September 23, 2002

Jonathan G. Katz
Secretary, United States Securities and Exchange Commission
450 Fifth Street NW
Washington, D.C. 20549-0609

            Re: Proposed Regulation Analyst Certification, File No.
            S7-30-02, Release Nos. 33-8119 and 34-46301

Dear Mr. Katz:

Thank you for giving the Securities Industry Association ("SIA")1 the opportunity to comment on the above-referenced proposal to strengthen the quality of research reports distributed by U.S. broker-dealers. This proposal should bolster public confidence in the U.S. equity markets by requiring research analysts to address whether their views have been modified by the investment banking interests of their firms. Strengthening investor trust and confidence is critical to our markets, and we believe that the Securities and Exchange Commission ("Commission" or "SEC") is taking a salutary step toward that end in proposed Regulation AC. For its own part, SIA last year adopted best practices to deal with this issue, recognizing that taking steps to disclose such conflicts is an important goal of the securities industry. Although we believe that best practices can often be effective in raising standards in the securities industry, we supported the efforts of the National Association of Securities Dealers ("NASD") and the New York Stock Exchange ("NYSE") to adopt regulations in this area superseding our Best Practices, and we strongly support this proposal by the Commission as well.

Introduction.

Proposed Rule AC represents a logical progression in the Commission's leadership in seeking to bolster the objectivity of analyst research. The past six months have seen the following regulatory steps:

  • The adoption on May 8, 2002 of new rules by the NASD and NYSE that, among other things limit communications between investment banking personnel and research analysts concerning research reports, restrict research analysts' personal holdings of the securities of companies that they cover, and require disclosure in research reports about potential conflicts posed by investment banking business;

  • A massive ongoing regulatory investigation conducted jointly by the SEC and state securities regulators, resulting to date in a May 21, 2002 settlement agreement between the Attorney General of the State of New York and Merrill Lynch, Pierce, Fenner & Smith Inc. in which the firm, without admitting any violations of state or federal law, agreed to a number of procedural reforms regarding its research practices; and

  • Enactment on July 30, 2002 of the Sarbanes-Oxley Act, Pub. L. No. 107-204, which contains a number of provisions intended to address potential analyst conflicts of interest.

SIA supports all of these regulatory and legislative actions. We also believe that it is important that regulations in this area be consistent and uniform with each other. For this reason, this letter will focus on areas where we believe the Commission should be mindful of interpreting and administering Regulation AC in a manner consistent with the work already done, particularly the quite comprehensive regulatory framework created by the NASD and NYSE. SIA agrees with the principle that those responsible for producing research reports should personally agree with the views expressed in those reports. Existing law supports this principle, and we applaud proposed Regulation AC for making this principle even more explicit. SIA therefore strongly supports the proposal. We have some requests for clarification and some suggestions for strengthening the requirement.

Discussion.

    A. The Function and Scope of Analyst Certifications.

  1. The Meaning of the Certification. The proposed certification requirement includes a statement attesting whether or not part or all "of the research analyst's compensation was, is, or will be, directly or indirectly, related to the specific recommendations or views expressed by the research analyst in the research report . . . ."2 This language is unduly broad. Insofar as analysts are paid to write research reports that contain specific recommendations or views, potentially all of an analyst's compensation can be said to be "directly or indirectly related" to those recommendations or views. We believe that the Commission's objective here is to get analysts to certify that they sincerely hold the views expressed in their reports, and to disclose whether any compensation is being paid or held out that is intended to induce the analyst to be unduly positive in his or her recommendations.

    The SEC's adopting release states that the rule is not intended to prohibit analysts from receiving compensation for covering issuers or for preparing research reports, nor is it intended to preclude analysts from being compensated based on the accuracy of their recommendations.3 This is an important principle, since compensating analysts based on the performance of their recommendations powerfully aligns their interests with those of investors. However, the SEC needs to go further to remove uncertainty as to exactly what sort of compensation must be disclosed. It is important for the SEC to either revise this broad language in the rule, or to state in the adopting release that the compensation that is required to be disclosed is that which is intended to induce an overly favorable particular view or recommendation, not compensation paid for writing a research report, compensation based on the performance of the analyst's recommendations, or compensation for other services provided to investors.

  2. Certification Should Not Impede Revision, Supervision or Oversight. The proposed rule could create some confusion within firms about how analysts and their supervisors can go about changing a research report. We do not believe that the Commission intends the rule to disrupt normal processes aimed at (i) avoiding potential conflict, liability and regulatory problems, or (ii) strengthening, prior to publication, a report's accuracy, insightfulness and prescience.

    1. Supervision. The SEC should clarify that certification is not intended to prevent research supervisors or review committees from seeking changes in the final research product. Expanded use of review committees was a key provision of a widely-noted recent regulatory settlement that several major firms have since adopted.4 Research supervision, including review committees, play an important role in ensuring accuracy, intellectual rigor, and objectivity in research reports. The Regulation should not impact the ability of an analyst to reconsider both factual and analytic aspects of a draft report in response to comments from a supervisor or review committee. Moreover, Regulation AC should not impede the ability of a supervisor or a review committee to require that a report follow a written firm policy on when a research report can upgrade or downgrade a rating, or when it must withhold a rating, to prevent conflicts or for other reasons. It would be helpful if the SEC's adopting release could clarify that following a direction made in accordance with a firm's policies and procedures to change or withhold a rating in a draft research report, or to reconsider some aspect of a draft report's analysis, does not automatically preclude an analyst from certifying that the views in the report accurately reflect his or her personal views. Rather, it would remain a matter for the analyst's individual judgment whether the views expressed in the final version of the report reflect his or her personal views, after taking into account supervisory comments.

    2. Compliance Oversight. Certification also should not preclude the ability of an analyst or a firm to modify the report to remove statements that create legal or regulatory concerns. For example, the legal or compliance department may have legitimate concerns that overly forceful statements might pose libel or defamation concerns for the firm. There may also be "gun-jumping" concerns with the content or timing of the report, or other compliance concerns posed by Securities Act Rule 139, if at the time the report is being prepared the firm has reached agreement to act as underwriter for the issuer. The SEC should clarify that a report can legitimately be influenced by such considerations, as long as the analyst is able to certify that the final product accurately reflects his or her personal views at the time of publication, as those views stand after consideration of compliance oversight comments.

    3. Revisions. Although not required by the rule, it would be helpful for the SEC to clarify that the certification does not preclude the ability of an analyst to refine his or her thinking between a first draft of a report and the final product as a result of receiving the benefit of the views of others in the research department. One's sincerely held views about a company or a security can change as additional information and perspectives from colleagues arrives, or as one reflects further on the draft report. The rule should not give analysts the misimpression that their view of a company must be frozen in place once it is first reduced to writing in a draft report.

  3. Certification should be required only of lead analyst on the report, not of everyone whose name is listed. Proposed Regulation AC would define research analyst as "any natural person who is principally responsible for the analysis of any security or issuer included in a research report." The development of the analysis contained in a research report can be based on the contributions of several persons and all such persons may or may not be named in a research report with respect to a specific security or issuer. Multiple certifications by all named research analyst staff in each report are likely to be confusing, result in lengthy and repetitive disclosures, and may occasionally delay the issuance of time-sensitive reports, without advancing the SEC's goal of strengthening investor confidence in research reports. Therefore, we recommend that the SEC clarify that only the most senior or lead research analyst responsible for a security or issuer would be required to provide a certification to be included in the research report.

  4. Third-Party Research. In situations where a broker-dealer redistributes research produced by another broker-dealer, the proposed rule and proposing release are not clear on whether the responsibility for obtaining the required certification, and for maintaining records about public appearances of a research analyst, lies with the broker-dealer that employs the analyst, or the broker-dealer that distributes the research. These obligations should rest with the broker-dealer that employs the analyst, since only that firm is in a position to ensure compliance. Moreover, if the regulation is interpreted to require broker-dealers that merely distribute third-party research to provide certifications and maintain records of public appearances, these firms would face substantially increased costs and therefore may discontinue providing this valuable service to their customers. It would be helpful if the Commission could clarify that the obligations of the rule apply to the firm originating the report, not to a firm that only distributes the final product.

  5. Private Liability. SIA appreciates the Commission's statement that, other than restating the existing privately actionable duty under the antifraud provisions of the federal securities laws, "[p]roposed Regulation AC . . . does not impose any new liability." As the Commission notes, analysts may currently be found to have violated Section 10(b) of the Exchange Act if they make baseless recommendations or recommendations that they disbelieve. The proposing release also states "Regulation AC is not intended to create duties under Section 10(b) of the Exchange Act." It would be helpful if the Commission could restate that its action in adopting Regulation AC is meant only to reinforce this existing legal obligation, and is not intended to create any other new duties actionable under any other federal law, nor under any state or common law claim.

    B. Consistency with SRO Rules.

The Commission's proposing release notes that proposed Regulation AC "is intended to complement other rules governing conflicts of interest disclosure by research analysts, including NYSE Rule 472 and NASD Rule 2711." In order to ensure that Regulation AC is applied in complementary fashion with these other rules, it would be helpful for the Commission to clarify in its adopting release that certain aspects of the rule are to be interpreted consistently with the NASD and NYSE rules, as well as with guidance that the NASD and NYSE have issued concerning their rules.5 We have identified the following issues under Regulation AC where it would be helpful to make minor adjustments to the proposed rule, or to have a statement from the Commission explaining that Regulation AC is to be interpreted in similar fashion to NYSE Rule 472 and NASD Rule 2711 in many respects.

  1. "Any and all securities of the subject securities or issuers." The proposed certification requirement requires some modification in order to make it consistent with the approach taken by the NASD and NYSE regarding reports that express views about multiple issuers. We have two concerns about the certification requirement in this context, the requirement as applied to compendium research, and the requirement as applied to quantitative analysis.

    Compendium Reports. With regard to compendium research reports, the NASD and NYSE provided exceptions from the general disclosure provisions of their rules for research reports covering six or more subject companies. Under the NASD and NYSE rules, these reports need not contain the required disclosures as long as they "direct the reader in a clear manner as to where they (sic) may obtain applicable current disclosures in written or electronic format."6 Another workable alternative to the SROs' approach would be for the SEC to permit a single certification in the compendium report signed by all of the lead analysts mentioned in the report, specifying which sections of the report each analyst is certifying. SIA would support either or both approaches.

    Quantitative and Technical Analysis. While the SEC does not specifically ask for comment on the application of the proposed rule to quantitative or technical analyses, this research also requires some accommodation by Regulation AC. We urge that the SEC carve out all such analysis from the rule. Quantitative analysis contains recommendations derived from quantitative mathematical models based on objective publicly-known measures of stock performance, or by comparing the relative performance of a security to a benchmark (such as a comparison between a number of equity securities and an index such as the S&P 500). These reports may recommend transactions in specific issuers based on this analysis, or they may recommend trading strategies, quantify value or offer forecasts for entire sectors. Technical analysis evaluates a security based on factors such as imbalances in supply and demand for the security or interest rate sensitivity of the security. Like quantitative analysis, this analysis is based on objective factors that have nothing to do with the specific nature of the issuer or its business, and so the report does not raise any conflicts concerning any potential investment banking opportunities with the issuer.

    Since these reports are based on objective criteria, rather than on judgments about the business prospects or business fundamentals about particular companies, it is hard to see the relevance of the certification requirement to these reports. At a minimum, SEC should follow the lead of the NASD and NYSE and offer guidance in its adopting release that the rule is not intended to cover reports of broad-based indices, reports commenting on economic, political or market conditions, or technical analysis concerning demand or supply for a sector, index or industry, statistical summaries of multiple companies' financial data, or reports that recommend increasing or decreasing holdings in particular industries or sectors but that do not contain recommendations or ratings for individual securities. The Commission should consider going farther than the SRO Joint Guidance in some respects. For example, we believe that quantitative analysis of individual companies, not just industry sectors, should be exempt from both the NASD and NYSE requirements and Regulation AC.

  2. "Clear and Prominent Certification". The SRO Joint Guidance indicates that disclosure is sufficiently prominent if it is on the front page of the research report, or the front page of the report refers to the page where the disclosures are found.7 It also provides that electronic reports can utilize hyperlinks to this disclosure, as long as the first screen that the investor sees clearly and prominently labels these hyperlinks. In the case of electronic reports that cannot employ hyperlinks (such as a report in PDF format) the guidance directs firms to follow the same approach as for paper reports. We recommend that the SEC offer similar guidance to the meaning of "clear and prominent certification" in Regulation AC.

  3. "Research Report." The definition of "research report" would be more accurate, easier to administer and better focused on information important to investors, if it were defined as "a written or electronic communication that includes an analysis of the securities of an issuer or issuers, provides information reasonably sufficient upon which to base an investment decision, includes a recommendation, and is branded or held out to investors, as an original research report on the securities of an issuer or issuers."

    This modified definition would clarify that the definition is intended to pick up electronic communications. It is likely that the Commission did intend to cover electronic as well as written reports, but such a clarification would avoid uncertainty. The SEC's adopting release should give guidance on the applicability of the rule to electronic communications that is consistent with the SROs' guidance.

    Including the concept of "branding" in the definition8 would avoid applying the Regulation to other types of materials that are already subject to comprehensive regulation as sales materials, and that are not produced by the broker-dealer's research department or held out to investors as research. Particularly for purposes of Regulation AC, which is focused on direct conflicts of analysts, it is appropriate to clarify that the rule is not intended to extend to sales and trading personnel who only repackage views, and do not present the product as original research. This would also avoid the application of the rule to purely internal communications within the firm. If the SEC chooses not to incorporate the "branding" element into its definition, it would be helpful if the SEC could clarify that the term "research report" is not intended to reach purely internal communications that are not distributed to customers. The NASD and NYSE, which adopted substantially similar definitions to the proposed SEC definition of the term "research report" in new NASD Rule 2711 and amended NYSE Rule 472, offered interpretive guidance that the definition is not intended to pick up such reports.9

  4. Application to investment advisers. SEC asks for comment on whether the rule should explicitly exclude investment advisers. Footnote 16 of the release indicates that investment adviser who is not principally responsible for preparing research reports is excluded from Regulation AC, even if dually registered. This is consistent with the guidance given by the SROs on NASD Rule 2711 and NYSE Rule 47210. We support this view.

    C. Public Appearances.

  1. Disclosures in Public Appearances. The proposing release poses the question "What disclosures, if any, should be required during public appearances?" We believe that the disclosures required by the existing SRO rules concerning public appearances are sufficient, and this is a situation where more would be less. The most widely seen public appearances by research analysts are probably on cable television business shows. These appearances are squeezed between commercial breaks and coverage of other events and stories. Consequently, such appearances range from "sound bites" measured in seconds, to a few minutes at most. We are concerned that requiring additional disclosures on top of those required by the recently adopted SRO rules would crowd out much of whatever substantive remarks an analyst has to make. Moreover, additional disclosures might clog these appearances with so much cautionary disclosure language that viewers would begin to consider the disclosures as mere boilerplate verbiage to be disregarded. This result would not be desirable for investors, the broadcast media, or the research analyst community.

  2. Certification. The Commission requested comment on whether an analyst should be required to list all public appearances over the prior calendar quarter as part of the certification statement submitted to the broker/dealer. This requirement already exists under the NASD and NYSE analyst rules, 11 and adding a similar requirement here only creates an additional recordkeeping burden on analysts without adding any regulatory oversight value.

Conclusion.

Thank you for giving SIA this opportunity to comment on proposed Regulation AC. We believe that the proposal will substantially improve public trust and confidence in research analysts and we urge its swift adoption. If you have any questions on any aspect of this letter, please contact either me or George Kramer, Vice President and Associate General Counsel, at 202-296-9410, or by e-mail to skaswell@sia.com or gkramer@sia.com.

              Sincerely,

              Stuart J. Kaswell
              Senior Vice President and
              General Counsel

Cc: The Honorable Harvey L. Pitt, Chairman
The Honorable Paul S. Atkins, Commissioner
The Honorable Roel C. Campos, Commissioner
The Honorable Cynthia A. Glassman, Commissioner
The Honorable Harvey J. Goldschmid, Commissioner
Annette L Nazareth, Director, Division of Market Regulation
James A. Brigagliano, Assistant Director, Trading Practices, Division of Market Regulation

____________________________
1 SIA brings together the shared interests of more than 600 securities firms to accomplish common goals. SIA member firms (including investment banks, broker-dealers, and mutual fund companies) are active in all U.S. and foreign markets and in all phases of corporate and public finance. The U.S. securities industry manages the accounts of nearly 80 million investors directly and indirectly through corporate, thrift and pension plans. In 2001 the industry generated $198 billion in U.S. revenue and $358 in global revenue. Securities firms have approximately 750,000 employees in the United States. (More information about SIA is available on our web site: http://www.sia.com).
2 Proposed Rule 242.501(b)(1) and (b)(2)(i)(emphasis added).
3 "The Commission . . . notes that the proposed regulation is intended to address analysts' beliefs about their expressed views and recommendations, not the accuracy of the recommendations or opinions regarding securities discussed." 67 Fed. Register 51510, at 51512 (August 8, 2002).
4 See http://www.oag.state.ny.us/investors/merrill_agreement.pdf.
5 See Joint Memorandum of the NASD and the New York Stock Exchange, June 26, 2002 ("SRO Joint Guidance").
6 NASD Rule 2711(h)(11), NYSE Rule 472(k)(2)(v).
7 SRO Joint Guidance at 12.
8 SIA also argued in connection with the SRO rules that "a recommendation or commentary that is not held out or `branded' as independent research by an equity analyst should not be covered by the rules." Letter to Jonathan G. Katz from Stuart J. Kaswell, April 11, 2002, at 11.
9 SRO Joint Guidance at 5.
10 Id. at 3.
11 Id. at 14.