December 18, 2001

Jonathan Katz
U.S. Securities and Exchange Commission
450 5th St. NW
Washington D.C. 20549

RE: Release No. 34-44568; File No. S7-14-01 [SEC Concept Release: Request for Comment on the Effects of Decimal Trading in Subpennies]

Dear Mr. Katz:

Island appreciates the opportunity to comment on the Commission's Concept Release on Sub-penny Trading ("Concept Release"). As an ECN currently operating within the Nasdaq market, Island now accounts for more than one out of every five trades on Nasdaq. Further, approximately 40% of the orders submitted and 35% of the executions on Island occur in sub-penny increments. Island currently accepts orders priced out to three decimal places. Accordingly, Island's continued growth casts substantial doubt on many of the claims of commentators that assert that sub-penny increments would cause investor confusion and harm transparency. To the contrary, Island's experience is that investors understand that sub-penny increments provide an opportunity to lower transaction costs and bring further efficiencies to the market. Based on our experience with sub-penny increments, Island believes that the Commission should not only continue to permit sub-penny trading but should also move forward expeditiously in requiring quotations of at least three decimal places in the publicly disseminated quotation.


Although the equity markets completed their transition to decimals in April of 2000, ECNs such as Island have been offering sub-penny trading for much longer. Specifically, prior to decimalization, ECNs traded in increments of 1/256th or $.00390625. Even though many ECNs traded in increments down to 1/256th, the public quotation only went down to 1/16th or $.0625. Thus, the mandated minimum quoted spread was approximately six cents. The six cent spread between the highest bid price and the lowest offer price provided a built in margin for certain market intermediaries.

At the time that a transition from fractions to decimals was being considered, there was substantial opposition from a large portion of the securities industry. It appears that much of the opposition, though cloaked in such arguments as investor confusion and capacity concerns, may have ultimately been due to the reduction in the mandated minimum spread. In particular, the transition from fractions to decimals also constituted a mandatory reduction in the minimum spread from $.0625 (1/16th) to $.01 (1/100th). The approximately six-month time period since April reflects the serious economic consequences of this reduction in the mandated spread. During that period, numerous leading market participants have announced lower earnings or shut down their trading desks altogether. Given the effects of decimalization on a large segment of the broker-dealer community, it is not surprising that much of the industry is against sub-penny trading.

The task of the Commission, however, is to determine what is in the best interests of investors and the continued long run health of our Nation's equity markets. Thus, in considering the issue of sub-penny trading, it is imperative to consider the broader policy implications of mandatory minimum increments. The first fact that must be recognized is that any effort to mandate a minimum increment constitutes price fixing. To understand why this must be true, consider the following analogy. Suppose that the soft drink companies successfully pushed through legislation requiring that soft drinks could only be sold in increments of $1. It is reasonable to assume that most consumers and economists would be against such a law. Consumers would be adamantly opposed because they would pay much more for soft drinks than they would absent the regulation. Economists would be opposed since the price of soft drinks would be set other than by a free market, distorting forces of supply and demand. Neither group would change their views if the soft drink companies argued that: 1) the impact of the law is less than a dollar per can; 2) vending machine companies incur substantial costs in stocking vending machines with change; 3) dollar increments would simplify soft drink pricing.

While most would agree that a minimum increment of $1 for soft drinks would be unacceptable, many fair-minded individuals appear to be considering prohibiting sub-penny trading for equity securities. The only difference between the soft drink scenario and the equity markets is one of degree. In other words, it is clear that the soft drink would cost consumers a "meaningful" amount of money relative to the price of a soft drink. Who decides, however, how much is "meaningful" to investors? For example, on a 1,000 share order the third decimal place can be worth anywhere from $1 to $9. Therefore, by limiting trading to 2 decimal places, somebody would be deciding that $9 is not relevant to investors. Given that investors change on-line brokers to save a few dollars on a trade, it is doubtful that anyone can honestly claim that $9 is irrelevant to investors.1 Since it is difficult to argue that $9 is irrelevant, the only argument is that the costs associated with sub-penny trading exceed the benefits which we know to be cost savings to investors.

At this point, therefore, it is useful to examine some of the "costs" that have been advanced in support of prohibiting sub-penny trading.

1) Adverse Impact on Market Transparency

According to this argument, sub-penny trading harms transparency because, with more increments, there will likely be more price points, obscuring the price at which market participants can purchase a "significant" number of shares. For example, assume the buy limit orders on the book are as follows:

Buy Orders
$20.003 1,000 shares







In this scenario, the highest bid price disseminated to the public would be $20.003 since only the highest bid is currently disseminated by most markets. Some commentators claim that the dissemination of a $20.003 bid is misleading because the majority of the trading interest is at $20. They complain that, as a result of the finer increments, that they cannot accurately gauge market depth and liquidity.

With respect to transparency and accurately measuring depth and liquidity, the actual cause for any potential confusion as to the "true" inside market, however, is the method in which quotation data is currently disseminated. The limitations associated with our current quotation dissemination conventions have only been magnified with the transition to decimal trading. Specifically, there are many ways to disseminate market data that would allow market participants to know the full depth of the book. Indeed, both the Nasdaq and NYSE have announced initiatives to provide investors with information concerning multiple price levels. Changing the way we view quotations or simply showing more depth of book information (Island currently shows its 15 best orders on its BookViewer) would eliminate this problem. Island believes that it would be a mistake to base a decision as to what is in the best interests of the long run health of our securities markets on the limitations of a quotation convention that has changed little in more than 100 years.

Thus, in the above example, if all markets disseminate depth of market information, a market participant that wants to sell 20,000 shares could send an order in to sell 20,000 shares at $20. The market participant would receive executions from $20.003 down to $20. The price improvement from the sub-penny increments would save the market participant $6. The extra depth of book information would allow the market participant to price the order to ensure that all 20,000 shares were immediately executed.

2) Investor Confusion

Many commentators argue that investors would be confused by anything more than two decimal places. This is an argument hard to prove either way but there is nothing inherently difficult in understanding base ten numbers. Certainly, it will be much easier than fractional pricing which was only recently eliminated.

3) Too Many Executions at Too Many Prices

Similar to the market transparency argument in that some market participants claim that multiple executions at multiple prices are too burdensome from an administrative standpoint. Some critics point, for example, to the possibility of multiple confirmations that could lead to confusion and higher costs. The issue of confirmations, like many of the objections to decimal pricing and now sub-penny pricing, relates to conventions in the industry that can be changed. By itself, an issue like confirmations should not be used to determine issues that affect our Nation's competitiveness in equity trading. Rules regarding confirmations, for example, can be amended or interpreted to allow average price confirmations rather than sending a separate confirmation for each and every confirmation. Changing the confirmation rules also may be appropriate since the key consideration for an investor is the total price of the transaction rather than the number of prices or executions it took in executing the full amount of the order. Other conventions such as charging for clearing based on the number of executions can be changed to a per share basis.

4) Technological Limitations

Some commentators have expressed concerns that sub-penny trading would strain capacity limits of both market participants and market data vendors. As with some of the other concerns, Island believes it would be a mistake to base a decision that could affect our Nation's global competitiveness based on the system limitations of some industry participants. As the technological advances of the past decade reflect, capacity limitations are, at worst, a short-term issue. The continual increase in processing power and bandwidth will more than alleviate any capacity concerns. Indeed, approximately 40% of the orders displayed on Island are in sub-pennies and Island has not experienced any capacity related problems. Island is also unaware of any difficulties experienced by Nasdaq market participants in accepting or processing executions from Island in sub-penny increments.

5) Discouraging Limit Orders and Stepping Ahead

Some commentators are concerned that sub-penny increments would reduce transparency by discouraging the use of limit orders. In support of this position, commentators cite studies on the effects of decimal trading on the New York Stock Exchange ("NYSE") that show the reduction in the number of limit orders displayed on the NYSE book. It is difficult, however, to draw any conclusions from these studies. One explanation is that these studies may merely be measuring the impact of decimals on the NYSE. In other words, given that the NYSE only displays its best price, NYSE members may believe that it is better to send their orders to floor brokers that are in a better position to assess actual supply and demand than to send limit orders electronically. If, however, all NYSE members saw all quotation information at the same time (the increased use of floor brokers suggests that floor brokers have better information that the public at large) perhaps the number of limit orders would not have decreased. Island, for example, which disseminates depth of book information at the same time, has seen its volume and the number of limit orders substantially increase since the introduction of decimal trading. This is despite the fact that Island allows orders to be entered out to three decimal places as compared to two decimal places on the NYSE.

Some commentators also raise the related issue of stepping ahead. Specifically, they are concerned that since the price of taking priority is less in a sub-penny environment that market participants posting limit orders are more likely to see another market participant step ahead. By analogy, imagine you went to a car auction and the minimum bidding increment was a penny. The cost of waiting to see the highest bid and merely bettering it is reduced.

In the securities market, however, it is currently not even necessary to outbid another market participant to take priority. For example, a market participant could post the highest bid on the NYSE but see numerous transactions occur on regional exchanges without receiving an execution. Thus, trading ahead can currently occur at the same price as a limit order. If trading ahead can occur at the same price, the minimum increment becomes irrelevant in terms of discouraging limit orders. Finally, Island has not experienced any reduction in the number of limit orders despite the fact that Island permits orders to be entered with three decimal places.


Island believes that the change to decimal trading has strengthened our securities markets and benefited investors. Island believes that further benefits can be realized by permitting sub-penny trading. Accordingly, Island strongly believes that the Commission should not only refrain from limiting sub-penny trading but should increase the number of decimal places displayed in the Consolidated Quote to at least three and possibly four places. Any regulatory effort to limit sub-penny trading would be to arbitrarily decide how much money is relevant to investors. Such a decision is better left to the marketplace than regulation. Island thanks the Commission for its consideration.


Cameron Smith
General Counsel


1 In addition to the monetary savings, another benefit of decimal trading has been the reduction in traditional payment for order flow arrangements. Prior to decimal trading, payment for order flow was financed, in large part, by a market participant that was simply rebating part of the 6 cent spread back to the broker providing the order. With 1 cent spreads now becoming commonplace for certain securities, payment for order flow has been reduced. Given the Commission's historical discomfort with the practice, this should be viewed as another benefit of decimal trading.