 |
Home | Previous Page | |
|
||
 |
|
|
|
||
 |
|
|
|
|
|||||||||||||||
|
|
|||||||||||||||
|
|
Panel 3 Transcript: Market Structure RoundtableJune 2, 2010Director Robert Cook:
The concept release described undisplayed liquidity as trading interests that is available for execution at a training center, but is not included in the consolidated quotation data that is widely disseminated to the public. Undisplayed liquidity is also commonly known as dark liquidity. The release also noted that key forms of undisplayed liquidity are dark pool, ATSs that do not display quotation in the consolidated quotation data, broker dealers that internalize orders, and undisplayed order types of exchanges and ECNs. And the release then asked for comment on three broad areas. First, the effect of undisplayed liquidity on order execution quality; second, the effect of undisplayed liquidity on public price discover as trading volume of undisplayed liquidity reached a sufficient significant level that has detracted from the quality of public price discovery and execution quality; and third, fair access to sources of undisplayed liquidity. So, we look forward to hearing from our panelists on these important issues and related matters. So sticking to our format from earlier today, we'll proceed first with opening statements by each panelist followed by a question and answer session. So, we'll try to keep our comments to about three minutes and we'll start with Brian. Brian Conroy:
As the Commission examines the current structure of the U.S. equity markets, we urge the Commission to consider two important preliminary points. First, the U.S. equity market is among the most competitive, liquid, and efficient markets in the world. Daily trading value is the largest and spread costs are the lowest of any global equity market. Second, we recognize the Commission to be careful — excuse me — we request the Commission to be careful and measured in their approach to ensure that any proposals it makes do not hinder the performance of the existing equity market and are based on sound empirical evidence. The fundamental responsibility of our trading desk is to implement the investment ideas of our portfolio managers by trading securities as efficiently as possible. To fulfill this responsibility, Fidelity uses a wide variety of trading venues and strategies to execute orders. It is important to note, however, that we do not favor one type of trading venue or model over others. Instead we support a market structure that encourages the innovation and efficiency that results when multiple trading venues compete for customer business in a variety of ways. And to best serve our shareholders, we think it is essential that the variety of services offered to investors includes access to both displayed and undisplayed liquidity. Sourcing liquidity is one of the institutional investor’s biggest challenges, and we find that liquidity exists in many forms. There is quoted liquidity that is represented in the NBBO, capital that is provided by a broker dealer or market maker, uncommitted liquidity that exists as positions with a portfolio manager, natural liquidity that is provided by another institutional order, and dark or undisplayed liquidity. A wide variety of liquidity sources is particularly important to Fidelity given that we regularly place block orders representing a significant portion of a stock's average daily volume, and our orders virtually always outsize the liquidity that is available in a quoted or visible market. The ability to trade anonymously without exposing our order to the quoted market is highly valuable when we are working these large block trades. If required to place these outsize orders in public limit order books, our shareholders would be vulnerable to short-term traders who could step ahead of our shareholders orders, potentially resulting in higher trading costs and a lower investment performance. There has long existed a symbiotic relationship between the displayed and undisplayed liquidity, and in our view a healthy market is one that strikes an appropriate balance between lit and dark venues. Exchanges and ATSs represent two different business models, both of which are necessary and important to institutional investors like Fidelity. We believe any examination of market structure should be done with a view towards fostering the appropriate balance between the traditional and alternative markets for the benefit of all market participants. If trading went largely dark, market volatility would increase, and price discovery would decrease to the detriment of long term investors. A trading market without meaningful dark venues, however, would drive up costs for long term investors and expose the trading intentions of mutual funds and other large investors who depend on the anonymity that the dark venues provide. Having a good balance of competition among trading centers helps ensure that the markets remain free, highly efficient, and liquid. I would like to thank the Commission for holding this hearing and look forward to discussing some of the specific issues concerning undisplayed liquidity that the Commission raised in its market structure concept release. Thank you. Director Robert Cook:
Larry Leibowitz:
It is tempting to use the word “innovation” as a magical justification for almost anything designed to give a competitive advantage, including practices that appear to be in the best interest of fair and orderly markets. Indeed there are many innovations which have arisen in the past 10 years in the financial markets more generally that regulators would undoubtedly like to have taken back or at least overseen differently. There's a natural tension between innovations which impact the subset of the market or market participants, and the goal of having a market which gives investors confidence and protects the concerns of issuers. Nowhere are some of these intentions clearer than on the topic of dark liquidity. Undisplayed liquidity has been a valuable and evolving tool for market participants for decades, whether taking the form of a broker keeping part of an order hidden from the marketplace, a block desk, or an exchange floor; a hidden order type on an exchange, executed in the dark through and ATS, or through a purely internalized execution. And while each of these structures provides benefits to the direct parties to the transaction, each allow different types of access and has different side-effects. In addition to looking at the impact of particular dark practices, the Commission should also look at the aggregate effect of dark activity on the overall marketplace, and determine how to best foster interaction among these in a way that creates a healthy environment for all participants. Indeed recent academic studies have highlighted the negative impact on spreads and volatility as a result of off-exchange trading level increases. Given the 600 basis point increase in off-exchange activity in the past year excluding direct edge, this is cause for concern. Additionally we must consider the toxicity level on exchanges as we continue to filter increasing levels of order flow before accessing public markets, disadvantaging displayed limit orders are the very orders we want to encourage. There are no easy answers to these questions. Turning back the clock and stifling a competitive and innovative marketplace is not the solution. However, there are some moderate steps that should be looked at. First is more consistent rule making and oversight for similar functions. We should make sure that volume isn't migrating to the dark for unfair structural reasons or regulatory arbitrage. Existing practices such as sub-penny price improvement should be examined to see whether they violate the spirit if not the specifics of existing regulation. Second is obligations to the market by liquidity providers and incenting of displayed liquidity to ensure deeper markets and lower trading costs, and help dampen volatility. Finally, there should be additional disclosure and scrutiny of order handling and execution practices both for institutional and retail orders by brokers, exchanges, as well as ATSs. Clearly, existing practices were a contributing factor to the May 6 crash. We must also not forget that the equity markets primarily exist to facilitate capital raising activity, yet we frequently neglect the issue in our market structure debates. The issuer community generally views our recent market structure evolution negatively, and May 6 reinforces this view. We need to refocus our markets on what's best for issuers and investors rather than slowly emphasizing speed and advantages aimed at selected participants. Thank you. Director Robert Cook:
Dan Mathisson:
The empirical results are impressive. Bid-Ask spreads in the U.S. are currently the tightest in the world with the S&P 500 averaging four basis point spread in the first quarter of 2010 which is the best out of 30 developed markets. Execution fees dropped quickly after Reg ATS and have continued down over the years. Institutional commissions followed them down and are down 65 percent since 1998. Retail commissions followed exchange fees down as well, and today retail commissions are the lowest they've ever been. And as a result of these lower commissions and a cleaner, faster, more fair system; volume has grown to 9.5 billion shares per day so far this year, which is up from 1.2 billion shares in the year before Reg ATS went into effect. Now, in this world of competition and free choice among investors, many investors have found that using dark order types has been beneficial. Dark liquidity is a small niche. It's currently about 10 percent of the market, but we feel it's an important niche for big institutional investors that sometimes chose to trade without displaying their bids and offers, without broadcasting their intentions via public bid or a public offer to every trader on the planet. Now, critics of dark liquidity suggested that dark order types have damaged price discovery, but there's no empirical evidence of this, and in fact, the empirical evidence indicates the opposite. Although dark pool volumes have steadily grown over the past six years, we found in a study that displayed size has actually increased 72 percent since 2004 while Bid-Ask spreads have grown tighter. So, we have a market where the bid and the ask gets tighter and size up keeps getting bigger. The empirical evidence supports what we have long believed at Credit Suisse, that markets work best when there is a vibrant competition among numerous types of trading centers with all investors given an equal opportunity to access all of the markets. With a diverse mosaic of exchanges, ECNs, ATSs, and OTC market makers all competing with each other for order flow, the investor inevitably wins as competition drives lower fees, innovative products, and improvements in reliability and customer service. Thank you for the opportunity to appear today, and I will be happy to answer any questions that you may have. Director Robert Cook:
Seth Merrin:
As we are focused on the integrity of the markets, it's important to understand the value that Liquidnet provides to this investment community. The average institutional order size is about 180,000 shares, which is about 600 times larger than the average execution size on the exchanges. The problem is that if the institution introduces even a 50,000 share order into the market that's designed for a 300 share execution, short-term traders take advantage of the supply demand and balance and buy ahead of the institution knowing that they could just sell back to the institution at a higher price. So, as you know it's the tens of millions of households that invest in mutual funds and pension funds that pay that higher price. Our system helps address this problem. The average size of our negotiated execution is more than 150 times larger than that of the average execution size on the exchanges. And third party validation from the largest transaction cost analysis provider in the industry has ranked Liquidnet number one in execution quality across all global brokers, across all categories for the past two years running. So, our number one ranking in execution quality means reduced trading costs and higher investment returns for the tens of millions of beneficiaries of the mutual funds and other accounts managed by our customers. I'm also pleased to report, since I know it's going to be a topic here, that during the events of May 6 our system operated without any problems as institutions generally stop trading when their orders hit their limit prices. We had zero customer issues on May 6. And since I am here, I would like to focus a little bit on our H2O system and address the use of institutional indications of interest. Our H2O system has been important in achieving the number one ranking in execution quality. With H2O institutional block orders can interact with other orders on the market including retail orders, and our H2O system interacts with the New York Stock Exchange, and the other exchanges, and ECNs and brokers as well. It's important to understand that H2O provides 100 percent price improvement for all the orders including retail orders, and we're concerned that the SEC proposal on IOIs as currently written would prevent retail from receiving the same 100 percent price improvement that our H2O system provides to our institutional customers. We proposed a minor modification to the SEC proposed block exemption so that institutional block IOIs could be sent to non-block recipients as long as the institution knows and consents. The institution is committed to trading that block and the execution venue provides significant price improvement to both parties which can be defined as a minimum of one cent price improvement where the spread is two cents or higher, and the midpoint execution on the spread is one cent. In this comment letter, Credit Suisse presents its data that internalizing dealers provide approximately five percent price improvement of customer orders and exchanges provide negative 10 percent price improvement of customer orders. Liquidnet H2O system which uses institution IOIs currently provides 100 percent price improvement for all customer orders including retail. The SEC can achieve the objectives of the proposed rule on IOIs while at the same time preserving the benefits of the system that was demonstrate actual cost savings for both institutional and retail investors, including institutional execution quality that has been ranked number one in the industry, and the level and price improvement for retail orders that could help set a new standard for the industry. According to the tab group, bi-set institutions execute over 70 percent of their flow electronically. So, we should not limit the bi-sets’ ability to use tools that lower the cost of trading and increase the returns of the hundreds of millions of Americans that they represent. Thank you again for inviting me to participate on behalf of Liquidnet, and I look forward to answering any questions that you have. Director Robert Cook:
Eric Noll:
Non displayed liquidity is valuable to some investors at some junctures. NASDAQ recognizes that the availability of dark trading venues is important for larger institutional orders, and it believes that the Commission’s proposals appropriately balance those needs with the dual goals of providing market efficiency and protecting investors. Internalization should be permitted, provided that the internalizing firm simultaneously displays the protected quote at the NBBO, or provides meaningful price improvement over the NBBO. NASDAQ supports the Commission’s proposal to include actionable indications of interest within the definition of bids and offers, to reduce to threshold for display of dark liquidity, and to establish post trade transparency for dark pool executions. Additionally, investors need modernized execution and routing performance metrics to account for changed market conditions. We should aim to recreate the quality competition that rule 605 and 606 initially created. Changes would include finer speed measurements, consistent definitions of market centers and covered orders, and coverage of the life of each order. All markets that trade the same security should be equally transparent about their operations including rules governing their trading systems, the criteria for admission into membership, and the pricing of comparable services. The Commission should reexamine the policy basis for existing difference in transparency between exchange markets and other markets, and determine whether such differences are justified, and whether they contribute to regulatory arbitrage. Finally, investing in NASDAQ's opinion is not a zero sum game. All market participants may have different investment horizons and different investment goals. Market structure serves investors best by ensuring the broadest possible mix of participants rather than attempting to rank or otherwise elevate some investors over others. The open interaction of diverse trading interests and strategies and trading platforms whether long-term, short-term, retail, institutional, proprietary promotes continued innovation and efficiency in our market place. Vilifying or excluding entire sets of market participants such as high frequency traders or algo traders undermines historic Commission policies favoring competition. I think it's incumbent on us as exchange operators and as regulators to create a market place where all of those trading interests can interact in a fair and open transparent matter. Thank you. Director Robert Cook:
William O’Brien:
As mentioned by Robert Cook at the intro of this panel, undisplayed liquidity takes many forms including various exchange order types, alternative trading system products, and various institutional and wholesale market making services. There are many legitimate economic and execution quality-driven reasons why investors and their intermediaries seek to use these alternatives. Investors and their chosen brokers are best equipped to decide how to balance competing execution priorities, and how to bundle execution services with other services that help them or their clients meet their overall financial objectives. The use of non-displayed order types and the use of non-exchange execution facilities merely reflect this. The benefits of greater execution flexibility apply not only to the institution and trader seeking to affect a large trade but also to retail investors seeking guaranteed executions of the full — size of their order at the NBBO or better, with meaningful opportunities for price and size improvement. Orders executed through undisplayed order types or on off-exchange liquidity facilities should not be viewed as an net reduction of overall market liquidity, but rather liquidity seeking its most natural and efficient form of execution. The existence of undisplayed liquidity does not raise fairness issues in an environment where investors have a broad-based ability to participate in and reap the benefits of these order types and execution facilities either directly or through their intermediaries. Restrictions placed on such venues will not eliminate undisplayed orders or off-exchange executions, nor alter the demand for them. Rather, it will just make such executions either more expensive or less accessible to certain categories of investors. Alternatively it could cause investors to turn to less regulated instruments or markets with greater flexibility, whether through derivatives transactions or seeking overseas markets without such requirements or restrictions. Investor choice has pushed all execution venues to complete for order flow with improved execution quality, product innovation, and lower fees. Exchanges continue to play a critical role in providing pre-trade transparency and price discovery, which helps to ensure a baseline level of execution quality for all investors. And exchanges are pretty fairly compensated for this, I believe. If the level of overall market share among exchanges were to fall precipitously below historical norms, it would be appropriate to examine what further steps would be needed to preserve or transfer the role that displayed exchange liquidity plays in our market. But with on-exchange liquidity consistently above 70 percent, we are just simply not near that point today. While the use of undisplayed liquidity will ebb and flow over time for reasons far beyond simple market structure mechanics, it has not approached a level that could undermine overall market quality at this time. To ensure this going forward and to allow investors to hold their intermediaries accountable for the execution decisions they make, improvements to existing rule 605 and 606 could be made to provide more detailed insight to investors. Once again, I'd like to thank the Commission for the opportunity to testify today, and look forward to a great Q and A session. Male Speaker:
Mark Ready:
I'd like to start by saying that my remarks today are solely my opinion and don't reflect the opinion of the faculty and staff at the University of Wisconsin. [laughter] A little inside joke. So, I just wanted to make a couple of points to try to help frame the discussion, if I can, and then make one sort of detailed point following up on the remark that Seth about price improvement, for example, at Liquidnet. So, the first point that I want to make is that there is sort of an irony here related to the discussion of buy-side firms and the way they think about avoiding front runners, avoiding people who anticipate their orders. So, when you think about it, the portfolio managers at these buy-side firms look at a stock that's currently trading at 20, and they think it's worth 30 or 40, so they decide to buy it. They don't buy it because they think it's worth 21. They buy it because it's worth a lot more, in their opinion, at least, than 20. So, they give the order to their trading desk, and the trading desk's objective is to try to get it for as close to 20 as possible. To the extent that the trading actions of the trading desk get ferreted out by the market, you'll see the price drift up maybe to 20 1/2, 21. And that's their lament that their volume is moving price towards, nowhere near, but towards the direction where they think the true price is. In some sense, what's happening is the market is causing price to be efficient by sorting out that there's somebody who believes that this stock is a really good deal. So, when you think about it, that sort of makes you think well, so, why is it that we're trying to have mechanisms that protect this information when the information coming out makes the prices more efficient? So, that's one way to think about. And an even higher level is to say well, if you don't protect that information then there's less incentive to go search it out in the first place and maybe less of it will be produced. But, it did strike me, as I was listening to a lot of the discussion of the evils of, say, short-term traders, who anticipate the orders of the buy-side firms, that the evil that we're talking about here is that the price is going to be driven in a direction that the firm, by its very actions, believes to be the correct price. So, the market is becoming more efficient. But, as I said with my other caveat, it's just something to keep in mind; it doesn't give you a regulatory prescription off the bat. The second thing that I wanted to point out is we're talking about different types of undisplayed liquidity here, and I think it's good to keep a clear distinction between two different kinds of undisplayed liquidity. One is the undisplayed liquidity that is related to this action by buy-side firms trying to keep their order from being revealed to the rest of the market, so that they can trade at a reasonable price. And so, that's the initial sort of business model, as I understand, for Liquidnet was to try and facilitate buy-side firms meeting directly with each other, so that there wouldn't be leakage of their trading interests and they would get reasonable prices. So, a second type, however, that's included in the discussion is internalization. And internalization is sometimes individual broker dealers trading as principal to facilitate a buy-side investor, but in many cases it's broker-dealers like Knight, for example, internalizing retail flow that they get from, say, TD Ameritrade. Now, that second type of internalization is quite different. To me, I think of it as what we used to call cream skimming. So, the idea is, you have order flow that comes from a group of retail investors who are by their nature not particularly informed. A lot of times they're trading just because they have portfolio reasons. They've got some money to invest, or they need to buy a car, or whatever, and so, their order flow has no real information in it. And so, it should be traded at a very, very narrow spread, even narrower than what you see across the market. So, one of the ways that internalization can be good for at least that subset is it gives them some sort of special deal, so that they can get even lower trading costs as a result of the fact that their order flow is not particularly informed. So, I was here during the original promulgation of what we used to call Dash 5, now called 605, and one of the reasons that I at least argued for realized spreads in that disclosure was so that people like Ameritrade could have a negotiation with Knight and say “Look, you need to give us better price improvement than you're giving other people because our flow is less informed. And we can see that you're making more on our order flow than you make on other people.” So, there is this other type of internalization that might be useful to separate, certainly, to think about it separately, but maybe even to sort of separate from the rest of the discussion. Finally, a minor point on price-improvement statistics. And sometimes by definition, Seth's model, when retail order flow trades in Liquidnet, it gets price improved. It has to. That's the way it works. The one thing that you have to keep in mind when you compare his execution statistics, which are 100 percent price improvement, with the execution statistics of another venue is that trading only happens for those retail orders in Liquidnet, if there happens to be a natural contra side; whereas, if you send your orders to Knight, Knight trades it no matter what, contra side or not. So, it's not an apples to apples comparison to compare his price improvements to, say, Knight's price improvements. So, thank you. Male Speaker:
Andrew Silverman:
Morgan Stanley's approach to trading electronic markets emphasizes the confidentiality of customer-order information while achieving best x. Our business model and trading engines are designed to limit the amount of order information communicated to third parties. Over the past several years, we have been very outspoken with respect to aggressive order-handling practices and routing practices that have evolved in the U.S. equity markets. We have advocated for increased transparency and meaningful disclosure, and have pointed out a number of practices that we have found troubling. However, we continue to believe that the fundamental condition of the U.S. national market system is sound and the equity markets, as a whole, do an excellent job. The issue of undisplayed liquidity is absolutely not new. Exchanges in trading crowds have always existed, and good traders have always sought to shield their true interests from the entire marketplace to get, really, the best price for their customers. Enhancements in technology have allowed undisplayed liquidity to continue to play a critical role in today's market structure. Undisplayed liquidity takes many forms, and serves the entire range of market participants, including retail, and institutional, and long and short-term investors. All forms of undisplayed liquidity should not be treated the same. We would like to highlight our view of three key issues relating to undisplayed liquidity. First, fair access: Morgan Stanley's dark liquidity is truly dark. Our liquidity pools are carefully constructed and monitored to provide a level playing field, while maximizing the potential benefit for our customers. We do not believe that dark pools should be forced to provide unfettered access, direct access to all forms of aggressive and potentially predatory order flow. Second, trade at and depth of book: trade at would essentially require a broker to take out all top-of-book protected quotes and more price levels if depth of book was included before it could execute an order at the same price for its own customer. Such an approach would effectively put most continuous-matched dark pools that execute at the NBBO out of business. It would also end a broker's ability to commit capital, prevent auto-execution of customer orders, effectively prohibit upstairs crossing by broker-dealers holding natural contra-side interest. Trade at is a draconian solution, I believe, in search of a problem, and should not be implemented. A trade-at regime would not automatically bring this flow back into the [unintelligible] markets or the exchanges. We think a far more likely result is that orders will move back upstairs to the buy-side trading desks. Third and last, low-priced stocks: under the SEC's current sub-penny rule, one would expect executions to occur at the bid, at the offer or midpoint. I just want to repeat that: most executions should occur at the bid, offer, or midpoint. We question why so many prints are going off in very small sub-penny price increments. The use of hidden or dark midpoint-pegged orders and other formulaic order types by some dark pools already provide the abilities to substantially reduce the overall spread for the stocks. The behavior opens the door to gaming and practices, and must be carefully examined before additional action is taken. So, to close, thank you very much for the opportunity. I commend the Commission for its willingness to tackle these very complex and important issues and look forward to answering your questions today and going forward. Thank you. Director Robert Cook:
Chairman Mary Schapiro:
William O'Brien:
So, I guess the question is as you look at some of the other anecdotal data points that the bid-ask spread is as narrow as it's ever been, the depth at the inside, and, if you want to think about it over the last 10 years, depth within the former minimum increment, right? Whether it's an eighth or a tini, it's greater than it's ever been. It's hard to say that we've compromised it. Now, there are some obviously, I think, logical data points that would at least give me some cause for concern, displayed-market paneling less than 50 percent of volume, for an example. But I think you have to look at both quantitative and qualitative measures to make a determination. And I think we need to monitor that, I think, vigilantly over time, but where we stand today, we're nowhere near the point. It's not even close. Larry Leibowitz:
William O'Brien:
Director Robert Cook:
Mark Ready:
Brian Conroy:
Director Robert Cook:
Mark Ready:
Male Speaker:
Dan Mathisson:
Now, if you change the rules so that OTC market makers weren't allowed to provide liquidity, then instead of the retail broker getting paid a rebate to send it to an over-the-counter market maker, they would have to go take it in the displayed markets, which would typically cost thirty mills, three-tenths of a cent. They would not get the price improvement of a tenth of a cent. You have like a four tenth of a cent swing, and since the retail broker isn't receiving the rebate, they would have to charge more than the eight or $10 a trade. So, you'd be looking at a higher commission. You'd be looking at a worse price for the retail investor. And so, it's an ideal setup right now. William O'Brien:
So, they're making individual decisions that work for them. And many of them have either fiduciary duties to or duties as agent to their underlying clients or just their own desire for self-good. And it's a question of how much do we restrict individual liberty with respect to procuring the best form of execution quality over the collective good, right, in the name of price discovery or some other buzz word. And I think if you look at how the market has operated as a whole, under normal conditions, and I think May 6 is an entirely different animal altogether and really shouldn't inform our discussion of these types of points, you see that it's generally working very well in providing a baseline level of execution quality for everyone, no matter how they choose to execute within a framework that gives everyone the liberty to decide what works best for them. You shouldn't force that to change. Seth Merrin:
Andrew Silverman:
Mary L. Schapiro:
William O'Brien:
Mary L. Schapiro:
William O'Brien:
Mary L. Schapiro:
William O'Brien:
Mary L. Schapiro:
William O'Brien:
Larry Leibowitz:
Seth Merrin:
So, the most frequently used high-frequency trading strategy today, and May 6 was really an exaggerated event of what happens on just about every institutional order. And that is, they create this supply and demand imbalance in the marketplace and high-frequency traders are looking for that. An algorithm is simply a way to take a large institutional order and slice it into little tiny pieces. What high-frequency trading computers do is they look for those algorithms that are slicing into little tiny pieces and they do the opposite. And they jump on the bandwagon. When you have something where allegedly E-mini futures start as a catalyst, and they start putting pressure on the entire S&P 500, which is the most liquid stocks, which is where high-frequency traders play, that created the catalyst for everybody piling on board and forcing the market down. The other market structure change was that when the New York Stock Exchange owned 90-plus percent of the market share, when the declared a halt or slowdown, it stopped. When they trade one out of ex-number of shares that they do today, the big number — 83 percent of the market share — when they declared a halt or slowdown, it didn't really matter. But the change that you guys have proposed, I think, is a great solution. Dan Mathisson:
Mark Ready:
Dan Mathisson:
Mark Ready: Again, I’m not against it, I’m just warning you that when the stock needs to move, halting it may not be a good thing. Brian Conroy:
David:
Dan Mathisson:
There are some significant ATSs that don’t allow broker-dealers in. We think that it would be a smoother market if everybody was able to interact with everyone else and there were non-discriminatory admission policies in place and let all the liquidity match up with each other. Seth Merrin:
Institutional trading has gone 70 percent electronic over the last few years, 10 years. And, you know, institutions are simply not going to display their liquidity on any venue. They never have and they never will. So, there are a couple points. One, when we asked what fair access meant, the response from the staff was as long as you have a standard set of rules that’s published that you abide by, then that’s fair access, which we do. But I’m very concerned about the level — anything that takes away tools that the institutions require to mask their intent into the marketplace, so any elimination of that is only going to increase overall transaction costs, which ultimately is going to disadvantage the 100 million Americans that invest with them. So, you know, it’s the large trade exemption that we’re really looking at. Andrew Silverman:
Dan Mathisson:
Andrew Silverman:
Chairman Mary Schapiro:
Seth Merrin:
So, we took that wholesale marketplace and we opened it up for everybody else to participate, and we were able to build a better venue, but one size will never fit all. And to try to get everybody to play in the same sandbox at the same time, there’s always going to be a loser. And up until this point the loser has always been the institutions, and the institutions are the ones that manage money on behalf of all the little guys. So, I think that we should say that it’s complementary. You know, this is the only industry that uses the word “fragmentation” to really mean competition. And there’s no instance that — I think we’re all agreed on this panel — where competition has not been good for the ultimate investor. So we should look at all these types, you know, providing everyone plays fairly by the same rules, as being very complementary in its design and its capabilities to the investors. William O’Brien:
And then what slowly happened over time is that as broker-dealers began to take use of similar technologies and automate pre-existing business models, they were swept up within the scope of that rule. And it was — from a regulatory perspective, it was accommodated for. So, whether it’s the automation of an Institutional Block Desk, which is effectively what Liquidnet is, or the automation of a broader based institutional execution and other financial services, which are services provided by Andy and Dan, that reflects that. And I think the difference — parroting back Andy’s comments — are — is it continued execution of a pre-existing business model or does being a broad-based liquidity pool become the business model in and of itself? So, I don’t think about fair access. When I think about Reg ATS, I think about what are the things that would warrant taking away the exemption all together. That is what Reg ATS is, right? It is an exemption from the registration requirements under Section 6 right? So, there are, there are — and Andy alluded to them — some order routing and order handling practices that you could start to look at, you know: connectivity, especially connectivity to multiple, unaffiliated non-displayed trading venues, getting past the size threshold that’s, you know, forget about bringing down the five percent fair access; what about bringing down the 20 percent threshold generally for the exemption? I think there are other ways to look at this outside of the context of what’s the right threshold for fair access when a non-display liquidity venue is really embarking on a strategy that is about broad liquidity provision and where selective provision to customers other than retail, other investors may warrant a change in status. Director Robert Cook: Larry Leibowitz:
Because Seth’s point that fragmentation should exist in its competition forgets about the fact that we do have a responsibility to create a good public market. It doesn’t mean that we shouldn’t have fragmentation; we just need to be careful about it. I think the Five Percent Rule is arbitrary and the real question gets back to the original question we asked a couple of questions ago about the total value of the number of exchange venues added up. But I think your last point hits it right on, which is that what we really need to do is say ATS was originally designed to allow competition with exchanges. Well, the truth is that the tables have been turned; it is very difficult for exchanges to compete with ATSs now, and in fact, we can’t even own an ATS because then it becomes a facility and is subject to the same regulatory burdens. I mean, that would be one way around this is, okay, exchanges are allowed to own ATSs. We are allowed to discriminate customers within those ATSs. We’re allowed to do some of those practices. But I think going through what the exemptions are for an ATS item by item and say, “Does this still make sense?” Because the broad category of ATS does make sense; it makes sense that Dan can run his business and he doesn’t have to file to be an exchange. The question is where is the balance between the difference between us and the practices that we have and the costs? But also how much it slows us down and hampers our ability to compete. Dan Mathisson: Larry Leibowitz:
Dan Mathisson: Larry Leibowitz: Dan Mathisson: Now, the trade off is, you are right, it takes a long time to get rules approved. It’s a different regulatory regime, but it needs to be weighed versus the economic advantages. You can’t level the playing field on one side and then not level it on both sides. You need to either actually completely level the playing field, or you need to just acknowledge that these are two completely different regulatory regimes currently with different costs and benefits. William O’Brien: [laughter] I left a good job at NASDAQ for Direct Edge; why would anybody do that? Eric Noll: And the interaction between ATSs and exchanges in other market facilities is a healthy competition and one that we should encourage, but what we should be after really is is are we maximizing public price formation and public price discovery? Are we maximizing liquidity discovery? And lastly, are we seeing order isolation? Right? Are we having orders that are either posted on a public market that are not getting executed because they can’t access or otherwise interact with orders that are taking place in the dark, and likewise, are there orders in the dark that are not accessing publicly displayed liquidity in some way because of the rules of that ATS? One final point on that I’d like to make, though, is that as we talk about fair access and access to ATSs, I think we should also keep in mind that pricing is part of that access debate. So, we often talk about membership and who we let in and who we don’t let in, but I think it’s important from an exchange point of view and I think it’s also relevant from an ATS point of view that we are not allowed to discriminate on price as an exchange operator; fairly so. But other marketplaces do discriminate based on who the customer is on price, and I think it’s very important that we keep that in mind as we debate access, that we not lose sight the pricing about access is also important, and a necessary condition about what fair access is. Brian Conroy:
So, it goes back to the original argument about competition between venues, competition between brokers. So, the stitching here, to a certain extent between competing venues, is the actual products that the marketplace participants use to source liquidity and achieve their fiduciary obligation to seek best execution for their clients. Larry Leibowitz: So, while you guys probably are at the extreme end of actually fulfilling that obligation, and maybe even developing your own tools to do it, there are many other brokers in the marketplace or buy-side institutions in the marketplace that aren’t as advanced as you are. Director Robert Cook:
Dan Mathisson:
Seth Merrin:
William O’Brien: Larry Leibowitz:
I mean, there is a whole lot of pieces in this that get towards just more transparency, recognizing the market for what it is today. I think in its time, it was a great tool. When it first came out, nobody had any clue as to what the actual quality that was occurring in these venues was. And for us as — I was at an internalizer at the time; we used it as a great marketing tool against our competitors, but I think we just need to revisit a lot of the details of it. William O’Brien:
Mark Ready:
Brian Conroy:
Larry Leibowitz: Chairman Mary Schapiro: William O’Brien: Larry Leibowitz:
And so what tends to happen is the limit orders get — if you look at what happens to their order — to the stock after the limit order, it will tend to move against the limit order more often than not. That’s why exchanges end up having to pay for limit orders, where as in the internalizer world, it is the opposite; they are paying for market orders, not for limit orders. It is the opposite. So, when we talk about toxicity, what we are talking about is the toxicity of market flow, which is exhaust, essentially, hitting public exchanges. William O’Brien: Larry Leibowitz: William O’Brien: Larry Leibowitz: William O’Brien:
Larry Leibowitz:
Brian Conroy: So, I think to the extent that there are limit orders displayed on exchanges, to draw the parallel that they are at the whim of toxic order flow is, I think, a little — is understating the real issue of there is a complex marketplace; 60 percent is arguably HFT or some part of an algo strategy where, yes, we may be searching in dark pools, at the same time posting orders in the marketplace, at the same time sending market orders to dark pools in the marketplace depending on how aggressive our strategy is. And again, that has to do with the tools we use and the brokers with whom we desire to work with, and so it’s not just retail order flow that is sitting posted on an exchange at the mercy of more sophisticated investors. Seth Merrin:
Male Speaker: Dan Mathisson:
So, I have an account on E*TRADE. I am allowed to — with a 30-day holding period, I am allowed to trade ETFs. It is — you get an instant execution for eight dollars at a price better than the offer. Now, most retail orders are market orders and they want an instant execution. They get that instant execution, so it would be more than the tenth of a cent. They would also have to then pay the take fee, so it would be about a four-tenth of a cent swing, plus the increase in commission if their broker-dealer is no longer getting the rebate. Seth Merrin: Andrew Silverman: Larry Leibowitz: William O’Brien: Director Robert Cook:
So, for example, there has been proposals that there be disclosure about whether order information of certain customers is made available to other customers, or about just having the form ATS be publically available. Should we consider pursuing those or other types of transparency initiatives with respect to the business operations of ATSs? Eric Noll:
Andrew Silverman: The panel before, High Frequency Trading, a person was talking about milliseconds; a blink of an eye is 200 milliseconds. You can shop an order far and wide in eight. Bottom line is — and I will just talk about Direct Edge for a moment — Direct Edge has an EOP program, and I would say Bill is 100 percent clear on what his value proposition is, what he does with an order. He is great. Seth, do you know what he’s doing with his order flow? But there are not a lot of people out there that have the transparency that they should have. And I have to tell you that this is all about the client. This isn’t about the economics; this is about the client. And I would say the transparency was mind boggling, what was going on, and that the clients weren’t informed. So, any business model is great. Competition I think is exceptionally important, but let’s just make sure that we’re competing and informing the client where order flow is going and what the economic benefit would be, you know, to the firm and why their order handling practices are the way they are. So, you know, this is something that I think we need to really clarify going forward, because without it, you lose trust with our clients. Brian Conroy: Andrew Silverman: You had information that order flow that was double and triple counted, and to me, we also need that post-trade transparency of some sort to make sure that the client is in formed. Now, it might be at the end of the day, it might be on T+1 basis, but again, there has to be some standards with regards to that, because the client cannot feel there is any monkey business with their order flow. Seth Merrin: Director Robert Cook: [end of transcript]
http://www.sec.gov/news/otherwebcasts/2010/060210marketstructure-trans3.htm
|