October 28, 2002
Mr. Jonathan G. Katz
Re: File No. 4-466 (Hearings on Market Structure Issues)
Dear Mr. Katz:
The American Stock Exchange ("Amex") would like to thank the Securities and Exchange Commission ("Commission") for the opportunity to present our views on the current structure of the U.S. securities markets. We also would like to commend the Commission for undertaking a necessary review of the market structure that has developed since the promulgation of Regulation ATS.
It is almost four years since the adoption of Regulation ATS and we have all had ample time to observe its effects. Simply put, we believe that Regulation ATS has proved to be a mistake, undermining the core principals enshrined in the Congressionally mandated national market system that the Commission has worked to develop for more than a quarter century - best execution, transparency, equal regulatory oversight and fair competition. We, therefore, urge the Commission to return to the fundamental values of the national market system and require all markets, no matter what name they go by, to satisfy and uphold the core principals.
Background of the American Stock Exchange
The American Stock Exchange has a long history of innovation and is unique among the U.S. securities markets in that we are the only market that actively lists and trades securities across three diverse business lines - equities, options and exchange traded funds, commonly referred to as ETFs. In equities, we focus principally on providing a well regulated auction market for small and mid-cap companies. Our options market is the second largest in the United States. Indeed, for the first time since beginning this business more than 25 years ago, we have recently had weeks where we are the most active equity options market.
What really sets the Amex apart from all other U.S. markets, however, are ETFs, probably the fastest growing, most innovative financial products offered by any financial institution over the last decade. After more than four years of working with the Commission and millions of dollars of R&D expense, the Amex pioneered ETFs in 1993 with the introduction of an ETF based on the S&P 500® index, known as Spiders (Amex: SPY). Since then, we have spent tens of millions more developing new products and educating the marketplace about the benefits of ETFs. Nine years later, the Amex remains the clear leader in ETFs, listing 121 of the 123 listed in the U.S. market today, including in addition to Spiders, the Nasdaq 100® ETF (Amex: QQQ) and the Dow Jones Industrial Average® ETF, known as Diamonds® (Amex: DIA). Outside the American market, we have been able to leverage our reputation in ETFs to create a global presence for the Amex, reaching agreements to trade Amex-listed ETFs in Europe and Asia.
Not content to rest on our laurels, we are busy creating the next generation of ETF products - innovative products that will provide investors even greater flexibility and new investment opportunities. For example, we recently launched fixed income ETFs and are getting ready to introduce leveraged ETFs, inverse ETFs and, most significantly, actively managed ETFs.
In short, the American Stock Exchange has emerged as a strong, innovative international competitor, precisely the type of market that the Commission hoped to foster by its promulgation of Regulation ATS. Significantly, these accomplishments have come without the need for the Commission to bend or waive the fundamental regulatory scheme that has shaped and fostered the national market system over the last 25 years. Rather, the American Stock Exchange has spent countless hours and millions of dollars either conforming these new products to the existing regulatory structure or laboriously working with Commission staff to change rules where appropriate. Notwithstanding the time, work and money spent developing these products in an environment where our competitors, because of unlisted trading privileges, can begin trading them virtually as soon as we do, the very accomplishments of the American Stock Exchange over the last ten years are now seriously threatened by Regulation ATS, which provides ECNs significant regulatory waivers from the burdens of exchange regulation, all at the expense of the integrity of the national market system, a system mandated by Congress to ensure public investors receive best execution of their investment decisions, or simply said, investor protection.
Topics for Market Structure Hearings
1. Market Data
The linchpin of the national market system has been the availability to all investors and market participants of real-time consolidated quotation and trade information from each market and each markets' members. True market transparency requires that every market's quote be reflected in the NBBO and that all members of each market provide quotation information for inclusion in the NBBO. The Commission asks if consolidated trade reports matter to retail investors. The answer is "Absolutely yes" and so do consolidated quotations. Retail investors may seldom look at last sale data, but, just as investors may seldom examine footnotes to corporate financial statements, the transparency and openness of market information provided by consolidated trade reports, or, financial footnotes, can be vital to protecting investors and promoting sound investment decisions. Real-time consolidated trade information provides a basis for comparing the execution price with prices available in other markets and is a by-product of the national market system for measuring best execution.
The movement by some markets towards display of supplemental market information such as display of "depth of book" improves market transparency. While the Commission should allow the dissemination of this information, the decision whether to make this information available and the nature and extent of that availability should be left to competitive forces in the marketplace.
Rather, Commission regulation and enforcement should be focused first and foremost on ensuring that real-time dissemination of the best bid and offer accurately reflects all quotation and limit order information. Over roughly the past 18 months, significant trading activity has occurred outside, and in flagrant violation of, the most important mechanisms of the national market system. These deviations from the rules are materially undermining the system. Yet, despite repeated objections and many proposals to address this problem by the Amex and other market participants, the abuses continue to this day. Further, through a variety of schemes, the revenues generated from CQ and CTA directly support the operations of those markets that steadfastly refuse to participate in most significant components of our national market system. As if this were not enough, this revenue is also being used to facilitate and, indeed, encourage fraudulent and misleading trade reporting. Even if the Commission determines to allow these markets to operate as they do - a concept that we find incomprehensible, the Commission must put an immediate end to having the national market system provide direct financial support to their operations.
Consistent with our position during meetings of the Market Data Advisory Committee, Amex sees no advantage in having multiple, competing consolidators of market data, as advocated by some Advisory Committee participants. The consolidation function is best performed by an exclusive consolidator under the CQ and CTA Plans. We are concerned that increasing the number of consolidators would simply introduce complexity and inefficiencies, such as the need for multiple disaster recovery sites and plans, without producing any real, offsetting benefits. Moreover, we believe that the competing consolidator model would create the possibility of differing data streams and thus is fundamentally at odds with the notion of a national market system.
As the exclusive information processor under the CQ and CTA Plans, SIAC has continually demonstrated the efficiency of its operations by its long record of providing reliable, real-time market data to the industry without significant disruptions. This record is all the more impressive if one takes into account the exponential growth in the amount of data that SIAC has had to collect, consolidate and make available for re-dissemination, especially the in area of options market data. We are not opposed, however, to explore opening the exclusive processing function to competitive bidding.
Market data rates are set by the CQ/CTA Operating Committee under the provisions of the CQ/CTA Plans. These Plans include significant procedural checks and balances. Fee changes must be approved the Operating Committee. Generally, however, before the Operating Committee votes, the proposals are presented to the boards of the market centers represented on the Operating Committee for consideration by the broader range of interests serving on those boards. After the Operating Committee approves fee changes, they are submitted to the Commission for public comment and approval, providing a further opportunity for broader industry input. We believe this is a sound procedure for setting rates based on consensus of the markets.
Revenues under the CQ and CTA Plans are allocated among market centers on the basis of the respective number of trades they each report, without regard to share volume or the quality of those markets. This approach disadvantages the market centers that provide greater liquidity by treating a single trade for 100 shares of stock the same as a single trade for 100,000 shares of the same stock in a different market. By the same token, the quality of quotation information is completely ignored in the allocation formula.
Though the current methodology has the virtue of being procedurally simple, its simplicity comes at the great expense of substance. Indeed, its reliance solely on counting the number of trades (a practice dating back to the mid-1970s) causes it to be overly simplistic and, unfortunately, subject to gaming and manipulation solely for the purposes of collecting market data revenue, not improving the quality of the market. The time has come to recognize that the current methodology has utterly failed to achieve what should be the major goals of the Commission and the CQ/CTA participants - to encourage the provision of maximum liquidity to, and the tightest possible spreads in, national market system securities.
This past Friday, the American Stock Exchange submitted to the other markets a proposal for a more sophisticated and substantive method for allocating market data revenues among the CQ/CTA Plan participants. Under this new approach, markets would be encouraged through the policies of the revenue sharing formula to improve market quality by increasing liquidity, tightening spreads and enhancing price discovery.
Two obvious and important objectives, especially in today's increasingly fragmented and volatile markets can be achieved by changing the current methodology. First, if the CQ/CTA Plans were to include a volume measure, in addition to a trade count, when calculating the revenue distribution related to executions, all marketplaces would be encouraged to trade in the maximum depth, providing the greatest liquidity possible.
Second, if the CQ/CTA Plans were to include as part of the revenue distribution mechanism an incentive to provide the National Best Bid ("NBB") or National Best Offer ("NBO"), weighted by volume and time, but limited in its impact to the volume traded in that security relative to all Network A or Network B securities (to prevent attempts to manipulate the revenue distribution methodology by putting up meaningless quotes), the CQ/CTA revenue pool would reward those Plan participants that are actively providing quotations that are meaningful (that is, quoting in size), contain the tightest bid/offer spreads and have the greatest liquidity, rather than those marketplaces (or their participants) that have learned how to manipulate the system solely with the aim of divesting tape revenue.
Under the Amex proposal, the distribution of CQ/CTA market data revenues would generally be 25% based on each of the following: number of trades; share volume traded; bids that are the NBB; and offers that are the NBO. We look forward to discussing this important issue with the other market participants and the Commission in the coming weeks.
Whatever the revenue allocation formula, however, the core national market system facilities for exchange-listed equities - CQ, ITS and CTA - are designed, among other things, to protect investors by providing mechanisms to ensure best execution. Market participants by whatever name that choose not to undertake these obligations should not share in the revenues derived from them. In addition, no market should have financial incentives under the CQ/CTA Plans to reward members, by means of cash payment for order flow, that operate outside the national market system with revenues directly resulting from the operation of national market system facilities.
Indeed, this past February a majority of the CQ/CTA Participants voted in favor of not giving a Participant credit for its members' trades if that member is not a full participant in the national market system (i.e., quoting in CQ, accessible through ITS and reporting to CTA). Unfortunately, the Commission determined that this action needed unanimous approval of the Participants and, therefore, we were forced to reverse the action. After a number of unsuccessful attempts to achieve unanimity, the Amex submitted to the Commission in May 2002 a petition to amend the revenue sharing formula to provide that CQ/CTA Participants will not be given credit for their members' trades if that member is not a full participant in the national market system. To date, however, the Commission has been unwilling to publish this proposal for public comment.
Market data revenue is an important revenue source for funding SRO operations, including regulation and technology development and operation. The value and integrity of the quotation and last sale information is the product of the markets' regulatory and technological efforts as a whole and not merely the cost of processing market data. However, market data revenue rebate programs adversely impact the national market system and self-regulation by encouraging the regulatory abuses, noted above, by members printing trades solely to accrue market data revenue. Indeed, such programs conflict with the national market system because they reward members merely for reporting trades in a particular SRO's market, regardless of whether a member has met its best execution obligations.
Ironically, if the Commission had not forced the reversal of the interpretation approved this past February, many if not all of the recent problems would have long been resolved. In particular, Island would have had no incentive to develop the scheme that has led to the proliferation of trade shredding and wash sales.1
2. Best Execution/Agency Issues
Proponents of alternative trading systems and Nasdaq recently have been advocating "speed of execution" as the primary consideration in best execution.2 The Commission, however, has clearly stated, "price is the predominant element of the duty of best execution."3 The Commission, moreover, has stated that a delay in order execution to achieve the best price is a precept of the national market system:
The Commission's view of a national market system consisting of "equally regulated, individual markets, which are linked together to make their best prices publicly known and accessible," presupposes that price is the most important factor in best execution and that speed of execution, like many other considerations, is a secondary factor. Linking markets to make their prices publicly known and accessible requires the transmission of orders between markets, which necessarily causes delay as orders are transmitted from one market to another. The Commission remains firmly committed to this policy, recently ordering the options exchanges to put in place an intermarket linkage system.
In our view, the agency auction market structure of the Amex provides the optimal combination of speed with price improvement opportunities, particularly for orders larger than 500 shares. Attached is a Rule 11Ac1-5 comparison for the month of August 2002 in the QQQ ETF - the Amex's most heavily traded security (see Exhibit A). The data compares the Amex to the other major market centers in both market and marketable limit orders. The data shows that the Amex provides the best effective spread in 3 of the 4 order size categories and in the 4th, we have the second best effective spread. In contrast, despite their speed, Island is last in 3 of the 4 order sizes and in the 4th, it is second to last. The data also demonstrates that the Amex's execution speed remains relatively constant among all order sizes, while Knight and CHX drastically slow down as they handle larger orders. These trends are consistent month-after-month in QQQ.
We believe that intermarket linkages such as ITS provide a useful mechanism for linking markets. We also believe that de minimis trade throughs in traditional common stocks should not trigger the application of the ITS trade through rule if the benefits of a speedy execution in one market center truly outweigh the possible harm of an investor missing an execution in another market center. Absent a showing of harm to investors, however, we do not believe that a market with automatic execution should be excused from such linkages merely because complying with linkages would slow its order processing.
For example, handling orders transmitted to the Amex from other market centers also may slow trading on the Amex, but that does not mean that we should be inaccessible to the national market system. That said, we agree that ITS requires revisions for securities such as ETFs to account for the rapid quote changes associated with derivatively priced securities. The case for modifying the trade through rule for common stocks, however, is less compelling than for ETFs. In our view, the national market system should collectively function as the price discovery market for common stocks and produce an efficient market price for the securities traded in it. In contrast, there are several possible price discovery markets for ETFs: e.g., stock baskets, index futures, or the ETFs. Thus, trade throughs in ETFs do not have the same impact on price discovery as they have on common stocks.
While we agree that some modifications should be made to ITS to better accommodate derivatively priced securities, we are disturbed by indications that the Commission is ignoring its longstanding lodestar that price is the most important element of best execution. We are particularly troubled by the Commission's recent approval, on an accelerated basis, of a Nasdaq pilot program to allow SuperSOES orders to trade through a UTP Exchange's quote without any consideration of the impact of such a rule on a broker-dealer's best execution obligation.5 The Commission staff has granted similar no action relief to the Chicago Stock Exchange to allow CHX specialists to trade through the Amex's quote in a Nasdaq stock (although the staff apparently stated that the no-action position did not affect the best execution obligations of CHX specialists).6
The relief given to the Nasdaq market and CHX is premised on the supposed inaccessibility of UTP exchanges' quotes. In the case of the Amex, this supposed inaccessibility is simply not true. For years the Amex has had electronic order routing and execution systems that, while not as fast as the fastest ATSs for small orders, are as fast or faster than other market centers for larger orders. We thus question whether the markets that seek to exclude the Amex's quotes are doing so to preserve their market share in Nasdaq stocks rather than to protect their systems. We observe no small amount of hypocrisy, moreover, in the Nasdaq's complaints of the Amex's inaccessibility, when Nasdaq shares millions of dollars with an invisible and inaccessible market (Island) through it market data revenue sharing arrangements. As noted in the Market Data section above, these cash payments for order flow by certain participants in the CQ and CTA Plans have led to widespread concerns over tape shredding and wash trades. We believe that the Commission should simply prohibit these schemes
3. Exchanges and the Self-Regulatory System
In the early 1990s, spurred by a desire to encourage innovation of new markets and to "avoid the strait jacket of exchange regulation,"7 Commission staff began to give operators of trading systems that did not enhance liquidity in traditional ways through specialists, market makers or a single price auction structure, "no-action" relief if those systems operated without registering as exchanges so long as they abided by conditions set forth in the no-action letters.
In 1998, reacting in part to its well-founded concerns documented in the Commission's 21(A) Report against the NASD8 that alternative trading system were leading to market fragmentation and harming market transparency by operating as private "hidden markets," the Commission adopted Regulation ATS.9 Specifically, the operation of these alternative trading systems was - and, unfortunately, still is - leading to a two-tiered market, an unofficial one only viewable and accessible by the alternative trading system's members and the official market being created by the national market system and used by public investors. The Commission also took this step to address the growing regulatory disparity between ATS's and other markets, disparities the Commission found negatively affected other securities markets and, most importantly, public investors.10
In adopting Regulation ATS, the Commission sought to establish a better balance between the regulatory needs of the Congressionally mandated national market system and the need to encourage the development of innovative new markets. The Commission sought to accomplish its goal by allowing, on the one hand, an ATS that operated below a threshold of 5% of the average daily trading volume in a security largely to escape the regulatory constraints placed upon registered exchanges. On the other hand, in an effort to bring ATSs into the national market system, Regulation ATS attempted to subject an ATS that exceeded the 5% threshold to an order display and equivalent access requirement and an ATS that exceeded a 20% threshold to a fair access and certain requirements relating to its operational system.
Since its adoption, it has become increasing clear that Regulation ATS has not resulted in the better balance between regulation and innovation sought by the Commission. While a single de minimis ATS (an ATS with less than 5% market share) may not have a significant impact on the U.S. securities markets, the Commission failed to anticipate that the trading of multiple ATSs operating under the de minimis exemption can, in the aggregate, have a very negative overall impact on the national market system's guiding principals of transparency, best execution, equal regulation and fair competition.
Indeed, the ability of an ATS to frustrate the Regulation ATS requirements designed to integrate ATSs into the national market system has recently been vividly demonstrated by The Island ECN's choice to "go dark", a tactic it adopted notwithstanding the Commission's grant of an unprecedented exemption to a core national market system principle designed specifically to accommodate Island. Thus, a market, like Island, that matches customer orders with other customer orders, does not display its customers' orders and reports its trades through the CT Plan avoids the most substantive provisions of Regulation ATS, including: (1) the order display and equivalent access requirement, (2) the limitation on fees that are inconsistent with the equivalent access requirement, (3) the fair access requirement, and (4) the requirements with respect to the capacity, integrity, and security of the ATS's automated systems.
By going dark, Island achieved precisely the result that the Commission sought to avoid with the adoption of Regulation ATS, namely the presence of a two-tired market - an unofficial one only viewable and accessible by the alternative trading system's members and the official market existing within the national market system that is available to investors. Island's actions have lead to the truly perverse result seen today of an alternative trading system (Island) and a facility of a national securities association (Nasdaq) sharing in revenue generated by two national market system plans (the CQ and CTA Plans) while the ATS is invisible and inaccessible to the intended beneficiaries of the national market system - the investing public. It should go without saying that allowing significant ATSs to opt out of the national market system because the Commission is reluctant to allow Regulation ATS to be enforced or because of the exception that allows markets, like Island, to go dark, undermines the core national market principles of transparency, best execution, equal regulation and fair competition.
In contrast to the "regulation lite" regimen applicable to ATSs, exchanges are subject to a raft of burdensome requirements. Exchanges are obligated to enforce compliance by their members with their rules and the federal securities laws.11 Pursuant to this obligation, the exchanges have spent heavily on technology and incur significant data storage costs in connection with the fulfillment of their obligation to surveil trading in their markets. Not only are these systems very expensive to create, maintain and revise as is frequently needed, but given their necessary limitations, they also require the exchanges employ large staffs to review the various reports created by them. The Commission, moreover, recently brought several enforcement actions against SROs when it believed that they were not fulfilling their SRO responsibilities. In contrast to the Commission's willingness to take enforcement action against SROs, we understand that the Commission staff recently directed an SRO in writing to desist from taking disciplinary action against an alternative trading system when the SRO was preparing an enforcement action for clear and open violations of the requirements of Regulation ATS.
In addition to the exchanges' wide-ranging regulatory responsibilities, they also are subject to a number of additional burdensome and costly requirements that are inapplicable to ATSs (even though ATSs and exchanges both meet the Rule 3b-16 definition of "exchange"). Among these additional requirements are obligations to file and obtain Commission approval of rule and system changes, file to adopt, change and even eliminate fees (which must be fair), provide for fair representation of members in the management of exchange affairs, have outside directors on the governing board, dual siting and system redundancy requirements and fair membership access rules. The numerous requirements applicable to exchanges stifle innovation and impede their ability to compete with the less regulated and, therefore, more nimble ATSs. As noted above, the Commission itself has characterized the pattern of exchange regulation as a "strait jacket." In fact, in today's regulatory environment - ranging from an exemption from the definition of exchange for some market participants that is so significant that it has swallowed the definition to the use of market data revenues by markets as a source of payment for order flow - there no longer is any meaningful reason to put on the exchange strait jacket.
The Commission's policy bias towards alternative trading systems not only conflicts with the national market system goals of equally regulated, linked markets making their best prices publicly known and accessible, it also damages the interests of investors by favoring markets that do not employ liquidity providers such as specialists and market makers with affirmative obligations to the market.12 We thus question the public policy basis for more burdensome regulation on markets that feature specialists and market makers with affirmative obligations as opposed to markets that lack such liquidity providers, particularly when there is no demonstration that markets that feature liquidity providers with affirmative obligations are more susceptible to trading practice abuses than markets that lack affirmative market making requirements.
While we appreciate - indeed, support - the Commission's desire to foster innovation in the development of our markets by removing the "strait jacket" of exchange regulation, we take exception with the unarticulated assumption that alternative trading systems are more innovative than markets registered as exchanges. We believe the American Stock Exchange's innovative leadership over the past 25 years belies this assumption, particularly while wearing the strait jacket. Indeed, the Amex continues to be a leader in innovation, even in a regulatory environment where participants such as Island operate in unabated and protected free riding of our innovation, to the direct detriment of the very reason we exist - the public investor.
To make matters worse, Island markets itself as faster and less costly than exchanges - an absurd claim to anyone truly familiar with their practices. The principal reason Island is faster is because it ignores the investor protection rules followed by the other markets that ensure investors receive the best available price in the market - that is, it refuses to participate in the Congressionally mandated national market system. Worse still, although it defies new Commission rules explicitly requiring them to join the national market system, they actually receive revenue generated from the national market system. In other words, we - the markets that comply with federal securities laws by fully participating in the national market system - are actually providing direct financial support to a competitor that is knowingly violating mandates of the federal securities laws.
Finally, we believe the national market system and, more importantly, public investors, would be far better served by ending the Commission's policy bias towards new, so-called innovative markets and developing a more evenhanded, contemporary approach to exchange regulation. The Commission should level the playing field between registered exchanges and alternative trading systems by repealing Regulation ATS and requiring the affected markets to register as exchanges. Further, we believe the Commission should give serious consideration to freeing exchanges from their strait jackets and apply a regulatory regime more relevant to the today's capital markets. In view of the development of ATSs in recent years and their acceptance in the U.S. markets, we see no reason to maintain two wildly dissimilar regulatory structures for organizations that perform the historical function of an exchange in providing a forum for trading and fit the Commission's definition of an exchange.
The Commission concluded in the NASD 21(A) Report that invisible and inaccessible markets damaged investors by degrading best execution and transparency. We hope that this lesson is not forgotten. In addition, we wish to assure the Commission that unequal regulation not only damages markets registered as exchanges, it also harms investors by stifling innovation at the registered exchanges and weakening market centers that provide investors with assured liquidity and price discovery. We therefore urge the Commission to recall the lessons of the past and restore the national market principles of best execution, transparency, equal regulation and fair competition by repealing Regulation ATS and modernizing exchange regulation.
* * * * *
We appreciate the Commission reviewing market structure issues at this time and would be pleased to further discuss our views with the Commission or staff.
Very truly yours,
cc: Chairman Harvey L. Pitt
Rule 11Ac1-5 Data for August 2002 for QQQ Exchange Traded Funds