Taking Control of Best Exectution sm

September 22, 2000

Mr. Jonathan G. Katz
Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, DC 20549-0609

Re: Release No. 34-43084

File No. S7-16-00

Dear Mr. Katz:

The Transaction Auditing Group, Inc. (TAG) appreciates the opportunity to respond to the above referenced release, Disclosure of Order Routing and Execution Practices. It has been several years since the Market 2000 Report proposed actions to enhance the competitiveness and efficiency of the U.S. Securities markets with emphasis on the utilization of technology to enable increased transparency, and an overall commitment to best execution. The Commission has promoted competition combined with fairness to investors by adopting the Order Execution Obligations (Release No. 34-37619A) as well as continually promoting the fairness and integrity of our markets as is evidenced in a speech by the Chairman "Best Execution: Promise of Integrity, Guardian of Competition." We view the proposed rules as a natural extension of those concepts.

As an integrated third party service provider, TAG has filled a need in the securities industry by delivering independent and in-depth statistical analysis and reporting of execution quality since 1996. Only as a result of our extensive and continuous investment of time and money, combined with the strong working relationships we enjoy with a large percentage of the broker-dealers, market makers, clearing firms and SROs, have we been successful in developing the methodology, infrastructure and measurement tools needed to accurately and independently measure execution quality.

Over the past four years, our customers have used the statistical information we generate to establish benchmarks for execution performance. With the benchmarks established, many of our customers, driven by the competitive advantage of superior performance results, redirected their efforts and resources toward the improvement of execution quality. TAG statistics have enabled broker-dealers to search for market centers that compete for order flow based on best execution. In our opinion, the proposed rules would establish uniform measures of execution quality and further evolve the best execution practices used in the securities industry. Our specific comments on the release are as follows:

Rule 11Ac1-5: Disclosure of Order Execution Information

Scope of Rule

Stock-by-Stock Burden

TAG has provided information to its customers on a stock-by-stock basis for over three years. Slight modifications to our systems would enable generation of the additional statistics required by this rule with little difficulty. A sample stock-by-stock matrix as proposed under this release is available at our Internet site www.tagaudit.com/rule5.html.

Uniform Statistical Measures

The approach of analyzing execution quality by security, order type and size is consistent with the approach that we have taken in providing execution quality reports to our customers.


Many of our customers have expressed the desire for TAG to provide publication assistance in addition to data gathering, archiving and analysis. We feel that our ability to provide the Internet posting functionality to our customers is essential. As long as dissemination is permitted via the Internet, the collection, analysis and publication of large volumes of information would be feasible.

Required Technology

Compliance with this release would require all market centers to be successful on four fronts: the ability to gather and archive trade execution data; the ability to gather and archive market information data; the ability to qualify, compare and analyze the two types of information necessary to generate statistics on a stock-by-stock basis; and finally the ability to post the results on the Internet. TAG's success in developing the methodology and infrastructure necessary to capture, archive, analyze and publish this information has required extensive investments in management, technical and financial resources. In the absence of third party technology and service providers, only the largest of market centers and SROs would be able to absorb such costs.

Scope of Rule - Market Center

Alternative Execution Systems

An SRO that operates an alternative execution systems such as the Small Order Execution System ("SOES") operated by Nasdaq, the OptiMark systems operated by Nasdaq and the PCX, and floor brokers who receive orders on the floor of an exchange and obtain an execution of the orders with little participation by a specialist, should be required to report. All orders should be subject to evaluation under this release to provide the most comprehensive decision data.

Broker-Dealer as Market Maker

It is our belief that a firm "or department" charged with the execution of an order should only be held accountable upon receipt of the order by their trade execution system. We therefore agree that the Commission's approach is fair and appropriate.

Low-Volume Market Centers

The proposed rule is designed to enhance disclosure of order routing and execution practices thereby promoting fair and vigorous competition among market centers. The exemption of low volume market centers or securities that are thinly traded would dilute the competitive nature of the disclosed information. TAG, as a third party technology and service provider, is able to provide information pursuant to this release to even the smallest of market centers at minimal cost.

Scope of Rule - Covered Orders


In general, we agree with the Commission that the benchmark against which trade execution quality should be measured is the consolidated BBO.

Unusual Market Conditions

We recommend the exclusion of trades executed during unusual market conditions. The inclusion of trades executed during unusual market conditions would distort execution quality results and cause inconsistencies from period to period. The list of unusual market conditions would include, but not be limited to, crossed markets, locked markets, news pending, fast markets and other quotes disseminated as "not firm."

Pre-Opening Orders

Market opening orders in Nasdaq securities represent a significant and increasing portion of orders being submitted for execution. Given that the handling of these orders varies dramatically between market centers, we recommend their inclusion with covered orders. The analysis of market opening orders should be against the first unlocked/uncrossed two-sided BBO occurring at or after 9:30:00 am.

Orders Requiring Special Handling

We agree with the exclusion of orders to be executed at the market closing price, stop orders, orders such as short sales that must be executed on a particular tick or bid, orders that are submitted on a "not held" basis, orders for other than regular settlement, and orders that are to be executed at prices unrelated to the market price at the time of execution. In addition, we would recommend the exclusion of orders containing modifiers such as "all-or-none" or "fill-or-kill."

Scope of Rule - National Market System Security

SmallCap and OTCBB

Consistent with the approach of measuring execution quality on a stock-by-stock basis, the inclusion of SmallCap, as well as OTCBB securities, would be appropriate. There would be no incremental cost associated with the preparation of this additional information.

Listed Options

It is possible to measure the quality of listed option executions on a class-by-class basis. Until such time as OPRA disseminates a consolidated BBO, executions can be analyzed against a BBO derived from the live OPRA feed, if the Commission finds this appropriate.

Required Information

Odd Lots

In an effort not to disadvantage any investor, we recommend the inclusion of an additional order size category for odd lots.

Near-the-Quote Limits

We propose the replacement of this order type with a new category, "outside-the-quote" limit orders. The outside-the-quote category would include all limit orders outside the consolidated BBO at the time of order receipt regardless of the dollar difference between the quote and the limit price.

A "trigger event" would initiate the execution quality measurement of an outside-the-quote limit order. The trigger event is defined as the time when the limit order becomes at-the-quote or better versus the consolidated BBO.

Once triggered, the statistics that apply to inside-the-quote and at-the-quote limit orders could be calculated for all outside-the-quote limits. Note that the cancellation of an outside-the-quote limit order would only be included if the trigger event has occurred.

An example of a triggered outside-the-quote limit order is as follows:

Limit Price
 10:30:3010 1/410 3/810 1/2   
Trigger->10:30:35 10 1/410 3/8  
 10:30:49 10 3/1610 5/1610 1/414 seconds

Missed Opportunities

Regardless of the limit order type, we recommend the inclusion of an additional statistical measure called "missed opportunities" defined as any buy (sell) order for which the best offer (bid) has moved to the limit price and the order was not executed prior to the best offer (bid) subsequently moving away from the limit price. In addition to the cumulative number of shares of covered orders executed or cancelled after a missed opportunity, we propose a statistic representing the share weighted average period from the time of the missed opportunity to the time of order execution.

An example of a Missed Opportunity is as follows:

Limit Price
 10:30:3010 1/410 1/410 5/16  
 10:30:35 10 1/810 1/4  
Missed Opportunity->10:30:49  10 1/410 5/16 14 seconds

Required Information - Information Required for All Types of Orders

Orders Routed Elsewhere

We recommend that orders routed away should not be broken out as a separate category. Order senders select a market center based upon the market center's performance against various statistical measurements - the "execution profile." It is the expectation of the order sending firm that a market center will execute orders according to this profile. By including statistics of orders executed routed elsewhere, the execution quality of the away market center will impact the original market center's statistics and therefore be reflected in its performance. In this manner, the market center which routes orders elsewhere will be incentivized to attain execution for all orders consistent with its profile.

Execution Speed

We agree with the Commission that using execution speed would be a valuable determining factor across all order types.

Consolidated Tape Prints

We would agree with the inclusion of the suggested additional statistical measure of share volume printed on the consolidated tape at prices that are equal to or worse than the limit order price.

Realized Spread

The generation of realized spread statistics would not be difficult or costly for third party providers who have developed comprehensive archival market data systems. It may be difficult, however, for a market center to fold the calculation of realized spread into an existing real-time trading system.

Time of Order Execution

Customer expectations are based on the immediacy of order execution from order placement, so we do not recommend any statistics based on time of order execution.

Rule 11Ac1-6: Disclosure of Order Routing Information

Scope of Rule

Covered Securities

As with our comments associated with Rule 11Ac1-5, Nasdaq SmallCap equities, OTCBB and listed options should be included for review and analysis.

Low-Volume Broker-Dealers

The argument to include all broker-dealers, regardless of size, made under Rule 11Ac1-5, should also apply to routing firms without regard to the number of orders routed on behalf of customers.


In our opinion, allowing routing firms two months after the end of the quarter to issue their quarterly disclosures is too much time. Technology should allow for the publication of routing information within a month of the close of the quarter.

Payment for Order Flow

We believe that execution quality is a valid and important measurement of performance as apposed to payment for order flow and shared trading profits. Therefore, we do not agree that that the disclosure of the aggregate amounts of payment for order flow and shared trading profits would benefit the investor.

Frequency of Reporting

We feel that monthly reporting would be more useful to the investor than quarterly. Our concern with quarterly reporting is that investors could be making decisions based upon information that is up to five months old.

Cost of Preparation

We do not believe that the cost of producing and publishing the information would be prohibitive. TAG, as a third party technology and service provider, is able to provide information pursuant to this release to even the smallest of market centers at minimal cost.

General Comment Rule 11Ac1-6

The Commission's approach to the disclosure of order routing information is an excellent first step towards increasing investor understanding of how orders are executed. It does not, however, provide sufficient information for an investor to evaluate the impact of each broker-dealers order routing choices. We believe it would be helpful to require that broker-dealers publish statistics similar to those required under proposed rule 11Ac1-5. Stock-by-stock level statistics would enable the investor to better judge the execution quality provided by firms. This information is a natural outgrowth of the regular and rigorous reports already provided to our broker-dealer customers. There would be no incremental cost associated with this rule enhancement.

The Commission is proposing two rules to improve public disclosure of order routing and execution practices. The publication of the statistical information outlined in this release supports that goal. Current technology makes the dissemination of the data feasible, and the potential benefit of these rules to the investor far outweighs the cost of implementation. We believe that the proposed rules are consistent with the Commission's mandate to promote the national market system by enabling competition. We strongly support their adoption.


Alan R. Shapiro
Chief Executive Officer

Howard Kohos
Executive Vice President
Chief Operating Officer