Pursuant to section 19(b)(1) of the Securities Exchange Act of 1934 (“Act”) and Rule 19b-4 thereunder, notice is hereby given that on February 22, 2000, the Chicago Board Options Exchange, Inc. (“CBOE” or “Exchange”) filed with the Securities and Exchange Commission (“Commission”) the proposed rule change as described in Items I and II below, which Items have been prepared by the Exchange. In this proposed rule change, CBOE seeks to Start Printed Page 19402extend a pilot program that was first approved by the Commission on November 22, 1999. On March 22, 2000, CBOE filed Amendment No. 1 to the proposed rule change. The Commission received eight comment letters on the pilot program. The Exchange's response to these comment letters can be found in Item IV. The Commission is publishing this notice and order to solicit comments on the proposal from interested persons and to approve the proposal on an accelerated basis for a 6 month pilot that will expire on August 22, 2000.
I. Self-Regulatory Organization's Statement of the Terms of Substance of the Proposed Rule Change
The CBOE proposes to extend, for a 6 month period, a pilot program that provides for certain orders to be rejected from RAES for manual handling in certain limited situations. The text of the proposed rule change is available at the CBOE and at the Commission's public reference room.
II. Self-Regulatory Organization's Statement of the Purpose of, and Statutory Basis for, the Proposed Rule Change
In its filing with the Commission, the CBOE included statements concerning the purpose of, and basis for, the proposed rule change. The text of these statements may be examined at the places specified in Item V below. The CBOE has prepared summaries, set forth in Sections A, B, and C below, of the most significant aspects of such statements.
A. Self-Regulatory Organization's Statement of the Purpose of, and Statutory Basis for, the Proposed Rule Change
The purpose of the proposed rule filing is to extend, for an additional 6 month period, the pilot program that provides (where the Exchange's Automated Book Priority (“ABP”) system has been implemented) for certain orders to be rejected from RAES for manual handling in the limited situation where the bid or offer for a series of options generated by the Exchange's Autoquote system becomes crossed or locked with the best bid or offer for that series as established by a booked order. The Exchange believes this limited kick-out situation provided by the pilot program is the best alternative currently available to the Exchange to address the particular risk presented by the unusual situation where the Autoquote crosses or locks with an order in the Exchange's book. In fact, as described further below, the Exchange has found that only 0.44% of all orders (in those classes where the ABP system has been implemented) routed to RAES would be rejected pursuant to the pilot program.
The Exchange's ABP system allows an order entered into RAES to trade directly with an order on the Exchange's customer limit order book in those cases where the prevailing market bid or offer is equal to the best bid or offer on the Exchange's book. The Commission approved the Exchange's rules implementing the ABP system in October 1999, however, these rule changes do not become operative in a particular class until the Exchange implements the ABP system in that class. In those classes in which the ABP system has yet not been implemented, orders are still subject to Interpretation .04, which requires an order to be rerouted from RAES in the event that an order in the book is establishing the prevailing best bid or offer (whichever one is relevant to the particular order). The Exchange is not proposing to provide this extra protection to orders that are rejected where the ABP system has been implemented for a number of reasons. First, as the Exchange noted in its original filing, in most cases where the order is kicked out due to an Autoquote inversion, the booked order already will have been traded in open outcry before the incoming RAES orders are received. In addition, the Exchange's systems have been designed such that a rejected order will normally be routed directly to the Exchange's electronic brokerage terminal (“PAR”) in the trading crowd and will appear on that PAR machine instantaneously. Consequently, these rejected orders will routinely be represented in the trading crowd within a matter of seconds of being rejected. These orders will be entitled, by virtue of the firm quote rule, to be executed at the bid or offer displayed when that order reaches the trading station.
As described in the prior filing, in the course of planning for the implementation of the ABP system, the Exchange became aware of an unintended consequence of the operation of the ABP system. That is, the Exchange realized that in situations where the best bid or offer for one or more series of a particular class is established by one or more orders in the Start Printed Page 19403book, the market makers logged into RAES for that class of options would be subject to a substantial risk in the event that the market in the underlying stock moved significantly and quickly in a direction that made the booked order substantially better than the price calculated by CBOE's Autoquote formula. In that event, while the booked order would quickly be executed, CBOE represents that the ABP system may not be able to react quickly enough to remove the executed order from the limit order book. As a result, once ABP is implemented, orders entered in RAES would automatically be executed against the stale bid or offer still being shown in the book notwithstanding the booked order having already been executed. CBOE contends that this result could cause direct and substantial economic disadvantage to the market makers who are obligated to participate in RAES executions. The Exchange believes there is no question that the consequence of implementing ABP without addressing this substantial increased risk is that (i) market makers may choose not to participate on RAES (thus, affecting the liquidity of those lower volume series traded on RAES and endangering the viability of RAES itself) and/or (ii) market makers may request the Equity Procedure Committee to either reduce the size of orders eligible for RAES or to take some series off of RAES (thus, eliminating significant advances in automatic execution that our customers have requested).
As mentioned in that prior filing, the Exchange expected the number of orders that would be rejected from RAES under this proposed rule would represent only a small subset of the orders that were rejected in those same classes before ABP was implemented in those classes. In fact, the Exchange has found that the number of kick-outs resulting from the implementation of this system is a remarkably small percentage of the RAES-eligible orders. Of the 150 classes in which ABP had been implemented as of February 14, the Exchange found that only 44 of those classes had an ABP order on that day. Over the course of that day, 5908 orders were routed to RAES in those particular 44 classes accounting for 41,102 contracts. Of those 5908 orders, 1054 orders (representing 9017 contracts) were handled by ABP, i.e. they were traded against orders in the book and in some cases also against market makers at the price of the booked order. In all 44 classes during the course of the day, there were only 26 orders (representing 130 contracts) rejected from RAES due to the Autoquote bid or offer crossing or locking with the price of the booked order. This is, on average, less than one order per day per class that was rejected pursuant to the pilot program and amounts to only 0.44% of the orders routed to RAES in those 44 classes and only 0.31% of all the contracts routed to RAES in those 44 classes. It should also be noted that if ABP had not been implemented in those classes, all 1054 orders that were handled by ABP would have been rejected from RAES for manual handling because of the situation in which the book touches the market. With ABP in place along with the limited kick-out, only 2.46% of the orders (and 1.44% of the contracts) that would have been rejected without ABP are now rejected with ABP.
The Exchange believes that the present alternative of rejecting RAES orders in the limited situation it has described is the most effective way to provide the benefits of the ABP system without creating such a great risk to Exchange market makers that they choose not to participate on RAES, or that they encourage the appropriate Floor Procedure Committee to offer only a few active series on RAES. During the 6 month pilot period, the Exchange will continue to seek other alternatives to having these orders rejected. Among the alternatives the Exchange is presently considering are: (i) Having the Autoquote system generate an order that will be traded on RAES in those cases where the Autoquote crosses with the book value and (ii) having an income order trade against the book order at the book price for the volume in the book and then having the balance of the incoming order trade at the next best available price whether it is another booked order or against the market makers logged onto RAES at the best market maker quote whether from Autoquote or verbalized by a market maker. The Exchange will continue to search for alternatives to develop its systems to provide the best opportunities for its customers. As it is, Exchange customers who enter orders in the RAES system in those classes where the ABP system has been implemented are much less likely to have their orders rejected for manual handling today than they were before the implementation of ABP along with the limited kickout provided by the pilot program.
The Exchange is committing to provide a study each month during the pilot program detailing the number of kickouts that the Exchange experienced pursuant to the pilot program during the previous month.
2. Statutory Basis
The CBOE believes the proposed rule change is consistent with and furthers the objectives of Section 6(b)(5)  of the Act in that it is designed to remove impediments to a free and open market Start Printed Page 19404and to protect investors and the public interest.
B. Self-Regulatory Organization's Statement on Burden on Competition
CBOE does not believe that the proposed rule change will impose a burden on competition that is not necessary or appropriate in furtherance of purposes of the Act.
C. Self-Regulatory Organization's Statement on Comments on the Proposed Rule Change Received From Members, Participant, or Others
No written comments were solicited or received with respect to the proposed rule change.
III. Summary of Comments
The Commission received eight comment letters on the pilot program. All of the commenters disapprove of the pilot program and ask the Commission not to extend it. Generally, the commenters assert that the pilot program protects CBOE market makers and disadvantages retail customers. A few firms commented on the linking of the options exchanges. The linking issues, however, is not the subject of this filing.
One commenter argues that the pilot program allows CBOE market makers to abandon their firm quote responsibilities. He states that CBOE's Autoquote system does not reflect public bids or offers, but only the activity of a CBOE computer. The commenters asserts that, for example, when this system locks or crosses CBOE's bid as established by a customer limit order, the pilot program will allow market makers to abandon the prevailing public quotation, and to reject all incoming sell orders which would otherwise be entitled to trade against the best published bid. The commenter goes on to state that after these sell orders have been redirected to the crowd, these orders will most likely be executed at an inferior price.
In addition, this commenter believes that CBOE's arguments supporting the pilot program are flawed. He notes that CBOE supports the pilot by arguing that without it, market makers might avoid participating on RAES or might widen their quotes, both in response to the risk created by potential arbitrage situations. He further notes, however, that CBOE also states that it does not anticipate that the potential arbitrage situation will occur that frequently and therefore, the pilot program will have a minimal impact on the market. In sum, he argues that CBOE's support of the proposal is flawed because it simultaneously argues that makers may be exposed to tremendous risk, but the situations creating this risk will occur very infrequently.
Another commenter also refutes CBOE's arguments supporting the pilot program. In particular, the commenter notes that CBOE's fear that market makers may not participate on RAES should be balanced with some of CBOE's other RAES initiatives, such as requiring all DPMs to participate in automatic execution systems and earlier attempts to decrease the number of market makers participating in RAES. Further, the commenter addresses CBOE's argument about market makers widening their quotes by asserting that CBOE already permits double-width quotes in many volatile options classes and also allows market makers to specify a RAES size limit that is less than the class maximum.
In addition, this commenter argues that the changes to various exchanges' automatic execution systems may create public confusion and unfairly restrict customers' trading opportunities. Before the approval of the ABP system and the pilot program, the commenter asserts that public customers knew how their orders would be handled when these orders reached the CBOE floor. When the ABP system was approved, the commenter notes that CBOE deleted Interpretation .04, which provided protection for kicked-out RAES orders, because CBOE believed that the ABP system would reduce or eliminate kick-outs. However, after approval of the ABP system, the commenter points out that CBOE subsequently expanded the situations in which RAES orders could be kicked-out through a series of rule filings, including the pilot program. According to the commenter, the effect of all of these changes is that CBOE still has the ability to kick-out orders, but it no longer has a rule in place which protects these kicked-out orders.
Two broker-dealers commented that the pilot program has an adverse impact on the trading strategies of their customers. In particular, these firms maintain that they have created order routing systems that send customer orders to the market with the best price, and these order routing systems rely on firm quotes and automatic execution. They assert that the kick-out feature of the pilot program hurts their order routing systems because the price displayed by CBOE might not actually be the price that their customers receive. Further, they argue that once an order is kicked-out, their customers lose the advantages of an automatic execution system such as RAES, which according to these commenters, include the ability to modify or cancel orders online. Three other individuals also share these comments.
All of the commenters argue that the pilot program does not allow customers to take advantage of certain trading opportunities, including arbitrage situations. For example, one commenter asserts that the essence of successful options trading, and of successful arbitrage, is the identification of a pricing disparity between the theoretical price of the option and the displayed best bid or offer. This commenter believes that the pilot program, with its kick-out feature, does not allow traders to take advantage of these opportunities.
Two commenters offer suggestions on how to eliminate the need for the pilot program. One of these commenters believes that if CBOE provides additional staff to help take out the booked order when the booked order is locked or crossed by the Autoquote price, the need for the pilot program would be eliminated. The other commenter suggests that when an Autoquote price touches the price of a book order, the system should automatically execute the book order against a market maker. The commenter believes that this would eliminate the need for the pilot program because it would eliminate the possibility of a book order being locked or crossed with the Autoquote price.
In the alternative, this commenter suggests that if the pilot program is to continue, then CBOE should be required to notify broker-dealers that automatic execution is not available in a particular options series. the commenter believes that CBOE should post this notification at least three seconds prior to removing the options series from the automatic execution system. In addition, this commenter believes that the pilot program should not be extended because it gives no incentive to CBOE to fix its systems.
IV. The Exchange's Response to the Commenters
Seven comment letters were submitted on the original proposed rule change: one by Interactive Brokers; one by James Gelbort; one by George Start Printed Page 19405Brunelle on behalf of a private investment firm client; one by Thomas Coyle; one by Linda S. Tors; one by John Rohde; and one from Datek Online. It should be noted that all but three of the letters—the IB Letter, the Gelbort Letter, and the Brunelle Letter 1—were sent to the Commission after the public comment period had expired; the Datek letter was sent more than one month after the comment period ended. Nevertheless, the Exchange is addressing the arguments raised in each of the letters.
Stripped of their rhetoric and inaccuracies, these letters all essentially argue that the Exchange's proposed rule should be disapproved because it does not allow, in their opinion, for the smooth operation of a certain business model of which they presumably want to take advantage. A central theme of many of the letters is that the type of kick-out provided for by this rule (and other procedures at other exchanges) is a step backward in a technological world that is providing quicker and better access for customers to automatic execution systems. What these letters ignore is that the Exchange has continually expanded access to RAES over the last few years by increasing the eligible RAES order size, and that with this new kick-out there are actually fewer orders rejected from RAES today (not more as these letters suggest) than there were just a few months ago before the ABP system was put in place. Before the implementation of ABP in a particular class, every incoming RAES-eligible order would be rejected from RAES in those cases in which a booked order was establishing the best price on that side of the market against which the order would be traded. In those classes where ABP is in place, an incoming RAES-eligible order is only rejected from RAES if the booked order is establishing the best price on the side of the market against which the order would be traded and if the Autoquote bid or offer (as appropriate) crosses or locks with that book price.
The letters also wrongly assume that there is no public benefit to this kick-out  and that the proposal was established merely to protect the Exchange's market makers from suffering losses or to protect the market makers' “advantages.”  Again, these letters ignore the fact that, unlike the professional traders who commented on the pilot program, market makers have become subject to ever greater obligations that have been imposed by Exchange rules. In fact, the ABP system obligates the Exchange's market makers to trade up to fifty (50) contracts (the maximum RAES order size) at a price that was established by a public customer and not by the market makers.
One of the commenters suggested that the book staff have an incentive to continue to display a book price that is crossed or locked with the Autoquote system. Of course, it should be apparent from everything the Exchange has explained why the DPM book staff has an incentive to take down the already traded book price as soon as possible. The longer the book price remains, the more orders that will be sent to the Exchange trying to trade at the erroneous price and the more orders that will subsequently be rejected due to the pilot program. The Exchange's DPMs have an incentive from a customer service standpoint and for the sake of running an efficient business to ensure the displayed prices are accurate and that the prices of orders that are traded are taken down as soon as possible.
While the above discussion addresses the arguments presented in all of the comment letters, the Exchange wanted to address individually some of the letters which raise some issues that are particularly troubling because they state inaccuracies and/or misrepresent the Exchange's intentions.
Brunelle Letter 1
The Brunelle Letter 1, which was sent on behalf of a “private investment firm” who chose to remain anonymous, states that the CBOE is arguing that “the public can have RAES, or they can have the Firm Quote Rule * * * but not both.” This statement is contrary to the Exchange's rules and to Exchange practice. In fact, the Exchange's firm quote rule, CBOE Rule 8.51, states in paragraph (a)(2) that “the appropriate Floor Procedure Committee * * * may establish a different firm quote requirement for a particular class of options that is not less than the RAES contract limit and no more than 50 contracts.” By virtue of this rule, every order entered for the maximum RAES eligible size or less is entitled to firm quote treatment. This means that every RAES-eligible order, including those that are rejected in the limited circumstance permitted by the pilot program, will absolutely receive firm quote treatment whether through RAES or after having been rejected from RAES. Because the Exchange has developed systems that route those rejected orders instantaneously to electronic PAR terminals in the trading crowd, in most cases these orders will be executed at the prevailing quotes within a few seconds of when they were entered.
The Gelbort Letter states that the “CBOE does not propose to expand the ABP system to insure that booked bids or offers are, in fact, rapidly executed by crossed or locked Autoquotes.” As the Exchange has stated herein, the Exchange has in fact considered and continues to consider expanding the ABP system to have the Autoquotes trade against the booked orders. It was simply not possible at the time ABP was implemented to change the system to allow for this to happen and so the method chosen for dealing with the problem was the one with the Exchange determined was the least disruptive of those feasible alternatives.
Mr. Gelbort continues by arguing that “[e]ven in an electronic world, on-floor traders continue to enjoy significant advantages.” In fact, what Mr. Gelbort completely neglects to point out is that any “advantages” that on floor traders may have once enjoyed have been eroded over the years as customers have gained access to computers that allow them to identify opportunities for trading and have allowed them to transit orders nearly instantaneously to the floor. In fact, the Exchange has facilitated the erosion of these “advantages” by remaining at the forefront of developing systems that allow for quick access, by increasing the order size eligible for automatic execution, and by guaranteeing that RAES orders will be filled at the NBBO if the NBBO is no more than the step-up amount better than the CBOE best quote. What Mr. Gelbort also conveniently neglects to mention is that in spite of the instantaneous access to the Exchange's markets, high speed computers, and a wealth of information at their fingertips, the professional traders enjoy one enormous advantage over Exchange market makers. They have absolutely no obligation to trade at a particular price, unlike Exchange market makers. CBOE market makers who are logged onto RAES, however, are obligated to trade incoming RAES orders at the disseminated price or better when they are assigned the trade even if that price was established by a Start Printed Page 19406small order in the Exchange's book that was better than the price any CBOE market maker was willing to pay for that particular series.
Mr. Gelbort also argues that the result of these rules is to lead to “needless public confusion.” As stated earlier, however, the Exchange has already pointed out that it has gone to great lengths to inform the public of those limited circumstances where an order may be rejected from RAES pursuant to the pilot program both by filing the proposal for pubic comment and by issuing regulatory circulars on the matter. The reasons why an order may be rejected from RAES pursuant to this proposal are clearly defined and have been clearly stated. Mr. Gelbort's final paragraph on the subject rule filing, at the bottom of page 4, is a series of inaccuracies and self-serving statements. Mr. Gelbort argues that if the keystrokes have not been made to trade a booked order it is due to “inattention rather than some inherent systems delay.” In fact, at most trading stations there are traders who specifically look for situations where the Autoquotes become crossed with a booked order and trade them immediately. However, even though it takes only “a few quick keystrokes” to trade the order, this is all the time it takes for the RAES system to be flooded with orders from multiple customers. This is particularly true when the DPM staff has to trade more than one booked order at the same time. As far as Mr. Gelbort's assertion that the CBOE has been willing to provide additional book terminals and trained personnel DPMs that request them (presumably to suggest that this could solve the problem without the need for rejecting RAES orders), while this is true and remains true, this is not a solution to the particular problem. The Autoquote system may become crossed with a booked order at any time in any options class across the floor and by the time the situation arises it will be too late to transfer staff as Mr. Gelbort no doubt knows.
Mr. Gelbort continues by correctly stating that DPMs have been assigned to all equity option classes and argues that this should eliminate any concern about market makers not participating on RAES if this particular kick-out were not employed because DPMs, at least, are required to participate on RAES at all times. Mr. Gelbort's conclusion is flawed for a number of reasons. First, the Exchange does not believe it is ideal in most instances for DPMs alone to participate on RAESs. Non-DPM market makers, however, are not required to log onto RAES unless they are present in the trading pit and they have logged on at a prior time in the particular expiration cycle. In fact, to the extent market makers are logged onto the RAES system, these market makers will have an incentive to ensure that the quotes are updated and accurate. In addition, regardless of whether a DPM is logged onto RAES, if the risk involved in trading over RAES becomes so great, the DPM will likely request the Floor Procedure Committee to remove all but the most active series from RAES.
Like the Gelbort Letter, the IB Letter draws faulty conclusions from failing to have access to a number of facts. Like the Gelbort Letter, the IB Letter suggests there are better alternatives than rejecting orders from RAES when the Autoquotes cross with the price of a booked order. Interactive Brokers makes this statement without knowing what alternatives the Exchange considered (and continues to consider) and without knowing what time and effort might be involved in instituting Interactive Brokers' preferred solution to deal with the issue. The simple fact of the matter is that the Exchange, Interactive Brokers and Mr. Gelbort all share the same ultimate goal, to have the CBOE's systems operate in the most efficient manner with the fewest disruptions. However, the Exchange is also concerned about providing market makers with the proper incentives to provide the best and tightest markets for the benefit of all customers. Until the Exchange is confident that the quality of its markets will not be compromised by subjecting market makers to undue risk for which they cannot reasonably account, it should not be forced to adopt any particular methodology for dealing with the issue at hand merely because it happens to more easily accommodate the particular system designed by one firm.
Interactive Brokers' entire first argument on pages 2-4 of the IB Letter is predicated on the notion that the number of exceptions to automatic execution is growing on the options exchanges. However, as discussed previously, the number of kick-outs that result from the current pilot program is only a very small subset of the orders that have been kicked out in situations before ABP was implemented on the Exchange. It is the Exchange's judgment, however, that although it is not ideal, it would prefer the limited number of kick-outs provided for by the pilot program than to risk losing liquidity on RAES or having series taken off of RAES.
Interactive Brokers, in fact, suggests an alternative solution on page 5 to deal with the Exchange's particular concern that the Exchange is already considering. Namely, Interactive Brokers suggests that when an Autoquote price touches the price of a booked order, the system should automatically execute the booked order against a market maker. The CBOE agrees that this may well be a longer term solution to the particular issue. In light of the complexities of the RAES system and the Exchange's other current system priorities (including a conversation to decimalization), “fixing” the problem would entail more than “a few of programming work” as Interactive Brokers suggests.
Finally, Interactive Brokers argues that in lieu of disapproving the proposed rule that the Exchange be required to post in electronic form, accessible to broker-dealer routing systems, a notification that automatic execution is not available for a particular option series. Interactive Brokers argues this notice should be accessible at least three seconds prior to such options series being removed from the automatic execution system. The Exchange is, in fact, exploring having a code placed next to its disseminated quotes that indicates when the best quote for a particular series is being established by a booked order. The Exchange believes it may be able to provide such notice in the near future and this would undoubtedly benefit Interactive Brokers' system. It would not be feasible to wait three seconds to remove the series from automatic execution, however, because the instant that a booked order becomes the CBOE's best bid or offer, the market makers become subject to the risk that the pilot program was designed to manage.
V. Solicitation of Comments
Interested persons are invited to submit written data, views, and arguments concerning the foregoing, including whether the proposed rule is consistent with the Act. Persons making written submissions should file six copies thereof with the Secretary, Securities and Exchange Commission, 450 Fifth Street, N.W., Washington D.C. 20549-0609. Copies of submission, all subsequent amendments, all written statements with respect to the proposed rule change that are filed with the Start Printed Page 19407Commission, and all written communications relating to the proposed rule change between the Commission and any person, other than those that may be withheld from the public in accordance with the provisions of 5 U.S.C. 552, will be available for inspection and copying in the Commission's Public Reference Room. Copies of such filing will also be available for inspection and copying at the principal office of the CBOE. All submissions should refer to File No. SR-CBOE-00-03 and should be submitted by May 2, 2000.
VI. Commission's Findings and Order Granting Accelerated Approval of Proposed Rule Change
After careful review, the Commission finds that the proposal is consistent with the requirements of the Act. In particular, the Commission finds the proposal is consistent with Section 6(b)(5)  of the Act. Section 6(b)(5) requires, among other things, that the rules of an exchange be designed to remove impediments to a free and open market and to protect investors and the public interest.
In extending this pilot, the Commission has balanced the commenters concerns with those expressed by CBOE. The Commission notes that CBOE has provided figures the show that kick-outs under this pilot program occur infrequently. Specifically, on February 14, 2000, CBOE conducted a study to determine how often kick-outs from RAES occurred as a result of this pilot program. On that date, CBOE found that out of the 150 classes for which the ABP system had been implemented, only 44 of those classes had an order executed through the ABP system, i.e., the RAES order interacted with an order on the limit order book. In those 44 classes, 1054 orders (representing 9017 contracts) were executed through the ABP system. In those same 44 classes, only 26 orders (representing 130 contracts) were rejected from RAES due to the Autoquote system locking or crossing CBOE's best bid or offer as established by the book. Moreover, the orders rejected from RAES as a result of this pilot represent a small percentage of the total amount of orders routed to RAES in these 44 options classes on February 14 (5908 orders representing 41,102 contracts). These figures support CBOE's position that kick-outs under this pilot program occur infrequently.
Nevertheless, the Commission is mindful of the commenters concerns. In particular, the Commission agrees with the commenters that there are other solutions than the one employed by CBOE in this pilot program. In this filing, CBOE listed two alternative solutions. One of these alternatives involves having an incoming order trade against the book order at the book price for the volume in the book and then having the balance of the incoming order trade at the next best available price—whether it is with another booked order or with a market makers logged onto RAES. This alternative would allow customer orders to interact with orders on the limit order book, but would eliminate the risk to market makers of executing a RAES order for the maximum eligible size when the limit order is for a smaller number of contracts. In this regard, the CBOE has represented that it will continue work on systems changes to address the situation when the Autoquote system locks or crosses CBOE's best bid or offer as established by the book and has assigned a high priority these systems changes. CBOE stated that it is confident that these changes could be implemented by the end of this calendar year, after it has completed the projects needed for it to convert to decimal trading.
In the meantime, the Commission agrees with one of the commenters that CBOE should provide protection to kicked-out orders in options classes where the ABP system has not yet been implemented. When the ABP system was originally proposed, CBOE represented that the ABP system, by allowing RAES orders to interact directly with orders in the exchange's limit order book, would reduce or eliminate the need for kick-outs. Because of this representation, CBOE eliminated Interpretation .04, which provided protection for orders that had been kicked-out. As of the date of this filing, CBOE has not implemented the ABP system on a floor-wide basis. The Commission therefore believes that Amendment No. 1, which re-adopts Interpretation .04, should help provide protection to orders kicked-out in those classes in which the ABP system has not been implemented. CBOE also stated that it would continue to roll out the ABP system in those classes in which it had not yet been implemented.
In light of the likely benefits to customer limit orders expected to be gained by the continued implementation of the ABP system, the Commission finds good cause for approving the proposed rule change prior to the thirtieth day after the date of publication of notice thereof in the Federal Register. Further, the Commission notes that the CBOE has agreed to provide monthly reports to the Commission regarding the number of times an incoming RAES order is rejected pursuant to this pilot.
It is therefore ordered, pursuant to Section 19(b)(2) of the Act, that the proposed rule change (SR-CBOE-00-03) is hereby approved through August 22, 2000.Start Signature
For the Commission, by the Division of Market Regulation, pursuant to delegated authority.
Margaret H. McFarland,
3. See Release No. 34-42168 (November 22, 1999), 64 FR 66952 (November 30, 1999) (File No. SR-CBOE-99-61).Back to Citation
4. In Amendment No. 1, the CBOE amended the filing to respond to questions from the Commission staff and to incorporate these responses into the text of the rule filing. In addition, the CBOE proposed to adopt an Interpretation that provides protection for orders kicked out of RAES when the prevailing market bid or offer is equal to the best bid or offer on the Exchange's book. This Interpretation, which was part of CBOE's rules until October 1999, would apply to option classes where the Automated Book Priority system has not been implemented (Interpretation .04 to CBOE Rule 6.8). See letter from Timothy Thompson, Director, Regulatory Affairs, CBOE, to Elizabeth King, Associate Director, Division of Market Regulation, Commission, dated March 21, 2000 (“Amendment No. 1”).Back to Citation
5. See letters from George Brunelle, Law Offices of George Brunelle, to Secretary, Commission, dated December 20, 1999 (“Brunelle Letter 1”); James I. Gelbort, to Jonathan G. Katz, Secretary, Commission, dated December 21, 1999 (“Gelbort Letter”); Thomas Peterffy, Chairman, and David M. Battan, Vice President and General Counsel, Interactive Brokers, The Timber Hill Group, to Jonathan G. Katz, Secretary, Commission, dated December 21, 1999 (“IB Letter”); Linda S. Tors, to Jonathan G. Katz, Secretary, Commission, dated January 6, 2000 (“Tors Letter”); Thomas Coyle, to Jonathan G. Katz, Secretary, Commission, dated January 3, 2000 (“Coyle Letter”); John Rohde, to Jonathan G. Katz, Secretary, Commission, dated January 9, 2000 (“Rohde Letter”); Brent Houston, Senior Vice President, Capital Markets, Datek Online, to Jonathan G. Katz, Secretary, Commission, dated February 1, 2000 (“Datek Letter”); George Brunelle, Brunelle & Hadjikow, to Jonathan G. Katz, Secretary, Commission, dated March 23, 2000 (“Brunelle Letter 2”). The Division of Market Regulation received Brunelle Letter 2 on March 28, 2000. In Brunelle Letter 2, the commenter generally reiterates the comments from his previous letter (Brunelle Letter 1) and also comments on another CBOE rule filing, SR-CBOE-99-57.Back to Citation
6. In the event that the order in the book is for a smaller number of contracts than the RAES order, the balance of the RAES order would be assigned to participating market makers at the same price at which the rest of the order is to be executedBack to Citation
7. See Release No. 34-41995 (October 8, 1999), 64 FR 56547 (October 20, 1999) (File No. SR-CBOE-99-29).Back to Citation
8. As of February 10, 2000, ABP has been implemented in over 150 classes of equity options on the Exchange floor, including many of the most actively traded option classes. ABP has been implemented in options classes at every trading station on the floor. As the Exchange has noted to Commission staff, the Exchange will continue to roll out ABP to the other option classes on the floor in any orderly manner—in a manner designed to ensure the continued integrity of the ABP system.Back to Citation
9. In those classes where ABP has not yet been implemented, when a RAES order is entered into the Exchange's Order Routing System at a time when the prevailing market bid or offer is equal to the best bid or offer on the Exchange's book, the order generally is routed electronically to a Floor Broker's terminal or work station in the crowd subject to the volume parameters of each firm. Today, the orders are routed to the Floor Brokers instead of being automatically executed in the crowd at the market price, because execution with the crowd would be inconsistent with CBOE Rule 6.45, which provides that bids or offers displayed on the customer limit order book are entitled to priority over other bids or offers at the same price. Until ABP is implemented in the particular class, the first such order rerouted from RAES due to a situation in which the book touches the market is entitled to be filled at the prevailing quote at the time the order was rerouted. See Amendment No. 1.Back to Citation
10. CBOE explains the potential risk market makers could be subject to by implementing the ABP system without the proposed “carve out” by way of example. Assume that in a volatile stock (where the maximum order size for RAES has been established at 50 contracts) small customer orders in the book are establishing the best bid in six different series. In one particular series, Series A, assume that the CBOE market is 5 (bid)—51/8 (offer), with a book order to buy 5 contracts at $5 (which establishes the best bid). Assume further that the price of the underlying internet stock drops precipitously in a matter of seconds. When the underlying moves, the Exchange's Autoquote system will also update CBOE market makers' quotes for the options overlying that stock. Assume with the drop in the underlying, the Exchange's Autoquote system establishes a bid and offer of 43/4-7/8 for Series A. (The same scenario would play out with the other five series whose best bid is established by an order in the book.) The order in the book representing the best bid will likely be immediately executed by the crowd in the auction market. For some period of time after the trade has been consummated in open outcry, however, the bid will still be displayed as CBOE's bid while the Order Book Official physically punches the keys to take the bid down from the display. During this period, the displayed bid of 5 in the book will be out of line with the theoretical bid of 43/4 generated by CBOE's Autoquote system. In the meantime, traders who have equipped themselves with the necessary computer equipment and communications facilities could have identified the pricing disparity between the theoretical price of the options and the displayed best bids, could automatically generate orders to sell the affected options and route those orders to RAES. If RAES is allowed to operate as it does under normal circumstances, each order to sell that arrives at the Exchange from these investors, for so long as the out-of-line book bid continues to be displayed, will be assigned to market makers in the trading crowd who are logged on to RAES. These market makers in turn will be obligated to buy at the $5 bid, which could now be significantly away from the theoretical bid. Of course, the same adverse consequences could be experienced in the other five series of the class in which the bid was established by a booked orderBack to Citation
11. In those 44 classes in which an ABP order was received, 26 orders were rejected. While there was a limited concentration of the kickouts in certain classes, no class had more than 5 kickouts for the entire day. Of the 26 rejects, 19 of them occurred in five classes as follows: CSCO (Cisco Systems)—5, YHOO (Yahoo! Inc.)—4, CMGI (CMGI Inc.)—4, AOL (America Online, Inc.)—3, QCOM (Qualcomm Inc.)—3.Back to Citation
12. Of course, a more revealing statistic might be the percentage of RAES orders rejected compared to all RAES orders received in those 150 classes in which ABP had been implemented, not only those classes in which an ABP order was received. The percentages for the 150 ABP classes would be significantly lower than they are for the 44 classes alone.Back to Citation
14. See supra note 5.Back to Citation
15. See, e.g., Brunelle Letter 1, Gelbort Letter, Tors Letter, Rohde Letter and Datek Letter.Back to Citation
16. See IB Letter, Datek Letter.Back to Citation
17. See Brunelle Letter 1.Back to Citation
18. See Gelbort Letter.Back to Citation
19. See IB Letter, Datek Letter.Back to Citation
20. See Tors Letter, Rohde Letter, Coyle Letter.Back to Citation
21. See Brunelle Letter 1.Back to Citation
22. See Gelbort Letter.Back to Citation
23. See IB Letter.Back to Citation
24. CBOE did not receive a copy of Brunelle Letter 2.Back to Citation
25. As described above, only 2.46% of the orders (1.44% of the contracts) rejected before the implementation of ABP are rejected pursuant to the pilot program. If the number of rejected orders were compared to all RAES orders in those classes in which ABP had been implemented these percentages would be even smaller.Back to Citation
26. See Brunelle Letter 1 at 1, “Without any countervailing benefit to the public markets. * * *”Back to Citation
27. Id.Back to Citation
28. See IB Letter.Back to Citation
29. The Exchange estimates that for one series it will generally take the DPM book staff 1 to 4 seconds to complete the transaction. Of course, there are some instances where more than one booked order may be traded at the same time. As soon as the booked order is traded, the book-Autoquote inversion will generally cease to exist and all incoming RAES orders after that point will be automatically traded and not rejected from RAES.Back to Citation
30. In addition, pursuant to Section 3(f) of the Act, the Commission has considered the proposed rule's impact on efficiency, competition, and capital formation. 15 U.S.C. 78c(f).Back to Citation
32. See Amendment No. 1 at 2.Back to Citation
33. The extension of this pilot should not be interpreted as suggesting that the Commission is predisposed to approving the proposal permanently.Back to Citation
[FR Doc. 00-8880 Filed 4-10-00; 8:45 am]
BILLING CODE 8010-01-M