U.S. Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, D.C. 20549
Attn: Jonathan G. Katz, Secretary
Re: Securities Exchange Act Release No. 42208, File No. S7-28-99
Ladies and Gentlemen:
As stated in our earlier comment letter, Charles Schwab & Co., Inc. ("Schwab") believes that the Commission's concept release on market data fees is a wake-up call to action for restructuring the market data system to promote the interests of all investors.1 The concept release has contributed greatly to the ongoing dialogue and debate regarding the future of our markets. What the Commission now chooses to do will substantially impact the structure of our market system in the decades to come. Because of this, we are taking the unusual step of filing this reply comment letter.
We could not agree more with the observations made by the New York Stock Exchange ("NYSE") in its initial comment letter that market data is a "critical subject" because securities prices are "important to millions of [investors] around the world."2 The NYSE is also right on target when it observes that individual investors are rapidly shifting from the telephone and registered representative channels for receiving market data to electronic channels and now demand the same analytical tools as the professional. It also follows, that it is "essential" for individual investors to have access to "a stream of electronically-delivered dynamically-updated data."3
But what conclusions should be drawn from these profound technological changes and shifts in investor expectations and behavior? Schwab believes that, in order to make the markets as transparent and fair as possible consistent with Securities Exchange Act § 11A, market data should be treated as a public good and all investors should have access to the same quality and quantity of market data at the lowest possible cost. In contrast, one economic theory holds that market data is like any other good, and because there is inelastic demand for it the market data networks ("Networks") and the exchanges are entitled to charge a "higher price compared with a service with more elastic demand - such as execution services."4 That school of economic analysis exemplifies the dangers of an unregulated market data monopoly.
The views and information provided by commenters in response to the concept release should form the basis for further Commission action. Accordingly, we felt that it was important to respond, in particular, to six "myths" that appear in some of the comment letters:
The Myth that the Call for Market Data Reform Is a Minority View.
Some commenters have asserted that a small minority of online brokers is responsible for the debate over market-data fees, and that these brokers are trying to evade their responsibility to pay for market infrastructure and services. This view reveals a deep misunderstanding of the state of the markets today. Online brokerage firms number well over one hundred5 and represent a substantial portion of the investing public.6 Online brokerage is the fastest growing segment of the brokerage industry, and most of the largest "traditional" brokerage firms now have an online channel.
Moreover, the Commission has received comments from representatives across a broad spectrum of the public and industry applauding the concept release and urging the Commission to take action to reform the current system. These commenters include full service financial firms,7 global market data vendors,8 market makers,9 economists,10 and
individual investors.11 The comments of the Securities Industry Association also make clear that the online brokers are not alone in the industry in their support of cost-based fees, greater representation in the fee-setting process, and improved disclosure of plan operations, including costs and revenues.12
The Myth of Equal Access and "Free, Unlimited" Market Data.
The NYSE's comment letter states that "100 million people ha[ve] on-line access to free real-time Network A data on public web sites"; "70 million cable television subscribers ha[ve] free access on their television screens"; and "high-quality market data [is] available on a very reliable basis to retail investors at zero cost."13 We do not know of any Network or exchange program that offers free real-time market data for online channels of investor access. As the concept release makes clear, the CTA and other Networks charge Schwab and other broker-dealers tens of millions of dollars a year to disseminate real-time data to our customers. Delayed quotes, however, are free and ubiquitous in the online world. But delayed quotes are simply not reliable or useful in today's fast-moving markets for making important investment decisions. This is why the exchanges give it away for free and why professionals do not use it.
Cable television screen real-time tickers may be entertaining, but market data flash scrolling across a television screen is not useful to investors who want to track particular stock prices and then act on the most up-to-date information. The ever-present market data on cable is not "access" to the data in the sense required by Section 11A of the Exchange Act. Only dynamically updated streaming data that the investor herself can control by specifying her "watch list" or portfolio of securities can level the playing field with professionals. Only access to Nasdaq Level 2 data and the CTA participants' own electronic books can make our markets transparent to all investors. But equal access to the same quality of market data for all investors is cost-prohibitive under the current "non-professional" licensing scheme and the pro-level fees charged for market data below the top of the book. If the commission proceeds with the "depth of book" limit order transparency, it is critically important that this information be available on a cost- effective basis to all investors, not merely to a few institutional investors.
Nasdaq's May 22, 2000 announcement that it plans to lower the monthly charge for non-professional access to Nasdaq Level 2 data from $50 to $10 per month is a very positive step, although the new fee would still be prohibitive for many individual investors. It does stand in sharp contrast to the CTA's inability to grant retail investor access to any of its members' electronic book information.14
The NYSE's comment letter asserts that real-time market data is "free" to retail investors because the costs are absorbed by most firms.15 Fees do not simply vanish because brokerage firms absorb them. Excessive market data fees affect our ability to maintain low commission schedules and high-quality services. In response to the Commission's attention to the market data issue, the Networks have all lowered their monthly non-professional subscriber fee caps, and two of the Networks have instituted enterprise caps. We applaud these developments. The current fees, however, still impose a significant barrier to providing investors with the best products and services technologically possible, such as dynamically updated portfolios.16 Charging "non-professional" fees and precluding all but the wealthiest or most active individuals from receiving Nasdaq Level 2 (and similar CTA data) prohibits firms from competing with one another based on innovative investor tools that incorporate streaming data. The result is a tremendous opportunity cost and burden on competition.17
The Myth of "Rate Making" Perils and Hazards.
We agree with the NYSE that an excessively detailed rate making approach to market data fee setting could be cumbersome and overly complicated. The 24-step rate making approach exhaustively detailed in footnote 11 of the NYSE's letter seems rather intimidating. Fortunately, we do not read the concept release as advocating such a rate making approach, and the Commission has other options at its disposal. For example, the Commission could take a three-step approach that, while requiring serious thought and work, would be relatively streamlined: (1) determine what costs the exchanges and their exclusive securities information processors may recover consistent with national market system requirements and goals; (2) based on independently audited cost data provided by the exchanges on an annual basis, calculate an appropriate market-wide fee recovery ceiling; (3) establish standards for how the exchanges may assess fees against categories of enterprises, such as broker-dealers and market data vendors who distribute the data to end-users.
A number of other arguments have been raised as to why a cost-based approach to market data fees will not work. For example, the NYSE says it "would have to" raise its listing and transaction fees if the Commission adopts a cost-basis approach to market data fees. But we see nothing in the NYSE's submission to support that claimed need. In fact, as our March 14 comment letter points out, the concept release documents that the NYSE has sufficient net revenues to pay for critical functions such as self-regulation even if market data were disseminated at cost.
Sensitive to any unintended spillover effects of a cost-based approach, the NYSE wonders whether it might lead to "cost padding" or even "outright waste and mismanagement" in the absence of any "incentives to reduce costs through operating or administrative efficiencies."18 We believe that the personnel at our registered exchanges have more integrity than that. The NYSE's comments do point to the need for independent and active oversight and audit of the market data process, especially once cost justification becomes part of the reasonableness standard.
Some have argued that, because Congress did not expressly mandate a rate making process, the Commission would exceed its authority to subject market data fees to a cost-justification process. That argument is simply unsupportable. Perhaps the most fundamental feature of Section 11A is that it confers considerable latitude to the Commission to regulate exclusive processors and to promote national market system principles as the markets evolve. The NYSE's comment letter identifies this flexible authority by pointing to the legislative history:
Any exclusive processor [of market data] is, in effect, a public utility, and thus it must function in a manner which is absolutely neutral . . . . [S]erious antitrust questions would be posed if access to this facility and its services were not available on reasonable and nondiscriminatory terms to all in the trade or if its charges were not reasonable. Therefore, in order to foster efficient market development and operation and to provide a first line of defense against anti-competitive practices, Sections 11A(b) and (c)(1) would grant the SEC broad powers over any exclusive processor and to assure the processor's neutrality and the reasonableness of its charges in practice as well as concept (emphasis added).19
The Myth of Fair Market Data Cost Allocation
The NYSE believes that the CTA already "fairly allocates" costs.20 We find this curious since the CTA members' position appears to be that market data fees are not, and should not be, cost based in the first place. More to the point, without full disclosure of the Networks and exchanges' costs, it is not possible to analyze whether there is a "fair allocation."
It is true that Network A "imposes the same market data fees on broker-dealers regardless of their SRO membership" and geographic location.21 But this proves little. The issue is whether the fee structure discriminates based on modes of investor access and broker-dealer business models. The NYSE letter candidly admits that "disparities" exist, but argues that this is permissible because the exchange boards approved the difference in treatment "as reasonable and appropriate."22 But on what basis did the boards determine that the fees were reasonable and appropriate? Under the current system, there is no way to review such claims because there is no objective standard against which to measure reasonableness. Similarly, on what basis was the current Network A enterprise fee cap approved? Although the enterprise fee has substantially benefited a few of the largest traditional firms by capping their professional-driven fees, that only means that the fees assessed against online investors now subsidize the traditional business model.
The NYSE letter mistakenly views Schwab and other online broker-dealers as "retreating" from past positions.23 We think the record speaks for itself that we have consistently asserted that the current market data fee structure discriminates against online investors and broker-dealers who must pay for each quote investors access in contrast to other channels. In our March 14 comment letter -- in response to the concept release's specific questions -- we also discussed in more detail the negative effects this fee system has on market data rationing, market access and transparency, and innovation. We also pointed out that broker-dealers who offer their customers multiple channels for receiving the same data stream are charged multiple times under the exchanges' "professional" and "non-professional" fee scheme.
The NYSE suggests that the Networks are justified in allocating additional or different fees for online retail investor access, because "the quality, nature and use of the data are different."24 But data usability and quality is the direct result of value-added services provided by the online firms and their vendors. The Networks should not be compensated for the value others provide. Instead, their compensation should reflect the cost in providing the data stream.
The Myth of a Constituent-Driven Process
In an ideal world, the boards of the participant SROs would make the pricing decisions, "where the customers and the representatives of the public reach consensus on pricing and other terms."25 Unfortunately, this does not seem to be the case, or at least the current process is non-transparent to us. Schwab likely pays one of the highest market data bills in the industry, representing millions of retail customer quotations, and we have never been invited to take part in any significant multilateral discussion on market data fees at the exchange level. Our only discussions have been bilateral with the Network administrators' staff as we have struggled to understand and implement the arcane pricing models the Networks have developed and deployed over the years. Market participants and the investing public must be adequately represented in the decision-making process for setting market data fees to instill public confidence in the system.
The exchanges' vested interest in market data profits clearly put them at odds with a significant percentage of their own members. This is another example of why profit-driven markets should not be in the business of regulating brokerage firms due to the unavoidable conflicts of interest.
The Myth that Disbanding the CTA Will Cure All Market Data Ills.
The NYSE has proposed a resolution to the market data fee debate - to withdraw from the CTA and distribute its data directly to information vendors. We agree that the distribution of market data would be improved if the exchanges and others would compete based on service, quality, and price. Nonetheless, under the NYSE's proposal, it would still have a monopoly over its members' pricing and trade data. So if the NYSE were to directly distribute its data to independent vendors rather than through the CTA, the NYSE would still be acting as an exclusive securities information processor subject to the statutory requirement to provide its data on fair, reasonable, and non-discriminatory terms. The NYSE would also still be subject to the principle well-stated in the concept release that "proprietary interests in [market data] are subordinated to the Exchange Act's objectives for a national market system." In short, allowing an exchange to sell its market data directly would not affect at all the Commission's responsibility to ensure that the exchange's fees for market data are fair, reasonable and non-discriminatory.
We agree that a competitive model for the consolidation and distribution of market data at its source should be explored as a long-term solution. In the meantime, however, the Commission should not delay in addressing the problems identified in the concept release and comment letters.
The current rationing of streaming and depth-of-book market data is the single greatest obstacle to market transparency. This rationing is caused by the current pricing and delivery structure controlled by the exchanges, which denies individual investors access to the same quantity and quality of real-time, market data available to professionals. The most important step the Commission could take to combat any negative effects from "market fragmentation" would be to ensure that all investors have inexpensive real-time, streaming access to all prices available for a particular security in all markets where that security trades. It is only through market data reform and a cost-based approach to market data fees that the Commission's goal of market transparency in the Internet age will be attained.
Very truly yours,
W. Hardy Callcott
[Senior Vice President & General Counsel]
cc: Hon. Arthur Levitt
Hon. Isaac C. Hunt, Jr.
Hon. Paul R. Carey
Hon. Laura S. Unger
Robert L.D. Colby
Daniel M. Gray
|1||Letter from David S. Pottruck, March 14, 2000 ("March 14 comment letter").|
|2||Letter from James E. Buck, April 10, 2000 ("NYSE comment letter").|
|3||Id. at 1, 17.|
|4||Id. at 9.|
|5||Securities and Exchange Commissioner Laura S. Unger, "On-Line Brokerage: Keeping Apace of Cyberspace," (November 22, 1999), Appendix 1 (identifying 157 online brokerage firms).|
|6||Id. at 1 (noting that there were 9.7 million online accounts in the second quarter of 1999, 15.9% of all equity trades in the first quarter of 1999 were made online, and online brokerage assets are expected to grow sevenfold from 1998 through 2003).|
|7||See, e.g., Letter from Eric D. Roiter, Fidelity Investments, April 12, 2000 (advocating competition instead of exclusive processors, transparency of market data collection costs, and restructuring fees to cover only direct costs and to remove "royalties" assessed against retail investors).|
|8||See, e.g., Letter from Lou Eccleston, Bloomberg L.P., April 11, 2000 (market data fees should be strictly limited to the direct costs incurred by the SROs in order to curb exclusive processors' monopoly power and increase availability of market data to improve price transparency and promote investor protection).|
|9||See, e.g., Letter from Michael T. Dorsey, Knight/Trimark Group, Inc., April 18, 2000 (Commission should adopt rules that prevent market data revenues from being used for competitive purposes and that create a framework to provide the most market information to retail investors as cheaply as possible to improve transparency and promote competition).|
|10||See, e.g., Letter from Gene L. Finn (former chief economist for the SEC and NASD) (current market data fees discriminate against online investors and restrain competition).|
|11||See, e.g., Letter from H. Pim Goodbody, Jr., March 28, 2000 (market data should be free; "[w]hat other product or service industry makes you pay to find out what its products cost?"); Letter from Dan Jamieson, Feb. 28, 2000 (market data should be treated as a public good, and economic analysis shows that the cost of supplying market data is much lower than the current rates charged).|
|12||See Letter from Marc E. Lackritz, President, The Securities Industry Association, April 11, 2000.|
|13||NYSE comment letter at 11, 13.|
|14||In the NYSE's second comment letter responding to the concept release, it states that "Network A does offer such a service, and at no incremental cost." Letter from James E. Buck, May 24, 2000 at 2 ("second NYSE comment letter"). We don't know of any Network A depth-of-book market data that is made directly available to retail investors.|
|15||NYSE comment letter at 15.|
|16||In our initial comment letter we stated that, based on the then current monthly fee caps applied to each retail investor, it would cost almost $200 million per year for Schwab to offer streaming quotes to all of our 3.3 million online account holders. The second NYSE comment letter is correct that, given recent changes in market data pricing, that number is now inaccurate. Taking into account Network A's enterprise cap of $500,000 per month, Network B's recently instituted enterprise cap of $500,000 per month, Nasdaq's recently reduced monthly retail subscriber fee of $1 per month, and OPRA's monthly retail subscriber fee of $1 per month, the adjusted cost would now be $91.2 million per year. That is still almost five times higher than the $19 million Schwab paid last year to offer quote-by-quote or "static" market data. Although Schwab advocates enterprise caps for all of the Networks in lieu of the current "non-professional" and "professional" fee categories, the enterprise caps must be reasonably related to the Networks' and exchanges' actual costs for providing the market data.|
|17||The NYSE "see[s] no evidence that [online firms] passed through the savings" of recent fee decreases to customers. NYSE comment letter at n.27. Schwab's revenue per $100 of customer assets has steadily declined from 119 basis points in 1993 to 67 basis points in 1999; its average commission per revenue trade dropped from $76.75 to $45.55 in the same period. While we cannot speak for other firms, Schwab has aggressively passed on the cost-savings associated with the rise of online investing to our clients.|
|18||NYSE comment letter at 9, 10.|
|19||Appendix B to the NYSE comment letter, "Legislative History and Subsequent Administration of Securities Exchange Act Section 11A," at 16 (quoting S. Rep. No. 94-75, 11-12 (1975), reprinted in 1975 U.S.C.C.A.N. 179, 189-190).|
|20||NYSE comment letter at 15.|
|21||Id. at 12.|
|23||Id. at 16.|
|24||Id. at 17.|
|25||Id. at 14.|