March 14, 2000

Jonathan G. Katz, Secretary
U.S. Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, D.C. 20549

Re: Regulation of Market Information Fees and Revenues

Securities Exchange Act Release No. 42208, File No. S7-28-99

Ladies and Gentlemen:

Charles Schwab & Co., Inc. (Schwab) is pleased to respond to the SEC's recent concept release on market data fees.1 Schwab is the world's largest online broker-dealer in terms of transactions and customer assets. In the aggregate, market data fees paid to the exchanges and Nasdaq cost our firm over $19 million last year. Those costs were passed on to our customers, both directly and indirectly in terms of higher commissions and fees. Because of our longstanding view that the current market data fee structure interferes with market transparency and access for individual investors, we submitted a rulemaking petition to the Commission on this subject on June 29, 1999.

We appreciate the Commission's thoughtful review of the issues raised in our petition that are critical for reforming the regulatory structure for market data. Serious reform is needed to bring market data into the Internet age. We urge the Commission to exercise its authority to adopt new rules that will (i) assure the fairness and reasonableness of market data fees, (ii) provide individual investors with the same quality and quantity of market data available to professionals, and (iii) further the National Market System goals of equal access, market transparency, and fair competition.

Last month the Commission issued a release on market fragmentation.2 The Commission identifies price transparency as the first and most essential element to preventing fragmentation. We agree. The most important step that the Commission could take to prevent market fragmentation is to make real-time streaming market data from all markets readily available at a reasonable price to retail customers. Only this step will put retail investors on a level playing field with market professionals. In our opinion this step is more important than any of other concepts proposed in the fragmentation release (on which we will comment in more detail in a separate letter).


Seven central findings in the release. Schwab strongly agrees with seven key points in the market data concept release that should form the foundation of new rules. First, state law concepts of ownership do not apply to market data, and any claim of market data ownership must give way to Exchange Act objectives and the public interest.3 Second, the Commission has broad oversight and rulemaking authority over the entire market data collection and distribution process. 4 Third, access to market data is the cornerstone of market transparency for all investors, and the lowest possible cost means the greatest transparency.5 Fourth, market data fees must be cost-based with a cap, to ensure their neutrality and non-discriminatory impact, backed by enforceable objective standards.6 Fifth, national market system plans and exclusive processors of securities information (the "Networks"), owned and operated by certain exchanges and Nasdaq (collectively, the "exchanges"), are monopolies and should be operated as public utilities for the public good and not private gain.7 Sixth, the data that the Networks provide has the same production costs, no matter how many users access it.8 Seventh, advancing technology has made the market data fee structure archaic. The Commission's release correctly notes that the revenues the Networks exact from "non-professionals" (i.e., ordinary retail investors) have "grown exponentially" by almost 1,000 percent over the last five years, simply because more investors are accessing market data in the most convenient way -- over the Internet.9

These are important principles, and the Commission is to be commended for its willingness to re-think the status quo. In the midst of so much forward progress, however, Schwab respectfully disagrees on one significant point: the concept release's proposal to allow the exchanges to subsidize their market and regulatory operations (so-called "common costs") from market data revenues. We do not believe there is any factual, policy, or legal basis for allowing profits from a monopoly business to subsidize the exchanges' competitive activities.

The Concept Release is a wake-up call to action for restructuring the market data system to promote the interests of all investors. The current market data fee structure poses a significant barrier to innovative uses of market information that would greatly benefit individual investors. Excessive and duplicative fees penalize broker-dealers and vendors that offer retail customers multiple channels for accessing real-time market data and innovative tools incorporating that data. For example, the exchanges charge Schwab "professional" fees to allow the firm and its employees to access market data to serve customers by phone or in person. On top of those fees, the four Networks together charge an additional $60 in "subscriber" fees annually per customer to allow individual investors to dynamically access the same market data directly over the Web, such as streaming quotes or portfolio valuations. Neither the Networks nor the exchanges incur any additional costs, yet an additional "non-professional" surcharge is added. Our firm is unable to absorb the extra $60 cost for the vast majority of our customer base, who on average trade only six to eight times per year at $29.95 per trade. As a result, these investors are deprived of the same quality and quantity of market data that professionals enjoy.

The fees charged today are not related to the exchanges' cost of providing market data, nor are they necessary to fund self-regulation even if that were permitted under the Exchange Act. The concept release contains convincing evidence that the exchanges do not need market data revenues to subsidize their market and self-regulatory operations. They collectively earned over $300 million in net revenues in 1998 alone.10 In contrast, the Commission's release makes clear that the only costs actually attributable to the collection and distribution of market data is the approximately $37 million in expenses reported by the Networks, which then mark-up the price of market data anywhere from 800 percent (Network A) to 2,000 percent (Network B) over their costs.11 These huge mark-ups demonstrate that the Networks are in fact monopolies and are earning monopoly profits. In an era where the exchanges compete fiercely with each other, broker-dealers, and ECNs for market share, the Commission has a responsibility to assure that market data revenues from the monopoly data providers are not used to subsidize the exchanges' competitive activities for order flow and listings. Even if it had not received a single penny of market data revenue, the supposedly not-for-profit NYSE would have made $67 million in profit in 1998 alone.12 As the release points out, for most of its history the NASD made do without any significant market data revenues.13 Market data revenues are not necessary to fund effective self-regulation.

We believe the Commission should adopt rules that vigorously enforce the Exchange Act requirements of fair and equitable fees, vigilantly guard against monopoly practices, and strengthen public confidence and accountability. The rules should:


But for the Market Data Fee Structure, Today's Technology Could Place a Wealth of Real-Time Market Data at Individual Investors' Fingertips, Dramatically Furthering the National Market System Goals of Equal Access, Transparency and Fair Competition.

The Commission's commitment to the National Market System goals of equal access, market transparency, and fair competition, together with twenty-first century technology, should guide the reforms in market data. The structures which modernized our markets in 1975 do not today achieve the goals they furthered then. Over the last decade, the Internet has revolutionized the way people access and use information. Investors no longer have to wait for the evening news or morning paper to monitor the markets. They can receive free, up-to-the minute business news through any number of Web sites or information providers. Through online investing sources, individuals are also able to access much of the same research and investing tools previously available only to professionals. Self-directed investors can use this information on their own or with the guidance of a market professional to better define and achieve their investment goals. Ready access to information has created the world's best-educated individual investors, while advances in information technology and dissemination have made our markets more efficient and fair.

Disparity in market data quality and availability is a barrier to equal access, transparency, and fair competition. In stark contrast to the positive developments brought forth by Internet technology, most individual investors today cannot efficiently and readily access the most vital market information of all: consolidated, dynamically updated real-time quotations. Although broker-dealers with an online channel offer customers the ability to request or "query" a real-time quote, typically they can only request one quote at a time for a single security after typing in that security's symbol. This essentially manual process is a barrier to accessing what could be a free-flowing data stream from the markets.

Moreover, the vast majority of individual investors can only see snapshots of the "top of the book," while the full extent of current buying and selling interest in a security that defines the current market is unrevealed.14 This is a far cry from the detailed and dynamically updated information that professional investors use and rely on, including Nasdaq Level II and the NYSE specialists' electronic book. Thanks to the Commission's Order Handling Rules, retail orders now compete with professionals' bids and offers. But retail investors cannot easily or inexpensively gain access to the same real-time data about depth and movement of the market that professionals have. The lack of retail investor access to the full range of market data available to professionals is the greatest barrier to market transparency and fair competition. That barrier is created by the current market data fee structure.

Market data fees significantly deter and penalize innovation and efficiency. The Commission's release asks the most important question: Do the current monthly non-professional subscriber fees "deter a significant number of retail investors from using real-time market information or preclude broker-dealers from providing enhanced information services to their retail customers?" The answer emphatically is "yes." But for the fees the Networks charge -- and the way they impose those fees -- Internet and wireless technology would enable dissemination of high-quality real-time data to millions of individual investors. Examples include streaming, dynamically updated quotes for multiple securities; the ability for investors to customize their own "ticker" for stocks in their portfolio; real-time securities and account valuations; and access to Nasdaq Level II and similar CTA data.

For a broker-dealer today to provide such real-time market data products and services would require capping each of its customers under the current "non-professional" monthly rate.15 But the fees the Networks impose simply make this cost-prohibitive. Here are the hard figures:

Multiplied across the industry, these costs constitute a massive tax on transparency.

Because of the fees imposed under the "non-professional" subscriber category, brokers with an electronic channel must ration the amount of real-time market information they make available to their customers. The reality is that "non-professional" fees are now largely a fiction. Retail investors generally do not pay those fees directly -- they are absorbed by broker-dealers and vendors, which must pay either the subscriber fee or the per-quote charge that is expensive and difficult to administer. Nonetheless, investors are significantly impacted by the high cost of market data and the resulting market data rationing. The non-professional fee category really just allows the exchanges to raise new revenues from Internet access. This is a tax on innovation and transparency that discriminates against electronic delivery to individual investors. The great harm to individual investors is that they are deprived of the fastest, highest quality real-time market data that otherwise should be available. The great harm to the National Market System is that the markets are less accessible and less transparent to the public.

The current fee structure results in triple billing for the same data stream. The release asks whether fees for individual investor access to market data should reflect the "superior" market data services online investors receive.17 There is no basis for the exchanges to benefit from the value-added (and paid for) by electronic vendors and broker-dealers. Yet they do benefit because the Networks effectively double and triple bill broker-dealers that offer their customers multiple channels for accessing real-time market data. For example, if a broker-dealer offers an automated telephone channel, the typical charge for market data is a fee for each telephone port. If that broker-dealer also maintains branch offices and offers telephone access to registered representatives, a second charge for market data is a fee for each desktop terminal. If the broker-dealer additionally offers a Web channel or wireless access to market data, a third charge for market data is the "non-professional" per quote or per subscriber fees.

Although the various channels essentially are fed from a single market data stream to serve the same customers, the broker-dealer is charged three different times in three different ways by each of the four Networks. The net result for Schwab is that our market data bills for 1999 were $600,000 for automated phone access, $8.3 million for desktop terminals, and $10.3 million for non-professional electronic fees, for a total of over $19.2 million.18 This multiple-channel penalty acts as a disincentive for innovation, efficiency, and convenience of investor access. The Commission's intervention is necessary to adapt the fee schedule to the changes sweeping the securities industry.

The Commission's Proposal To Cap Market Data Fees under a Cost-Based Approach Would Result in Reasonable Fees, But Only If the Fees Are Based on Actual Costs for Collecting and Disseminating Market Data.

We wholeheartedly support the Commission's proposal to create a cost-based approach to setting market data fees, with a cost limit cap. As monopolistic providers of consolidated market data, the Networks operate what in effect are public utilities. The Networks have captive suppliers of data (i.e., investors and broker-dealers) who are not paid, and consumers who have no negotiating power. The Commission's regulations require broker-dealers to report their best bids and offers (including customer limit orders) and sale prices to the exchanges. The exchanges in turn are required to consolidate that information and make it available to the public through vendors. Allowing the exchanges to capture excess market data profits to fund competitive operations is antithetical to the core purpose of our National Market System.

Cost-based fees are the only way to meet the statutory requirements of fair, reasonable, and non-discriminatory fees. To be effective, fair, and enforceable, cost-based standards must be based on the true cost of collecting and distributing market data. No less, and no more. The Networks and exchanges should recover their costs, including costs that the Networks report now,19 as well as any expenses they can show they incur specifically for maintaining, expanding, testing, and auditing their market data systems, and upgrading those systems to prevent outages and delays.20 But accountability is critical for industry and public confidence in this process. The Networks' and exchanges' accounts for market data costs must be independently verified and audited.

We recognize that decimalization requires expanded capacity, and a cost-based approach appropriately should reflect the cost of necessary system upgrades. However, retail investor access to the same quality and quantity of information that professionals enjoy will be even more important once the market is quoting in decimals. A cost-based approach should also reflect that improved technology has generally decreased the cost of data production and distribution. As more investors access market data over the Internet, the growing user base should result in lower fees, because the costs remain relatively fixed. Unless the current fee structure is changed, the growing "non-professional" subscriber base will result in ever-increasing profits for the exchanges. This is already the case. As the concept release notes, retail fees have expanded almost ten-fold, while professional fees have increased only about 1½ times over the four years from 1994 to 1998.21

It would be inappropriate and unfounded for the Commission to allow market data profits to fund "common costs." The Commission is correct when it observes that "[f]inding an appropriate basis for allocating common costs . . . is an extremely difficult task."22 A common cost approach would unavoidably and impermissibly allow cross-subsidization of the exchanges' competitive activities.

We do not believe there is any legal or policy reason why market data revenues should pay for so-called "common costs" in the amorphous categories of "market operation" and "market regulation." Exchange operation and oversight are not undertaken for the purpose of collecting and disseminating market data. "Market operation" activities are required for exchanges to stay in business as trading markets, and are not undertaken to consolidate members' bid and ask quotes and sales prices for distribution to the public. "Market regulation" activities are required for the exchanges to remain qualified to do business. They are a condition of exchange registration without which a market loses the privilege of being a "securities exchange."23 Self-regulation is essential to attracting issuers to list on the market and investors to use the trading facilities. In recent congressional testimony, Dick Grasso articulated the relationship of market oversight to the "brand name" value of the New York Stock exchange.24

Allowing market data fees to fund exchange operations would give official sanction to competitive subsidies. This is especially true because the allocation of common costs is a highly subjective and imprecise undertaking. The release suggests using "the historical experience" of the exchanges in funding between 30 and 40 percent of their operations from market data, citing only recent NASD data for support.25 Using this number as a baseline or rationale to accept the status quo would be unjustified for many reasons, including: (1) for most of its existence the NASD made do without significant market data revenues; (2) the exchanges enjoy annual profits -- $300 million in 1998 alone - that could and should fund their market operations; and (3) it is the exchanges' monopoly control of market data revenues that triggered the need for the concept release in the first place. Most importantly, allowing the exchanges to subsidize 30 to 40 percent of their competitive operations from market data revenues contravenes the Commission's holding in the 1984 Instinet case,26 prior Commission releases on clearing agencies,27 the legislative history of the Securities Act Amendments of 1975,28 and legal principles governing monopolies.29

The release asks whether the exchanges are entitled to a rate of return from market data revenues. We do not believe that the exchanges should earn an unchecked rate of return for this utility function pursuant to a congressional mandate.30 Notably, market makers and ECNs that compete with the exchanges for order flow and that provide the information that is incorporated into the consolidated stream of market data do not share in market data revenues to defray their expenses. Nevertheless, if the exchanges were allowed a rate of return, monopolies are generally restricted to only 7-10 percent net profit because they face no market risk. Clearly, an objective standard would result in much lower profits than the amounts the exchanges currently demand. The mark-ups on market data are enormous. In 1998 Network A incurred only $18 million in costs but charged $143 million in market data fees, a mark-up of almost 800 percent; Network B incurred only $5 million in costs, but charged $99 million in fees, a mark-up of 2,000 percent; while OPRA incurred only $4.8 million in costs, but charged $43.6 million in fees, a mark-up of 900 percent.31

There is no basis for funding self-regulation from market data fees. The concept release observes that the exchanges have important regulatory functions that must be funded "one way or another."32 We share the Commission's concern that changes in market structure not lead to decreased funding for self-regulation. We submit, however, that it is very important how regulation is funded to avoid monopoly subsidization of competitive activities. Market data fees are the only category of exchange fees that are derived from a monopoly operation and not subject to competitive pressures. They are also the only fees that, when reduced, have a positive effect on market regulation. Cheaper fees result in wider distribution of market data, thereby increasing investors' opportunities to detect problems in order handling and executions.

Some of the exchanges have argued for maintaining the market data status quo, warning that market data fees are essential to fund self-regulation. The concept release itself belies that claim, showing that the exchanges are in no danger of a funding shortfall that could impact self-regulation. As noted above, the exchanges collectively enjoyed $300 million in net profits in 1998. The two exchanges with the greatest self-regulatory responsibilities had the largest share of the profits: NYSE profits were $178.7 million, NASD profits were $76 million.33 Nothing in the concept release explains how the exchanges spend or invest these profits. Surely these surplus revenues are available to support the exchanges' regulatory functions. Combining the examination and disciplinary functions of the different exchanges (in a hybrid single SRO) would produce additional cost-savings that may be used to fund self-regulation. In any case, the costs of member self-regulation should be borne by exchange members directly, not transferred to individual investors seeking access to market data.

The concept release asks whether the market data rebate programs some exchanges have adopted demonstrate that data revenues exceed regulatory funding requirements.34 The rebate programs on their face show that market data revenues exceed regulatory funding requirements. In fact, these programs show that surplus market data revenues are not earmarked for regulatory purposes at all but instead impermissibly subsidize competitive efforts to attract order flow.35

The Commission Should Mandate a Cost-Based Enterprise Fee To Replace "Non-Professional" Fees that Impose Huge Burdens on Investors and Broker-Dealers and Discriminate against Electronic Methods of Distributing Market Data.

As noted above, the "non-professional" per quote and subscriber fees only apply to electronic channels and effectively serve as a tax on innovation that impedes investor access to market information. These fees result in double- and triple-billing for the same data stream. The Networks have statutory responsibilities to collect and disseminate a stream of market information. Market data fees and requirements should not vary based on the method of distribution vendors and broker-dealers competitively use to re-package and deliver market data to their customers.

Equally significant, the non-professional fees impose an enormous paperwork and contractual burden on public customers that should be weighed under the SEC's obligation to consider the impact of rules on efficiency, competition, and capital formation.36 For example, Nasdaq requires that online broker-dealers present a lengthy and confusing licensing agreement to our customers on the Web. We estimate that it takes a customer 30 minutes to read and try to understand its terms and conditions. Assuming a million of our customers make it through this "click-through" agreement, this mandated process wastes 500,000 hours of our customers' collective time. In the agreement, Nasdaq claims ownership over market data, disclaims responsibility for the market data's accuracy, and requires our customers to waive numerous rights.37 One form of the agreement even purports to give Nasdaq the right to enter customers' homes to see how they are using their market data. Not surprisingly, we have received many complaints from our customers who do not understand why they must become contractual licensees just to access real-time quotations. The agreement stands in stark contrast to the Commission's plain-English initiatives in other areas.

In addition to the burdens placed on individual investors, the non-professional fee also requires a tremendous amount of accounting, tracking, auditing, and paperwork for broker-dealers. We estimate that Schwab personnel spent approximately 4,000 hours last year performing the necessary monthly accountings and reports required by the non-professional fee, negotiating with the Networks' exchange administrators, and implementing and administering the online market data retail "subscriber" agreements. The Commission should weigh these significant burdens in reviewing the current market data structure, the same way it must review the burdens of its own rulemaking proposals under the Paperwork Reduction Act.

A cost-based enterprise fee would solve the dual problem of exponentially-growing retail investor fees and the wasteful administrative and contractual burdens of the current system. A broker-dealer or vendor could be assessed a simple monthly fee based on objective criteria that do not require vendors and broker-dealers to report all the different ways they use market data and how many quotes their customers consume. Such a fee could be based on vendor and broker-dealer size. An enterprise fee ceiling would be appropriate in recognition that market data costs do not correspondingly increase in relation to the market data a firm consumes.38

Independent Accounting and Ongoing Review of Market Data Costs Are Necessary To Assure Cost-Based Neutral Fees that Serve National Market System Goals.

The Commission should take concrete steps to regulate the Networks' monopoly powers, promote the fairness of market data plan and fee changes, and shed much needed sunlight on the market data process.

Independent audits and examinations to verify costs. For the Commission's cost-based approach to work, it is essential that the Networks' and exchanges' costs attributed to market data be subject to independent audits to verify their accuracy and reasonableness. The Commission requests comment "on whether direct funding would create an inappropriate incentive for the exchanges to increase these [market data] costs beyond reasonable levels."39 Without such audits, no mechanism would be in place to prevent this. Moreover, the current lack of audits makes it impossible to evaluate the use of, and alleged need for, market data fee revenues to cover costs. As the Commission notes, the exchanges' current financial statements "provide little information concerning the exchanges' internal cost structures."40 We therefore urge the Commission to require annual audits to be filed with the Commission and to make the audit results available to the public. To ensure the reliability of the process, audits should be performed according to Commission-approved uniform engagement procedures, a mechanism the Commission successfully adopted for reports on industry Y2K remediation efforts.41

Ongoing Commission review of market data fees and practices. To further public and industry confidence, the Commission should include the Networks' and exchanges' market data operations in regular SRO examinations. The Commission should review all of a Network's market data fees together, instead of one isolated filing at a time, in order to assure reasonableness and neutrality in the overall fees charged to vendors, broker-dealers, other professionals, and individual investors. The Networks should file their entire fee schedule for periodic review so that the Commission can assure that the fees are still reasonable and to guard against the market data structure again being outpaced by fast-moving technological and market developments.

Independent management and operation of the Networks. As the concept release recognizes, the era of for-profit exchanges means that the exchanges administering the Networks will be obligated to their equity holders to exploit every economic advantage.42 Just as one SRO can spin-off its market operations, the exchanges should be required to spin-off their market data operations to an independent administrator. The SIA's recent white paper "Reinventing Self-Regulation" proposes a hybrid single-SRO structure that could be an appropriate place for market data administration to reside.43 To further ensure that the operation of the Networks remains fair, neutral and transparent, the boards and committees that govern the Networks should include public and industry representatives.44 In the very least, Network control and administration should not rest in the hands of a particular profit-making exchange once demutualization occurs.

Prior public notice and comment on rule changes and pilot programs. The process by which the Networks file rule changes must be modified. Proposed fee changes should be accompanied by detailed financial analyses, including assessment of the costs to be covered by the fees, as well as evaluation of the impact of the fees on broker-dealers, institutions, and retail investors. The Commission should eliminate rules allowing fee changes to become "effective-upon-filing" and pilot programs to the extent they would result in increased fees for any category of market data user or establish any new category of fee, user, or subscriber.


Schwab applauds the Commission for reviewing the issue of market data for the first time in 25 years. Properly resolving the market data debate is a prerequisite for most of the market structure challenges the Commission currently faces, such as market fragmentation, market linkages, and SRO demutualization. We encourage the Commission to take this opportunity to ensure that affordable, real-time, dynamic, full-depth-of-book market data is readily available to all investors, including retail investors. Ensuring equal and affordable access to the same quality and quantity of real-time market data for all investors is the single greatest step the Commission can take to increase the transparency of our markets, increase the liquidity of those markets, and thereby improve capital formation in our markets. We believe that these goals outweigh any arguments in favor of the status quo or the continued use of market data revenue to subsidize competitive activities. The Commission's choice between these alternatives will have an enormous impact on the future competitiveness and vitality of the securities markets in the twenty-first century.

Very truly yours,

David S. Pottruck

cc: Hon. Arthur Levitt

Hon. Norman S. Johnson
Hon. Isaac C. Hunt, Jr.
Hon. Paul R. Carey
Hon. Laura S. Unger
Annette Nazareth
Robert L.D. Colby
Belinda Blaine
Daniel M. Gray


1 Exch. Act Rel. No. 42208 (Dec. 9, 1999), 64 Fed. Reg. 70613 (Dec. 17, 1999) ("Concept Release").

2 Exch. Act Rel. No. 42450 (Feb. 23, 2000), File No. SR-NYSE-99-48.

3 Id. at 70615.

4 Id. at 70614.

5 Id.

6 See, e.g., id. at 70619, 70621, 70629-30.

7 Id. at 70619 & n.48, 70621 & n.77. The four Networks are Tape A (NYSE listed stocks) of the Consolidated Tape Association (CTA), Tape B (Amex listed stocks) of the CTA, Nasdaq, and OPRA (for options).

8 Id. at 70630.

9 Id. at 70631.

10 Id. at 70615 & n.6, Tables 9-17 (income before income taxes line item).

11 Id. at Tables 5-8 (compare total revenues and operating expense line items). Because Nasdaq did not separately account for market data expenses, the figure we have used assumes Nasdaq's expenses are the average of the three other Networks. The exchanges themselves apparently do not separately account for market data expenses.

12 Id. at Table 9 (compare income before taxes and market information revenues line items).

13 Id. at 70625.

14 Some broker-dealers do offer their more active customers streaming quotes and Nasdaq Level II information, typically for an additional fee.

15 This is because reaching the per-quote ceiling before the subscriber fee kicks-in (just 200-400 quotes per month, depending on the Network) happens very quickly with dynamic real-time market information.

16 $2 for Nasdaq, and $1 each for Network A (NYSE), Network B (Amex), and OPRA.

17 Concept Release at 70632.

18 Moreover, to consolidate the four Networks' data streams to feed Web servers and desktop terminals, broker-dealers rely on market data vendors. The fees noted above do not even include the significant amounts vendors charge for aggregating the raw data streams and delivering it to broker-dealers in a consolidated reliable form.

19 Concept Release at Tables 5-8 (total expenses line item).

20 For example, under the Commission's Automation Review Policy (ARP), the exchanges are required to periodically report their system changes and upgrades. See id. at n.109 (citing ARP releases). Disclosures on ARP filings relating to quotation and trade reporting should include audited costs and be made public.

21 Id. at 70614. The release also observes that the growth in market data fees has kept pace with other exchange revenues during the same period. This fact, however, is unrelated to the question of the current fees' reasonableness. Similarly, the fact that the volume of quotations and transaction report inquiries has increased more than 150% between 1994 and 1998 is also unrelated to the fees' reasonableness compared to actual costs. Id. at 70617.

22 Id. at 70628.

23 See Exchange Act §§ 5, 6 and 15A. We think the concept release exaggerates the relationship between market surveillance and accurate market data. Securities fraud and manipulation perpetrated by false or misleading market data -- as opposed to improper trading activities, sales practices, or false and misleading statements to the market -- is negligible. Few, if any, cases over the years have involved members' knowingly publishing false market information. In any event, such cases would arise from member regulation, not market regulation.

24 See Testimony of Richard A. Grasso, Chairman and Chief Executive Officer of the NYSE on Public Ownership of the U.S. Stock Markets, Committee on Banking, Housing and Urban Affairs, U.S. Senate (Sept. 28, 1999).

25 Concept Release at 70630 & n.130.

26 In re Institutional Networks Corp., Exch. Act Rel. No. 20874 (Apr. 17, 1984), 1984 WL 52727 ("Instinet case").

27 See, e.g., Exch. Act Rel. No. 14531 (Mar. 6, 1978, 43 Fed. Reg. 10288, 10302 (Mar. 10, 1978) (rejecting the use of clearing agency fees for one service to "subsidize certain [other] services in order to encourage participation in a registered clearing agency").

28 See, e.g., S. Rep. No. 94-75 (1975) ("S. Rep."), reprinted in 1975 U.S.C.C.A.N. 179 (SEC "must assume a special oversight and regulatory role" to remove burdens on competition caused by exclusive processors and other SRO projects "which enjoy an effective monopoly, or which are merchandised to members on a basis other than cost and quality of services.").

29 For example, local telephone monopolies are prohibited from subsidizing competitive long distance operations with their local monopolies. See, e.g., Bellsouth Corp. v. FCC, 144 F.3d 58 (D.C. Cir. 1998).

30 In 1985 the NASD claimed a rate of return was not necessary for the fees it charged to information vendors on the grounds that it had no shareholders and thus no need to fund dividends. See Concept Release at n.128 (citing Instinet case).

31 Id. at Tables 5, 7 and 8. Nasdaq did not report separate market data cost information to calculate its mark-up.

32 Id. at 70628.

33 See id. at Tables 9 and 10.

34 Id. at 70633.

35 Schwab's July 1999 comment letter on Nasdaq's market data fee rebate program expands on this point. Exch. Act Rel. No. 41499, File No. SR-NASD-99-25.

36 Exchange Act § 3(f).

37 There may be sound public policy reasons to protect the Networks from non-negligent market data delivery problems and inaccuracies. But that policy should be carried out by statute or regulation, not by consumer contracts of adhesion.

38 Network A's current enterprise fee of $500,000 per month has no cost basis to support it and, therefore, does not meet Exchange Act requirements.

39 Concept Release at 70633.

40 Id. at 70634.

41 See Rules 17a-5 and 17Ad-18. See Exch. Act Rel. No. 40608, 63 Fed. Reg. 59208 (Nov. 3, 1998); Exch. Act Rel. No. 40587, 63 Fed. Reg. 58630 (Nov. 2, 1998).

42 Concept Release at 70629.

43 Securities Industry Association, "Reinventing Self-Regulation," (January 2000) available at

44 The SEC in recent years has sought to increase public representation on SRO boards. See, e.g., Report Pursuant to Section 21(a) of the Securities Exchange Act of 1934 Regarding the NASD and The Nasdaq Market (August 8, 1996); Order Granting Approval to Philadelphia Stock Exchange Proposed Rule Change, Exch. Act Rel. No. 38960 (Aug. 22, 1997) (requiring 50 percent board representation by public governors).