0001 1 U.S. SECURITIES AND EXCHANGE COMMISSION 2 3 4 5 6 7 EQUITY MARKET STRUCTURE 8 ADVISORY COMMITTEE 9 10 11 12 13 14 Tuesday, August 2, 2016 15 9:30 a.m. 16 17 18 19 20 21 22 23 U.S. Securities and Exchange Commission 24 100 F Street, N.E., Washington, D.C. 25 Multipurpose Room 0002 1 PARTICIPANTS: 2 3 Mary Jo White, SEC Chair 4 Michael S. Piwowar, Commissioner 5 Kara M. Stein, Commissioner 6 7 Mark Flannery, Director, Division of Economic and Risk 8 Analysis 9 Gary Goldsholle, Deputy Director, Division of Trading and 10 Markets 11 John Roeser, Associate Director, Division of Trading and 12 Markets 13 David Shillman, Associate Director, Division of Trading 14 and Markets 15 16 Matthew Andresen 17 Reggie Browne 18 Kevin Cronin 19 Brad Katsuyama 20 Richard Ketchum 21 Manisha Kimmel 22 Mehmet Kinak 23 Andrew Lo 24 Joseph Mecane 25 Eric Noll 0003 1 PARTICIPANTS (CONT.): 2 3 Maureen O'Hara 4 Joe Ratterman 5 Nancy Smith 6 Chester Spatt 7 Gary Stone 8 9 PANELISTS: 10 Bill Alpert, Senior Editor, Barron's 11 James Angel, Associate Professor, Georgetown University 12 McDonough School of Business 13 Stacey Cunningham, Chief Operating Officer, NYSE Group 14 Frank Hatheway, Senior VP and Chief Economist, NASDAQ 15 Chris Nagy, Director, Healthy Markets Association 16 Venu Palaparthi, Senior VP, Compliance, Regulatory and 17 Government Affairs, Virtu Financial 18 John Zecca, Senior VP, MarketWatch, NASDAQ 19 20 21 22 23 24 25 0004 1 C O N T E N T S 2 3 PAGE: 4 Welcoming Remarks 5 5 6 Market Quality Subcommittee 5 7 8 Customer Issues Subcommittee 86 9 10 Regulation NMS Subcommittee 148 11 12 Trading Venues Regulation Subcommittee 150 13 14 15 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. SHILLMAN: Okay, good morning. Thank you 3 all for joining us for today's meeting of the SEC Equity 4 Market Structure Advisory Committee. I believe we have a 5 quorum so I will call the meeting to order. 6 My understanding is Reggie Brown is 7 participating by telephone, so I would like to confirm 8 Reggie is there. 9 MR. BROWNE: I am, good morning. 10 MR. SHILLMAN: So why don't we kick things off 11 by having Chair White make her opening remarks. 12 CHAIR WHITE: Thank you, David. And thank you, 13 everybody. I think everyone is here in this August 14 meeting, so I appreciate that very much. And welcome to 15 everyone, members, panelists, for your continued hard 16 work. You've already had a busy summer, which is great. 17 Just last month, you made recommendations to 18 the Commission concerning an access fee pilot and trading 19 venue regulation reforms. I expect that the Commission 20 will consider a recommendation regarding an access fee 21 pilot later this year. I have also asked the Staff just 22 to update you to evaluate the trading venues 23 recommendations, and they're exploring possible 24 rulemakings to enhance the NMS plan governance process. 25 Today's meeting features proposed 0006 1 recommendations from the Customer Issues and Market 2 Quality Subcommittees for the full committee to consider. 3 These recommendations build on the updates provided to 4 the committee by these subcommittees at your April 5 meeting. 6 The Market Quality Subcommittee's proposed 7 recommendations focus on three topics. The national 8 market system plan to address extraordinary volatility, 9 commonly referred to as the Limit Up/Limit Down plan; 10 market-wide circuit breakers; and the market opening. 11 The subcommittee's work in this space complements the 12 work the Commission and the SROs have been doing over the 13 past several years on equity market volatility 14 moderators. I've been asked just to quickly summarize 15 those efforts to help put the subcommittee's 16 recommendations in context. 17 The Commission and SRO regulatory efforts have 18 focused on the unique issues that arise from electronic 19 trading. While advancements in trading technology have 20 obviously provided demonstrable and significant benefits 21 to investors in the markets, they also pose certain 22 risks, including unchecked market volatility. 23 In our modern markets, electronic order books 24 are used to aggregate trading interests and market 25 participants use algorithms and other automated 0007 1 technologies to submit and cancel orders with great 2 speed. This method of interaction in the market can lead 3 to sudden liquidity imbalances and initial price moves 4 which substantially reverse after liquidity balances are 5 restored. We have seen this many times, both market wide 6 and at the individual stock level, and this type of 7 volatility has occurred in nonequity markets as well. 8 Events such as the Flash Crash of 2010, the Treasury 9 market volatility of October 15, 2014, and August 24, 10 2015, have highlighted the volatility risks posed by our 11 modern markets, which require regulators to consider 12 measures to mitigate those risks. 13 Our equity markets currently rely on two 14 primary volatility moderators, market-wide circuit 15 breakers and Limit Up/Limit Down. These mechanisms were 16 created and modified in an attempt to address concerns 17 that were highlighted by the equity Flash Crash of May 6, 18 2010. 19 The late-day market volatility on May 6, 2010, 20 did not trigger the then-existing market-wide circuit 21 breakers, which at the time were set at 10, 20 and 30 22 percent. In the wake of that event, close consideration 23 to the adequacy of market-wide circuit breakers was given 24 by the SROs, regulators and the joint SEC-CFTC Advisory 25 Committee on Emerging Regulatory Issues. That scrutiny 0008 1 led to several changes. 2 First, the reference index used for measuring 3 market decline was changed from the Dow Jones Industrial 4 Average to the S&P 500 Index. This change was made 5 because the S&P 500 Index, as you know, is comprised of a 6 broader group of securities and was seen as providing a 7 better measure for market-wide volatility. The S&P 500 8 Index was also considered to be a better measure of 9 cross-market volatility, as it is closely correlated to 10 financial derivatives products such as the E-Mini and 11 SPY. 12 Second, the trading halt duration was shortened 13 from 30, 60 or 120 minutes to 15 minutes. And finally, 14 the market decline thresholds were reduced from 10, 20 15 and 30 percent to 7, 13 and 20 percent respectively, as 16 the prior circuit breaker levels had shown during the 17 flash crash that they were not adequate to safeguard 18 against unusual volatility, the prior circuit breaker 19 levels. The lower thresholds and reduced trading halts 20 duration were designed to balance the need to halt 21 trading for significant declines, while also minimizing 22 any disruption resulting from a trading halt. 23 Today, this committee will consider potential 24 recommendations to further refine our market-wide circuit 25 breakers. These recommendations arise out of the Market 0009 1 Quality Subcommittee's thoughtful assessments of the 2 events that occurred a year ago on August 24, 2015, and 3 reflect concerns that a market-wide halt in trading would 4 have exacerbated that particular volatility event. 5 I am keenly interested in the views of 6 committee members and our expert panelists concerning the 7 proposed refinements. A challenge with any market-wide 8 mechanism like these circuit breakers, as you know, is 9 that it needs to be optimized for a variety of different 10 market conditions and different volatility events. May 11 6, a late-day volatility event, failed to trigger these 12 circuit breakers, leading to a subsequent narrowing of 13 the thresholds. Trading on August 24, a volatility event 14 at the market open, would have triggered the circuit 15 breakers, had stocks in the S&P 500 Index opened in a 16 timely fashion, potentially exacerbating volatility that 17 day. 18 Some questions the committee may want to 19 consider include: How can market-wide circuit breakers 20 be optimized, given the range of volatility events that 21 can occur in our modern markets? What considerations 22 should be balanced and how might those considerations 23 weigh in favor of one version of a circuit breaker over 24 another in different market conditions? Would widening 25 the trading bands or using futures market data cause 0010 1 foreseeable issues for other volatility events or market 2 conditions? 3 The committee will also consider potential 4 recommendations on refinements to Limit Up/Limit Down. 5 The committee's consideration of these recommendations is 6 particularly timely. Limit Up/Limit Down has been in 7 effect on a pilot basis for several years, to allow the 8 Commission and the SROs to evaluate its operation and 9 whether it has achieved its objectives of dampening 10 extraordinary market volatility and reducing erroneous 11 trades. Extensive analysis related to Limit Up/Limit 12 Down has already been done and further analysis continues 13 that will inform our collective assessment of Limit 14 Up/Limit Down for the future. 15 As you know, Commission Staff has published 16 research papers on the events of August 24, evaluating 17 the operation of Limit Up/Limit Down, and the specific 18 reasons behind the significant trading pauses that 19 occurred on that day. Among other things, that analysis 20 showed that many of the pauses occurred as a result of 21 prices moving back to their original trading range after 22 the initial significant declines. One recommendation 23 being considered today is intended to address those 24 particular pause events. I am interested in the 25 discussion around this recommendation, including how to 0011 1 determine the appropriate price a stock could revert to, 2 without triggering Limit Up/Limit Down halts. 3 As the Committee knows, the SROs continue to 4 evaluate Limit Up/Limit Down and potential improvements 5 that may be warranted, based on observations regarding 6 how effectively Limit Up/Limit Down has operated thus 7 far, including obviously on August 24, 2015. That work, 8 among other things, is considering harmonization of 9 current clearly erroneous execution rules with Limit 10 Up/Limit Down, and the advisability of coordinated 11 reopening procedures. I am glad to see that these two 12 important topics are among the subjects of the 13 recommendations being considered by the committee today. 14 I am very interested in committee member and 15 panelist views regarding the potential limit state and 16 trading halt modifications that the subcommittee has 17 recommended. The Limit Up/Limit Down analysis done to 18 date suggests that reopening auctions may not be working 19 as well as intended. Unlike the opening and closing 20 auction mechanisms that are successful at aggregating 21 liquidity, it appears that market participants have not 22 been participating at the same rate in the Limit Up/Limit 23 Down reopening auctions. 24 In addition to hearing more detail about how 25 these potential alternative processes would work, I look 0012 1 forward to hearing opinions from participants today about 2 whether the market may be solving these particular issues 3 on its own. In other words, as market participants 4 continue to gain experience with Limit Up/Limit Down in 5 their own trading protocols during specific volatility 6 events, such as August 24, are they changing their 7 behavior and becoming more willing to participate in 8 these reopening auctions? 9 This afternoon, the committee will consider 10 recommendations from the Customer Issues Subcommittee. 11 These recommendations touch on two topics. The first 12 relates to the creation by the SEC of what I will call an 13 investor sentiment benchmark. The second relates to 14 additional enhancements to Commission Rules 605 and 606. 15 Again, as you know, the Commission recently 16 proposed amendments to Rule 605 and 606, focusing on 17 customer-specific institutional order routing disclosures 18 and targeted revisions to the retail disclosure 19 provisions of Rule 605. I welcome this committee's input 20 on further improvements to these important disclosure 21 rules. 22 The Customer Issues Subcommittee's focus 23 complements the Commission's proposal, as the 24 subcommittee's proposed recommendations primarily address 25 potential changes that go beyond the scope of the 0013 1 Commission's recent proposal. Today's recommendations, 2 for example, focus on broader improvements to the 3 existing retail disclosures, including expanding 605 4 public reporting to broker-dealers, expanding 606 public 5 reporting to exchanges and ATSs, and including additional 6 reporting mechanisms in the existing 605 reports. 7 The Market Quality subcommittee has noted that 8 it will provide specific comments on the Commission's 9 recent proposal, and today's recommendations before the 10 committee also touch on several discrete issues the 11 Commission requested comment on in its recently proposed 12 606 amendments. These include dividing 606 data by S&P 13 500 Index stocks and other stocks, rather than listing 14 market, including a new section for over-the-counter 15 equity data in the 606 reports, and improving the 16 accessibility of 605 and 606 reports. Again, I look 17 forward to the committee member and panelist views on 18 these proposed recommendations, as well as on the 19 proposal for the SEC to monitor individual investor 20 sentiment to help inform really all of our ongoing 21 regulatory efforts. Investor confidence in our markets 22 is obviously of utmost importance, and we should strive 23 continuously to seek new and innovative ways to receive 24 their full input. 25 So let me just close by saying that your work 0014 1 today again demonstrates that this committee continues 2 really to march through and tackle some of the most 3 complicated and important issues facing our equity market 4 structure today. So thank you very much for your work 5 and your continuing commitment. Thank you. 6 MR. SHILLMAN: Thank you, Chair White. 7 I guess before we move on, if I could ask 8 Reggie to please mute your phone? You may have done so 9 already, but we have very sensitive microphones, and I 10 think that would help. Thanks. 11 Commissioner Piwowar, would you like to make an 12 opening statement? 13 COMMISSIONER PIWOWAR: Sure. Thank you. Thank 14 you, Dave. 15 Good morning and welcome. Let me start by 16 thanking the members of the committee, especially those 17 serving on the Market Quality and Customer Issues 18 Subcommittees for all of your efforts. It's obvious that 19 the recommendations being presented today are the result 20 of considerable thought and analysis. 21 At earlier meetings, I have encouraged the 22 committee to follow its own agenda and establish for 23 itself goals and metrics for achieving success. I hope 24 you will continue to do so, and I have no intention of 25 interfering. 0015 1 I do, however, want to raise an issue with 2 respect to the framing of one of today's discussion 3 topics. On the committee's webpage, the Customer Issue 4 Subcommittee's stated objective is to "Consider 5 initiatives to protect investor interest and promote 6 investor confidence." In furtherance of that endeavor, 7 one of the subcommittee's recommendations to be 8 considered and discussed at today's meeting is that the 9 SEC benchmark and monitor so-called investor confidence 10 in U.S. equity market structure to inform rulemaking and 11 educational efforts. 12 I have previously challenged the notion that 13 the Commission should promote investor confidence, 14 because it is not part of our core mission and it is a 15 nebulous concept. And, most importantly, focusing on 16 investor confidence creates the wrong incentives. Quite 17 the opposite of promoting investor confidence, I believe 18 the Commission should be encouraging a healthy dose of 19 investor skepticism, in order to fulfill our investor 20 protection mandate. 21 We should urge investors to ask questions 22 about, for example, how their orders are treated and 23 whether that is consistent with their personal investment 24 goals and decisions. Self-reliance in situations where 25 investors are empowered with material information can be 0016 1 a more powerful protection than government dependence. 2 Nonetheless, the subcommittee's recommendation 3 regarding investor confidence has embedded within it 4 concepts which I have been a great proponent. 5 Specifically, it is suggested the SEC test the usability, 6 clarity and effectiveness of disclosures on investors and 7 that investor educational materials themselves be tested 8 and revised accordingly. I completely agree. 9 In my first public speech as a commissioner, 10 for example, I called on the Commission to commit itself 11 to a robust investor testing program that examines the 12 efficacy of not only proposed rules but our existing 13 rules as well. In my mind, the investor testing 14 contemplated by the subcommittee's recommendation would 15 be valuable not because it could improve so-called 16 investor confidence, but because it could facilitate 17 investors making informed choices about investment 18 venues, strategies and products, such as a key to the 19 Commission fulfilling our investor protection mandate. 20 Again, I congratulate the committee for its achievements 21 to date and I look forward to what comes next. 22 Finally, I extend a special thank you to the 23 distinguished panelists and the SEC Staff for your 24 contributions to today's event. Thank you. 25 MR. SHILLMAN: Thank you, Commissioner Piwowar. 0017 1 I would now like to introduce my colleagues in 2 the Division of Trading and Markets. To my right is Gary 3 Goldsholle, Deputy Director of the Division. At the 4 other end of the table is John Roeser, Associate Director 5 in the Office of Market Supervision. And next to John is 6 Mark Flannery, the Chief Economist and Director of the 7 Division of Economic and Risk Analysis. 8 Standard disclaimers apply to today's meeting. 9 Any views expressed by the Staff in this forum are ours 10 alone and cannot be attributed to the Commission or to 11 the commissioners. 12 The last two advisory committee meetings 13 focused on the recommendations of the Regulation NMS and 14 the Trading Venues Regulation Subcommittees. As a result 15 of those efforts, the full committee has now made formal 16 recommendations to the Commission with respect to an 17 access fee pilot, as well as changes to NMS plan 18 governance and related matters. 19 Today, we turn to the work of the other two 20 committees, subcommittees, the Market Quality 21 Subcommittee and the Customer Issues Subcommittee, which 22 will be presenting their preliminary recommendations to 23 the full committee. Much work has gone into developing 24 the subcommittee recommendations being considered today. 25 As you may recall, at the February advisory 0018 1 committee meeting, Commission Staff presented a research 2 note on the August 24, 2015, equity market volatility, 3 which included analyses relevant to the operation of the 4 Limit Up/Limit Down Plan. Commission Staff also 5 presented a memorandum addressing certain issues 6 affecting customers in the current equity market 7 structure. 8 At that same meeting, expert panelists 9 representing a range of perspectives discussed their 10 views on the issues and participated in open dialogue 11 with committee members, the chair, commissioners and 12 Commission Staff. Since that time, the subcommittees 13 have held a number of meetings and have invited outside 14 experts to provide their views on the various issues 15 under consideration. 16 Today, we will begin by asking the subcommittee 17 chairs to introduce their recommendations. We will then 18 hear the views of our expert panelists before turning it 19 over to the full committee and commissioners for 20 discussion. 21 So let's move on to the first agenda items, the 22 recommendations of the Market Quality Subcommittee. And 23 I would like to ask Eric, as the chair of that 24 subcommittee, for an overview of those recommendations. 25 MR. NOLL: Thank you, Dave. Thank you, Chair 0019 1 White, Commissioner Piwowar. 2 As Dave mentioned, our subcommittee really 3 considered changes to Limit Up/Limit Down, market-wide 4 circuit breakers and the market opening for our first 5 three recommendations. 6 Within Limit Up/Limit Down, we really looked at 7 three areas that we think are important that we think 8 there might be some room for modification in the Limit 9 Up/Limit Down process. The first one was really a 10 recognition that the auction reopening process following 11 a Limit Up/Limit Down halt has not really worked since 12 inception. So the issues that we saw, particularly on 13 August 24, which was many halts and reopening of 14 securities on little or no volume at each reopening, and 15 then subsequently triggering a following halt almost 16 immediately thereafter. And so as we looked at this and 17 we tried to consider, you know, what we should do about 18 this, we came rapidly to the conclusion that the auction 19 itself probably was not fixable. 20 We agreed and believe that the Limit Up/Limit 21 Down mechanism in a very macro sense performs the role 22 for which it was intended, which it prevents stock from 23 runaway prices. And the mechanism actually does that 24 fairly effectively. But what it doesn't do, it doesn't 25 reopen the securities in a way to discover the new fair 0020 1 trading price. And there are a number of reasons why 2 that might be that we considered. The first and most 3 obvious one is during the trading day, orders are 4 dispersed across multiple venues, multiple exchanges, 5 multiple market makers and ATSs. And in order to get 6 those orders back onto a primary exchange for an auction 7 seems to be an onerous burden that very few firms or 8 people in the marketplace choose to take on. 9 Another obvious reason is that most people's 10 electronic algorithmic execution mechanisms simply stop 11 working during an auction process, when an auction is 12 called intraday. So they don't participate in the 13 auction, they just wait until the stock reopens. And so 14 when you -- when you get that kind of condition, you just 15 don't have the level of participation to discover prices 16 in the way that you would hope. 17 A factor that recently came to my attention as 18 well is, in other markets like the London Stock Exchange, 19 for example, who have tried to pioneer a midday auction 20 in a much less fragmented market than our own has very 21 little participation, again, intraday in their auction 22 process. That may develop over time but historically, 23 intraday auctions have not seen a great deal of success 24 in any market, including the United States. 25 So what our proposal has been is really to 0021 1 acknowledge that the Limit Up/Limit Down process works, 2 again in a macro sense. And that when we hit a limit 3 condition, that limit condition be extended for two 4 minutes, at which free trading can take place around that 5 limit condition, with inside the bands. But if it 6 remains, if you will, stuck on that limit condition, 7 trading -- order matching does not take place at that 8 point. But what does take place is order imbalances and 9 information about orders are distributed into the 10 marketplace. But no orders are canceled, no auction 11 process takes place. And then after that, the stock 12 would reopen with a new adjusted band around that closing 13 price. So the bands would, if you will, slide with the 14 limit conditions. That's very similar to the way the 15 futures market works, and the Limit Up/Limit Down 16 conditions that work in the futures market. 17 We think this would do two things. One, it 18 would preserve the ability to stop runaway stocks. It 19 would avoid the problems of halting securities from 20 trading and forcing people into reopening auction where 21 we are not successfully getting to a good reopening 22 price. And it would not force any market participants to 23 change any of their behavior. They won't have to cancel 24 any orders, they won't have to do anything different. 25 They just have to participate when the markets reopen. 0022 1 And so we think that that's perhaps the -- while not 2 perfect, the best cure we can get for one of the problems 3 that we saw, particularly on August 24, but more 4 generally that happens on a frequent basis during the 5 day, every day. 6 The other two Limit Up/Limit Down 7 recommendations that we would like the Commission to 8 consider is one, and I think this one is already under 9 way, which is conforming clearly erroneous rules to Limit 10 Up/Limit Down bands. This again was a problem that 11 developed on August 24, where many market makers or 12 market participants were unclear as to whether trades 13 that they made either inside or outside the bands would 14 stand following the events of August 24, and caused many 15 of them to withdraw from the marketplace because of that 16 uncertainty. And we think the easiest way to make sure 17 that that doesn't occur is to make sure that those bands 18 are clearly aligned with the Limit Up -- the clearly 19 erroneous rules are clearly aligned with the Limit 20 Up/Limit Down bands. And again, I think that's probably 21 under way. 22 The last one is a little more, perhaps, 23 complicated to put into place. But it's this idea of 24 mean reversion around the Limit Up/Limit Down bands. 25 I've heard people describe it as making them symmetrical. 0023 1 I don't think that's quite the terminology we would use 2 here. 3 But one of the things that was clearly noticed 4 on August 24, and as also an observer of the way the 5 markets work on a regular basis, there are often halts on 6 the way up and on the way down or on the way down and on 7 the way up, where you have prices that are triggering 8 halts. But because the bands adjust based on that last 9 trade, you could trigger a stock on the way down and then 10 immediately it wants to return to the price from which it 11 came and will trigger a halt on the way back up. And so 12 we'd like the Commission to consider the idea of 13 embedding a concept of mean reversion into how the bands 14 get set. So an example would be if a stock is trading 15 for 100 and it trades down to 90 and triggers a band, 16 that it can trade back to 100 without triggering the band 17 at 98 and causing -- or 99 and causing a halt back when 18 it was heading back to where it was trading before. And 19 that's to catch what we would call anomalous trades that 20 are bouncing around a price, where there really is a, if 21 you will -- a fair trading price is probably 100 but the 22 stock is bouncing around that, maybe in a volatile way, 23 but without interrupting that with a halt or a limit 24 condition. 25 The other recommendations we made and spent a 0024 1 lot of time considering were -- one was on market-wide 2 circuit breakers. As Chair White suggested in her 3 remarks, on the morning of August 24, had all stocks been 4 open in the S&P 500, we would have triggered a circuit 5 breaker at the 7 percent level. We spent a lot of time 6 talking about this and I think there's still some 7 discomfort among the committee -- subcommittee about 8 this, and in the market in general around this. But the 9 prevailing view across everybody we spoke to who we 10 invited into the committee meetings and among the 11 committee members, was the concept of thank God we didn't 12 have a halt that day. 13 And one of the things that that immediately 14 suggested to most of us, if that's really the case, that 15 does suggest that the bands are too tight. That if the 16 concept of most market participants, and this was fairly 17 uniform, there were very few people who said the 18 opposite. That we should thank God we didn't halt, that 19 suggests that the benefits of that tighter range are 20 probably not material and that people would like a little 21 more latitude in how the index is triggered market-wide 22 halts. 23 That being said, we agree that there should be 24 market-wide circuit breakers. There was some concept of 25 trying to figure out whether declines in the index could 0025 1 be predicated on some sort of qualitative measure. We 2 could not reach any sort of agreement about how -- one, 3 who would do that and, two, what those qualitative 4 measures would be that this would make sense. So at the 5 end of the day, we decided that widening those bands out 6 would probably make the most sense. 7 There was some conversation about using the S&P 8 futures values, but there was not agreement among the 9 subcommittee that that was a viable choice. People still 10 thought we should use, at the end of the day, the cash 11 values. But I did raise it, because it was a topic that 12 came up with some vigor. 13 And then the last topic that we considered and 14 that we want to make a recommendation around is around 15 market opening. And again, I think there has been a lot 16 of movement around this already from the SROs and from 17 other market participants. But I think our -- our firm 18 recommendation without commenting on how a market opens 19 or what anyone's business model is or should be, that 20 markets should open at the same time, and they should 21 open as close to 9:30 as possible. And that any kind of 22 mechanism that delays that market opening creates 23 uncertainty and volatility in the marketplace, and that's 24 not to the benefit of any market participants. 25 And that's our report. 0026 1 MR. SHILLMAN: Okay, thanks very much, Eric. 2 Before we move on to the panelists, I'd like to 3 invite Commissioner Stein to make opening remarks. 4 COMMISSIONER STEIN: I will try to keep my 5 remarks really brief. I just want to welcome everyone 6 again. I -- I think my fellow commissioners are very 7 grateful that the pro bono time that you're giving to the 8 Commission to help inform our decisions on market 9 structure -- I started to hear what Eric was saying about 10 the Market Quality Subcommittee's report and the 11 recommendations on Limit Up/Limit Down. I've said this 12 before. I think we're all aware the events on August 24 13 highlighted the need for us to refine our national market 14 system plan to address extraordinary volatility, which is 15 another name for our Limit Up/Limit Down process, and how 16 we deal with that. 17 As we know, on August 24, there were over 1,200 18 trading halts. And since then, we have had a great deal 19 of discussion regarding what caused the trading halts and 20 how to best handle extraordinary volatility in the 21 future. Clearly, this is a topic of concern to everyone 22 in this room. And the subcommittee has the two 23 recommendations that you started to talk about, Eric, the 24 first being a traditional -- an alternative to the 25 traditional halt, which is sort of intriguing. You know, 0027 1 under this recommendation, a stock stuck in limit up or 2 limit down would not be subject to a traditional halt but 3 instead enter a two-minute monitoring period before 4 moving to a pre-open state. 5 The second recommendation is to allow stocks to 6 trade back up to their original price without triggering 7 halts. So I do look forward to the discussion of the 8 pros and cons of both these recommendations. 9 I like you have also consistently heard that 10 the opening time should be as close to 9:30 as possible. 11 That is in line with what I have been hearing from every 12 single market participant I have talked to. 13 In addition to the recommendations on Limit 14 Up/Limit Down, I note that the Market Quality 15 Subcommittee's report states that the subcommittee is 16 going to suspend future discussion of market quality for 17 small to mid-size companies. And I just want to reflect 18 back to you that this concerns m. Small to mid-sized 19 companies need access to capital to survive, grow and 20 create jobs. And market quality and liquidity are 21 critically important factors in how these companies 22 access that capital. 23 The recent regulatory initiatives we've 24 embarked on, such as Reg A-plus and Reg Crowdfunding, 25 stand to benefit from future study and discussion about 0028 1 the factors affecting market quality and liquidity for 2 small to mid-size companies. Failure to address issues 3 related to market quality for these company securities 4 could limit capital formation and affect their ability to 5 grow. 6 The members of this entire committee are 7 uniquely qualified to understand the cost of illiquid 8 markets to these companies and to propose potential 9 solutions. I hope that you will reconsider and continue 10 to look at these issues for the good not only of the 11 small to mid-size companies, but also for the good of our 12 entire marketplace and our entire economy. 13 I am also looking forward to hearing the 14 recommendations of the customer issues subcommittee. The 15 recommendations for monitoring investor confidence in the 16 United States equities market structure intrigues me. 17 However, as we all know, investors must trust the 18 markets. If they do not, they won't invest their capital 19 and businesses will find it harder to thrive. Thus, it 20 is important that we understand what market structure 21 attributes contribute to increased investor confidence. 22 So I think it's a really interesting idea and I look 23 forward to hearing today's discussion. Thank you. 24 MR. NOLL: Dave, if I could perhaps explain our 25 comments about the small and mid-cap stocks to 0029 1 Commissioner Stein and other people? 2 It was our belief when we held our meetings 3 that there might be some market structure things that we 4 could change or suggest to the full Commission that might 5 make discovery of liquidity better for small and mid-cap 6 stocks. We invited a whole host of small-cap managers, 7 people from OTCBB, the TMX venture market, other people 8 to come and speak to our subcommittee. 9 And I will say that I was shocked and I think 10 others were shocked that not a single person that came to 11 that meeting had a viable suggestion about changes to our 12 current market that would suggest a better liquidity 13 environment for small and mid-cap stocks. Most of their 14 comments were very much focused on the regulatory 15 environment for those companies to go public. The Reg A 16 things, Sarbanes-Oxley, other issues around the 17 challenges of a small company to make itself go public. 18 And almost no one had any what we would consider to be 19 interesting suggestions around what it would mean to have 20 liquidity in a small-cap name. 21 And so at the end of those conversations, we 22 did consider whether we could think about an industrywide 23 standard for market making. We believed and I think 24 continue to believe that those are probably left in the 25 hands of the SROs and that should not be a mandated 0030 1 function on how to incent liquidity in a market maker 2 system. But that different people could come up with 3 different business models that could do that. But there 4 wasn't a fix or a set of fixes that we felt comfortable 5 recommending. 6 And I would say, and I would certainly defer to 7 some of the buy-side-only people here on the panel, but I 8 think we were almost all shocked at how adamant the small 9 cap managers were of saying, there isn't anything that 10 you can do about this, we're not making any suggestions, 11 we're much more concerned about the way companies get 12 public and their issues about getting public than we are 13 about the way they trade. Which we thought was 14 interesting. 15 COMMISSIONER STEIN: I'm just wondering if it 16 might make sense to talk to our small business and 17 emerging growth company advisory committee or investor 18 advisory committee to sort of talk about folks who are 19 not already, you know, I'll say -- it's easy to say 20 perhaps on Wall Street, but are struggling with this more 21 regionally, or folks in Silicon Valley. 22 But it might be we need to reach out to a 23 different group of folks, you know, than are already in 24 the marketplace. I will just suggest that that might be 25 fruitful. 0031 1 MS. O'HARA: I just wanted to emphasize sort of 2 the point that Eric was making. I think everyone on our 3 committee felt strongly that the issues of illiquidity 4 for small stocks is very important. The question is, is 5 there a market structure solution? And that is what we 6 were spending our time looking for. And I think one of 7 the things that we found particularly interesting was 8 looking at how other, you know, venues, other countries, 9 have tried to find market structure solutions. And the 10 kind of scary history of suggesting that most of those, 11 you know, separate markets, separate structures, have not 12 worked well. That when you look at sort of how those 13 companies have traded, they have not -- their challenges 14 appear equal if not similar to the ones we're dealing 15 with in our market. 16 So I think again, just to underscore what the 17 committee felt, we had hoped that we would come up with 18 market structure changes and perhaps we have to cast a 19 wider net. But the net we cast was pretty wide and 20 unfortunately nothing, you know, apparent in our market 21 structures have emerged. 22 COMMISSIONER STEIN: So, Maureen, was this more 23 a question of folks needing to become public and have 24 that fulsome and robust disclosure before investors were 25 willing to help make -- liquify those markets? I mean, 0032 1 do you have any sense of what that issue is? 2 MS. O'HARA: you know, I think it's a really 3 tough challenge. I mean, part of the challenge when you 4 look at small stocks is that, you know, there simply 5 aren't enough investors in them right now, which in turn 6 creates a lack of liquidity at a basic level. And, you 7 know, the market makers, if you look at some of the 8 recent work in looking at sort of the smaller stocks, 9 market makers say, for example, on the NYSE, do step up, 10 the high frequency guys, tend to make markets more in 11 their smaller stocks. And obviously they're bigger in 12 the NYSE than in others. 13 But it does appear that part of the challenge 14 is how do you make companies go public now? You know, 15 it's a chicken and egg problem. The liquidity isn't 16 there, the regulatory, you know, barriers are high. And, 17 you know, the -- the challenge doesn't appear to be, per 18 se, the market. It appears to be how do we attract the 19 investors to the market. And I'm not sure what the 20 answer is. 21 MR. NOLL: Joe just reminded me that, of 22 course, with the tick size pilot, we are going to have a 23 potential objective test on some market structure changes 24 there. And, you know, we might see some results coming 25 out of that as well. 0033 1 I will just say, as an aside on the tick size 2 pilot, and I know we want to get on to the other topics, 3 I think one of the challenging things that we at the 4 committee have talked about and haven't really considered 5 whether we want to make any recommendations about, is 6 that many firms are choosing not to participate in the 7 tick size pilot because of the costs of preparing to 8 participate in the tick size pilot, knowing full well 9 that there isn't going to be enough volume in those 10 securities to make it worthwhile or bear the costs of 11 making the necessary regulatory and technology changes to 12 participate and are therefore opting out. 13 So it does raise in our heads some of the idea 14 that it may not -- we may not get good results out of it 15 just because you're not going to see enough market 16 participation. But that's not -- that's not a 17 recommendation or a firm opinion. It's just a discussion 18 point that we started to think about. 19 MR. SHILLMAN: Thanks, Eric and Maureen. And 20 thank you, Commissioner Stein. 21 Moving on to our panelists, let me quickly 22 introduce them and then ask each of them to provide their 23 views on the recommendations. 24 Going from my left to my right, we start with 25 Jim Angel, an associate professor at the Georgetown 0034 1 School of Business. Next to him is Stacey Cunningham, 2 Chief Operating Officer of the NYSE Group. Next to 3 Stacey is Venu Palaparthi, Senior Vice President of 4 Compliance, Regulatory and Government Affairs at Virtu 5 Financial. And finally, we have John Zecca, Senior Vice 6 President of Market Watch and head of Market Regulation 7 at NASDAQ. 8 So, Jim, could I ask you to start it off? 9 MR. ANGEL: Thank you. It's an honor to be 10 here. And let's see if I can make the clicker work. 11 But anyway, as most of you know, I am a finance 12 professor at Georgetown, who studies market 13 microstructure. You may not be aware, I also have an 14 engineering background, and I am a co-inventor on 12 15 patents on financial technology. I am also the person 16 who informed or rather warned the SEC in writing five 17 times in the year before the Flash Crash that our markets 18 were vulnerable to these kind of disruptions. So I have 19 been very concerned about this issue of market volatility 20 for a very long period of time. 21 Now -- the clicker is not clicking here. But 22 anyway, so, first of all, I would like to thank the 23 subcommittee for its excellent work. I concur with 24 almost everything you say. So just good work. 25 I'd like to make three points in my 0035 1 preparation, and there's a lot more details in my written 2 comments. 3 But point number one, Limit Up/Limit Down does 4 work to prevent runaway prices, just like the 5 subcommittee said, and it can be improved. Point one. 6 Point two, our market-wide circuit breakers are 7 dangerously defective. And again, I concur with what you 8 said, Eric. You know, thank God they did not trigger on 9 August 24. 10 And third point, I want to echo what 11 Commissioner Stein just said, don't give up on small cap 12 liquidity. There are things that can be done. I'll talk 13 about those in a moment. 14 First of all, with Limit Up/Limit Down, it does 15 prevent runaway prices. And if the clicker here were 16 working, I would show you a very nice graph that would 17 show you an example of Limit Up/Limit Down in action, 18 that basically -- okay, next slide. Okay. Thank you. 19 Okay. 20 And okay. So, as you can see, this is what 21 happened to a stock on August 1, 2014. And the top line 22 is the upper price band. Then we get the reference 23 price, used for Limit Up/Limit Down. Below that, we have 24 the busting price, at which -- below which trades were 25 canceled. And then we have the lower price band. 0036 1 So, as you can see, you know, in a matter of a 2 few seconds, there was volatility in the stock and Limit 3 Up/Limit Down prevented the stock from going to zero. 4 And that's a good thing. But it did not prevent numerous 5 trades that had to be canceled. And that's a problem. 6 And, as was pointed out, if people think their trades 7 might be canceled, they won't step up to provide 8 liquidity. 9 Again, I concur with the subcommittee. The 10 reopening is quite problematic. We have an auction. But 11 guess what? Auctions are very problematic. 12 Now, there are some pundits who view auctions 13 as the panacea for all the market's ills. They're not. 14 They are vulnerable to gaming, they are vulnerable to 15 manipulation. If there is a liquidity imbalance, they 16 are vulnerable to that, too. Their design is very 17 complicated. There are numerous design decisions. It's 18 not just like you have an auction. It's what order 19 types, when do you allow them, do you allow 20 cancellations, what kind of information do you give out, 21 how do you determine the price, how do you allocate 22 shares? There are numerous little decisions to make. 23 And as we saw on August 24, the auctions don't always 24 work. 25 In particular, here's an example of what 0037 1 happened to one large cap ETF that holds only domestic 2 U.S. companies on August 24. You know, there were 12 3 Limit Up/Limit Down halts. 4 Now, the solid line at the top of the graph is 5 my calculation of the true net asset value for the ETF. 6 You know, I took the constituents and I looked at the 7 actual trading prices at which those constituents were 8 trading in consolidated trading at that time. So that 9 was the true price. 10 The pluses, it looks like a scatter shot on the 11 chart. Those are the actual trading prices. And you can 12 see all the gaps when the market was closed because of 13 Limit Up/Limit Down. Humm. We can see that the auctions 14 did not discover anything close to the actual underlying 15 true price of that ETF. And, if you're in the audience 16 and you can't really see the scale, you can see that at 17 times, that ETF was trading more than $20 away from its 18 fundamental value. So we can do a lot better. 19 You'll notice that it wasn't until trading 20 resumed in the continuous market that the price converged 21 with its fundamental value. So what can we do? 22 I really like the idea of gradually widening 23 the bands. I think that proposal is a great improvement. 24 We don't have to halt trading. We can let the prices 25 naturally recover to where supply meets demand. 0038 1 Also, again I concur with the committee, that 2 we need to harmonize the clearly erroneous execution 3 rules with Limit Up/Limit Down. 4 Now, when we talk about market-wide circuit 5 breakers and trading halts, it's useful to think about 6 why halt the market in the first place? What are we 7 hoping to achieve from a trading halt? 8 Now, back in the old manual days, like we had 9 in the crash of '87 when, you know, the systems couldn't 10 keep up, the printers were jamming, the phones weren't 11 being answered, well, it made sense to give the systems 12 time to catch up. You know, in a manual environment, 13 nobody knew what their position was. They needed to know 14 what that was so they could trade. Those conditions no 15 longer apply. 16 Now, clearly, if there is a news pending event, 17 fairness dictates we need to, you know, call a halt, let 18 everybody get the information. If there is a major 19 malfunction in the market, well, we need to turn the 20 machine off if it's producing bad prices. 21 But if the market is under stress, the last 22 thing it needs is more stress when people don't know if 23 it's going to stay open or not. And as we saw in China, 24 the potential closing of the market actually led to even 25 more harm than good. 0039 1 So our current design is defective. We need to 2 fix it. You know, in particular, I concur, 7 percent is 3 too small. There was basically no serious economic 4 analysis when 7 percent was put in. You know? And if 5 you look at it, it would have triggered on the open, 6 after we reopened after 9/11. It would have triggered 7 way too many times in 2008. And was anybody going, oh, I 8 wish we'd halted? No. 9 And in particular, it should be based -- any 10 market-wide circuit breaker should be based on the 11 opening price, not the close. It doesn't make any sense 12 to close the market immediately after it's been closed 13 all night. We have a very gentle and orderly open in the 14 market. You know, the futures trade all night. They 15 tell us the macro price level. Preopen trading by 16 professional traders gives us a really good sense of 17 where the market ought to open. So, you know, our 18 opening mechanism works really well on a daily basis. 19 So if we have had an overnight or a weekend 20 event and the market will close -- or, rather, open with 21 a major change, it should do so. It makes no sense to 22 reclose the market immediately after we've just opened 23 it. 24 Now, another issue, I'm glad you mentioned what 25 was done in the futures markets. There really needs to 0040 1 be close coordination with the derivative markets. We 2 need to really think how we reopen in that kind of 3 situation. And I never thought I'd say this, but where's 4 the FSOC when you need them? 5 Okay, we also need to rethink the reopening 6 process. You know, do we want the market to be totally 7 shut down, or do we want to have the gentle but effective 8 opening procedure we have now, where the derivatives, the 9 futures go first, then the professional traders step in 10 in preopen trading, and finally we open with battle 11 hardened prices. 12 And finally, I want to echo what Commissioner 13 Stein just said. You know, we have a crisis in the small 14 cap sector. That we have roughly half as many U.S. 15 public companies as we had 20 years ago. Now, I know 16 there are people in this building who don't like small 17 companies. There are people in this building who think 18 it's a good thing you keep them out of the hands of 19 investors. That's a mistake. 20 If you look at the Wilshire 5000 index, which 21 basically says it's everything in the U.S., it used to be 22 around 7,000 companies in that index. Now it's down to 23 3,556 as of the end of June. We are shutting off, you 24 know, a very important source of capital for small, 25 growing companies. 0041 1 We are also shutting off an important source of 2 investment opportunities for retail investors. You know, 3 because I don't have access to private equity. So I have 4 access to public companies. I have access to mutual 5 funds. So if we close off this opportunity for small 6 companies to raise funds and go public in the public 7 market, what that means is less capital, that means less 8 investment, that means less economic growth, that means 9 fewer jobs. 10 Now, there's no silver bullet. There are a 11 number of things that are driving this, everything from 12 our messed up litigation system -- talk to CEOs, they'll 13 tell you, if I go public, I get sued, end of story. We 14 have all the compliance burdens that, you know, trying to 15 cut down a large cap compliance regime for a small cap 16 company, it's like, you know, taking your daddy's 17 oversize suit and trying to sort of cut it down for a 18 small kid. It never really works. 19 We also do have market structure issues. And 20 in the old days, we -- issuers used to have a very strong 21 choice. They could list on the high-cost NASDAQ market. 22 You know, those bid/ask spreads were so wide, we thought 23 it was a scandal. But those higher transaction costs 24 supported, you know, a lot of research that created an 25 information environment that investors traded in. Or, 0042 1 you know, they could choose the low-cost AMEX market. 2 But they didn't. Okay. 3 There are things we can do. You know, for 4 example, we can adopt the European style approach of 5 allowing issuers to subsidize market markets, to 6 subsidize liquidity in their shares. There are other 7 things we should look at but, you know, I want to let my 8 fellow panelists also talk. 9 Thank you. 10 MR. SHILLMAN: Thanks, Jim. 11 Stacey. 12 MS. CUNNINGHAM: Hi. Thanks, Steve. 13 My name is Stacey Cunningham and I am the chief 14 operating officer of the New York Stock Exchange with 15 responsibility for our markets. 16 I would like to thank the Commission and the 17 committee for the opportunity to participate in today's 18 Equity Market Structure Advisory Committee meeting. I 19 particularly appreciate the ability to share NYSE's 20 initial views on the Market Quality Subcommittee's recent 21 recommendations. 22 Market quality is constantly at the forefront 23 of our minds at NYSE, not simply as a reaction to August 24 24, 2015, but because high quality markets benefit our 25 listed companies, exchange members and investors. I will 0043 1 begin by sharing our thoughts on market quality generally 2 and then in the context of the subcommittee 3 recommendations, provide an update on the progress the 4 Limit Up/Limit Down Operating Committee has made toward 5 revising the LULD plan, as well as an update on NYSE's 6 opening processes. 7 A primary tenet of market quality is providing 8 stable displayed quotes with meaningful size and a narrow 9 bid/ask spread. High quality markets are achieved by 10 bringing together committed liquidity providers with a 11 wide variety of liquidity seeking investors that have 12 diverse investment goals and time horizons. This becomes 13 increasingly important but increasingly challenging in 14 today's fragmented marketplace. 15 On recent days, more than 40 percent of trading 16 occurred in the dark. And as a result, liquidity 17 providers on exchange don't have the benefit of 18 interacting with diverse order flow. Yet the market is 19 still dependent on them to be setting -- establishing 20 benchmark prices. Those prices are used broadly, from 21 retail investors who are making investment choices, to 22 the off-exchange venues using them to match trades. 23 As the Equity Market Structure Advisory 24 Committee and all of its subcommittees, not just the 25 Market Quality Subcommittee, consider market quality, 0044 1 promoting displayed liquidity must be a primary focus. 2 And finding ways to encourage market makers to provide 3 displayed liquidity at all times, especially during times 4 of volatility, is a means to that end. 5 At NYSE, our market models are constructed to 6 create that right balance between market maker benefits 7 and obligations. The designated market makers on NYSE 8 have more significant obligations than any other class of 9 market maker in U.S. equities. Among these obligations, 10 they guarantee all interest in the opening and closing 11 auctions is satisfied. They also have to contribute to 12 the NBBO, not just have a two-sided quote out in the 13 market, but be on the NBBO a certain percentage of time 14 of the day, and they have to provide depth continuity as 15 multiple price points throughout the trading day. 16 And unlike most market maker incentive 17 programs, these aren't voluntary obligations that they 18 can ignore when market conditions warrant it. They are 19 rule-based mandatory obligations that carry regulatory 20 risk. And in exchange for that risk, they are rewarded 21 with a set of incentives that come in the form of parity 22 participation and enhanced economics. And we quite 23 literally are investing in market quality for our 24 issuers. 25 On August 24, it became clear though that the 0045 1 balance of market making benefits and obligations had 2 underestimated another factor, and it's the market 3 structure risks and uncertainty that we've talked about a 4 number of times. And just to highlight a few of them, it 5 was uncertainty that market makers had around a market 6 wide circuit breaker trigger, you know, potentially 7 triggering that, around hedging opportunities that were 8 reduced, around clearly erroneous executions or around 9 opening prices that weren't yet established. And it's 10 with the goal of reducing that uncertainty of liquidity 11 providers and increasing transparency for all market 12 participants that we approach the industry dialogue for 13 improving market quality. 14 The Market Quality Subcommittee notes that the 15 primary function of the LULD plan works, but that the 16 reopening process doesn't function well. We don't 17 disagree with that. The subcommittee sites a lack of 18 market participation during the reopening process 19 following a trading halt compared to similar opening and 20 closing auction processes on primary listing venues as an 21 issue and believes that making changes to address that 22 issue would be too cumbersome for the industry. 23 For these reasons, the subcommittee has 24 recommended focusing on a futures-like extension, as Eric 25 highlighted, rather than pursuing trading pauses or 0046 1 pausing trading. We agree that moving to a structure 2 that -- where limits dates would be extended rather than 3 halting is certainly worth exploring. But any such 4 overhaul of LULD warrants an extensive review before it 5 could be finalized. We believe the markets shouldn't 6 lose changes that we can make in the short term to 7 improve the operation of the current LULD plan while 8 we're assessing more significant changes. 9 As one example, amendment of the plan was 10 recently implemented on July 18. It revised the 11 reference price used in the absence of an opening 12 auction. And based on initial results, it's led already 13 to a roughly 75 percent decline in spurious trading 14 pauses. And we think there are more things that we can 15 do in the short term while evaluating bigger changes. 16 Earlier this year, the LULD Operating 17 Committee, in tandem with the LULD Advisory Committee 18 identified two obvious areas that required in-depth 19 review and established working groups dedicated to 20 addressing those areas. The recommendations have been 21 based on empirical observations as well as extensive 22 dialogue with market makers, institutional investors, 23 retail investors and issuers. It's been a very 24 collaborative and productive process. 25 The first working group was the trade 0047 1 resumption working group and spent time comprehensively 2 assessing that reopening auction following a trading 3 pause. And we've agreed to concrete changes that we 4 believe would improve the process. 5 Based on the feedback that we' have received, 6 the primary markets have committed to adopting a 7 standardized approach regarding reopening auctions that 8 would be implemented through both amendments to the LULD 9 plan as well as exchange rules. Specifically the plan 10 participants intend to propose an amendment to the LULD 11 plan that would prohibit market centers from resuming 12 trading in a security until the primary reopens and until 13 price bands have been disseminated and received by the 14 SIP. In connection with the changes to the LULD plan, 15 the primary markets would propose changes to their own 16 rules to adopt consistent standards for automated 17 reopenings following a trading pause. The goal of these 18 changes is to make sure that all market orders will be 19 satisfied in a reopening auction. Satisfying all 20 marketable interest was unanimously viewed as a critical 21 requirement among retail broker-dealers. 22 Our proposed solution also addresses concerns 23 around mean reversion, which were included in the 24 subcommittee recommendation. The Trade Resumption 25 Working Group has made significant progress and we 0048 1 believe the SROs would be ready file both the proposed 2 amendments to the LULD plan and primary market rule 3 changes this month, to have them to the SEC for review 4 this month. 5 The second working group was on re-tiering. 6 And that group has been working on identifying methods to 7 re-tier the securities in the NMS plan. And key aspect 8 of the analysis is to identify tiers that would obviate 9 the need for clearly erroneous execution rules. Which 10 would also eliminate another area of uncertainty for 11 market makers, which leads to better market quality. 12 At a high level, we're looking to move away 13 from index participation or product type as the reasons 14 what would classify a certain product or security into a 15 tier and instead focus on the volatility profile of a 16 security to determine which tier it should fall into. 17 While the proposed changes of the working 18 groups would require technology changes by the exchanges 19 and the SIP, they wouldn't be onerous changes for market 20 participants. They're also not mutually exclusive to 21 analyzing a futures-like approach to LULD, which we think 22 we should continue to explore. 23 Both the Trade Resumption Working Group and the 24 Re-Tiering Working Groups received heavy support from the 25 LULD Advisory Committee, as well as extensive industry 0049 1 feedback. And we'd like to thank all of those who 2 contributed. 3 The subcommittee also recommended that the 4 exchanges and the Commission adopt policies and 5 procedures to ensure all stocks open as close to 9:30 as 6 possible. And as Eric indicated, it's not opining on a 7 certain market model, but saying that they all need to 8 open at 9:30 does, indeed, opine on a certain market 9 model. We challenge the value in emphasizing this as a 10 market quality objective. 11 We believe issuers should be able to choose 12 what matters most to them. And NYSE listed companies 13 choose price discovery, combining human judgment with 14 automation leads to better result for both the issuer and 15 their long-term investors. Prioritizing price discovery 16 over speed is especially important, because many retail 17 investors participate in that opening auction in the 18 first few minutes of the day. And they can be most 19 benefitted by an opening that's well priced. 20 A measure of the success of the NYSE opening 21 process is the price volatility that follows the opening. 22 For example, on Brexit Day, NYSE listed names experienced 23 50 percent less volatility in aggregate after the opening 24 compared to fully electronic venues. Given the 25 fragmented nature of today's markets, we recognize though 0050 1 the importance of opening as quickly as possible, and we 2 have been working with our DMMs to encourage finding the 3 right price. We've already seen progress on this front. 4 In July 2016, 99.8 percent of NYSE stocks were 5 open by 9:35, on average, compared with 96.4 percent in 6 January of this year. Additionally, we recently received 7 SEC approval on rule changes that will allow for opening 8 automation parameters and increased transparency rules 9 that adjust for volatility. These changes were designed 10 to facilitate faster openings, and as well as to provide 11 a lot of transparency into the opening pricing process 12 while still protecting price discovery. And they're 13 scheduled to be implemented this month. 14 I would be remiss if I didn't briefly touch on 15 the subcommittee recommendations relating to customer 16 issues. Although we won't be participating in this 17 afternoon's panel, we do want to state our support of 18 those recommendations. With respect to recommendation 19 number 2 in particular, if the SEC is going to continue 20 to use Rule 605 and 606 reports as a basis for execution 21 quality and to disclose conflicts of interest, we believe 22 the Commission's recent proposal is a step in the right 23 direction -- I'm sorry, the committee's recent proposal 24 is a step in the right direction. We also believe the 25 Commission should expand those proposals or develop 0051 1 comparable proposals to provide more transparency of the 2 fixed income markets. 3 Thank you again for the opportunity to present 4 the views of NYSE. 5 MR. SHILLMAN: Thanks, Stacey. 6 Venu. 7 MR. PALAPARTHI: Thank you. Thank you for the 8 opportunity to provide Virtu's perspective. Virtu is, as 9 you know, a leading market maker across the globe. We 10 trade on multiple exchanges and multiple asset classes. 11 As a technology enabled market maker, we view 12 anything that hurts the market as having to hit the 13 control alt del, as opposed to just a refresh. For us, 14 for most of us in the room, a refresh is much easier to 15 handle than a complete reboot. And especially on days 16 like August 24, when you have 773 limit up and 505 limit 17 downhauls, that's a lot of rebooting. 18 So we view this as -- this committee's work as 19 consistent with our views in the marketplace. We feel 20 that anything that extends the limit state as opposed to 21 halts the market is beneficial from a market maker's 22 perspective. To process that new information and 23 generate that better price, which is our function, 24 primary function. 25 We think that the committee's recommendation to 0052 1 extend the limit state is definitely very useful. But, 2 in addition to that, we do agree with what James 3 suggested, to widen the bands, you know, as we go down. 4 And that kind of helps us to continue to stay in the 5 market and process the information and generate a price. 6 So those are particular recommendations of the 7 committee we are 100 percent in agreement with. 8 With respect to mean reversion, I do agree that 9 needs more study. But once again, it is consistent with 10 the overall objective of keeping the markets open and 11 reducing the number of halts. So that is definitely 12 something that we would support. 13 Nobody can argue with harmonizing rules across 14 exchanges. We do trade on many exchanges and in multiple 15 asset classes. And therefore, for us, price discovery is 16 not limited to a certain exchange. And when, you know, 17 it's impacted by what is going on in one asset class or 18 one exchange, that kind of leaves us wondering why we are 19 forced to refresh and, you know, we should not have to do 20 that. 21 Synchronizing clearly erroneous with LULD, I 22 think James made a compelling case for that. There's no 23 argument there. 24 With respect to market-wide circuit breakers, 25 once again, we would definitely urge the consideration of 0053 1 widening the bands. We feel that anything that brings 2 the market to a halt is definitely disruptive in the 3 price discovery -- to the price discovery process. We 4 appreciate all of the efforts that NYSE and NYSE Arca has 5 taken with respect to bringing about more transparency, 6 especially with preopening indicators. And I want to 7 emphasize that opening all markets as close to 9:30 as 8 possible is something that we definitely appreciate. And 9 they have made efforts to do that. They are making 10 continuing efforts to open as close to 9:30 as possible, 11 and we are appreciative of those, and we urge them to 12 consider even more measures to open as close to 9:30 as 13 possible. 14 That, you know, we are very speedy. And that's 15 my speedy response. Thank you. 16 MR. SHILLMAN: Okay, great. Thanks, Venu. 17 John. 18 MR. ZECCA: Thanks, David. And good morning, 19 and thank you to the Commission and Staff and committee 20 for inviting me. It's an honor for me to be here on 21 behalf of NASDAQ. 22 Just by way of background, I'm the head of 23 market regulation for our 11 trading venues in the United 24 States and Canada. And I have had the privilege of 25 working in one form or another in the securities industry 0054 1 for about 25 years, although I'd say it's even longer. 2 If you'd allow me one quick personal story, my 3 father, a former stock broker, taught me to read the 4 stock tables and said I could buy a few shares once I did 5 my own research. Of course, the stock I bought was Pan 6 Am, which went bankrupt within about five years, so 7 that's probably why I'm a lawyer and not an investment 8 advisor. But I think it does illustrate the critical 9 role that the markets play in the lives of Americans of 10 all ages. 11 NASDAQ has always viewed our highest priority 12 as running well regulated markets that serve the needs of 13 investors, market participants, and now 3,700 listed 14 companies around the globe that contribute so much to the 15 growth of the world economy. To that end, we've been a 16 leader in realtime surveillance of equity markets, which 17 gives us a unique perspective on the operations of the 18 market in real time and how interconnected and complex 19 our market is. So it's from this perspective that we 20 join you today to discuss the recommendations of the 21 market quality subcommittee. 22 As we've stressed in our previous panels, there 23 can be no doubt of the need to reassess U.S. markets to 24 ensure that they continue to provide accurate priced 25 discovery, coupled with efficient and effective trade 0055 1 executions to serve the needs of the broader economy, 2 which is our ultimate goal. 3 The subcommittee's recommendations are 4 tentative steps in that direction, focused as they are on 5 market disruptions. And indeed, since the 2010 Flash 6 Crash, as the Chair noted, the Commission and 7 self-regulatory organizations with substantial input from 8 the industry have adopted a number of safeguards, 9 including first circuit breakers then Limit Up/Limit 10 Down, Reg SHO circuit breakers and updated market wide 11 circuit breakers. 12 Yet the disruptions on August 25, 2015, 13 demonstrated that further modifications are needed. Many 14 stocks had downward corrections before recovering were 15 unnecessary Limit Up/Limit Down halts, while other stocks 16 did not open on time on their primary listing market. 17 NASDAQ, as chair of the committee administering 18 the national market system plan to address extraordinary 19 market volatility, took a leading role in coordinating 20 input from market participants to address this concern. 21 This culminated in implementation, within three months 22 after SEC approval of the proposal, of a plan to reduce 23 the number of Limit Up/Limit Down halts triggered upon 24 the open, which Stacey mentioned, including the very 25 impressive results we saw of a 75 percent reduction. 0056 1 The work will not stop there. As Stacey 2 described, there are a number of other features before 3 the committee. I'd like to focus on a couple of things 4 regarding clearly erroneous and Limit Up/Limit Down. I 5 do think that it's important that we consider the 6 difference between market disruptions and ordinary 7 activity, whether it's mean reversion on Limit Up/Limit 8 Down or the clearly erroneous bands themselves. 9 You know, let's take the LULD piece. Certainly 10 in a market disruption, you have a sense that the right 11 market was probably somewhere before the panic started. 12 But in a news event with an individual stock, there may 13 not be a price to go back to. Because obviously, the 14 price now has to factor in that news. And sometimes 15 stocks do bump up and down just as people try to digest 16 the news. So I do think our discussion has to take those 17 nuances between everyday and market disruptions into 18 account. 19 However we do that though, I think the market 20 structure changes highlight the valuable role played by 21 national market system plans in fostering collaboration 22 among the exchanges and market participants and allowing 23 experimentation in a more flexibility way than would be 24 the case solely through SEC rulemaking. So we thank the 25 Commission for supporting this process. 0057 1 While we have not had sufficient time to 2 completely assess the subcommittee's recommendation 3 regarding eliminating traditional trading halts and for 4 LULD, we welcome the idea as an additional option for 5 consideration. Before adoption of this recommendation, 6 however, it would be critical to understand how the 7 resumption in trading would work to ensure that 8 uncertainty and volatility are not exacerbated if the 9 auction process is avoided. Considerable testing and 10 input from market participants are essential for any 11 changes. 12 The NASDAQ listing exchanges process all 13 opening auctions immediately at 9:30. And we strongly 14 agree with the subcommittee's recommendation that all 15 stocks listed on U.S. exchanges open as close to 9:30 as 16 technically possible. August 24 illustrated the risks 17 when underlying securities do not open on time on the 18 primary listing market, impacting users who rely on data 19 from the primary markets and making it difficult to 20 assess the value of individual companies and the market 21 as a whole. This can also impact, of course, derivative 22 products, including exchange traded products in which 23 many retail investors rely. 24 While steps have been taken to reduce this 25 problem, trading on the day after Britain's vote to leave 0058 1 the European Union indicates that the problem persists. 2 As we argued recently in another context, while 3 participants benefit from a choice in market models, 4 those models should not impede overall market efficiency 5 and price discovery. 6 The subcommittee's comments on market-wide 7 circuit breakers also merit attention. Circuit breakers 8 sort of remind me of oxygen masks on an airplane. You 9 know, they are comforting to know they exist but panic 10 inducing when they're deployed. The subcommittee is 11 right to recommend review in light of recent experience, 12 to ensure that they deploy only when absolutely needed. 13 Although we are still reviewing the 14 subcommittee's comment, which I understand was not a 15 recommendation regarding the reference index, we do have 16 some initial concerns about the use of the S&P futures 17 contracts as a measure, because these instruments 18 introduce the potential risk of contagion since they do 19 represent technically a different asset class than 20 equities. And perhaps work should be done to ensure that 21 the equities version is in fact a good measure. 22 One note of strong dissent in echoing comments 23 by Commissioner Stein and others, NASDAQ is disappointed 24 with the subcommittee's unwillingness to consider 25 recommendations on ways to improve the trading of small 0059 1 and mid cap companies. It appears from the 2 recommendations that, while the subcommittee solicited 3 feedback from the trading and investment community, it 4 may not have spoken directly to smaller issuers for 5 feedback. And again, we only know what we saw so if this 6 isn't correct, then we are concerned that the 7 subcommittee ignored the community most familiar with the 8 problem. 9 As we've previously stated, the 10 one-size-fits-all market structure we suffer with today 11 has not served these issuers and their investors well, 12 and puts the capital formation engine of our country at 13 risk. In fact, Congress has focused most of its review 14 of market structure in the context of the needs of small 15 companies, and this committee owes those companies a 16 similar focus. 17 As the listing venue of choice for many 18 emerging growth companies or EGCs, NASDAQ believes that 19 there are indeed solutions that can improve the markets 20 for EGCs. I think Jim did a great job of outlining a few 21 of them. Sort of in that vein, a couple of these include 22 reducing fragmentation in EGCs by limiting or eliminating 23 unlisted trading privileges to create a single center of 24 liquidity where the vast majority of supply and demand 25 can meet. Relating to that proposal, allowing directed 0060 1 orders in EGCs so that brokers can ensure the execution 2 quality experience of their customers. And also 3 considering order delivery for EGCs to reduce the risk 4 associated with posting quotes and orders across multiple 5 markets. In other words, relaxing autoexecution where 6 necessary to encourage market participants to post more 7 marketable volume. 8 On a broader level, NASDAQ encourages the 9 subcommittee and the Commission to accelerate the 10 consideration of a wide range of market structure issues 11 before us. We need to find a faster way to experiment 12 and ultimately implement needed changes. The U.K.'s 13 financial conduct authority recently announced that it 14 will encourage product innovation through a regulatory 15 sandbox initiative and, while it's not identical to what 16 we're speaking about, we think the concept could be 17 extended to market structure initiatives. We have also 18 advocated using market simulations to expedite 19 consideration of some proposals. 20 It is through the audacity to challenge the 21 market structure status quo that we will ensure that the 22 benefits of equity ownership are clear to the next 23 generation of young investors seeking to put their 24 savings into the market. 25 I thank the Commission and the Staff and the 0061 1 committee for their time and look forward to a fruitful 2 discussion. And my full statement is on the website. 3 Thank you. 4 MR. SHILLMAN: Thanks very much, John. So I 5 will open up now to discussion and questions from the 6 full committee and the commissioners. 7 MR. KETCHUM: Just to follow up your very 8 interesting remarks, you spoke about -- and Stacey, I 9 know you separately spoke, both of you, about the need 10 for additional evaluation of the subcommittee's 11 recommendation off a more futures limit approach to 12 reopening. You mentioned a particular concern off of how 13 the reopening would be managed. 14 Is your concern off of the manage of a buildup 15 of market orders over the two minutes? Or could you talk 16 a little bit more as to what exactly your concern is from 17 the standpoint of a futures limit type approach? 18 MS. CUNNINGHAM: Our view is that there isn't 19 necessarily -- we don't necessarily eliminate the need 20 for a trade resumption in some form. So that's why we 21 wanted to address that in the first step. 22 I wasn't sure from the details within the 23 subcommittee's proposal how exactly trading would resume 24 and how orders would be processed coming out of that 25 period. 0062 1 MR. ZECCA: And I think to that end, the reason 2 we have opening auctions and things like that is to try 3 to aggregate the volume. The feeling was that just 4 allowing the trade to go out, even though it may be in 5 line, is not necessarily the best form of price 6 discovery. So again, we're not entirely sure how that 7 would work here. But in the end, you're sort of playing 8 with the concept of a halt without calling it a halt and 9 I think the question is in doing that, are you in fact 10 limiting the opportunity for orders to come in and meet 11 and go through the auction process or not. So I think 12 that would be a fundamental question. 13 MR. SHILLMAN: Joe? 14 MR. RATTERMAN: Yeah, I'm pretty excited about 15 the recommendations put forward. I agree with Stacey 16 that we should probably let the exchanges and their 17 recent standardization, harmonization efforts give the 18 new auction standardized process a chance. I like the 19 idea of extending the limit state for two minutes or 20 maybe longer. But eventually you may -- no matter 21 mechanism you come up with, you may never, you know, 22 actually get to the right price and an auction may be the 23 only way. So given that that backstop is probably always 24 there, I think giving this new harmonized effort some 25 time in the market to see if that improves things would 0063 1 be wise. 2 So extending the time by two minutes, 3 absolutely. I like the idea of mean reversion or 4 basically allowing the stock to recover in the space that 5 it went on its volatility run without re-halting on the 6 way back is quite vital. So, yeah, my personal opinion 7 is let's let the new standardization that the committees 8 have come up with try it out in the market for a while. 9 We can always come back to this. All the other changes 10 with regards to Limit Up/Limit Down, we're fully in favor 11 of. 12 MR. SHILLMAN: Joe? 13 MR. MECANE: Sure. One clarification along 14 those lines maybe for Eric in terms of how the 15 subcommittee thought about the extension of the 16 two-minute pre-open period. I agree with all the 17 recommendations. The one reaction I had though was if 18 participation in the auction is one of the issues that 19 has been identified and something that we're trying to 20 figure out how to better address, in your conversations 21 with market participants, with the extension of that 22 pre-open state to two minutes and letting people add, 23 subtract interest in the auction, was the feedback that 24 people would be willing to participate for longer periods 25 of time with that mechanism going on? I'm just curious 0064 1 what the thinking was in terms of how that would balance 2 out? 3 MR. NOLL: Yeah, the timing issue actually, I 4 think, was a subject of some controversy in our 5 conversations. Some of the market participants really 6 want it very short, 10 seconds, 15 seconds. Some of the 7 other market participants argued that it could take as 8 long as five or 10 minutes for them to adequately 9 participate in a reopening process because of the need to 10 engage with other people or to think about it, to assess 11 where they thought the stock should trade. The two 12 minutes really ended up being a compromise, a sense that 13 we probably needed to extend that limit state for a 14 little bit longer to let -- at least maximize 15 participation from people who currently do not 16 participate in reopening processes, one because they 17 either don't have a mechanism to do it, or the timing is 18 just too short for them to do so. 19 MR. MECANE: And along those lines, is the 20 thinking that at the end of that two minutes, there would 21 be some form of an auction to reopen, or would the idea 22 be that it would just begin trading -- 23 CHAIR WHITE: Joe, could you just stay on your 24 mic? Thanks. 25 MR. MECANE: Yeah, I think the idea is that, 0065 1 unlike an auction process where everybody would halt and 2 the primary market would then reopen the security, as 3 long as there were indications of interests and bids and 4 offers and available book on the data feeds for everyone, 5 that the market would just simply reopen and start 6 trading again. 7 MR. SHILLMAN: Chester. 8 MR. SPATT: So overall, I like the 9 recommendations of the -- of the subcommittee. But I did 10 have a few -- a few comments. 11 First, on the issue of clearly erroneous 12 trades, I think this is actually -- I think this is very 13 important. I think it's very important to have as much 14 harmonization as possible in that space. I think, at 15 least with hindsight, some of the most problematic 16 actions taken by the agency over the last half dozen 17 years have been to rule trades as clearly erroneous where 18 it wasn't clear, ex ante, that they were erroneous. And 19 it seems to me having clarity, having as much clarity as 20 possible and ideally aligning with other rules as is 21 proposed here, is just very, very important because of 22 the impacts on ex ante liquidity provision. And what we 23 really want to have in these -- in these awkward times, 24 in the instances like the Flash Crash or August 2015, we 25 want to promote liquidity provision. And that's why I 0066 1 think that's just so important to have clarity about 2 execution. Because those are just exactly the times 3 where people are going to otherwise stand back. So I 4 think that's extremely important and I think the 5 committee is -- the subcommittee is to be commended for 6 taking that on. 7 One point where I have some disagreement, 8 although this may be largely with the discussion of the 9 committee, I'm not sure, is on the issue of the 10 market-wide circuit breakers. So I am in agreement on 11 the usefulness of widening the bands. But I'm not sure 12 that I quite understand the case for widening the bands 13 from 7 percent to 10 percent, per se. It seems to me 14 there's a lot of fighting the last war in this. So one 15 is sort of looking back at a few episodes. There's a 16 very small number of these episodes, like maybe one and 17 one, and 7 percent kind of worked in a particular way and 18 10 percent would have worked in another way in these 19 episodes. 20 Well, this is just only ex post. This is not 21 any kind of ex ante description of what would have been 22 better, what would have been worse. And that even leads 23 me to conclude maybe the bands ought to be even possibly 24 wider. But I don't -- I don't know. It just seems to me 25 there's not an empirical -- there's not a compelling 0067 1 empirical case for 10 percent versus 7 percent. And I 2 think that's potentially an important issue to reflect 3 more about. There's basically -- you know, the fact that 4 we have, in effect, just an empirical observation or two, 5 that's not really data in the way we usually think of 6 data. I mean, the fact -- I mean, the fact it's kind of 7 applied across the market, those are not independent 8 events. You know, you're basically talking about one or 9 two events here. 10 So the final point that I wanted to raise was 11 about small companies. And as somebody who wasn't on the 12 committee, the subcommittee, I want to weigh in in 13 support of the subcommittee, not knowing the details of 14 the due diligence that they did. But as I just reflect 15 on this, you know, it seems to me that problems in the -- 16 problems in small companies wanting to go public, if you 17 just step back and reflect on where are the costs likely 18 to be and what's likely to be first order versus second 19 order, it seems to me any kind of awkwardness about the 20 specifics of our trading process is likely to be second 21 order compared to going -- the costs of being a public 22 company. Those costs seem to be must more first order 23 from a company's perspective. 24 The active decision companies are making is, 25 are they big enough to incur those costs. This was a big 0068 1 issue indeed when I was on the Staff of the Commission in 2 some of the discussion about Sarbanes-Oxley. And 3 ultimately the margins on which the regulatory actions of 4 the Commission started to evolve with respect to 5 Sarbanes-Oxley were where were the lines with respect to 6 small companies, what was the nature of risk in the 7 context of 404 compliance, et cetera. And it seems to me 8 these are likely to be where the first order issues and 9 challenges are, and where potentially the SEC might have 10 the most leverage. 11 I'm sort of -- I start out quite skeptical that 12 the details of how we set up our trading process and 13 whether the tick size is a penny -- for example, in a 14 slightly different context, whether the tick size is a 15 penny or a nickel is going to influence the decisions of 16 whether these companies are going to go public. I would 17 imagine those decisions are much more heavily influenced 18 by what is the size of their -- what is the size of the 19 potential capitalization of these companies, what are the 20 annual costs of being a public company and what are the 21 tradeoffs associated with it. That seems to me much more 22 where the first order costs are likely to be. 23 And if there is any context where perhaps 24 investor confidence may be an issue, it may actually be 25 also about whether to invest in small companies 0069 1 individually. And it may be here that the lessons from 2 -- even from 2000 and/or perhaps 2008, but perhaps even 3 going back to 2000 may still be resonating with some of 4 the public and, in particular, with the assessment of 5 these firms as to whether it's worth taking on retail 6 investors. Because, you know, the flip side, of course, 7 is that part of -- meanwhile, part of the agency's 8 mission, of course, is protecting investors. And I don't 9 mean from the point of therefore excluding investors from 10 participating in what are desirable investments. But it 11 may be it's just hard to have the due diligence done in 12 that sort of way. And private equity may be the -- may 13 be the right alternative. 14 MR. SHILLMAN: Kevin and then Rick. 15 MR. CRONIN: I just wanted to make an 16 observation with respect to the timing issue. As we had 17 the conversations at the subcommittee level, whether it 18 was two minutes or 20 minutes or 20 seconds, I think one 19 important consideration is to make sure that the 20 information is available. Right? One of our big 21 concerns has been that if the information is only 22 available via proprietary data feeds, it doesn't have 23 enough penetration in the marketplace for rational people 24 to do rational things. 25 And I think to the observation around 0070 1 algorithms, algorithms will take a step back in 2 environments of uncertainty, just as market makers will, 3 and that exacerbates the problem. And part of that I 4 think starts from just a lack of information. 5 So I think the table stakes for this is that 6 the information has got to be available, not to a select 7 few but to everyone. And whether that, again, is two 8 minutes or 30 seconds, I think to the extent that the 9 information is better available, we'll get better 10 answers. 11 MR. KETCHUM: Just two reflections on 12 system-wide circuit breakers. The second piece to a 13 question that Professor Angel -- interesting point he 14 made. 15 First, I think the subcommittee's 16 recommendation on system-wide circuit breakers couldn't 17 be more correct. As one along with the two Joes who were 18 involved in this shift in recommendations to the 19 Commission on a lower circuit breaker. 20 Demonstrating I've been around way too long, if 21 you look back to the actual reflection when the circuit 22 breakers were created after the '87 crash, they were 23 focused on two things, an effort to try to interdict in 24 situations where trading had become truly disorderly. 25 And equally important, where that disorderly trading had 0071 1 been so extreme as to create credit concerns, which 2 contributed to greater disorderliness. 3 We've moved a long way away from that with 4 respect to 7 percent. I think we did it perhaps as an 5 overreaction but a logical reaction to the poor way the 6 markets were operating in a new electronic environment, 7 and unclear what the impact of single stock circuit 8 breakers would be. 9 It seems to me now that there's enough clarity 10 as Eric points out that single stock circuit breakers 11 basically work, that having -- having a system-wide 12 circuit breaker at a low level no longer is justified 13 anyway. It may not have been justified initially, it 14 certainly is no longer justified. And it creates far 15 more risk and it needs -- I agree with Chester, there is 16 no quantitative analysis to determine that 10 percent is 17 right. What seems to me to be absolutely right is that a 18 quick move is absolutely right. And there is lots of 19 quantitative information on that, as well as a reflection 20 back on the initial pieces off of the performance of the 21 market since then. 22 So I would say move it anywhere up and then 23 continue to do more study as to what the right number 24 would be, but move it up. 25 The second piece though that I thought 0072 1 Professor Angel was interested but I have some concerns 2 of is the suggestion that any system-wide circuit breaker 3 should be keyed off of opening prices which, in many 4 ways, makes great sense. But I guess my concern would be 5 that opening prices tend to be driven by a futures market 6 that is not subject to effective arbitrage and sometimes 7 overshoots, and tends to be driven by initial openings 8 which, however it ends up, if indeed one assumes the 9 effort is to push to 9:30 very much electronic type 10 openings is likely to be, if anything, faster transmitter 11 of the potential both correct pricing information and 12 potentially incorrect pricing information. 13 And I guess my question would be so, if it goes 14 down 25 percent at the opening are you still up for it to 15 go down another 20 percent? Or would you have lower 16 levels if you're going to tee everything off the opening 17 price? That would be my worry there. 18 Returning back to the disorderly trading at a 19 level that caused credit risk, if you open the market 20 down 20, 25 percent, I think having the stomach to go 21 another 10 to 20 percent would spook me. 22 MR. SHILLMAN: Jim and then Joe Mecane. 23 MR. ANGEL: Thank you. I think those are 24 excellent points that really need to be discussed. But I 25 think the thing to remember is it isn't just the futures. 0073 1 You know, yes, the futures are trading all night. But we 2 also have the pre-open trading. And if a, you know, 3 information event is occurring overnight, there has been 4 plenty of time to digest that, plenty of time to do a 5 proximate arbitrage between the pre-open and the futures. 6 So I would think, you know, the -- what price 7 is discovered by 9:30 is probably a pretty accurate price 8 of where the market is at 9:30. 9 Now, but another very important point, every 10 one of these circumstances is different. And if we try 11 to come up with a one-size-fits-all rule, chances are it 12 won't fit the next market tsunami. At some point, we 13 need to have some kind of market judgment so that if 14 there is some kind of either information event or other 15 situation, somebody needs to be able to, you know, 16 basically, you know, push the button and say, hey, let's 17 have a halt. I can't think of what that circumstance 18 would be. But we can't think what all the circumstances 19 would be either. 20 So at the end of the day, we do need market 21 judgment. And I think working off the open, you know, is 22 a lot better than working off the previous close. 23 Because if you think about, you know, the morning after 24 9/11 -- actually September 17, you know, we had a pretty 25 fair and orderly open, the market was down a little over 0074 1 7 percent, which is approximately where it closed that 2 day. Now, if you think about one of the reasons for a 3 halt is to change market mechanisms, to go from a 4 continuous auction to a call auction, and the idea is you 5 give people time to assess information, to check their 6 trading positions, to put in orders. But that may have 7 made sense in the manual days but, as Venu pointed out, a 8 refresh is easier than a reboot for a lot of people in 9 today's electronic markets. 10 So if you do want to switch market mechanisms, 11 you just had a call auction. So the fact that you didn't 12 like the results of the call auction doesn't mean you 13 should immediately have another call auction. So that's 14 why I think we should definitely work off the open rather 15 than the stale price from a previous close. 16 Oh, and I would like to add one more thing. 17 You know, we do Limit Up/Limit Down one way, but the 18 market-wide circuit breakers are basically done not 19 through a national market system plan but through 16 20 separate SRO filings. I think that is a prime example of 21 the fragmentation of our regulatory system and it should 22 either be done through an NMS plan or through a single 23 SEC rule, not through 16 separate SRO filings. I think 24 that's very inefficient. 25 MR. MECANE: Just a few follow-on comments from 0075 1 Rick's point and Professor Angel's points, without 2 necessarily a strong conclusion. But just for context, 3 when we picked the 7 percent, and a lot of people were 4 involved in that discussion around this table, it was 5 partly to address two things. One was a criticism or a 6 concern that if we had a very large number of individual 7 either single-stock circuit breakers or limit up/down 8 triggering and didn't have a broad enough halt across the 9 market, would that be the right result? And so the 7 10 percent was again not necessarily right or wrong, but the 11 feeling at the time was that since 10 percent had never 12 been triggered since the '87 crash that, you know, there 13 was a concern that using that as the benchmark going 14 forward wouldn't address that particular issue. So 15 whether that's right or wrong, that's just one point of 16 clarification. 17 The other thing on the idea of benchmarking 18 versus the open instead of the close, I mean, Jim makes 19 good points. Rick raises one of the two points why I 20 think we ended up recommending the close which was, you 21 know, the validity of the futures prices. But to respond 22 to Jim's point, the problem when we looked at pre-open 23 trading was that even on very volatile days, the number 24 of individual securities that trade in the pre-open 25 session is somewhat random and not always broad enough to 0076 1 give a strong enough sense of where the market is. And 2 the volatility, once you get outside the top 50 or 100 3 most active names of the pre-open price compared to where 4 the continuous market opened was disparate enough that it 5 was hard to rely on. So we couldn't come up with a good 6 mechanism to kind of link the two things together. 7 Again, maybe not the right answer, but just context. 8 The last point I was going to make around small 9 caps, and I wasn't on the subcommittee that talked on 10 this particular subject, but maybe a semantical point but 11 just a perspective is I don't think the subcommittee, not 12 to speak for them, seems to be saying that the 13 effectiveness or the importance of small cap trading and 14 how those companies trade isn't of the utmost importance. 15 But I would go back to the evolution of the tick pilot 16 and there's obviously been a lot of opinions expressed on 17 the tick pilot pro/con debate over the last few years. 18 One thing I would say is that leading up to the 19 passage of the tick pilot and even subsequent to the tick 20 pilot, I don't believe that there has been a better 21 suggestion made about something to potentially do with 22 market structure around small and mid-cap trading. So 23 even if you think the tick pilot is a bad idea, I'm not 24 sure at least I've heard a better suggestion. In which 25 case, my point would be, it's a pretty dramatic change to 0077 1 how securities trade and I wouldn't think we'd want to 2 try additional experiments on top of the tick pilot. So 3 even if you don't think the tick pilot is a good idea, my 4 suggestion would be that, as the committee, we closely 5 monitor the effect of the tick pilot and if it doesn't 6 yield any positive results, we then continue to study 7 whether there should be another recommendation made on 8 top of it. But I wouldn't think we'd want to pile on to 9 something so significant at this point in the -- in the 10 experiment. 11 MR. SHILLMAN: Joe Ratterman and then Brad. 12 MR. RATTERMAN: Yeah, I couldn't agree more 13 with Rick. We were both there and we were faced with 14 what was a pretty amazing event in the markets. And I 15 think what we did, appropriately so at the time as an 16 industry was a shotgun approach to fix the problem that 17 we saw. Let's get something done with individual stocks 18 because that doesn't exist. Let's do something with 19 these market-wide circuit breakers because obviously 20 they're too wide. We don't want that to happen again. 21 We were all stinging. 22 And I think it was absolutely appropriate to 23 kind of shotgun and blast and fix everything we possibly 24 could and stand back and watch what worked and what 25 didn't. With Rick, I agree that the Limit Up/Limit Down, 0078 1 which originally was single-stock circuit breakers, 2 evolved into Limit Up/Limit Down, actually has gone a 3 long way, probably potentially all on its own, to fix 4 what we saw on May 6. 5 So in hindsight, looking back, with 6 single-stock circuit breakers and then Limit Up/Limit 7 Down, I also completely agree that the market-wide 8 circuit breakers can be widened back and softened and 9 readdressed, now that we know that the Limit Up/Limit 10 Down works. 11 And so historically, we pushed down to 7 12 percent and made a bunch of changes because we didn't 13 know that the single-stock approach would work. It did. 14 Now with that and we're optimizing that, let's stand back 15 and re-look at the market-wide circuit breakers. Because 16 really, you want the market-wide circuit breakers only to 17 kick in when the systems will not allow price discovery 18 to happen. And in hindsight, I'm not sure where there's 19 too many events where any of us wanted the markets to 20 have been halted across the board over the last several 21 years. 22 And so, you know, being able to test when we 23 would have wanted it to happen with the systems that we 24 have, I think is appropriate to look at that and to say, 25 no, we didn't want them to happen. So did we get too 0079 1 close with the current rules? If so, let's adjust those 2 rules. 3 MR. KATSUYAMA: So the state Stacey and Jon 4 quoted, the 75 percent decrease, what was the time frame 5 around that? 6 MR. ZECCA: Well, it's only been live for two 7 weeks. So I think you did the same way as we did. So 8 the methodology is by market center. So it would be Arca 9 New York and old AMEX and then on our side NASDAQ. 10 MS. CUNNINGHAM: We compared the first week of 11 July to the week subsequent to the change in reference 12 price. 13 MR. KATSUYAMA: Got it. 14 MS. CUNNINGHAM: And to that point, we moved 15 away from using premarket trading to Joe's point that 16 there is a lot of volatility there. So leveraging the 17 close from the day before singlehandedly reduced the 18 number of halts. 19 MR. KATSUYAMA: Okay. And so it doesn't 20 include prior event -- like August 24 is not in the data 21 set? 22 MR. ZECCA: Right, it was just a week-on-week 23 comparison. 24 MR. KATSUYAMA: Just a week-on-week thing? 25 Okay. 0080 1 Yeah, I mean, I think in general, a lot of the 2 -- I think the recommendations from the subcommittee, I 3 think, are all sensible. I think that we had an event. 4 We were able to use that event as a learning experience. 5 I think the things they've suggested make sense. And 6 taking a broad range of perspectives, I think, you know, 7 again, any time you can get a collective participant 8 recommendation as opposed to one specifically from any 9 particular trading centers, I would always lean more to 10 the -- I would lean more to the broadest sense of 11 perspectives than I would to the narrowest. So I think 12 what they've suggested is sensible. 13 I think that, you know, again, a very narrow 14 sample size and a very narrow statistic, although it's 15 great that they've reduced 75 percent, I don't think that 16 the sample that we're dealing with really reflects on the 17 event that we're trying to prevent; i.e., August 24. So 18 again, fully supportive of what I read and I just think 19 it's a good reflection and a sensible reflection of a lot 20 of the concerns that I've heard since then, since that 21 date. 22 MR. SHILLMAN: Stacey? 23 MS. CUNNINGHAM: Thanks. I actually just 24 wanted to highlight that the point of raising the 25 amendment 10 change was just that it was one incremental 0081 1 thing that had a significant impact. The changes that 2 we're proposing that we intend to file short term would 3 have the much more significant impact. And, in fact, we 4 saw a significant reduction in LULD halts, even looking 5 at the day following the Brexit referendum. And part of 6 that was also because of changes that we had made as an 7 industry by adjusting caller behavior. And that was just 8 one incremental step. And we didn't see the number of 9 LULD halts that we saw in August 24. 10 So our point is that we can make a number of 11 improvements that will have significant impact and these 12 are intermediate steps. We can still consider extending 13 the limit state. We think that's certainly worthwhile. 14 But each on their own, they also have impact. 15 MR. SHILLMAN: Gary. 16 MR. STONE: So, Eric, I just have a question. 17 I think it's the same one that Chair White had in the 18 beginning. What is the revert price or the snap-back 19 price and how do you determine what it is? 20 MR. NOLL: So it's a good question. But I 21 think where it comes from is, you know, it's obviously 22 the reference price that is used to trigger the band. So 23 if a stock is trading for 100 and the band is then 91.10, 24 around 100, 100 is the revert price. So if it would 25 trade down and hit 90, it should be allowed to trade back 0082 1 up to 100 without triggering another halt. 2 I had a question, it's probably a similar 3 question to Brad. When you said you reduced LULD 4 conditions 75 percent, was that in total? Someone said 5 75 percent is spurious LULD, and I just want to make sure 6 we're talking about the same thing. 7 MS. CUNNINGHAM: Yeah, I think the number was 8 from 40 down to 10 in the early trading coming out of the 9 opening. 10 MR. ANGEL: I can attest those numbers are 11 real, because I monitor those things on a daily basis. 12 And my inbox is no longer filled with spurious LULD 13 halts. 14 MR. SHILLMAN: Go ahead, Gary. 15 MR. STONE: Just another question, when we 16 talked about coordinating market-wide halts, Professor 17 Angel, you talked about the fact that there are 16 18 different SROs. Is -- would something like that be 19 addressed under the LULD NMS plan, or is that something 20 that has to be done separately as another initiative by 21 the Commission or another plan? 22 MR. ZECCA: Certainly, we can talk about it. 23 You know, at the end of the day, however it's done, I 24 think it's going to require the discussion of the SROs in 25 collective agreement. So, you know, sometimes that's 0083 1 done by doing separate rules, as has traditionally been 2 with clearly erroneous. Market-wide circuit breaker 3 sometimes is done through a plan. 4 Could we move it under something like LULD? 5 Obviously, we would have to talk about it. I think it's 6 theoretically possible, but it would require an amendment 7 to the plan. So it would be itself something of a 8 complex process, versus just fixing the rules. 9 MR. SHILLMAN: Any other questions from the 10 committee, the commissioners or the Staff? 11 If I could just ask a factual question, Eric, 12 on the quote only period, it's different from a -- 13 trading can't occur during the quote only period, but 14 it's different from a halt because interest that is 15 present in the market stays in the market, imbalance 16 information is disseminated. But then at the end of the 17 period, the gate is lifted and trading just resumes, 18 there's not a formal opening? 19 MR. NOLL: Exactly. That's what we have in 20 mind. Which is very similar to the way the futures 21 market works today. 22 MR. SHILLMAN: And the experience is -- 23 MR. NOLL: -- unfragmented market. But we 24 think it solves the problem of when you have dispersed 25 trading interest across multiple venues, continuing to 0084 1 reflect that in the marketplace, most market participants 2 would be reading that in from all the different various 3 venues, they'd be able to see that and make a 4 determination on how they would want to participate. And 5 so essentially just lift the gate and start trading 6 again. 7 MS. CUNNINGHAM: I guess one of the things that 8 we just were considering as we thought about that process 9 as well is just the impact on retail investors. So if 10 algorithms step away and market makers step away and 11 you're left with retail investors and you trade resume 12 with just that complexion there, how does that work out 13 for retail? And I think our retail investor profile in 14 the equity markets is more significant. And certainly 15 the feedback that we heard from them was that they wanted 16 to know that they had a fair place. 17 So we think, we just want to explore what that 18 trade resumption would look like and what the impact on 19 retail would be as well. 20 MR. NOLL: And I think that that's a good idea. 21 I will say that we did have retail firms come and talk to 22 us about this and they were comfortable with there not 23 being an auction process to reopen in an LULD condition. 24 That may have changed. But we certainly had ample 25 representatives from the retail firms there and gave them 0085 1 every opportunity to participate in this process. 2 MR. CRONIN: And again, the two conditions. 3 Time, how much time is available for them to react, and 4 the information. Right? If retail investors, like any 5 participant in the market, have those two things, again, 6 we'll get to a better answer. Today, I think that it's a 7 stretch to say that that's appropriately in place. 8 MR. RATTERMAN: And just to comment on the 9 time, you know, we talk about 10 seconds, 20 seconds, 20 10 minutes, two minutes. I think almost any amount of time, 11 you know, two to five, 10 minutes would be fine, because 12 we still have the option to call a news halt, for 13 legitimate reasons to halt a security. And so if you 14 don't have that pending, then it's just simply a price 15 formation mechanism that takes time to heal and the 16 amount of time that you allow it to heal, I think, could 17 be pretty long, significantly longer than today, because 18 you can always halt the stock for legitimate reasons for 19 a news halt or other serious condition. 20 MR. NOLL: That's a good clarification, Joe. 21 Because I think one of the things that we wanted to be 22 clear about, this has no impact on an exchange's 23 regulatory ability to make a pending news halt or other 24 regulatory reason to halt the trading of a stock. This 25 is just around the Limit Up/Limit Down mechanism. 0086 1 MR. SHILLMAN: Okay, well, it's been a good 2 discussion. I guess as far as the next step, I'd leave 3 it to Eric and the subcommittee to determine whether you 4 want to revise your recommendation to the full committee, 5 or go straight to a draft recommendation, from the full 6 committee to the Commission, that could then be voted on. 7 But we could then plan on doing what we did -- 8 similar to what we did for the other two subcommittees. 9 Which is, at the next meeting, whatever document you 10 produce, which would of course be publicly available and 11 people could comment on it, could then be acted on at the 12 next full committee meeting. 13 Okay, so I think with that, we're a little bit 14 early but we're at the lunch break. And why don't we 15 convene back here at 12:30? Thanks. 16 (Whereupon, a luncheon recess was taken.) 17 A F T E R N O O N S E S S I O N 18 MR. SHILLMAN: Okay, welcome back. 19 This afternoon's session will follow the same 20 format as this morning's with an overview of the 21 subcommittee's recommendation, panelist statements and a 22 full committee discussion. The focus this afternoon, of 23 course, is a Customer Issues Subcommittee and Manisha, as 24 chair of that committee, would you like to give an 25 overview of your recommendation? 0087 1 MS. KIMMEL: Great. So, as you recall from our 2 April status report, we focused on three -- four primary 3 topics. One is understanding the retail investor 4 perspective on equity market structure. The second was 5 the use of stop order types by retail customers. The 6 third was a review of execution quality statistics. And 7 the fourth was payment for order flow. 8 Since our April meeting, we note that FINRA has 9 released Regulatory Notice 1619 relating to the use of 10 stop orders in periods of extreme volatility, and that 11 this guidance is consistent with the subcommittee's 12 recommendations presented in April. 13 With respect to payment for order flow, the 14 subcommittee has not come to any specific 15 recommendations. However, discussions spanned several 16 topics, including prohibiting the use of payment for 17 order flow, managing the conflicts of interest, and 18 opportunities for enhancing disclosures. 19 We note that the recently issued order handling 20 proposal does include new disclosures relating to payment 21 for order flow, and we believe that that's the 22 appropriate place for those comments to be -- to be 23 addressed. As part of its deliberations, the subcommittee 24 has sought input from market participants including 25 retail broker-dealers, industry groups and members of the 0088 1 SEC's Investor Advisory Committee. We've received and 2 reviewed statistical information on retail execution 3 quality and note that retail execution quality has 4 improved over the last several years. 5 While we have begun reviewing the Commission's 6 order handling proposal and are positive that the 7 comments relating to the need to modernize Rule 605 and 8 Rule 606, we do believe that additional modifications to 9 Rule 605 and 606 are warranted as part of that process. 10 Additionally, the subcommittee has discussed 11 performing surveys in order to benchmark and monitor 12 investor confidence, as described in our first 13 recommendation. 14 So our first recommendation is to benchmark and 15 monitor investor confidence in the U.S. equities market 16 structure. And this, our goal here is really that the 17 SEC have an informed and current view of the state of 18 individual investor confidence in our equity markets, and 19 that they should periodically sponsor or conduct surveys 20 of individual investors to establish benchmarks and 21 trends over time for investor confidence in the U.S. 22 equity markets. 23 We note that there are existing academic 24 studies and industry studies on confidence, stock market 25 participation levels and investor trust in the markets 0089 1 that should inform the SEC's view. And, you know, we 2 have reviewed several -- several of the academic research 3 related to this and -- and recognize that it is a fuzzy 4 concept. And so as part of that, we really think that 5 the survey information should be used, you know, to be 6 very specific, to understand views on market structure, 7 on disclosures and volatility and the implications of 8 those views on investor behavior, asset allocation 9 decisions, readiness for retirement and perceptions on 10 advisability of risk associated with investing in 11 equities. 12 So make those -- to make recommendation one, 13 you know, more concrete, the last piece of this is that 14 we believe that these surveys should include testing the 15 usability, clarity and effectiveness of disclosures with 16 a representative sample of investors and that, as part of 17 the feedback from those surveys, that both disclosures 18 and rulemaking would be revised in response to the 19 testing results. That's our recommendation one. 20 Our second recommendation is around modifying 21 Rule 605 and 606 to provide meaningful execution quality 22 and order handling disclosures from a retail perspective. 23 We note that the recently released order handling 24 proposal covers two more aspects than is currently in 25 these recommendations. Those include the definition of 0090 1 institutional and retail orders, as well as institutional 2 order disclosures. At this time, we are providing 3 recommendations on retail order disclosures. 4 So our first recommendation is to improve the 5 accessibility of Rule 605 and 606 reports. We believe 6 that the SEC should be the central place where Rule 605 7 and 606 reports are made available. And that not only 8 should these reports be available, but they should be 9 incorporated into the Rule 605 -- into the currently -- 10 the SEC's data visualization tool, so that you would have 11 the ability to see aggregate information across retail 12 broker-dealers as well as market makers, et cetera. 13 Our next part of the recommendation relating to 14 605 and 606 is expanding the scope of Rule 605 by 15 requiring every broker-dealer to report -- to report 605 16 data with a de minimis exemption for -- with an exemption 17 for broker-dealers with de minimis order flow. So today, 18 Rule 605 requires that market centers publish these 19 reports. Broker-dealers who route order flow to other 20 execution venues only produce Rule 606 reports. 21 While we recognize there would be both 22 compliance and implementation costs, we believe that the 23 use of third party vendors may mitigate some of these 24 concerns. By having all broker-dealers report 605 data, 25 there would ben an opportunity for market participants, 0091 1 academics and the press to evaluate these statistics in a 2 consistent manner. And consistent with the 3 recommendation made earlier of being able to have all 4 that data in one place, you could also have the 5 aggregated statics which we've heard from retail 6 broker-dealers, you know, and from the use of this data 7 in the press is beneficial. 8 To further improve standardization, the 9 consistency of reporting, the SEC could consider 10 centralizing report creation in an unbiased trusted 11 source such as FINRA. This again could also make 12 implementation potentially easier. 13 For market data driven calculations in Rule 605 14 and 606, we recommend requiring that calculations be 15 based on either SIP or proprietary data feeds, depending 16 on what feeds were used for routing and execution. The 17 disclosure should include identifying what type of market 18 data is used in the report. 19 We also recommend modifying Rule 605 to reflect 20 changes in market structure since the rule was initially 21 adopted. That would include the following changes: 22 Inclusion of odd lots; segregation of immediate or cancel 23 orders, including the following new data elements, quoted 24 spread and enhanced liquidity; expanding the execution 25 time buckets to include more granular measurements, 0092 1 including less than 50 milliseconds, 50 milliseconds to 2 500 milliseconds, 500 milliseconds to one second. And 3 the time, where we got the 50-millisecond structure is 4 based on the clock synchronization standards established 5 by both FINRA and the future consolidated audit trail. 6 We also believe that the FAQs should be revised 7 regarding the methodology to assign an NBBO for orders 8 received in the same second as a quote change such that 9 all firms reference the NBBO from the closest millisecond 10 prior to the receipt of a covered order. Today, the FAQ 11 says that any neutral algorithm can be used. 12 We believe we should expand the scope, as it 13 relates to 606 now, expand the scope of Rule 606 14 reporters to include exchanges and ATSs that route out. 15 And then specifically, with respect to 606 modifications, 16 also to reflect changes in market structure since the 17 rule was initially adopted, we recommend instead of 18 dividing data by listing markets, divide by S&P 500 and 19 other NMS equities. We note that the proposal, the SEC 20 proposal currently just recommends breaking out NMS 21 stocks, although they do ask a question about dividing by 22 S&P 500 or other NMS equities, we believe that should be 23 done, including a new section for OTC equity data. And 24 then segregating data currently in the other categories 25 such that there would be separate columns for market 0093 1 open/close orders, stop and stop-on-quote orders and, 2 based on those attributes, put odd lots in the right 3 category, so that you're really minimizing the use of the 4 other category to the greatest extent possible. 5 Next, we would recommend segregating marketable 6 limit orders from nonmarketable limit orders. That's 7 also in the current proposal. We agree with that, 8 recognizing that the use of market data in determining 9 limit orders and nonmarketable orders is not something 10 that 606 reporters currently do today. 11 We'd also recommend consistently identifying 12 routing destinations across Rule 606 reporters to allow 13 for better aggregation. Including plain English 14 descriptions, definitions and FAQs. And finally, to 15 include execution quality statistics by routing 16 destination. For example, effective spread over quoted 17 spread. 18 You know, recognizing that different decisions 19 are made with respect to routing decisions, we think that 20 the segregation of S&P 500 and other NMS securities might 21 help to, you know, avoid any anomalies there. 22 As mentioned, there's two other aspects to the 23 order handling proposal that are not in this current 24 recommendation. Those include the definition of retail, 25 institutional and the institutional order routing 0094 1 disclosures. We expect to have recommendations on those 2 forthcoming. 3 Thank you. 4 MR. SHILLMAN: Thanks, Manisha. Why don't we 5 move on to our panel and let me quickly introduce them. 6 This afternoon we have, again going from left 7 to right, Bill Alpert, Senior Editor at Barron's. Next 8 to Bill is Frank Hatheway, Chief Economist at NASDAQ. 9 And next to him is Chris Nagy, director of the Healthy 10 Markets Association. 11 So why don't we start with Bill? 12 MR. ALPERT: -- speak. So I'm not speaking for 13 my employer, unlike most of the people here. 14 A quick intro. We've been around for almost a 15 hundred years. I've been around for more than a third of 16 that. I think my career stuck. Let's see. 17 When Clarence Barron owned the Wall Street 18 Journal and started us the year before that, he had 19 exposed Charles Ponzi. Investor protection has always 20 been an important part of what we do and he did it by 21 doing the math, which was kind of cool. 22 I've tried to send work the SEC's way. After 23 this 2010 story, they started an internal task force to 24 investigate U.S.-listed Chinese companies that might have 25 been abusing the reverse merger process to come here. I 0095 1 know that post hoc doesn't mean propter hoc. But that's 2 what happened. 3 I'm a data dilettante. I've long made noise 4 about how the SEC shouldn't just collect comments from 5 lawyers like me, but should do back test simulations. 6 And as a demo, we commissioned a couple of Yale 7 researchers in 2011 to do a back test of the proposed 8 stock-by-stock circuit breakers. And they applied the 9 Limit Up/Limit Down rules to the previous three years of 10 trading in U.S.-listed stocks. It didn't require 11 statistical rocket science. It did involve about 8,000 12 hours of cloud computing processing on about 60 13 processors to go through the 24 billion trades. But the 14 point was that if we kibitzers could do it, then the U.S. 15 government could do it. 16 So I'm still trying to figure out why I'm here 17 talking to all you experts. I did write this story last 18 year and so I think I'm here to show you what a lousy job 19 someone could do comparing brokers under the existing 20 rules, and then maybe you'll feel more compelled to 21 modify them. 22 So my editors asked me to look at the best 23 selling claims that mom and pop investors were being 24 scalped by high-frequency traders. And I recognized that 25 as an entrenched battlefield that I was wary of walking 0096 1 into. And as a reporter, you're always worrying about 2 being somebody's patsy or shill and trying to see past 3 the staged factory tours that people take you on. So I 4 didn't want to take anybody else's work. I stupidly 5 decided to try to use the Rule 605 and Rule 606 data. 6 And to take things a step further, I wanted to come away 7 with something useful for our readers. So I thought I'd 8 try to rank the brokers and market makers using what has 9 been available. 10 I open sourced my analysis on GitHub, so that 11 our calculations would be available for replication and 12 criticism. I had to make very rough inferences about the 13 brokers' executions because of the gaps in the disclosure 14 requirements and it's those gaps that we're here this 15 afternoon to discuss. 16 The market makers were straightforward to 17 compare. It required some data processing, but Rule 605 18 disclosures allowed you to compare their price 19 improvement, their effective over quoted spreads. 20 As you know, the Rule 606 reports give no 21 direct information on execution quality. So they leave 22 out lots of kinds of orders. They are just not suited 23 for direct measurements of a broker's execution 24 performance. 25 Some of the brokers voluntarily showed 0097 1 execution quality measures on their websites. But each 2 one used a different measure, so you couldn't compare 3 them. So the only effective, objective way to use the 4 available disclosures was to score each broker with a 5 weighted sum of their order flow fractions from the 6 routing reports and then weight those with the effective 7 overquoted measures of the market makers that they were 8 sending their orders to. 9 And, of course, you know, we're all discussing 10 here proposals to directly measure the now-exempt 11 brokers. And that will save future journalists from 12 having to write thousands of lines of computer code. So 13 I would recommend that on that basis alone. 14 So, but more seriously, a fundamental problem 15 with the inferences that you can make on broker 16 performance is that a market maker's average execution 17 across all of its sending brokers may be better or worse 18 than its performance on a particular broker's flow. And 19 almost every broker claimed that, like the kids in Lake 20 Wobegon, their executions at a particular market maker 21 were above average for that market maker's general 22 measure. And at least one broker scolded us for even 23 trying to use the existing information to do a comparison 24 like this. 25 You also know that brokers don't report 100 0098 1 percent of their order routing on the 606 forms. They're 2 looking at market orders, market limit orders. So one 3 broker got a lot of market orders while others had more 4 limit orders. And so that also kind of taints the 5 comparison. 6 I was happy to see that even, given that rough 7 and ready calculation, when a few months later a few 8 brokers and market makers started their voluntary 9 disclosures under the auspices of the Financial 10 Information Forum, the rankings were roughly right. But 11 there's a nicer lesson from those voluntary disclosures. 12 Our security regulation here in the United States is 13 based on nicely agnostic belief that, with adequate 14 disclosure, the chips will fall where they may and 15 they'll usually fall in the right place. So here's 16 evidence for that belief. 17 The voluntary disclosures were the first time 18 that odd lot orders had been reported. And those are the 19 little, bitty light blue bars that you see under the 20 arrows there. And as you can see, the price improvement 21 on those were much inferior for most of the voluntary 22 reporters than the other long-disclosed light blue 23 categories, other order sizes. I show a couple of them 24 there. So that shows you how disclosure improves 25 people's behavior. 0099 1 So the dark blue bars are a year later, this 2 past March quarter. And so you can see that, for this 3 particular broker, and the same was true for all the 4 other brokers and market makers, there's much bigger 5 price improvement on those odd lots. And indeed, every 6 category is better a year after they started these 7 voluntary disclosures. 8 So, you know, when everybody is looking at your 9 Fitbit score, you work harder. 10 These are selective disclosures. Only a few 11 brokers and market makers are making them, so a mandate 12 would be nice. 13 And I think the proposals look pretty good to 14 me. I'm just a hack writer. 15 Some suggestions from my ordeal in trying to 16 make use of the old disclosures. Text files should 17 always be available for any data you're providing to the 18 public. They're nonproprietary, they're future proof, 19 they can be read by anyone and anything. It allows 20 firms, academics, journalists, citizens to use them in 21 modern data analytic software. The Rule 606 disclosures 22 at a lot of firms now are on PDF files. And, you know, 23 when data aren't machine readable, it discourages 24 analysis and comparison. 25 And even on the 605 filings, market centers use 0100 1 idiosyncratic names. There are empty columns and other 2 peculiarities in the tables that make them difficult to 3 process with a computer script. And, of course, they're 4 scattered all over the Internet and ought to be 5 centralized somewhere at the SEC or FINRA, somewhere. 6 In the case of one of the biggest brokers, they 7 weren't available anywhere, the 605 filings, even after I 8 reminded them over the course of weeks and weeks. 9 The whole history should stay online. Right 10 now, only about three months are left on there. And data 11 storage is cheap. There's no reason not to leave 12 everything up there. 13 So, thanks. 14 MR. SHILLMAN: Thanks much, Bill. 15 Frank. 16 MR. HATHEWAY: Thank you, David. Good 17 afternoon, Chair White, commissioners, committee members. 18 I thank the Commission and the committee for inviting me 19 here today to discuss customer issues. NASDAQ 20 appreciates the Commission's and committee's thoughtful 21 consideration of the subcommittee's work and looks 22 forward to actively participating in the dialogue of the 23 specific proposals that will emanate from this process. 24 As head of the team that produces data on our 25 U.S. and European markets, and also analyzes data from 0101 1 many other markets, I look forward to talking about the 2 topics we have before us this afternoon. Investors must 3 have confidence that our markets operate in a manner that 4 serves and protects their interest. This confidence 5 enables our listed companies to raise capital, finance 6 economic growth. 7 Since its earliest days, the Commission has 8 rightly concluded that earning investors' confidence 9 requires a level of transparency beyond that which is 10 required in other industries. NASDAQ supports this 11 transparency and respect for our many operations, whether 12 it takes a form of producing reliable market data, 13 describing our operations in our rule book, regulating 14 the disclosure obligations of listed companies, or 15 producing detailed reports to be consumed by the 16 investing public. 17 The subcommittee's recommendations include some 18 needed enhancements in the disclosure requirements of 19 Rule 605 and 606 of Regulation NMS. Changes we have 20 previously supported for updating Rule 605 include finer 21 speed measurements, consistent definitions of market 22 centers and covered orders, and coverage of the life of 23 each and every order that an investor submits, or a 24 retail investor submits. 25 Rule 606 also needs to better reflect today's 0102 1 markets, and we welcome the subcommittee's efforts to 2 modernize Rule 606 disclosures. We will also digest and 3 respond to the Commission's recent proposing release on 4 disclosure of order handling information. Finally, 5 NASDAQ takes a more cautionary note toward the 6 recommendation of the SEC benchmark and monitoring 7 investor confidence in market structure. 8 We also share the subcommittee's view that 9 payment for order flow is an important customer issue in 10 today's markets. Like the subcommittee, we know that the 11 order handling disclosure proposing release addresses 12 payment for order flow and includes a number of questions 13 about potential disclosure obligations. We'll reserve 14 comment on this issue until we've reviewed the 15 Commission's release. 16 In our view, Rule 605 and 606 have had a 17 positive impact on execution and routing practices since 18 their implementation. Bill gave a quick example about 19 odd lots. 20 Beyond that, the rules have provided valuable 21 comparative tools for brokers that handle orders on 22 behalf of retail investors. While individual retail 23 investors generally don't review 605 statistics 24 themselves, for reasons Bill alluded to, the existence of 25 the reports appears to provide precisely the form of 0103 1 discipline that the Commission envisioned when it adopted 2 Rule 605 and 606. 3 Based on our review of the Rule 605 and 606 4 reports, however, it's not clear that the reports under 5 these rules continue to provide the level of transparency 6 necessary to exert pressure on market centers to provide 7 superior execution quality and routing. The rules have 8 lagged behind technological advances, enhanced market 9 quality and, consequently, in some cases they render the 10 metrics less useful than they might otherwise be. 11 NASDAQ, as co-chair of the Tick Size Pilot Data 12 Subcommittee -- that's a mouthful -- has been deeply 13 involved in the current effort to produce execution 14 quality statistics on the stocks covered by the tick size 15 pilot. The pilot will result in the most detailed 16 execution quality statistics ever to be made public by 17 the SROs and their members. While the requirements of 18 the pilot should not be taken as a template for updating 19 Rule 605, the creation of the execution quality 20 statistics for the pilot has been very instructive with 21 respect to obtaining and processing order and execution 22 data from a wide range of trading centers. 23 Turning to the specifics of the subcommittee's 24 recommendations, there are two recommendations with 25 respect to Rule 605 data from recommendation two. The 0104 1 first is the scope of Rule 605 data to be aligned with 2 Rule 606 by requiring every broker-dealer, except for 3 those with de minimis order flow, to report Rule 605 4 statistics. We support this recommendation. Rule 605 5 reports miss important information about the overall 6 execution quality of a covered order, because they only 7 cover routing handled by market centers. Orders are now 8 handled by smart order routers, as Bill mentioned, that 9 may or may not sit within a market center. Any price 10 slippage or delays may occur as the order is received by 11 multiple non-executing market centers or broker-dealers 12 is not captured in the data. 13 The second recommendation is to expand the 14 statistics required under Rule 605 and increase the 15 granularity of Rule 605 reports by segregation of 16 immediate or canceled orders from covered orders. I want 17 to take a minute on this point. 18 At a very high level, I question whether 19 continuing to mandate disclosure of aggregate statistics 20 is a worthwhile endeavor. As Bill mentioned, the goal in 21 evaluating market centers is to evaluate their execution 22 quality performance for orders that essentially are like 23 mine, either as an individual or as a broker-dealer. 24 That evaluation is very difficult with aggregate data, 25 which only offers statistics on the average order. 0105 1 The pilot will produce detailed, disaggregate 2 information on every market and every marketable limit 3 order received by a trading center. 4 You're going to need your programmers again, 5 Bill. 6 The pilot will also produce aggregated 7 statistics with a much greater level of detail and 8 granularity than Rule 605 or, frankly, in the 9 subcommittee's recommendations. The marketplace of ideas 10 is going to show which type of data is more informative. 11 I encourage the Commission to consider the 12 relative usefulness of data produced under the pilot in 13 the different guises it has, as they decide whether to 14 add more detail and granularity to Rule 605 aggregate 15 data or pursue an alternative based on more disaggregated 16 data. If the Commission decides to increase the level of 17 detail and granularity of aggregated data, however, 18 rather than require disclosure of disaggregated data, 19 then we support the subcommittee's recommendations. 20 The subcommittee makes three recommendations 21 applicable to both Rule 605 and 606 data. The first is 22 to improve accessability. Bill mentioned how difficult 23 it was to get the data. It's also difficult to come up 24 with ways to visualize the data. That said, the SEC's 25 data visualization tool essentially addressed a market 0106 1 failure where exchanges' order book data required a 2 substantial infrastructure investment to process. That's 3 not the case for Rule 605 and 606 data, where a number of 4 data vendors offer visualization tools. Before replacing 5 existing vendors' offerings, the Commission may want to 6 consider other alternatives for making the data widely 7 available and accessible. 8 The second recommendation is to improve 9 standardization and consistency of 605 and 606 data by 10 considering centralized report creation by an unbiased 11 and trusted source. NASDAQ has previously stated and 12 reiterates today that, to make informed decisions, 13 investors need consistent information on similarly 14 situated orders being handled by competing venues. Based 15 upon our review of 605 reports, it appears that market 16 centers do not define covered orders consistently and do 17 not handle their preparation of statistics consistently. 18 Experience with the pilot, however, also shows 19 consistency will not be 100 percent achievable even when 20 report production is highly centralized -- excuse me, is 21 highly coordinated or even centralized. This is the case 22 for the pilot. Consistency may be a goal that the 23 Commission and the committee determine can be better met 24 in the context of the consolidated audit trail 25 initiative. 0107 1 The third recommendation is for the Rule 605 2 and 606 calculations to depend on the feeds used for 3 routing and execution. Like the previous recommendation 4 on consistency, my work on the pilot reveals that this 5 recommendation has -- while it has a strong, intuitive 6 appeal is at odds with the incumbent IT infrastructure 7 for many trading centers. Furthermore, a recent academic 8 study from UC Berkeley has shown the differences between 9 using SIP and proprietary data feeds to measure execution 10 quality are exceedingly small. The Commission and the 11 committee should carefully evaluate whether the benefits 12 of mandating the use of routing and execution data feeds 13 for execution quality measurement outweigh the costs. 14 There are two specific recommendations in 606. 15 To an extent, this commendable effort has been superceded 16 by the proposing release on disclosure of order handling 17 information. As a general matter, NASDAQ supports 18 consistency between reporting obligations pursuant to 19 Rule 605 and 606, and so we support the recommendation 20 and the scope of Rule 606 disclosures be included -- be 21 expanded to include exchanges and ATSs. 22 The second recommendation for Rule 606 data 23 calls for removing the listing market classification, 24 increasing granularity with respect to OTC equities and 25 various order types that are out there, improving 0108 1 consistency of destination identification, plain English 2 documentation, and linking information in the Rule 606 3 reports with information in Rule 605 reports. We are 4 broadly supportive of all these recommendations. 5 Finally, and to return to the first 6 recommendation, NASDAQ shares the committee's belief that 7 the Commission would benefit by knowing whether investors 8 are operating in their interest. In my view, surveys of 9 individual investors, even sophisticated individual 10 investors, are strongly influenced by recent investment 11 performance, market volatility, news headlines, 12 macroeconomic conditions and other factors which are best 13 only loosely related to market structure, business models 14 or corporate disclosure. 15 Furthermore, a great deal of information on 16 investor sentiment can be gleaned from existing public 17 sources and surveys. I encourage the SEC to fully 18 explore these alternatives as a cost effective means to 19 achieve the subcommittee's first recommendation before 20 sponsoring or conducting its own survey. 21 I thank the Chair, the commissioners, the 22 committee for your time and attention. We appreciate the 23 committee's thoughtful consideration of these issues and 24 welcome the opportunity to continue to work with the 25 Commission and the committee to consider important 0109 1 changes in market structure for the benefit of investors 2 and listed companies. I look forward to your questions. 3 MR. SHILLMAN: Thanks, Frank. 4 Chris. 5 MR. NAGY: Thanks, David. 6 Chair White, Commissioner Stein, Commissioner 7 Piwowar and members of the MSAC committee, thanks for 8 inviting me here to speak today. This is an issue that 9 has been near and dear to my heart for a long time, so 10 I'm excited that I have this opportunity. 11 I'm speaking to you today as director of the 12 Healthy Markets Association, but I would like to point 13 out that my remarks are mine and not necessarily those of 14 the members or the funders. I would also add that the 15 written testimony that I've submitted is quite a bit 16 longer than what I've spoken here, and that's probably to 17 your relief. So I will invite you all to review my 18 written testimony as well. 19 So after the SEC adopted the forerunner to Rule 20 606 almost exactly 16 years ago, I witnessed firsthand 21 the beneficial impact of the reports on competition and 22 behavior in the marketplace. Unfortunately, fundamental 23 changes in how securities are traded over the past two 24 decades have rendered these reports nearly obsolete. 25 As many of you know, I think this is where Rule 0110 1 606 comes in. I have long argued that modernizing Rule 2 606 could promote transparent competition, reduce 3 conflicts of interest, promote beneficial behavior 4 changes. I have attached with my written testimony the 5 Healthy Markets proposals contained within that as well, 6 too. I think it's Exhibit 1. 7 Essentially, all investors deserve to know how 8 their orders are being handled and executed. The SEC and 9 your proposal, MSAC's proposal, could also help with 10 that. But I also worry that the SEC's proposal has some 11 flaws in it that, if not properly remedied, may severely 12 undermine its efficacy. 13 The first issue would be institutional versus 14 retail orders, which was touched briefly upon. So the 15 Commission proposes to dramatically expand the coverage 16 of Rule 606 reporting, which is great. But it would also 17 create a new distinction between institutional orders and 18 retail orders. 19 We believe the SEC should abandon this 20 artificial distinction and apply its proposal equally to 21 all investors. I think investors of all types and sizes 22 deserve to know how their orders are handled and 23 executed. Worse, this misguided dichotomy between retail 24 and so-called institutional is also poorly implemented. 25 While the apparatus surrounding the execution of 0111 1 so-called retail orders has evolved differently than the 2 apparatus for trading of large institutional orders, the 3 results should be the same for both, best execution. 4 Further, the arbitrary distinction misses two 5 key facts of the modern marketplace. First, a large 6 number of smaller institutions may trade through retail 7 brokerage channels, which they use for order execution 8 and custody. These institutions, who have the same 9 obligations and responsibilities as their larger 10 brethren, should not be denied access to the same 11 important information as their larger peers. Second, and 12 perhaps just as importantly, the SEC's proposed 13 distinction assumes that large institutional investors 14 send relatively large orders to their brokers, such as 15 large block trades common in decades gone by. 16 While some of this still does occur, the 17 instances are far less than years past. Today, many of 18 the most sophisticated investors often feed brokers 19 smaller orders over time, or simultaneously utilize more 20 than one broker in advanced order routing strategies to 21 avoid information leakage and higher trading costs. 22 These orders may often be individually less than the 23 $200,000 proposed within the proposal, and so they would 24 not be treated as institutional orders under the proposed 25 rule. Even further, investors may seek to deliberately 0112 1 avoid being identified as institutional orders, and so 2 may use this arbitrary distinction as a way to do that. 3 We firmly believe that any share of 4 dollar-based distinction between customer order types is 5 fraught with these perils, which are likely to make the 6 resulting report of limited utility. I urge the SEC and 7 the MSAC not to go down this path. 8 The second issue I would like to discuss is 9 centralized reporting and historical data. Perhaps one 10 of the greatest challenges to anyone seeking to analyze 11 and review 606 and 605 data, as we heard earlier, is the 12 fact that the information relevant to a large investor or 13 broker is likely scattered on hundreds of diverse 14 websites. Brokers and increasingly investors are likely 15 to want to use this information as part of their best 16 execution review practices. By leaving this information 17 scattered to the corners of the Internet, the laudable 18 goal is rendered nearly impossible for all but the most 19 well resourced firms. 20 Centralized reporting will also help reduce 21 incorrect or missing data, which could then be easily 22 spotted and checked. Currently, these reports are often 23 laden with errors. In fact, it was mentioned earlier, 24 back in 2010, or of the 397 firms I reviewed in 2010, 25 which it was submitted as part of my letter on the 0113 1 concept release, I found only that 22 percent of those 2 firms were accurately meeting their disclosure 3 requirements. Even as I prepared for my testimony today, 4 I found firms with incorrect links to old and obsolete 5 data or incorrect naming conventions. 6 So while the question of who would own such a 7 repository is always a thorny one, I think one 8 possibility could be a centralized reporting structure in 9 a manner similar to how FINRA recently began publishing 10 the OTC and ATS statistics. And I see Rick smiling over 11 there. 12 Finally, just in terms of increased 13 flexibility, I wanted to talk a little bit about this. 14 So the SEC's proposal addresses many -- many of the 15 obvious shortcomings with the existing reporting 16 requirements. But it doesn't provide for further 17 enhancements or improvements in the years ahead. And I 18 think this is a really important topic. 19 The Commission should recognize that, as well 20 as specific reporting requirements are structured today, 21 they will quickly be outpaced by both technological 22 advancements and regulatory expectations. Many brokers 23 and investors are increasingly looking to perform 24 comprehensive analysis such as parent to broker to algo. 25 to venue-level analysis, but are having an increasingly 0114 1 difficult time getting the data they need. While Rule 2 606 reports could help, the proposed rule wouldn't help 3 many investors, because that customer-specific 4 information would only be provided for so-called 5 institutional orders and only upon request. 6 I recommend that these reports be reframed to 7 provide all investors with all of the relevant data on a 8 periodic basis, while also making targeted tweaks over 9 time. 10 Finally, I wish to highlight just a few of the 11 many issues which are addressed in greater detail again 12 in my written testimony. And obviously we'll also be 13 submitting a very detailed comment letter in the coming 14 weeks. 15 First off, indications of interest, as Frank 16 mentioned here. The SEC should include disclosures and 17 statistics for IOIs by broker-dealers. But these should 18 not be -- these should be reported separately from orders 19 to cut down on the noise. And I know this is a thorny 20 issue. It's something that we certainly want to spend a 21 lot more time with, and I think the Commission accurately 22 recognizes some of those issues. 23 Second, order routing strategies. The SEC has 24 proposed to categorize institutional disclosures under 25 different order routing strategies which are passive, 0115 1 neutral and active, kind of green, red, yellow light type 2 stuff. While this may make sense, this categorization 3 should be as quantitative as possible. 4 The disclosure metrics themselves. So we 5 applaud the Commission for driving the integration of 6 order routing and execution quality metrics in both the 7 individual and aggregate institutional reports. Although 8 we believe that the reporting obligations should be 9 unified between institutional and retail to the extent 10 possible. 11 Then disclosure of net execution fee or rebate 12 or other routing inducements. So we fully support many 13 of the Commission's recommendations regarding the 14 disclosure of fees, rebates and other routing 15 arrangements within the proposal. 16 And then finally just a note on the options 17 markets. We'd urge the MSAC and the SEC to include 18 disclosures about the options markets. These markets 19 continue to be strikingly opaque and are in need of 20 transparency, both from the 605 and further 606 21 disclosures. 22 So in concluding, if we are to modernize and 23 expand 606, we must first get the foundational aspects 24 correct. My above comments are intended to restore 606 25 reports as effective tools for protecting investors and 0116 1 promoting competition amidst changing markets. I 2 encourage you to explore the Healthy Markets 3 recommendations and promise to continue offering my 4 suggestions in the months and years to come. 5 Thank you very much for the opportunity to 6 present and I look forward to any questions that may come 7 up. Thank you. 8 MR. SHILLMAN: Thanks, Chris. 9 I will now open it up to the committee for 10 questions and discussion. 11 MS. KIMMEL: I have a question for Bill. You 12 talked about the reporting you've done on the aggregate 13 statistics related to execution quality of retail 14 broker-dealers and market makers. Can you talk to how, 15 you know, big of an audience you see for those types of 16 stories versus some other stories you write on? Is there 17 a big interest from the retail investor audience that you 18 serve? 19 MR. ALPERT: (Off mic.) 20 MR. SHILLMAN: Bill, you may want to turn on 21 your mic. 22 MR. ALPERT: -- compared to the trading 23 commission. So I think the public has to be brought 24 along and will need to learn about these things. I think 25 that there can be a virtuous circle, as these measures 0117 1 become disclosed, reported, and hopefully the subject of 2 competition between brokers -- they might start 3 advertising them, and that can generate an interest and a 4 demand. 5 So the honest answer is that there's not great 6 pent-up interest yet, I think, among the public. And 7 over time, it should develop. 8 MR. MECANE: I might be asking the same 9 question but, in case's a different answer, I was going 10 to ask what type of reaction -- 11 MR. ALPERT: Maybe there will be a different 12 answer. 13 MR. MECANE: -- what kind of reaction you got 14 from your readership based on the study? So I think 15 you're saying it wasn't as big a reaction as you 16 expected. But did you get any interesting feedback? 17 MR. ALPERT: I think the most interesting 18 feedback was tangential, perhaps from diehard conspiracy 19 theorists who believed and will forever believe that the 20 high-frequency traders are tricking us in ways that we'll 21 never be able to measure. We're living in the matrix and 22 the computers are controlling all of our perceptions. 23 So a lot of bitter complaints. You know, we're 24 not interested in the numbers. We're sure that the mom 25 and pop retail investor is being front run somehow. 0118 1 And then again the discussion about proprietary 2 feeds versus the SIP and whether these measures are all 3 spurious because they're all mostly off the SIP, so there 4 must be something wrong. 5 MR. MECANE: I think we've had a few of those 6 people on the panel. 7 MR. HATHEWAY: Bill, if it makes you feel any 8 better, within a day or two of your Limit Up/Limit Down 9 story running, I was on the phone with Casey King to find 10 out if there was any back testing that he was doing that 11 we should be doing. 12 MR. ALPERT: He was the Yalie. 13 MR. HATHEWAY: Yes, he was the guy at Yale. So 14 at least somebody follows up your stories. 15 MR. ALPERT: Thank you. 16 MR. SHILLMAN: Okay, I think Gary and then 17 Rick. 18 MR. STONE: So, Manisha, you and the committee, 19 great recommendations. I disagree with the first one, 20 though. I'm not sure if we do a survey we're actually 21 going to test what we think we're testing, in that in the 22 academics on the -- you know, the committee can help out. 23 But I've seen studies that talk about how these studies 24 have really correlated with either how the economy is 25 going or how the stock market -- whether or not it's 0119 1 going up. So I'm not sure that that would be a great 2 place, especially in an era of a tight budget, where I 3 would like to see the Commission, in terms of priorities, 4 where they spend their money. I'd like to see them spend 5 more money on Midas and getting futures data and other 6 correlated data in there to look at stuff. 7 But the other -- the other ones are really, 8 really good. 9 My only question to the panels, probably Chris 10 and Frank, more to Frank probably, is that if we take a 11 forward looking at the consolidated audit trail, would 12 that make 605, 606 collection by the market participants 13 redundant? Or could it? 14 MR. HATHEWAY: I think it could. And again, 15 this is drawing out of what we're doing with the tick 16 pilot, which is collecting both within what the exchanges 17 do and within what the over the counter world does under 18 FINRA supervision. There's a hell of a lot of richness 19 that's going to come out of those reports and those 20 statistics, and the enhancements that Rick and his team 21 has had to make to have the tick pilot work. 22 Whether the consolidated audit trail goes in a 23 direction that can replicate that, we'll have to see. 24 And how useful these statistics turn out to be, we'll 25 have to see. But standardization enables me to do things 0120 1 like -- you know, it's 1:25 in the afternoon. Suppose I 2 put in a 300-share market order for Priceline and I get a 3 fill from Fidelity. How good was that? Well, I could 4 conceivably query a data base for market orders that are 5 about that size at 1:30 in the afternoon on a Tuesday and 6 see how other people did. And you can't do that with the 7 aggregate stats. 8 MR. NAGY: So, Gary, I would say that yes and 9 no. I think it helps from the data gathering 10 perspective. So today, we do -- at my consulting firm, 11 we actually do a lot of data gathering for the buy side 12 in terms of helping them analyze their execution quality. 13 Many times, we have great difficulty in terms of getting 14 the algo. and venue-level specific information with 15 respect to that. I think certainly CAT will help in that 16 endeavor. But where CAT won't help is within 606, when 17 you look at some of the fee arrangements, some of the 18 payment order flow arrangements, the disclosure 19 arrangements. There's nothing in CAT that will help with 20 that whatsoever. 21 Furthermore, within CAT, there's not a lot of 22 the quantitative metrics that are going to be available. 23 While you could certainly extrapolate them from some of 24 the CAT data, I don't think it provides a comprehensive 25 framework, is designed for that. And, you know, it still 0121 1 remains to be seen when CAT will actually be in place. 2 So it seems like we might be able to have this in place 3 before CAT. 4 MR. SHILLMAN: I'll also note, CAT will be 5 regulatory only. It won't -- 6 MR. NAGY: Yeah, that's a great point, David. 7 MR. KETCHUM: Chris, first, thank you for your 8 detailed analysis, which I found very helpful. It was a 9 little less detailed from the standpoint of the 10 subcommittee's recommendations, which were in many cases 11 supplemental to the Commission's proposal and more 12 focused on a distinction you don't find useful from the 13 standpoint of institutional versus retail. But more 14 focused, more broadly, with respect to the retail 15 activity. 16 It would be helpful to me at least if you could 17 give a little bit more of a feel of what out of the 18 supplemental pieces that the subcommittee suggests the 19 Commission should consider you think help address some of 20 your critiques of the initial proposal and what don't? 21 MR. NAGY: Sure, Rick. And thank you for that. 22 So I'd start off by saying that the proposal is 23 very comprehensive. It's 309 pages long, it has 183 24 questions. Many of those are very detailed. So, you 25 know, I'm still working through a lot of those in terms 0122 1 of understanding what some of the unintended consequences 2 may be, and so on. 3 In terms of the proposal, I found that there's 4 a lot of areas within the MSAC proposal that we're very 5 agreeable on. I would -- my thought is that, you know, 6 centralized reporting is probably a very key 7 consideration with respect to that. I think it's really 8 important because I think what it will do is help to 9 foster the modernization on a fairly regular basis of the 10 statistics, which I think has been lacking with respect 11 to it. 12 Does that help answer your question? 13 MR. SHILLMAN: Chester. 14 MR. SPATT: So first, with respect to 605 and 15 606, I think the recommendation of the committee is -- of 16 the subcommittee is a great recommendation. I think it 17 kind of pushes both in the direction of trying to provide 18 for more disaggregated data done also focuses a fair bit 19 on how this information is disseminated and actually 20 accessible to the public. So I think that's -- so I 21 think potentially that's very important and will help 22 push and promote competition in the markets and in the 23 routing space. So I'm fully supportive of that. 24 With respect to the first recommendation, I'm 25 quite a bit more skeptical. It's not -- it's not fully 0123 1 clear to me what the -- what is the specific -- it's not 2 so clear to me how doing the survey of confidence will 3 ultimately inform the work of the Commission. I could 4 well imagine that the responses with respect to 5 confidence reflect recent market movements, reflect the 6 populace's frustrations with Washington, with financial 7 regulation in Washington, and a whole range of other 8 factors. And, you know, even with respect to things like 9 asset allocation and investor behavior, which is sort of 10 an area of subspecialty of mine and sort of other fronts, 11 it's not -- it's not clear then how the SEC's mission 12 will be -- will be enhanced by the survey. 13 I do think if one goes down this route, it's 14 going to be important to figure out in advance what is it 15 one is trying to use the survey to really kind of get at 16 and design the survey to enhance that, rather than to 17 some degree looking to the survey to then figure out what 18 the questions -- what the questions are. I think it's a 19 little bit backwards. So, you know, that summarizes sort 20 of the source of my skepticism about the first 21 recommendation. 22 MR. SHILLMAN: Nancy. 23 MS. SMITH: Yeah, I think that needs some 24 clarification. When we were talking about this in the 25 subcommittee, the whole idea is that investors 0124 1 participating in our markets is needed to have the 2 capital formation and to have the liquidity we all want. 3 So the end, I think we all agree with in terms of you 4 want -- you want to know if investors are feeling, well, 5 this volatility we're seeing, the excessive levels of 6 volatility that may come with some decisions we make 7 about market structure issues, is it worth it? How do 8 you weigh it in the balance, because you're always 9 weighing tradeoffs. And if you don't know what impact 10 you're having on retail investors, individual investors, 11 how do you appropriately weigh the impact of the 12 decisions you're making so that, for instance, would you 13 decide differently if you knew investors have stepped 14 back from the market a percentage because they think it's 15 too risky? They've decided that I can't participate 16 here. I'm panicked. I'll lose my hard-earned retirement 17 savings. 18 Are we finding that investors are not 19 optimizing their asset allocation decisions because they 20 decide that equity investing is too risky, probably 21 riskier than many of us would say? We do know that 22 there's a lot of research on individuals being more loss 23 averse than risk averse. So what are the implications of 24 that? 25 So this is getting to an area where, when we 0125 1 started asking questions about, well, what do we know 2 about investor confidence? What do we know about what 3 individuals think about these issues and when the market 4 goes through these big swings, what are the impacts? 5 And I agree at times, it would be a sentiment 6 issue. But we quickly discovered and I discovered it 7 wasn't so much the sentiment issue as how does it affect 8 my perception of the wisdom of investing in our equity 9 markets. So it's a pretty fundamental issue. 10 I certainly think it's hard to parse out these 11 elements. But I do think it's something that is 12 worthwhile. I remember the first meeting here how we 13 talked about you really need to think about individual 14 investors. And somebody -- we all need to put ourselves 15 in their shoes and represent their interests. And I 16 think if we don't know what they're thinking, how do we 17 begin to do that? 18 So I don't think it's a silver bullet or a 19 panacea to know this. But in talking to people who are 20 studying these issues about stock market participation, 21 finding that it's going down overall a bit, why is that? 22 And what are the impacts of what we do and the volatility 23 levels we feel are okay? I mean, at some level, we have 24 a tolerance for more or less volatility. And as we make 25 those decisions, what are the tradeoffs? 0126 1 Does that help, Chester? 2 MR. HATHEWAY: Yeah, if I can make sure I was 3 clear on what I was saying, it was more that the -- not 4 necessary, in my view, to sponsor another survey. Every 5 quarter for our CFO, I put together about six different 6 metrics on investor confidence, with a focus on retail. 7 So it's important to know it. It's just there may well 8 be existing sources that do a pretty good job on both 9 survey level and on conduct. 10 MS. SMITH: And that is something that we 11 discovered, is there is information out there. But how 12 do we bring it together and start looking at it and 13 seeing how we can utilize it in making decisions overall. 14 MR. SHILLMAN: Andrew, Brad and Chester. 15 MR. LO: So I'd like to respond to Chester's 16 comment but also make a clarifying remark about the first 17 recommendation, and also respond to Commissioner 18 Piwowar's opening statement. 19 We're not talking about investor confidence in 20 the same way that I think, Frank, you're talking about 21 it. Investor confidence typically indicates some form of 22 sentiment about market direction. That's not the kind of 23 investor confidence that we were focused on in our 24 subcommittee. Our focus is on investor confidence in 25 market structure, in the integrity and fairness of 0127 1 markets. That's fundamentally different than asking 2 about sentiment with regard to market direction, which 3 may have no bearing on what the SEC does. 4 But facilitating capital formation is central 5 to what the SEC's mission is. And if investors feel that 6 the market is rigged, as certain journalists have 7 claimed, then you might think that's a cause for concern. 8 Rather than letting the SEC's mission and mandate be 9 dictated by the press, might it not be useful to hear 10 from investors directly, as well as institutional 11 investors -- it doesn't have to be limited to retail -- 12 as to whether or not they feel the market is rigged or 13 too volatile or unfair or inefficient in certain ways? 14 So the idea behind the survey is to ask the 15 ultimate end user of financial markets what their 16 thoughts are regarding whether or not markets are fair, 17 equitable and efficient. That was really the motivation 18 behind the survey. 19 With respect to cost, this is not something 20 that the SEC necessarily needs to take on. They can 21 partner with universities or other third party -- neutral 22 third parties to conduct these surveys. But to date, I 23 know of no survey that focuses on investor confidence in 24 market structure and integrity, and I think that's really 25 the point of the survey. 0128 1 To Chester's point about being focused on 2 specific issues with regard to investor confidence, 3 absolutely. That's a very important point. If you don't 4 spend time focusing on the questions, you'll get back 5 garbage. So it does take time to structure these kinds 6 of surveys. But the point is that these surveys should 7 start being done in a thoughtful way to inform the 8 Commission with regard to these issues. 9 Finally, with regard to Commissioner Piwowar's 10 comments, absolutely, investor skepticism should be 11 encouraged. But again, what we're talking about is 12 investor confidence in market structure, not investor 13 confidence in markets themselves. 14 MR. PIWOWAR: But that's different from what 15 Nancy just said. And she was part of the same 16 recommendation. So my point is, it's a nebulous concept 17 here. What are we actually trying to measure here? 18 And Nancy brought up volatility in the markets, 19 right? I mean, that's -- there's two sources of 20 volatility. There's fundamental volatility, which is the 21 volatility of the economy, cash flows, earnings of 22 individual companies. And then there's transitory 23 volatility. 24 And so how will consumer surveys tell us what, 25 you know, is driving consumers about their, quote, 0129 1 unquote, confidence in the market structure, which is 2 more about the transitory volatility? 3 MS. SMITH: Yeah, one of the organizations I 4 talked to was DALBAR. And they were talking about the 5 data they have, which is mostly mutual fund data, and 6 correlations to market movements where I think we've all 7 talked about days that we're studying, where we're saying 8 it's not fundamental movement. But I understand your 9 point about how do you separate, you know, the 10 fundamental versus transitory. 11 But you can look at this data and say, is there 12 a relationship between, you know, people making ultimate 13 decisions about not investing in the equity markets. But 14 this is an area that, as Andrew has said, there isn't a 15 lot out there. And just pulling this together, we don't 16 have the answers today. I can't tell you what the answer 17 is. But we haven't looked at it, and I think that's the 18 fundamental point that we're trying to make. 19 And I thought Andrew did a great job of 20 summarizing what we were thinking as well. 21 MR. ALPERT: Now I realize why I'm here. As a 22 journalist, I'm supposed to be complaining about the 23 market being rigged. And so I guess you found me because 24 I'm an apologist, right, for the status quo. 25 (Laughter.) 0130 1 MR. KATSUYAMA: So a couple thoughts. So I 2 think, you know, the subcommittee's recommendation, Chris 3 followed up on this idea of centralized reporting. And 4 in fairness to Rick, I don't think he's ever said 5 anything about it and it wasn't his idea, even though 6 FINRA was the one that was suggesting it. 7 MR. KETCHUM: We do look forward to having more 8 opportunities to do things we're not compensated for. 9 Particularly as I leave, as many of those as we can do, 10 the happier we'll be. 11 (Laughter.) 12 MR. KATSUYAMA: So on the ATS disclosures, you 13 know, just a quick anecdote, I mean, we found when we 14 launched our ATS, that we were selling against smoke. 15 People were making statistics up. We were trying to 16 decide where we ranked. We would go in and find out we 17 were ranked on varying scales from, you know, 25 to 5, 18 and could never get a straight answer. And the minute 19 that report came out, it kind of clarified a lot of 20 issues and allowed people to make decisions according to 21 what they saw from, you know, a centralized, unbiased, 22 trusted source. And I think that it improved things 23 quite dramatically. And I think that the same thing can 24 happen here. 25 I think one thing we've probably learned over 0131 1 the last couple years is that data can be massaged in 2 many different ways. It was interesting to hear Bill 3 call his reporting lousy, but that was his own words, not 4 mine. 5 No, in all fairness, I think that it's garbage 6 in, garbage out. I think the more that we can improve 7 the quality and the standardization of the inputs, the 8 more we can rely on what comes out of that. 9 I was interested, Bill, in the chart that you 10 put up, the bar charts, the comparative bar charts. Was 11 the point of that chart, just if I can remember it 12 correctly, that odd lots previously were not disclosed 13 and then when they were voluntarily disclosed, the price 14 improvement on those odd lots jumped, it seemed like, 5-, 15 6X? Is that -- was that -- 16 MR. ALPERT: Yeah, but initially the price 17 improvement on them was lousy. It was sort of like 18 somebody had turned the spotlight on before they were 19 ready. And so you could see that in that order category, 20 the price performance lagged the stuff that had been in 21 the spotlight for years. That was my main point. 22 MR. KATSUYAMA: So if you had seen that study 23 before you had written your report, would you have come 24 to a different type of conclusion or -- I mean, the 25 question is, is it better now than it was 20 years ago? 0132 1 I think the answer to that is quite obvious. Another 2 answer is, is it as good as it could be, or remotely as 3 good as it could be? And I think that begs an entirely 4 different question and that takes out factors like 5 technology and automation and all sorts of other things. 6 So I think that chart you put up, you know, 7 kind of underscores the whole point of what we're trying 8 to do with 605 and 606, is to improve the quality of the 9 metrics going in, improve the outputs coming out. And 10 disclosure does improve performance for those both who 11 are looking for that disclosure and those who aren't. 12 And I think, you know, if what I read on that chart was 13 correct, I think that statement in and of itself is the 14 underlying basis for what we're doing here. 15 And again, I was part of the subcommittee, so 16 I'm slightly biased. But I do think there are 17 improvements here that could be made. We talked a lot -- 18 I think we had a lot more suggestions and came to a set 19 of them that I think we could universally agree on. But 20 again, you never want to go through all this effort and 21 then leave a comment open like, oh, well, what data 22 source did they use or, oh, who's creating the data. The 23 more we can standardize that, I think centralized 24 reporting is absolutely a key part of ensuring that the 25 integrity of the data is sound and we can stop talking 0133 1 about that and talking more about the outputs. 2 And again, I just want to make sure that people 3 know that it was actually a person we invited to a 4 subcommittee discussion brought it up and it got a lot of 5 -- it got a lot of traction on the subcommittee. And I'm 6 not trying to volunteer Rick's organization for more work 7 they do not get compensated for. But I just want people 8 to be aware of how it came into our discussion and I 9 think it's a great idea. 10 MR. SHILLMAN: Chester and then Joe Mecane. 11 MR. SPATT: So both -- I want to both follow up 12 on sort of the points I made before but also have one 13 additional comment. So, you know, as I reread the 14 recommendation in light of the recommendation in light of 15 the discussion, it seems to me the recommendation is sort 16 of very broad about investor confidence. I read the 17 first bullet and see reference to individual investor 18 confidence in our equity markets. Then the bullet goes 19 on, to establish benchmarks in trends over time for 20 investor confidence in the U.S. equity markets. There's 21 no reference to market structure, there's no reference to 22 how trading occurs. And even if there was a reference to 23 how trading occurs, still the overwhelming thrust of this 24 would be to focus upon the confidence of America's 25 investors and America's capital markets. That doesn't 0134 1 mean in the market structure. 2 You know, I go down and look at the third 3 bullet. There's references -- well there is reference to 4 views on market structure. But there's also references 5 to views about volatility, investor behavior, asset 6 allocation, readiness for retirement. So, you know, this 7 suggests a pretty broad kind of focus on confidence sort 8 of broadly. And I think there needs to be kind of a lot 9 more thinking about how the survey is going to be 10 conducted, what are the goals of the survey, even 11 articulating the goals of the survey before the 12 recommendation is really ready. That's my view. 13 I also would emphasize, you know, given the 14 cost factors and the like, it might be helpful to refer 15 to more concretely, there is reference to existing 16 academic and industry studies on confidence. But the 17 group may want to sort of flesh that out further. You 18 know, looking at the specifics of Frank's and others' 19 studies, and to see to what extent do they actually 20 address issues of market structure. Perhaps they do, 21 perhaps they don't. I don't know. 22 And then I have one comment that's unrelated to 23 the specifics of this. But it's about the broad view, 24 about the broad issue of confidence. I think one issue 25 that it's important to stress is that ultimately, I 0135 1 think, the SEC's goal should be to promote fairness. At 2 least one of the main goals should be to promote fairness 3 in the markets for our country's investors. That seems 4 to me to be absolutely fundamental. 5 You know, in some of my own writing about the 6 financial crisis, as I look back, I wrote one piece about 7 conflict between regulators. Now, I kind of took a more 8 of a hypothetical view. But, you know, I tried to 9 reflect on some of the places where, in my mind, there 10 would have been differences in perspective between what I 11 called a fairness regulator and a regulator that had more 12 orientation of trying to fix things. So, in effect, the 13 SEC versus the banking regulators. 14 And, you know, I think clearly the banking 15 regulators felt that their role was to help make things 16 better. You know, I think the SEC's view was, from an 17 outsider's perspective, was a little bit more focused on 18 what would promote the fairness in the markets. And you 19 sort of see this in a number -- a number of issues, 20 issues involving conflict about things like Bank of 21 America, for example. Where, on the one hand, the 22 Commission imposed a sanction and, on the other hand, 23 arguably the Fed was trying to suppress the possibility 24 of disclosures. Or at least that's from an outsider's 25 perspective. 0136 1 So I do think it's important -- I do think it's 2 very important for the SEC -- I think part of what gives 3 the SEC so much power and clarity is if it's clear that 4 its mission is not to kind of artificially buoy 5 confidence, but ultimately to promote fairness. And so I 6 just would want to emphasize that. And that's not 7 necessarily -- that's not an argument, by the way -- to 8 be clear, that's not an argument against doing the 9 survey. But it does sort of really reinforce, I think, 10 the importance of what the objectives of the survey are 11 on being clear about that. 12 MR. MECANE: I just want to clarify one thing 13 on recommendation two. So if I understand it correctly, 14 what it basically would require is a retail broker-dealer 15 that, let's say, routes to three market centers would 16 publish their 605 results specifically with those market 17 centers? Is that what that second part of 605 is? 18 MS. KIMMEL: Yeah, so it's basically they would 19 be publishing the executions that they did or that were 20 done on their behalf by market makers, would be part of 21 their 605. So you would be able to have that direct 22 link, instead of trying to put together 605 and 606, 23 you'd have a 605 for retail. 24 MR. MECANE: And when you vetted that 25 recommendation with the retail firms that you spoke to, 0137 1 was there any negative reaction to that idea? 2 MS. KIMMEL: Yes, they don't love it. 3 MR. MECANE: And what's the -- what's the -- 4 MS. KIMMEL: So I think their argument is, it's 5 the aggregate statistics that are more important for the 6 retail investor, and that retail investors aren't going 7 to look at 605 reports. The counter argument to that is, 8 if everybody is doing the 605, then you could have all 9 sorts of aggregation based on that, based on a standard, 10 based on everybody doing it. 11 The other issue with aggregation is also the -- 12 what gets put into the aggregate statistics. And 13 depending on your firm's business model, you may want 14 different things put in to those aggregate statistics. 15 Which is part of the reason why the voluntary effort, you 16 know, doesn't have a universal appeal. 17 MR. MECANE: And are the panelists supportive 18 of the idea of brokers disclosing their 605 statistics 19 for the venues that they route to? I know, Bill, I think 20 that was one of your recommendations. Chris, I think you 21 were supportive of it. 22 MR. HATHEWAY: Right, I think the way I'd put 23 it is the 605 stats should follow the life of the order. 24 So if it originated with the retail broker, the retail 25 broker would be reporting on that order. 0138 1 MR. NAGY: Yeah, I completely agree with that. 2 You know, in this day and age, there's kind of no reason 3 why we shouldn't have that data. In fact, FINRA -- which 4 this is in my written testimony -- recently, in 1546, 5 talks about order-by-order reviews. And one of my fears 6 is that, you know, whatever gets cemented into stone with 7 these proposals today, will they kind of be thinking 8 about that future of the order-by-order reviews, which 9 are becoming increasingly more important? I can actually 10 envision a world where a retail consumer could go to a 11 website and see the quality of their specific order. 12 Today, and I think as Bill has realized, you 13 know, anybody that goes to look at a 606 report, just the 14 raw data, is in for quite the shock. You know, there's 15 no header data. It's pipe delimited. And you have to 16 understand what the nomenclature is within it, which is 17 very difficult. So I think that's one of the big reasons 18 why we don't see retail consumers go into that. 19 MR. SHILLMAN: Andrew and then Gary. 20 MR. LO: Just a very minor point in response to 21 Chester. Recommendation number one does read, "Benchmark 22 and monitor investor confidence in U.S. equity market 23 structure." That's actually in the title of the 24 recommendation. 25 But to your point, I think it is -- it is a 0139 1 good idea to be more focused when you do the surveys. 2 But the fact is that there are many different aspects of 3 the surveys that we think should be done. And so that's 4 really the intent of the recommendation. 5 The bottom line is that you can't manage what 6 you don't measure. And so this is just another way of 7 measuring some of the issues that the SEC might want to 8 consider. 9 MR. STONE: Andrew, I totally agree, you can't 10 manage what you can't measure. I'm just not sure that 11 the Commission is actually the place that should be doing 12 it. Sort of part of me sits here and thinks that the 13 people who are really incentivized to doing this would be 14 the ETF issuers, would be the retail brokers. Because 15 this strikes at the heart of their business model. And 16 so I think it's better placed with them or an academic 17 association, or them sponsoring an academic study for it, 18 or an institution like a newspaper doing it, because 19 that's what they do all day. 20 I want to come back to one of the bullet points 21 that we sort of glossed over, which was on recommendation 22 number two. Because I completely agree, and I just want 23 to make sure that we look at the incorporation of 605, 24 606 into the SEC's data visualization tool. 25 Originally when I read that, I did not agree 0140 1 with it. Until I realized that I needed to step into the 2 individual investor's shoes. And the reason why that's 3 so important is an individual investor needs to have the 4 SEC provide that data to them. He can't rely on vendors 5 who are going to charge for that service. 6 And so, from that respect, I hope that that 7 doesn't get lost. It's nice that FINRA is going to be 8 benevolent and put it up on their website. But I still 9 think that the Commission really does need to provide 10 visualization tools so people can understand what they're 11 looking at. 12 MR. ALPERT: And I think it would be salutary 13 to have a competition between vendors and the SEC and the 14 press to develop easier to use tools and better 15 presentations. 16 MR. SHILLMAN: Great. Are there any other 17 comments or questions? 18 CHAIR WHITE: Just to go back for a second to, 19 I guess, recommendation one. I mean, I do think it's 20 important that investors have well founded confidence in 21 the fairness of the equity market structure. And that, 22 to me, encompasses both kind of the fact and appearance 23 of fairness in those markets. And, I mean, one of the 24 things you worry about is even if, you know, 25 hypothetically you could build a perfect fairness model 0141 1 in the equity market structure. But if investors don't 2 think it's fair, and what you get from the survey 3 suggests that they don't, we need to know about that. 4 That puts aside -- and clearly, you know, precision in 5 the survey is very important, and putting aside questions 6 of who should do it. But I do think that's awfully 7 important to us. 8 MR. SHILLMAN: Maureen? 9 MS. O'HARA: Can I just add a footnote? I 10 think one of the interesting challenges, if you think 11 about this survey, is if you only survey investors, you 12 only survey people who have already decided that they're 13 fair enough to trade in. And so you now have the other 14 problem -- you're really going to have to survey people 15 who didn't do it. And there's an awful lot of them. And 16 how are you going to find them and what are you going to 17 do? I have to admit, I am a little skeptical on this. 18 But, you know, Andy has made, I think, a very strong 19 point that knowing something can be very helpful. 20 I would suggest, if the decision is to go 21 forward and do this, that the survey be done under very 22 stratified sorts of mechanisms, such as active investors, 23 ETF-only investors, you know, or people who have opted 24 not to invest. Because I know when you look at the 25 surveys in the retirement area, of which there are now 0142 1 many, you know, one of the things that comes out is, you 2 know, why don't people save for retirement? Well, you 3 know, some of them have bizarre views of the nature of 4 how retirement products work and that annuities are just 5 designed to steal your money. And so I think these 6 surveys have to ben done extremely carefully or you run 7 the risk that, you know, someone looks at a survey and 8 says, 40 percent of the people say the markets are 9 rigged, they really are. I used to think they were fine. 10 Now I don't. Now I leave. 11 So I just speak carefully to the fact that, you 12 know, the goals to me are not as clear as they have to be 13 to decide who you're going to be surveying, how you're 14 going to survey them, and what you're going to do with it 15 at the end of it. 16 MR. SHILLMAN: Bill. 17 MR. ALPERT: Would I be out of line to seek 18 ideas from the experts here on the way that these 19 proposals distinguish institutions from retail? People 20 have criticized the dollar difference there. And I think 21 other people suggest using the order distinctions of held 22 orders and not held orders as a way to distinguish retail 23 from institution. 24 Is it possible on the institutional orders to 25 get information on which ones have venue and time 0143 1 discretion, which ones just have venue discretion? What 2 ideas do people have on a more useful way to classify? 3 MR. STONE: I just want to ask the subcommittee 4 a question. Talked about dividing -- instead of dividing 5 by listed markets, divide by S&P and other NMS equities 6 and then doing OTC as a separate category. Was there any 7 consideration to doing it by ADV bucket? 8 MS. KIMMEL: No. I mean, it was really focused 9 on how retail thinks about the market and the S&P 500 10 rang true, you know, in addition to the retail brokers we 11 talked to. 12 MR. SHILLMAN: Brad. 13 MR. KATSUYAMA: Just one last point. So we had 14 in there the idea of if you use SIP to execute, you know, 15 evaluate yourself against the SIP. Or if you use direct 16 feeds, evaluate against direct feeds. And then Frank 17 brought up the point that it would be actually 18 complicated to do that. 19 I'm not exactly sure what -- like kind of the 20 methodology. I mean, when you're routing, as an example, 21 you're using your own feeds to make the determination on 22 where to route. Which means that it's actually easier, 23 at least from our standpoint, to use that as the 24 benchmark to kind of evaluate your data, rather than 25 using an external source that you're not using to fuel 0144 1 your decision around. So I thought, you know, that point 2 -- again, if we're trying to patch up a lot of the loose 3 ends, I thought that that point was sound. So I'm not 4 exactly sure where -- 5 MR. HATHEWAY: Let me give you a simple 6 example. ISO orders, you don't capture third market 7 quotes, other market quotes in your -- in the data record 8 for that order, because it's an ISO. So it's simply -- 9 it's not that it can't be done. It's just what is in the 10 incumbent infrastructure. That's what we found out going 11 through the tick pilot work, doesn't necessarily for all 12 trading centers, even the ones that use prop feeds, 13 enable you to say this is what the prop feed quote was at 14 the time order received, order execution, because people 15 capture what they need to capture. 16 MR. KATSUYAMA: Does the SIP solve the ISO 17 issue? 18 MR. HATHEWAY: No, it doesn't solve the ISO 19 issue. 20 MR. KATSUYAMA: So it's not solved either way. 21 MR. HATHEWAY: Right. So then it gets into the 22 point of the academic study from the guys at Berkeley. 23 What benefit do you get at what cost for having people 24 having to re-architect their -- 25 MR. KATSUYAMA: Well, the study inferred direct 0145 1 feeds without actually taking direct feeds in. So it was 2 a SIP direct feed study without actually taking direct 3 feeds. 4 MR. HATHEWAY: Right. If we had time, we could 5 talk to the study. But, yes. 6 MR. KATSUYAMA: So in terms of the complication 7 of using a separate data source relative to the ones used 8 to execute, I would beg to say it's less complicated to 9 use the same source. 10 MR. SHILLMAN: Joe. 11 MR. MECANE: I'll just give a quick answer to 12 Bill's question. I think that, you know, my impression 13 is that the 606 proposal tries to be respectful of the 14 existing 606 framework and build upon that. But I do 15 think, as you're alluding to, that it's not necessarily a 16 clean distinction between customer types. So my sense 17 would be the industry is still formulating views on 18 different ways to perhaps make that distinction, to go 19 back to, you know, the SEC with more formal comments 20 around. I think there's some inefficiencies in terms of 21 orders from a certain customer falling into both of those 22 distinctions. So I think the industry is trying to think 23 about how to make a thoughtful recommendation about 24 different ways to potentially distinguish. But I think 25 it's probably early to know, you know, what that -- what 0146 1 that recommendation looks like. 2 MS. KIMMEL: I would agree with Joe. And I 3 think from a retail broker-dealer perspective, there's a 4 lot of concern that, you know, you may have order flow 5 that's over that $200,000 threshold that isn't being, you 6 know, managed, isn't being order handled, really, and how 7 does that work? 8 MR. SHILLMAN: Chris. 9 MR. NAGY: Just in reply to that, and maybe a 10 little bit to Rick's comment earlier to me, but I was 11 surprised to see that your committee didn't have any 12 recommendation on the delineation between the 13 institutional and retail, such that is in the proposal. 14 I think it's something that is worthy of consideration. 15 And I thought, while not optimal, there is a secondary 16 solution that the SEC has proposed in terms of customer 17 classification, be it a registered investment adviser or, 18 if not, a FINRA broker-dealer. I think the problem with 19 that though is you don't accurately capture the 20 nonregistered advisers, and there is a considerable 21 amount of those that manage under $100 million that are 22 out there. 23 But I think it's an issue that should be 24 considered very importantly. Because you're talking 25 about providing one class of investors significantly more 0147 1 in terms of information about their order handling and 2 order execution, versus another class of investors. 3 MS. KIMMEL: Yeah, and I think as we mentioned 4 in our recommendation, those are the -- there are two 5 topics that we are yet to cover, given the recent release 6 of the proposal. And that is the definition of 7 institutional versus retail, as well as the institutional 8 disclosures. 9 MR. SHILLMAN: Okay. All right, well, this has 10 been another very good discussion. I think, as with the 11 first subcommittee, Manisha, leave it to your discretion 12 to, you know, refine your proposal and, at the 13 appropriate time, convert it into a recommendation -- a 14 draft recommendation from the full committee to the 15 Commission that could then be voted on at a subsequent 16 meeting. When you deliver that draft recommendation to 17 the full committee, then it will be made public and 18 people can comment on it. 19 At the same time, I understand you'll be 20 commenting on our recent institutional order routing 21 proposal, and would look forward to your developing that 22 and submitting it as a -- I think your plan would be to 23 submit it as a draft comment from the full advisory 24 committee. And so you would go through the same process. 25 Is that right? 0148 1 MS. KIMMEL: Yes. 2 MR. SHILLMAN: Yeah. So again, hopefully both 3 of those issues could be addressed and voted on at the 4 next meeting. 5 So I think with that, we are scheduled for a 6 15-minute break now. My sense is the updates from the 7 subcommittees may be relatively short, so is there an 8 appetite for just plowing through, or would people prefer 9 to take a break? 10 Keep going? Five -- okay. Why don't we take a 11 five-minute break and meet back at 10 after 2:00. 12 (Recess.) 13 MR. SHILLMAN: Welcome back. The last order of 14 business for today is an update from the Regulation NMS 15 Subcommittee and the Trading Venues Regulation 16 Subcommittee. 17 I'd like to ask each of the subcommittee chairs 18 to provide a brief report on the issues they will be 19 focusing on next. So let's start with the Regulation NMS 20 Subcommittee. Kevin, can I turn it over other you? 21 MR. CRONIN: Thanks, Dave. 22 So we thought it would be disingenuous for our 23 Regulation NMS Subcommittee to not be focused on 611. 24 And as we have now had some progress on 610, we will move 25 to 611. 0149 1 Specifically, there has been conversation. As 2 you probably remember, the full committee had a long 3 discussion on the order protection rule, as well as a 4 fair amount of discourse on locked and cross-markets, and 5 maybe to a lesser extent on sub-penny pricing. But we 6 would now look to really focus our attention on 611, with 7 no prescribed thoughts going into it, other than to say 8 there's a number of different ways that we might approach 9 it. 10 It could be enhancements to the existing rule, 11 certainly. We've talked about before adding, perhaps, a 12 block exemption. It could be that we would prescribe 13 some changes, or maybe scrapping the rule in particular. 14 Again, no preconceived notions walking in, but we do want 15 to look at this and have the same amount of vigor and 16 effort that we put forth in the tick pilot examination 17 and proposal we've put forth. So we will look to include 18 others within the industry as we formulate our ideas and 19 suggestions on what needs to be done. 20 I might also take the opportunity to suggest 21 that there are a couple of other issues that are out that 22 our group has talked about that really do have an overlap 23 with other groups. And, Dave, I've mentioned this to you 24 and to Steve before as well. You know, things like 25 market data, I think, probably touch some element of each 0150 1 of the subcommittees. And while I think there has been 2 some great work done on the regulatory side, the 3 governance side, I think there is some work to be done 4 yet on things like competition within SIP and the value 5 of the SIP and even maybe perhaps pricing. 6 We also think that there is probably some value 7 in having further conversations around inducements more 8 generally. We talked a little bit about that with 9 respect to maker-taker rebates and that sort of thing. 10 But there are a number of inducements in the industry 11 that exist that probably warrant further consideration, 12 for example, like payment for order flow. So that might 13 require us to perhaps rethink some of the structure of 14 the subcommittees. But as it pertains to the Regulation 15 NMS Subcommittee, we are going to now focus on 611, and 16 we will look forward to bringing the full committee some 17 of our ideas in time. 18 MR. SHILLMAN: Great. Any questions or 19 suggestions for Kevin? 20 Rick. 21 MR. KETCHUM: Thank you, Dave. 22 The Trading Venues Regulation Subcommittee also 23 has been meeting recently to discuss the status of the 24 two recommendations is presented on a preliminary basis 25 at the April 26 MSAC meeting, and as well as discussions 0151 1 with regard to other related topics that we mentioned 2 that are within our mandate. And I would just take a 3 moment to thank the commenters who have provided the 4 subcommittee with helpful perspective to take an account 5 on those areas. 6 I think the subcommittee is making good 7 progress in advancing each of its two pending 8 recommendations. The first of these recommendations 9 concerns a review of SROs' rule-based liability limits. 10 The second concerns centralization of common regulatory 11 functions, a topic that's had some discussion in the last 12 week. 13 And I'll just remind the committee and the 14 Commission representatives that I'm recused with respect 15 to the recommendation on regulatory centralization, so I 16 will, when I finish wrapping up, will pass it on to 17 Professor O'Hara. 18 On the first pending recommendation, there is a 19 general consensus among the subcommittee members to move 20 forward in the area of exchange rule-based liability. 21 The thrust of a recommendation, I would expect, would be 22 a recommendation for the Commission to perform a 23 comprehensive review of current exchange rules to 24 evaluate whether the rules should be revised to make the 25 application and amounts of liability limits consistent 0152 1 and to evaluate the sufficiency of liability limits 2 across exchanges, particularly listing exchanges. 3 Consistent with the preliminary recommendation, 4 the subcommittee also has continued to discuss but is not 5 prepared to make any formal recommendations at this time 6 with respect to regulatory immunity. We have not reached 7 a consensus conclusion, but we will continue to have 8 discussions in those areas. 9 The additional topics that the committee has 10 been focused on concern potential conflict of interests 11 at alternate trading systems and other executing 12 broker-dealers. The subcommittee reiterated its 13 directional support for the SEC's proposed revisions to 14 Regulation ATS. And the subcommittee also noted the 15 SEC's recent proposal that is meant to enhance 16 transparency for brokers' order handling practices. 17 Overall, given the SEC's significant ongoing efforts, the 18 subcommittee at least at this time does not presently 19 expect to make recommendations in those areas. 20 The subcommittee continues to engage with 21 questions about the functioning and utility of 22 consolidated market data feeds, similar to Kevin's 23 subcommittee, as they currently operate under the 24 framework of the SIP plans. We are focused on a number 25 of questions, including a review of recent developments 0153 1 from the standpoints of reports about increasing SIP 2 speeds and operational enhancements, and evaluating how 3 that may impact any recommendations for change or 4 increased transparency, as well as looking at whether 5 there are ways for the industry and public to more 6 clearly understand the operation of the SIPs, for 7 example, with agreed-upon statistics about SIP latency or 8 performance. 9 And finally, we look with interest with respect 10 to the discussions of an experiment in its early stages 11 in Europe to pursue a competing SIP model. And we are 12 continuing to review the efficacy and issues around a 13 competing SIP model versus the present model that exists 14 in the United States. From that standpoint, we are 15 planning to hold a panel discussion with select market 16 participants, including buy-side representatives, to get 17 a better sense of the need for potential recommendations 18 in this area. 19 MR. SHILLMAN: Thanks, Rick. 20 Any questions or suggestions for Rick? 21 Any other issues or topics anyone would like to 22 discuss at today's meeting? 23 Maureen? 24 MS. O'HARA: I should pick up the second part 25 of our recommendation. So we actually do have two things 0154 1 that we have been looking on, and the second one is the 2 centralized -- centralization of common regulatory 3 functions. The subcommittee continues to believe this is 4 extremely important. 5 We mentioned this before when we were at our 6 last meeting. We raised the fact that we think, 7 particularly in the era of centralization of these -- 8 these activities with CAT and everything else coming on 9 is extremely important. 10 We have received comments, most notably from 11 NASDAQ and the NYSE, who actually both agree that some 12 centralization and consistency is desirable. And 13 particularly as it will serve to reduce any sort of rule 14 arbitrage. So that's the good news. They also raise 15 concerns, however, that you have to be careful about what 16 it is that you centralize and not. 17 We did stress before that the subcommittee does 18 believe that there are some regulatory functions that are 19 best left at the SRO level. So we did not say up front 20 that everything should be centralized and we continue to 21 believe that that dichotomy should exist. But the 22 question then becomes, what are these activities that are 23 best centralized? And this is what we are going to be 24 spending time on over the next coming months. 25 We think that it's important to be more 0155 1 specific about the types of activities that would benefit 2 from common centralization. We're going to be soliciting 3 more input on what those particular activities would do. 4 We are contemplating potentially a task force approach to 5 addressing these sorts of issues. But this is something 6 we will spend time on and we're not really ready today to 7 exactly say what our recommendations will be. 8 MR. SHILLMAN: Thanks, Maureen. 9 Any other topics anyone would like to discuss 10 today? 11 Okay, well, I think this was another very 12 constructive meeting. I think we made a lot of progress 13 on the issues raised by the two subcommittees. 14 So to sum up, the Market Quality and Customer 15 Issues Subcommittees will continue to refine the 16 recommendations with the goal of having a final committee 17 vote at the next meeting. The Regulation NMS and Trading 18 Venues Subcommittees will begin work on their new areas 19 of focus. And, for our part, Commission Staff has been 20 assessing the first set of committee recommendations on 21 an access fee pilot and NMS planned governance, and will 22 be working diligently, as the Chair has instructed, to 23 develop recommendations for consideration by the 24 Commission. 25 So with that, I will entertain a motion to 0156 1 adjourn. 2 MR. KETCHUM: So moved. 3 MR. SHILLMAN: And a second? 4 MR. MECANE: Second. 5 MR. SHILLMAN: All in favor? 6 Thanks again for all your hard work, and we're 7 adjourned. 8 (Whereupon, at 2:25 p.m., the meeting was 9 concluded.) 10 * * * * * 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0157 1 PROOFREADER'S CERTIFICATE 2 3 In The Matter of: EQUITY MARKET STRUCTURE 4 ADVISORY COMMITTEE 5 File Number: OS-0802 6 Date: August 2, 2016 7 Location: Washington, D.C. 8 9 This is to certify that I, Nicholas J. Wagner, 10 (the undersigned), do hereby swear and affirm that the 11 attached proceedings before the U.S. Securities and 12 Exchange Commission were held according to the record and 13 that this is the original, complete, true and accurate 14 transcript that has been compared to the reporting or 15 recording accomplished at the hearing. 16 17 _______________________ _______________________ 18 (Proofreader's Name) (Date) 19 20 21 22 23 24 25