A Framework for the Future
December 1, 2000
TO: Joel Seligman, Chairman, SEC Advisory Committee on Market Information
FROM: Carrie Dwyer, Executive Vice President, Corporate Oversight, Charles Schwab & Co., Inc.
RE: A Framework for the Future
The need for reform of the market data system is driven by the simple change in the end user over the past 25 years; the investing public has become the American public, with more individuals represented in the market than ever before. Their interest in the markets is not dependant on whether they are self-directed Internet traders, or their money is managed on a discretionary basis by a money manager. The key point is that the day-to-day movements of the markets and individual securities affect our fiscal, social and political well-being as never before. We need to know, and market data is how we know. Today, market data and the analysis of its meaning is the most popular news in America. The source of that news is not the stock markets themselves, but the individual investment decisions of millions of people, trading around the world in a diverse group of new and old markets. Yet, despite the increased breadth of participation, all the data is still funneled to the same small group of markets that were in place 25 years ago. This group still controls the format, the speed, who can receive it, what it costs. If this were any other type of news, would we be comfortable with such absolute control? Would we want one newspaper for all of America? One editorial board? One printing plant? And yet for the key component of the news of our financial well-being for a larger and larger slice of America, that is what we have.
The first meeting of the SEC's Advisory Committee on Market Information indicated broad support for reform of the current system of distributing real-time market data. We at Schwab have long advocated more efficient and fair processing and dissemination of market data, by talking about the fundamental principals that should govern the process. We have not advocated any particular model, believing that there could be several ways to achieve the national market system goals set forth by Congress, as long as the governing principals are correct. While it may be a useful exercise for the Committee to consider alternative designs for a market data system, we think the Committee should consider first, as with any design project, what is wrong with the current design, and how to what guiding principles would produce a better design that addresses the current system's flaws. Since many of the thoughts that follow were raised around the table at our meeting, I hope that this may be a useful synthesis of many different, but similar, points of view.
A. Failure of the Status Quo
The governance structure of the current industry plans for collecting and distributing market data impedes innovation and competition.
By creating monopolies over consolidation and sale of market data, these plans engender irreconcilable conflicts between the exchanges' roles as competing markets, regulators, securities information processors, and information vendors. While approved by the SEC in response to congressional directive, the plans have become artificial impediments to the achievement of a national market system. For example:
- Only exchanges and the NASD may be participants in these facilities, granting exclusive control over vital market facilities to those with the greatest interest in maintaining the status quo.
- The governance structures of these plans vest the dominant market in each plan with substantial authority as plan administrator, allowing the administrators to thwart innovation by competitors and other market participants.
- The unanimous voting requirement gives each plan sponsor veto power, which can be used to thwart changes to the plans that threaten the competitive positions of plan sponsors.
- Plan governance is limited to plan participants and does not include the voice of other significant market participants, such as retail and institutional investors, ECNs and other broker-dealers, data vendors and the financial media.
Under the current structure, market data is only made available in a few limited, pre-set formats.
As both consolidators and vendors of market data, plan sponsors are in the exclusive position to determine the formats in which market data is sold, thus preventing private entities from developing new and more useful formats for data dissemination. This has become a critical customer protection issue with the advent of decimals: the limited formats currently made available are now wholly inadequate given dramatically increasing quote updates and reduced depth over a greater number of price increments. The plan sponsors have been slow to react, and even the proposals presented to date (Nasdaq's SuperMontage and the NYSE's look at the book) are only a starting point for addressing this problem.
As monopolists, plan sponsors have no incentive to upgrade capacity.
Under the current system, there are no market incentives compelling plan sponsors to upgrade capacity to ensure accurate and timely data dissemination. This too has become a critical customer protection issue with the advent of decimals and the continuing expansion of listings and options series. Stale quotes and other capacity related problems have plagued the Nasdaq market for several years now, and are an increasing problem in the options markets.
The plans impose an onerous administrative burden on quote vendors and broker-dealers.
The plans' current pricing model requires vendors and broker-dealers to expend enormous resources on counting customers' quote requests and accounting for the types of end users. Another unwieldy administrative burden is the extensive and confusing click-through agreement that each plan requires retail customers to review and agree to before accessing quotes over the web.
The current structure effectively requires a reluctant SEC to engage in rate-making.
The exchanges' incentive to maximize income for the plans (and consequently their own earnings) is at odds with the NMS goals and investor interests that these plans were created to serve (namely, wider dissemination of market data to improve execution quality and enhance market competition). To protect against this inevitable conflict, Congress interposed the SEC as a de facto rate-maker, a statutory responsibility it nonetheless has assiduously avoided.
The current structure results in pricing inequities.
The lack of broad industry and investor representation in plan governance, and the discretion of plan administrators to strike side deals with certain customers, or institute long-term pilots without public notice and comment that favor certain users, result in an inequitable allocation of prices. For example, fees are much higher for data delivered over the Internet by online brokers than for the same information delivered over the phone by traditional brokers.
Market data has become a competitive weapon.
Plan participants use their monopoly over market data to advance their own competitive interests. For example, several markets use market-data revenues to pay for order flow. Nasdaq has proposed paying a portion of market-data revenue to users of its PRIMEX trading system, giving PRIMEX a competitive advantage over competing systems precluded from plan participation. Market data revenue also subsidizes the exchanges' marketing programs, and the NYSE and Nasdaq have indicated they will use their privileged access to real-time market data to develop marketing features for their web sites.
Plan administrators wield their authority under the plans to maximum competitive advantage.
Plan administrators have enormous discretion over who is allowed to "license" market data and under what terms. Each market data agreement is customized, with vendors and broker-dealers required to disclose in an exhibit how, where, and to whom they plan to distribute the data. Plan administrators have the authority to accept, limit or reject uses of the data. Because the plan administrators compete with other market participants (data vendors, ECNs and other broker-dealers), the agreement process provides them with competitive advantages in terms of information and the ability to control future uses of market data. Plan administrators also wield their authority against other plan participants. There are many examples of plan administrators blocking development of competing products or acting unilaterally to impose anti-competitive terms on other plan participants.
Investors are forced to buy back their own information at a mark-up not subject to competition or rules.
Markets are created by investors, but the current system forces investors to pay for information about the market they have created with their orders and trades. Regulation has created a profit center at investors' expense by requiring ECNs and other broker-dealers to collect investors' price and sales information at their own expense and turn it over to the plans for free. Investors must then repurchase this information from the plans to execute their trades and monitor their investments, and their brokers are forced to repurchase this information to comply with various confirmation and reporting rules.
B. Framework for the Future
In light of the failures identified above, we recommend an alternative model that incorporates the following principles and goals:
A Level Playing Field
Regulation should not prevent the retail investor from accessing critical market information on the same terms as other market participants. Denying retail investors equivalent access to the same real-time market information available to the professionals only reinforces the professional's advantage.
No model will eliminate the need for SEC oversight, but to the maximum extent possible, the new structure should incorporate market incentives to drive innovation and operational efficiency, provide market discipline over pricing, and guard against anti-competitive practices.
To prevent undue influence and the potential for anti-competitive activity by plan administrators, consolidation of market data should be separated from the markets. Plan participation would still be required of markets trading a security, but one or more independent processors would be responsible for consolidating data from the participant markets, licensing data and establishing fees, and allocating revenue on the basis of pro rata market share.
To ensure fair governance, independent processors should be governed by balanced boards that include participant markets and as well as other significant market constituencies: retail and institutional investors, alternative markets, broker-dealers, market data vendors and the financial media should all have a voice in plan governance, and deliberations should be public. As has been the experience with the SROs, public participation would infuse governance of the information processors with a greater sense of objectivity and impartiality. In any event, governance should be by majority vote, thus limiting opportunities for competitors to block innovations on other markets.
Multiple, Competing Quote Products
To introduce market discipline and foster innovation, the SEC should provide markets and individual market participants with the opportunity to disseminate unconsolidated market data (including multiple levels of price and depth information) independently through third-party vendors. This structure would distinguish between the inside NBBO quote and richer quote formats reflecting depth at multiple price levels. Plan participation would only be required insofar as the consolidated NBBO quote, leaving the markets and individual market participants free to license unconsolidated, albeit richer market data independently to third-party vendors. Introducing opportunities for competing quote products has the potential to transform the dissemination of market data. We are already seeing the availability of unconsolidated market data for Nasdaq stocks, and although this data does not offer a complete picture of the best prices available, in important respects (timeliness and depth) it exceeds the consolidated data feed and better serves the needs of certain types of customers. The existence of multiple, competing quote streams could impose competitive pressure on plan sponsors as well as individual market centers to ensure that consolidated data is readily accessible in a useful and cost effective format.
Because customer and ECN orders comprise a significant portion of the market data disseminated by exchanges, ECNs and other broker-dealers with significant customer orders should be eligible to participate directly as members of the independent processor. Direct participation would provide an incentive for ECNs and broker-dealers to make order information more widely available, thus increasing transparency, and would limit the ability of the current plan participants to use market data to gain competitive advantage over non-participants.
Cost-Based Pricing of Regulatory Information
Broker-dealers and ECNs, who are required by the SEC to report pricing and sales information to an SRO at their own expense, should not have to pay fees above cost for this same information when it is consolidated. Information that firms are required by regulation to submit should not be treated as a profit center, especially in the absence of any price competition.
An Objective Fee Structure
To avoid onerous record-keeping burdens and inequitable distinctions over how quotes are delivered to investors, pricing should be set at the enterprise level without distinctions between "retail" or "professional" use. Enterprise fees should be reasonably related to the costs of consolidation and should be equitably allocated based on objective factors such as trading volumes or number of customers (regardless of whether they be professional or retail). To ensure flexibility for vendors to distribute market data in the most useful formats possible, data should be licensed to vendors in raw form rather than pre-set formats such as today's one-dimensional Level I quote streams.
Regulatory Oversight of Fees
Although imposing market discipline should mitigate pricing inequities, close scrutiny of fees will still be required to the extent that distribution of market data remains subject to monopoly control. Scrutiny is especially critical where data must be purchased to satisfy regulatory requirements. (For example, with the dwindling utility of the NBBO in light of decimal increments, firms are already having to license richer quote formats to ensure customers are adequately informed before they trade, and the NBBO quote increasingly is relevant only to a handful of SEC rules on execution quality benchmarking.) Objective oversight of fees will require the SEC to establish precise guidelines for determining costs, and to review fee changes on a prospective basis. The SEC's current approach (post-effective reviews based on flexible guidelines) is susceptible to cursory rubber stamping. The only way to ensure observance of the statutory principles of fair, non-discriminatory and equitably allocated fees is to set fees in accordance with those principles from the outset. The SEC should establish precise guidelines consistent with those principles, to be implemented by the balanced boards of the independent processors.