Securities and Exchange Commission
IN THE MATTER OF:
Department of Justice
I. Introduction and Summary
The Securities and Exchange Commission ("Commission") has requested public comment on four rule proposals designed to improve the handling and execution of customer orders. The request for public comment was published on October 10, 1995 in the Federal Register. The Commission has also requested public comment on the proposed rule change filed by the National Association of Securities Dealers ("NASD") relating to implementation of the NAqcess System ("NAqcess") as a replacement for the Small Order Execution System on the Nasdaq Stock Market, Inc. ("Nasdaq"). This request for public comment was published on December 1, 1995 in the Federal Register.
As the executive branch agency responsible for promoting competition, consistent with other statutory goals, the U.S. Department of Justice ("Department") is pleased to have the opportunity to comment on the Commission's rule proposals. Reflecting the Department's expertise in competition policy, our comments focus on the likely effects of the Commission's proposals on competition for customer orders.
Our comments are specifically directed at the likely effects of the proposed rules on Nasdaq; we do not address the effects of the proposed rules on exchanges. In offering these comments, the Department recognizes that the Commission may have to balance competition-policy concerns against other regulatory concerns such as market fragmentation and transparency. In this context, the Department agrees with the Commission's statement regarding the importance of competition in securities markets:(1)
"Many of the innovations in the markets during the past 20 years have been generated by competition. It is not surprising that the introduction of new technologies which benefit investors has been the product of competition between and among markets and market participants."
In its introduction to the rule proposals, the Commission points out that "Congress saw competition as the primary source of change and innovation in achieving an NMS (National Market System) and directed the Commission to use its rulemaking authority to remove impediments to competition and facilitate the development of an NMS." (Federal Register, vol. 60, No. 195, 10/10/95, p. 52793) The Department agrees that competition is the most powerful and efficient way to improve market conditions and applauds the Commission's efforts to enhance competition in the marketplace. The Commission's proposals are designed to "enhance transparency in our markets and improve the handling and interaction of customer orders." (Id.) Increased transparency and standing of public limit orders have been recognized by numerous market participants -- including the NASD -- as an objective likely to pay large dividends in terms of increased competitiveness. The Department's primary purpose in these comments is to assess whether the Commission's rule proposals enhance the overall competitiveness of the market.
Our comments are addressed to three of the Commission's four proposals: (1) the "Limit Order" proposal requiring specialists and over-the-counter ("OTC") market makers to display customer limit orders priced better than the specialist's or OTC market maker's quote; (2) the "Electronic Communications Networks" proposal, requiring exchange specialists and OTC market makers to quote to the public any better prices that they privately quote through certain electronic communications networks; and (3) the "Price Improvement" proposal, requiring that specialists and OTC market makers provide customer market orders some opportunity for price improvement before executing such orders.(2) We comment on the NASD's NAqcess proposal in the context of our discussion of the Commission's Limit Order proposal.
The Department strongly supports the Commission's Limit Order proposal. We believe that adoption of this proposed rule will have a highly beneficial impact on Nasdaq, a stock market with which we have recently gained considerable familiarity. In effect, the Limit Order proposal will allow customer limit orders to compete more effectively with market makers' quotes, injecting additional competition into the Nasdaq market. The primary beneficiaries of this added competition will be the investing public, in the form of narrower bid/ask spreads and thus a reduced cost of trading. Indeed, the Department believes that competition would be enhanced even further were the Commission's Limit Order proposal to be strengthened in several ways, as discussed below.
The Department also supports the Commission's Electronic Communications Networks proposal. The proposal may reduce the possibility of collusion and may also serve some of the Commission's other goals, such as promoting transparency and reducing market fragmentation. We do, however, have some concerns with the proposal. The proposed rule could impair the use of electronic communications networks, and could have the unfortunate effect of slowing the adoption of new technology and reducing the variety of trading options available in the market. If the proposal is adopted, then we urge the Commission to continue to monitor developments in Nasdaq, and to review from time to time the continued need for the rule.
The Department's view on the Price Improvement proposal is one of caution. If the other proposals are effective in narrowing spreads and promoting competition, the Price Improvement proposal may not be necessary. Offering market orders the opportunity for short-term price improvement in and of itself will tend to narrow bid/ask spreads, but at the expense of encouraging customers to enter market orders rather than limit orders. Since public limit orders have the potential to play a key role in providing competition for market makers, shifting customers away from limit orders is not without its costs. All of this suggests to us a wait-and-see approach: defer adoption of the Price Improvement proposal until it is possible to assess how the Nasdaq market operates subject to the other proposals. The complexity of the Price Improvement rule, which could well impose significant compliance and enforcement costs, bolsters this conclusion.
In connection with the Price Improvement proposal, the Commission requested comment on rules regarding excess spread parameters. The Department is concerned that such rules may result in wider spreads, and therefore recommends that the Commission consider repealing excess spread rules.
II. Limit Order Proposal
This proposal would give standing to public limit orders that are placed with market makers. Market makers would be required to display a customer's limit order price if it were superior to the market maker's quote, which is, in effect, also a limit order. The public limit order price would then be eligible to be the national best bid or offer on the Nasdaq screen.
The Department welcomes and supports this proposal. The proposal will increase the transparency of public limit orders and enable public limit orders to compete with market makers' quotes. In so doing, the proposal will encourage the entry of public limit orders, especially since the proposal allows customer orders to interact with one another. The increased competition between public limit orders and market makers' quotes is likely to reduce inside spreads and to undermine market makers' willingness and ability to collude.(3)
Of the Commission's four rule proposals, this is the most significant in terms of its competitive effects. We are quite confident that it will be highly beneficial in reducing the cost to the public of trading Nasdaq stocks. These benefits are likely to be more pronounced over time, as market participants become familiar with the improved treatment of limit orders and increasingly place such orders with their brokers.
Since customer limit orders have the potential to play such an important pro-competitive role in the market, we encourage the Commission to consider strengthening the Limit Order proposal in three respects.
First, it would be desirable if the public could enter a limit order without using a market maker. The proposal does not provide for such a possibility. The subsequent NAqcess proposal would enable order-entry firms, under certain conditions, to enter public orders directly into a central limit order book ("CLOB"). Although we have concerns with specific elements of the NAqcess proposal,(4) its principal innovation -- a CLOB that does not require the use of a market maker to place orders in it -- is commendable for the very same reasons we favor a strong version of the Commission's Limit Order proposal: (1) it facilitates competition between public limit orders and market makers' quotes; (2) it enables public limit orders to interact with one another; and (3) it enables intermediaries other than market makers to enter public limit orders into such a facility.
The ability of the public to enter a limit order without using a market maker is important; market makers have an incentive to discourage public limit orders. After all, public limit orders directly compete with market makers' own quotes. Historically, market makers have not executed public limit orders within the inside spread, but have instead waited until the market price has moved sufficiently that they are no longer within the inside spread.
Second, we encourage the Commission to consider giving time priority for public limit orders. Without this protection, market makers may be able to discourage such limit orders by strategically choosing when to match the prices of customer limit orders. More specifically, a market maker could match and trade ahead of customer limit orders when and only when the market maker perceives that the market price is moving in sympathy with the limit order. Assuming that market makers have an informational advantage over the public, such strategic treatment of public limit orders could do much to discourage these orders and thus undermine the Limit Order proposal. Affording time priority to public limit orders would go far toward preventing such strategic behavior. It is useful to recognize that giving time priority to public limit orders will inevitably mean that time priority would have to be given to all limit orders -- those of market makers as well as those of the public. Giving time priority to market makers' quotes may have the added benefit of enhancing competition among market makers by creating incentives for market makers to improve their quotes.
We recognize that there may be difficulties in ensuring that limit orders indeed receive time priority. For one thing, the Commission would have to determine how to treat limit orders entered in vertically integrated firms, i.e., firms that serve as both broker and dealer. In addition, affording time priority may not be feasible without the use of an automated transmission and execution system. We do not profess expertise on the implementation of such systems, nor on how existing technology, such as that used by Instinet, could be applied to the Nasdaq system. We merely note that it would be highly desirable on competition grounds to provide time priority to public limit orders, especially in the context of an automated transmission and execution system, and we encourage the Commission to explore this possibility.
Third, the Department recommends that the Commission consider implementing an opening auction that reflects the accumulated market and limit orders at the market opening. An opening auction would help achieve the Commission's goal of "[increasing] market transparency and efficiency by ensuring that prices fully reflect overall supply and demand (Federal Register, vol. 60, No. 195, 10/10/95, p. 52794)."(5) At present, the opening price is set by the first market maker who happens to trade. In addition, an opening auction would allow customer order interaction without the participation of a market maker, thereby providing some new competition to the continuous Nasdaq market. It is possible that the NAqcess proposal could be modified so as to achieve a single price opening. The Department encourages the Commission to explore this possibility, both now and in the future.
III. Electronic Communications Networks Proposal
This proposal would require market makers to quote a price on the Nasdaq screen that is no worse than a price they quote on any other electronic communications network, notably Instinet. The proposed rule would apply to market makers in their dual roles as principals and agents for customer orders.
The proposed rule is intended to decrease costs associated with market fragmentation and the splintering of quotation information. Market fragmentation is defined as "the inability of an order in one market to trade with an order in another market."(6)
Market fragmentation could indeed be costly to retail investors if their transactions were executed on Nasdaq at prices inferior to those on other electronic communications networks. But if the effect of the rule were to induce market makers to improve their Nasdaq quotes to match their quotes on Instinet, then the Department believes retail investors would benefit.(7)
Our support for the proposal is based on an analysis of the dynamics of possible collusive agreements in the trading of Nasdaq securities. Electronic communications networks may be used by market makers to better their offers and make some trades while avoiding price improvement on Nasdaq. Were this option to be removed, then the incentives for market makers to cheat on a possible collusive agreement on Nasdaq trades in order to make some relatively desirable trades that could no longer be directed to Instinet could be sufficient to cause any agreement to collapse.
The nature of the service offered by market makers on Instinet or other electronic communications networks differs from the service they offer on Nasdaq. Different customer segments tend to use different trading mechanisms. Large, institutional customers that have relatively many trading alternatives, or that are known to be trading for liquidity purposes, may be in a good position to negotiate favorable prices. But if these customers prefer to use Instinet, and if market makers are not required to match their quotes on Instinet on the Nasdaq screen, then the presence of large, institutional customers in the market for a security will have only a limited effect in disciplining pricing on Nasdaq. As a result, it may be easier for market makers to achieve and maintain a collusive agreement on Nasdaq.
The issues raised by the Electronic Communications Networks proposal are highly complex, and we are less confident of our prediction of the effects of this proposal than we are of our expectations regarding the Limit Order proposal. The most significant concern we have is that, by constraining the ways in which market makers can use electronic communications networks, the proposed rule may impair the usefulness and competitive impact of such networks and impede their further development. For example, the proposed rule could decrease liquidity on Instinet by reducing its use by market makers. This could impose a cost on institutions that currently make significant use of Instinet, and also on other market participants who benefit from information on Instinet prices and quotes. Our concern about the possible effect of the proposed rule on innovation and the development of new, efficient networks is amplified by the current rapid pace of technological change in information systems, and the significant benefits that could result from application of this new technology to securities markets. While we think, as discussed above, that the benefits likely to flow from the proposed rule outweigh these concerns, these possible adverse consequences should be carefully considered before the proposed rule is adopted.
Another concern is that the proposed rule may reduce the variety of choices available to at least some traders. For example, there are significant differences between services offered on Nasdaq and on Instinet. In the Nasdaq market, the identity of the market maker is displayed along with its quotes; in Instinet, the identity of the offerer is not revealed. Anonymity assists market makers in working large positions; there is less price impact if their identity is not revealed. A second difference is that market makers' must display a quote on Nasdaq continuously, whereas orders on Instinet are single orders only (i.e., once they are hit, they are removed automatically from the system). Third, the minimum tick size is smaller on Instinet than on Nasdaq. In general, regulations that force the prices of different products to be equal generate inefficiencies and reduce competition.
Given some unpredictability in the effects of the Electronic Communications Networks proposal, the Department urges the Commission to closely monitor developments on Nasdaq and to continue to analyze the effects of the proposed rule if it is adopted. We expect that implementation of the proposed rule will, on balance, enhance competition in the trading of Nasdaq securities.(8)
IV. Price Improvement Proposal
The Commission is proposing a price improvement rule for customer market orders in a covered security. Covered securities would include the 250 Nasdaq stocks with the highest average daily trading volume over the previous quarter. The Commission has set this threshold in an attempt to identify the securities for which price improvement opportunities are appropriate. In addition, the Commission is proposing a non-exclusive safe harbor as one means by which an OTC market maker can be assured that an order has received sufficient opportunity for price improvement for purposes of the rule.
The Department supports the Commission's goal of narrowing inside spreads by affording price improvement to market orders. And we certainly appreciate the potential benefits associated with the (brief) narrowing of the inside spread under the Commission's non-exclusive safe harbor.
However, the Department is far from certain that the proposal will deliver significant competitive benefits, especially in the context of the other proposals. For example, if the other proposals are successful in forcing inside spreads to competitive levels, the Price Improvement proposal may impose more in compliance and enforcement costs than it delivers in benefits. Moreover, and of significance, the Price Improvement proposal could operate at cross purposes with the Limit Order proposal to the extent that the Price Improvement proposal induces customers to enter market rather than limit orders. Furthermore, the non-exclusive safe harbor provision in the proposal gives market makers an opportunity to quote less favorable prices than otherwise, knowing that they have an opportunity to step in and take a market order when it is entered if conditions at the time warrant it. As a result, market makers' quotes may become less informative and less competitive.
For all of these reasons, we encourage the Commission to consider deferring adoption of the Price Improvement proposal until more experience is gained with the operation of the Nasdaq market under the other proposals.
Excess Spread Rules
The Commission requested comment on several issues raised by the proposal, including rules regarding excess spread parameters. The Department believes that excess spread rules may result in wider spreads. A market maker that is contemplating narrowing its spread for a given stock might anticipate that other market makers would respond by narrowing their spreads. If a few other market makers matched or bettered the narrower spread of the first market maker, then the excess spread rule could come into play, and prevent subsequent widening of the spread. In some instances, this possibility could be a sufficient reason to discourage a market maker from narrowing its spread in the first place. We therefore recommend that the Commission consider repealing excess spread rules.
The Department recommends adoption of the Limit Order and Electronic Communications Networks proposals, deferral of the Price Improvement proposal, and repeal of excess spread rule on Nasdaq.
Dated: January 26, 1996
1. Statement by the Securities and Exchange Commission Upon Release of the Market 2000 Report, January 27, 1994.
2. The fourth proposal would require exchanges and over-the-counter market makers in listed securities to publish quotations for listed securities where the exchange or OTC market maker trades more than 1% of the aggregate trading volume for that security.
3. Berkman, using actual data on limit orders, shows that limit orders narrow the inside spread, in "The Market Spread, Limit Orders, and Options," Journal of Financial Services Research, January, 1993, pp. 399-415. In addition, empirical studies find that spreads are lower in auction markets (in which limit orders are exposed) than in dealer markets (in which limit orders are not exposed). See Huang and Stoll, "Dealer versus Auction Markets: A Paired Comparison of Execution Costs on NASDAQ and the NYSE", forthcoming in Journal of Financial Economics.
4. The Department's concerns with the NAqcess proposal relate to limitations on the size of orders entered in NAqcess, and limitations on which entities can have their orders entered in the system. We recognize that the system is designed to facilitate small retail orders, and that the limitations would not effect, at least not directly, small retail orders. But the Department believes that relaxing the limitations would enhance competition and market efficiency, including the efficiency with which small retail orders are executed.
The NAqcess proposal has a maximum order size of 1000 shares for limit orders, except for limit orders in the securities that comprise the Nasdaq 100 index, for which the maximum size would be 3000 shares. The proposal has a maximum market order size of 1000 shares for active Nasdaq stocks and lower limits for less active stocks. The Department encourages the Commission to consider deleting any size limitations for limit orders and market orders entered into NAqcess.
The NAqcess proposal, with one or two exceptions, does not allow brokers and dealers to enter their own principal orders into the NAqcess limit order file. We encourage the Commission to consider allowing brokers and dealers to enter limit orders as principals in NAqcess.
5. The benefits from the enhanced market efficiency would spillover to derivative markets; several stock indices, including the S&P 500 (which contains 50 or 60 Nasdaq stocks), have a cash settlement at the market opening.
6. Hans Stoll, "Organization of the Stock Market: Competition or Fragmentation?" in Lehn and Kamphuis (eds.), Modernizing U.S. Securities Regulation (Burr Ridge, Ill: Irwin Professional Publishing, 1992).
7. According to economic theory, this is not the only possible outcome: in general, if different consumers face different prices for the same product or service, then regulating the outcome to require a single price may result in some consumers paying higher prices, without necessarily causing prices paid by any consumers to fall. Even if prices paid by a few consumers do fall, consumers may be worse off overall. In addition, by preventing market makers from placing orders in Instinet within the inside spread, the proposed rule may increase market makers' costs of adjusting inventory. Any increased costs would likely be passed on to traders. In the case of the Electronic Communications Networks proposal, however, given the nature of the Nasdaq market today, the Department expects that consumers will benefit overall.
8. A review of evidence from the London Stock Exchange may help inform this analysis. The London Stock Exchange, which is organized as a dealer market, was subject to Rule 4.18, which barred market makers from displaying a superior price (or quantity) on a public system other than SEAQ, the Stock Exchange Automated Quotation system. This rule was recently jettisoned, following an analysis by the Office of Fair Trading that concluded that the rule impeded the development of competing trading systems and restricted price competition. It is not yet clear what effect this change in the rules will have on the London Stock Exchange, nor is it clear exactly what the parallels are between the respective rules or between the London Stock Exchange and the Nasdaq market, but these questions may merit further attention.