September 13, 2000

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

Re: SEC File No. SR-NASD-99-53

Dear Ladies and Gentleman:

During a decade of remarkable change and innovation, American Century1 has been a strong, persistent advocate for structural reform in U.S. and international equity markets. This effort has been both aided and driven by technological advances, which have spurred innovation and created significant competition in the marketplace. We calculate that these advances, highlighted by the emergence of ECNs and other innovations, have saved our shareholders more than $1 billion dollars in the last five years.2 We have lowered our implicit (market impact) and explicit (commissions) trading costs by using new trading platforms which protect our anonymity and allow our institutional orders standing in the marketplace. These innovations have leveled the playing field for institutions and their shareholders by curtailing the necessity of brokerage intermediation that has been an embedded part of historic market structure.

Supermontage Seeks to Negate Recent Market Structure Improvements

NASDAQ's Supermontage proposal and subsequent amendments threatens to negate the technological advances (ECNs) and market reforms (Order Handling Rules) that have created a trading environment providing American Century shareholders and others with enormous savings. Now NASDAQ, through Supermontage, seeks to utilize new technology and a new set of protocols to preserve and rescue an old model based on the central position of the traditional broker in the trading process.

Supermontage, as proposed, represents a significant setback for institutional investors and should not be approved.3 Supermontage compels ECNs to publish tradeable quotes to meet obligations under the Order Handling Rules. Because NASDAQ is the only securities information processor, ECNs have no order display alternatives. It is not a voluntary structure.

Supermontage Tainted by Series of Flawed Amendments

American Century first commented on the Supermontage proposal in January 2000 and characterized it as a small step in the right direction. Depth of book, anonymous routing of orders, and trade through protection were viewed as positive first steps in the evolution of NASDAQ's market structure. As Supermontage's proposed structure evolved in a series of amendments to the proposal, the details revealed a different agenda. NASDAQ's goal of good sound market structure reform has been supplanted by market structure that sounds good to only certain and established classes of market making NASDAQ firms. We believe Supermontage's amendments are not an attempt to innovate and, in fact, represent a thinly veiled attempt to appropriate and supplant the economics of major, efficient competitors in the brokerage business.5

Supermontage Will Widen Spreads and Compromise Institutional Anonymity

Supermontage abrogates price and time priority due to its complex order routing priorities, which we understand to be the following:

1) A market participant against its own quote;

2) Market makers and ECNs that do not charge access fees;

3) ECNs that charge access fees;

4) Reserve size of market makers and ECNs that do not charge access fees;

5) Reserve sizes of ECNs that charge access fees; and

6) UTP interest.

The structure of Supermontage asserts a priori that all orders are not created equal. This inequity places market participants using ECNs in a no-win situation. They must choose between anonymity and their standing in NASDAQs's order queue. It clearly establishes market maker priority over ECNs that charge access fees and UTPs. NASDAQ's design drives orderflow back into the hands of traditional market makers, the group most protective of the status quo and least interested in the progressive evolution of equity market. During the last half of 1999, American Century traded 85% of its eligible OTC orders on ECNs that charge access fees. We traded over 170mm shares that represented $8.8 billion of our shareholders assets. We were able to save them millions of dollars in transaction costs because of both technological innovation and regulatory reform.

The order handling rules were designed to protect the investing public's interests. The structure proposed in Supermontage divides the order book into separate classes and establishes price over time in priority queues and hence threatens the ability of new entrants to compete. Let us explain: For example, an ECN charging an access fee enters an order in PDLI that betters the NBBO (National Best Bid and Offer).

Bid Ask

PDLI 92 92-1/2

ECN1 92 95

MM1 91-1/2 93

MM2 91-1/2 93

ECN2 90 92-1/2

A market maker then enters a similarly priced order.

Bid Ask

PDLI 92 92-1/2

MM1 92 93-1/2

ECN1 92 95

Under the Supermontage algorithm, the market makers order goes to the top of the order book. This structure establishes prima facie that NASDAQ market makers are granted a right of first refusal in competition with ECNs and other markets. Furthermore, Supermontage establishes internalization as the de facto NASDAQ market standard.

Supermontage Threatens Competition and Innovation in NASDAQ

The Supermontage algorithm contains a critical feature that allows an exception for price and time priority and, while compulsory for all ECNs, is voluntary for all market makers who may choose to internalize order flow by simply matching the NBBO. If this rule were allowed as proposed, market makers would have no incentive to improve upon the NBBO before executing customer orders. This tendency will be further exacerbated by market makers who maintain the right to gain immediate priority at the same price on Supermontage if that price is posted by an ECN with an explicit pricing model. Market makers receive an economic incentive to quote wide spreads under the current proposal. They need only improve quotes when a competing order from an ECN or UTP exchange arrives at a better price. This provides a free option against all orders entered by access fee ECNs and UTP exchanges, which currently comprise a large majority of NASDAQ's order flow. If ECN access fees are rolled up as proposed in the amendments, spreads will widen and those participants internalizing order flow will see their margins increase due to wider spreads. As such alternative execution venues are slowly strangled, spreads on NASDAQ's Supermontage will surely widen. This outcome is supported by analysis of recent data:

Internet Stock Trading
Friday, January 8, 1999

Percent of Time

 

ECN Alone
at Inside

MM Alone
at Inside

ECN and MM
at Inside

AMZN

49.0%

27.3%

23.7%

BCST

47.2%

31.1%

21.6%

EBAY

52.6%

30.5%

17.0%

YHOO

68.6%

16.8%

14.7%

The data above shows that:

The proposed structure of Supermontage threatens to unwind the SEC's order-handling rules by pushing a significant majority of ECNs to the back of NASDAQ's priority queues despite a record of publishing the market's best prices with far greater frequency than NASDAQ market makers. Finally, Supermontage would allow market makers to trade for their own accounts ahead of the public's orders currently entered by an ECN or by another market maker.

If Access Fees are Bad for ECNs, How Will NASDAQ Charge for Supermontage?

In no section of the Supermontage proposal or its amendments does NASDAQ address the fee structure associated with the use of Supermontage. Which market participants will be paying for access to Supermontage? Will fees be charged and a portion rebated back to introducing firm depending on volume? If so, why should customers transacting business early in the year subsidize those investing later in the year at a lower implicit rate? This fee structure should be made public prior to approval of the Supermontage proposal.

Market Structure that Makes Sense for Institutional Investors

American Century believes any valid market reform should center on strict price and time priority within a market center. We believe all orders should be obligated to interact with each other within a market center. We believe that market centers should have linkages, which allow for seamless transactions between them. In essence, the creation of a virtual central limit order book with no single point of failure.

Introduced under the guise of innovation, the seventh iteration of Supermontage regresses to the period prior to the enactment of the Order Handling Rules. Rather than roll back the clock to the "good old days," NASDAQ should strive for fair investor-oriented solutions. As NASDAQ expands its global alliances, it has publicized its attempt to create a global stock market. Is Supermontage the kind of market structure to form the foundation for a linked global marketplace? Why should global markets, many with price/time priority and trade through protection, choose to partake in such an inferior structure?

We believe Supermontage is bad public policy and bad for U.S. investors. We need sound, simple market structure that levels the playing field for all investors. We need reforms that spur competition through increased transparency, not reforms that protect self-interested intermediaries at the expense of the investing public. The Commission should promote a structure that regulators, practitioners and the public can equally understand.

Sincerely,

Steven C. KleinGregory H. Bokach
Head of Global TradingSenior U.S. Equity Trader
NASDAQ Quality of Markets Committee
 
Harold S. Bradley John Wheeler
Senior Vice PresidentManager, U.S. Trading


September 1, 2000

The Honorable Phil Gramm
Chairman
Committee on Banking, Housing, and
Urban Affairs
370 Russell Senate Office Building
Washington, DC 20510

The Honorable Tom Bliley
Chairman
Committee on Commerce
United States House of Representatives
2125 Rayburn House Office Building
Washington, DC 20515

Dear Chairman Gramm and Chairman Bliley:

We noted with great interest your letter to SEC Chairman Levitt, dated August 17, 2000, concerning the NASD's SuperMontage proposal. We share your views that the "issues raised by this proposal are significant and must be considered and addressed only with the greatest of care."

Advances in securities trading technology have brought significant benefits to investors by fostering innovation in the provision of financial services and promoting full and fair competition among market participants. ECNs, ATSs and other developments have provided investors with fairer, more efficient, and cheaper execution of orders while preserving competitive incentives for innovation in trading services. The result has been better performance for the investors who have entrusted their retirement savings and college funds with us.

We generally have been supportive of the NASD's efforts to effect reforms to Nasdaq that increase its efficiency as a marketplace and enhance competition in the trading of Nasdaq stocks. We do not necessarily agree on every element of the SuperMontage proposal. Nevertheless, in the spirit of your August 17 letter, we believe that there are some important aspects of SuperMontage that need careful review to ensure that the proposal meets the test of enhancing competition and increasing efficiency in our securities markets.

For example, the SuperMontage order execution algorithm appears to discourage price competition. Unlike the strict price/time priority that applies on ECNs, and which encourages and rewards price competition, the SuperMontage algorithm puts both market maker quotes and orders entered directly into SuperMontage ahead of investor orders entered on ECNs, even in circumstances where the ECN orders are better priced. Market participants would be able to jump ahead of and free ride on investor orders placed on ECNs.

The SuperMontage algorithm also puts orders from competing exchanges last in line for execution, even worse treatment than ECNs receive. Thus, market participants would have even less incentive to enter orders on a competing exchange than on an ECN.

Moreover, the SuperMontage algorithm would enshrine internalization as a central feature of the Nasdaq market, reducing quote competition among market makers, reducing order interaction, and disadvantaging investors whose orders are internalized.

Finally, by forcing the best-priced investor orders from all ECNs into Nasdaq's ECN, SuperMontage may work to inhibit our ability to freely choose where our investors' orders will be represented

We believe that the NASD must address the issues raised in this letter before SuperMontage is approved. We appreciate your active involvement to ensure that we have the necessary time to carefully evaluate and understand the implications of SuperMontage on the US marketplace.

As you stated in your earlier letter, "with changes of this nature, if it is not done right the first time, there may not be a second chance." We agree. In this increasingly competitive global marketplace - we must get it right the first time.

Sincerely,

Steven C. Klein
Head of Global Trading

Footnotes

1 American Century Investment Management is an investment advisor managing more than $120 billion for more than 2 million investors in mutual funds, separately managed accounts and other investment vehicles.

2 See American Century Investment Management comment letter on market fragmentation on our website (www.americancenturyventures.com)

3 See attached letter from institutions to Senator Graham and Representative Bliley concerning institutional concerns about proposed Supermontage structure.

4 Discussions with various ECNs confirmed NASDAQ's technology outage caused numerous errors resulting in losses.

5 ECNs, in fact, constitute the first opportunity for buyers and sellers of stock to interact without the intermediation of a dealer as was envisioned in the 1975 Amendment to the Securities Exchange Act.