Chicago Board Options Exchange, Incorporated
September 16, 2003
Via Facsimile and DHL
Mr. Jonathan G. Katz
Re: Securities Exchange Act Release No. 48355; File No. SR-BSE-2002-15
Dear Mr. Katz:
This letter and the attached document set forth the comments of the Chicago Board Options Exchange, Incorporated ("CBOE") with respect to the referenced release and rule filing regarding a proposal by the Boston Stock Exchange, the Bourse de Montreal Inc., the Interactive Brokers Group, and, to a lesser extent, other brokerage firms, to trade listed options via an entity called the Boston Options Exchange ("BOX"). CBOE previously commented on this controversial proposal in a letter dated February 14, 2003, from the undersigned, to the Secretary of the Securities and Exchange Commission ("Commission").1 In that letter we observed, among other things, that the cornerstone of the proposal was to garner business by adopting rules that would greatly increase the practice of "internalization".2 This would be accomplished, in large part, by allowing internalized orders to trade in a very limited, three-second auction in a one-cent increment that is not used in options trading today and that is not proposed to be used for quoting and trading on BOX except in the case of internalization. The recent amendment to the proposed BOX rules does not correct the deficiencies previously identified by CBOE.
The attached comments address numerous concerns we have with the proposal, but our primary focus is the danger posed to the marketplace when internalization is allowed to become even more prevalent than it is today in options trading, as well as the danger of allowing trades in one-cent increments for internalization purposes when five and ten-cent increments are not only used universally throughout the options industry, but are the proposed BOX standard. These are complicated issues, and the way they are handled by the Commission will have far-reaching implications for the options industry and for investor protection. In fact, Chairman William Donaldson and previous Commission Chairmen have expressed concern with the growing practice of internalization:
Because of the complicated nature of the issues involved and the fact that our comments will be publicly available on the Commission website, the attached 9-page document will attempt to explain what internalization is, how the internalization component of BOX will work, and why the issues posed by BOX need serious examination before certain aspects of BOX are allowed to proceed.
Although we are troubled that what amounts to a new options exchange is being proposed through a standard rule change filing (as opposed to an exchange application) by a stock exchange that has never traded options and that only partially owns the proposed new exchange, we are not necessarily opposed to the addition of BOX as an options exchange. Instead, our concerns primarily relate to the BOX's proposed internalization mechanism, the Price Improvement Period ("PIP"), and we strongly believe that the PIP should not be approved. At a very minimum, the issues raised by the PIP, namely increased internalization, increased payment for order flow, and the use of penny pricing for internalization purposes only, need to be further examined before the consequences of approving the PIP and the various competitive responses to the PIP that will be filed by other options exchanges are unleashed on the marketplace.
CBOE believes it would be unwise to allow the PIP to proceed (even as a pilot-program) before the Commission follows through on its previously stated intention to issue a concept release on the effects of internalization and payment for order flow in the options markets. It does not make sense to allow what are universally considered controversial practices to manifest themselves even more in the options markets before a planned study on those practices is complete. Only the Commission is in a position to adopt a comprehensive policy concerning these complicated issues, and that policy should not be dictated by piecemeal decisions in response to different exchange rule filings. Indeed, in an interview with Institutional Investor magazine (August 2003, Volume XXXVII, No. 8, page 37) Chairman Donaldson indicated that, with respect to market structure issues, the Commission should adopt an overall philosophy so that individual decisions do not lead the markets in unintended directions. We couldn't agree more.
The consequences of allowing the PIP go beyond approval of a rather novel but anti-competitive and non-transparent internalization scheme that will likely increase options volume on the BOX exchange at the expense of best execution obligations of broker/dealers who introduce their order flow to BOX. It is likely that other exchanges, including CBOE, will be incented to respond with their own internalization mechanisms to satisfy the demands of internalizers thereby continuing BOX's escalation of the "conflicts crisis" on Wall Street.3 The Commission will then be forced to either approve these proposals, with the concomitant degradation of execution quality for investors, or be forced to engage in ad hoc line-drawing as each new internalization iteration is proposed. Either set of actions would certainly be damaging to the markets.
Similarly, exchanges will rush to quote and trade options in pennies (or subpennies) to negate the one-cent increment component of the PIP. With the effects of a penny minimum price variation in the stock market (which was initially only intended as a pilot program) still being debated, and the looming issue of subpennies facing the Commission, it is dangerous to thrust the options market into these trading increments through approval of the PIP. We strongly believe that the issues faced in the stock markets will be compounded significantly in the options markets where hundreds of thousands of options series exist, all with their own quotes. Moreover, the potential impact of one-cent increments on options message traffic and capacity has not been studied since the conversion to decimals, and the potential adverse impact to quote vendors and the options markets generally should be considered before BOX is allowed to introduce pennies into the options markets.
We recognize that the Commission's task in considering the BOX proposal is a difficult one. Nevertheless, the dangers from the PIP are so clear and real that we urge the Commission and its staff to require that the BOX proposal be amended to eliminate the PIP before BOX can be approved.4 Then the Commission can consider these significant market structure and investor protection issues in a comprehensive manner rather than through an isolated filing.
As the original options exchange we fervently believe that approval of the PIP will lead the options markets into a place where internalization is the rule and fulfilling best execution obligations is the exception. We stand ready to work with the Commission and its staff to address these difficult issues. Any questions regarding our comments should be directed to Joanne Moffic-Silver, General Counsel & Corporate Secretary, at 312-786-7462, Angelo Evangelou, Senior Attorney, at 312-786-7464, or the undersigned.
cc: Chairman William Donaldson
September 16, 2003
Chicago Board Options Exchange Comments on File No. SR-BSE-2002-15
The Boston Options Exchange ("BOX") consortium (comprised primarily of the Boston Stock Exchange ("BSE"), the Bourse de Montreal Inc., and the Interactive Brokers Group) proposes to add a sixth options exchange which would be characterized as a "facility" of the BSE. BOX would be an all electronic options marketplace with no real "membership" structure, no specialists or primary market makers with greater obligations than regular market makers,1 and, generally, a price-time priority sequence that makes no distinctions between public customer and professional orders when matching orders (CBOE currently requires customer priority). If these were the only unique components of the BOX proposal, the approval of BOX might well add new and interesting features to the active competition that already exists among options exchanges.
The primary purpose behind BOX, however, is to maximize trading volume and profitability by embracing a controversial practice called internalization. Recognizing that certain brokerage firms are increasingly eager to capture the value from their customers' options orders by either taking the other side of such orders when it would be profitable for the firm to do so based on pricing models and market conditions (i.e. internalize the orders) or by selling the execution responsibility for the orders in bulk to certain liquidity providers/market making firms,2 the BOX proposed the Price Improvement Period ("PIP") which would allow internalizing brokerage firms and market making firms that pay for order flow to take internalization to a new level.
This would be accomplished, in large part, by allowing such firms the ability to internalize orders in a rapid-fire three-second process that can occur at a price that is only one cent better than the prevailing market price (as opposed to more meaningful and standard five and ten-cent increments). This "improvement" is how BOX is marketing its internalization machine as a "win" for customer orders while masking the detrimental effects the PIP will have on such orders and the options markets in general. In explaining those detrimental effects, CBOE believes it is important to fully understand not only how the PIP works but how internalization works.
With the growth in the listed options business over the past several years, certain brokerage firms have developed significant options trading expertise that has motivated them to attempt to capture the returns available from options market making without taking on the various obligations required of market makers. Most notably, their efforts in this regard have been concentrated in their taking of the other side of customers' orders.
Internalization arose, initially, in response to periodic liquidity needs driven by large customer orders. While this type of internalization still exists, the practice has since expanded to now take on one of three different forms: accommodative internalization, selective internalization or systematic internalization.
Accommodative internalization occurs when a broker/dealer responds to the needs of a customer to effect a large or complex options transaction that exceeds the liquidity available in the market. The broker/dealer steps in to provide additional depth by offering to take the other side of the customer's order, performing what is known as "facilitation," which is the committing of firm capital to execute a customer's order. It is important to note that the primary purpose of this type of internalization is not to pursue a profit opportunity commensurate with the contra-position, but instead to assist a customer in filling an order. The customer may or may not be charged a fee of some kind, depending on the relationship.
Selective internalization is a practice that has emerged more recently as broker/dealers realized that they could pre-select transactions that hold greater profit potential for the contra-party (a practice characterized by some market participants as "cherry picking") and take the other side of those orders. In making this assessment, broker/dealers may be increasingly focused on a profit-oriented approach. A consequence of this practice has been to increasingly exclude professional liquidity providers (exchange market makers with statutory obligations to make markets and provide liquidity for all eligible orders whether potentially profitable or not) from a meaningful opportunity to participate in trades.
Further, since most option classes are listed on more than one options exchange, broker/dealers have the ability to play one marketplace off against another by seeking a greater share of participation from market maker trading crowds, indicating that in order to receive any part of an attractive order, market makers have to compete with other exchanges in terms of how much of an order they would be willing to give up to the broker/dealer. More recently, this practice has evolved to include so-called "if-then" orders, in which broker/dealers may delay transmission of an order to any marketplace until they have determined in which venue their participation will be greatest. The concerns raised by this practice have been exacerbated by a lack of uniformity among the exchanges in dealing with "if-then" orders thereby creating a regulatory arbitrage situation that is not favorable to investors.
Systematic internalization is the newest form of the practice and is similar to the practice of selective internalization described above. However, in this case the broker/dealer's assessment of profit potential is not limited to a single order, but includes an entire stream of orders (for example, all small retail orders from a specific source). This order flow may be directed, en masse, to a liquidity-providing firm either owned by the broker/dealer or one with which a profit sharing agreement exists. Typically, order flow is also routed to a venue where the participation level for that affiliated liquidity provider is the greatest, whether by rule or by virtue of the paucity of competing participants. BOX proposes to offer the most lucrative participation level via the PIP. Indeed, the BOX proposal evidences a trend among new options entities to create business models predicated on encouraging, institutionalizing, and servicing internalization.
Internalization on Existing Options Exchanges
While internalization is far more limited on the existing exchanges than it would be on the proposed BOX exchange, it is currently a growing practice and CBOE urges the Commission to conduct a comprehensive review of selective and systematic internalization as well as payment for order flow before they are allowed to expand further. Nevertheless, because internalization does exist today, it is constructive to explain how it works on the existing options exchanges. The options exchanges generally have adopted some form of a 20/40 standard for internalization based on what has been approved by the Commission. This means a firm can internalize at least 20% of a customer order (for which it is supposed to try to get the best execution) at the prevailing market price (the best bid/ask quoted by market makers and other participants) or a firm can automatically internalize at least 40% of the order if it improves on the prevailing market price by at least one minimum increment. Today, the minimum increments are five cents for options priced under $3 and ten cents for options $3 and over.3 The 20/40 percentages are minimum guarantees, and it is possible for the firm to internalize a larger percentage if no other market participants are interested in matching the internalized price . Also, currently, only orders larger than 50 contracts may be subject to the 20/40 guarantees. Here is an example of how these guarantees work:
Example: Assume a brokerage firm has determined that it stands to profit by trading against a 100-contract order to buy option XYZ it has received from a public customer. The firm's broker requests a market in the options series. The CBOE market in that series, which happens to be the national best (NBBO), is a $3.50 bid and a $3.80 offer. The firm could participate in 20% of the order at 3.80. In order to attain the higher participation amount of 40%, the firm would have to quote an offer of 3.70 which could be matched by other market participants. Thus, to get a higher percentage, a ten-cent improvement is required.
Alternatively, a brokerage firm may choose to not utilize the 20/40 rules and may attempt to internalize orders in larger percentages by identifying the quote prices and sizes at the various options exchanges trading the product and assessing the likelihood of market makers at each exchange expressing an interest in participating with the firm. If it determines that little resistance will exist at a particular price point on a particular exchange, it may seek to internalize as much as possible on that exchange. This may result in a larger percentage for the firm than had the firm invoked the 20/40 rule, but making such assessments on a case-by-case basis involves a significantly slower and more cumbersome process than what is proposed by BOX. This sort of practice also calls into question whether a firm is living up to or ignoring its obligations as agent for the customer. It is important to note that, in connection with these practices, CBOE has initiated disciplinary actions in instances where firms failed to exercise due diligence in handling these types of customer orders. Nevertheless, without the involvement of the Commission in defining the ground-rules, the result is regulatory arbitrage among the exchanges.
Under either alternative, it is possible that brokerage firms that internalize are not always obtaining the best price possible for the customer order (e.g. why not start trying to get the customer to buy at 3.60 instead of 3.80, 3.70, or 3.79 as we will see in the case of BOX). This raises the question of why would brokerage firms have an interest in seeking the most favorable price for the customer if they are on the other side of the trade. It is evident that the practice is fraught with conflicts of interest. BOX now seeks to expand and systematize this conflict.
Like every other options exchange, the BOX will trade and quote in five and ten-cent increments. However, when it comes to internalization, BOX will allow internalized orders to trade in one-cent increments. The process will be very fast, secretive, automated, and free of many of the impediments to internalization that exist today on other exchanges. Here is how it will work and save the internalizer (while costing the customer) 4 cents or 9 cents per "improved" order.
Assume the best market in option XYZ on BOX is $3.50 - $3.80 and is also the best price nationally. A brokerage firm (or a market making firm that has purchased order flow from a brokerage firm) that seeks to internalize a customer order to buy 100 contracts of option XYZ may initiate a PIP.4 Instead of exposing the customer buy order at 3.60 to gauge interest at a price more favorable to the customer, the PIP would allow the firm to enter a 3.79 offer for 100 contracts. This offer would not be posted in the BOX quote which is visible to the public (nor would this liquidity be available to the public or anyone else). Instead it would only be visible to certain BOX participants including BOX market makers registered in XYZ. Further, this limited "exposure" would last for exactly three seconds. During that three-second period, the BOX participants that are privy to this particular PIP would be allowed to match the 3.79 PIP offer or improve on it in one-cent increments. With the exception of a "Market Maker Prime",5 they participate at the end of the PIP in price/time priority.
As long as the introducing firm matches the final price by the end of the three-second time period, it is guaranteed no less than 40% of the order. Because the process is so fast and because it takes place in penny increments, the introducing firm could easily end up with 100% of an order at only one cent better than the best offer (i.e. $3.79). A customer paying 3.79 would rather pay 3.70, but that is irrelevant to BOX. Even though a customer may be said to have benefited by paying 3.79 instead of 3.80, the PIP will have deprived the customer of an opportunity to have his order exposed for a meaningful time period in a public auction market that could have resulted in the customer's paying 3.70 or even less.
To complete the example above, assume two BOX market makers (A and B) match the 3.79 offer within the three-second period each for 100 contracts with Market Maker A getting there first. The 100-contract customer order would be executed at 3.79 with the introducing firm trading 40 of those contracts and Market Maker A trading 60. Market Maker B gets nothing because its computer system is not as fast as Market Maker A's. In many cases, it may be that no market makers match a firm's PIP offer (the fastest market making firms need not quote every product listed on BOX) and the firm will get 100% of the order by only bettering the quote by a penny. Either way, the introducing firm will always have the last "look" and be able to match. While time priority will apply to other BOX participants, it will not apply to firms seeking to internalize as they can always match at the end of the three-second period to participate.
Effects of the PIP
1. Growth in Internalization.
As just illustrated, the PIP takes internalization to a new level by allowing firms to take at least 40%, and sometimes 100%, of a customer order in only three seconds at one penny better than the prevailing market price. Thus, it is easy to see why the PIP will be wildly popular with firms that seek to internalize (some of which are BOX owners). Obtaining a minimum of 40% at 3.79 is much better than only getting 20% at 3.80 or 40% at the conventional 3.70 increment. Plus, the firm gets to characterize the execution as price improvement to its customer. What is next, 3.7999?
With the approval of the PIP, internalization will grow in leaps and bounds because it will be easier than ever to systematize the practice. The more systematized and entrenched internalization becomes, the greater the impact to the options industry. CBOE believes the impact will have an adverse effect on the quality of the options markets. Stated below are a number of specific concerns raised by the practice of internalization that can have detrimental effects on the future state of the market:
Erosion in the Brokers' Level of Fiduciary Efforts - While accommodative internalization can contribute to the liquidity of the marketplace, it is questionable whether either selective or systematic internalization does so. In fact, it is likely that the latter two bring about the unintended consequence of reducing the ability and availability of dedicated market makers to fulfill their liquidity providing functions when the marketplace needs it most.
As more options order flow is intentionally routed to liquidity providing firms having proprietary relationships with introducing brokerage firms, or internalized selectively by the brokerage firm, the potential for a conflict of interest with customers will increase. There may be little, if any, incentive for a broker/dealer to route orders to an unaffiliated market, or propose a customer order facilitation to a market with a meaningful order exposure (i.e. vibrant price competition), even if that market holds greater promise of price improvement for the customer. The fact that a broker/dealer only needs a penny better than the prevailing NBBO to trade against that order may result in a lack of meaningful order exposure and may, over time, lead to an environment where price improvement is rarely found because price improvement is rarely sought. Such an evolution would clearly not be in the best interests of customers.
A Decline in the Power of Price Competition - Continuing consolidation in the options industry is creating the risk that a small number of firms, which possess both order flow and proprietary trading facilities, will eventually dominate options trading, leading to a reduction in the opportunity for real price competition and order exposure. Similarly, increasing selective internalization will further reduce exposure of a variety of orders to liquidity providers. As the value of competitive pricing erodes - due to limited exposure of orders, payment for order flow, and reliance on the published NBBO as a key measure of best execution - the business models of liquidity providers without captive order flow will simply collapse. Those liquidity providers are needed more in the options market because unlike the stock market, due in part to the large amount of options series, a majority of options orders end up trading with market makers or other professional participants. There will be no incentive to quote if, no matter how competitive the quote, participation is so limited as to be unprofitable. This will lead to a reduction in overall liquidity and the opportunity for price competition in the marketplace.
A Reduction in Market Efficiency during Volatile Periods - Among the underlying principles that make exchange derivative markets efficient is that risk avoiders seek out other parties willing to assume risks in return for a profit opportunity. In the options marketplace, this structure is encouraged through the use of market maker privileges and affirmative obligations. Options market makers are required to provide constant pricing and liquidity, during all market conditions, in order to assure that investors have the ability to trade efficiently. In exchange for these obligations, market makers are granted certain privileges. These privileges allow these participants to focus on the business of assuming risk and managing that risk. In order to do so, they must have reasonable assurances of fair competition, access to order flow and the fair opportunity to realize a profit.
As introducing broker/dealer firms take more of the opposite side of their customers' trades, the detrimental risks to general market efficiency are two-fold. First, selective and even systematic internalization may lead to a reduction in the number of liquidity providers. The broker/dealers, without market-benefiting obligations, will avoid unprofitable trades and uncertain market conditions. The marketplace, having driven out the primary liquidity providers, would experience dramatic reductions in overall liquidity at those times when it is most needed. Second, to the extent the growing internalization trend leads to exchanges establishing or allowing crossing procedures that result in the erosion of meaningful order exposure and price discovery, there will be a fragmentation of price discovery in the marketplace.6 Approval of the PIP would accelerate these unfortunate consequences.
2. Penny Increments.
By allowing non-public quoting to take place in increments that are smaller than the standard nickel and dime increments that will be used on BOX for publicly disseminated quotes, BOX is damaging the credibility of its public quotes. Only qualifying BOX participants who desire to trade in non-standard penny increments will be allowed to do so without providing penny quotes to the Options Price Reporting Authority ("OPRA") for dissemination to the public. Thus, the disseminated OPRA quote feed from BOX will generally not represent the highest bids and lowest offers available on BOX. This will render the publicly disseminated BOX quote meaningless. The PIP will allow certain BOX participants to penny-jump the prevailing BOX public quote. This means that by improving the quote by a penny without even disseminating the improved price, a firm can jump ahead of orders in the BOX book and shut them out from participation. Penny jumping is an existing problem in the stock market today, but at least in the equity markets the penny-better quotes are disseminated to the world. Here, a firm will trade ahead of orders on the BOX book by establishing priority over the quote without having to disseminate an improved quote or taking on any obligations. The options markets will be hurt by a lack of price transparency and this certainly will degrade intermarket quote competition.
One thing that is certain is that if the Commission approves the PIP, the other exchanges will be forced to respond with competitive initiatives of their own in this race to the bottom. If fact, the International Securities Exchange ("ISE"), eager to win that race and as the only existing all-electronic options exchange, has already filed to implement its own internalization program imaginatively called the "Price Improvement Mechanism" or PIM. The PIM will also allow internalized trades to take place in one-cent increments. Mimicking the PIP will be more difficult for the floor based options exchanges because a three-second exposure to a trading crowd in an open outcry auction may not be practicable. This will drive the floor based exchanges to seek other competitive alternatives.
The CBOE will likely respond to the PIP by changing the minimum increments for bids and offers on CBOE to one cent.7 This will undoubtedly prompt other options exchanges to move to penny quoting (even BOX). Aside from the significant issues that the stock market is dealing with as a result of penny quoting, it remains to be seen if the options industry will have sufficient capacity to handle the massive message traffic that will result from penny quoting. When the industry converted to decimals several years ago, the options markets converted to nickel and dime increments for a reason (a penny pilot for options was not requested). We will now be forced to test the soundness of that decision. Perhaps OPRA will be able to handle the incredible message traffic attendant with constant streaming quotes in thousands of options series from numerous sources, but vendors that sell the OPRA feed may be pushed beyond their capacity limits so that instead of disseminating all quotes, the public will only see quotes that are selectively throttled.
This response by CBOE is not a cavalier one. After analyzing the PIP and forecasting its probable effects on our trading volume, we believe that moving to penny increments may become the only way to remain competitive with the proposed internalization machine. If BOX were so keen on trading in pennies, why didn't it propose to regularly quote and trade in penny increments? We believe it is because the BOX model is not predicated on competitive quoting, but rather on obtaining order flow via internalization at the expense of customers. We further believe that the issues attendant with quoting and trading in penny increments need to be studied before one-cent increments are allowed as proposed by BOX. Indeed, these issues need to be considered on a variety of fronts. First, the manner in which message traffic capacity planning is conducted by OPRA and the exchanges is still unresolved. CBOE and other exchanges have no Commission-approved method to plan for additional capacity needs that likely will result from the competitive initiatives that will be proposed in response to the PIP, which may include converting to penny and/or sub-penny quoting and trading. Second, trading issues that will come with penny and subpenny increments including "penny-jumping" and lack of depth in quotes should also be considered.
Other PIP and BOX Problems
CBOE has identified other significant concerns with the proposed PIP and BOX as stated below.
The CBOE continues to have these and other serious concerns11 about the BOX proposal. We believe the operation of the PIP is designed to permit virtually unimpeded internalization and would lead to a dangerous mutation of the options markets into a fragmented system of separate internalization mechanisms. Such a structure would serve the interests of professionals to the detriment of customers and would undermine the integrity of the options markets. We urge the Commission to disapprove the PIP aspect of the BOX filing. We stand ready to discuss these concerns further with the Commission and its staff.