   trading. 
23             If we find a mechanism that an issuer can 
24   choose to be in there that centralizes liquidity, that's 
25   worth looking at, and I just would say that, as an 
0172
 1   operator of three listing venues, it would be good for us 
 2   to say let's get rid of UTP, but we don't want to do that 
 3   if we don't think it's the right thing for our issuers. 
 4             MR. FRAMBES:  I would just reiterate what I 
 5   said in my opening statement.  The reason small caps are 
 6   illiquid securities or hard to trade is because they're 
 7   usually small and it's a scarce resource. 
 8             To the degree we can find a model where we can 
 9   have investors meet each other in one or two places, a 
10   centralized lit place, and then we still have dark 
11   available to us, I think is very important. 
12             You know, again, to reiterate, dark is just 
13   very important.  Exchanges obviously have a fair access 
14   rule.  We understand that.  We agree with it. 
15             And then, you know, bringing it back even a 
16   step to -- I think a lot of people mentioned here taking 
17   a look at Reg NMS and how it applies to small caps may 
18   make a lot of sense. 
19             MR. REDFEARN:  Brad? 
20             MR. KATSUYAMA:  I think issuer choice is a 
21   really important theme.  I think issuers making an 
22   informed choice, I think, is most important, which is why 
23   I think they need to understand the benefits and the 
24   effects of concentrating liquidity on a smaller number of 
25   venues or a single one. 
0173
 1             I think giving them that choice and being 
 2   counseled by the exchange that would benefit the most 
 3   from such a monopoly, you know, might not be the best 
 4   step.   
 5             So, I think, you know, getting information in 
 6   the hands of these companies, perhaps you make the pilot 
 7   opt-in as one example of not forcing them to do 
 8   something, but at the end of the day, I think that the 
 9   issuers need information to make a choice that will 
10   impact how their stock trades, and I just don't think 
11   that exists right now.  This is a great opportunity to 
12   get it. 
13             MR. REDFEARN:  Thank you, Brad. 
14             Joe? 
15             MR. MECANE:  Two quick things.  I'm going to 
16   slightly disagree with the theme of choice, and I 
17   understand why there's an incentive to let issuers have 
18   part of the decision-making process, and we speak to a 
19   lot of issuers, but I would offer that if we -- given 
20   that we don't know what the right answer is and we don't 
21   know what the outcome is and we don't know what the data 
22   looks like, if we -- whatever we put in front of an 
23   issuer I think would just result in sort of a random 
24   decision getting made, and I would offer that that would 
25   probably not give us really good evidence to make any 
0174
 1   conclusions around market structure. 
 2             The second thing I would say is, you know, 
 3   people use the word "pilot" in different ways.  I would 
 4   just offer that, if we were going to go down this path 
 5   and take out a fundamental rule that's been part of the 
 6   market structure since 1975, that we'd want to do it in a 
 7   way where, if we saw unintended consequences or negative 
 8   outcomes, we'd want the ability to revoke the change 
 9   back. 
10             So, whether we call it a pilot or not, I just 
11   think not having that kind of mechanism would be the 
12   wrong way to proceed. 
13             MR. REDFEARN:  Thank you. 
14             Bob? 
15             MR. SCHWARTZ:  UTP is a very complicated and 
16   interesting issue that I think, in part, it's substantive 
17   and, in part, it's a means of getting somewhere.  I would 
18   suggest that the primary question is what should the 
19   structure be for these small microcaps.  That's what this 
20   is about.  And when you think of the small microcaps, we 
21   can learn something from what we know about markets, but 
22   they are a special challenge.  The limit order book is 
23   going to break down.  Ecologically, it's not going to 
24   work.  I think what you need is a proper combination 
25   between -- and I'll say it again -- call auctions and 
0175
 1   dealers. 
 2             MR. REDFEARN:  Thank you, Bob. 
 3             Owain? 
 4             MR. SELF:  I think I would echo the comments, 
 5   which is that I think a multi-tiered market structure 
 6   approach, the different segments of the market, makes 
 7   sense.  Whether that needs to be done, you know, by, you 
 8   know, removing the UTP, I think is a bit of an extreme.   
 9             However, I am sympathetic that the ability to 
10   innovate at certain market structures while in 
11   competition with other exchanges or venues, sorry, which 
12   may not have the same objectives may lead a race to zero 
13   and constantly moving all the way back to continuous. 
14             So, I think it's problematic and, as I said, 
15   not trying to help bring the solution, but ultimately, I 
16   think there is a balance between the flexibility in the 
17   regulation without necessarily forcing it to go to one 
18   venue. 
19             MR. REDFEARN:  Okay. 
20             Well, we are now -- not according to that clock 
21   but the other clock -- about 3 minutes over, so thank you 
22   all very much for this very engaging discussion.  We very 
23   much appreciate your thoughts. 
24             Again, I just want to remind those folks who 
25   are either on webcast or in the audience that we do have 
0176
 1   a website where comments are solicited, so for any 
 2   comments or issues people have, please feel free to 
 3   submit them to us.  So, thank you all very much. 
 4             (Whereupon, at 1:33 p.m., a luncheon recess was 
 5   taken.) 
 6       A F T E R N O O N  S E S S I O N 
 7             MR. REDFEARN:  Thank you all very much for 
 8   being here, and welcome.   
 9             This is about to start our last panel today, 
10   which is focusing on thinly traded exchange traded 
11   products, specifically, in contrast to what we've been 
12   discussing this morning so far, which has been mostly on 
13   corporate stocks. 
14             I think the best way to start today would be 
15   simply to let each of you introduce yourselves, and 
16   please feel free to say what you do, how you're 
17   interacting with the market, and if you have any high-
18   level thoughts for us to think about, take it away. 
19             So, Josh, why don't we start with you? 
20             MR. KULKIN:  Hi, everybody.  I'm Josh Kulkin.  
21   I'm head of trading at Jane Street Capital.  I want to 
22   thank the Commission for having us here today to talk 
23   about this important issue. 
24             Jane Street is a global liquidity provider and 
25   a market maker, and we actually price just about 4,000 
0177
 1   ETFs on a constant basis.  We're a registered market 
 2   maker and liquidity provider in essentially all the 
 3   products we're talking about today.  We trade about 30 
 4   percent, by dollar, of the ETPs that are under 100,000 
 5   shares a day of average trading volume. 
 6             The most noteworthy thing, in my opinion, about 
 7   illiquid ETPs is that, counterintuitively, the spread 
 8   gets much tighter, the larger the trade is.  This is due 
 9   to creation/redemption sizes and minimum 
10   creation/redemption sizes, which seems like something we 
11   should certainly talk about today. 
12             For small trades, there are lots of issues to 
13   think about as a market maker.  Certainly, the 
14   creation/redemption sizes, creation/redemption fees, AP 
15   availability, all these going to us figuring out how wide 
16   to be in an ETF. 
17             What should be somewhat reassuring is that, in 
18   almost all cases, even in ETPs that rarely trade or 
19   sometimes don't trade for days on an exchange, market 
20   makers like us are able to provide markets that reflect 
21   the liquidity of the underlying basket for institutional 
22   size. 
23             Find a way to bridge the gap between what we're 
24   able to show for institutional size to the smaller sizes 
25   that retail investors and other small investors want to 
0178
 1   transact in, I think is probably -- should be a big focus 
 2   of what we talk about today. 
 3             MR. REDFEARN:  Awesome.  Thank you, Josh. 
 4             David? 
 5             MR. LaVALLE:  I want to again thank the 
 6   Commission, Commissioners and staff, for the opportunity 
 7   to include issuers in this conversation. 
 8             My name is Dave LaValle.  I'm the head of the 
 9   SPDR ETF capital markets team in the U.S. for State 
10   Street Global Advisors. 
11             State Street Global Advisors is a global asset 
12   manager and one of the largest ETF issuers in the world. 
13    We have 134 products in the U.S., all ETFs, over 675 
14   billion in assets, providing exposure to a broad range of 
15   asset classes and serving a broad range of investors.  
16   That's retail investors, both self-directed and through 
17   intermediary channels, and also institutional investors. 
18             Of particular importance to include issuers in 
19   this conversation, specifically given the alignment that 
20   we have with end investors, and therefore also the 
21   alignment with the SEC's mission of protecting those 
22   investors -- we're not interested in the routing strategy 
23   that a particular investor uses, what exchange they go 
24   through, what executing broker they choose to use, but 
25   more importantly that they're able to invest in our 
0179
 1   products with a high degree of confidence and with 
 2   minimal impact. 
 3             And so, we encourage exchange traded products 
 4   to be part of larger and broader equity market structure 
 5   conversations, given that ETFs represent roughly 30 cents 
 6   out of every dollar that's traded in the U.S. equity 
 7   markets.  I think Ari made a good point earlier that 
 8   we're about a third of the market, so we get a third of 
 9   the panels today, so -- and maybe a third of the crowd 
10   now that I look out there. 
11             Nonetheless, I think the point that I want to 
12   make -- two points here.   
13             First, for ETFs, thinly traded or less actively 
14   traded does not mean illiquid.  Traditional liquidity 
15   measurements like average daily volume, size of the 
16   inside quoted spread, while one measurement of liquidity, 
17   are not necessarily the best measurements of liquidity 
18   for an exchange traded product. 
19             You know, the liquidity profile and the 
20   tradability of the underlying constituents of an ETF is 
21   really what needs to be focused on, and I think Josh made 
22   some allusions to that in his opening comments. 
23             Additionally, I think we've spent an inordinate 
24   amount of time focused on the supply side of the 
25   equation, but we have to recognize that exchange traded 
0180
 1   products are sold.  These are products that are sold, not 
 2   bought, and so, we also need to focus an appropriate 
 3   amount of attention on the demand side of the equation, 
 4   which is what the issuers are focused on driving. 
 5             Once again, thank you very much for the 
 6   opportunity. 
 7             MR. REDFEARN:  Thank you, Dave. 
 8             Phil Mackintosh. 
 9             MR. MACKINTOSH:  So, my name is Phil 
10   Mackintosh.  I recently joined Nasdaq as a senior vice 
11   president and global head of economics and statistical 
12   research, but I've spent most of my career at Virtu and 
13   KCG and Credit Suisse, where each time I was the global 
14   head of trading strategy, so doing trading research.  In 
15   those roles, I produced research on trading costs, 
16   liquidity, market structure, and ETFs. 
17             So, I would like to thank the Commission and 
18   all the staff for having me, as well, and for leadership 
19   on this issue.  It's an important topic, improving the 
20   market structure for thinly traded ETPs. 
21             There's a few points that I would like to make, 
22   just to set the context for Nasdaq's point of view on 
23   this, and firstly, as my colleagues noted, Nasdaq's focus 
24   on this issue has been to find a way to enhance capital 
25   formation, focused on corporates and bringing more 
0181
 1   companies into U.S. public markets from the private 
 2   markets, and to help the institutional investors be able 
 3   to execute size trades in those stocks to diversify the 
 4   ownership of corporate stocks so that Mr. and Mrs. 401(k) 
 5   can participate in their growth. 
 6             I would also highlight that the U.S. ETP market 
 7   is, by many metrics, the most successful ETP market in 
 8   the world. That's thanks, in large part, to the structure 
 9   of the market and the current ETP ecosystem. 
10             Having said that, we're supportive of any plans 
11   that would help ETP issuers to grow new products and help 
12   investors to trade them with more confidence. 
13             As the Commission's recent white paper noticed, 
14   there are an even higher proportion of thinly traded ETPs 
15   than there are stocks. 
16             I would argue, though, that, although this 
17   problem is related, ETPs are different. 
18             So, specifically, as my colleagues here have 
19   said, ETPs benefit from liquidity providers who can 
20   provide and access liquidity of the underlying baskets at 
21   or near the basket valuations, making trading in size 
22   relatively inexpensive, a big contrast for the problem 
23   that we just discussed with trading corporates. 
24             The prevalence of this is highlighted by the 
25   fact that, in the Commission's data, you see a lot more 
0182
 1   OTC market share in the ETPs.  Okay?  However, some 
 2   thinly traded ETPs you'll see have excessively wide 
 3   spreads, and as Josh referred to, that's going to make it 
 4   more expensive for investors that are trading smaller 
 5   size on the lit markets. So, that mostly affects smaller 
 6   retail investors, either by increasing their costs or 
 7   maybe even by turning them away from potential 
 8   investments. 
 9             So, recent industry research suggests that this 
10   is mostly a problem for ETPs with low assets and volumes 
11   below creation unit size. 
12             So, I think Josh's comment about creation unit 
13   size is very relevant to this discussion, and what we 
14   found, what I found when I did this research in my past 
15   job was it was primarily because of the costs of holding 
16   and hedging whole creation units and holding the hedge on 
17   the other side that facilitating these really small 
18   retail trades actually ended up costing a large 
19   percentage of the trades. 
20             So, the spreads themselves aren't unreasonable 
21   given the costs of actually holding creation unit size 
22   positions. 
23             So, given that, I guess one thing we could 
24   potentially discuss would be to look at the minimum 
25   creation unit size as a way to reduce the cost for market 
0183
 1   makers of holding the positions in these new thinly 
 2   traded ETPs. 
 3             MR. REDFEARN:  Thank you, Phil. 
 4             Laura? 
 5             MS. MORRISON:  Thank you. 
 6             It's a pleasure to be here today.  My name is  
 7   Laura Morrison.  I am the SVP and global head of listings 
 8   for CBOE global markets exchange.  I started my career in 
 9   the equity market structure environment many years ago, 
10   working on the trading floor of the NYC, and then worked 
11   with corporate issuers for many years, market makers for 
12   a number of years, while the NYC was rolling out the 
13   hybrid NMS model, and then in 2007 took the opportunity 
14   of understanding both those aspects of the business to 
15   build the ETF listings business. 
16             About three years ago, I was tapped on the 
17   shoulder by a competitor BATS at the time, who said all 
18   we're doing are ETF listings, and this is what -- we are 
19   looking to change things up in a bit of a disruptive way. 
20    I found that to be intriguing, and I decided to make the 
21   move but have been a student of market structure and 
22   price discovery for many, many years. 
23             Having the opportunity to build the business at 
24   CBOE is most interesting, and understanding what it is 
25   that a market maker or an issuer is looking for, or then, 
0184
 1   therefore, the end investor is looking for in a listing 
 2   venue, in an exchange. 
 3             As David said -- I can't agree with him more 
 4   that the involvement of the issuer in this conversation 
 5   is critical, because our issuers are building products to 
 6   be listed on an exchange, financial products.  They're 
 7   not building autonomous vehicles or they're not building 
 8   the next new media social outlet or music delivery 
 9   system. 
10             It's critical that issuers and market makers 
11   and exchanges be involved in this discussion, in what 
12   makes ETFs unique and how we're so similar in many ways, 
13   how ETFs are so similar. 
14             So, with that, as it relates to the topic of 
15   this discussion, I strongly recommend let's keep things 
16   simple.  I was impressed with the rollout of our 
17   amendment number 12 to limit-up/limit-down post August 
18   24th that relates to harmonization amongst exchanges in 
19   terms of what it takes to resume after a product has gone 
20   into a halt intraday. 
21             There are instances where exchanges certainly 
22   should compete, and do, but there are also instances 
23   where keeping matters simple provides a benefit of, 
24   certainly, confidence in the marketplace. 
25             So, while we think about potentially treating 
0185
 1   products that don't trade as frequently different, I also 
 2   think there's other solutions that we could consider, 
 3   some of which we've been looking at, like the periodic 
 4   auctions, potentially, or even a trade at NAV solution, 
 5   in addition to the auctions that we're running on the 
 6   open intraday and on the close. 
 7             But most importantly, we look at statistics of 
 8    when a market maker is involved in a new launch, because 
 9   frankly, I've been involved in over 2,500 launches in my 
10   career, with 100-plus issuers, and when there is a launch 
11   without a market maker, a dedicated market maker, it's a 
12   significantly different experience versus not having that 
13   dedicated market maker who is committed to providing that 
14   liquidity not only on the first day but as the issuer is 
15    able to then get out there and sell it and distribute it 
16   and market that product. 
17             It does make a difference.  Our ability to pay 
18   that market maker and also to come up with different 
19   unique programs that will incent different types of 
20   traders to come to our market and do just the same. 
21             So, once again, thank you for this opportunity. 
22    I appreciate having this session focused on ETFs, in 
23   particular. 
24             MR. REDFEARN:  Thank you, Laura. 
25             Greg? 
0186
 1             MR. SUTTON:  Thank you for having me.  My name 
 2   is Greg Sutton.  I'm at Citigroup Global Markets.  I 
 3   oversee our low touch businesses, and germane to this 
 4   conversation, I also manage the global program trading 
 5   business, and housed within that at Citigroup are a 
 6   number of different ETF focuses for us. 
 7             So, we have creation/redemption business, which 
 8   is largely done on behalf of our institutional clients, 
 9   as well as market makers; the sales of ETPs, largely, 
10   again, to our institutional client base; and we also 
11   manage the ETP provider relationships across that desk, 
12   as well. 
13             So, it puts us in, I think, a very unique 
14   footing where we have the full life cycle of the ETF 
15   product in terms of -- you know, from beginning to end, 
16   we're involved in all facets or can be involved in all 
17   facets of it. 
18             Respective to being here today, again, thank 
19   you for that, and we certainly think that ETPs should be 
20   part of the discussion with respect to UTP and 5250, and 
21   I look forward to having a conversation on that today. 
22             Thank you. 
23             MR. REDFEARN:  Charles. 
24             MR. THOMAS:  Thank you for having us here 
25   today.  It's a great program that you put together.  I'll 
0187
 1   say I learned a lot from the first two sessions, so 
 2   hopefully we can continue that momentum in the session on 
 3   ETFs. 
 4             My name is Charles Thomas.  I lead our ETF 
 5   capital markets function at Vanguard. 
 6             As many of you may know, Vanguard is a global 
 7   investment management firm.  We offer mutual funds, ETFs, 
 8   and a variety of investment services in more than 20 
 9   countries and manage over $5 trillion of our client 
10   assets. 
11             Our mission is clear and straightforward, to 
12   take a stand for all investors, to treat them fairly, and 
13   to give them the best chance for investment success. 
14             With that in mind, when we think about market 
15   structure developments and changes to market structure, 
16   we always tie that back to our end investors -- retail, 
17   mom and pop investing in their 401(k), financial advisors 
18   trying to improve outcomes for their clients, pension 
19   funds investing on behalf of their employees.  So, we do 
20   try to tie that experience back to our end investors. 
21             Specific to ETFs, we have over 800 billion in 
22   assets under management across our global ETF platform.  
23   We have ETFs listed in five different regions.   
24             My team oversees the U.S. listings.  We have 77 
25   products listed across the three major exchanges 
0188
 1   represented here today, and collectively, those 77 
 2   products trade in excess of a trillion dollars in annual 
 3   trading volume. 
 4             We support those ETFs by working closely with 
 5   the market participants, the exchanges, the market 
 6   makers, everyone that's involved in trading those 
 7   products.  We facilitate an efficient creation/redemption 
 8   mechanism to get those ETFs out into the marketplace.   
 9             We're very thoughtful about product design, and 
10   we work closely with our clients to help them understand 
11   and educate them around ETF issues. 
12             Given that experience, our perspective is that 
13   the ETF market is both healthy and well functioning.  
14   ETPs and ETFs, exchange traded funds, in particular, 
15   benefit from a pretty healthy regulatory framework that 
16   is very well functioning. 
17             First, ETFs are governed by the '40 Act, so 
18   they're regulated investment products, which encourages 
19   diversification and investor protection, and then the 
20   exchange traded nature of the funds ensure that they 
21   benefit from all of the enhancements in equity market 
22   structure that we've seen over the past few decades, 
23   everything from decimalization to electronification of 
24   markets and all the protections introduced with the Reg 
25   NMS. 
0189
 1             So, given the healthy, well-regulated 
 2   marketplace, we feel that ETFs generally are serving 
 3   investors well.  Over the past few decades, investors 
 4   have been able to achieve exposure to diversified 
 5   investment vehicles at a very low holding cost, with very 
 6   low trading costs. 
 7             So, we approach this conversation from the 
 8   perspective of continuous improvement and trying to 
 9   further improve those investor outcomes. 
10             With that in mind, I will echo some of the 
11   comments that my colleagues have made and will focus on 
12   one key consideration maybe for discussion on the panel, 
13   and I would frame that as the definition of liquidity for 
14   ETFs versus common stocks. 
15             We heard a lot in the prior two panels around 
16   defining liquidity.  Maybe there's some nuance, but the 
17   study that Alex walked us through earlier today generally 
18   focuses on trading volume as the main metric, and that 
19   makes sense for common stocks.  If you think about the 
20   primary market issuance, the issuer comes to market maybe 
21   once or rarely overtime, and then the main metric of 
22   liquidity is how often investors are buying and selling 
23   those shares in the secondary market. 
24             There's a lot of commonalities between stocks 
25   and ETFs when it comes to equity market structure.  
0190
 1   Liquidity is more nuanced when we think about the 
 2   difference, and ETFs are different, and that comes back 
 3   to the primary market issuance. 
 4             ETFs, unlike common stocks, have a regular 
 5   daily creation/redemption function that some of my 
 6   colleagues have mentioned. 
 7             So, if you're an ETF investor or an ETF market 
 8   maker, you can approach the fund on a daily basis to 
 9   either increase or decrease the supply of ETF shares out 
10   there in the marketplace. 
11             So, we would argue at Vanguard that ETFs have 
12   effectively unlimited liquidity and an ETF can both be 
13   thinly traded and very liquid at the same time because of 
14   that underlying creation/redemption process. 
15             That process also provides a very robust 
16   reference price for arbitraging the ETF, and generally 
17   ensure investor confidence that they'll be able to 
18   transact at the value of the fund. 
19             So, overall, we think the ETF marketplace is 
20   healthy, well functioning, and to the extent that there's 
21   opportunities for improvement, we would encourage 
22   acknowledging that ETF liquidity is a little bit 
23   different. 
24             MR. REDFEARN:  Thank you, Charles. 
25             Kumar. 
0191
 1             MR. VENKATARAMAN:  Thank you, Brett, for the 
 2   kind invitation.  It's a pleasure to be here.   
 3             My name is Kumar Venkataraman.  I'm a professor 
 4   of finance at the Cox School of Business at Southern 
 5   Methodist University, and I am the academic on the panel, 
 6   and so, my area of research is market microstructure, and 
 7   I have written about the structure of the markets in the 
 8   fixed income space, in the commodity space, in the equity 
 9   market space. 
10             And as I was listening to the first panel, I 
11   was reminded of my PhD dissertation, which I completed in 
12   the late 1990s, looking at trading on the New York Stock 
13    Exchange versus the European electronic markets, and one 
14   of the articles that I published, part of my 
15   dissertation, made the point that the key difference 
16   between the old floor-type market structure and the 
17   electronic markets is anonymity.  
18             That reputation and trader relationship plays a 
19   role in the old floor-based system whereas it does not in 
20    the electronic world, or it's difficulty to replicate 
21   that  in the electronic markets, and so, it was 
22   interesting for me to hear Jason and Frank and others 
23   bring up that distinction today. 
24             So, my views on the markets are formed based on 
25   my analysis of data, and as an academic, I have had the 
0192
 1   benefit of looking at some very detailed sort of data at 
 2   exchanges outside the U.S., and this data has account-
 3   level information, so we can actually look at the trading 
 4   of sort of each participant in the market and follow that 
 5   over time. 
 6             And so, some of the observations that I will 
 7   share today is going to be based on some of the research 
 8   that I have completed, as well as other research that 
 9   I've read that have been completed in academic studies. 
10             I would like to make one observation to just 
11   set  up my thoughts for market structure of less liquid 
12   securities, which is that if you look at sort of this 
13   very detailed account-level data and you study the 
14   behavior of market makers, one thing that's very clear is 
15   that, if you have market makers without any affirmative 
16   obligations, then their participation tends to be highly 
17   correlated with each other. 
18             So, when market conditions are favorable, then 
19   you find that there's a lot of participation, and if 
20   market conditions are unfavorable for market making, then 
21   you find that they all scale back together, in unison, 
22   and this makes complete economic sense. 
23             When market conditions are unfavorable, on a 
24   risk-adjusted basis, it's not very profitable to make 
25   markets, and if you're a market maker with that 
0193
 1   obligation, you are there to make money, so obviously, 
 2   that's exactly the way that we would expect that you 
 3   would behave. 
 4             But what it means from sort of a regulating 
 5   perspective or from the perspective of an issuer or an 
 6   investor in securities is that it results in sort of 
 7   fragility in liquidity supply, which means that when you 
 8   are trying to trade, you may not have the counterparty on 
 9   the other side, and this is particularly relevant for 
10   small cap stocks, because these stocks -- for these 
11   stocks, these unfavorable conditions -- low volume, one-
12   sided flow, etcetera -- tends to happen more often. 
13             So, I think some of the conversations we are 
14   having today is consistent with what we would expect in 
15   this market structure, where predominantly you have 
16   market makers without affirmative obligations, which is 
17   that there is sort of more participation, generally, 
18   because thinly traded stocks are just not really 
19   profitable from a market making perspective and that 
20   there is fragility in liquidity supply, that there is 
21   high variation. 
22             So, this brings in the idea of a designated 
23   market maker, and in my remarks later on, I would like to 
24   share what we observe in markets around the world, where 
25   they have these  types of market making arrangements. 
0194
 1             Thank you. 
 2             MR. REDFEARN:  Thank you, Kumar. 
 3             Doug? 
 4             MR. YONES:  I'll kick off things.  My name is 
 5   Douglas Yones, and I would like to first thank the SEC 
 6   for having me here today.  I'm the head of exchange 
 7   traded products at the New York Stock Exchange.  As a 
 8   result of alphabetical order, I'm going to try and find 
 9   new comments to make that haven't already been stated by 
10   my colleagues. 
11             For those that aren't aware, the New York Stock 
12   Exchange -- in particular, NYC Arca -- the Arca exchange 
13   has about 83 percent of the assets under management for 
14   the ETF industry. 
15             Out of the 2,130 exchange traded products that 
16   exist today, 1,497 are listed with us.  So, we have a lot 
17   of data, and you know, I think, if I need to reiterate 
18   anything or some of my opening points that I'll narrow it 
19   down to is alongside Mr. Thomas' point, which is exchange 
20   traded products are very different from single 
21   securities, and where that difference lies is the fact 
22   that their value is a derived value, and at any given 
23   moment, market participants, liquidity providers, lead 
24   market makers, or market makers have the ability to 
25   compete against one another for that derived value. 
0195
 1             And where issuers have some control in this 
 2   space is they can derive the overall costs of the 
 3   creation and redemption process. 
 4             So, whether it be through their custody fees, 
 5   whether it be through their basket sizes, whether it be 
 6   through the fees they charge, let's say, on cash 
 7   components or delivery of components, there is an impact 
 8   that an overall issuer can have on their particular 
 9   security, which, again, differs from the single stock 
10   marketplace. 
11             And so, I hope part of our conversation will 
12   have an opportunity to explore not just the idea of UTP 
13   but maybe some of the other ways in which all of us as 
14   market participants can help to drive better liquidity 
15   for all ETFs, not just low-volume ones, although as the 
16   paper discussed, the majority of our exchange traded 
17   products today are what would be considered low-volume. 
18             So, I think, in many cases, what we'll talk 
19   about here today is representative of almost every ETF 
20   that we have listed on our overall marketplace. 
21             The final comment, I think, was very important 
22   from the first two sessions.  We kept hearing the 
23   question of what is the issuer trying to solve for, and I 
24   loved the one comment that said market structure won't 
25   change investor interest, and that's no different in the 
0196
 1   world of exchange traded products. 
 2             When a product is new, it's up to the issuer 
 3   and overall community to help go and explain what the 
 4   investment product is to the end investor and help 
 5   convince them that this either does or doesn't make sense 
 6   for an investment portfolio.  That conversation happens 
 7   outside of whether or not the size of average daily 
 8   volume exists. 
 9             So, as we continue our conversation here today, 
10   I hope we give a chance for some of those smaller 
11   products to take root, because even the most popular and 
12   highest volume products today weren't always so a number 
13   of years ago. 
14             Thank you. 
15             MR. REDFEARN:  Thank you very much.  I 
16   appreciate all those opening comments.  So, let's get 
17   into some of the discussion here. 
18             I think this issue of thinly traded but highly 
19   liquid is an interesting one and certainly one that 
20   strikes us differently than when we look at corporate 
21   stocks.  So, I want to explore that a little bit. 
22             So, if we wanted to measure liquidity in 
23   exchange traded products, what would you say is the best 
24   way that we could sort of assess the liquidity of the 
25   product? 
0197
 1             And I'm going to maybe start -- Josh, do you 
 2   have any thoughts on that?  And then I'm going to put it 
 3   out to anybody else who wants to comment. 
 4             MR. KULKIN:  I can tell you how we think about 
 5   it as market makers, because at the end of the day, when 
 6   we're providing liquidity, we're going to have to go out 
 7   and either hedge with the ETF that we've just sold or 
 8   bought or with the underlying basket. 
 9             So, a good example of an illiquid ETP is like 
10   MDYV, which is a mid-cap value ETF that Dave and his 
11   folks over at State Street put out that trades probably 
12   about, I don't know, 5 or 10 thousand shares a day, and 
13   if you use Bloomberg's TCA or trade cost analysis tool, 
14   it will tell you that the impact that you should have if 
15   you want to buy $150 million of it or something like that 
16   is something like 14 percent. 
17             So, when you think about what impact you should 
18   have based on the size of your order, you look at average 
19   daily trading volume and historical volatility, and we 
20   all have models that look about the same for figuring out 
21   how to take those two inputs and come out with an 
22   expected impact. 
23             At Jane Street, what we do is we take a look at 
24   the impact that that would imply in the basket, and we 
25   take a look at all of the individual stocks in the 
0198
 1   basket, weighted by their weight in the ETF, and say if 
 2   we had to go out and hedge the basket, what should the 
 3   impact be, and that would give something like a 4- or 5-
 4   basis-point impact number for that same order. 
 5             So, when you see somebody comes to the market 
 6   and wants to trade an ETF like that, even though it's 
 7   very thinly traded, if they want to trade it in the 
 8   institutional size and none of the sort of structural 
 9   problems of these really small unit sizes or, you know, 
10   the cost of creations being relevant to the overall cost 
11   of the transaction, you can sort of just use the 
12   liquidity of the underlying basket as your proxy for how 
13   much impact that order should have, and that's going to 
14   be our number one input when we're thinking about how to 
15   come up with a width for that. 
16             MR. REDFEARN:  And is the tool you're talking 
17   about on Bloomberg looking at the impact, also, of -- 
18             MR. KULKIN:  I don't think Bloomberg's tool 
19   does the fancy thing for ETFs.  I think it does the naive 
20   thing.  But as a firm and when we go out and our 
21   institutional broker talks to customers, we're trying to 
22   explain that -- look at the liquidity of the basket when 
23   you are looking at how much it should cost you to 
24   transact. 
25             MR. REDFEARN:  Greg. 
0199
 1             MR. SUTTON:  Just to follow on with that, so we 
 2   have a platform that we've created that would give to 
 3   institutional clients -- actually, ETF providers get it, 
 4   as well, and it's borne out of, really, just program 
 5   fitting theory, and it's looking at a basket, what it's 
 6   going to cost, you know, whether it's global, whether 
 7   it's U.S., emerging market, developed frontier, whatever 
 8   it might be. 
 9             You can put in the ETF.  You can put in the 
10   notional value of the ETF.  It automatically deconstructs 
11   it into the basket, and you can immediately figure out -- 
12   so, similarly to what Josh is saying, we're looking at it 
13   more from what the impact is to the underlying, what we 
14   can equate what the impact is on the underlying 
15   securities, and then have a conversation with the client 
16   or the client can look at themselves and figure it out. 
17             Our institutional market making team uses the 
18   product, as well, and again, it's just a way to look at -
19   - you know, you could trade the ETF in the marketplace, 
20   but obviously, you have to be very, very aware of where 
21   the underlying is and where you can source the underlying 
22   and at what cost you can source the underlying. 
23             So, that's a vehicle that is available for our 
24   client base and providers. 
25             MR. REDFEARN:  Any other thoughts on that? 
0200
 1             Phil. 
 2             MR. MACKINTOSH:  Yeah.  I would just add that, 
 3   a lot of the time, the market makers will also use 
 4   derivatives to hedge.  Whether they're trying to get a 
 5   basket trade done and there's two or three stocks that 
 6   are just too illiquid to finish fast, you could still use 
 7   futures and get most of your risk hedged. 
 8             But taking an ETF like HYG, high-yield 
 9   corporate bonds, often what you'll find is people will 
10   trade in the futures market to hedge the interest rate 
11   curve and then use the CDS to get their credit exposure. 
12             So, just because the underlying doesn't look as 
13   liquid as you think doesn't mean there's not another 
14   alternative way to get your risk hedged and to be able to 
15   facilitate large size. 
16             MR. REDFEARN:  I'm sure Kumar could speak to 
17   the fact that we have had this discussion in the FIMSP, 
18   the fixed income market structure panel, where we've 
19   looked at, you know, a lot of ETFs that actually have 
20   corporate bonds underlying them, as well, and there's a 
21   different way of thinking about liquidity in those 
22   metrics, as well. 
23             MS. MORRISON:  I think it's also important to 
24   note that, when an ETP comes to market, there's really a 
25   deep dive into what that underlying liquidity is and does 
0201
 1   it, in fact, meet our obligations for listing? 
 2             So, the generic obligations, if, in fact, it 
 3   does meet those generic listing obligations under our 
 4   rules approved by the SEC, or if not, we go for a 
 5   specific product approval, and one of the main key 
 6   conversations is what is the liquidity of the underlying 
 7   asset, and is the information that's available on the 
 8   valuation something dependable?  So, there's an extensive 
 9   review that happens. 
10             In addition, we also -- all of the exchanges 
11   enhance the continued listing obligations to make sure 
12   that that diversification and that liquidity is available 
13   on a continuous basis, not just at the time of listing. 
14             MR. REDFEARN:  So, Greg, I have a question for 
15   you. When you run your model across the ETP -- and maybe 
16   I should ask you, you know, what is your universe of 
17   traded -- exchange traded products, because we are 
18   looking at a world now where there are over 2,000 ETPs 
19   listed on the marketplace. 
20             By the other standard, I think some 76 percent 
21   is trading less than 100,000 shares.  Okay.  So, that's 
22   not the best metric. 
23             We have different ways of looking at the 
24   underlying. 
25             I haven't seen any good, you know, analysis of 
0202
 1   ETPs across the board. 
 2             What is the trend right now?  Are there more 
 3   illiquid ETPs by this metric coming out?  How do you see 
 4   the liquid versus illiquid dynamic?  How do we get some 
 5   feeling for what is the illiquid universe versus the 
 6   liquid universe, as opposed to having metrics that are 
 7   kind of hard to measure? 
 8             MR. SUTTON:  That's a good question.  I think, 
 9   again, it's a -- so, first of all, we do -- we'll 
10   transact across the entire universe of securities. 
11             Again, I bring up equities earlier as they are 
12   --roughly 75 percent of AUM is in an equity underlying.  
13   Obviously, there's been a lot of growth in the fixed 
14   income space. 
15             But you know, as we look at liquid versus 
16   illiquid in ETFs, again, from a creation/redemption 
17   standpoint, we really focus on what we can do on the 
18   underlying, and so, whether it's a true replication, so 
19   it's a relatively liquid basket that's accessible in U.S. 
20   market hours, if it's a  U.S. client, or if it's a non-
21   dollar product or something a little bit more bespoke and 
22   we have to come up with proxy hedging, to us it's all 
23   about -- so, there's two different components. 
24             We have a client business that's doing agency 
25   creation/redemption for clients, and then we also have 
0203
 1   the market making business. 
 2             So, similar to Josh, you know, we're creating 
 3   proper hedges that we can get ourselves into a position, 
 4   and it's all about the amount of tracking error that we 
 5   are willing to accept when we're taking into a position, 
 6   and that's where the math is for us around providing 
 7   liquidity. 
 8             So, whether it's providing liquidity as a 
 9   principal block trade on an ETF or we're providing 
10   liquidity in the form of an ETF credit redeem, they're 
11   two different conversations, but there's a lot of 
12   parallels that, you know, you can certainly say, well, if 
13   I traded, you know, using the baskets with maybe a proxy 
14   hedge or even if it was a full replication, I think it's 
15   going to cost me X.  If you want to buy it right now, 
16   it's going to cost you Y, and then it's a decision for 
17   the client to make. 
18             But I mean, back to your question around 
19   illiquid versus liquid ETFs, I would say that's tough to 
20   say.  Again, it's all around the underlying and the 
21   hedge-ability of the underlying so we can provide a 
22   proper market. 
23             MR. REDFEARN:  So, are there some, then, when 
24   you look at the cost, they're too illiquid to touch? 
25             MR. SUTTON:  I wouldn't say they're too 
0204
 1   illiquid to touch, because you know, we generally think 
 2   there's a price for everything, but it might be at a 
 3   significantly wider spread if I can't -- you know, if I 
 4   can't borrow something, or it's a much difficult -- it's 
 5   a much difficult tracking error for us to estimate. 
 6             If there's much more volatility in that number, 
 7   then we're going to have a wider -- we are going to have 
 8   wider pricing. 
 9             I mean, Josh might have better thoughts around 
10   that from -- as being -- as an electronic market maker, 
11   but that's from the institutional side. 
12             MR. KULKIN:  Yeah.  I mean, we make markets in 
13   every ETP that's listed in the United States.  You know, 
14   again, sort of as Greg was saying, there's nothing we're 
15   not willing to make a market in. 
16             The only times we -- the only times that we're 
17   not just going to reflect the liquidity of the underlying 
18   asset, be that, you know, high-yield bonds or some of the 
19   -- there are ETFs that reflect -- that are composed of 
20   less liquid high-yield bonds.  I mean, those are going to 
21   be wide but no wider than the high-yield bonds 
22   themselves. 
23             The only time that we sort of are going to be 
24   wider than the basket is when one of the structural 
25   issues that I alluded to in my opening remarks and other 
0205
 1   people seem to agree with come into play.   
 2             If there are minimum creation/redemption sizes, 
 3   if we sell an ETF and we have obligations, even though we 
 4   get six days, if there's not going to be order flow in 
 5   the other direction, I'm going to have to create it, and 
 6   oftentimes those positions will and have sat on our books 
 7   for multiple years. 
 8             We're going to ask what seems like unreasonable 
 9   amounts of spread on trades like that, but it's just, 
10   again, reflecting our underlying costs. 
11             So, to the extent that we can ameliorate those 
12   underlying costs either by lowering creation/redemption 
13   unit sizes or if issuers waive creation/redemption fees 
14   on their less liquid products, I think those are sort of 
15   the things that -- those are the only real times when we 
16   feel like we're not able to truly reflect the liquidity 
17   of the underlyings. 
18             MR. REDFEARN:  Dave? 
19             MR. LaVALLE:  Sure.  I think another way to 
20   think about this -- and we sit in a unique spot where we 
21   look across a broad range of investors, and we also look 
22   across a broad range of liquidity providers, both in the 
23   form of, you know, Jane Street being a market maker or, 
24   you know, Citi being a global bank, both in principal 
25   market making function, also in the primary market, so 
0206
 1   secondary and primary market functions. 
 2             There's two layers of liquidity here, and I 
 3   think you need to be keenly focused on the reality that 
 4   whatever  is trading in the ETF shares and liquidity of 
 5   the ETF shares is the kind of primary function through 
 6   which you can source liquidity in trying to trade, either 
 7   buy or sell that product, and then that secondary level 
 8   liquidity is in the form of, you know, the basket's 
 9   liquidity profile, which I think Josh and Greg have done 
10   a great job of articulating. 
11             What we find in our less actively products that 
12   are also quite liquid by basket terms is that, 
13   oftentimes, the secondary market spread profile is kind 
14   of quoted for the worst case scenario. 
15             So, if -- I won't speak for Josh, but what we 
16   hear from all of our market makers is if there was more 
17   transparency into the type of order that was coming in 
18   the front door, if it was 300 shares of a security that 
19   trades 10,000 shares a day, you know, they're going to 
20   quote that in one way. 
21             If it's 50,000 shares or 100,000 shares of a 
22   security that has a 50,000-share creation mid-size, they 
23   will oftentimes be a lot tighter than the actual quoted 
24   spread that's -- you know, if you go out on any market 
25   data screen, and we have countless examples where we'll, 
0207
 1   you know, show clients examples of, you know, maybe a 25-
 2   basis-point spread in the product but 10, 15, 20 times an 
 3   average day's worth of volume is trading, you know, at or 
 4   near the midpoint based upon the fact that there's kind 
 5   of reliable understanding of what the risk profile is and 
 6   therefore how the market maker or the liquidity provider 
 7   is going to respond or going to be able to manage that 
 8   risk. 
 9             And so, you know, I think it's important to 
10   kind of understand that that quoted spread is somewhat 
11   kind of quoted for the worst case scenario, because in 
12   many instances, if you sell 500 or 1,000 shares of a 
13   product and you do have a Reg SHO obligation or you're 
14   not able to source those shares, you know, it will result 
15   in a creation unit, which means, you know, a large 
16   capital commitment and, you know, impairment on your 
17   balance sheet. 
18             MR. REDFEARN:  I'll just say, if anybody wants 
19   to speak, just hit your button and turn the light on.  
20   That will help me get to you. 
21             Charles. 
22             MR. THOMAS:  Yeah.  So, Brett, I'll circle back 
23   to your original question.  We kind of proposed a 
24   different definition of liquidity for ETFs, and you 
25   asked, okay, how can we characterize that? 
0208
 1             I agree with Dave's points that there's two 
 2   layers of liquidity.  So, looking at the secondary market 
 3   trading for ETFs is one element of ETF liquidity, but the 
 4   real liquidity backstop for an exchange traded fund is 
 5   the creation/redemption mechanism. 
 6             So, maybe to put some color around your 
 7   question, Greg had mentioned that about three-quarters of 
 8   the products listed in the U.S. are equity-based.   
 9             If we think about that landscape and what that 
10   looks like, most of those products are going to be 
11   market-weighted, tied to more or less plain vanilla index 
12   benchmarks, which means that the portfolios are tilted 
13   towards the larger-cap, heavily-traded end of the equity 
14   market spectrum. 
15             And if we think about what happens when you're 
16   trading an ETF and then showing up in the primary market, 
17    you're taking one trade and breaking that across, in 
18   most cases, hundreds, if not thousands of securities, and 
19   splitting up that liquidity, trading it in very small 
20   sizes in each underlying name. 
21             So, given that that's the market -- the 
22   structure that we're dealing with, I think we would argue 
23   that it would be tough to find an ETF that we would call 
24   illiquid.  There may be some one-off exceptions, niche 
25   products, but given that most products are tied to 
0209
 1   broadly diversified plain vanilla benchmarks and you're 
 2   breaking up that trade size, they're very liquid. 
 3             MR. REDFEARN:  Do you think that's the same -- 
 4   I mean, it's interesting, because we do look at the data 
 5   and you see a lot of these names that trade 5,000 shares 
 6   or 10,000 shares, not even trading every day, so -- and 
 7   you wonder about the retail market.   
 8             So, for those stocks, it sounds to me like 
 9   you're saying even if it's 1,000 shares, 5,000 shares, 
10   generally, it's just simply not an issue because of the 
11   other dynamics that we're witnessing here. 
12             MR. KULKIN:  Yes.  I think there's definitely 
13   improvement.  I mean, I do feel, if there were smaller 
14   creation/redemption unit sizes, retail would have a 
15   better experience with some of the less liquid products 
16   if they do find their way to invest in them. 
17             So, I definitely think it is -- it's not a 
18   gigantic problem, but it is certainly less than optimal 
19   than how it could be. 
20             MR. REDFEARN:  So, it's less a function of the 
21   market maker and the market-specific activity as it is 
22   the primary issuance. 
23             So, let's talk about that, right?  We are 
24   looking at ways in which we can enhance liquidity in the 
25   marketplace. What is the optimal solution here in terms 
0210
 1   of changing create/redeem sizes? 
 2             MR. KULKIN:  Sure.  I mean, I think giving 
 3   issuers more flexibility to set lower creation/redemption 
 4   unit  sizes. 
 5             Right now, it's my understanding that they're 
 6   required under their exemptive relief that -- that they -
 7   - from the '40 Act -- and hopefully, in the new ETF rule, 
 8   they'll have more flexibility around creation and 
 9   redemption unit sizes, so that me, as a market maker, I 
10   can facilitate smaller sizes on less frequently traded 
11   ETPs, let's call them, as opposed to illiquid, and still 
12   not have that obligation to then create to meet 204 rules 
13   and hold that on my balance sheet for a long, long time. 
14             I think, you know, I think a lot of the 
15   products that don't trade a lot just -- issuers do a good 
16   job of getting a good variety of products out there that 
17   might not meet all the market conditions, and even though 
18   they don't trade doesn't mean that they are not excellent 
19   products for certain market conditions. 
20             So, to the extent that issuers like market 
21   makers making good markets in their products even when 
22   they're not under high demand, having them accommodate 
23   getting in and out of smaller sizes would probably be 
24   good for retail investors, for the issuers themselves, 
25   and certainly for market makers. 
0211
 1             MR. LaVALLE:  I would say that when we're 
 2   talking about, you know, creation/redemption size, it's 
 3   certainly a different lever that we can utilize, but 
 4   there's two sides to that coin. 
 5             The first side is obviously you need to ensure 
 6   that, you know, our portfolio managers are principally 
 7   focused on tracking error, while capital markets teams 
 8   are principally focused on kind of liquidity profile or  
 9   secondary market and primary market performance of these 
10   products, and so, there's a balance to strike there to   
11   ensure that you have a creation/redemption unit size that 
12    can effectively allow the portfolio manager to manage 
13   the creation/redemption. 
14             So, you could say that -- and we applaud the 
15   Commission's work with John Hancock that got the approval 
16   for a 10,000-share creation/redemption unit size, which 
17   is fantastic and offers some flexibility. 
18             However, the reality is that the portfolio 
19   manager can't effectively manage that portfolio and have 
20   optimal tracking, that decreases the ability for our 
21   investors to invest with confidence.  So, there's 
22   striking a balance between those two. 
23             I think the other reality is, is when we're 
24   bringing products to market, we need to ensure that 
25   whatever the underlying portfolio is, that that 
0212
 1   underlying portfolio gives enough transparency to members 
 2   of the liquidity-providing community and market making 
 3   community to ensure that there won't be long sustained 
 4   instances where the second market trading value is 
 5   decoupling from the intrinsic value of the product. 
 6             And so, we're constantly focused on bringing 
 7   products to market that allow our market participants to 
 8   kind of engage with the product with a high degree of 
 9   confidence, therefore keeping spread as tight as they 
10   possibly can be. 
11             I think it's also worth mentioning that, at 
12   least in our case, we've closed more products in the last 
13   three years than we've launched, and so, the dynamics of 
14   bringing products to market are also changing, and the 
15   stigma -- negative stigma associated with closing 
16   products is something that I think has been changing over 
17   the past several years. 
18             MR. REDFEARN:  So, we do see a lot of new 
19   issuance coming to market, and there have been a lot of 
20   names, and it's interesting you bring up the closing 
21   product, because some of them that don't trade, it 
22   appears -- you know, it makes you wonder about the level 
23   of interest that's been generated in that. 
24             So, you know, how should we think about all 
25   these products that are on the marketplace where 
0213
 1   sometimes they just don't seem to trade? 
 2             Is that a one-year, a two-year, a three?  I 
 3   mean, how do we think about what is the gestation period 
 4   that's given for these products to sort of grow into a 
 5   product where there's investor interest or not? 
 6             Doug. 
 7             MR. YONES:  I'm not going to jump in -- I would 
 8   like to just finalize the previous topic before we move 
 9   too far away. 
10             MR. REDFEARN:  Sorry, I didn't see your light. 
11             MR. YONES:  No worries. 
12             In addition to Reg M relief, which I think was 
13   referenced a number of times, there are other levers we 
14   could pull as an overall industry in terms of relief for 
15   issuers that would help to reduce overall 
16   creation/redemption sizes. 
17             As Dave mentioned, Reg M relief that was issued 
18   to John Hancock, which could be applied elsewhere -- it 
19   does require for many issuers to have a 10,000-share 
20   basket that perfectly strikes across their entire lineup, 
21   and that's just not possible for many issuers, as they 
22   have not been awarded exemptive relief that allows for 
23   either customs or customized baskets, and so, bringing 
24   customized baskets to other issuers I think would help to 
25   alleviate some of that stress. 
0214
 1             The other place would be around the liquidity 
 2   rule, so for those issuers that create/redeem in cash, 
 3   they often then are subject to other behaviors as a fund 
 4   company to publish their liquidity results, and so, 
 5   again, looking at some sort of ETF exemption in those 
 6   places, I think, could help additional issuers reduce 
 7   their creation/redemption size and help to narrow 
 8   spreads. 
 9             Did you want to circle back to the other 
10   question? 
11             MR. REDFEARN:  Yeah.  I mean, on the other 
12   question, I am interested in this, because we see a lot 
13   of products that come to market, and you know, we're just 
14   wondering what is that process? 
15             MR. YONES:  Yeah.  I think it's an interesting 
16   one, because there's probably no set standard incubation 
17   period.  What I mean by that, even in my former role at 
18   an ETF issuer, many times an asset manager, just like in 
19   the fund space, may create a fund thinking that the 
20   investment thesis for that fund might not occur for many 
21   years. 
22             It might be after the end of a bull market or 
23   the beginning of a bull market, and so, it's very hard 
24   for the  builder of an investment product to know exactly 
25   when that product will be successful. 
0215
 1             So, I think it is difficult for any of us to 
 2   prescribe a set range in which we would expect to see a 
 3   fund at a certain size or a certain level of trading. 
 4             The natural market dynamics are there that a 
 5   fund company has to sustain a certain level of cost to 
 6   keep an investment product open, and so, there are 
 7   natural forces at work, as Dave mentioned. 
 8             We continue to close, you know, many ETFs each 
 9   year, alongside opening new ones, and so, the natural 
10   environment is working. 
11             MR. LaVALLE:  I would just also say, in light 
12   of my opening comments, that this really talks to the -- 
13   driving the demand side of the equation, and I think it's 
14   very difficult to avoid that reality. 
15             We've spent a lot of time talking about supply, 
16   quoted spread, size of the inside, all these notions of 
17   how we can make the market trade more efficiently from 
18   the supply side, but I can promise you that I could 
19   probably get Greg and Josh to battle over any product 
20   where I bring a client to actually trade, you know, 
21   whether it's 50,000, 100,000, or any sort of 
22   institutional size of a trade. 
23             And so, when we're bringing new products to  
24   market, there's a trans-fund alignment in our firm, 
25   across our product, you know, development team, capital 
0216
 1   markets team, and also our distribution team, and when we 
 2   go to  raise seed capital and we go to work with members 
 3   of the liquidity-providing community, we explain the 
 4   marketing  plan, the distribution plan that we're going 
 5   to support this product launch with, and it's met with 
 6   open arms, as opposed to, you know, any sort of 
 7   proverbial, you know, opportunity of just kind of 
 8   throwing products out there and seeing what sticks. 
 9             And so, we have to focus on driving the demand 
10   side of the equation, and we're committed to doing that, 
11   and if we do do that, we'll have better likelihood of 
12   success of products coming to market, and that will lead 
13   to having, you know, greater attention being paid to 
14   those products by the liquidity providing community. 
15             MR. MACKINTOSH:  I've got a couple of data 
16   points to kind of tie the last two discussions together. 
17             So, just talking about the proliferation of 
18   ETPs, as Brett said, there's over 2,000 ETPs now.  What's 
19   interesting is the last 10 years.  We've kind of been 
20   stuck in a range in terms of value traded in ETPs.  
21             So, although the number of ETPs has grown and 
22   although the assets in them has grown about six-fold, 
23   we've actually been stuck in a sort of 60 to 80 billion 
24   dollar a day range in terms of volume or value traded, 
25   and I think one of the reasons for that is one of the 
0217
 1   things you see with retail flow. 
 2             Retail has been buying more and more of their 
 3   gross trading going into ETFs, and also, a lot of the new 
 4   products aren't index funds anymore, they're smart beta 
 5   funds.   
 6             So, they're actually a little bit more like a 
 7   classic mutual fund.  They probably wouldn't be expected 
 8   to have as high a turnover of the investor base.  They're 
 9   kind of longer-term strategies that are much more focused 
10   to retail investors, but they're providing retail 
11   investors with really cheap access to some interesting 
12   investment strategies, as well. 
13             MR. REDFEARN:  Go ahead, Greg. 
14             MR. SUTTON:  Actually, I agree with what Phil 
15   said on the -- in terms of the length of ownership.  The 
16   other component to be aware of is when we do a creation 
17   basket or redemption basket for a client, we're printing 
18   the stock, the primary stock in the underlying.  We don't 
19   print the ETF again. 
20             So, that volume is lost to the ETF itself.  
21   However, it is reported in the underlying stock. 
22             So, it is a little misleading at times, but 
23   those are the rules that we have right now. 
24             MR. YONES:  I would like to add one point to 
25   that, because again our measures around average daily 
0218
 1   volume are just that.  They're taken, usually, from a 
 2   downstream data vendor that may not collect all trading 
 3   prints for the day. 
 4             So, particularly new ETFs that tend to have, 
 5   you know, a conversational trade with institutional 
 6   volume, many times that will trade principal, maybe 
 7   they'll print in the after-hours market, and most data 
 8   vendors won't pick up that volume in their -- on their 
 9   tape, and so, you may have very large substantial 
10   institutional trades happening in an ETF in its early 
11   days, yet it doesn't show up as overall volume. 
12             MR. REDFEARN:  So, I want to pivot just for a 
13   second to another discussion.   
14             We've talked about the create/redeem process, 
15   Reg M relief, but when we think about the discussion that 
16   we had earlier today for the -- for corporate stocks, I 
17   wanted to ask your views about, if something were to be 
18   done there where we contemplated, you know, exclusive 
19   listings or no UTP for stocks, the question is, when you 
20   think about, in the ETP market, first of all, would that 
21   be something where, if there was a universe included, 
22   that you would definitely want them outside of that 
23   particular program. 
24             And secondly, what would be the impact, 
25   potentially, if any, on the trading for any of these 
0219
 1   underlying, presuming that you have many ETPs that have 
 2   any of these stocks in the actual ETP, and I don't know 
 3   if anybody has looked at that, specifically. 
 4             MR. LaVALLE:  I'll start from the issuer's 
 5   perspective.  I feel like the exchange-rated products, 
 6   you know, being derivatively price instruments, you know, 
 7   will have an impact on their pricing based upon, you 
 8   know, any sort of impact on the underlying constituents. 
 9             So, to the extent that any sort of pilot 
10   program, you know, results in an impact on spread quality 
11   or, you know, liquidity quality of an underlying 
12   constituent, that will also be realized in the ETF 
13   shares, and so, you know, I -- although I haven't looked 
14   at the exact, you know, kind of data driver of that, but 
15   the way that I say it is it could have a potential to be 
16   kind of double taxation. 
17             If, in fact, you have an ETP that's part of a 
18   program and the underlying constituents are part of the 
19   program, you could have an effect where there's kind of a 
20   double result. 
21             MR. REDFEARN:  Or a double benefit. 
22             MR. LaVALLE:  Or a double benefit. 
23             MR. REDFEARN:  Kumar. 
24             MR. VENKATARAMAN:  So, I think some of the 
25   definitions of liquidity that have been described by the 
0220
 1   panels refer to institutional investors, but when you 
 2   think about retail investors and you go back to the 
 3   classic definition of liquidity is the ability to convert 
 4   a security into cash quickly, without a significant price 
 5   concession, and so, from that perspective, given the 
 6   participation in this market, sort of the traditional 
 7   definition of bid/ask spreads and trading volume as 
 8   measures of liquidity, I think, are still applicable in 
 9   the ETF space. 
10             The second observation I have is with respect 
11   to sort of market structure improvements, and this 
12   applies both to the conversations in the second panel on 
13   corporate issues but also with respect to perhaps some 
14   things that can be tried in the ETP space, and it goes 
15   back to this idea that, in many of these thinly traded 
16   securities, sort of the natural flow of buyers and 
17   sellers do not arrive naturally. 
18             So, there's always a flow problem that a market 
19   maker solves, but market maker participation, as I 
20   explained earlier, based on a lot of data analysis that 
21   has been done, tends to be sort of correlated, so there 
22   is fragility, and so, the way you sort of solve t his 
23   problem is by having someone with an affirmative 
24   obligation. 
25             So, many European markets -- and this was 
0221
 1   discussed in the previous panel -- in many European 
 2   markets, we have looked at data where a designated market 
 3   maker has been introduced.   
 4             The introduction of the designated market maker 
 5   is associated with lower trading cost in the form of 
 6   tighter bid/ask spreads, lower transitory volatility, 
 7   improved price discovery, high trading volume. 
 8             So, investors are pleased, because they have 
 9   tighter spreads, and they also have greater confidence 
10   that their liquidity will not vanish, so there's less 
11   fragility  in liquidity supply, but importantly, what we 
12   observe is that, because liquidity is priced around the 
13   introduction of these designated market makers, their 
14   stock price actually goes up by 3, 4, 5 percent around 
15   the time of the introduction.  
16             So, this brings up an interesting compensation 
17   mechanism which I think has been tried in the case of 
18   ETPs many years ago, which is that the right model 
19   perhaps to consider is whether the issuer can, in doing 
20   their liquidity arrangement, facilitate a designated 
21   market maker arrangement, because if you think about the 
22   role of a CFO or a corporate executive, it's to generate 
23   -- it's to invest in positive net present value projects 
24   that increase the stock price, and if you think about 
25   liquidity improvements as that sort of a positive NPV 
0222
 1   investment, then it makes sense for a corporation to 
 2   internalize this liquidity benefit. 
 3             So, it seems to me that a mechanism where the 
 4   corporate issuer can participate in this sort of a 
 5   liquidity -- and it's very common in markets outside the 
 6   U.S. -- would be a arrangement that the Commission could 
 7   consider. 
 8             MR. REDFEARN:  Are you talking specifically 
 9   about where the issuer is paying the market maker, or are 
10   you talking about something else? 
11             MR. VENKATARAMAN:  That's right.  So, that 
12   would be -- and I believe there is at least one of the 
13   regulations that I came across -- it's a FINRA 5250 
14   regulation that prohibits this type of arrangement in the 
15   U.S., but my understanding is that the concern is with 
16   respect to price manipulation, and to the extent that -- 
17   so, in these European markets, there's a very robust 
18   surveillance program, which I'm sure can be introduced in 
19   the U.S., as well. 
20             So, to the extent that the compensation is sort 
21   of distant arrangement, and there is no exchange of 
22   information between the issuer and the designated market 
23   maker, it makes sense for the main participant who 
24   benefits from higher liquidity, which is the issuer or 
25   the shareholders of the company, to actually be involved 
0223
 1   in improving the liquidity of the underlying security. 
 2             So, that's something to consider. 
 3             MR. REDFEARN:  Great.  Thank you. 
 4             Doug, I see your light on there.  I am 
 5   interested in this discussion about the market maker 
 6   discussion and this particular issue. 
 7             MR. YONES:  Yeah.  I thought I would start with 
 8   the first part of Kumar's remarks around a designated 
 9   market maker. 
10             I do feel as though we do have a committed -- 
11   we call them, you know, lead market makers.  Each 
12   exchange has their own program around a lead market 
13   maker, but their roles do require them to participate in 
14   auctions, again, dependent on exchange, require them to 
15   be at the inside a certain level of time, and you know, 
16   with a certain level of spread quoting depth, and we've 
17   seen these programs work, work consistently, and actually 
18   increase market quality over time by the exchanges 
19   continuing to compete and improve against those programs, 
20   which does leave me with some concern, if we were to take 
21   away the ability for exchanges to compete with one 
22   another and revoke UTP, you know, it does raise the 
23   question of, you know, will the programs continue to 
24   evolve the way they have? 
25             I'll leave it up to counsel if you want to 
0224
 1   continue on the 5250 remarks. 
 2             MR. REDFEARN:  Yes.  Anybody have any comments 
 3   on that one? 
 4             MR. LaVALLE:  So, we've spent a lot of time 
 5   thinking about FINRA's Rule 5250, and full disclosure, 
 6   when I was at Nasdaq, I proposed a market quality 
 7   program, which was the first program that achieved SEC 
 8   approval and also an exemption from FINRA Rule 5250.  It 
 9   was never adopted by any issuers.  It fostered a fruitful 
10   conversation, I believe. 
11             In this round of comments around FINRA's Rule 
12   5250, I think there's a couple of questions that we need 
13   to ask.  What problem are we trying to solve?  And I 
14   think there's a couple of different answers to that 
15   question. 
16             The first answer to that question is that 
17   there's 200 products in the marketplace.  The second 
18   answer to that question is oftentimes that the cost 
19   associated with being a market maker in less actively 
20   traded ETPs outweighs the benefits associated with that 
21   obligation. 
22             And the reality is there's a couple of things 
23   that haven't been answered, which is what's the actual 
24   cost going to be, and so, if we actually talk about it, I 
25   bring a $5 million product to market in terms of the 
0225
 1   initial seed capital at 10 basis points that's going to 
 2   generate $5,000 of actual revenue, right?  And I don't 
 3   think if I gave Greg or Josh $5,000 to make a market in a 
 4   less actively ETP they're going to jump at that 
 5   opportunity. 
 6             So, in fact, whatever the cost is, it's higher 
 7   than that, and how is the issuer actually going to fund 
 8   that?  Well, they're going to fund that through their 
 9   management fees. 
10             So, to the extent that we do believe that 
11   payment to a market maker will tighten spreads, which is 
12   something that we don't have a lot of data on, but to the 
13   extent that it will tighten spreads, that will be in 
14   exchange for a higher management fee. 
15             There's the key that we think is really a 
16   problem, is the impact or potential impact on the 
17   resilience of the market, and this is a really important 
18   one. 
19             Currently we have five lead market making  
20   partners, and those five lead market making partners are 
21   critical to the liquidity profile of the products that we 
22   bring to market, and we enjoy those relationships with 
23   our lead market makers, and we appreciate the support 
24   that they offer. 
25             To the extent where we go to a kind of payment 
0226
 1   for market making regime, whatever the number is, it's 
 2   unlikely that we're going to pay each one of our five 
 3   market makers $10.  We're likely to go to a regime where 
 4   we pay one of them $25 and a second $15, whatever. 
 5             The point is, is you're going to have 
 6   concentration of liquidity provision, and there's also 
 7   going to -- you know, put yourself and your products in 
 8   the position where it's unlikely that you're going to get 
 9   additional support from those market makers that were, so 
10   to speak, left to the side. 
11             And additionally, if I'm going to pay a 
12   particular market maker, I might not be that excited for 
13   my payment to also be subsidizing their liquidity 
14   provision in other products. 
15             And so, I think we need to think long and hard 
16   about a pay for market making regime and understand all 
17   the potential unintended consequences, despite the fact 
18   that we do enjoy the support of a lead market making 
19   regime currently. 
20             MS. MORRISON:  So, at CBOE, we're dedicated to 
21   creating unique and different market making programs that 
22   will improve the quality of our market and the price 
23   discovery process for the end investor. 
24             To the extent that we have flexibility in this 
25    area -- for example, we launched -- in addition to the 
0227
 1   lead market making program, which we also have, we have a 
 2   liquidity management provider program, LMP program, that 
 3   essentially attracts a crowd around an LMM so the LMM 
 4   isn't standing there by themselves during times of 
 5   volatility. 
 6             LMP program is more attractive to market 
 7   participants who don't actually want to sign up as a 
 8   market maker, with those additional obligations that come 
 9   along with that, as well as compliance oversight. 
10             So, as I mentioned earlier in my remarks, it 
11   attracts a different group of market makers in and around 
12   each one of the products that we have. 
13             So, that's proven to be successful, and in  
14   addition to that, as it relates to then stepping over to 
15   the Rule 5250, while we're definitely committed to this  
16   offering, we've had issuers, not the issuers at the table 
17   here, but we have had other issuers come to us to say we 
18   enjoy the opportunity to be able to pay our market makers 
19   directly in other countries, and we see it as another 
20   lever to pull. 
21             We're not going to do it for every single one 
22   of our products, and we're not going to do it for every 
23   single one of our market makers, but in an instance, in a 
24   situation where we have a new product that we'd like to 
25   bring to market, spending our money in this area, we 
0228
 1   felt, was the better way to spend it versus, you know, 
 2   all the other reasons why you'd be spending money on 
 3   marketing and distribution. 
 4             So, it's something that hasn't been available 
 5   here in the U.S.  It's something that, with proper 
 6   oversight and engagement, it may be an additional tool, 
 7   if you will, again, not for every single issuer and not 
 8   for every single product that's out there, but an 
 9   additional tool that would help promote, particular as a 
10   product is just coming to market and gaining traction and 
11   maturing, it may help bring that product above the 
12   waterline. 
13             It's not meant, however, on the other hand, to 
14   be the last ditch effort for these extremely less liquid 
15   products in hopes that you can, you know, resuscitate, 
16   but that wasn't the intention of our comment letter or 
17   our advocacy for this. 
18             I just think it's important to make that part 
19   clear. 
20             MR. REDFEARN:  Thank you. 
21             Charles? 
22             MR. THOMAS:  Just a few thoughts on the idea 
23   behind issuers paying for market making services. 
24             First, to reiterate Dave's earlier point, we 
25   would view that as a payment that's a cost incurred by 
0229
 1   the fund.  So, over time, that would show up in our 
 2   expense ratio or the price that investors pay to invest 
 3   in the ETF. 
 4             So, from a philosophical perspective, we would 
 5   think about that being long-term investors that have 
 6   bought and are holding that fund for a long period of 
 7   time, subsidizing via the higher expense ratio short-term 
 8   traders that are coming in and out of that fund by 
 9   compressing the bid/ask spread. 
10             So, on a principled basis, we don't necessarily 
11   agree that that's the right approach for market 
12   structure. 
13             Second, we don't view this as necessarily 
14   having a clear tie to investor outcomes.  If you think 
15   about some sort of fixed issuer payment to market makers, 
16   we would think about that as subsidizing the fixed cost 
17   or the startup costs that a market maker incurs to start 
18   quoting a particular name, and I would be curious to hear 
19   reactions from those folks that are closer to trading 
20   about this, rather than helping at the margin when an 
21   investor is looking to trade a particular security. 
22             So, it's not clear to us that investor outcomes 
23   would be impacted in a significant way. 
24             Laura mentioned a number of different issuers 
25   that have been exploring this.  We actually think there's 
0230
 1   a lot of complications with any potential implementation 
 2   of this, particularly with access to data. 
 3             In order for an issuer to structure these types 
 4   of payments or contracts, you need to be able to 
 5   effectively monitor how your ETFs are trading and then 
 6   have a conversation about enforcement and what the 
 7   obligations of that market maker actually are, and if 
 8   they're fulfilling those obligations. 
 9             So, you need access to data.  It's not clear 
10   that many of the smaller issuers would necessarily be 
11   able to afford that cost. 
12             We view this issue -- and I don't want to go 
13   back to the transaction fee pilot, but we view it as 
14   somewhat related to the discussion around rebates. 
15             So, to us, the best possible outcome for the 
16   marketplace is to have what we would call a clean bid/ask 
17   spread.  In other words, what the investor is paying is 
18   exactly what the market maker is earning in providing 
19   liquidity. 
20             The current model with the rebate structure 
21   actually creates a wedge between that clean bid/ask 
22   spread in terms of what the investor is paying and what 
23   the broker is earning. 
24             An additional payment from an issuer for ETFs 
25   widens that wedge even further.  So, it's not clear as to 
0231
 1   what the cost of liquidity is in the marketplace. 
 2             And then just to quickly touch on the European 
 3   experience, which I think Laura was alluding to, market 
 4   structure in Europe is very different, very fragmented 
 5   across numerous different countries, numerous exchange 
 6   venues.  So, although the payment for market making has 
 7   arisen in Europe, we would say there are good reasons for 
 8   that, given the difference in market structure. 
 9             Given that experience in Europe, we actually 
10   have some concerns about moving in this direction in the 
11   U.S., potentially some unintended consequences of 
12   allowing for payments, particularly with the 
13   consolidation of market making services, where you 
14   introduce a little bit of an anticompetitive environment 
15   that makes it more difficult for new entrants to provide 
16   liquidity. 
17             MR. REDFEARN:  Thank you.  Not that we want to 
18   slip into another transaction fee pilot debate, but was 
19   that a no rebate argument you were making? 
20             MR. THOMAS:  We're very excited about the 
21   transaction fee pilot and very excited about the 
22   inclusion of a no-rebate bucket, so we're excited to see 
23   what the results are. 
24             MR. REDFEARN:  We are just about out of time, 
25   but I just want to open it up for a second, and if 
0232
 1   anybody has any final parting thoughts, the question that 
 2   we have is -- we've heard a few things here today. 
 3             We've heard that we should continue to look at 
 4   Reg M, that there are issues with respect to 
 5   create/redeems and potentially having a smaller size 
 6   there, so we've heard that before, and I'm not surprised 
 7   to hear it today.  We have this issue of payment for 
 8   market making in 5250 and then potentially other market 
 9   making enhancements that might strengthen that. 
10             Let me just -- you know, in one minute or less, 
11   if we could just go and tell us what do you think we 
12   should be focusing on to make the environment for trading 
13   exchange traded products and the liquidity better? 
14             MR. KULKIN:  I think giving issuers the freedom 
15   to have more flexibility around creation/redemption 
16   sizes, I think, is, by far, the most important thing, at 
17   least from our perspective. 
18             I would urge that, if you are going to allow 
19   payments, you know, the idea that they be fair between 
20   market participants, the idea that I want to be the best 
21   market maker, because I'm making the best price, not 
22   because I'm getting paid the most or have the best 
23   relationship with Dave. 
24             So, I would much rather reduced fees or smaller 
25   creation sizes than payments any day. 
0233
 1             MR. REDFEARN:  Thank you. 
 2             Dave? 
 3             MR. LaVALLE:  Revocation of UTP could help.  I 
 4   don't know if there's consensus around that, but what 
 5   there has been consensus across three panels today is 
 6   that a one-size-fits-all approach to anything in the U.S. 
 7   markets is difficult to see being perfect, and so, I 
 8   think we should be equally as careful about any sort of 
 9   pilot program or regime we're going to put in that's also 
10   one size fits all, and recognize that, again, exchange 
11   traded products are 30 cents out of every dollar that's 
12   traded in the U.S. equity  markets, and we need to make 
13   sure that we're paying close attention to that as we put 
14   in new regimes and new changes. 
15             Thank you. 
16             MR. REDFEARN:  Thank you, Dave. 
17             Phil? 
18             MR. MACKINTOSH:  Yeah.  So, I would say there's 
19   about 100 ETFs that trade more than 100 basis points 
20   wide, and when you look at them, they are the thinly 
21   traded and also thinly capitalized ETFs, and it really 
22   does put together what Josh and Dave were saying before. 
23             Your ETF that's going to charge $5,000 of 
24   management fee a year, it's going to be Josh paying that 
25   to you, because he seeded it for you, and that's kind of 
0234
 1   where this whole discussion of payment -- direct payment 
 2   for some of these ETFs comes through. 
 3             At least it would put the economics better in 
 4   line with whether you're going to hold that ETF for the 
 5   longer term, and that's probably something that's easier 
 6   to implement with like the revocation of UTP, just 
 7   because you would be able to focus on one venue, one set 
 8   of rules, maybe a slightly different market structure for 
 9   those ETFs. 
10             MS. MORRISON:  Taking a slightly different 
11   tack, I look forward to the SEC filing the SEC rule on 
12   ETFs, hopefully this year. 
13             I think that's going to do a lot for promoting 
14   competition within the issuance space and level the 
15   playing field of exemptive reliefs that are out there. 
16             So, that is something that is going to make a 
17   difference this year. 
18             MR. REDFEARN:  Great.  Thank you. 
19             Greg? 
20             MR. SUTTON:  Again, what Josh said on the 
21   creation/redemption size obviously is an important 
22   component of it.  I think we support suspending UTP, as 
23   well, but we definitely are adamant around dark trading 
24   volume continuing to remain for ETFs. 
25             I mean, in the illiquid ETF space right now, 
0235
 1   over 60 percent of the volume comes in the dark. 
 2             So, it's obviously a very significant component 
 3   of it and needs to be -- that needs to be understood. 
 4             MR. REDFEARN:  Thank you. 
 5             Charles? 
 6             MR. THOMAS:  With respect to ETFs, I'll end 
 7   where I started. 
 8             So, ETF liquidity is a little bit different to 
 9   measure than for common stocks, so any proposal should 
10   hopefully acknowledge that and address that. 
11             If we think about the revocation of UTP, 
12   relaxing Reg NMS, which was proposed earlier, or even 
13   repealing FINRA Rule 5250, we would view all of those as 
14   fairly significant changes to current market structure, a 
15   market structure that's functioning reasonably well, so 
16   we would want to be thoughtful about evaluating any 
17   potential changes there. 
18             With respect to ETFs, some of the conversation 
19   around Reg M providing better, more efficient, or lower 
20   cost access to the ETF primary market, where liquidity is 
21   very, very strong, we would suggest that would be a good 
22   place to start and probably less impactful to the broader 
23   market structure. 
24             MR. REDFEARN:  Thank you. 
25             Kumar? 
0236
 1             MR. VENKATARAMAN:  Just a last observation on 
 2   designated market makers. 
 3             If we want to pursue a designated market maker 
 4   model, then it's important to think about compensation, 
 5   because by definition, a designated market maker 
 6   participates when others are not, it's not profitable to 
 7   do so.  So, you have to think about that. 
 8             And so, in that context, I feel the issuer pay 
 9   model makes sense, because the people who benefit from 
10   better liquidity is the issuing firm, and so, it's simply 
11   a way for the issuing firm to invest in a positive net 
12   present value project, which is investment in liquidity. 
13             I think in a fragmented market structure, it's 
14   difficult for an individual exchange, if we go with an 
15   exchange pay model to provide a market maker with 
16   sufficient incentives to accept meaningful quoting and 
17   trading obligations. 
18             MR. REDFEARN:  Thank you. 
19             MR. YONES:  In closing, we do not think 
20   reducing the number of lit markets would help raise 
21   liquidity for low-volume ETFs. 
22             If the goal is to increase -- if the goal is to 
23   reduce fragmentation, then we would say the Commission 
24   should look at all trading venues, and that would include 
25   everything over the counter, because as mentioned, you 
0237
 1   know, the numbers are over 60 percent of the low-volume 
 2   ETF as trading is actually over the counter. 
 3             If the Commission does decide to move forward 
 4   with the proposal, we would highly recommend offering an 
 5   opt-in/opt-out choice for issuers, either before or also 
 6   during the pilot, depending on the overall results, and I 
 7   think the point you said, is there anything else that 
 8   could be done, I think we could continue to look at some 
 9   of the various  market maker rules, as Ari had mentioned 
10   in the previous session, whether it be around exemption 
11   for, you know, short sell trades for hedging, Rule 204 
12   obligations, etcetera.  So, we could host another session 
13   around market making exemptions. 
14             MR. REDFEARN:  Fantastic. 
15             Well, listen, thank you all very much.  It's 
16   been a very illuminating discussion today.  It's always 
17   good to be able to make sure that, as we consider these 
18   things, we think about and consider the unique aspects of 
19   exchange traded products and funds. 
20             So, thank you all very much for your time 
21   today.  We very much appreciate it. 
22             (Whereupon, at 3:32 p.m., the meeting was 
23   concluded.) 
24                           * * * * * 
25              
0238
 1               PROOFREADER'S CERTIFICATE 
 2   
 3   In the Matter of:  ROUNDTABLE ON MARKET STRUCTURE FOR 
 4                      THINLY-TRADED SECURITIES 
 5   File number:  OS-0423 
 6   Date:         Monday, April 28, 2018 
 7   Location:     Washington, D.C. 
 8    
 9        This is to certify that I, Christine Boyce, (the 
10   undersigned) do hereby swear and affirm that the attached 
11   proceedings before the U.S. Securities and Exchange 
12   Commission were held according to the record, and that 
13   this is the original, complete, true and accurate 
14   transcript, which has been compared with the reporting or 
15   recording accomplished at the hearing. 
16    
17    
18   __________________________________   _________________ 
19   (Proofreader's Name)                    (Date) 
20    
21    
22    
23    
24    
25