Simon Johnson (MIT)

May 7, 2001

To: SEC Advisory Committee on Market Information

From: Simon Johnson, MIT

There has been considerable discussion during our meetings and in various memos about the potential benefits and costs of switching to a very different system of distributing market information (i.e., some form of multiple consolidator model without the Display Rule).

We have talked at length about potential benefits. Let me elaborate on some of the economic risks.

The Regulatory Mechanism

1. Market regulation in the US relies on a delicate balance between decisions and interventions by the SEC, rule-making and enforcement by SROs, self-policing by other market participants, and of course ultimate reliance on the court system.

2. The US system has been uniquely successful in that it allows competition between market centers (including a remarkable degree of entry by new centers) while maintaining overall high standards, for example price transparency from the perspective of individual investors. Moving to a model of multiple consolidators without the Display Rule could upset this finely tuned regulatory mechanism in several ways.

3. If changing the model of consolidators undermines the ability of SROs to adequately fund their self-regulatory functions, this damages a key piece of the regulatory mechanism. How exactly will rules be enforced and by whom? Will the quality of SRO oversight be impaired? The committee still needs to address these issues directly.

4. In addition, for effective regulation there need to be clear and transparent minimum standards that the regulators, the courts, the SROs and other market participants can observe and enforce. Clearly the Display Rule (and particularly the NBBO) plays this role in the current system. If moving to multiple consolidators is accompanied by abolishing the Display Rule/NBBO, the regulatory mechanism will be much harder to operate. Satisfying "best execution" obligations also may become more complex.

5. The US securities industry currently benefits from a "light touch" regulatory system. If you remove the tools that make this possible, either the regulator (and legislator) will have to mandate a much more intrusive and restrictive model of regulation or there will be less regulation. Less regulation may seem appealing, but it has potentially dangerous economic consequences in terms of both market power and quality.

Market Power

1. Market power arises in any situation where there is price setting rather than price taking. In a market with differentiated goods or imperfect substitutes there is always some degree of market power. The question is to what extent the price setter feels constrained by potential competition, the response of consumers, and regulators.

2. Much of our discussion has focussed on actual competition, i.e., the alternative providers to whom consumers can turn if they are not satisfied with a product or service. The intensity of competition in US securities markets is an important outcome of the current regulatory system, but there is no guarantee it would survive the transition to multiple consolidators without the Display Rule.

3. All price setting is "local", in the sense that it is relative to the competitors and consumers who are offering/looking for the same service at the same moment in time. In general it is costly for consumers to search between segments at one moment or even over time. It is these search costs that unavoidably create differentiated products and potential market power.

4. Many members of the committee seem to assume that multiple consolidators will necessarily lead to more competition. But this is definitely not the case if it leads to fragmentation of market information (for example, because the Display Rule is abolished) and greater search costs for consumers (e.g., small investors).

5. In any market, firms naturally move to sell goods that cannot easily be compared, each carving out its own niche, and fragmentation is the natural result. While this is tolerable in many situations (outside of financial markets), fragmentation is not conducive to overall financial development and to the fair treatment of small investors. Note that innovation may reduce search costs (and has under the current National Market System), but it can also increase them – for example if the innovators with deep pockets have an incentive and opportunity to create more fragmented markets.

6. Current constraints on market power do not necessarily imply the moderation of market power with multiple consolidators (particularly without the Display Rule). For example, there may be a period of intense competition, followed by the development of market power. This is the typical experience of deregulation.

7. In fact, under a new system with multiple consolidators, it may become harder for new market centers to enter, for example if there are no requirements to distribute their information. In this case, the longer-term effects would be to reduce actual and potential competition. Note that innovation is likely to decrease rather than increase in this scenario.

8. If this happens, what would be the right response of the regulators and legislators? There would be a strong economic case for further regulation and even rate setting. It is much more appealing and significantly easier to set rates when there are relatively few producers with considerable market power and when there is no new entry.

9. We should keep in mind the following points about the economics of market power (which have already arisen during our discussion).

10. Market power can exist even if prices are falling. This is possible if costs are also falling, for example due to innovation on the part of information technology providers. In a complex business with differentiated goods, there is always some degree of market power.

11. Market power can exist even if customers are represented on the governing body of the organization setting prices. It may be helpful to have customer representation, but given that all the customers cannot be represented, this cannot eliminate market power. In fact, strong representation of large customers could lead to rules or practices that favor them relative to small investors. Just because a large firm has small investors as its customers does not necessarily imply that it will always and everywhere act in their best interests.

12. Market power does not imply setting infinite prices. It means setting prices above the purely competitive level. A for-profit organization typically sets prices at the most profitable level and has a problem with its shareholders if it does not.

Quality

1. There is an impression among some that market participants can be relied upon to ensure good behavior without much regulatory oversight. This is not consistent with the available evidence on how financial markets develop.

2. The problem does not lie with the majority of market participants. They obviously gain substantial value from guarding their reputations. Great names in finance only occasionally engage in questionable practices (usually when their overall business is in trouble or if there is a problem with internal controls.)

3. The main problem lies with some of the relatively small fringe players. It is often relatively hard for some these organizations to compete on the basis of having a great reputation (although others will build reputations and become the great names of the future). Some of these firms will offer a product with lower quality. This offer will attract less informed consumers who will either consistently overpay or pay too much for a while and then drop out in frustration.

4. In many markets high and low reputation players can coexist. The danger for financial markets is that relatively few dishonest players can ruin the market for everyone. If small investors feel that prices are not transparent or trading is in some other sense "unfair" they will withhold their business. This has happened in the past in the US and could happen again.

5. Although many small investors are well informed, some are not. It is this second group that would really lose out if information is fragmented and not of consistently high quality.

6. Just relying on the duty of best execution is unlikely to be enough to maintain consistent quality across all markets and all customers. In particular, without a Display Rule, the ability of the consumer (or the regulator or the courts) to measure the quality of execution may become much more difficult.

7. Given the rapidly falling costs of providing information and the move to decimalization, the best way to protect individual investors is probably to increase, rather than reduce, the mandatory minimum of information that must be provided (and displayed).

Most Likely Outcomes

1. Moving to a model of multiple consolidators would likely not change much as long as the Display Rule is kept.

2. Abolishing the Display Rule might initially lead to more competition. However, this would likely lead rapidly to fragmentation of market information and greater market power in particular segments. There would be less entry by new market centers.

3. The quality of some services would also most likely be compromised, particularly for relatively small investors. Given that it would be much harder for the SEC and other parts of the regulatory system to function, there would be a mounting sense of frustration.

4. The result would be growing pressure for re-regulation. This kind of pressure is typically stronger when the economy is weaker and when there have been well-publicized problems.

5. In all likelihood the ultimate response would be greater regulation, including potentially more detailed controls and rate setting of the form that exists in other industries.

Moving to a new system of market information may convey benefits as well as these costs. However, the benefits are likely to be incremental and captured by particular market players. The costs are likely to be systemic and borne largely by some small investors. The costs could also potentially be very large, both in the short-term and if the final outcome is a less efficient system. Given that the current system for distributing market information works well (i.e., is consistent with innovation, lower prices, and consistently high quality services for small investors), is there really a case for radical change?