What Do Institutional Investors Want in a Securities Trading System?

October 21, 2003

Unedited Transcript Prepared from a tape recording.

2:45 p.m.	Registration
3:00	Presenter:	John Colon, Greenwich Associates
	Discussants:	Kenneth M. Lehn, University of Pittsburgh
		Steve Sachs, Rydex Global Advisors
	Moderator:	Peter J. Wallison, AEI
5:00	Adjournment

Proceedings:
MR. WALLISON: The concept that all investors want the best price is embodied in what the SEC considers the concept of best execution of a securities trade, and it is also the key concept that appears to underlie the SEC's support for maintaining the New York Stock Exchange as a dominant and centralized market venue.

It's clearly true, for example, that if all trading in New York Stock Exchange securities were centralized, large numbers of investors would find the best prices because that would be a highly liquid market where all buyers and sellers are meeting.

However, there were strong indications in the May conference that institutional investors do not consider best price, as I indicated before, at any moment in time to be the most important criterion for a satisfactory security trading system. Although price is certainly a consideration, they seem to place a higher priority on other factors, primarily such things as market impact of their trades, trading costs, anonymity, and speed of execution.

The NYSE, they argued, as a centralized, open-outcry auction market dominated by specialists, was not meeting their needs in these other dimensions. If true, this calls into question the entire securities market structure in the United States because it is underpinned by the idea that all investors always want the best price available at the time they trade, and that can only be achieved through a centralized market such as the New York Stock Exchange.

It's important to note that the U.S. securities market has not assumed its current form because of the demands or needs of investors. This form is the result of choices made by the SEC as the market's regulator. In other areas of the economy--we might cite food sales, for example--market structure is dictated by what consumers want. When transportation became easier because of automobiles, the corner grocery story gave way to the supermarket. In effect, consumers created this structure because they preferred the lower prices and the broader selection available at supermarkets.

In general, where there is no regulation of market structure, we can be reasonably sure that the market will assume a shape that meets the needs of consumers. That is not always true of regulated markets, which are frequently shaped by regulatory decisions. And since regulation is a political process and has a tendency to freeze in place outmoded structures, we should be wary when we hear that any significant class of consumers--in this case, institutional investors--is displeased with the services they are receiving from a regulated market. That's why the views of institutional investors exposed at the May conference were so important. These views suggested that the SEC might be supporting an outmoded and inefficient market structure based on a faulty view of what constitutes best execution.

However, until the Greenwich Associates study, the one we will be considering today, there was no way to ascertain whether the views of the institutional investors expressed at the May conference were unique to them, at least to those who appeared, or otherwise reflected the views of the institutional investor community as a whole.

As a survey of a cross-section of this community, the Greenwich study goes some distance toward answering this question and for that reasons deserves careful consideration. If, as the study seems to show, the New York Stock Exchange in its current form is not meeting the needs of institutional investors, a number of questions arise.

Can the NYSE alter its form so that its services do satisfy the demands of institutional investors? Can it do so and still serve the investors who find its services fully satisfactory? If it can't do both, would it be better to open up trading in New York Stock Exchange-listed securities so that other trading venues, such as ECNs, would be better able to compete with the New York Stock Exchange? Or would it simply be better to maintain the NYSE as the central trading venue in the securities markets, despite the dissatisfaction of institutional investors? These are questions we will be considering at future conferences during this study.

Now, for today's panel, we have that study prepared by Greenwich Associates, and John Colon of Greenwich Associates will be presenting it. He is in charge of the international business--I'm sorry. He's in charge of the investment banking and institutional equity brokerage practices for Greenwich Associates in the United States, Asia, and Europe. Before joining Greenwich Associates, he was an analyst in the corporate finance department of Lehman Brothers (?) in New York.

You have in your packages bios of the others who will be on this program, but let me just mention a few things about them.

Ken Lehn is the Samuel A. McCullough Professor of Finance at the Katz School of Business at the University of Pittsburgh, where he teaches courses on business valuation and corporate structuring. Ken is a professor of law at the University of Pittsburgh Law School in addition. He joined the faculty of the University of Pittsburgh in 1991 after serving as chief economist of the U.S. Securities and Exchange Commission for four years. He has also published many articles in academic journals.

Steve Sachs, who is right next to me, is the director of trading for Rydex Global Advisors. He has 12 years of experience trading securities in the financial markets. In 2002, he began at Rydex as director of trading, where he oversees all securities and derivatives trading.

I think what we will do is begin with John. John will make his presentation, and then we'll have comments first from Steve and then from Ken.

John?

MR. COLON: Thank you. First, we'll see if we can get the technology to work properly. That's one of major challenges.

Obviously, the topic is a fairly timely one and a fairly broad one. From Peter's comments and from others', it's fairly clear that the concept of best execution remains a fairly elusive concept. It tends to differ quite significantly depending upon the particular investment organization, its objectives, the particular orders that it's looking to execute, the particular market conditions that exist at a given point in time. Even the guidelines provided by an organization such as AIMR, which are fairly widely followed, describe the pursuit of best execution as much as a process as opposed to necessarily a specific set of outcomes. In other words, you have a process as an investment organization designed to help produce the best outcome, but it's not a transaction-by-transaction sort of process necessarily.

But anonymity, price improvement, certainty of your orders being executed, market i mpact, speed--they're the motherhood and apple pie of order execution. They're all considered to be important. But, again, how you prioritize them or what is the optimal mix of those ingredients can differ quite significantly from order to order to order.

To try to get some additional insight into this, we were actually initially approached by Instinet and asked to speak to institutional traders to get their views as to whether or not particular trading venues were more likely to deliver on these particular requirements or basics behind best execution.

To that end, we conducted a set of interviews in the first half of September of this year. Those we interviewed were senior traders from 103 institutions. They collectively managed over $2.5 trillion in assets, or an average of about $25 billion per institution. About half of the institutions would have had assets under management in excess of $10 billion. Most of the rest would have been between 1 and 10. Relatively few would have fallen below, say, a $1 billion asset hurdle. So these do tend to be larger, more active investors, and as Peter points out, the focus of this is the institutional marketplace. These represent larger institutions in that marketplace, and they're certainly not representative necessarily of the retail marketplace.

The traders from these institutions represent multiple investment styles. In fact, in one of the more detailed DECs(?) you have in the folder, you find that when we asked the question as to what were your investment styles, it adds up to well over 100 percent. That's simply because these large active managers do tend to have portfolios following multiple different strategies. It is predominantly, however, active investors, so not passive or quantitative portfolios. And it tends to be a mix of value, growth, both large cap, small cap. But it does tend to be a group of active investors as opposed to indexed funds that may simply be looking to really track an index.

On average within these institutions, about two-thirds of their volume is in exchange-listed stocks; the remainder would tend to be in Nasdaq stocks. They'll have small amounts in ADRs, exchange-traded funds, et cetera, but the bulk of the activity is in basic New York Stock Exchange-listed stocks as well as Nasdaq stocks.

There are a series of top-line findings that we've put forward. In my own view, they're not necessarily definitive, but they help delineate the lines of some of the arguments. And I think as you go through, you'll find that the arguments tend to revolve around the benefits of technology versus human touch, and they also tend to revolve around the perceived benefits of centralized markets versus having multiple competing venues. Quite often, you can tie a number of the comments or seeming discrepancies in the information in terms of where do traders come out in terms of the various camps on these issues.

I did once have a colleague who admonished one of our clients that their clients were not encumbered by conviction. Now, what that means is that, again, given the specific circumstances, the specific order, the traders that we're speaking to might give quite different answers. Essentially, they may pursue whatever is most opportunistic and advantageous at a particular point in time. But, nonetheless, the basic boundaries of the argument I think begin to become fairly clear.

There is one finding that obviously stands out to anyone flipping through the results of the study that we've done, and that is clearly the view that the specialist is likely to be more of a hindrance than a help in terms of the overall process of traders trying to achieve their objective. Obviously, that's not a particularly surprising finding given what's recently been published in the press, but certainly our interviews would tend to confirm that.

So in terms of the top-line findings, low market impact, anonymity, price improvement, certainty of execution are all major requirements when executing orders. There is a view that ECNs hold a substantial advantage or, I should say, greater likelihood of being able to deliver anonymity than are either a broker providing upstairs order execution or through an exchange. They're more likely to deliver low market impact relative to an exchange or an upstairs broker.

On the other hand, the exchange and brokers hold modest advantages over ECNs in terms of certainty that a transaction will, in fact, be executed in a timely fashion and in full. Over three-quarters of traders, however, would consider proprietary trading by the specialist to be a conflict of interest. Two-thirds of the traders that we spoke to felt that specialist Nasdaq market makers do not add value in trading large liquid stocks. Again, you've got really two central elements coming into play between those two statements. One is a question as to whether or not specialists can really fulfill the central role of maintaining an orderly market. If a stock is liquid, active, then you don't really necessarily need a lot of help to maintain good order. If the market is plummeting, or racing forward, perhaps, does the specialist really have the capital and the will power to stand in front of that tidal wave?

So you've got the two opposite ends of the spectrum. Are they really needed in a good market environment? And are they really capable of providing their basic function in a truly awful market environment? If you put on top of that the question mark about the objectivity of the specialists, and it's a one-two punch.

Nearly half of the respondents that we spoke to, the traders that we spoke to, would prefer to trade more volume off of the exchange floor. Again, the strength of the response was fairly clear, with the main reason being cited mistrust of the specialists. But other areas were cited as well in terms of lack of visible liquidity, order depth, anonymity, market impact, the need to break up orders into smaller blocks in order to get them executed.

There are, however, a significant number of traders who would state that they prefer not to move volume off of the exchange floor, and the most frequently cited reasons tend to relate to the perceived benefits of centralization and human touch. So, again, you've got certain people looking at the specialists and saying that's a problem to have that set of humans involved in the process. You've got others who say that it's actually a positive to have humans involved in the process, and we'll touch on that a bit more in some of the detail.

A majority of the traders did not see multiple venues as a negative. That said, a significant minority do see it as a negative. I think it does raise the issue that traders are not only looking at best execution, but they're also looking at what makes for an efficient process. There have been a number of efforts by organizations to put forward new trading systems that have foundered simply because they were almost impossible to use--too complex, too time-consuming, trading desks didn't have the ability or the wherewithal to deal with them.

So certainly one of the areas that traders are looking for is efficiency in their own work flow in addition to the actual quality of execution they receive, and there are certainly quite a few that appear to view multiple venues as a complexity in the process, even if they believe that those venues may be able to match a centralized exchange in terms of liquidity. There are, nonetheless, a series of changes which institutions would put their hands up and say, Of course, if these existed, we'd be favorably disposed. So taking away the competition from specialists, having a liquid electronic marketplace in listed stocks, I think again you have to underline the word "liquid." If it's liquid and it's efficient, of course, why wouldn't they want it? Real-time limit order book, integrated quote display, again, looking for transparency in the trading process.

So to touch on some of these different elements, we asked traders to just rate simply the overall quality of execution in listed stocks, so it wasn't specifically Nasdaq--or, excuse me, it was specifically listed, not general execution or Nasdaq, as to which was providing the best executions. I think you can view this a couple of different ways. One is that there is a slight leaning towards ECNs as being the most effective venue for executions in listed stocks. But it is not dramatically different from the upstairs brokers. You do find, again, somewhat of a preference for ECNs and upstairs brokers relative to orders sent to the exchange, but you certainly don't see a powerful negative coming through as it relates to the perceptions of exchange-based executions.

So, again, you've got some view towards ECNs as a more efficient mode, but you don't necessarily have a complete dismissal of exchange-oriented trading in terms of its ability to provide best execution.

Again, a bit of motherhood and apple pie. We asked traders: What are the most important attributes in executing orders in listed stocks? Again, the question was phrased in terms of listed stocks, but, frankly, we would have gotten exactly the same answer in terms of executing any kind of stocks: anonymity, price improvement, the certainty of execution, minimizing market impact, all cited by a majority as being either very important or essential in terms of their trading process. Speed of execution comes up perhaps a little bit more in the mid-range. Frankly, I think that's simply because speed of execution can actually run counter to some of these other requirements, depending on a particular order. In other words, in order to obtain speed, I might have to suffer significant market impact. So I might prefer to actually have an order gradually worked into the marketplace as opposed to taking the risk of that market impact.

So, again, I would certainly not take this to mean that speed of execution is unimportant. Depending on the nature of a particular order, it may, in fact, be the most important element.

But, again, it tends to say that they're looking for everything, and with the individual transaction, however, the prioritization of these factors certainly may change.

We asked traders where they felt they had the highest probability of achieving these objectives across different market venues. By about a three-to-one margin, they felt that they had a higher probability of achieving a low market impact and executing orders through ECNs or alternative trading systems than they did in executing orders through an exchange. They were also somewhat more likely than dealing through a broker. There's certainly a dramatic advantage perceived to exist with the ECNs and alternative trading systems in terms of speed of execution. Frankly, none were seen to necessarily have a significant advantage in terms of the potential for price improvement. Certainty, the likelihood of success, tended to differ or to move a bit more towards the exchange or to brokers. And, again, speed of execution, the advantage tended to shift back, again, to the ECNs or alternative trading systems.

So in a series of different elements that are, again, the basics that people would look to in terms of best execution, ECNs or alternative trading systems are seen to be more likely to deliver against those objectives than are other venues.

Some of the reason behind the views on the exchange certainly may relate to perceptions of the role of the specialist. As I mentioned previously, you've got two different aspects. One is fulfillment of their basic function, and then the other being the potential for conflict of interest. So the question: Do specialists and market makers still add value in highly liquid stocks? Roughly 70 percent of the traders we spoke to said no, they don't. So, again, if a stock is, in fact, highly liquid, do you, in fact, need somebody to keep an orderly market, or do the markets tend to be orderly by their very nature?

Now, again, we did not ask the specific question around their ability to step up in a disorderly marketplace, but, again, these are organizations which are profit driven, and the surefire way to not be successful in maintaining a profit is to try to stand in the way of a tidal wave should the markets be in the more difficult situation.

You get a much more powerful statement around the perception of whether the s pecialist's ability to trade on a proprietary basis constitutes a conflict of interest with over three-quarters of the traders that we spoke to expressing the view that it did, in fact, constitute a conflict of interest. When asked whether or not traders would then move more business or prefer to move business away from the exchange floor, frankly, you could interpret this a couple of different ways. You could say that 47 percent saying that they'd like to do more volume away from the floor is a fairly powerful indictment. But then if you also have three-quarters or over three-quarters complaining about the specialist structure, you might argue the flip side that 44 percent saying they don't plan to move says that there's a considerable number who feel that the process is fine. But, nonetheless, you do have traders splitting into a couple of camps in terms of those who are, relatively speaking, content with the current process versus those who are expressing dissatisfaction.

Why do they want to move business away from the floor? Again, the reasons cited--and there are significantly more quotes in the detailed packages that you have, but I think these are fairly representative: The specialist is a well-funded competitor; the primary problem with the specialist is (?) , in other words, stepping in front of the customer orders, and lack of visible liquidity; a view that liquidity is drying up in the marketplace; and then, finally, there are too many intermediaries who don't have my best interest at heart.

So you have a number of traders essentially viewing the exchange and, in particular, the specialist as being a competitor rather than necessarily a facilitator in terms of the trading process.

Now, again, when we asked where they were executing their volume in exchange-listed stocks, it's about 80 percent on the exchange; 12 to 13 percent off the exchange, some business going through Nasdaq in specific; and then you've got on the order of 9 percent accounts would say that they are executing in listed business through ECNs or alternative trading systems. So, again, you have this conundrum. Over half or roughly half of the people we speak to would say they'd like to move business away from the exchange floor, but this would imply it's not happening, it's staying where it is. Why?

Typically, the reasons that traders cite as to why they would not move business away from the exchange floor relates to centralization. But there is also this element that comes into play, again, which I would describe as the human touch. Central location makes it faster for buyers and sellers to meet. On the floor, there's more interaction, interaction of buy and sell orders, which provides a better chance for price improvement. One trader simply saying, "I don't like fragmentation of the market; I want to see all the players in one place." And, finally, "I prefer to deal with people; there's more control."

Now, it's not within the group of slides that I've chosen to look at here, but, again, with the more detailed package that you have, we did ask a question in a slightly different context, which was about the relative merits of putting transactions through Superdot versus floor brokers. And while it didn't speak specifically to the difference between floor brokers and specialists, I did find it quite interesting that there was not the vehement negativism around the floor brokers that there seems to be around the specialists. And if you have a moment at a later time to look through that portion of the research, you'll find comments such as: Human intervention of the broker, with proper instruction, adds value; I'll be protected by the guy who is working with me.

Basically, what you're seeing is that the issue of information and who has it and how do they use it is key. And there appears to be a perception that the specialist isn't always using that information in the best interest of the ultimate trader; whereas, with the floor brokers, you have a different mechanism in that, first of all, there is not a monopoly in that position. And when you read through the quotes, you'll get a fairly strong sense that the trading desks of major institutions feel like they have more control in terms of interactions with floor brokers because they have actually the carrot and the stick in terms of how much volume do they put through that particular floor broker. They feel that that floor broker is not acting on their behalf or somehow leaking information, trading to their disadvantage. They have any ability to discipline that. Or, again, with the specialist system, I believe they don't feel that they have that degree of control or ability to discipline the process.

Again, one of the central reasons cited for maintaining business on the exchange are the benefits of a centralized interaction of buyers and sellers. When asked explicitly whether or not competing venues affect traders' ability to achieve best execution, nearly half simply said it had no effect; 15 percent said it actually makes it easier; 37 percent said it makes it harder.

So, again, you've got, arguably, a majority at least in a position of saying having multiple competing venues either has no effect or, in fact, might make trading somewhat easier. But, again, it's a fairly considerable minority who would say that multiple competing venues make it harder. Again, we did not ask a specific follow-on question around that topic, but I believe that part of what you're seeing there is, again, traders looking not only at simply best execution, but also what is the most efficient process in terms of their management of their order flow and their own work flow. And at times, if they've got to be concerned about multiple different venues, how do I tape into them, how do I communicate and route the orders, et cetera, that can become somewhat of a burden and I believe underlies some of the differences in perception.

But when we asked simply about the overall quality of execution between the New York Stock Exchange-listed stocks and Nasdaq, again, I believe that it doesn't paint a particularly clear picture. You had roughly 39 percent of those we spoke to said that listed executions were either substantially or somewhat superior; you had 31 percent saying that listed execution was somewhat or substantially inferior; and you had 30 percent saying it was about the same. So while it's not exactly thirds, it's fairly close to it.

What, then, are some of the things that, again, people would look at and say that they would like to see as far as changes? Again, you would get a fairly high proportion, over three-quarters of those we spoke to, tending to say that it would be valuable or very valuable to have an electronic market for listed stocks with sufficient liquidity to compete with the New York Stock Exchange. Again, I think it's important to bear in mind that the key words there are "sufficient liquidity." So part of the difficulty in this whole debate is you don't necessarily have a liquid central electronic market. You have a central market, which is not electronic, and then you have a series of multiple venues operating in an electronic marketplace. So you don't have the third alternative really in place for people to draw a comparison.

But certainly if you had an electronic system, it could provide the liquidity, which is a key requirement. People would be favorably disposed.

You do certainly see a vast majority say that they'd like to see changes to the rules to prevent specialists from competing with customer orders. So, again, they'd like to see the elimination of conflicts.

You've got less clarity around changing the rules to allow stocks to be represented by more than one specialist. Now, again, arguably, that would be a means to introduce some competition and disciplining into the process if, in fact, a trader could say, well, I won't trade with this specialist, I will trade with that one. But, again, I'm not sure that the people we were speaking to necessarily focused specifically on that question as a way to introduce competition and, therefore, somewhat more of a disciplining mechanism into the process.

Would people tend to be favorably disposed to a facility that simultaneously exposed orders on the floor into electronic venues? Yes, they would tend to be favorably disposed in that direction. Would they like to see an integrated single-order display? The answer tends to be yes. Would they like to see a real-time order book? Yes. A little bit less clarity around changes to some of the other related systems such as the ITS system.

So, again, I think that the results that we have help frame the debate. I don't think that they provide necessarily a definitive answer. Best execution is, again, to some extent, clearly in the eyes of the beholder. There are certainly a number of different aspects on which ECNs and alternative trading systems are seen to deliver a greater likelihood of a positive result. But you, nonetheless, have a significant number of investors who, A, like a centralized marketplace and are concerned about fragmentation; and you also have a significant number of people who still like the human touch, as long as they believe that that individual is working for them.

It's a little bit like a poker game. I'd love to have you tip your hand as long as I don't have to tip mine, and I'm going to be really upset if the guy standing behind me is tipping my hand.

So people are looking for information flows. They're looking to gain an advantage, but they want that to be done, as they would see it, on a fair and evenhanded basis.

MR. WALLISON: John, thanks very much.

Steve?

MR. SACHS: Good afternoon. The question of what do institutional investors want in a trading system, very good question. The answer is probably greatly dependent upon what kind of day the trader is having when you ask him the question.

A recent article on best execution in the Journal of Portfolio Management stated that improving market quality is the overriding objective for a market center; the important question is how to implement the objective.

I think this statement really hits the heart of the matter that we're discussing. There are numerous changes that need to be made to the current listed trading structure, but, more importantly, these changes need to be implemented in a way that allows institutions to manage money successfully. The bulk of American investors still get their exposure to the markets via mutual funds and professional money managers. So why not design a structure that will benefit the most number of people?

I thought I would discuss the top-line findings from the study and just give you some of my thoughts on each as opposed to digging into the real detail of each of the questions, and then hopefully get some discussion in the Q&A session going.

Low market impact is the most important requirement when executing orders in listed stocks. Yes, without a doubt. Low market impact for me personally is always the objective. This point goes to the topic of price discovery and quantity discovery. Decimal pricing has greatly reduced the depth of orders at each price level. Bidding or offering for large size is not rewarded in the current decimal environment. The use of VWAP orders by institutions, both, you know, in the U.S. and in Europe, has even further reduced the need to bid or offer for large size at any particular price point.

MR. WALLISON: Would you define that, VWAP?

MR. SACHS: Volume weighted average price, so the idea being that your execution price over the course of a time period is, in fact, the average price based on the volume that's traded over that time period. It's an exceptionally passive way to trade in the marketplace.

The average trade size and average posted bid offer size or resting liquidity size, as they call it, have both decreased since we moved to decimal pricing.

So point two, ECNs are three times more likely than an exchange to deliver low market impact and twice as likely as an upstairs broker. Yes, impact can certainly be controlled better with the use of an ECN. It's why they're such valuable tools for institutions.

It makes sense if you're the only one in control of your trade, you're going to have less impact, less information leakage in the marketplace. But you're also going to get fragmentation of the market, and this leads to a whole host of other issues, you know, not the least of which is opportunity costs. And one of the things that we'll generally give up for lower impact cost is the risk of higher opportunity costs. With this fragmented market in an ECN environment, if the liquidity is now, for example, on the floor of the New York Stock Exchange, I now have to access the floor. If I want to do that in an anonymous way, the only way for me to do it currently is via Superdot. Now my order gets exposed to the trading crowd at the post and the specialist before it gets executed. It's sort of a Catch-22, you know, the basic problem behind fragmentation.

Point number three, two-thirds of traders do not think NYSE specialists and Nasdaq market makers add value in trading large liquid stocks. Without a doubt, I would agree with this. If you could provide an environment where you could match up orders electronically anonymously between institutions and say, you know, the top 100 liquid names in the marketplace, there would be no need for a specialist or market maker.

For example, the question is IBM. If you were to remove the specialist and/or competitive market makers from the market in IBM, would the institutional order, you know, demand--supply and demand be able to maintain a fair and orderly market in all market environments?

My answer to that question is yes. I think that, you know, any given day you can look at the depth of activity in IBM and make that statement.

Crossing networks, Posit, Liquid Net, the various ECNs that are out there I think prove out this point that, you know, given an electronic environment for liquid securities, you know, the market can be maintained.

The second point is that 78 percent of traders considered proprietary trading by the specialists to be a conflict of interest. I would agree with this statement as well. The problem is the only way for a market maker or a specialist to maintain an orderly market is if it's profitable for him to do so. I believe John mentioned that in the study.

The need for this profitability creates the conflict, so the question becomes, you know, Is there a need to maintain a fair and orderly market in IBM? My answer to that question is no. Well, what about the rest of the Russell 3000, for example, the small-cap and mid-cap securities? On any given day at any given moment, there isn't a liquidity there to maintain the market. You may be looking at, you know, spreads of multiple dollars, even ten dollars wide, with, you know, no retail or institutional order flow posted in the limit order book.

Number four, institutions would like to see changes in the listed market. In particular, they strongly endorse not allowing specialists to compete with customers. Again, this question goes back to what do you do with the less liquid stocks in order to maintain the market with no specialists? Real-time order books and quote displays work well, again, with the top end of the liquidity scale, but not small-cap and mid-cap stocks.

How do you reconcile this, the need for the specialists to be profitable and the need to maintain a fair and orderly market along with the desire for institutions to have this anonymous electronic environment or the natural matching of buy-and-sell orders?

The answer in my mind lies somewhere in between the structure that we currently have within Nasdaq and within the New York Stock Exchange. If I could have anonymous electronic access to an open order book that had equal standing at the trading post as opposed to what we have now, that type of structure I think is what most of the respondents in the survey were probably referring to when they were talking about access. Specialists need to maintain markets. They need to interfere, if you will, or they need to compete with customer orders at some point in time. The question is limiting--or the answer is limiting this interference. We currently have rules in place to do this, so the actual question is: Are the rules wrong or is simply the oversight not there in the current rule structure?

Most importantly--and, you know, I'm not sure where I stand on this with the rest of the institutional world--I think the move to decimal pricing needs to be reversed. I would be a proponent of, for example, nickel increments in stocks that trade, say, above $10 per share, the idea being that it would encourage liquidity and depth of liquidity at particular price points, which is exactly what most institutions are looking for.

Again, if we can find the correct price, find the price that we're willing to trade at and find the depth at that price, you know, we are then able to limit the impact and at the same time hopefully limit the opportunity costs that we're exposed to in the marketplace.

You know, again, the move to decimal pricing I think was very beneficial for individual investors, but at the expense of institutional investors. And, again, the idea is that if most individual investors are still getting their exposure to the market via mutual funds, wouldn't it make more sense to make the orders less expensive for those institutions that are managing the money?

The last one, number five, nearly half the respondents would prefer to trade more off the exchange floor, the most common reason mistrust of the specialists. This goes back directly to my last few points in that, you know, we need a structure that encourages both price and quantity discovery and allows for natural order flow interaction, either in an electronic environment or without interference from a trading crowd or a specialist at the post. You know, that being said, I think the most likely scenario is an electronic environment where we can get that type of anonymity and liquidity that we're looking for.

And my last point on that is really implementation. I think this is key. Over the last 18 months, NYSE launched both the institutional express system and the liquidity quote display system. Liquidity quote was a good idea on paper. The idea was that the specialists would post an institutional size market around the retail or limit order book market on the NYSE. Exactly what institutions are looking for: the ability to trade inside the particular bid or offer that may not necessarily be the tightest market, but we know at a nickel or a dime spread, we can actually gain the liquidity that we're looking for.

That was the general idea behind the institutional express system, the ability to give traders like myself the ability to execute against that quote immediately, with no interference from anyone else in the crowd. The problem with that is that there has to be certain conditions that exist for an order to be considered institutional express.

I can tell you that over the course of the last 18 months I've only found one condition where that actual express condition existed, and I was able to access that liquidity. Again, it goes back to the point of the structure right now is not allowing us the depth of liquidity and the price and quantity discovery that we're looking for.

I think there is an answer to this question. There's no doubt about it. You know, we've spent the last decade restructuring the over-the-counter markets, and I think that this is really no different of a situation.

The point with the NYSE or with the listed markets really comes down to oversight as well. You know, one of the things that we're lacking right now with the listed markets, and I think it is reflected in some of the answers from the respondents, is integrity. And that's the cornerstone of the--should be the cornerstone of the exchange, and ultimately, you know, the free markets, the ability for corporations to raise money and for stocks to trade actively in a national market system is the foundation of the structure of our economy. I think that we're in a very precarious position right now with the current environment that we're in.

Those are the comments I have.

MR. WALLISON: Thank you, Steve.

Ken, why don't you go ahead.

MR. LEHN: I may be the academic analogue of the specialist [inaudible, off microphone] -- in the opening comments, and that is that the traditional exchanges and marketplaces have not been, as Mr. Zarb (ph) said, keeping pace with the needs of the market. And, you know, we've seen with the changes in technology and changes in customer demands a lot of longstanding organizations and different industries be upheaved a bit by start-ups. If we look at the airline industry, the old trunk carriers are having their clocks cleaned by new entrants who seem more responsive to customer demand. We're seeing that, of course, in telecommunications. We see in the computer industry. So this is not a unique situation.

The Greenwich Associates survey is certainly good news for the ECNs, and just briefly summarizing again what John indicated, the three major points that I saw coming out of this survey is that the ECNs trump the NYSE on anonymity, speed of execution, and market impact; number two, the respondents believe specialists add no value in liquid stocks, such as IBM, and that proprietary trading by specialists does constitute a conflict; and then, third, almost one-half of the respondents want to trade more away from the exchange floor. And if it already hasn't received enough wake-up calls, I would think that this survey would provide another wake-up call to the NYSE.

The Greenwich study is also consistent with academic evidence. There are a number of studies that are beginning to emerge on comparative analyses of the ECNs and other venues. And the two major findings that come out in the literature are, number one, that execution costs are lower for ECN trades than for broker-filled orders, and there are three studies that I cite on the slide, and certainly if anyone is interested in the precise slide, I'm happy to give that to you. And then, second--which I also find interesting--is that the ECNs have been documented to contribute to price discovery. A study by Roger Wong (ph) that documents that for liquid Nasdaq stocks, and then Hendershot and Jones' finding that the island ECN contributed to price discovery for ETFs, at least until September of 2002, when for regulatory reasons they no longer posted quotes. And since then the performance has been diminished.

And I should add that I'm not aware of any studies, at least in the academic arena, that have compared the price discovery of the NYSE with the price discovery of ECNs. That seems to be a gap in the literature, as far as I can see.

I'd like to focus my brief comments on two areas. One would be a set of comments on the study itself, methodological type issues that often can cause the eyes to glaze over, but, nonetheless, I'll raise a few; and then, second, some implications for both the NYSE and for regulators.

The first comment I have on the Greenwich study is a question as to how representative the sample is of all institutional investors. You know, as John indicated, it certainly would not be representative of retail investors, who still constitute a fairly large section of the market, but there are a lot of large institutional investors that were not surveyed, and I listed a few here: American Funds, Delaware Investments, Dreyfus, Lord Abbott and so forth, Vanguard--although I think if John Bogle had been interviewed, probably the results would have been roughly the same. But it does raise the question, as there always is in survey analyses, as to whether there is any potential selection bias.

The fact that the study was commissioned by Instinet, if that was known to the people who were targeted for the survey, then there could be a selection bias where those that are more favorably disposed to the NYSE might have declined to participate. I'm certainly not arguing there was any intentional selection bias, but certainly it would be worthwhile to know if some institutions were asked to participate and declined to do so.

A second question or comment would be whether the responses are different for different sub-samples of institutional investors. You know, do large institutions answer differently than small institutions? And that raises some issues as to whether the responses maybe should be valuated in some sense so that you provide more weight to institutions that are relatively more important, at least in terms of the assets under their control.

Also, do actively managed funds look for different things than indexed funds? And there has been a growth of indexed funds over time, so it may be interesting to know whether or not there are differences of that sort.

And then third is whether the responses are different for the large-cap versus the small-cap, and obviously also the mid-cap funds. And, again, small caps generally operate in less liquid markets. My guess might be that institutions that manage small-cap funds look for things that are different from institutions that manage large-cap funds.

And then, finally, comment three would be a few other questions that I found myself asking that were not asked in the Greenwich survey. I think John alluded to some of this in his discussion, but I don't think the questions were actually asked of the institutions.

One would be which marketplaces do the best job of executing the so-called hard trades that require large capital commitments.

Second, which marketplaces do the best job of handling trades during stressful periods, for example, after 9/11?

Third--and Steve, I think, made reference to this, at least indirectly--that is, has decimalization made it more attractive to trade on ECNs? You know, has there been a regulatory subsidy in some sense that has contributed to that.

Then, fourth, more questions on the price discovery performance of the different marketplaces. The twin pillars of market quality are the efficiency with which a market provides transaction services, and that's certainly the focus of the survey. But the second dimension of market quality -- [tape ends].

-- discovery function. And one of the issues of concern, of course, is whether ECNs are in some sense free riding off the exchange. The exchange makes investments that contribute to price discovery, and when the limit orders are submitted on the ECNs, the basis for those limit orders might very well be presumably are to a large extent, the prices that are actually formed on the NYSE. And my friend Jim Overdahl (ph) back there has done a study that, among other things, discusses the evolution of the telegraph and how the telegraph in the 1800s created similar problems where there was potential for free riding of the primary market that was making investments in price discovery.

And as I indicated earlier, there are academic studies showing that ECNs do contribute to price discovery for liquid Nasdaq stocks and ETFs. I don't know what the answer is empirically with respect to price discovery on the NYSE versus the ECNs.

What are the implications for the NYSE? As I read the study, and I think more generally about ECNs, it clearly sends the message to me that the raison d'etre of the specialist for certain types of transaction is eroding, namely, transactions by large institutions in high liquid securities. It's very questionable as to whether or not the role of the specialist still exists. I think for less liquid stocks, the intellectual defense of the specialist system is much stronger. But I don't think we should exaggerate the problems at the NYSE, and there are four bullet points I have here. One is there is a large body of academic evidence that does document the NYSE does a good job in terms of execution costs and price discovery. The studies here to date have largely focused on execution costs of the NYSE versus Nasdaq. So it's still an open question, NYSE versus the ECNs. But also in terms of price discover, at least the academic evidence does indicate that it obviously serves a very important price discovery function.

It still accounts for the bulk of the order flow in listed stocks. You know, as John pointed out in the survey, notwithstanding some of the survey responses, the bulk of the order flow is still on the NYSE-listed stocks, and that provides it obviously with an enormous competitive advantage. The network externalities of trading, i.e., that traders want to go where other traders are trading, does provide a competitive advantage that should not be underestimated. In my judgment, again, without empirical evidence, it still has a very strong advantage in price discovery.

And then, finally, the proof's in the pudding to a large extent. If you look at seat prices--and I've gathered some data on NYSE seat prices from 1970 to 2003, and I adjusted them for inflation. There's always a question whether you're adjusting for inflation or adjusting them for movements in the market. I've done it both ways. It doesn't really matter. But what you see is that the seat prices since the growth of ECNs in the mid-1990s, roughly, has been quite good; that notwithstanding the growth of ECNs, the seat prices have certainly held their own, suggesting that market participants have not certainly sounded a death knell of the NYSE.

The main challenge, as I see it, is that the NYSE needs to adapt more effectively to changes in technology and changes in customer demand. The analogy that I always think of is IBM in the late 1970s, 1980s, with the advent of the personal computer, and suddenly at IBM, which was a highly bureaucratic, highly centralized run company, operating in an industry where there were technological changes and changes in consumer demand that were causing the market to change very, very rapidly, and a bureaucratic, highly centralized organizational structure doesn't cut it in that environment. And IBM, of course, had their clocks cleaned with respect to personal computers during that period.

And I see very similar things going on with the NYSE. It's still a very bureaucratic organization, a lot of centralization. Much of it is an artifact of the regulation, and one of the recommendations I would make is that the SEC be more enabling of changes in procedures and allow this organization to become a more nimble organization that can adapt more effectively.

The governance forums that are underway and some of the discussion about changes in governance that are designed largely--at least it appears they're designed largely to deal with some of the fallout from the Grasso compensation issues, I think may have a more important effect, and that is, by affecting the business model that the NYSE will use.

A couple of options that they have. One would be conversion to a for-profit status where there are actually shares rather than seats and a market in which those shares can transact. And, you know, again, there's plenty of academic evidence that would suggest, and common sense, that when you have that kind of a governance structure, you are more subject to monitoring by shareholders; it becomes less of a clubby environment and, I think in many cases, a more efficiently run organization; and you also have the discipline of the corporate control market.

If there are shares out there that are transferable, if Instinet thinks they have a better way of trading stocks, then they can acquire control of the NYSE. Or if Bloomberg or if Bill Gates wants Microsoft to make a takeover attempt at the exchange, then that's much easier than it is, obviously, under the present situation where you have a not-for-profit status with seats, which are somewhat transferable but in a very bureaucratic way.

And then a smaller board, as John Reed has advocated. You know, the NYSE board--I saw Sunday's reports that it had 27 members. If you look at the annual report for 2002, there are 24 members. Just by comparison, Instinet, according to my research, has eight directors, a much smaller board. And there's certainly an argument that when you have eight people as opposed to 24, those eight are more likely to know about the compensation of the CEO; they're more likely to know about the business strategy, the business model. And just as we know when we go out to dinner with friends, it's a lot easier to get eight people to agree on a restaurant than 24, so it's much easier to coordinate policy and react more nimbly.

And in terms of implications for regulators, you know, again, somewhat consistent with what Peter was saying at the beginning, I think one is it has to avoid the one-size-fits-all approach to market structure. I think it's the wrong question to ask which is the better way to trade securities. There may be multiple ways to trade that satisfy different clienteles, and that should be encouraged, not discouraged. We shouldn't focus on trading costs to the detriment of price discovery. Again, I think the price discovery issue is an important one in this whole controversy.

And, third, enable the NYSE to transition to a more nimble governance structure. In my judgment, there's far too much micromanagement of exchange procedures by the SEC. The exchange operated very well from 1792 until 1934 before federal regulation of their internal procedures. And I think that the bureaucratic process, the whole--you know, when I was at the SEC in the 1980s, I remember--I think it took years discussing the issue of whether or not to allow floor brokers to use hand-held telephones. You know, why such an important role by the Federal Government in micromanaging those procedures? I think the government does a much better job on focusing on enforcement of rules, throwing more resources at the Enforcement Division, and really having sizable penalties that catch the attention of people that violate the rules of the exchange, but allow those procedures to largely be determined by the exchange.

And then, finally, to revisit the issue of decimalization, as Steve was indicating. You know, there's a case to be made that decimalization has done more harm than good, and I certainly think that should be part of the policy debate.

MR. WALLISON: Thank you very much, Ken.

I think we have a little time for discussion among the panel and then for questions, so let me first of all ask Ken, can you put up your comment three again? Comment three. It would be useful to get John to answer some of the questions that Ken has up there because those are important, plus the question of how the sample was chosen and the different responses, if you know, between large-cap and small-cap institutional managers.

MR. COLON: The sample is, I think, only biased in the sense that it is towards large active investors. Now, there may have been some self-selection in the process, but in the more detailed materials, in fact, the list of institutions that were initially approached.

A majority of those who responded were willing to let us say that they had, in fact, participated, and that's indicated by an asterisk in that review. But you'll see that there are a significant number of institutions of all different stripes. Again, they typically are large. They're typically active. But they're mutual fund managers, pension fund managers. There are a number of hedge funds within that sampling as well.

We did, in fact, disclose that we were doing this together with Instinet, so, again, there may be some selection in the process where people who might have felt they had a particular bone to pick with the process were more eager to jump in there and take a couple of shots than some others. But I generally find that the trading community is not particularly shy one way or the other. And I think that most of them do carry a genuine interest in trading and its nuances and process and take the process quite seriously.

We have not gone back to look at more fine segmentations. Again, at some point we will get into a question of statistical reliability, but actually it's an interesting question and something we should and will do to see whether there's a substantial difference between the very largest and those who are, nonetheless, large but not enormous. Again, I think if there's a size bias, it tends to be that these are typically large institutions engaged in the process.

I'm not sure that there's a style bias, and the reason, again, that I say that or I don't believe there's a style bias is that most of these institutions, because they are larger, are multi-style managers. Again, if we could have taken more of their time, we may have had more questions around the particular attributes of trading in large-cap stocks or trading in small-cap stocks or whether they have different perceptions whether they're trading for different funds that they manage. But most of these were, in fact, traders who were managing or trading on behalf of an institution that ran multiple styles.

So I feel overall quite confident that we have a fairly representative view as to the marketplace.

I should note that, again, it does not include a significant number of passive or indexed funds. We do find that quite a large volume of that business is actually being done as portfolio trades or basket trading, so that some of the issues may not pertain in quite the same way. You'd see, coming back to Steve's comment, that an enormous amount of that business is done with VWAP as a target, and so in the portfolio business, you may not have quite the same concerns that exist, I think, with some of the active investors. But, again, Steve would probably have a somewhat better sense of that than I would.

But some of the specific questions here, again, I think they are quite useful ones. We did not ask specifically about capital commitment. We certainly do find that in other studies that we do as a firm, the capital commitment does come up as a critical function provided by the brokerage community, but one that is typically concentrated with 50 to 75 of the largest institutions. It's not at all unusual to have the major brokers talking about loss ratios, which means what proportion of the gross commissions they receive from an institution do they expect to lose in providing capital commitment to help facilitate that institution's business.

So certainly the hard trades, the large trades requiring capital commitment, in our experience we've seen the majority of that actually going to the upstairs market to the brokers, and, again, it's transferring the risk of a market movement--those are the trades that do really require speed, transferring that risk really more to the upstairs broker. And, again, they are looking to undertake that, A, as the profit driver but also as an accommodation to their clients. Many of them are doing it in the belief that somehow if they take a loss, that client will make it up to them later with agency or commission business. Not an easy proposition but still one that they're doing.

We certainly did not look into the specifics around stressful periods such as 9/11. Again, I'd be interested in others' views on decimalization. I think that in the particular case of Instinet, if I'm correct, they actually found that to be quite a stressful event because the spreads in Nasdaq stocks evaporated. What's interesting to note is most of the major institutions are now actually paying commission on their Nasdaq business, which previously they made their money simply off the spread, ECNs had somewhat of an advantage in that the spreads at times were wide, and the ECNs acted on an agency basis in what was principally a market-maker-driven marketplace.

Well, with decimalization, the spreads absolutely collapsed. You had a number of major brokers saying we're going to stop making markets because there is no spread. So now you've got institutions actually paying commissions to make up for the lack of spread that exists as a result of decimalization.

Again, others may have some additional views on this.

MR. WALLISON: Steve, do you want to follow up on any of those?

MR. SACHS: Yes, I think that certainly the capital commitment and the fact that you left out indexed funds I think is an interesting point in the survey in that there might be a little bit of a misperception there in that the indexed world or the program trading world has changed substantially, I think, in the last four to five years. And the problem that you run into with program trading with listed securities, you don't have the same issues necessarily with depth of liquidity, but what you have is the electronic access, in that a lot of program trades are being done by automated execution strategies that are simply accessing ECNs or Superdot to the exchange. So you run into the issue of competition for your order and/or the (?) issue from the specialist in that, you know, the algorithms that people design for program trading are driven behind the idea of being able to obtain, you know, price or quantity discovery, and that if I show size, you know, I will receive reward for that, in that if I'm a liquidity provider I will be rewarded for that as opposed to a liquidity demander.

So I guess I would be interested in seeing the same survey of not only the traditional indexed world, but I think there are more than a few shops out there that are the more sort of quasi-quantitative active shops that are running depths that are trading large quantities of orders and program trading as well as single-stock trading.

MR. WALLISON: Steve, I'm interested in why the ECNs might offer less market impact. This is quite puzzling. They have--they appear to have less liquidity. The institutional investor is interested in depth at a price. What in your view is--other than anonymity, which clearly can have some effect on market impact, simply you know who is coming in.

MR. SACHS: Right.

MR. WALLISON: And so you're not sure whether that person has information that you don't have and all kinds of things like that can have market impact and there's a crowd and so forth. But what is your theory for why the ECNs score so well on market impact versus New York Stock Exchange specialists?

MR. SACHS: My theory regarding the survey on ECNs may be a little different in that, you know, again, I think they're very valuable tools. I use them regularly. But I think one of the problems in having conversations with other institutional traders is I think control of an order, which is what you have with an ECN, with effectively no information leakage, I think control of the order is sometimes confused with quality of execution.

For example, while I may have extremely low market impact using an ECN, there are a host of other things that I have to deal with, namely, fragmentation, which will lead to higher opportunity costs. So I think that you--I don't know if a bias in the answer is, you know, the correct phrase, but I believe that there's, you know, a bit of confusion in the institutional trading world, and that if I'm having low impact, you know, I'm automatically getting a better execution because not all institutions are necessarily measuring opportunity costs the way they should be in their execution formula. And it really depends, again, wholly depends on the style of the institution and that opportunity cost is extremely important to me but not as important as impact costs. You know, leaving no footprint in the market is my first goal.

Actively traded hedge funds may not care about impact. It's about opportunity costs. Statistical arb traders, you know, certainly don't care about impact in most cases. Their initial concerns are speed and, you know, opportunity costs. And if they want to capture the order as quickly as possible at, you know, a price or near a price they desire, it's what their model is generally driven from. It just varies greatly from institution to institution, but I think, again, the overriding theme here is that there's quite possibly a little confusion between quality and control in using ECNs.

MR. WALLISON: There seems to be some--when people talk about the specialist, there is some concern about the fact that the specialist is, in effect, a monopolist, that is to say, the specialist is right in the middle, everyone comes to him or her, there isn't any way to--there isn't any way to control what the specialist is doing, and that may be a factor that concerns institutional investors.

But there have been suggestions--and I think it was mentioned here--that maybe there should be more than one specialist, multiple specialists.

Now, doesn't that offer--if it were done through the New York Stock Exchange structure, wouldn't that raise the question of fragmentation as a solution to the specialist problem? Introducing that kind of competition, wouldn't that introduce fragmentation?

MR. SACHS: It would. You know, in my mind, that's very similar to, you know, what Nasdaq currently looks like today in that you've got multiple competing market makers in every security. There are certainly advantages of it. You get obviously price competition, which generally benefits, you know, not only retail investors but institutional investors as well. But, again, the by-product of that is fragmentation of the market, and it goes right back to the question of most institutional investors will tell you, if I have one central marketplace where I could trade, you know, all of my securities and have one big pool of liquidity, that's what I'm looking for.

You know, unfortunately, I don't know that we can get one big pool of liquidity and still have a competitive market-maker system. But competition among specialists is certainly an issue worth looking at. I can certainly tell you from my experience that there are particular specialists on particular stocks that I will not trade with. I will look for alternative execution venues. Generally speaking, it's either an ECN or a capital commitment trade from a broker simply to avoid the particular specialist in that security.

MR. WALLISON: Ken, what do you think of the idea of multiple specialists? Have you given much thought to that idea and how it would work and what effect it would have?

MR. LEHN: Well, again, I think the tradeoff would be the fact that the advantage--

MR. WALLISON: Can you move the microphone a bit closer to you?

MR. LEHN: Sure. The advantage of the one specialist is that all the orders and all the information sort of comes into one market maker, and there's a fellow at the University of Minnesota, Larry Benvineste (ph), who has argued that that actually enhances the efficiency of markets to have one person there seeing all the information.

But then, as you say, that obviously then confers a monopoly position on that one market maker, so you end up in the classic tradeoff as to whether or not the costs of market making in some sense are lower by having one, and the extent to which investors realize some of those cost savings versus, you know, hidden markups due to the fact that you have less competition for making market (?) .

The empirical evidence I think generally would suggest that it really doesn't matter. There is some work done in the options market where equity options trade in different venues, and I believe the American Stock Exchange traditionally has had one specialist and CBOE has had multiple market makers, if you will, within that market. And I know we have experts here who know more about this than I, but my recollection is that the transaction costs did not differ significantly across those markets.

MR. WALLISON: John, do you have any comment on that question from your survey or your other experience?

MR. COLON: I'm not sure that people focused in looking at that question specifically. It wasn't addressed as do you think that this would actively promote competition and had a disciplining mechanism. But, again, I think it's a little bit reading the tea leaf, but certainly from the open-ended comments that people were making and the difference they seemed to--and, again, this is an inference--between floor brokers and specialists and why they're less concerned about one versus the other tended to be a sense of is there a disciplining mechanism. So whether competition would be the right disciplining mechanism, but, you know, I think, again, you have the problem where individual traders may not feel that they're getting (?) enough or that it's detrimental enough to really be worth making a big fuss about. And so you potentially get a lot people standing on the sidelines and nobody really willing to charge forward.

It's interesting that only recently have you had a couple big institutions speak out loudly and publicly on the topic, and it's obviously not a new--

MR. WALLISON: Yeah.

MR. SACHS: I'm sorry. I personally think that's a very interesting observation in that we've had multiple periods over the last decade or so where institutions could really positively effect change in the marketplace and market structure, and really for the most part institutions chose to stay on the sideline until, as you said, just recently with the developments at the NYSE.

MR. WALLISON: Well, now we will turn to the experts. I hope all of you have been thinking of questions, things that are unclear or things you'd like to get before our panel.

What we will do is I'll ask you to identify yourself, if you have a question, and we'll go from there. Turen (?).

MR. REDDY: Turen Reddy with Wall Street Letter. Given the fact that this meeting is talking about best execution, I was at the Security Traders Association meeting last week, and I asked their president, Should the SEC come up with a definition of best execution? And he said, no, they shouldn't because that's just more meddling in our business and it's only going to further complicate things and that the market should figure this out.

I wonder if the panelists--you know, what their thoughts are. Should there be a clear definition of what best execution is so everybody's playing by the same rules?

MR. WALLISON: Okay, panel. Steve, why don't you start.

MR. SACHS: No, the SEC should not define best execution because the fact of the matter is--again, it goes back to what we talked about earlier. That really depend on the particular institution, the style or the manner in which they manage money, the clients that they're managing money for, whether it be, you know, public funds, private funds, limited partnerships, whatever it may be.

Personally, I'm in favor of--the AIMR guidelines that were set forth I think were very helpful to the institutional community, and I personally think that this is a topic or this is a particular part of the market structure that institutions need to define for themselves. And the fact of the matter is that a good solid process is a good beginning point, which is effectively what, you know, AIMR has stated in their policy, because at the end of the day, you know, the outcome is really going to vary greatly between the different types of institutions.

MR. WALLISON: Anyone else want to comment?

MR. COLON: I'd absolutely agree with that. I believe I mentioned AIMR guidelines as well. And, again, I'm not actively a trader, as Steve is, but I think very much those outline a process rather than trying to define an outcome. And within that process, as you read through those guidelines, you can take into account all sorts of things beyond simply price. There is opportunity cost. There is--what are the other things that that institution is accessing as part of its investment process? So not to raise a whole other frightening issue, but you've got the use of commissions to pay for research product, which some would argue is inherently in conflict with the notion of best executions. If I can do business on an electronic basis at two cents a share, why am I doing business through an upstairs broker at five cents a share?

Well, there are actually a host of reasons for it that are quite legitimate as part of the investment process. And, again, I think the AIMR guidelines emphasize the process rather than necessarily trying to define the specific outcome.

MR. LEHN: And I agree with the prior to comments.

MR. WALLISON: Other questions? Oh, come on. Mike?

MR. EISENBERG: Mike Eisenberg, Securities and Exchange Commission. These are my own views and not necessarily those--

[Laughter.]

MR. WALLISON: My own question and not necessarily...

MR. EISENBERG: Two things. First of all, I think the underlying premise that Peter laid out at the beginning that the commission is biased in favor of the central market, specifically the New York Stock Exchange, I think is probably an argument for another time, but I think that can be reasonably disputed in terms of what the commission has done, not just over the past several years but going all the way back to, in effect, breaking up the fixed commission rates, and there are a whole array of matters which the commission I think favored competition among the markets, certainly with the NASD and the ECNs. I think that's an argument for another time.

The other thing is that regulation tends to freeze outmoded structures that Peter started with, freeze outmoded structures like the fixed commission rate in which the commission took the initiative to break, fixed minimum commission rate, some of the other things that have gone on more recently in terms of changing what has gone on in the market. It seems that it sometimes tends for institutions that are operating in the market not to want to change things because they're very happy with the way things are and they don't want the government to go in and change things. So it works both ways.

My real question is with respect to the thing that was just brought up, and that is, you can go to an ECN, get it for two cents a share, one-and-a-half cents a share, and many pension funds and other advisors do that at no suffering to what they consider best execution, which includes a number of the things that you've talked about, opportunity and so on.

The question is in terms of your sample. One of the reasons that some mutual fund groups go to full service, quote-unquote, New York Stock Exchange is the opportunity to allocate three cents a share, or whatever it is, for distribution--they do it both outside of 12(b)(1) and inside of 12(b)(1)--the opportunity under 28(e) to allocate money for research, and there are other things that they allocate [inaudible]. They use various devices like step-outs and other types of trading methods, which, in effect, permit them to send executions at prices that are higher than best execution, which includes price discovery and taking capital into account and some of the other things.

Doesn't your sample, by using those funds that distribute through the brokers that do this, is that distortive of what your sample really is in terms of best execution? In other words, there are other things that they want to do which add to the price of execution which have nothing to do with price discovery and stability and the quickness of execution and the other things that we normally think of as best execution?

MR. COLON: Well, again, I think looking at things like AIMR guidelines and the SEC's own regulations, there obviously is the permission of quite a wide range of activities in terms of how an institution utilizes its order flow to obtain services, and that some of those are really pure execution services and others are a slightly more diverse array of services.

I think in many cases an institution's willingness to pay a larger commission doesn't mean that they stop paying attention to things like price discovery, market impact, et cetera. So that they may be willing to pay four cents a share to a, quote, full-service firm to also obtain that firm's research. But there will certainly be hell to pay if they feel that that firm then leaks information all over the marketplace that causes the prices to move and they get at the end of the day a crummy price for that agency execution.

So I don't think the fact that they're willing to pay an increased amount in some cases on a commission basis, which is part of the aggregate execution cost, necessarily takes their eye off the other aspects of execution.

Again, I've made a couple of allusions to the fact that Greenwich Associates does other studies within the equity marketplace, coming back to the notion of VWAP and execution measurement, et cetera. Among the very largest institutions, there appears to be a clear movement towards using order entry price or, in some cases, decision price as the manner in which they benchmark the effectiveness of the executions that they're receiving. It tends to be mid-size, smaller institutions that are less sophisticated that are perhaps sticking somewhat more with VWAP as the primary means by which they're benchmarking themselves.

So, again, they're very much paying attention to did the market move, particularly the larger institutions, between the time that they made the decision and the actual execution. So, again, I do think they, in fact, have their eye quite on the ball in terms of quality of execution, even if they are paying for other services in that process.

MR. : Even if these services go to the management company and not [inaudible].

MR. COLON: Yes.

MR. WALLISON: Other questions? Yes, sir?

MR. : (?) at the Investment Company Institute. I just wanted to make more of a comment than a question. I think our members, again, a lot of whom were surveyed in this--and actually the list of some of those who weren't surveyed, I've heard their opinions as well. I think I do agree with Steve on most of the things he brought up. Just a couple points.

I think it's important to note that the institutional investor community, mutual fund community, has not been silent, so to speak, on a lot of these issues. They haven't been speaking out in the press until very recently, but we've put in letters going back to March 2001 to Dick Grasso recommending changes to the New York Stock Exchange system, recommending automatic execution for large blocks, changes to institutional express and liquidity quote. Again, you know, we've put these in for a while now. They've been largely ignored by the New York Stock Exchange and by the specialists, but they've been out there. Again, they're coming more to light now with the specialist investigation and Dick Grasso's leaving the exchange.

But, you know, these are important issues to institutional investors. They've been out there. We've been pushing them. You know, I think it's important that the exchange starts to listen to some of these things.

MR. WALLISON: Any other questions?

[No response.]

MR. WALLISON: Well, let me try one more. Since we have a little bit of time, maybe I can stimulate some discussion.

What is to be done about the lower-cap stocks? I gather that there's a sense that you don't need a specialist for a security that has a lot of depth in liquidity. But for the mid- and perhaps lower-capitalization stocks on the New York Stock Exchange, would you continue to have a specialist there? And if so, is there enough compensation for a specialist to handle lower-capitalization stocks?

MR. SACHS: I think that's a good question. I think, in my opinion only, I think the specialists would probably respond to no, there isn't enough profitability in small-cap stocks in particular that they need the large-cap stocks with the volume and the opportunities to trade, you know, consistently throughout the day in order to make their operations profitable.

But what would--you know, just thinking out loud, what would be wrong with, again, a multiple specialist or a multiple competing market-maker scenario for those small-cap names and providing--you know, attempting to provide the order for the market stability in that manner?

MR. WALLISON: Any other thoughts? Ken? John?

MR. LEHN: Well, I think there's a very strong intellectual defense for specialists for low-cap stocks that are not particularly liquid. I think one other thing, too, that is worth pointing out is that often the issue of the specialist system and technology commingle that one could have a virtual marketplace with specialists, and I think something of that order, perhaps more technologically sophisticated but still having the protocol of the specialist for less liquid stocks would be a very viable outcome.

MR. WALLISON: John, any point?

MR. COLON: Again, I'd agree I think to a large extent with those comments. I think that one of the interesting roles that brokers and market makers have played in the process is that in some cases institutions will seem to recognize that it's difficult for somebody to make money in trading small-cap stock. There just isn't enough volume. Those larger institutions at times will actually, quote, pay for executions in smaller-cap stocks by using some of the larger-cap order flow.

Now, again, it's not necessarily contrary to the demand to achieve best execution, but if there are three or four venues that they feel offer very similar quality of execution, what may act as the tie breaker? That at times can be one of the tie breakers? We've been willing to be there from them in some of the smaller, less liquid names.

MR. WALLISON: Andy? First, wait for the mike please.

MR. KLEIN: I'm Andrew Klein at Shiff, Harden and Waits (ph). Steve, you mentioned earlier in your comments that there were some specialists, some New York Stock Exchange, with whom you would not deal or you would attempt to go somewhere else just to avoid them. I wonder, are you able to disclose and describe what it is that those specialists did compared to the ones that you liked that elicited that response from you?

MR. SACHS: I won't disclose who they are. You know, it's a relatively small list of specialists on the NYSE these days, anyway, so I think that you can probably narrow it down. But the general activity is really what I have found, I think, to be a consensus among the traders in the business that I talk to on a regular basis.

Certainly, you know, everybody's complaint in pin-(?) but, you know, there are ways around that, and I think, you know, the bulk of the trading community has, for the most part, figured out ways to circumvent that.

Really, what I have found to be the main issue with the few particular stocks I refuse to deal in are a couple of things. One is what the definition of a fair and orderly market is and effectively the negative obligation or affirmative obligation rule, which has been obviously in the press as of late. And then the--a lot of people will boil it down to style differences between the individual specialists, which is also another issue inherent in the current market structure and that one particular specialist acts in one way to maintain an orderly market in a stock, another specialist acts in a much different way. There are really--if you look at most of the specialist firms, there is not an internal policy, if you will, within those firms of how a specialist is to maintain his orderly market. Generally speaking, P&L is what's driving their activity in the selection of the specialist for those particular stocks.

The biggest issue that we have found, generally speaking, is around market open and market closes, and then activity around exceptionally large trades that are attempting to be put on in the marketplace. And what we're finding most often is the inability or the unwillingness of the specialist to perform one of his main functions, which is, in fact, if there is an order imbalance at any given point during the day, the specialist has the capability to contact any of the participants from that day's trading activity or even previous days' trading activity to solicit, if you will, demand or supply, whichever side the imbalance may be, for that particular security. We find that that happens very rarely, and we are talking in the most liquid of securities.

And we have a situation in the market now where we've gone to decimal pricing and we have a great number of stocks that are trading below $20 a share, below $15, below $10 a share that are very large liquid stocks that are the cornerstone of a lot of particular portfolios and indexes, that you are having 100, 150 basis point moves in order to transact what we would consider normal volume. Thirty, forty, fifty, sixty thousand share blocks are moving stocks like Motorola, you know, 100, 150 basis points, either around opening close or just simply around large transactions. You know, we don't find that to be a very efficient system either for the specialists in particular or even for us. And given today's technology, there's got to be a way to communicate that information to the marketplace better, if you will.

MR. KLEIN: And do you find that by going to other forums, to ECNs or ATSs or regional markets, that you're finding better satisfaction there in those same stocks?

MR. SACHS: Generally speaking, we will either--we will first search an ECN or an ATS, but more often than not, we're simply using a capital commitment trade. And, fortunately, with the repeal of--I think it's 319, don't quote me--the trade-away rule and that all, you know, listed securities can now be traded in the third market as opposed to only at the post, you know, we find that a very easy way around that particular situation.

The question is: It's an easy fix, but should I have to do that? And does that benefit the rest of the market? Because that liquidity or the opportunity for price improvement is now not going to exist at the post. It's going to exist somewhere else.

MR. KLEIN: You referred in your response to one other factor, and that was pin-(?) by the specialist. How can you tell when it is the specialist, in fact, I assume dealing for his own account, who's doing that as opposed to someone else competing with you from another angle for the same trade?

MR. SACHS: That's a good question, actually. On every transaction, you cannot tell. We have actually made an effort to--for example, we will send an order via Dot to the floor or give an order to a broker. We will actually, you know, then send another broker--effectively, we generally use two dollar brokers, independent floor brokers, without any proprietary trading capability, to the post to examine the book for us. And more often than not, we find that it is, in fact, a specialist bid that is in place of us.

You know, to be fair, quite often we find it is a stat arb program, if you will, that is keying off of liquidity that we are trying to demand or provide. But, you know, at the same time there's just as many specialists that are putting the bid or the offer in front of us. And I think if you look at the data that was gathered just after we went full implementation into decimals, I believe specialist trading activity increased somewhere on the order of 40 percent in the months following the move to decimal activity--or to decimal pricing.

MR. KLEIN: That latter one couldn't be surprising. It became a whole lot less risky for them to trade ahead of other people.

MR. SACHS: Absolutely.

MR. KLEIN: I think they said in congressional testimony, actually, before that occurred, that that's exactly what they would do.

MR. WALLISON: Any other questions?

[No response.]

MR. WALLISON: Well, if not, I want to thank all of you for coming. I want to thank our panel for an excellent set of presentations, and we hope to see you at our next conference.

Thank you.

[Applause.]