Hearing on "Maintaining Leadership in the Financial Marketplace of the Future"


9:00 a.m., Monday, May 8, 2000
3rd Floor Conference Center - Federal Reserve Bank of Chicago
230 South LaSalle Street, Chicago, Illinois

Good morning Chairman Gramm, Chairman Grams, and Senator Fitzgerald. I am co-founder and Chief Executive Officer of Archipelago, which is headquartered in the heart of the Loop in downtown Chicago. Today, it is a leading electronic communications network, or “ECN,” that serves a varied client base and executes over 3% of Nasdaq’s daily volume. (Further background on Archipelago can be found in Exhibit A, attached hereto.)

Additionally, as you know, Archipelago is entering into an alliance with the Pacific Stock Exchange (“PCX”) to create the first fully electronic national stock exchange for both listed and Nasdaq securities. This new exchange market will be fully integrated into the National Market System (“NMS”) and will compete toe-to-toe with the New York Stock Exchange (“NYSE”), the American Stock Exchange (“Amex”), and the Nasdaq Stock Market (“Nasdaq”). We have been actively discussing our market structure with the Securities and Exchange Commission (“SEC”) and, within weeks, will be filing our rules with them. When published for comment, we trust that you will see that Archipelago has laid the foundation for the first for-profit, technology-driven exchange that levels the playing field for all investors by combining greater transparency, faster speed and lower cost, all based on the guiding principles of integrity, credibility and fairness.

I. Intermarket Linkages Are Essential To Healthy Market Structure

Prior to starting Archipelago, I worked for a number of years in the options industry and can say that I am very encouraged to hear that options exchanges, under the direction of SEC Chairman Levitt, are moving forward on establishing linkages. And in connection with Mr. Brodsky’s testimony, we strongly agree that a system of bifurcated regulation, generally, and as to equities options and futures on equity products, specifically, results in competitive inequities and leads to regulatory arbitrage. Regulators should speak with one voice to avoid confusion; we don’t want to have regulators conflicted over what’s right.

In the Nasdaq marketplace, where Archipelago does the vast majority of its business, we have learned a thing or two about intermarket linkages. First, it has been Archipelago’s long-standing position that robust competition among market centers should be promoted by improving information and order delivery linkages. We believe that this is best accomplished by allowing different market centers to build proprietary links among themselves based on their own business judgment and experience. A system of competing linkages would redirect the monopolistic control of existing market centers, assure price competition and efficient order execution, and allow for global expansion. This is what Congress ultimately envisioned and mandated in 1975.

Second, we believe that the establishment of proprietary intermarket linkages must be accompanied by rules that mandate levels of quality, reliability, and openness. Importantly, these rules should be unambiguous in prohibiting one market center from shutting or freezing out another market center based on arbitrary or anti-competitive reasons. A dramatic example of this took place when the NYSE strong-armed OptiMark, a facility of the PCX, by compelling them to use the NYSE Super DOT linkage as the direct result of the NYSE shutting OptiMark out of ITS.

Third, these linkages should not and cannot be aggregated at one single point. The lessons of the Information Age teach that “single-point” entities expose themselves to tremendous risk of failure. Thus, the idea that intermarket linkages should be aggregated at one point, as is the case today with ITS and SelectNet, runs counter to every business tenet developed from the dawn of the Internet. If AT&T’s system were to temporarily fail, would we want that failure to affect the service of Sprint, MCI-Worldcom, Qwest, etc.? If AOL’s system were down, would we want that to bring down the systems of other Internet Sevice Providers? Of course not; and the same lesson applies to our capital markets.

A. Deficiencies of ITS and SelectNet: Both Are Single Point Systems

Given this point, Archipelago views warily the sudden expressed interest of the NYSE to tear down and rebuild ITS on its singular terms and rules. Because we have been shut out of ITS, we have had more experience with Nasdaq’s single point of failure, SelectNet, but think the lessons apply equally across markets. SelectNet, which is essentially an “email” linkage between market participants that helps us execute nearly 30% of Nasdaq’s daily volume is continually plagued with shortcomings, delays during heavy trading, and even outages. SelectNet has a history of technological shortcomings and experienced major outages during the market stress just last month. 1

ITS meanwhile, which is dominated by the NYSE, operates in an environment that is fraught with conflicts and is far slower than the sub-second increments required for new competitors, like ECNs. Moreover, without competition, it is internally driven: there is no clear mission for the system, no incentive to innovate, and limited accountability.

1. More Conflicts: SuperMontage and Nasdaq’s UTP Plan

The problems of a controlled system in our securities market are not limited to ITS and SelectNet. Other examples are the NASD’s proposed “Order Display Facility,” known also as “SuperMontage,” and the Nasdaq’s Unlisted Trading Privileges Plan (“UTP”).

Potential conflicts of interest exist when a marketplace, such as Nasdaq, both aggregates information and executes orders. If approved, SuperMontage would harm competitors by granting competing exchanges “second class” access by giving UTP exchanges priority only after market makers, and then ECNs. On its face, this proposal disrupts competition.

The governance of Nasdaq’s UTP, like ITS, requires a unanimous vote. As a single information aggregator, Nasdaq’s UTP invariably creates barriers to entry to new competitors. For example, Nasdaq has established a “lock out” period where it prevents competitors from sharing equally in information revenue for 4 1/2 years after entry into its system. Nasdaq reaps gratuitous gains from its windfall tariff on transactions executed by its competitors.

2. Single Anything: Bad Idea!

The overall point that we are striving to make is that a “single anything” eliminates choice, engenders conflicts, and produces a vehicle that will stymie innovation and competition. Thus, we are also very wary of Nasdaq’s proposal to impose its “Order Display Facility,” known as Super Montage, as the single point of information aggregation and order execution in the Nasdaq marketplace.

B. Solutions: Proprietary Linkages and Competing Aggregators

Archipelago respectfully suggests that the way around the “single point of failure” is through the establishment of proprietary linkages and good rules. For instance, Archipelago has built private high-speed links, based on the highest levels of quality, reliability, and openness, with Instinet, Island, and Redibook outside of Selectnet, and will soon establish links with additional ECNs and market makers. 2 It is Archipelago’s ordinary course of business to “talk,” if you will, during and after market hours with these ECNs through these proprietary links, but such links are particularly important when Selectnet fails or is under-performing. When that occurs and other market participants are shut out, Archipelago and its linked ECNs continue to execute orders.

1. A Better Way

Similar to the idea of choice and competition regarding linkages, Archipelago believes that if multiple information aggregators were permissible, competitive barriers would be eliminated, and that through the power of choice, the information systems employed would be constantly improved. Archipelago suggests, therefore, that a new class of market participant, one that is qualified to disseminate quote/trade information by the SEC, should be established. In this manner, market transparency would be maximized, and the barriers to entry and the problems resulting from designated monopolies would be minimized.

II. Decimals …They Won’t Be Wished Away

As you are all very familiar, the NASD’s latest statement on decimalization is that they will be ready to go no earlier than March 2001. Let me make Archipelago’s position very clear: we do not want to force something on the NASD that they perceive as reckless or dangerous. However, when we read that the SuperMontage, which will mark Nasdaq’s entry into the competitive order execution business, is scheduled for implementation around or before Nasdaq’s implementation of decimals, we wonder why their technical resources have not been prioritized to their decimalization project. On the same theme, we have read the NASD’s recent announcements about their involvement in several international deals, one of which with “iX” will create an exchange that could conceivably pose a competitive threat to U.S preeminence in world capital markets. We wonder how much of the NASD’s technical resources will be drawn down on these projects and whether the exchange entity created from its deal with iX will trade in nickel and penny increments before the Nasdaq market does in the United States.

It seems apparent that entry into a new line of business via SuperMontage and doing deals around the world have greater priority at the NASD than supporting the industry-wide commitment to decimals. In the end, because decimalization will not be implemented by the NASD for the foreseeable future, U.S. investors will lose hundreds of millions. This delay is yet another example of what can happen when a well-entrenched monopoly controls a marketplace and its would-be competitors.

III. We Have Come A Long Way But Must Be Eternally Vigilant:

Monopolistic Forces Will Not Cede Control Easily__________

The rapid development of our capital markets – of which Archipelago and other ECNs have played a critical part – and the optimism that it has spawned must be tempered by the ever-present reality that the forces of anti-competition and status quo will not yield their monopoly fortresses easily; and, through guile and influence, attempt to erect new barriers against the change agents who have begun to make inroads. How ironic that the NYSE and, even, Nasdaq – to date essential mechanisms that allow capital to be raised for the drivers of the information and technology revolution like Intel, AOL, Lucent and Cisco – themselves stand with their feet firmly planted in the past. We would respectfully suggest that the NYSE and Nasdaq could learn a thing or two by reading the business plans and prospectuses of many of the companies they list. “Time waits for no man” and, paraphrasing from Senator Charles Schumer’s Op-Ed article in last Friday’s (May 5, 2000) The New York Times, “innovation waits for no exchange, not even the NYSE.”

“30 and 5.” What do these numbers represent? They are the respective market shares of ECNs in Nasdaq and NYSE-listed trading. ECNs execute over 30% of the volume on Nasdaq and are the inside market for the vast majority of Nasdaq’s liquid stocks. Contrast that with NYSE-listed trading, where ECNS make up less than 5% of the volume. Why? Let me suggest that the reason for this huge disparity has nothing to do with the differences in liquidity characteristics between stocks like AOL that are traded on the NYSE and ones like Yahoo that are traded on Nasdaq. Rather, it has to do with the fact that the NYSE today is a monopoly and aggressively wields its monopoly power over innovators who attempt to challenge and compete with them. Case in point: just last year, the NYSE tried unsuccessfully to coax the SEC to approve a rule that prohibited NYSE specialists in many cases from using an ECN. This rule proposal was advocated in spite of empirical fact that ECNs have brought greater transparency to the Nasdaq marketplace and saved investors billions of dollars. Chairman of the NYSE Richard Grasso’s admission in The Wall Street Journal (March 1, 2000) sums it up perfectly, “[o]ne reason the NYSE hasn’t been as ambitious [in innovating]. . . as Nasdaq, is that it has lost far less volume to ECNs.”

We commend Senator Schumer for his forward-thinking and candid comments in his recent Op-Ed article. When he says, though, that “it may seem hard to believe, but for two centuries the New York Stock Exchange has operated mostly as a monopoly,” from the perspective of Archipelago, Instinet, Island and others, it is not hard to believe because we have been operating in a regulatory environment that essentially has precluded us from the listed securities business. Unfortunately, although competition has been foist upon almost every segment of our economy, even in our formerly monopolistic utility industry, to date, consumer choice has been absent from the listed market. Protectors of the monopoly may rejoin, “but the NYSE is a national treasure,” to which we respond that our strength is our national securities markets as a whole and not an individual anointed entity.

IV. Conclusion: Enhance Choices Through Intermarket Linkages

To sum up, a system of competing linkages, based on rules that mandate levels of quality, reliability, and openness, would: (1) redirect the monopolistic control of existing market centers; (2) circumvent single points of failure; (3) assure price competition and efficient order execution; and (4) allow for global expansion. Without such a system, our ability to remain the global leader in capital markets will be jeopardized.

Archipelago looks forward to continuing to work with the Committee to enhance competition in our securities market. I will be glad to respond to any questions that the members of the Committee may have at the appropriate time.


Archipelago was approved to operate as an electronic communications network (ECN) in January 1997, in conjunction with the SEC’s implementation of the Order Handling Rules. Among other things, the Order Handling Rules dramatically augmented the development of ECNs by enabling them to compete directly with other market makers for order execution flow in Nasdaq securities. Like other ECNs, Archipelago electronically matches and executes customer orders for securities transactions and submits its best bid and offer to Nasdaq for reflection throughout Nasdaq market. Unlike many others, however, Archipelago’s business model is based on its “best execution” order routing system, which externalizes orders to superior quotations outside of Archipelago itself. For example, if our best buy order is priced at $20 and another trading venue has a buy order at $20 1/8, we will route our marketable sell orders to the $20 1/8 bid. In this manner, Archipelago facilitates “best execution” for investors.

Archipelago was started by software entrepreneurs MarrGwen and Stuart Townsend and me and has grown substantially in its brief three-year history: today we execute about 60 million shares a day. Archipelago is headquartered in Chicago, but most recently added a second headquarters in lower Manhattan to solidify its presence as a truly national enterprise. Our Manhattan headquarters includes a state-of-the-art technology facility that will provide fail-safe redundancy and additional capacity to our trading systems.

In 1999, a diverse group of premier blue-chip institutions representing many market constituencies invested significantly in the company. They include Goldman Sachs, E*TRADE, J.P. Morgan and American Century, Instinet, Merrill Lynch, CNBC, and BNP/Cooper Neff. These investments validate our passionate belief that the time has come for Archipelago to graduate from an ECN to a for-profit, technology-driven exchange to bring innovation to U.S. capital markets.

1 -- The NASD proposal to move more Nasdaq trading from SelectNet to the SOES trading system (“Super SOES”), which was approved by the SEC but whose implementation has been delayed several times by the NASD, will not fix these problems either. In fact, many, including Archipelago, think that Super SOES will make these problems actually worse. Also, Super SOES is structured in a manner that is anti-competitive to ECNs and ATSs.

2 -- The Archipelago Integrated Book at www.tradearca.com offers investors the opportunity to look at the cross-populated limit order books of Archipelago, Island, Redibook, and soon other ECNs, through a single portal via the Internet.

Home | Menu | Links | Info | Chairman's Page