SEC Advisory Committee on Market Information Subcommittee on Alternative Models Summary Minutes
April 16, 2001 Meeting
The Subcommittee on Alternative Models ("Subcommittee") of the SEC Advisory Committee on Market Information ("Advisory Committee") held its second meeting on April 16, 2001. Professor Donald Langevoort, Chairman of the Subcommittee ("Chairman"), began the meeting by noting that four submissions had been made (from Datek, Nasdaq, the NYSE, and Schwab) in response to his request, at the first Subcommittee meeting, for descriptions of how a competing consolidators model might address various concerns that had been identified. The Chairman suggested that, at the second meeting, the Subcommittee discuss those submissions, and continue its review of the issues surrounding alternative models. The goal of the second meeting would be to identify issues, and isolate areas of agreement and disagreement, so that they could be presented, in agenda form, for deliberation by the full Advisory Committee at its May 14 meeting.
The Chairman first confirmed one of the conclusions from the first meeting - that Subcommittee members saw no technological barriers to a competing consolidators model. He noted that the FISD was surveying market data vendors on concerns they might have with a competing consolidators model - including technological concerns - and that the results of that survey should be available to the Subcommittee in the next week or two.
The Chairman then turned to the appropriate role of the SEC in overseeing competing consolidators, and asked that the group "firm up" its tentative conclusions on this issue from the first meeting. He noted that Nasdaq, in its submission, recommended that the SROs - and not the SEC - set minimum technological and data quality standards for the competing consolidators. The SROs would consult with an "advisory committee" of consolidators in developing these standards. If a consolidator failed to comply with the minimum standards, the SROs could cease providing it with market data.
Mr. Ketchum explained that Nasdaq felt the SROs, which represent members with a direct and significant interest in the quality of market data, are in a better position to craft technical standards than the SEC. The SEC would have an oversight role, through the rulefiling process, and may wish to conduct limited inspections of consolidators. Mr. Roiter noted that, if an SRO offered its data in a technological format that favored a "legacy system" developed by the existing exclusive consolidator, barriers to entry could be created. Perhaps the advisory committee of potential consolidators could work with the SROs on technological issues to address this concern. Mr. Bernard (NYSE) had no objection to this structure, but questioned whether competing consolidators would in fact emerge.
Mr. Ketchum suggested that multiple consolidators would increase the potential for technological risk. Mr. Haley (NYSE) disagreed, noting that multiple vendors exist today, and they merely would be shifting to a different position in the pyramid in a competing consolidators model. Mr. Quick cautioned that these issues impact the integrity of the marketplace, and we should be reluctant to create a situation where consolidators could fail.
The Chairman then asked whether members could agree that, while the technological risks of a competing consolidators model were marginally greater than the current model, they were manageable. Ms. Dwyer was unclear there was consensus technological risks would increase at all. Mr. Ketchum explained that Nasdaq's proposal that the SROs set minimum standards was not intended to control or interfere with the business of the competing consolidators, but merely to provide some regularity to the process. Ms. Dwyer questioned the need for additional standards for the competing consolidators - so long as the SEC enforces the non-discrimination provisions, market forces likely would be sufficient to control quality. Ms. Friedman (Nasdaq) countered that there might be a need for objective standards (e.g., formats, sequencing, performance), particularly in the short-term when market forces were less likely to be effective. Ms. Dwyer responded that, while there might be some role for standard-setting with respect to the competing consolidators, we should proceed very cautiously in this area.
Mr. Roiter believed that, while standard-setting might be necessary at the point of entry, he was uncertain how well it would work on an ongoing basis. Perhaps if the SEC were to institute an expedited process for "denial of access" proceedings in this context, ongoing standards might be workable, because there would be an efficient mechanism to check possible discriminatory disciplinary action by the SROs. An alternative to "pulling the plug" on an underperforming consolidator might be to require it to post a notice ("red flag") of its quality problems. Ms. Dwyer noted that a regulatory bureaucracy still would be needed to oversee the notice-posting process. Mr. Mohr (Datek) believed that, if the "red flag" notice regime were adopted, oversight would more appropriately be with the SEC than with the SROs.
Mr. Bernard suggested that the concerns about competing consolidators might be overstated. The process of "picking up the pipes" by the consolidators likely would create little additional risk. The real concern was that the consolidation function would no longer be performed jointly by the SROs, but by multiple consolidators. Ms. Dwyer cautioned that it would be a very large task to sort through all of the quality control issues with market data and establish appropriate standards. Ms. Friedman said that Nasdaq currently is compiling minimum standards for a consolidator in connection with its development of an RFP for the new independent SIP under the UTP Plan. Ms. Dwyer expressed concern at the possibility of the SROs acting as "gatekeepers" and setting standards for entry by competing consolidators - she would prefer that any entry standards be limited to non-discrimination, and nothing more. Mr. Ketchum countered that there was a role for the SROs to set standards on how the competing consolidators connect to them, and analogized to the right of any supplier to set connection standards for the receipt of its product. Mr. Bernard was of the opinion that quality control matters would largely take care of themselves, without active regulatory intervention - regulators merely should play a "backstop" role in the event market forces fail. Mr. Roiter suggested a process-oriented approach, whereby the SROs might come to the SEC with proposed standards that would then be published for public comment - this would put the SEC in more of a "referee" role, and avoid having it wade into technical details.
The Chairman concluded the discussion of appropriate standards for competing consolidators by noting that the Subcommittee's report would indicate that there was a divergence of views on this issue - some believe there should be a greater specification of technical standards and others do not.
Turning from technological to economic/pricing concerns, the Chairman asked about the minimum additional level, if any, of regulatory intervention (beyond the existing "fair and reasonable" and "non-discriminatory" standards), that would be necessary to address the economic risks of moving to a competing consolidators model. Mr. O'Kelly suggested that the discussion begin with the merits of retaining the Display Rule, since the resolution of that issue would impact all of the others. Some members started expressing views on the Display Rule. Mr. Bernard said the NYSE's position on the Display Rule was "agnostic," although they had some concern with the power the rule confers on small markets. Mr. Roiter thought the suggestion to limit the operation of the Display Rule to discrete points in time (e.g., immediately prior to an order entry decision), might be difficult to implement in practice. Mr. Ketchum noted that a positive effect from that limitation would be the flexibility it would introduce into the system, since most data requests are for the purpose of pricing portfolios rather than making investment decisions. Ms. Dwyer said Schwab favored elimination of the Display Rule.
The Chairman, however, preferred to begin the discussion with the assumption that the Display Rule would be retained, and then move to scenarios where it would be modified or eliminated. Mr. Feuer (Reuters) suggested that, if the Display Rule were retained, there would be few economic risks, but few economic benefits, from moving to a competing consolidators model. (Their preference would be to modify the Display Rule.) Ms. Dwyer argued that, if the Display Rule were to be retained, enormous market power would be conferred on each of the SROs in a competing consolidators model. There would need to be some fundamental checks on this market power, such as "most-favored nation" pricing or strict non-discrimination standards. But Ms. Dwyer expressed less concern about the pricing power of the dominant markets (which would still exist in a world without a Display Rule), because of other checks in the marketplace on this power. If the dominant markets priced their data too high, alternatives would arise, such as derivative pricing (e.g., synthetic quotes) and new market entrants, and there would be objections from members, listed companies, and other constituents. She believed the dominant markets would have good business reasons to put out high-quality data at reasonable prices.
Mr. Bernard stated that the NYSE probably would continue to fund approximately 17% of its costs with market data revenues in a competing consolidators model. But he noted that today the regional exchanges derive a much higher percentage of their revenues from market data, and he would expect individualized negotiations between the regionals and the competing consolidators to be more difficult. The existing cross-subsidies and rebate programs could very well be reduced or eliminated, and incentives might be created for the regionals to quote more competitively. Mr. O'Kelly did not disagree, but felt the SEC ultimately would need to decide what constitutes a "reasonable" fee - fee disputes inevitably will arise and the SEC will have to resolve them. Ms. Dwyer questioned the assumption that there would be a reduction in the regionals' data revenues with a competing consolidators model - to effectively operate "smart" order routing systems, users may very well want data from all of the regional exchanges.
Mr. Smith (Datek) believed market forces generally would work to solve the economic issues. They would love to see the NYSE raise its data fees too high because it would create a competitive opportunity for Island. All that is needed from a regulatory perspective is a non-discrimination standard. Mr. Bernard pointed out that the SROs need the flexibility to make "reasonable" discriminations among users' business models in pricing data. Mr. Ketchum believed a "most-favored nation" standard is appropriate for pricing data for comparable users, if this allows taking into account how the data is used. Also, he generally agreed with the NYSE that pricing abuses are effectively checked by the membership of the SROs, although backstop SEC oversight is necessary for those occasional times when this model fails. (On the merits of the Display Rule, while there are arguments either way, on balance Mr. Ketchum believed it should be retained in some form.)
Mr. Roiter argued that "most-favored nation" pricing currently is required by the Exchange Act's non-discrimination standard - the real issue is determining whether users are similarly-situated for the purposes of this provision. He also posited a market data model where revenues would be derived from advertising (e.g., Yahoo might post NYSE data on one of its Web pages and, in return, give the NYSE a percentage of the advertising revenues from that Web page). This model would have the advantage of making data free to users. Mr. Bernard pointed out, however, that this type of advertising-based model currently seems to be failing for online businesses.
The Chairman then asked for conclusions on whether, in a world with the current Display Rule and competing consolidators, additional SEC rules would be needed to address economic/pricing concerns. Ms. Dwyer believed there would have to be more precise regulations to ensure there is non-discriminatory, most-favored nation pricing. She argued that the SROs' existing practice of discriminating among users based on end-use of the data was unacceptable, and in fact was "unreasonably discriminatory." Flat license fees, or fees based on the volume of transactions on a particular exchange, would be acceptable, but not fees that distinguish how data is used, including whether the data ultimately is conveyed to investors over the telephone or by computer. Mr. Bernard countered that the NYSE has not found a better way to allocate fees than the existing consensus-based process - whenever adjustments are made there will be winners and losers. Ms. Dwyer noted that the other way to go would be to eliminate the Display Rule and defer to market forces. But so long as market data is a "regulatory product," it would be unfair for its price to vary depending on how the data is used.
Mr. Roiter suggested that, ultimately, the users as a group will bear the full costs of the data - the real issue is how those costs will be allocated among them. Mr. Bernard argued that the data producers legitimately should have some business discretion when it comes to these allocation decisions, since they have to make judgments on how to structure fees in a manner most likely to achieve certain revenue goals. Ms. Dwyer pointed out that the way in which data is used makes no cost difference to an SRO. Mr. Bernard responded that there inevitably will be disagreements about allocation decisions and that, while the current consensus-based system is imperfect, there does not seem to be a better one. Ms. Friedman added that allocation decisions are very difficult - many different factors are weighed, such as the value of the data to the user and the means of delivery. If one of Ms. Dwyer's preferred allocation methods was adopted, other users likely would complain of discriminatory treatment. Ms. Dwyer stressed that there is a need to balance the goal of adequately funding of the markets with the goal of achieving the broadest possible data dissemination.
After the lunch break, the Chairman summarized the discussion by suggesting there seemed to be consensus that, in a world with competing consolidators and a Display Rule, no additional SEC rulemaking would be necessary to address economic/pricing concerns, such as the reasonableness of fees and discriminatory pricing - the existing residual SEC oversight authority would be sufficient. Mr. Roiter clarified that, while at the outset no new regulations should be implemented, the full impact of competing consolidators could not be predicted, and there might be a need to step in later with additional regulations if necessary. Ms. Dwyer asked that she be noted as dissenting from the consensus view.
The Chairman then asked the members to discuss the potential benefits of moving to a competing consolidators model. Possible benefits might include increased innovation in the provision of market data, elimination of the governance inefficiencies of the existing consortia, and the introduction of competition among market centers in the market data arena. Mr. Bernard reminded members that, in the case of SIAC, the SIP function is only a $7 million per year business, so the potential economic benefits might be relatively small. In fact, many view SIAC as being quite efficient, so injecting competition could actually increase costs.
As to the potential for increased innovation, Mr. Bernard indicated this might not be a significant potential benefit. Ms. Friedman suggested any increased innovation might result from eliminating the inefficiencies (e.g., the cumbersome negotiation process) of the existing joint plans. Mr. O'Kelly pointed out that any efficiency gains in this regard might be offset by the need to have multiple negotiations for data - with each exchange - in a competing consolidators model. The Chairman concluded by noting that members appeared "underwhelmed" by the increased innovation potential benefit.
Turning to the elimination of the inefficiencies of the existing consortia, Mr. Bernard argued that, under the Exchange Act, burdens cannot be placed on competition without showing that they are necessary to achieve other regulatory goals. The existing joint SRO consortia burden competition, and can no longer be justified - on technological grounds - as necessary to assure the wide dissemination of market data. Mr. Bernard also believed that the consortia are dysfunctional - as the markets have developed and issues have become more complex, consensus has become more difficult to achieve.
The Chairman asked Mr. Roiter to outline possible arguments for retaining the consortia. Mr. Roiter suggested that the SEC wants to encourage market competition, in part through widespread dissemination of the NBBO, and one could argue that maintaining the joint SRO consortia is an effective and reliable means to achieve this. Mr. Bernard countered that an NBBO would be produced simply if the Display Rule were retained, and perhaps even as a result of best execution obligations alone. Mr. O'Kelly argued that, while achieving consensus in the consortia can be a difficult process, one significant benefit is that they permit users to deal with several exchanges through one point of contact. And, on balance, the consortia have worked quite well for the past 25 years.
Mr. Feuer noted that one rationale for the Display Rule was to promote competition among the exchanges. He believed the level of competition among exchanges has stayed approximately the same over the last 20 years, so that the Display Rule has not achieved that goal. Mr. Bernard agreed that the Display Rule has not engendered quote competition among the exchanges, and noted the irony that markets are rewarded based on transactions, not quotes, under the existing plans. Ms. Friedman pointed out that, regardless of the merits of the Display Rule, it provides no justification for retaining the joint plans, since the Display Rule can be complied with without them.
The Chairman then turned to the final issue of the day - whether the Display Rule should be modified or eliminated - and he asked the members to focus on the costs, risks, and benefits of any recommended change. As a possible sub-question, Mr. Roiter suggested that the group address whether market forces alone would produce an NBBO and, if not, whether that would be a problem. The Chairman added that members also might want to discuss, in the event an NBBO was not produced, what additional steps (if any) the SEC might need to take to assure best execution.
Ms. Dwyer favored elimination of the Display Rule, and believed that market forces would produce an NBBO without it. As a practical matter, broker-dealers today would not want to rely on anything less to satisfy best execution obligations - to do so would risk litigation. Mr. Bernard said the NYSE viewed this as a very difficult issue. Prior to the Display Rule, the regional exchanges were giving their data away and vendors weren't taking it. In the absence of a Display Rule, it might be difficult to get the regionals' data out to the market, but that data (at least the quotes) has very low value because the regionals rarely are at the best. Mr. Mohr stated that Datek supports elimination of the Display Rule - transmission costs have decreased since the time the Display Rule was adopted, and the marketplace readily could produce an NBBO to the extent it was desired.
Mr. Roiter expressed concern that, were the Display Rule to be eliminated, best execution obligations alone might not produce an NBBO. Best execution is not necessarily analyzed on a trade-by-trade basis and, in any event, can take things other than price into account (e.g., speed, reliability). Mr. Feuer responded that, to the extent best execution means more than price, there may not be a justification for retaining the Display Rule - investors should be able to decide what information they need.
The Chairman then asked for final thoughts on the Display Rule issue. Mr. Roiter found this to be a very close call. If the SEC were prepared to "ratchet up" the best execution obligations, he might be prepared to support elimination of the Display Rule, but absent that his inclination would be to support its retention. He had not reached any final conclusions, however. Ms. Dwyer reiterated Schwab's support for the elimination of the Display Rule. In practice, the duty of best execution is becoming stronger, and she felt confident an NBBO would be produced without the Display Rule. Mr. Bernard agreed that, at least with respect to broker-dealers with small retail customers, an NBBO naturally would be demanded, and that technological advances have made its production relatively easy. Mr. Smith suggested that Datek supports elimination of the Display Rule because it would free the marketplace from a "one-size-fits-all" mentality. Ms. Friedman reiterated Nasdaq's support for retaining the Display Rule, but with the qualification that it might be acceptable to limit its application to certain circumstances. Mr. O'Kelly expressed the Chicago Stock Exchange's support for retaining the Display Rule, although modifications to address problems such as flickering quotes might be appropriate.
The Chairman concluded the meeting by suggesting that there seemed to be more consensus on the technology-related issues than the economic issues. There clearly was no consensus on whether the Display Rule should be retained, modified or eliminated, and the Subcommittee's report would reflect that. A draft report will be circulated to Subcommittee members shortly for review and comment. The report will act primarily to set up the various issues concerning alternative models, so that they can be deliberated by the full Advisory Committee at its May 14 meeting. To the extent consensus was reached in the Subcommittee on a particular issue, that will be noted.