May 26, 2000

Mr. Jonathan G. Katz
Secretary
U.S. Securities and Exchange Commission
450 5th Street N.W.
Washington, D.C. 20549-0609

Re: Exchange Act Release No 42450; File No. 502-NYSE-99-48 -
Notice of Filing of Proposed Rule Change by the New York Stock
Exchange, Inc. to Rescind Exchange Rule 390; Commission Request
for Comment on Issues relating to Market Fragmentation

Dear Mr. Katz:

CIBC World Markets Corp. ("CIBC") is pleased to comment on the New York Stock Exchange, Inc. (the "NYSE") proposal to rescind NYSE Rule 390 as well as on certain issues set forth in Exchange Act Release No. 42450 (the "Release").

As an NYSE and NASD member who makes markets in over 500 securities on Nasdaq, CIBC embraces the concept of providing a market foundation that fosters competition and innovation as a means of providing liquidity, depth and best execution opportunities for all types of orders and diverse market participants. Accordingly, and as more fully set out below, CIBC strongly endorses Securities and Exchange Commission (the "Commission") approval permitting the rescission of NYSE Rule 390. CIBC also advocates a policy of regulatory initiatives that are designed to maintain and enhance competition for order flow by providing the market place with execution choices and alternatives.

1. CIBC Strongly Favors Rescission of NYSE Rule 390

NYSE Rule 390 prohibited NYSE members from effecting certain transactions in NYSE-listed securities off an exchange. The Commission recognized the anti-competitive effect of Rule 390 when it adopted Rule 19c-3 in 1980, which prohibits the application of any off-board trading restrictions to securities newly listed on an exchange after April 26, 1979. Since 1979 we have witnessed such developments as advancing technology providing faster and better executions, ECN's and regulatory initiatives such as the Limit Order Handling and Display Rules. These developments have resulted in a more efficient and competitive market. We strongly believe that the rescission of Rule 390 is appropriate and consistent with those developments. The Rule was, in fact, contrary to the competitive spirit and structure of today's markets.

2. Regulatory Initiatives that Foster Market Competition and Execution Alternatives Best Serve Current Markets

"Varied markets competing for order flow are consistent with the longstanding Congressional and Commission objective of enhancing competition in the equity markets."1 CIBC believes that any future rulemaking by the Commission or any Self-Regulatory Organization should be consistent with the above-quoted objective. In fact, as the Commission noted in its Market 2000 study in 1994, "The Commission should be reluctant to impose a single design on the markets absent evidence of a significant market failure."

We see no evidence of market failure. Quite the opposite, improved technology, certain regulatory initiatives and increased investor education and sophistication have resulted in perhaps the healthiest market in history. While we acknowledge there is always reason for improvement, we do not agree that the market has "several significant flaws,"2 nor do we agree that a single system or market is the solution. Such a solution would be inconsistent with Congressional intent, the Commission's own stated beliefs and the principle that capitalism fosters competition. Accordingly, we do not advocate as set forth in the Release a National Linkage System ("NLS") providing time and price priority and immediate execution, if it is the only market structure available to investors. That proposal fails to take into account the substantial diversity that exists in the current market.

We believe the current regulatory structure supporting the listed and over-the-counter markets is wholly adequate and is generally serving investors' best interests. Commission initiatives like the Limit Order Display Rule and SRO initiatives such as the Manning Interpretation3 have greatly raised the level of investor protection and investor confidence. Such confidence is reflected, in part by the current record trading volumes. We also believe such confidence and investor participation stems from the current menu of execution choices available to investors ranging from market makers "working" large orders to committing capital to ECN's that provide liquidity for smaller orders. We doubt such levels of volume would exist if significant structural flaws based on fragmentation were evident in this market.

Since the Market 2000 study, the market has witnessed a huge growth of the mutual fund industry as well as the explosion of the on-line trader. Both participants and all in between have different needs and priorities. A single market structure could not address all needs and priorities. If the Commission chooses to act, we believe a better solution would be a series of regulatory initiatives that continue to offer execution alternatives to the markets' diverse interests.

Consistent with a more flexible regulatory approach, we believe that the Commission should advocate a varied market structure that could be accessed by any participant, linked by market makers, dealers and specialists and their respective duties of best execution as opposed to a single structure based upon a mandated linkage predicated on time and price priority of certain orders. Regarding the duty of best execution, as the Commission noted in the Release, "A critical factor for non-marketable limit orders, however, is that they be routed to the market center that provides the greatest likelihood of execution. A broker must take into account many factors to properly make a routing or determination." A market structure that provides for varied execution alternatives provides brokers the best framework to make such a determination.

3. Certain Proposed Regulatory Initiatives Set Forth in the Release Need to Take Into Account the Size of an Order and the Liquidity of Individual Securities Markets

In our opinion, the primary factor in determining best execution alternatives for any order is size of the order. Accordingly, we believe a varied market structure that provides for display and execution of orders based upon an automated or matching system as well as providing for the execution of orders through dealers who commit capital and otherwise "work" orders, would serve the needs of all market participants and all sizes of orders.

In the release adopting the Limit Order Handling Rules,4 the Commission noted that the ultimate source of market liquidity is the customer order.5 In the Release, the Commission, seeking comment on the concept of intermarket, time and price priority for limit order or quotations that improve the NBBO, refers to the first limit order or dealer quotation that improves the NBBO as a "Price Improver." While we believe that limit orders provide some liquidity, they do not provide liquidity for all markets and for all investors. For example, a customer placing a 500,000 share sell order would rather be able to sell quickly 50,000 shares at 10 ¼ than quickly sell 5,000 at 10 3/8, particularly if hitting that 3/8 bid would jeopardize the larger liquidity at 10 ¼. A mandated limit order book with time and price priority will likely not provide incentive to dealers to provide the liquidity for those larger orders. In fact, the 5,000 share order in those circumstances will likely not result in overall price improvement for that 500,000 share order. We believe the Release has overlooked those large orders, often executed for the benefit of mutual funds and their individual customers.6 The Commission may be failing to recognize that a major provider of market liquidity may be the large "not held" order. It is that order that drives many markets and results in the liquidity necessary so that market makers may effect fast execution of limit and market orders.

As noted previously, we believe that protection of customer limit orders is important to market structure. Regulatory initiatives such as Manning and the Display Rule help insure that such orders are treated fairly. Continued vigilance to insure limit order protection is necessary in any market structure. However, we believe that any attempt to establish a single limit order book would discourage market makers from taking risk in many securities less liquid than, for example, the Nasdaq 100. In many circumstances, the Commission places too great an emphasis on the customer limit orders as the primary source of liquidity.7 While in the markets of well known, highly liquid securities, which may be more limit order driven, mechanisms such as the proposed intermarket time/price priority for limit orders or quotation that improve the NBBO may be successful, for many less liquid securities, such a market could create increased volatility and lack of depth. This is primarily because the proposal fails to take into account the size of the limit order. The Commission no doubt recognizes that a 100 x 100 (10,000 share) market is not driven by a 100 share order or a (1 x 1) quote even if that order or quotation offers price improvement.8 A customer placing a large order is likely to want substantial execution at a price as opposed to multiple executions at different prices if only a fraction of the order is price improved.

We believe that large orders, whether they come from individual or institutional investors may not be best served by a single market based on time and price priority, particularly for securities with less liquid markets. These large orders often need capital commitment from market makers and dealers. A single time/price priority based market could result in inefficient executions through multiple break ups of that large order. By requiring a market maker, no matter how technologically advanced, to expose that large order to one or more 100 share orders, increases time and expense of execution and likely results in a priority of matching orders over capital commitment. In many cases this would be the wrong priority. One need only look to NYSE rules regarding block trades and crosses to see that even the NYSE understands that large orders may have needs better met "upstairs."

In addition, any regulatory initiatives must be based on an understanding that not all specific securities markets are the same. While a matching system with limited dealer interaction may be effective in certain liquid, high volume markets, it would likely be ineffective and lead to greater volatility in less liquid markets. In such markets, a large displayed buy limit order could result in faded offers that would ultimately require the buyer to pay significantly more. Further, under these conditions, there is a greater need for a dealer's capital commitment. Dealers are more willing to commit capital in a liquid environment where there exists a potential of profit. A single system requiring time/price priority for a few small limit orders will not provide the liquidity or the incentive necessary for a dealer to commit capital nor for an investor to be comfortable exposing a large order.

Conclusion - The Current Regulatory and Market Structure has been Effective in Addressing the Needs of a Diverse Market

CIBC strongly believes that the market's diversity of participants is its greatest strength. This diverse investor base should continue to have a "menu to choose from in the equity markets.9 We believe the current market and regulatory structure provides the framework necessary for that menu to co-exist with the protections necessary to insure best execution for all orders, and for continued market evolution primarily based upon competition. We believe that the Commission through its rescission of Rule 390 acknowledged the need for competitive markets. One also need only to look to the growth of the dealer competitive Nasdaq market to realize such competitive markets are preferred by investors as well.

We support any structure that provides investors with choices. CIBC believes that a structure that establishes a market where small market and limit orders can interact with each other for often immediate execution could exist along side a market for larger not-held orders. Elements of price improvement for small and large orders should be predicated upon a broker's duty of best execution, where price improvement is but one factor to consider, as well as the market for the particular security. In addition, careful attention should be paid to large not-held orders and their important position in the market. This approach, we believe, best fosters competition among market makers and dealers. With the proper regulatory framework in place, which currently exists, a market maker's decision to price improve, offer immediate execution or execute orders for size should be driven by its desire to compete for order flow. Competition for order flow best exists in a market with execution alternatives, minimal regulatory constraints and the efficiencies such an environment creates. Cost savings in such a market would be shared by all. In a way, perhaps the ultimate in price improvement.10

Commission concerns regarding internalization and overreaching,11 although reasonable in such a market, have already been addressed by the Commission in 1980. In response to the concern for potential market fragmentation resulting from the adoption of Rule 19c-3, the Commission stated its belief that its concerns were ameliorated by certain initiatives at that time such as the existence of accurate trade reporting and quotation information. In addition, the Commission also stated that regulation to curb potential fragmentation and overreaching, in particular, was not necessary because integrated firms assumed a fiduciary relationship with regard to its customers' securities transactions.

The Commission was confident such duty would result in a maintaining of best execution standards.12 We believe the Commission should be more comfortable today than in 1980 with regard to fragmentation and overreaching. SRO and Commission initiatives and improved technology have gone far beyond transaction reporting to help insure market integrity. Similarly, initiatives such as Manning and the Display Rule have helped further clarify a firm's fiduciary duty and best execution responsibility.

As stated previously, a single market system based on time and price priority, for example, cannot meet the needs of a diverse market. Such implementation would be a step backward and undo many of the positive benefits of rescinding Rule 390. In addition, such a system would be significantly dependent on cutting edge technology uniform throughout the market that is not currently present across market participants. As evidenced by the recent commission release regarding delaying the implementation of decimalization,13 even certain SROs may not be ready to participate in a market so dependent on technology.

Given the strong regulatory and market structure that currently exists, the Commission has the opportunity to deliberately examine all relevant issues relating to the Release. For example, questions regarding technology and decimalization and their impact should be addressed prior to adopting any of the initiatives set forth in the Release. While there may always be room for market improvement, we suggest the

Commission move cautiously. After all, as the Commission itself has stated, the "Commission should be reluctant to impose a single design on the markets absent evidence of significant market failure."14

Thank you for the opportunity to comment on the Release and the important issues it raises. Please call the undersigned at (212) 667-7387 or Elliot Levine at (212) 667-7308 if you would like to discuss the issues raised by this letter or the Release.

Sincerely,

Thomas Gallagher
Vice Chairman
U.S. Equities


Footnotes
1 See SEC, Statement of the Securities and Exchange Commission on the Future Structure of the Securities Markets (Feb. 2, 1972), Rep. No. 75, 94th Cong., 1st Sess. (1975); H.R. Rep. No. 3, 94th Cong., 1st Sess. 44 (1975); and Exchange Act Section 11A(a)(1), 15 U.S. 78R-1(a)(1)(1988).
2 Letter dated February 29, 2000 to Chairman Arthur Levitt - "A Plan for Achieving a True National Market System."
3 See NASD Rule IM-2110-2.
4 See Exchange Act Release No. 37619A (September 6, 1996).
5 Id. at 43-44
6 The Commission previously stated that a money manager, when considering his best execution obligations, may obtain better overall execution by dealing with a broker-dealer that will commit its own capital than by paying lower commission rates to another broker-dealer without comparable execution facilities. (See Letter from Richard G. Ketchum to Ronald H. Hoenig (October 15, 1990)).
7 SEC Exchange Act Release No. 42758 (May 5, 2000) approving the rescission to NYSE Rule 390. In that release the Commission noted that, "Displayed limit orders are perhaps the most significant source of price competition in the securities markets." In addition, the release states that a price/time priority system can enhance depth and liquidity of a market. Such a view does not address the needs of less liquid markets or of larger orders seeking a degree of anonymity and dealer capital commitment for best execution.
8 The Commission should also consider such proposal in light of decimalization. In addition to the technological strain on the system continuously updating orders and quotations, we do not think it is in the market's best interest to be required to trade in a structure that values small orders that improve the price by one cent over larger orders needing capital commitment.
9 See Market 2000 Study. Even in 1994 it was noted a "menu" of choices were available in the equity markets.
10 In numerous releases on market structure, the Commission noted that a benefit of adopting Rule 19c-3 is that it provides an opportunity for lower cost execution through an integrated broker-dealer's internalization of order flow (See, for example, Exchange Act Release No. 16888 (June 11, 1980) Fed. Sec. L. Rep (CCH) PP 82,608 at 83,250).
11 Id. at 83,252 at n33. The term overreaching refers to the possibility that firms may take advantage of their customers by executing retail transactions as principal at worse prices than if they acted as agent.
12 Id. at 83,256
13 See Exchange Act Release No. 42685 (April 13, 2000).
14 See pg. 3. supra.