May 31, 2000

Mr. Jonathan G. Katz
Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, DC 20549

Re: Market Fragmentation Concept Release

Release No. 42450; File No. SR-NYSE-99-48

Dear Mr. Katz:

The New York Stock Exchange, Inc. (the "NYSE" or the "Exchange") appreciates the opportunity to comment on the market fragmentation issues brought into focus by the Securities and Exchange Commission (the "Commission") in its publication of the NYSE's filing to rescind Exchange Rule 390, the "Market Responsibility Rule" (the "Filing").1

In the Release, the Commission asks for public comment on the effect of the repeal of Rule 390 on fragmentation of the listed securities market in general, and how the practices of internalization and payment for order flow relate to market fragmentation and impact the best execution of investor orders.2 The Commission also asks for comment on six options for addressing fragmentation:

The Release also raises questions about the Intermarket Trading System ("ITS").

The Special Committee on Market Structure, Governance and Ownership

On October 7, 1999, the Exchange's Board of Directors formed the Special Committee on Market Structure, Governance and Ownership (the "Public Directors' Committee"). The Public Directors' Committee is composed exclusively of public directors with no member organization affiliation who qualify for election as CEOs of listed companies and institutional investors, or whose public sector or public service experience qualifies them as representatives of individual investors.

Since October 1999, the Public Directors' Committee has received oral and written presentations on market structure issues from investors, listed companies, member organizations, NYSE specialists and floor brokers, and industry experts. In addition, the Public Directors' Committee reviewed publications containing the views of a wide array of market participants and commentators, the legislative history of relevant parts of the securities laws, and Commission rules, releases and reports. The Public Directors' Committee also carefully considered the Release, considering each of the possible options put forth by the Commission for addressing fragmentation.

The Public Directors' Committee unanimously approved and, on April 6, 2000, presented to the Exchange's Board of Directors its Market Structure Report. The Board voted unanimously to endorse the overall principles of the Market Structure Report, to embrace the Report's recommendations, and to accept the Report overall. The Market Structure Report serves as a substantive part of the Exchange's response to the Release and it is attached hereto and incorporated by reference.

During the market structure hearings conducted by our Public Directors, market participants repeatedly noted that "one person's fragmentation is another person's competition." Those participants recognized that fragmentation and competition are more than just two sides of the same coin. They create a dilemma: Promote intermarket competition and fragmentation increases; address fragmentation and orders concentrate in a single market.

Since the enactment of the Securities Acts Amendments of 1975, the listed market has tried to solve this dilemma with the linking of transparency and market linkage in the hope of maximizing both order interaction and intermarket competition. But we have never managed a complete solution because intertwined in the goals of order interaction and intermarket competition are the conflicts of interests arising from internalization and payment for orderflow.

In the Release, the Commission renews the search for an answer, presenting six options for discussion. The Commission's request for comment issued as our Public Directors completed nine days of hearings on these and related market structure issues. In its Market Structure Report, our Public Directors' Committee charted a course for addressing these issues that takes into account how changing technology offers prospects for resolving the fragmentation/ competition dilemma.

In this letter, we discuss in detail the implications of our Public Directors' findings and recommendations for the questions posed in the Release. This letter seeks to explain the course that the Market Structure Report lays out by posing and answering the following three questions implicit in the Release.

How should the Self-Regulatory Organizations ("SROs") and the broker-dealer community:

For the listed market, we believe that the Commission can answer all three questions by taking, in conjunction with the SRO and broker-dealer community, a handful of investor protection, disclosure and leadership steps that will unleash the market forces that will hasten consumer empowerment. And in consumer empowerment, the dilemma would be resolved.

Disclosure of Market Practices

The Commission asks in the Release whether it should require greater disclosure by market centers and broker-dealers concerning their trade executions and order routing. The Exchange supports the concept of more effective disclosure to investors of trade execution and order routing practices by market centers (including broker-dealers' paying for order flow), and by broker-dealers (and market-maker affiliates), in their fiduciary capacity to their customers. The Exchange has also championed the idea of market participants developing best execution measurement standards so that broker-dealers and their customers would have comparable best execution data for use in making order routing decisions.3

The Commission suggests an array of items that markets would need to disclose on a uniform basis, including the ratio of limit orders to market orders, effective spreads for market orders in different types of securities, percentage of market orders that receive price improvement, speed in displaying limit orders and average time to fill different types of limit orders. Similarly, broker-dealers should be required to provide disclosure concerning, for example, the proportion and types of orders that are routed to different market centers, their arrangements with market centers for routing customer orders, and the results they have obtained through these arrangements.

The Market Structure Report of the Public Directors' Committee states:

Until internalization, preferencing and payment-for-order flow practices are reformed by regulatory action, or are made less prevalent as a consequence of decimalization, significant agency costs will continue to be borne by uneducated investors.

The Public Directors' Committee found that, while institutional investors understand the issues underlying quality executions and tend to monitor their brokers carefully, retail investors do not. For these reasons, the Exchange believes that informing investors about market structure and order execution issues is vitally important. The Exchange already provides price improvement data to member firms in the aggregate on a monthly basis, and on an order-by-order basis under the NYSE PRIME program.4

The Public Directors' Committee also found that retail investors generally do not understand how their orders are executed. So that investors may make more knowledgeable choices about how they want their orders to be executed, the Public Directors' Committee recommended that the Exchange develop a communications plan to inform investors about order-execution and market structure issues. The Board strongly endorsed this recommendation.

The Exchange is mindful, however, that education is a long-term process and is not a panacea for the present problems engendered by and associated with economic inducements for order flow. Similarly, disclosure, while appropriate, cannot by itself provide a complete solution. All too frequently, disclosure translates into "boiler plate" text similar to the fine print on a car rental agreement. It is there for the consumer to read, and the consumer should read it, but the agreement is hurriedly signed and often not read or understood.

While disclosure and education are important, the root causes of the problem -- economic inducements -- must also be addressed. Therefore, we believe that disclosure can and should be significantly enhanced; however, it is not by itself a viable alternative to banning internalization and payment for order flow or, in the alternative, instituting a Price Improvement Rule. The bottom line is that investors are the principals, and their brokers owe them the fiduciary duty associated with that agency relationship.

Fragmentation, Economic Inducements and Price Improvement Rule

The Exchange agrees with the Commission's statement in the Release that the elimination of Rule 390 raises the potential for increased fragmentation of the trading interest in exchange-listed equities. The Release defines "fragmentation" as "... the trading of orders [in the same security] in multiple locations without interaction among those orders."5

Congress determined in 1975 that market competition should be a decisive factor in allocating order flow. Congress resolved that it is in the public interest to assure fair competition among brokers and dealers, among markets and between exchange markets and the over-the-counter market, which would lead to greater technological innovation and reduced costs to investors. Since 1975, the markets and the Commission have continually dealt with the appropriate balance between the costs of order fragmentation and the benefits of market competition.

Rule 390, subject to limited exceptions, required NYSE members to execute principal trades (and crossed agency orders) on an exchange -- an anti-internalization, customer protection rule. As noted in the Market Structure Report, the Rule maximized the opportunity for investors' orders to interact with one another in agency-auction markets for potential price improvement and without unnecessary dealer interpositioning. Critics of Rule 390 viewed the rule as anti-competitive, even though Exchange members could have traded or competed as specialists on any other national securities exchange or foreign stock exchange, and outside of Exchange trading hours, over-the-counter in a foreign country.6

The Exchange believes that the Commission has correctly focused the broader discussion of the cause of fragmentation on the "direct and indirect economic inducements to brokers in return for the broker agreeing to route all or part of its order flow". In addition to internalization, the Commission also identifies the practice of payment for order flow as an economic inducement leading to fragmentation.

The potential for increased fragmentation with the elimination of Rule 390 has two dimensions. The first is the possibility of redistribution of order flow in listed stocks from the NYSE auction to competing dealer markets, which if it goes too far, would impair the price discovery mechanism. The second dimension is internalization, whether it be a broker's own order flow or order flow a dealer has paid for. The issue here is the bid-ask spread, and whether it belongs to public customers or to broker-dealers.

With respect to impairing price discovery, the NYSE is still the locus of price discovery for its listed stocks.7 In the Release, the Commission notes that the degree of fragmentation in NYSE-listed stocks is "fairly low", stating that in September 1999, the NYSE printed 74.4% of trades and 83.9% of share volume in its listed stocks. In fact, after an initial modest drop following the adoption of Rule 19c-3 and the inclusion in 1982 of the third market in ITS, the NYSE's market share of trade and share volume has remained fairly constant over almost 20 years. We also note that the NYSE's share of trading volume in Rule 19c-3 stocks, those in which member firms have been able to internalize since 1980, has remained fairly stable at 80-82%, only slightly below the Exchange's share of trading in volume Rule 390 stocks. Furthermore, the NYSE's share of trades 5,000 shares or greater is in the 90% range. In contrast, our share of trading volume in 100 to 4,099 share trades is in the low to mid-70% range. This data suggests that members have, for some time, been using regional exchanges to internalize retail investor order flow in Rule 390 stocks.

Using the Commission's benchmark measure of fragmentation -- market share -- we agree with the Commission's assessment of the NYSE-listed market. Going further, the Exchange notes that the degree of fragmentation has remained fairly stable despite the introduction of significant initiatives approved by the Commission, designed to further competition with the NYSE, each of which also had the potential to fragment the listed market.8 In each instance, the Exchange voiced its concerns on the potential negative impact that further fragmentation would have on the execution of orders -- particularly those orders of individual investors. At each juncture, the Commission concluded that these initiatives would further competition, which in turn would inure to the benefit of the customer. However, as the statistics above demonstrate, the market fragmentation impact has been most pronounced in the order flow of retail investors -- that class needing the most protection. The Exchange believes that not enough has been done to afford that protection and supports the Commission's decision to re-evaluate those initiatives with respect to internalization and payment for order flow.

As noted above, Rule 390 required members to execute principal trades with customer orders on an exchange, but permitted members to route their customer orders on an agency basis to any market in the world. In one sense, the Rule had no impact on deterring the practices of internalization and payment for order flow. Member firms desiring to internalize -- whether for their own order flow or order flow purchased from others -- have been able, since 1980, to internalize in Rule 19c-3 securities in their offices. For Rule 390 securities, they widely availed themselves of the ability to register, or have affiliates register, as specialists on regional stock exchanges where, as a practical matter, order flow may also be internalized.9 Similarly, the Rule imposed no restriction on members selling their customer order flow to over-the-counter dealers or to specialists on regional exchanges. However, other member firms have reconciled their fiduciary duty and best execution obligations by determining not to engage in the practices of internalizing and payment for order flow in the spirit of employing the customer protection principles that had been embodied in Rule 390.

Because the listed market has long been somewhat fragmented by these practices of member and non-member firms, it is difficult to precisely gauge to what extent the repeal of Rule 390 will further fragment the market. However, some additional degree of fragmentation can be expected to occur in the absence of enhanced best execution standards.

The prospect for growth of fragmentation would then seem to depend on several factors: additional member firms engaging in the practice of internalizing -- with large, integrated firms having the most potential for impact; additional dealers engaging in the practice of payment for order flow; and, increases in the size of customers' orders that firms are willing to internalize and dealers are willing to pay for. Access to the NYSE liquidity pool, of course, supports the ability of those engaging in these practices to minimize the risk of doing so; free access through ITS is presumably a benefit from their point of view, although we do not know enough about their businesses to understand if this subsidy is essential to them.

As the Public Directors' Committee notes, if all brokers-dealers internalized or sold their order flow, the Exchange agency auction model would largely disappear, in favor of a fragmented dealer and ECN order match market similar to the over-the-counter market -- reducing overall market transparency, impairing price discovery and thus, we believe, harming investors. By way of example, in the five NYSE-listed stocks with the greatest volume in March 2000, 78% was executed on the Exchange, with six other ITS Participants executing 22%. In comparison, the NASD's Nasdaq Trader web site shows that the top five over-the-counter stocks had an average of 98 market-makers each, with the top ten market-makers accounting for about 50% of the reported share volume. The largest single market-maker in each stock accounts for 7.5% to 10.6% of share volume. The very nature of the competing dealer system in the over-the-counter market means that it is a fragmented market.

The other dimension of internalization is its impact on customer orders. With the recision of Rule 390 (and its analogs on the other exchanges), there is no specific rule to address this aspect of the conflict between a broker-dealer's fiduciary duty to his customer, and his dealings with that customer as principal to his own economic advantage. Accordingly, the Filing asked the Commission to adopt an industry-wide Price Improvement Rule to limit internalization to those situations where investors are given improved prices.10 Specifically, the Commission should prohibit broker-dealers from trading as principal against their customers' orders, or purchased orders, unless they provide a price to the orders that is better than the NBBO against which the order would otherwise be executed.

The NYSE has consistently urged the Commission to take regulatory action to address the significant investor protection concerns raised by current economic inducement arrangements.11 The Commission states in the Release that economic inducements for dealing against captive order flow, such as internalization and payment for order flow, pose customer protection concerns. Indeed, the Public Directors' Committee concluded that the Exchange should "vigorously oppose" these practices, and the Exchange actually would prefer that the Commission ban such inducements. However, if the Commission does not ban them, the Exchange believes that its proposed Price Improvement Rule, while admittedly a weak substitute for a ban, would help ameliorate some of the negative effects of the inducements on customers. Customer orders that are deprived of the ability to interact with other orders will at least enjoy some degree of price improvement. The Exchange considers it vital that, in a regulated industry committed to customer protection, those charged with the fiduciary obligation of best execution for their customers should put those customers first.

Order Exposure for Price Improvement

The Commission asks for comment on a possible requirement that all market centers expose their market orders "in an acceptable way to price competition". The Commission offers two examples of acceptable exposure: First, an order could be exposed in a system that provides price improvement to a specified percentage of similar orders over a specified period of time,12 and second, a market-maker, before executing an order as principal in a security whose quoted spread is greater than one minimum variation, could publish for a specified length of time a bid or offer that is one minimum variation better than the NBBO. The Release notes that the Commission proposed an order exposure rule such as that embodied in the second example in 1995 when it proposed the Order Handling Rules.13 The proposed rule contained a non-exclusive safe harbor for best execution, which would have required that an order be stopped at the proposed execution price and quoted at a better price for 30 seconds.

The Exchange believes that each customer at-the-market order should be afforded the real opportunity for price improvement.14 This happens in the agency-auction market sponsored by the Exchange. The auction market, of course, does not guarantee that each order will be price-improved. But it does guarantee that each order is exposed to that potential.

Some markets claiming a high incidence of price improvement employ artificial methods to generate this price improvement, i.e., improvement not the result of interaction among orders. The Commission recognized this when it stated in the release adopting the Order Handling Rules that:

[S]ome markets or market makers may continue to offer price improvement opportunities, based upon internal order flow or execution algorithms.

It is our understanding that such markets evaluate their dealers' performance on methods they employ to achieve, in percentage terms, goals established by the market. However, in these schemes, every customer order is not afforded an equal opportunity to be price-improved. Only those in the order flow sequence governed by the algorithm are eligible for price improvement. Those not at that sequence spot are not afforded any opportunity for price improvement.

Markets and market-makers employing these artificial methods are principally engaged in internalization and payment for order flow. To the extent the executions are against dealers trading with captive customer orders subject to such inducements, the only practical way to assure that each order is afforded an equal opportunity for price improvement is to guarantee that each order is price-improved.

In the Exchange's January 15, 1996 comment letter on the Commission's 1995 proposed Order Exposure Rule, we noted that the Commission's proposal exempted blocks. We expressed concern about the Commission's sanctioning of a hidden market for block orders. The Exchange also stated that the safe harbor could result in the dissemination of a significant number of quotes, often of short duration. The NYSE urged the Commission to evaluate the effect that this rule would have on the market before finalization. In addition, the Exchange commented that the proposed 30-second exposure might not be a sufficient time frame to provide a realistic opportunity for price improvement. The Exchange also warned of the difficulty of providing price improvement in markets that generally have limited interaction between customer orders, and that a rule which nominally provides opportunities for price improvement could mislead investors into thinking that the execution they receive in such markets is similar to the executions that are available in the agency auction market.

Five years later, our views on mandated exposure remain unchanged. We believe that the significant increase in trade and quote volume since 1995 renders the "stop and expose" option even more problematic. If, however, the Commission decides to adopt an Order Exposure Rule, the Commission should rethink the minimum period for quotation exposure to provide a real opportunity for execution at an improved price.

National Linkage System/Consolidated Limit Order Book

The Commission asks for comment on the creation of a national linkage system, wherein all displayed orders and quotations of all market centers would be displayed and executed automatically in strict price-time priority. The proposal appears to be largely modeled on the structure discussed in the "White Paper", published on the Wall Street Journal Interactive Web Site, that was reportedly submitted on a confidential basis to the Commission on behalf of several large NYSE member firms.15

The essence of consolidated limit order book ("CLOB") proposals is that all registered broker-dealers would route all their customer limit orders in listed securities to a new centralized market system. Market participants would be permitted to access and execute against orders in the CLOB. The system would operate on absolute price-time priority, accompanied by a rule that would prohibit executions on the exchanges and in the third market without first satisfying any identically or better-priced limit orders residing in the CLOB. A CLOB would be governed by a new utility jointly operated by market participants under Commission oversight.

The Market Structure Report carefully considered various proposals for a CLOB, including the proposals in the White Paper, and came to the following conclusions: a CLOB would not reduce fragmentation, but in fact would have the opposite effect; a CLOB would bifurcate the market between large and small orders since the CLOB, by its very nature, would not attract significant institutional order flow,16 and its proponents seek a block exemption;17 smaller limit orders would be executed in one market at one set of prices while large institutional orders would trade in a parallel market at different prices, causing serious disruption of the price discovery mechanism; a CLOB would eliminate the specialist's affirmative obligation, resulting in wider spreads; and, moreover, by eliminating quote matching, a CLOB would dramatically impair intermarket competition and innovation, counter to the express goals of the 1975 Securities Acts Amendments. 18

The Exchange has proposed making the specialists' books of limit orders visible to its members and the broader public. Chairman Levitt has called on all markets to move toward open books as a voluntary, private sector initiative. Vendors then could consolidate the data and package it in a form useful to their customers.

The Division of Market Regulation has been given the responsibility to take the lead by hosting public roundtables to facilitate dialogue on the issue, and an initial roundtable was held on May 4, 2000. The Exchange believes that this is an excellent approach, and welcomes the opportunity to continue to participate in such a dialogue. As discussed below, we also believe it a useful approach for dealing in part with price-time priority issues the Commission raises.

CLOB-Derived Proposals

The Commission raises a series of questions regarding the establishment of intermarket time priority for the first limit order or quotation that improves the NBBO and the prohibition of dealers' engaging in "quote matching", that is, dealers stepping up to trade at the same price as a previously displayed investor limit order quoted in another market.19 The Release notes that:

The current market structure allows price-matching rather than requiring that orders be routed to the market center that is displaying the best price, thereby isolating the orders of different market centers. Moreover, there is no intermarket time priority - the market center that was first to display the best price will not necessarily receive any order flow. Thus, the market participant (whether investor or dealer) who publicly displays an order or quotation at a better price than anyone else is offering is not entitled to any assurance that the order or quotation will interact with the next trading interest on the other side of the market.

This and prior Commissions have long recognized that best execution obligations require a broker to quote match or ship an order to the market with the best price. The regional exchanges and the third market employ, among other practices, quote matching as a competitive tool to attract and execute order flow. Exchange specialists also quote match.

The Exchange notes that the proposals are similar to those put forth in the late 1970's in connection with earlier consideration of a CLOB. We have been quite clear that we believe that a CLOB is not workable.20

The Exchange believes, however, that an excellent way to address this issue is to focus on the broker-dealers' responsibilities to obtain best execution for their customers' orders, insisting that they utilize technology to develop search engines that can route on an order-by-order basis to the best market.21 This is not to say that the Exchange believes that simple routing to the best and priority quote should be mandated. In this regard, we agree with the Commission's statement in the Release that:

Price is not the sole factor that brokers can consider in fulfilling their duty of best execution with respect to customer orders.

Rather, it seems logical that, while brokers would look principally at price, they would also consider other factors such as a market center's history of affording price improvement, execution rates, liquidity, speed, cost, certainty, size, etc. in making decisions on how search engines should seek out the market to which an order would be routed.

In our discussion below on ITS, we set forth some relevant data on the System's operation, which supports the notion that best execution entails more than routing to the best displayed price.22 We note that the two ITS Participant Markets that provide automatic execution against their quotations, the Cincinnati Stock Exchange and the NASD's ITS/CAES, have ITS execution rates of 59% and 71%, respectively. This data suggests that, if neutral order routing based upon price alone became the NMS norm, there would be an unacceptably high rate of orders "missing the market". This data also supports the premise that quote-matching, rather than routing orders to a CLOB for execution on a strict price-time priority basis, constitutes an element of healthy competition.

We believe that the Commission has already begun to address this issue in an appropriate way in its pronouncements on the factors broker-dealers must consider in fulfilling their best execution obligations. In its release adopting the Order Handling Rules, the Commission stated that:

Similarly, in evaluating its best execution procedures for handling limit orders, the broker-dealer must take into account any material differences in execution quality (e.g., the likelihood of execution) among the various markets or market centers to which limit orders maybe routed.23

The Commission reiterates this best execution principle in the Release.

NYSE quotations are the NBBO well over 90% of time. It is for this reason that many have observed that it would be in the Exchange's competitive interest to strongly support the underlying price-time priority concept in the Commission's proposals. However, the Exchange believes that quote matching is a legitimate form of competition, and that the Commission should not act to inhibit it or to mandate to which market an order must be routed.

Another troubling aspect of the Commission's constructs is the extent to which it mandates the use of automatic execution. If automatic execution is a useful competitive tool that appears to meet at least some customers needs, then markets will offer it. Even today, however, markets that offer automatic execution have size limits and other restrictions on the type of orders that are eligible. Each market should be able to design its competitive market structure in response to market forces, subject to Commission oversight, but free from Commission mandates geared toward homogeneous markets.

Clearly, the Exchange strongly believes that competition should drive the markets, and that the responsibility for best execution should reside with broker-dealers. When a customer or a broker-dealer makes a choice to route an order to a particular market, that choice should be the result of a cost/benefit analysis. Whenever there are competing services, there are always both costs and benefits to the choice that is made. Implicit in the Commission's best execution pronouncements, which deal principally with bulk order routing decisions of broker-dealers, is the fact that every order will not always achieve best price execution. Similarly, the choice of the market to which to route a limit order even when the limit order price establishes a new NBBO will not always result in that order being the first to be executed. The relative costs and benefits of routing to a particular market will drive the routing decision, but no decision will necessarily be cost-free.

The options proposed by the Commission appear to be designed to elevate intermarket quote competition above all other forms of competition. However, 25 years of experience has demonstrated that market forces have rejected that priority. As an historical matter, a primary consideration by the Commission in adopting the Firm Quote Rule in 1978 was to promote intermarket quote competition -- that is, each market competing for order flow via its quotes, with the mandate that market-makers in each market publish continuous two-sided quotations in the stocks in which they were registered.24 In 1982, the Commission amended the Firm Quote Rule to permit market-makers on the regional exchanges and in the third market to disseminate quotes on a voluntary, rather than a mandatory basis.25 The Commission recognized that:

. . .the mandatory dissemination requirement has not achieved its objective of fostering increased competition for order flow.... While certain regional exchange specialists do actively compete for order flow by disseminating competitive quotations in a limited number of actively traded stocks, regional exchanges (and their specialists) by and large prefer to continue to use other methods to compete with the primary markets ... for order flow.26

For the most part, non-primary market-makers opt to meet their quote obligations by disseminating nominal quotations that track outside the primary markets' quotations, and then by quote matching the primary market when an order is received. The Commission, rather than relieving these market-makers of the mandatory quote obligation, could have promoted quote competition by requiring these dealers to publish to the NMS their real quotations -- that is, the price and size at which they hold themselves out to their private customer base as willing, indeed obligated, to automatically execute market orders. We appreciate though that the Commission saw the folly in the idea of trying to force market-makers to publicly quote at prices and sizes not of their choosing.

The Exchange believes in quote competition, whether that be by orders or market-makers establishing the best quote or by quote matching. The Exchange also believes that enhancement of quote competition will naturally result as broker-dealers employ current technology to develop search engines to route order flow to the best market. More importantly, the Exchange believes that quote competition, indeed all competition, should evolve naturally from market forces rather than being designed by the Commission. The Commission has been most successful in the past when it focused on assuring adequate disclosure and the best execution obligations of broker-dealers, and we believe it should continue to address the issue in that way.


As our Public Directors pointed out, if the Commission is going to let the practices of internalization and payment for order flow continue, then price improvement becomes imperative. However, adopting a price improvement rule provides only the bare minimum protection to the retail customer whose order is being internalized or sold. It does not assure that the customer has gotten the best possible price or has otherwise received best execution. Nor does it deal with the displacement of the order of the customer which would have constituted the other side of the trade had a dealer not intervened.27

To lay the regulatory basis in the listed market for customer protection and for fair competition for retail orders, we believe the Commission must weave the price improvement rule into a broader set of broker and SRO responsibilities. We recommend that the Commission:

We believe that, if the Commission defines best execution to include price improvement and meaningful order exposure, works with the SROs and broker-dealer community to enhance the dissemination of execution quality information, and augments enforcement as to these two areas and the order handling rules, market forces are already in play in the listed market that will take care of limit order display and accessibility. At the same time, the Commission will have laid the foundation for the ultimate resolution of the conflicts of interest in this arena: the empowerment of the consumer to make his or her own order routing decisions.

Democratization of information through technology is already leading to the empowerment of this class. One day, perhaps in this decade, technology and competition will place the order routing decision in the hands of consumers, allowing them to make informed order routing decisions that include weighing the trade-off between price improvement, time, predictability and other attributes of execution quality. They will make those decisions free of the conflicts of interests that surround the making of those decisions by their agent fiduciaries who are engaged in internalization or payment for order flow.

As it has already begun to do in regard to encouraging limit order transparency, the Commission can hasten consumer empowerment by challenging broker-dealers to develop business models around such empowerment. For our part, we will heed our Public Directors' call to contribute to that empowerment through investor education. Joined with the steps we have outlined above regarding better definition of best execution, better disclosure, and better enforcement of best execution, disclosure and order handling responsibilities, the Commission will have done the necessary.

Competition will do the rest.

Intermarket Trading System

The Release states that:

The ITS linkage has weaknesses that must be addressed, including restricted ECN access and slow and inefficient execution procedures.

In the White Paper, as well as in recently well publicized Congressional testimony, ITS has been criticized as archaic for not staying abreast of current technology and for its governance procedures.

The real issue is not whether ITS has kept pace with technology, which it certainly has; rather, it is whether ITS remains necessary. ITS operates on the same type of technology platform and operating system as the other NMS systems for listed securities -- the Consolidated Tape System ("CTS") and Consolidated Quotation System -- as well as the Exchange's own order processing systems. These are all state of the art systems with unparalleled reliability and capacity, and are all decimal-ready. Some may want to redesign the functionality of ITS into a system for purposes other than those for which it is designed; that, however, is a different discussion.

It is unclear to the Exchange why the Release characterizes ECN access to ITS as "restricted" and a "weakness" of ITS. All ECNs are, by Commission regulation, registered broker-dealers, and they are NASD members. Broker-dealers in all markets access ITS via the self regulatory organizations ("SROs") of which they are members.28 The Commission recently approved an NASD rule change to provide ECNs with the same access to ITS as enjoyed by NASD market-makers.

Similarly, the ITS execution procedures utilized by the ITS Participant markets are generally the same procedures employed in the execution of their own order flow against their own bids and offers. To the extent that a market's execution procedures are weak, the Participant's constituencies will competitively force resolution of those weaknesses. However, we are not aware that the Commission has previously criticized the execution procedures of the exchange markets generally, or specifically, the Exchange's procedures. The Exchange's procedures for executing incoming ITS commitments and executing SuperDOT market orders are relatively the same, as both are processed via the specialists' Electronic Book workstations.29

The Exchange's market structure combines a human interface with the most sophisticated technology available, which we employ equally as to both ITS and SuperDOT. As we have improved our speed and efficiency for handling SuperDOT orders, ITS has improved.

Some suggest that ITS commitments should receive either automatic or default execution. We do not agree. We believe that it is not appropriate to afford to the ITS commitments of our competitors an execution priority higher than the SuperDOT and other order flow of our members.

The Commission is aware of the Exchange's position on the governance of ITS.30

The principal concern, among others, that the Exchange has always had with altering the unanimous vote required to change the ITS Plan is our ability to protect how non-members can access the Exchange's market. Changing the unanimous voting requirement would effectively delegate to our competitors the ability to make such determinations.31

The Release correctly notes that ITS does not guarantee time priority. ITS was never designed or intended for that purpose. As discussed below, ITS was designed to protect customer limit orders quoted in one market from a transaction occurring at a less favorable price in another market, and to assure that customer market orders sent to one market can reach a better quote in another market. ITS was created at a time when the Commission's goal was that the markets would employ vigorous quote competition to attract business. However, that goal has never been achieved, and regional markets and third market dealers have devised other methods to compete.

The Release notes that ITS handles a relatively small portion of trading in listed stocks.32 Given the manner in which the regional and third markets have chosen to compete, it should not be surprising that ITS volume is de minimis. Their dealers use ITS mainly to lay off principal positions on the primary markets.33 These markets have generally competed by providing to small sized public investor orders fast automatic execution guarantees at the NBBO. They fulfill their guarantees principally by quote matching the primary markets.34 Furthermore, such dealers, either through their own NYSE memberships or correspondent relationships with NYSE members, all have access to so-called "DOT machines" -- terminals linked to the NYSE members' order routing systems that are, in turn, linked to SuperDOT. We have been told that they use the DOT machines far more extensively than ITS to reach the NYSE floor. They also use SuperDOT to place so-called "marker orders" on the Exchange as a means of fulfilling their self-designed obligations to guarantee customer limit orders.35

ITS was created in 1978 by the exchange markets in direct response to a Commission proposal to develop a CLOB for listed securities. The CLOB proposal then was similar in all significant respects to recent proposals to reconfigure ITS into a CLOB. For the reasons addressed in more detail above, the Exchange, other national securities exchanges and the industry opposed the CLOB proposal and embraced ITS as an alternative. The ITS alternative was designed to allow the markets to continue to compete among themselves to attract order flow, and at the same time to address Congressional goals set forth in the 1975 Securities Acts Amendments. Specifically, ITS was designed to operate under two governing principals identified above. ITS, the ITS Plan, and the Plan's price protection rules were intended to achieve these goals, and we believe they have.

Inherent in the ITS proposal was a radical change in the concept of membership on national securities exchanges, and more particularly in the primary exchanges. As the Commission is aware, only Exchange members may have access to the NYSE market. ITS changed that concept in a limited but material way. The ITS Plan provides that a member of one ITS Participant market can access another ITS Participant market without being a member of that other market. And, significantly, that access is free of any cost to the non-member.36 However, embodied in the ITS Plan is the concept that access to each other's markets is limited in purpose, i.e., access is only to the disseminated quotations in another market and not to the full range of access enjoyed and paid for by members. Because the Exchange is the primary market for its listed stocks, this limited purpose access to the Exchange was, and remains, particularly important. ITS was not and is not intended to be a substitute for Exchange membership.

Since the original six ITS Participants were all floor-based auction markets, the operational parameters set forth in the ITS Plan were fairly straightforward.37 Because ITS contemplates non-member access to each other's markets, the Plan provides that any change to the Plan requires unanimous agreement of all the Participant markets. Since the Commission approved the ITS Plan in 1978, there have been over 25 amendments to the Plan. Because of the unanimous voting requirement, significant amendments are vigorously vetted. Some amendments have been straightforward and relatively easy to negotiate. Others have been difficult. In two instances, differences among the Participants have led to the Commission adopting amendments on its own initiative.

While ITS operates today in the same functional manner as when it was formulated in 1978, the environment today is much different. The process of change began in the early 1980's with the addition of the NASD and its ITS/CAES automated linkage. Then the CSE transformed from a floor-based Participant to a fully electronic one, with the ability to access other markets by computer generating ITS commitments. In the past several years, the Participants struggled with the sharing of the costs two new initiatives -- a backup system and a stand-alone test system -- not provided for in the ITS Plan, but which our competitors argued should be funded pursuant to the Plan's cost-sharing formula.38 Today the BSE and the PCX are undergoing metamorphosis that will lead to floor-less markets.39 The CSE, for reasons unrelated to ITS (but related to its appeal to the Commission on a CTS matter) refused for an extended period to execute the Plan amendment necessary for the BSE and PCX to go forward.40 In addition, the options antitrust investigations have become an impediment to some of the ITS Participants' willingness to discuss issues inherent to the ITS Plan without Commission written authorization and/or the presence of Commission staff and anti-trust lawyers. At the same time, the Commission staff has been prodding the Participants to prepare to deal with the potentials of ECN access to ITS and the certification of new ATS exchanges -- even though none of these entities has made substantive approaches to the Participants about access to ITS. In sum, ITS was designed to address market structure issues of a different era. The Exchange believes that a different approach to deal with today's environment is appropriate.

In the course of its review of market structure issues, the Public Directors' Committee carefully reviewed the role of ITS. The Public Directors' Committee came to understand how extremely complex governing, operating, funding, modernizing and expanding a linkage among markets based on different market models has become: it involves varying degrees of floor-based and screen-based trading and pursuing different business models, and that it would become even more complex with the addition of new Participants. The Exchange also believes that any system that requires (admittedly by our own choice and for good reason) nine competitors to arrive at unanimity regarding competitive business issues has, in this day and age, become more burdensome, unworkable and unnecessary. This is particularly true when (1) an unknown number of new voting Participants are likely to apply to join the ITS Plan, and (2) technological advancements afford better ways to achieve the intended purpose of ITS.

The Public Directors' Committee concluded that ITS continues to serve the function for which it was designed. However, the Public Directors' Committee also concluded that managing the complexities may no longer be justified because significant developments in communications technology allow broker-dealer order routing systems to seek out best execution of customer orders without the need for market-to-market linkages. The Public Directors' Committee determined that ITS may be replaced by broker-dealers' pursuing best execution through their own direct linkages to markets, rather than the less efficient process of linking markets to markets.41 In this way, broker-dealers will be able to fulfill best execution responsibilities more efficiently.42

As noted in the Market Structure Report, when the 1975 amendments to the Securities Exchange Act of 1934 were passed, broker-dealers were limited in their ability to find the best quote and quickly route an order to it. Hence, broker-dealers were able to make routing decisions only on a wholesale, bulk basis. The ultimate responsibility for best execution on an order-by-order basis was shifted from the broker-dealer entering an order to the broker handling the order in the market, and ITS is the mechanism used to accomplish this.

The Public Directors' Committee noted that, today, many broker-dealers have available to them systems and search engines that are capable of making rapid order routing decisions, which should obviate the need for market centric intermarket order routing to achieve best execution. The implementation of order routing algorithms based at least principally on the price-time priority of quotations (and thus limit orders) could also effectively address those price-time priority issues that are raised in the Release.43 It would also address the Commission's concerns that price competition through quotations should drive the markets, as market centers desiring to compete for order flow would have to step up and offer competitive quotations available to all participants in the NMS.

The Exchange has long recognized the need to continually monitor our environment and, as appropriate, to "reinvent" the NYSE marketplace to be responsive to our constituencies. We have been at the forefront in discarding outdated thinking, and rethinking our traditional business model. On a number of fronts, ITS has been characterized as "archaic" and not up to today's technology needs of the markets (albeit differing political agendas may drive those arguments). The Public Directors' Committee has concluded that there are newer, smarter, and more flexible technologies in existence than the "hub and spokes" ITS model developed in the 1970s. In recommending the elimination of ITS, our Public Directors did not propose to abandon intermarket linkage. They sought only to "reinvent" the way linkage is accomplished using modern technology.

In the context of clearly-defined and vigorously-enforced best execution obligations and each market's display of limit orders, the listed market would be well positioned to effect that change. We believe that migration from the older hubs-and-spokes technology of the 1970s to brokers and service bureaus employing new order-routing technologies can be effected in an orderly way that does not place best execution at risk. We would work with the industry to assure that appropriate technology is available before the hubs-and-spokes utility is dismantled. In the case of our own trading floor, we would technologically enable our "executing brokers" - - the specialist when he or she acts as agent and the floor brokers - - to have electronic access to other markets through those markets' members. We believe this can be accomplished for listed stocks by the end of this year.

Enabling direct "front door" access between participants and markets has several collateral benefits. First, it ends the undermining of the economics of trading rights embodied in the concept of membership. Second, it allows the listed market to escape the hidden subsidies of the present ITS utility.44 Third, it reconciles the tension between the competitive heterogeneity of the different markets and the conformity fostered by a hubs-and-spokes system.

* * * * *

The Exchange believes that the best market structure for all participants is one that responds to competitive forces driven by the demands of institutional and retail customers. The Exchange also believes that all investors would be best served by:

Furthermore, the Exchange believes that all investors would be ill-served by:

The Exchange is pleased that the Commission is raising these important questions and would be pleased to discuss our response with you.


James E. Buck

c: Arthur Levitt, Chairman
Isaac C. Hunt, Commissioner
Paul A. Carey, Commissioner
Laura S. Unger, Commissioner
Annette Nazareth, Director, Division of Market Regulation
Robert L.D. Colby, Deputy Director, Division of Market Regulation


1 Self-Regulatory Organizations; Notice of Filing of Proposed Rule Change by the New York Stock Exchange, Inc. to Rescind Rule 390; Commission Request for Comment on Issues Relating to Market Fragmentation, Securities Exchange Act Release No. 34-42450; SR-NYSE-99-48 (February 23, 2000)(the "Release").
2 The Commission also asked for comment on the then-proposed repeal of Rule 390. On May 5, 2000, the Commission issued an order approving the rescission of Rule 390. See Securities Exchange Act Release No. 34-42758 (the "Rule 390 Order").
3 At a conference sponsored by the Exchange in December 1999, George Sofianos, Vice President, NYSE, presented statistics on best execution, including price improvement, for market orders on the NYSE and called for the establishment of uniform market-wide best execution measurement standards. U.S. Equity Markets in Transition, December 9-11, 1999. The presentation was based upon a NYSE Working Paper by Jeffrey Bacidore, Katherine Ross, and George Sofianos, "Quantifying Best Execution at the New York Stock Exchange: Market Orders." NYSE Working Paper 99-05, December 1999.
4 See Securities Exchange Act Release No. 34-38963 (August 22, 1997) approving NYSE PRIME.
5 Release, p. 3.
6 In addition to this intermarket competition, intramarket competition on the floor of the Exchange also helps keep quotation spreads and trading costs low.
7 Hasbrouck, Joel, "One Security, Many Markets: Determining the Contributions to Price Discovery," The Journal of Finance, September 1995. Using data from 1993, when the NYSE's portion of tape share volume was 81.9%, Hasbrouck determined that "price discovery appears to be concentrated at the NYSE: the median information share is 92.7 percent". The NYSE's share of tape volume has risen since 1993.
8 In 1980, Rule 19c-3 was adopted. (Release No. 34-16888 (July 18, 1980)). In 1981, the Commission approved the automated interface to the ITS/CAES Linkage. (Release No. 34-17744 (April 21, 1981)). In 1995, after review, the Commission determined to continue to permit payment for order flow, as long as there was mandated disclosure. (Release No. 34-34902 (April 3, 1995)). In 1996, the Commission approved the preferencing programs of the Boston Stock Exchange ("BSE") and the Cincinnati Stock Exchange ("CSE"). (Releases No. 37045 and 37046 (March 29, 1996) (BSE Competing Specialist and CSE Preferencing Approval Orders, respectively)). In 1998, the Commission adopted new rules to allow alternative trading systems to choose whether to register as national securities exchanges, or to register as broker-dealers and comply with additional requirements under Regulation ATS, depending on their activities and trading volume. (Release No. 34-40760 (December 9, 1998)).
9 Indeed, two Regional exchanges have structured their markets to permit multiple members to register as specialists in the same security, with each allowed to preference its own order flow to itself and to receive order flow from those whom they pay. Other markets use revenues from market data fees to make rebates to their market-makers to use to purchase order flow.
10 As stated above, in the Exchange's view, "captive customer order flow" includes both a broker's internalization of its own customer's orders (including the routing of orders to an affiliate of the broker) and the orders a dealer buys from another broker (its customer) to internalize
11 See Letter dated December 9, 1993 from James E. Buck, Secretary, NYSE to Jonathan G. Katz, Secretary, Commission, in response to the Commission's proposal in Release 34-33026 of a rule requiring disclosure of payment for order flow practices. See also letter dated November 24, 1992 from James E. Buck, Secretary, NYSE to Jonathan G. Katz, Secretary, Commission, in which the Exchange urged the Commission to ban the practice for listed securities (Market 2000 Comment letter). See also letter dated April 5, 1995 from Daniel Parker Odell, Assistant Secretary, NYSE, to Jonathan G. Katz, Secretary, Commission, in response the Cincinnati Stock Exchange's filing seeking permanent approval of its Preferencing Pilot. See also letter dated January 15, 1996 from James E. Buck, Secretary, NYSE, to Jonathan G. Katz, Secretary, Commission, responding to the Commission's release on Order Execution Obligations (SEC File No. S7-30-95). The Exchange incorporates by reference its comments stated therein.
12 Commission staff has explained that the use of the term "system" in this example is meant to connote market structures with a proven record of price improvement for market orders, such as the NYSE auction market. The staff stated that a theoretical possibility of price improvement would not be sufficient under the first example. Staff also stated that the Commission might consider the institution of a minimum standard of actual price improvement if this proposal were to be adopted. Telephone conference of March 30, 2000, between Daniel Gray, Senior Special Counsel, Office of Market Structure, Division of Market Regulation, Commission, and Karen Lorentz, Director, Intermarket Relations, NYSE.
13 Securities Exchange Act Release No. 34-36310 (October 6, 1995) (proposed Rule 11Ac1-5 - price improvement for customer market orders).
14 Customers, however, should be free to elect different treatment (such as choosing an automatic execution alternative) if it fits their investment needs.
15 "Responding to Chairman Levitt's Call: A Plan for Achieving a True National Market System," (February 29, 2000). (
16 In describing the interaction among large and small orders in the NYSE auction market, the Commission, in the Rule 390 Order, recognizes that investors with large trading interest "...typically will not display their full interest in a limit order because it likely would move the market against them, thereby increasing their transaction costs or even precluding any execution at all."
17 Economic inducements for order flow bifurcate the market to some degree; a CLOB would do so in significant degree.
18 We note that, in testimony on April 13, 2000, before the Senate Banking Committee, Federal Reserve Board Chairman Alan Greenspan also cautioned against a government mandate for any particular form of a central limit order book.
19 A related issue that will be presented with the implementation of decimalization is so-called "penny jumping", that is, a dealer bidding higher or offering lower, by the permissible minimum price variation ("MPV"), than the pre-established national best bid offer in order to trade as principal with incoming order flow. If the Commission perceives that there is a problem with the practice, it has the authority to establish a national MPV at a level that would make "penny jumping" uneconomical. Moreover, the Exchange would monitor specialist activities and take appropriate regulatory action if abuses are identified.
20 An alternative that would give priority to agency limit orders would not only require amending the Firm Quote Rule, but would also require each market center to modify all its quote processing systems to disseminate and receive both a dealer and agency quote stream. Vendors would also need to modify their displays as well.
21 For example, as discussed above, if all markets (and market-makers as appropriate) publicly displayed their limit order books and the industry developed a uniform best execution measurement standards, brokers would have more information to make better order routing decisions.
22 See footnotes 29 and 32.
23 Securities Exchange Act Release No. 34-37619A (September 6, 1996).
24 Securities Exchange Act Release No. 34-14415 (January 26, 1978).
25 The Commission recognized that the ITS Plan would continue to require that, as a prerequisite to access in a particular security, ITS Participant markets would continue to provide continuous two-sided quotations. The ITS Plan also provides for the nominal autoquoting of 100 share markets, which is the practice of many regional specialists and third market-makers.
26 Securities Exchange Act Release No. 34-18482 (February 11, 1982) ("Final Rule Amendment"); Securities Exchange Act Release No. 34-17583 (February 27, 1981) ("Proposed Rule Amendments").
27 In 1999, shares bought and sold by specialists for their own account represented 13% of the total shares bought and sold at the NYSE.
28 Additionally, while certain ECNs have been lobbying for direct access, apart from a SRO, Chairman Levitt, in his December 22, 1999 letter to Senator Charles Schumer made it clear that, under Commission rules and policies, direct access to NMS plans and systems is confined to SROs.
29 The average times to execute and turn around an ITS commitment and a SuperDOT order have been comparable since the implementation of the Electronic Book -- 16.3 seconds for ITS and 19.9 seconds for SuperDOT, during a sample period in March 2000. Furthermore, the Exchange's ITS execution rate is routinely the highest among the ITS Participant markets -- 94% for January and February 2000, compared to a 69% rate for the regional and third markets.
30 Letter dated August 31, 1998, from James E. Buck, Secretary, NYSE, to Jonathan Katz, Secretary, Commission.
31 By way of example, the Commission is aware of the protracted negotiations between the Pacific Exchange (the "PCX") and the NYSE concerning the PCX's proposal to link the OptiMark System to the NYSE via ITS by making it a PCX facility. At the outset, the Exchange noted that the very nature of OptiMark raised the potential that it could be used as an unfettered access vehicle to the NYSE.

Ultimately, because of the unanimity requirement, the parties came to agreement and the Plan incorporated measurements - the first was set at 15%, which if exceeded provides PCX up to two years before consequences are imposed; the second was set at 30%, which if exceeded provides that PCX not use ITS for OptiMark for a three month period. At the first measurement date, PCX reported that 81.5% of OptiMark business had been sent to other markets via ITS. Of that, 96% had accessed the NYSE. Without the unanimous voting requirement, the NYSE would have been unable to prevent OptiMark from enjoying the privileges of NYSE membership without bearing a fair share of the costs of operating the Exchange.

32 For the months of January and February, 2000, only 2.0% of share volume and 1.9% of trade volume in NYSE listed stocks traded over ITS. 68% of the ITS share volume and 67% of the ITS trade volume were executed on the NYSE. The Exchange accessed other Participant markets for 24% of total ITS executed share volume and 20% of total ITS trade volume. The NYSE's use of ITS represents only 0.59% and 0.53% of NYSE reported shares and trades, respectively.
33 We also note that members on the national options exchanges have use of advanced technologies linked directly or indirectly to SuperDOT to route orders to offset options positions in substantially greater volume than the ITS Participants use of ITS. Thus there is ample evidence to demonstrate that there exist other forms of intermarket linkage, apart from ITS, that fill the needs of market-makers.
34 We note that when it adopted the Order Handling Rules, the Commission addressed the issue of "hidden limit orders", requiring those that equal or better the market to be quoted. In essence, regional and third market operate on the basis of "hidden quotations", i.e., publishing nominal quotations generally away from the market, but yet offering better quotations to their public customers (i.e., matching the NBBO). These are, in effect, the firmest of quotes but are not published to the NMS for the benefit of all market participants.
35 For example, a regional specialist who receives, an order to buy 1000 shares at 50, may send to SuperDOT an order to buy at 50 some or all of the 1000 shares -- the regional specialist is said to have "marked" the NYSE specialist's book. When regional specialists receive execution reports from SuperDOT, they in turn give executions to their customers on their exchange. Regional specialists also report their trades to CTS, even though their role in the two trades is, in effect, as a "riskless principal". This "double printing" of the transaction, of course, results in additional CTS revenues to the regional.
36 The cost of direct access to the Exchange in the early days of ITS was much more significant than today. Then the Exchange fees charged to members were 1% of their gross commissions on listed stock trading - a time when commission rates were much higher. Today, access to SuperDOT is free of Exchange fees for orders below 2,100 shares, and only $1.90 for orders of competing dealers.
37 The original Participants in the ITS Plan were the American, Boston, Midwest (now Chicago), New York, Pacific and Philadelphia Stock Exchanges.
38 The Plan provides generally that development and production costs are shared based upon a formula tied to Participants share of CTS trade reports. This results in the Exchange funding 70% to 75% of ITS costs. Its is the most explicit of the subsidies to the other markets, which use ITS to access us far more than we use it to access them. See footnote 32.
39 A Plan amendment has been drafted and agreed upon in principle, subject to CSE ratification, by the Participants to accommodate the BSE and PCX. However, we have learned that the Commission staff has raised concerns about the specificity of operation contained in the amendment, and in other provisions of the Plan.
40 Appeal of CSE Administrative Proceeding File#3-9967.
41 This new linkage paradigm would also obviate all issues relating to the governance of ITS.
42 Members on the regional exchanges, the third market, and the options exchanges all use derivate pricing based on the NYSE market to conduct their business. Members on the options exchanges have hedging needs at least as compelling as the lay-off and best execution needs of the regional exchange and third market members for intermarket access. We estimate that approximately 10% of NYSE's share volume is derived from activity related to options trading -- significantly more volume than received from ITS. However, options members have developed sophisticated technological means to efficiently and rapidly enter orders into SuperDot to satisfy their access needs. Furthermore, options members use SuperDot as customers and receive no subsidies from the Exchange for that use. On the other hand, the regional exchanges, the third market and their members -- our competitors --receive subsidies in the forms of free ITS access to the NYSE market, and the NYSE funds a disproportionate share of ITS costs. See footnotes 29 and 32.
43 The Commission recognized technological advances in order routing capability in the proposed Order Handling Rules Release. (Release No. 34-36310 (September 29, 1995)). The technology is also available for broker-dealers to be able to evaluate their order routing decisions on an historical basis. For example, a broker-dealer may route orders to a market that appears to improve the NBBO on a regular basis, but if a historical analysis subsequently shows that orders sent to that market receive the quoted improved price less frequently than on other markets, the broker-dealer can fine-tune these routing decisions to fulfill its best execution obligation.
44 As we discussed in our recent response to the Commission's concept release on market data, hindsight shows that market structure intervention invariably carries with it unforeseeable costs resulting from market distortion. Indeed, as is clearly the case with the options allocation plan, the cure can sometimes be worse than the disease. (See letter dated April 10, 2000, from James E. Buck, Secretary, NYSE, to Jonathan G. Katz, Secretary, Commission re: File S7-28-99)