PDF Version

NOTICE OF PROPOSED NATIONAL INSTRUMENT 23-102

USE OF CLIENT BROKERAGE COMMISSIONS AS PAYMENT FOR

ORDER EXECUTION SERVICES OR RESEARCH SERVICES

AND COMPANION POLICY 23-102CP

I. INTRODUCTION

The Canadian Securities Administrators (the CSA or we) are publishing the following revised documents for a 90-day comment period:

• Proposed National Instrument 23-102 -- Use of Client Brokerage Commissions as Payment for Order Execution Services or Research Services (Proposed Instrument); and

• Proposed Companion Policy 23-102 CP (Proposed Policy).

We seek to adopt the Proposed Instrument as a rule in each of British Columbia, Alberta, Manitoba, New Brunswick, Nova Scotia, Ontario and Québec, as a Commission regulation in Saskatchewan and as a policy in each of the other jurisdictions represented by the CSA. The Proposed Policy would be adopted as a policy in each of the jurisdictions represented by the CSA.

II. BACKGROUND

On July 21, 2006, the CSA published the following documents for comment (collectively, the 2006 Documents){1}:

• Notice of Proposed National Instrument 23-102 -- Use of Client Brokerage Commissions as Payment for Order Execution Services or Research ("Soft Dollar" Arrangements) (2006 Notice);

• Proposed National Instrument 23-102 -- Use of Client Brokerage Commissions as Payment for Order Execution Services or Research ("Soft Dollar" Arrangements) (2006 Instrument); and

• Proposed Companion Policy 23-102 CP (2006 Policy).

The CSA invited public comment on all aspects of the 2006 Documents and specifically requested comment on fifteen questions. Forty-three comment letters were received. We have considered the comments received and thank all the commenters for their submissions. A list of those who submitted comments, as well as a summary of comments and our responses to them, are attached as Appendix "A" to this Notice.

III. RECENT DEVELOPMENTS

Also in 2006, the U.S. Securities and Exchange Commission (SEC) issued guidance on client commission arrangements. The transition period for implementation of the SEC's 2006 interpretive release (SEC Release){2} ended early in 2007. The final rules of the Financial Services Authority{3} had already taken effect by the time the 2006 Documents were published.

More recently, statements have been made by various representatives of the SEC that suggest that SEC staff continue to work on recommendations to their Commission that may help to increase transparency and improve oversight in relation to the use of client commissions. We will continue to monitor the developments in the U.S.

IV. SUBSTANCE AND PURPOSE OF THE PROPOSED INSTRUMENT AND PROPOSED POLICY

In response to comments received, and after further consideration by the CSA, the 2006 Documents have been materially revised. The purpose of the Proposed Instrument and Proposed Policy remains the same although their content has changed.

The Proposed Instrument continues to provide a specific framework for the use of client brokerage commissions by advisers. It clarifies the broad characteristics of the goods and services that may be acquired by advisers with these commissions and also describes the advisers' disclosure obligations in relation to such use of client brokerage commissions.

The Proposed Policy gives additional guidance regarding the types of goods and services that may be obtained by advisers with client brokerage commissions, as well as non-permitted goods and services. It also gives guidance on the disclosure that would be considered acceptable to meet the requirements of the Proposed Instrument.

V. SUMMARY OF THE PROPOSED INSTRUMENT AND PROPOSED POLICY

A. Common Themes from Comments on the 2006 Documents

The common themes that emerged from the comments received on the 2006 Documents were: (1) difficulties could arise regarding the application of the 2006 Instrument to principal transactions in securities where there is no independent pricing mechanism; (2) the requirements should be harmonized to the greatest extent possible with those in the U.K. and U.S., with preference for harmonization with the U.S.; (3) the proposed disclosure requirements would be difficult to meet and may not be useful to many clients; and (4) a transition period should be considered.

As noted above, we have considered the comments and have made substantive changes to the 2006 Documents (reflected in the current Proposed Instrument and Proposed Policy). These changes are summarized below. Several non-substantive changes have also been made in response to the comments received. These changes and the reasons for them are discussed in the summary of comments and responses included at Appendix "A".

B. Summary of Substantive Changes to the Proposed Instrument and Proposed Policy

The following summary of the substantive changes to the Proposed Instrument and Proposed Policy is divided into five parts: (i) application of the Proposed Instrument; (ii) the definitions of order execution services and research services; (iii) the framework for client brokerage commission practices; (iv) disclosure of client brokerage commission practices; (v) transition period.

(i) Application of the Proposed Instrument

We are now proposing a narrower application of the Proposed Instrument in response to comments regarding difficulties in meeting the requirements if the Proposed Instrument were to apply to all trades in securities. These comments suggested that:

• fees associated with securities traded on a principal basis are imbedded in the price of these securities and cannot be easily measured;

• the lack of pre- and post-trade transparency in the OTC markets makes it difficult to separate the price of a security from the additional services provided; and

• consideration should be given to limiting the application of the proposed instrument to trades in securities where an independent pricing mechanism exists in order to help harmonize with the scope of the SEC and FSA requirements.

Section 2.1 of the Proposed Instrument provides that the application of the Proposed Instrument will be limited to any trade in securities for an investment fund, a fully managed account, or any other account or portfolio over which an adviser exercises investment discretion on behalf of third party beneficiaries, where brokerage commissions are charged by the dealer. Additional guidance has been proposed in subsection 2.1(1) of the Proposed Policy to clarify that the reference in the Proposed Instrument to "client brokerage commissions" includes any commission or similar transaction-based fee charged for a trade where the amount paid for the security is clearly separate and identifiable (e.g., the security is exchange-traded, or there is some other independent pricing mechanism that enables the adviser to accurately and objectively determine the amount of commissions or fees charged).

Subsection 2.1(2) of the Proposed Policy has also been added to provide clarification regarding the basis for limiting the application of the Proposed Instrument, and to clarify that advisers that obtain goods and services other than order execution in conjunction with trades such as principal trades where a mark-up is charged (e.g., fixed income traded in the OTC markets), will remain subject to their general fiduciary obligations to deal fairly, honestly and in good faith with clients, but will not be able to rely on the Instrument to demonstrate compliance with those obligations.

(ii) The Definitions of Order Execution Services and Research Services

Generally, commenters indicated that we should harmonize requirements with the U.S. and U.K. in relation to the definitions of order execution services and research services, and the interpretations of those definitions in relation to the eligibility of certain goods and services. Many of these commenters may have overlooked the differences between these two jurisdictions regarding such definitions and eligibility. Those that noted the differences favoured harmonization with the U.S.

In response to the comments received, we have made changes to the definitions and corresponding guidance. The substantive changes relate to the following:

• The temporal standard for order execution services;

• The definition and characteristics of research services; and

• Views on the eligibility of various specific goods and services.

(a) The temporal standard for order execution services

There were no changes made to the proposed definition of order execution services. The definition remains consistent with that contained in the existing OSC Policy 1.9 and AMF Policy Statement Q-20{4} (Existing Provisions). However, we have made amendments to clarify the proposed temporal standard for order execution services in light of various comments received, which included suggestions that "order execution services" start from the point at which an order life cycles begins (after the investment decision is made), and would generally include those goods and services that are used to decide how, when or where to place an order or effect a trade.

Comments received in relation to questions asked on the eligibility of specific goods and services also indicated that different interpretations of the starting point for the temporal standard exist. For example, comments received relating to the eligibility of post-trade analytics indicated that some parties considered certain uses to be "order execution services" while others considered those same uses to be "research services". This may have been a result of the temporal standard proposed in the 2006 Documents that started at the point after which an adviser makes an investment or trading decision, but did not provide any further clarification as to delineation.

As a result, section 3.2 of the Proposed Policy has been revised and now proposes a temporal standard for order execution services which would generally include goods and services provided or used between the point at which an adviser makes an investment decision (i.e., the decision to buy or sell a security) and the point at which the resulting securities transaction is concluded.

We have also amended the definition of "research services"{5} in the Proposed Instrument by removing reference to "the advisability of effecting securities transactions in securities" and replacing it with language that is intended to help to avoid any future misinterpretation of the proposed temporal standard.

We think that clarifying the starting point for the temporal standard for order execution services would help to ensure consistency in the categorization of goods and services involved in the execution process regardless of the extent to which the adviser relies on the dealer for execution decisions, or contributes to or makes these decisions itself.

While we believe the temporal standard may be different from that included in the SEC Release{6}, we do not believe the difference would cause any issues regarding the eligibility of particular goods or services between jurisdictions. Rather, this should only result in differences in how an eligible good or service has been categorized between the two jurisdictions; for example, a good categorized as research under the SEC's temporal standard might be categorized as order execution services under the Proposed Instrument.

Question 1:

What difficulties might be caused by a temporal standard for order execution services that might differ from the standard applied by the SEC, especially in the absence of any detailed disclosure requirements in the U.S.? In the event difficulties might result, do these outweigh any benefit from having a temporal standard that results in consistent classification of goods and services based on use?

(b) The definition and characteristics of research services

We have made substantive changes to both the definition of research services and the associated guidance as a result of comments received regarding the 2006 Documents. These comments included that the characteristics of research services proposed, combined with the proposed obligation for advisers to ensure that research received adds value to investment or trading decisions, do not allow for eligibility of those goods and services that might not contain the specific proposed characteristics, or may not on their own add value to the investment trading decision, but do add value when used by an adviser as an input to its own analyses and research processes. We also re-examined whether an approach more consistent with that taken in the SEC Release, which places more focus on the use of the goods and services, should be adopted.

As a result, the following substantive changes were made:

• The proposed guidance included in section 3.3 of the Proposed Policy was revised to reduce the focus on the characteristics of research.

• The obligation proposed in paragraph 3.1(2)(b) of the 2006 Instrument for the adviser to ensure that research services add value to the investment decision was also removed in conjunction with amendments to place more focus on the use of goods and services for determining eligibility for payment with client brokerage commissions. (Other reasons also contributed to the removal of this obligation and these are discussed below in the section: The Framework for Client Brokerage Commission Practices.)

(c) Views on the eligibility of certain goods and services

We considered and re-examined the eligibility, as research services, of goods and services such as raw market data, proxy-voting services, and mass-marketed or publicly-available information or publications, and the eligibility, as order execution services, of order management systems and post-trade analytics. In response to comments, we also considered the eligibility of other goods and services such as seminars, telephone / data communication lines, expert opinions, pre-trade analytics, as well as databases and software.

Commenters provided various compelling reasons for why certain goods and services should be considered eligible, whether as order execution services or research services. These reasons generally included a concern relating to not being harmonized with the views in the SEC Release.

As a result, we have made the following substantive changes:

• The proposed definition of "research services" in the Proposed Instrument now includes databases and software to the extent they are designed mainly to support the other services referred to in the proposed definition of "research services", as is currently included in the definition of "investment decision-making services" in the Existing Provisions.

• The proposed guidance in subsections 3.2(3) and 3.3(2) of the Proposed Policy, which provide examples of goods and services that might be considered order execution services and research services, respectively, has been amended.

• The proposed guidance in section 3.5 of the Proposed Policy, which provides examples of goods and services that we would consider to be clearly outside the permitted goods and services under the Proposed Instrument, has been amended.

The summary of comments and our responses included at Appendix "A" provide more information regarding our views on various specific goods and services, and the reasons for the amendments made or not made to the Proposed Policy.

We emphasize that it is not feasible to attempt to include in the Companion Policy a comprehensive list of all possible goods and services that might be considered eligible as order execution services or research services. The examples proposed are intended solely to help an adviser with its assessment of whether a good or service meets the definition of order execution services or research services. Even if certain goods or services were specifically mentioned in a final Companion Policy, the adviser would still have to meet the obligations under Part 3 of the Proposed Instrument in order to be able to justify its use of client brokerage commissions as payment for those goods or services.

(iii) The Framework for Client Brokerage Commission Practices

In response to comments received, we have also made changes to the obligations proposed for advisers that use client brokerage commissions as payment for order execution services or research services. The substantive changes relate to the following:

• The relationship between the use of goods and services and the obligation to ensure such use is for the benefit of the client(s);

• The relationship between benefits received and particular clients;

• The ability to assess value received in relation to value paid; and

• Unsolicited goods and services.

There were no significant comments received relating to a dealer's obligations under the 2006 Instrument that resulted in substantive changes.

(a) The relationship between the use of goods and services and the obligation to ensure such use is for the benefit of the client(s)

As noted earlier in this notice, we have made amendments to the proposed definition and characteristics of research services in order to place more focus on the use of the goods and services for determining whether payment could be made for these with client brokerage commissions.

In conjunction with these amendments, we reassessed the general framework for the use of client brokerage commissions. Paragraph 3.1(2)(a) of the Proposed Instrument continues to require an adviser that uses client brokerage commissions as payment for order execution services or research services to ensure that the services benefit the client(s).

Additional guidance has also been proposed in subsection 4.1(2) of the Proposed Policy that indicates that in order to benefit a client, the goods and services obtained should be used in a manner that provides appropriate assistance to the adviser in making investment decisions, or in effecting securities transactions. The guidance also indicates that the adviser should be able to demonstrate how the goods and services paid for with client brokerage commissions are used to provide appropriate assistance.

Further, as a result of changes made to the proposed guidance regarding the characteristics of research services, and because of the refocus of the proposed framework towards the use of the goods and services, we have also removed the obligation proposed in the 2006 Instrument requiring the adviser to ensure that the research received adds value to investment or trading decisions. We believe that the additional proposed guidance relating to the use of goods and services in a manner that provides appropriate assistance should be sufficient.

(b) The relationship between benefits received and particular clients

In order to clarify that it is not our intention to require advisers to ensure that a direct connection exists between each specific good or service received and particular clients, we have made amendments to the proposed guidance.

Subsection 4.1(3) has been added to the Proposed Policy to acknowledge that a specific order execution service or research service may benefit more than one client, and may not always directly benefit each particular client whose brokerage commissions were used as payment for the particular service. The proposed guidance also indicates that advisers should have adequate policies and procedures in place to ensure that all clients whose brokerage commissions were used as payment for these goods and services have received fair and reasonable benefit from such usage.

(c) The ability to assess value received in relation to value paid

We considered those comments that suggested it might be difficult to ensure that the amount of client brokerage commissions paid is reasonable in relation to the value of goods and services received when there is a lack of cost information provided by dealers that bundle goods and services with order execution. We also considered those suggestions of adopting the SEC approach by instead requiring that a good faith determination be made of the reasonableness of the amounts paid.

We have therefore amended subsection 3.1(2) of the Proposed Instrument to now propose that the adviser must ensure that a good faith determination has been made that the amount of client brokerage commissions paid is reasonable in relation to the value of the order execution services or research services received. Additional guidance has been proposed in subsection 4.1(4) of the Proposed Policy regarding how the adviser might make this determination, including that the determination can be made either with respect to a particular transaction or the adviser's overall responsibilities for client accounts.

(d) Unsolicited goods and services

From the comments received, we note that a level of uncertainty exists regarding the treatment under the Proposed Instrument of unsolicited goods and services, and of access to goods and services provided by dealers, when the goods and services provided or offered are either not eligible under the Proposed Instrument or not used by the adviser. We also note concerns associated with the lack of control over what goods and services a dealer might send or provide access to in return for client brokerage commissions.

To address these concerns, we have proposed guidance in subsection 4.1(4) of the Proposed Policy to clarify that the relevant measure for any good faith determination under paragraph 3.1(2)(b) of the Proposed Instrument is the reasonableness of the client brokerage commissions paid in relation to the goods and services received and used by the adviser. This means an adviser that, by virtue of paying client brokerage commissions, is provided with access to goods and services, or receives goods or services on an unsolicited basis and does not use such goods and services, will not be considered to be in violation of this obligation if it does not include these in its assessment of value received in relation to commissions paid. The proposed guidance also indicates that if an adviser uses the goods or services, or considers their availability a factor when selecting dealers, the adviser should include these in its assessment.

We think this approach could also be extended to the situation when an adviser is making allocations with respect to a mixed-use good or service. We would not expect an adviser to allocate cost to, and pay with its own funds for, an ineligible portion of a good or service received on an unsolicited basis that was not used. However, the adviser would still have the obligation to make a good faith determination that the amount of client brokerage commissions paid was reasonable in relation to the value of the eligible portion of that good or service received.

(iv) Disclosure of Client Brokerage Commission Practices

Numerous comments were received in relation to the disclosure proposed in the 2006 Instrument. There were a number of arguments received for why the detailed proposed disclosure would be overly onerous to produce, and why it might be of questionable use to clients. However, we maintain the view that additional disclosure relating to the use of client brokerage commissions is necessary in order to increase the transparency to clients regarding such use, to help clients understand the services they are receiving, and to ensure appropriate rigour in the processes of all advisers.

To respond to the comments, though, we have made changes to the proposed disclosure requirements that we think provide an appropriate balance between the need for transparency and accountability, the associated burden and costs that might be imposed on advisers, and the aim for consistency with disclosure in the U.S. The substantive changes relate to the following:

• Clarification of the meaning of "client" for purposes of disclosure;

• The scope of the proposed narrative disclosure;

• The scope of the proposed quantitative disclosure; and

• Additional details to be maintained and made available upon request.

We do not believe any changes are necessary in relation to the form or frequency of disclosure.

(a) Clarification of the meaning of "client" for purposes of disclosure

As a result of the uncertainty evident from the comments regarding the meaning of "client" for purposes of disclosure, we have proposed guidance in section 5.1 of the Proposed Policy to clarify that the recipient of the disclosure should typically be the party with whom the contractual arrangement to provide advisory services exists. For example, for an adviser to an investment fund, the client would typically be considered the fund, unless the adviser is also the trustee and/or the manager of the fund, or is an affiliate of the trustee and/or manager of the fund, in which case the adviser should consider whether its relationship with the fund presents a conflict of interest matter under National Instrument 81-107 Independent Review Committee for Investment Funds that requires review by the Independent Review Committee established in accordance with that National Instrument, and whether it would be more appropriate for the disclosure to be made instead to the Independent Review Committee.

(b) The scope of the proposed narrative disclosure

We have revised the proposed disclosure requirements to increase the scope of the narrative disclosure to be provided so that clients will be better able to understand how their brokerage commissions are used by advisers as payment for goods and services other than order execution.

In formulating the new proposed narrative disclosure requirements we considered the suggestions received from commenters, and re-examined the current narrative disclosure included in Part II of the SEC's Form ADV and in the Investment Management Association's Pension Fund Disclosure Code.

The narrative disclosure requirements proposed in paragraphs 4.1(a) through (e) of the Proposed Instrument would essentially maintain requirements proposed in the 2006 Instrument for disclosure of the nature of the arrangements entered into relating to the use of client brokerage commissions as payment for order execution services or research services, as well as disclosure of the names of dealers and third parties that provided goods and services other than order execution, and the types of goods and services provided. However, we have also proposed that each dealer or third party named through this disclosure that is an affiliated entity should be separately identified, along with separate disclosure of the types of goods and services provided.

Additional narrative disclosure requirements that we have proposed include a description of the process for, and factors considered in, selecting dealers to effect securities transactions; the procedures for ensuring that, over time, clients receive reasonable benefit from the usage of the brokerage commissions charged to them; and the methods by which the determination of the overall reasonableness of client brokerage commissions paid in relation to order execution services and research services received is made.

Additional proposed guidance to help the adviser understand the expectations with respect to the proposed narrative disclosure requirements is included in subsections 5.3(2) and (3) of the Proposed Policy.

(c) The scope of the proposed quantitative disclosure

We have also revised the proposed disclosure requirements by decreasing the scope of the quantitative disclosure that was proposed in the 2006 Instrument. As an initial step in increasing accountability and transparency through quantitative disclosure, we have proposed in paragraph 4.1(f) of the Proposed Instrument to reduce the client-level quantitative disclosure requirements to disclosure of the total client brokerage commissions paid by the client during the period. In addition, in paragraph 4.1(g) of the Proposed Instrument we have proposed requiring disclosure on an aggregated basis of the total client brokerage commissions paid during the period, along with a reasonable estimate of the portion of those aggregated commissions that represents the amounts paid, or accumulated to pay for, goods and services other than order execution. Guidance has also been proposed in subsection 5.3(4) of the Proposed Policy in relation to the level of aggregation of client brokerage commissions for these disclosure purposes. The proposed guidance allows advisers flexibility to determine the appropriate level of aggregation based on their business structure and client needs.

We believe the quantitative disclosure proposed is consistent with that currently required to be made by investment funds to clients under NI 81-106, except that the proposed disclosure requires the adviser to make a reasonable estimate of the amounts paid or accumulated to pay for goods and services other than order execution, as opposed to requiring disclosure of these amounts to the extent ascertainable.{7}

We are also of the view that the scope of the quantitative disclosure requirements currently being proposed should not create any unreasonable burden on advisers, or that any apparent lack of harmonization between the quantitative disclosure requirements in the Proposed Instrument and those currently required in the U.S. and U.K. will cause any significant issues. Regardless, we will continue to monitor the developments in the U.S., including whether amendments to their disclosure regime are proposed, and are prepared to revisit the approach we have taken at that time.

Question 2:

What difficulties might be encountered by requiring the estimate of the aggregated commissions to be split between order execution and goods and services other than order execution? What difficulties might be encountered if instead the requirement was for the aggregate commissions to be split between research services and order execution services?

Question 3:

As order execution services and research services are increasingly offered in a cross-border environment, should the Proposed Instrument allow an adviser the flexibility to follow the disclosure requirements of another regulatory jurisdiction in place of the proposed disclosure requirements, so long as the adviser can demonstrate that the requirements in that other jurisdiction are, at a minimum, similar to the requirements in the Proposed Instrument? If so, should this flexibility be solely limited to quantitative disclosure given that the issues associated with differences in quantitative disclosure requirements between regulatory jurisdictions are likely greater than the problems associated with differences in narrative disclosure requirements? In addition, should there be limitations on which regulatory jurisdictions an adviser may look to for purposes of identifying suitable alternative disclosure requirements and, if so, which jurisdictions should be considered eligible and why?

(d) Additional details to be maintained and made available upon request

We have removed the requirement proposed in subsection 4.1(2) of the 2006 Instrument that would have required the adviser to maintain specifics about each good or service received in the event that a client were to make a request for such information. We are of the view that disclosure of the provider names and types of goods and services currently proposed under paragraph 4.1(c) of the Proposed Instrument should generally provide clients with sufficient detail relating to the specific goods and services paid for with client brokerage commissions.

Despite removal of this explicit requirement, advisers are reminded of the general requirement to maintain adequate books and records in order to be able to demonstrate compliance with the Proposed Instrument.

(v) Transition Period

In response to commenter concerns regarding the need to include a transition period, in particular those concerns relating to the need for time to meet the disclosure requirements proposed in the 2006 Instrument, we have proposed an effective date for the Proposed Instrument of six months from its approval date. This is included in section 6.1 to the Proposed Instrument.

We believe that the amendments made to Proposed Instrument since those proposed in the 2006 Instrument, including the removal of some of the more onerous reporting requirements, should address many of the commenter concerns, and therefore a longer transition period should not be needed.

Question 4:

Should a separate and longer transition period be applied to the disclosure requirements to allow time for implementation and consideration of any future developments in the U.S.? If so, how long should this separate transition period be?

VI. SPECIFIC REQUESTS FOR COMMENTS

In summary, we specifically request comment on the following issues:

Question 1:

What difficulties might be caused by a temporal standard for order execution services that might differ from the standard applied by the SEC, especially in the absence of any detailed disclosure requirements in the U.S.? In the event difficulties might result, do these outweigh any benefit from having a temporal standard that results in consistent classification of goods and services based on use?

Question 2:

What difficulties might be encountered by requiring the estimate of the aggregated commissions to be split between order execution and goods and services other than order execution? What difficulties might be encountered if instead the requirement was for the aggregate commissions to be split between research services and order execution services?

Question 3:

As order execution services and research services are increasingly offered in a cross-border environment, should the Proposed Instrument allow an adviser the flexibility to follow the disclosure requirements of another regulatory jurisdiction in place of the proposed disclosure requirements, so long as the adviser can demonstrate that the requirements in that other jurisdiction are, at a minimum, similar to the requirements in the Proposed Instrument? If so, should this flexibility be solely limited to quantitative disclosure given that the issues associated with differences in quantitative disclosure requirements between regulatory jurisdictions are likely greater than the problems associated with differences in narrative disclosure requirements? In addition, should there be limitations on which regulatory jurisdictions an adviser may look to for purposes of identifying suitable alternative disclosure requirements and, if so, which jurisdictions should be considered eligible and why?

Question 4:

Should a separate and longer transition period be applied to the disclosure requirements to allow time for implementation and consideration of any future developments in the U.S.? If so, how long should this separate transition period be?

VII. AUTHORITY FOR THE PROPOSED INSTRUMENT

In those jurisdictions in which the Proposed Instrument is to be adopted as a rule or regulation, the securities legislation in each of those jurisdictions provides the securities regulatory authority with rule-making or regulation-making authority in respect of the subject matter of the Proposed Instrument.

In Ontario, the Proposed Instrument is being made under the following provisions of the Securities Act (Ontario) (Act):

• Paragraph 2(i) of subsection 143(1) of the Act allows the Commission to make rules in respect of standards of practice and business conduct of registrants in dealing with their customers and clients, and prospective customers and clients.

• Paragraph 2(ii) of subsection 143(1) of the Act allows the Commission to make rules in respect of requirements that are advisable for the prevention or regulation of conflicts of interest.

• Paragraph 7 of subsection 143(1) of the Act allows the Commission to make rules prescribing requirements in respect of the disclosure or furnishing of information to the public or the Commission by registrants.

VIII. RELATED INSTRUMENTS

The Proposed Instrument and Proposed Policy are related to the Existing Provisions. The AMF and OSC intend to revoke the Existing Provisions and to replace them with the Proposed Instrument and the Proposed Policy, if and when adopted. The revocation of the Existing Provisions is not intended to take effect until the effective date of the Proposed Instrument.

IX. ALTERNATIVES AND ANTICIPATED COSTS AND BENEFITS

Most of the alternatives considered, and the anticipated costs and benefits of implementing the Proposed Instrument, are discussed in the cost-benefit analysis entitled Cost-Benefit Analysis: Use of Client Brokerage Commissions as Payment for Order Execution Services and Research. An updated cost-benefit analysis is being published together with this Notice and is included at Appendix "B".

An additional alternative was proposed by the British Columbia Securities Commission (BCSC) with the 2006 Notice. The BCSC suggested that the existing duty for advisers to act fairly, honestly and in good faith, together with guidance and the use of other regulatory tools including compliance reviews and education, would be an appropriate way to regulate client brokerage commission arrangements. Although the BCSC is participating in this republication, the BCSC Board has not yet decided whether the BCSC will adopt the Proposed Instrument. The BCSC looks forward to reviewing further comments in response to the Proposed Instrument.

X. UNPUBLISHED MATERIALS

In developing the Proposed Instrument, we have not relied on any significant unpublished study, report, or other material.

XI. COMMENTS AND QUESTIONS

Interested parties are invited to make written submissions with respect to the Proposed Instrument, Proposed Policy, and the specific questions set out in this notice. Please submit your comments in writing before April 10, 2008.

Submissions should be sent to all securities regulatory authorities listed below in care of the OSC, in duplicate, as indicated below:

Submissions should also be addressed to the Autorité des marchés financiers (Québec) as follows:

A diskette containing the submissions should also be submitted. As securities legislation in certain provinces requires a summary of written comments received during the comment period be published, confidentiality of submissions cannot be maintained.

Questions may be referred to:

January 11, 2008

{1} Published at (2006) 29 OSCB 5923.

{2} The SEC Release was issued on July 18, 2006 under Exchange Act Release No. 34-54165. These were effective July 24, 2006 with a six-month transition period to January 24, 2007.

{3} The FSA's final rules were published in July 2005 in Policy Statement 05/9, Bundled Brokerage and Soft Commission Arrangements: Feedback on CP 05/5 and Final Rules. These were effective January 1, 2006 with a six-month transition period.

{4} AMF Policy Statement Q-20 gained the force of a rule in June 2003 through Section 100 of An Act to amend the Securities Act (S.Q. 2001, chapter 38).

{5} The term "research services" replaces the term "research" used in the 2006 Instrument and 2006 Policy.

{6} For its temporal standard, the SEC Release states that "brokerage begins when the money manager communicates with the broker-dealer for the purpose of transmitting an order for execution and ends when funds or securities are delivered or credited to the advised account or the account holder's agent" (SEC Release, pp. 40-41).

{7} Consideration will be given to the need for harmonization between the disclosure requirements in the Proposed Instrument and those in the National Instruments governing disclosure by investment funds.

APPENDIX A

Proposed National Instrument 23-102 Use of Client Brokerage Commissions

as Payment for Order Execution Services or Research ("Soft Dollar" Arrangements)

and Companion Policy 23-102CP

Summary of Comments and Responses

I. Response to Questions

Question 1: Should the application of the Proposed Instrument be restricted to transactions where there is an independent pricing mechanism (e.g., exchange-traded securities) or should it extend to principal trading in OTC markets? If it should be extended, how would the dollar amount for services in addition to order execution be calculated?

The majority of commenters were of the view that the Proposed Instrument should be restricted to transactions where there is an independent pricing mechanism (exchange-traded securities). The reasons given were as follows:

• the fees associated with securities traded on a principal basis (such as fixed income securities) are imbedded in the price of these securities, cannot be easily measured, and the increased costs associated with the enhanced record-keeping needed to separate execution-only and research costs would not be justified given the lack of precision in the data;

• the lack of pre- and post-trade transparency in the OTC markets makes it difficult to separate the price of a security from the additional services provided;

• it is difficult or impossible to break out the commissions from the total transaction costs for securities traded on a principal basis;

• as long as commissions are not explicitly delineated by dealers, advisers will have to make their own estimates that will likely differ and lead to inconsistent disclosure;

• it is important to remain as consistent as possible with the FSA (whose requirements apply only to equities and related instruments) and the SEC (whose requirements apply to commissions on agency transactions and fees on certain riskless principal transactions that are reported under NASD trade reporting rules);

• it would be especially difficult to break down commissions for foreign fixed income securities because dealers in those countries are not be subject to the same requirements; and

• for securities traded on a principal basis there is limited scope for research and other services besides pure execution, so there is little value in "unbundling" the cost of execution in that case.

A few commenters, however, thought that transactions done on a principal basis should also be included in the scope of the Proposed Instrument, for the following reasons:

• soft dollar information should not be hidden from investors because of the type of product, transaction or market;

• there are proprietary broker-based fixed income research services paid for via the commissions implicit in bond spreads, and the calculation of the dollar amount is straightforward: that is, dealers place specific prices on each research service, and after the execution of the trade has been agreed to, an extra amount is added and identified as a research service payment;

• if principal transactions are excluded from the Proposed Instrument, unscrupulous advisers with both fixed income and equity mandates may shift non-eligible expenses defined by the instrument from equity soft dollars towards soft dollars related to principal transactions; and

• it is unfair to closely monitor commission expenditures in public markets and not OTC markets; at the very least, participants in OTC markets should begin to disclose the amount and type of goods and services procured through the dealers.

However, there was acknowledgement of the difficulty in determining the dollar amount for bundled services received in conjunction with principal trades. Some commenters suggested that, if a decision is made to expand the applicability of the Proposed Instrument beyond transactions where there is an independent pricing mechanism, it should apply to any transaction where a transaction-based fee can be determined or reasonably estimated.

- - - - - - - - - - - - - - - - - - - -

Response:

We agree that the lack of transparency regarding fees imbedded in the price of trades conducted on a principal basis in the OTC markets makes measurement of those fees difficult. The application of the Proposed Instrument is limited to certain trades in securities where brokerage commissions are charged. We have amended the guidance in the Proposed Policy to clarify that the reference to "client brokerage commissions" includes any commission or similar transaction-based fee charged for a trade where the amount paid for the security is clearly separate and identifiable (e.g., the security is exchange-traded, or there is some other independent pricing mechanism that enables the adviser to accurately and objectively determine the amount of commissions or fees charged).

The Proposed Policy also clarifies that advisers that receive goods and services other than order execution in conjunction with trades such as principal trades where a mark-up is charged (e.g., fixed income traded in the OTC markets), will remain subject to their general fiduciary obligations to deal fairly, honestly and in good faith with clients, but will not be able to rely on the Proposed Instrument to demonstrate compliance with those obligations. An adviser could likely apply many of the principles outlined in the Proposed Instrument and Proposed Policy in these situations to assess whether its general fiduciary obligations have been met, but this assessment may be more difficult and less supportable when information is not readily available to assist with a determination of value received for value paid (e.g., the security is not exchange-traded, or there is no other independent pricing mechanism to help identify the amount paid for the security versus the amount paid for execution plus any other services).

- - - - - - - - - - - - - - - - - - - -

Question 2: What circumstances, if any, make it difficult for an adviser to determine that the amount of commissions paid is reasonable in relation to the value of goods and services received?

The majority of respondents thought that the main difficulty in assessing the reasonableness of the commissions paid relative to the value of goods and services received for transactions involving execution and research was the lack of information provided by dealers on the cost components of bundled services. Some noted that, unless dealers are required to unbundle execution charges from charges for proprietary research, any attempt by advisers to determine the costs of execution and research, and whether they are reasonable in relation to the value of goods and services rendered, is merely an estimate.

One commenter, however, anticipates that the 2006 Instrument would cause "execution-only" trades to become more commonplace; in which case, industry norms would evolve as to what represents a competitive "execution-only" commission, and there will be far greater clarity as to the price being paid for goods and services relative to their value. Another commenter supported the view that "execution-only" trades may become more commonplace as total research costs come under more scrutiny, and limits are placed on the total spent for research.

Other reasons supporting the difficulty in assessing the reasonableness of value received for commissions paid included:

• while theoretical pure execution costs may be determined for a particular trade, the value of research is dependent upon the specific nature of the services provided and the circumstances under which it is provided;

• it would be difficult to determine reasonableness for an adviser that is small or just starting up, and/or if an adviser tends to execute transactions with only one dealer;

• there is a continuum of service levels ranging from low service direct market access to low to medium service algorithmic trading, to high-service execution involving liquidity search, monitoring and reporting the status of an order, feedback, execution advice and the provision of capital, all of which require different commission rates;

• in almost all cases, research received by an adviser is used for the benefit of more than one client, and a specific allocation of the benefits of research to one client would be nearly impossible;

• dealers often send advisers unsolicited research that is not used by the adviser; receipt of such research should not imply that the adviser is using commissions to pay for it; and

• advisers consider the reasonableness of commissions paid to dealers over time, and in context of the overall business relationship, not on the basis of individual trades.

In conjunction with the comments regarding the difficulties in determining whether commissions paid are reasonable in relation to the goods and services received, some commenters suggested that an approach consistent with that of the SEC, as described in their July 2006 Release should be taken: i.e. advisers should be required to make a good faith determination that commissions paid are reasonable in relation to the value of the research or brokerage services received, either in terms of the particular transaction or the manager's overall responsibilities for discretionary accounts.

Three commenters suggested that use of a robust independent commission management system would help monetize the value of bundled research or execution services paid for with commissions. They noted that new software solutions for evaluating soft dollar arrangements would help buy-side firms quantify the services received from dealers without additional administrative burden.

- - - - - - - - - - - - - - - - - - - -

Response:

We understand the concerns relating to the difficulties in determining if commissions are reasonable in relation to the order execution services and research services received, particularly in relation to bundled services. We still think it is important for an adviser to make a determination of whether the value of the goods and services received is reasonable in relation to clients' commissions paid to help ensure that clients are receiving adequate value.

We have made changes to the Proposed Instrument to require the determination to be made in good faith, and to the Proposed Policy to clarify that such a determination could be made in terms of either a particular transaction or the adviser's overall responsibilities for client accounts.

- - - - - - - - - - - - - - - - - - - -

Question 3: What are the current uses of order management systems? Do they offer functions that could be considered to be order execution services? If so, please describe these functions and explain why they should, or should not, be considered "order execution services"?

Some respondents indicated that order management systems (OMSs) and order execution/ execution management systems have become so intertwined that it is difficult to separate the order management system from the execution process.

Various respondents provided examples of the current uses of order management and order execution / execution management systems. In general, commenters indicated that these systems track the progress of an order from its initiation to completion. More specific examples included:

• modeling trades / execution strategies and portfolios;

• order entry, routing and messaging;

• collection of orders for multiple point entry;

• bulking of smaller orders;

• order and trade allocation;

• direct contact from the advisers to the trading desk;

• algorithmic trading functions and direct market access;

• analytic tools to assist in the investment decision-making process, including pre- and post-trade analytics;

• facilitating the expediency of the execution process;

• analyzing portfolio strategies;

• evaluating execution quality;

• post-trade matching;

• routing of settlement instructions;

• report generation;

• security-master information;

• compliance;

• portfolio administration; and

• record keeping.

The majority of commenters generally agreed that OMSs contain portions that are used to assist in the order execution process that should be considered order execution services, such as:

• modeling trades and execution strategies;

• order routing and messaging;

• direct market access and algorithmic trading functions; and

• settlement functions such as post-trade matching, and the routing of settlement instructions to custodian banks and clearing agents.

Others added that portions of OMSs could be considered research to the extent they assist in the investment decision making process. Examples included:

• market data integration tools;

• analytic tools; and

• portfolio and strategic modeling tools.

One commenter suggested that features such as managing trade allocations, monitoring portfolio risk, or certain compliance features should qualify for soft dollar reimbursement, but should be judged on their individual characteristics as to whether they are execution or research oriented.

Many of the commenters also indicated that there are portions of OMSs that are used for administrative purposes which should not be eligible, such as compliance, accounting and recordkeeping functions.

A few commenters were of the view that OMSs should not qualify as order execution services. The reasons were:

• since the main trading function of an OMS is routing orders to venues, platforms and sell-side participants which provide order execution, the functionality that improves the quality of order execution typically resides outside of the OMS and the primary benefits of OMSs accrue to the investment manager and not the asset owners;

• tools of the trade such as the basic hardware, software, reports, communication links and other resources needed to competitively and compliantly run a contemporary mutual fund should not be considered order execution services and the costs should paid for through the management fee;

• order management services provide a strategic advantage to firms that use them, and should therefore not be paid using client brokerage commissions.

- - - - - - - - - - - - - - - - - - - -

Response:

We agree with commenters that order and execution management systems can include functions that could be considered either order execution services or research services. For example, to the extent that they provide analytic and modeling tools used in the research process, or are used to assist in arranging or effecting a securities transaction, these portions may be eligible providing the adviser meets its obligations under Part 3 of the Proposed Instrument.

We also think that it would be difficult to argue that the portions of these systems used for administrative functions such as compliance, accounting and recordkeeping would sufficiently benefit the client by providing appropriate assistance in making investment decisions, or in effecting securities transactions, to justify their payment with client brokerage commissions. As a result, we think these systems would generally be considered mixed-use in nature.

- - - - - - - - - - - - - - - - - - - -

Question 4: Should post-trade analytics be considered order execution services? If so, why?

The majority of commenters thought post-trade analytics should be considered order execution services for the following reasons:

• assessment of past trading is a key part of the process of achieving best execution;

• they aid an adviser in making future decisions about how trades should be allocated among the brokers who provide execution services and the method of execution that is most appropriate (e.g. trader-managed; agency/principal blocks; algorithms; direct market access, etc.);

• they can influence how, when and where an adviser decides to trade;

• post-trade analytics are a key part of how an adviser reviews the order execution process and improves it -- through analysis of past trades to uncover problems in, or validate, a trading strategy, execution method or venue, dealer capabilities, etc; and

• they are all part of a continuous process, and a key part of analyzing the indirect or slippage costs within the trading process.

A number of respondents believed that post-trade analytics should be considered research. The reasons were:

• they are received and considered by the adviser before making further trading decisions, even if they are received after certain trades have been concluded;

• they include information about how well a broker conducted a particular transaction or series of transactions for an investment manager, as well as advice on liquidity and market-related timing, negotiation of the terms of a trade and other aspects of order handling;

• they assist advisers in assessing trading effectiveness;

• they assist in achieving best execution for clients; and

• they feed into an adviser's trading decisions and help promote competition between execution platforms.

One commenter noted that post-trade analytics are more properly characterized as research than order execution services, and that even though post-trade analytics are received after certain trades have been concluded, they should be considered research to the extent they help determine a subsequent investment or trading decision.

A number of commenters noted that post-trade analytics should be mixed-use products because they contain components that do not assist in making subsequent decisions, and are not received during either of the temporal standards contemplated for research or order execution services. For example, some of these commenters noted that post-trade analytics should not be eligible for payment with client commissions when used to evaluate portfolio performance for marketing purposes, recordkeeping, administrative and compliance purposes.

- - - - - - - - - - - - - - - - - - - -

Response:

Many of the reasons given by commenters for why post-trade analytics might be considered order execution services are the same as those given in support of their eligibility as research services. This appears to be a result of differing interpretations of the temporal standard for order execution services. We have made amendments to the definition of research services in the Proposed Instrument and to the guidance in the Proposed Policy that should serve to clarify that the temporal standard for order execution services starts after the adviser has made its investment decision (i.e., the decision to buy or sell a security). The amendments made would therefore allow for consideration of post-trade analytics as order execution services to the extent they are used to determine a subsequent decision of how, when or where to place an order or effect a trade. These amendments relating to the temporal standard are discussed in more detail later in Section II of this response to comments.

As suggested by the guidance provided in the Proposed Policy, we also think that to the extent that post-trade analytics are used for administrative or compliance purpose, it would be difficult for an adviser to argue that these uses provide appropriate assistance, and to therefore justify paying for these portions with client brokerage commissions. As a result, we think post-trade analytics would generally be considered mixed-use in nature.

- - - - - - - - - - - - - - - - - - - -

Question 5: What difficulties, if any, would Canadian market participants face in the event of differential treatment of goods and services such as market data in Canada versus the U.S. or the U.K.?

The overwhelming majority of commenters thought that the Canadian approach should be harmonized with the U.S. and U.K. approaches. The following reasons were given:

• adopting conflicting regulatory requirements would put Canada at a severe competitive disadvantage and encourage regulatory arbitrage;

• while a foreign adviser will be able to use commissions to pay for certain services, the Canadian adviser will have to absorb those costs as fixed-costs or by charging an increased fee; this may result in loss of business for Canadian advisers and any long term-decline in profitability will encourage Canadian advisers to move to other jurisdictions where the regulatory regime does not impair their ability to compete;

• if raw data feeds are excluded for Canadian advisers and not U.S. advisers, quantitative money managers in Canada would suffer a disadvantage compared to their U.S. counterparts because their data would cost more; they will have to charge higher investment management fees to international and U.S. clients than their U.S. peers, which will result in the loss of non-Canadian clients;

• if an inconsistent approach is taken, firms with offices in multiple jurisdictions would have to choose between adopting the strictest standards for all offices or suffering the inconvenience and costs of having different processes applicable to different clients' commission dollars, depending on the jurisdiction;

• differential treatment will result in additional costs for advisers in Canada who use sub-advisers in the U.S. or the U.K., as the sub-advisers will be forced to pay for the development of systems required to track the information required by Canadian regulators; and

• as Canadian mutual funds increase their holdings in foreign securities, now that the foreign content restrictions on RRSPs have been lifted, they increase their reliance on non-Canadian sub-advisers; inconsistent rules would make it difficult or nearly impossible for foreign firms to comply with Canadian rules, and foreign advisers may decide that dealing with Canadian advisers is more trouble than it is worth, effectively reducing Canadian access to necessary international expertise when it is needed most.

A number of commenters acknowledged that differences exist between the U.S. and the U.K. regulation, and noted that it is more important to harmonize the Canadian requirements regarding soft dollars with the U.S., for the following reasons:

• Canadian market participants are more familiar with U.S. standards;

• the SEC approach of focusing on how a given good or service is being used by the adviser is a preferable basis for determining eligibility for payment with soft dollars, rather than the detailed and complex categorization underlying U.K. rules;

• U.S. advisers are Canadian advisers' true competition for institutional investment management;

• U.S. domiciled advisers that work on behalf of Canadian funds and institutional clients would have a significant advantage under the Proposed Instrument as they would be able to pay for additional items (e.g. raw data feeds) with commission dollars (Canadian advisers would have to pay for these services from their operating budget, leading to lower management fees for U.S. advisers and a flight of capital away from Canadian advisers); and

• Canadian market participants that engage in cross-border business will likely try to ensure that their practices comply with SEC requirements.

- - - - - - - - - - - - - - - - - - - -

Response:

We think that those commenters that suggested we harmonize our requirements with the requirements and guidance of both the U.S. and U.K. may have overlooked the differences between the requirements and guidance items in these two jurisdictions which precludes harmonizing with both. These differences were highlighted in our notice that accompanied the 2006 Instrument.

We agree that harmonization with other jurisdictions is appropriate to the extent it is justifiable in our view to do so and are aware of the importance of harmonizing with the requirements and guidance in the U.S. We have taken all the comments into consideration and have made amendments to the Proposed Instrument to harmonize requirements with those in the U.S. to the extent it is justifiable to do so.

- - - - - - - - - - - - - - - - - - - -

Question 6: Should raw market data be considered research under the Proposed Instrument? If so, what characteristics and uses of raw market data would support this conclusion?

The majority of the commenters were of the opinion that raw market data should be considered research. Reasons given included that:

• raw market data is used to evaluate research generated by others;

• raw market data is a valuable input to advisers that perform their own research, whether on a general basis, or if used in quantitative models and for back-testing of those models;

• quantitative managers and advisers that perform their own research would be put at a competitive disadvantage if they cannot pay for raw market data to use as an input for their own research, compared to advisers that use commission dollars to purchase others' research based on the same market data; and

• allowing raw market data to be considered research would be consistent with the position taken by the SEC, and would ensure a level playing field between U.S. and Canadian managers.

In addition, some commenters stated that the proposed definition and guidance regarding research are inadequate as research does not need to contain original thought, and that data does not need to be analyzed or manipulated to express an opinion, as data can be used by advisers in forming their own opinions and therefore add value to the investment decision making process.

A couple of commenters suggested that although raw market data does not, in and of itself, add value to an investment or trading decision, if it is used as an input to analytics, or with tools for research purposes, it should be considered research. One of these commenters stated that it is incongruous to allow quantitative analytical software as research, but to not allow raw market data which is an input to that software, and added that reasoning should not be separated from the supporting data on which it was based.

Some commenters also argued that raw data has great value, otherwise Bloomberg, Reuters and their competitors would not spend a great deal of money collecting it and selling it to arms-length parties if advisers could do so themselves at a lower cost. Two other commenters added that efforts expended in sorting, ordering and presenting the data in a usable format manifests the thought, knowledge and expression of reasoning necessary to elevate raw data to the status of research. One commenter suggested that while simple quotes and volume information should not be allowed because they are cheap and readily available, some market data that is more difficult and expensive to obtain such as historical depth of market data used in the development of trading algorithms should be classified as research.

Some commenters raised a concern that if raw market data were not permitted as research, advisers would be encouraged to purchase raw data that has been slightly manipulated in order to be able to continue to pay for the underlying raw data with commission dollars. A couple of these commenters noted that the interjection of an intermediary in these circumstances would also likely result in higher costs for the raw data.

However, there were some commenters that did not believe that raw market data should be considered research if it is not analyzed or manipulated. A couple of commenters also indicated that that there is generally no value added from raw market data but that, if the data is used to support modeling applications that provide analyses used to support investment decisions, it should be permitted as there is a clear benefit.

Most of the commenters also agreed that raw market data should fall within the definition of order execution services to the extent it assists in the execution of orders.

- - - - - - - - - - - - - - - - - - - -

Response:

We agree that there are situations where raw market data is used by advisers as an input to their own research efforts, and that such uses could add value to the investment decision-making process. We also agree that to view raw market data as not eligible as research services could put these advisers at a competitive disadvantage relative to those advisers that use commissions to pay for others' research based on the same market data. As a result, we have amended the examples of eligible research services in the Proposed Policy to include market data from feeds or databases that has been or will be analyzed or manipulated by the adviser to arrive at meaningful conclusions -- this would include raw market data.

In making this amendment to the Proposed Policy, we also recognize that the definition of research services, and the guidance provided in relation to the characteristics of such services, would not accommodate the inclusion of raw market data and other potentially valuable inputs to the research process. We acknowledge that goods and services do not necessarily need to contain original thought, or need to be analyzed or manipulated prior to receipt, in order to be used for the benefit of clients by assisting in the investment decision-making process. We have made amendments to the definition of research services in the Proposed Instrument, as well as to the guidance on research services provided under section 3.3 of the Proposed Policy to reflect these views.

We have not changed our previous position that raw market data may also be eligible as order execution services.

- - - - - - - - - - - - - - - - - - - -

Question 7: Do advisers currently use client brokerage commissions to pay for proxy-voting services? If so, what characteristics or functions of proxy-voting services could be considered research? Is further guidance needed in this area?

Four commenters indicated that they use, or are aware of the use of, client commissions to pay for proxy services, while five indicated that they do not use, or are unaware of the use of, client commissions to pay for proxy services.

Most of the commenters that addressed this question believed that proxy services could be considered research to the extent used to support investment decision-making. Examples of the characteristics and uses of proxy services that support this position included:

• proxy voting services assist advisers in assessing the impact of mergers and acquisitions, proxy contests, takeovers, and other proxy proposals on shareholder value;

• they provide analysis of matters to be voted on, along with a recommendation on how to vote proxies;

• they provide research on an investee company's standards of corporate governance or research that assists in monitoring trends in corporate governance; and

• they assess the quality of the issuer's management team or provide analyses, reports or information about the issuer.

Some of these commenters also added that although proxy services should be considered research, there are functions provided by these services that may not be considered research, such as the administrative functions of receiving, voting and returning ballots. These commenters therefore viewed proxy services as mixed-use.

Three of the commenters did not believe that proxy services should be considered research at all. Arguments included that:

• proxy services have administrative and non-research uses that should not be paid for with client brokerage commissions;

• there is no value-added component for proxy services; and

• inclusion of proxy services as research could stimulate undue, costly trading.

One of the commenters suggested that further guidance should be provided on whether components of proxy services that are used to decide how to vote proxy ballots are analogous to traditional "maintenance research" and eligible for payment with client commissions. Two commenters did not feel any additional guidance was necessary.

- - - - - - - - - - - - - - - - - - - -

Response:

We agree that proxy services include products and services that could be considered research services; for example, if they provide information on corporate events such as mergers and acquisitions or constitute an analysis on corporate governance. We also agree that proxy services include functions that would not be considered research services, such as the administrative functions of receiving, voting and returning ballots.

Advisers that have determined that certain proxy services meet the definition of research services should also ensure that the services are used to benefit clients by providing appropriate assistance in making investment decisions for clients. For example, it may be difficult to support the claim that using research services provided by proxy service providers to assist with the administrative function of voting proxies (including if used to assist with decisions on how to vote proxy ballots) on behalf of clients provides appropriate assistance in making investment decisions.

As a result, we think proxy services could be viewed as mixed-use goods and services depending on both content and use. We do not believe any additional guidance is necessary at this time.

- - - - - - - - - - - - - - - - - - - -

Question 8: To what extent do advisers currently use client brokerage commissions as partial payment for mixed-use goods and services? When mixed-use goods and services are received, what circumstances, if any, make it difficult for an adviser to make reasonable allocations between the portion of mixed-use goods and services that are permissible and non-permissible (for example, for post-trade analytics, order management systems, or proxy voting services)?

Eight of the commenters, accounting for approximately half of the respondents, indicated that they use, or are aware of the use by their constituents of client brokerage commissions as partial payment for mixed-use goods and services. Some of the more common types of such goods and services included:

• data services such as Bloomberg and Reuters;

• proxy services;

• order management services; and

• trade analytics.

Two commenters indicated they did not use client brokerage commissions to pay for mixed-used services. One of these indicated that costs for any mixed-use items are treated as corporate operating expenses which are paid for with "hard" dollars. The reasons given were that the allocations would require extensive documentation and could be subject to differences in opinion on the appropriateness of the allocation.

Two commenters indicated that they use, or would use, client brokerage commissions as partial payment for mixed-use goods and services only if they could achieve an objective allocation of costs, for example, if a service had separate identifiable components to which separate prices were attached. One suggested that the criteria for determining whether a mixed-use item may or may not be paid for in part with client commissions should be simple and flexible enough to allow the adviser to make a reasonable determination as to whether a given item is being used to make investment decisions.

Circumstances that can make it difficult for an adviser to make reasonable allocations between the portion of mixed-use goods and services that are permissible and non-permissible included:

• when such goods and services are received as part of a bundled services offering without any cost information from the dealers or any reliable mechanism for separating the component parts, it would be difficult and costly to estimate the value received;

• without prescriptive rules on what is permissible and non-permissible, it would be difficult to make allocations because of the subjectivity involved; and

• there is potential for divergence among dealers in the industry regarding eligible items.

Some commenters suggested approaches to deal with the difficulties in making a reasonable allocation between the permissible and non-permissible portion of mixed-use goods. For example, advisers:

• could make a good faith determination, and keep adequate books and records regarding the allocations;

• could make allocations as judiciously as possible and include their underlying rationale as part of their disclosure to clients; and

• should seek assistance from mixed-use service providers in order to break down the service into component parts that qualify or do not qualify, and obtain a separate costing for each of these components.

One respondent, however, thought that the allocation process is becoming easier as vendors are providing more guidance regarding the research, brokerage and administrative components of their products and services.

- - - - - - - - - - - - - - - - - - - -

Response:

We continue to think that a mixed-use approach is appropriate. We acknowledge that making allocations can be difficult, particularly in relation to goods or services obtained in exchange for bundled commissions. However, client brokerage commissions should not be used to pay for goods and services an adviser obtains that do not meet the definition of order execution services or research services, or that are not used by the adviser to assist in the investment decision-making process or with the arranging and effecting of securities transactions.

Therefore, we think that if an adviser obtains mixed-use services with client brokerage commissions, it should make a reasonable allocation of those brokerage commissions paid according to the use of the goods and services. We have provided additional guidance in the Proposed Policy that for purposes of making a reasonable allocation, an adviser should make a good faith estimate supported by a fact-based analysis of how the good or service is used, which may include inferring relative costs from relative benefits. Factors to consider might include the utility derived from, or the duration the good or service is used for, eligible and ineligible uses.

We also continue to think that advisers should maintain adequate books and records concerning the allocations made in relation to mixed-use items in order to be able to demonstrate their good faith determination of the reasonableness of value received for commissions paid, and to demonstrate that clients have not paid for goods and services from which they do not receive benefit.

While we support efforts being undertaken by vendors to delineate the costs associated with various eligible and ineligible components, the additional guidance provided in the Proposed Policy suggests that an adviser should also be considering its use of the eligible components to assess the extent of its reliance on the vendor-provided cost allocations. For example, an adviser would have difficulty justifying its reliance solely on a vendor's cost allocations to determine the amount that could be paid for with client brokerage commissions if the adviser were to use that portion classified by the vendor as meeting the definition of order execution services or research services for purposes other than making investment decisions or arranging and effecting securities transactions (e.g., if used for administrative or compliance purposes).

- - - - - - - - - - - - - - - - - - - -

Question 9: Should mass-marketed or publicly-available information or publications be considered research? If so, what is the rationale?

The respondents' views were mixed regarding the treatment of mass-marketed or publicly-available information. Specifically, 11 commenters believed that the CSA should follow the SEC's approach and focus on the target of the mass-marketed or publicly available information. That is, information and publications such as newspapers, magazines, or online news that are aimed at a broad audience should not be considered research, but certain information and publications that cater to a narrower audience, such as trade magazines, technical journals, or industry-specific publications may add value to the adviser's investment or trading decisions and should therefore be permitted. Reasons given were:

• mass-marketed information does not have a value-added component that would qualify it as research, but certain publications that are trade, industry, sector or investment specific may be used for further investment decisions;

• mass-marketed information such as newspapers, magazines, periodicals, and online news should not be considered research as they relate to a routine expense for which hard dollars should be paid;

• certain newsletters and trade journals, although publicly available, serve the interests of a narrow audience and can provide an important foundation for unique and independent research; and

• trade magazines, technical journals or industry-specific publications are particularly relevant for managers and traders when conducting research.

One of these commenters suggested, however, that mass-marketed publications in foreign countries should be allowed, as they are not immediately available to Canadian advisers. This would avoid advisers having to rely on foreign brokers to relay this information to them.

Seven commenters indicated that mass-marketed or publicly-available information or publications should not be considered research. Reasons included:

• mass-marketed or publicly available information does not contain sufficiently sophisticated analysis to add value to investment or trading decisions; and

• while there may be some specialized publications that could qualify as research, the CSA should be concerned if some specialized publications that should be considered part of an advisor's continuing education or professional development are included in this category.

Six commenters thought that any publicly available information or publications, whether they are mass-marketed or not, should be considered research. The reasons were as follows:

• mass-marketed or publicly available information may provide valuable information to those knowledgeable enough to draw conclusions from them -- for example mass-marketed material from a European source (possibly in another language) is often not generally known, especially among English-speaking North American analysts;

• the fact that some information is mass-marketed and/or has a lower cost is reflective of the efficiency of the market, not whether it has value to an adviser and, therefore, if an adviser can obtain market and corporate information from such publications versus paying more to a dealer via commissions to obtain the same information, it is better for the client;

• publications like Barron's and the Wall Street Journal can, and do, include exhaustive analysis and research relevant to the investment decision-making process, and also provide information that can move markets; and

• if permissibility is only based on how widely available information is made, then it may run up against issues concerning "insider" information.

Two commenters thought that additional clarification is needed regarding the phrase "publicly available" information given that all publications that are considered to be research are "publicly available".

- - - - - - - - - - - - - - - - - - - -

Response:

We agree with commenters that suggest that publications marketed towards a narrow audience, such as trade magazines, technical journals, or industry-specific publications could provide valuable assistance in making investment decisions and could therefore be paid for with client brokerage commissions.

We continue to think that mass-marketed publications, which are those that are marketed towards a broad, public audience, and are typically of low cost, are more like overhead of an adviser's business and should generally be paid for with an adviser's own funds. Further, we believe many of these types of publications often contain a wide range of information, much of which would either typically not be sufficiently related to the subject matter of the definition of research services (i.e., not related to securities, portfolio strategy, issuers, industries, etc.), or would not provide appropriate assistance in making investment decisions. For these reasons, we believe it would be difficult for an adviser to justify paying for mass-marketed publications with client brokerage commissions.

We have amended the guidance provided in the Proposed Policy to reflect these views. We have also removed reference to the term "publicly available" in relation to these types of goods and services. Even if a publication that is marketed to a narrow audience with specialized interests is publicly available to a broad audience, its availability does not make it ineligible to be paid for with client brokerage commissions.

- - - - - - - - - - - - - - - - - - - -

Question 10: Should other goods and services be included in the definitions of order execution services and research? Should any of those currently included be excluded?

Two commenters did not believe any other goods and services, other than those discussed in the 2006 Instrument and 2006 Policy, should be included.

Other commenters provided examples of other goods and services for which guidance could be provided, as described below.

Seminars

Various commenters believed that seminars should be eligible for payment using client brokerage commissions. Reasons included that:

• seminars are simply an alternative medium by which to communicate information which may otherwise constitute research;

• seminars provide advisers with opportunities to refine their investment decision making process and to generate new analytical methods or investment ideas;

• blanket removal of seminars would hurt small advisers, especially those specializing in exotic areas or high tech areas where the fast pace of change requires constant innovation and learning;

• it is often cheaper for an adviser to pay for one conference and obtain access to multiple analysts than to pay commissions to each of their firms for access;

• some industry leaders only address the adviser community through these events; and

• allowances exist under NI 81-105 for mutual funds to provide seminars and conferences to dealers at no charge, or for mutual funds to pay for these on behalf of dealers, subject to certain conditions relating to the payment for the costs of travel, accommodation and personal incidental expenses.

It was suggested by one commenter that investor conferences sponsored by dealers should be eligible for soft dollar expenses so long as these expenses are reasonable in nature: for example, a trip to New York or Atlanta for a North American media conference is reasonable, while a trip to Aruba for a North American mining conference is probably not reasonable. This commenter also suggested that a compromise solution may be to allow only conference fees to be paid for with commissions.

Another commenter suggested that seminars with more social content than research could be disqualified.

- - - - - - - - - - - - - - - - - - - -

Response:

We agree with commenters that seminars are one method to convey information that may otherwise constitute research services. On this basis, we have amended the Proposed Policy to reflect the view that seminars and conference fees that, in the adviser's judgement, will benefit clients and otherwise meet the requirements of the Proposed Instrument may be paid for with client brokerage commissions. The amendments to the Proposed Policy also would suggest that it would be difficult for an adviser to argue that incidental costs incurred in attending seminars or conferences, such as travel, accommodation or entertainment, could be eligible.

- - - - - - - - - - - - - - - - - - - -

Telephone / Data communication lines

Four commenters supported including dedicated communication lines as an eligible order execution service for the following reasons:

• although the provision of such lines may be solely incidental and not a consideration in an adviser's order routing system decision, the lines nevertheless may be deemed to satisfy the temporal standard for order execution services;

• the lines assist advisers with the timely and accurate entry, handling or facilitation of an order by a dealer and are therefore directly related to order execution;

• banning connectivity hardware used to facilitate electronic trading and direct market access is unfair because it favours dealers and discriminates against advisers -- dealers will charge the adviser for direct market access through commissions expense, but if an advisor were to choose to build a direct connection to the exchange to achieve direct market access and bypass the dealer (a very common occurrence in the U.S.), the hardware costs associated with achieving full connectivity would be precluded from order execution services; and

• such services are permitted by the SEC.

Two commenters argued that if the decision as to what goods and services can be purchased with commissions were based on their use, then eligible goods and services should also include hardware and communication lines as long as the adviser can demonstrate dedicated usage in the order execution or research processes.

One commenter was of the view that