SULLIVAN & CROMWELL
TELEPHONE: 1-212-558-4000
FACSIMILE: 1-212-558-3588
WWW.SULLCROM.COM
125 Broad Street
New York, NY 10004-2498


LOS ANGELES • PALO ALTO • WASHINGTON, D.C.
FRANKFURT • LONDON • PARIS
BEIJING • HONG KONG • TOKYO
MELBOURNE

April 18, 2002

Jonathan G. Katz, Secretary,
Securities and Exchange Commission,
450 Fifth Street, N.W.,
Washington, D.C. 20549-0609.

Proposed Rules Regarding Research Analyst Conflicts of Interest-
File Nos. SR-NASD-2002-21 and SR-NYSE-2002-9

Dear Mr. Katz:

             We are writing in response to the Commission’s request for comments regarding Release No. 34-45526 (the "Release"), in which the Commission solicited comments on the proposals by the National Association of Securities Dealers, Inc. (the "NASD") to adopt new NASD rule 2711, and by the New York Stock Exchange (the "NYSE") to amend NYSE rules 472 and 351.

             We have set forth below certain comments on specific aspects of the proposals that we believe, based on our review of the NASD and NYSE rules, as well as discussions with our clients, can be improved or clarified in a manner consistent with the purpose of the proposals.

             As an initial matter, which is reflected in further detail in many of our comments below, we believe the rules should be made identical to the extent practicable. Many broker-dealers will be subject to both of the proposed rules. If they are, they will need to comply with the more onerous of the two regimes. If they are not, they may be at a competitive advantage or disadvantage to other firms.

Imposition of Quiet Periods

             NASD rule 2711(f) and NYSE rule 472(f) would establish "quiet periods" during which NASD and NYSE members that manage or co-manage securities offerings would be prohibited from issuing research reports regarding the issuers in such offerings.

             As an initial matter, we do not believe that an additional limitation needs to be placed on the timing of research reports following non-IPOs above and beyond that contained in Rule 139 under the Securities Act of 1933. Rule 139 permits research to be issued only if it is "distributed with reasonable regularity in the normal course of business". Research of this type is unlikely to raise the same level of concern about conflicts that "start-up" research might raise. Also, unlike start-up research, Rule 139 research is most likely to have an audience that expects the regular flow of research to continue without interruption. As for Rule 139(b) research, we think the rationale for imposing additional restrictions is particularly weak. We think it is unlikely that an analyst or a firm motivated by a desire to promote a particular issuer’s secondary offering would use research with an industry-wide, balanced focus (as required by Rule 139(b)) to do so. More generally, Rule 139 has been developed by the SEC over many years of experience with offerings, disclosure issues and information flows and should not be disrupted by new rules that conflict with them without addressing the exceptions that have developed for seasoned issuers set forth in that Rule.1 It is also inconsistent with a goal of increasing information available to all investors, particularly relating to seasoned issuers.

             If the NASD and the NYSE determine to retain the concept of quiet periods, there are some important corrections that should be made. The spirit of the quiet-period proposals, both through the definition of research reports, as well as the definition of initial public offering in the NYSE rule, appears to be directed at offerings of equity securities.2 However, the drafting of these two rules could be read to apply to offerings of debt securities as well. As a result, we believe that the limitations set forth in NASD rule 2711(f)(1) and (2) and NYSE rule 472(f)(2) should be clarified to only apply to equity offerings.3 We also believe that only offerings of equity securities by the issuer should be covered, because equity offerings by security holders do not present the same conflict issues as offerings by the issuer.4 In many cases, the issuer has no influence over the underwriter that is used, or the timing or frequency of the offerings by selling security holders.

Disclosure of Conflicts

             NASD rule 2711(h)(1)(C) and NYSE rule 472(k)(1)(i)(c) would require members to disclose in research reports, and analysts to disclose in public appearances, any "actual, material conflicts of interest" of the member of which the member or analyst knows or has reason to know. 5

             First, the proposed rules include no guidance as to what would constitute an actual, material conflict of interest. While firms may have a workable sense of what is material in the context of an issuer or its securities, what is material in the context of issuing research is much less developed or clear. For example, if a firm has maintained proper "Chinese wall" procedures so that investment banking relationships with an issuer that is the subject of a research report are not known to the analyst preparing the report and members of the investment banking department have not influenced the issuance of that report, would those relationships be material? We do not think they would be, and we urge the SROs to make that clear. Also, a large member firm may have many relationships with a company, and in an effort to ensure compliance with the rule may be over-inclusive in its disclosure. We believe that over-disclosure not only would place an undue burden on member firms, but also would dilute the value of truly material information. We also think that many of the client relationships that firms may feel compelled, in the absence of adequate guidance from the SROs, to disclose would be of interest only to their competitors and those of the clients.

             Second, and perhaps most importantly, these rules could create a serious problem regarding disclosure of material, non-public information about a firm’s investment banking clients. If by "actual, material conflict of interest" the SROs have in mind a significant investment banking relationship with a company that is covered by a research report, then the proposed disclosure requirement could require the firm to disclose such engagements even when they involve significant, unannounced transactions, such as a public offering or a merger. Disclosure of confidential engagements to research clients raises serious issues of selective disclosure and perhaps even tipping. It also could violate a firm’s duty to its investment banking clients to maintain the confidentiality of their affairs. Of course, a firm could elect to prohibit the issuance of research at any time when such a relationship exists, but that step, too, is likely to signal the research department and its clients that "something is up", especially in circumstances where those clients would normally expect to receive research (e.g., after significant news or a significant event). Unless the disclosure requirement is written in such a way as to permit a firm to avoid disclosing its confidential investment banking relationships, we think the requirement is contrary to the "Chinese wall" procedures that firms are required to follow and is unworkable.

             One way to avoid this problem would be to permit a firm to conclude that a confidential investment banking relationship was not "an actual, material conflict of interest" as long as it had not been disclosed to the research analyst and the investment banking department had not tried to influence the research analyst with regard to the report. We note that the proposed rules permit an analyst to appear in public without disclosing any conflict of the firm of which the analyst is unaware. Why should the standard be any different when the analyst publishes a research report? Another way to address this problem would be to provide an exception from the requirement that actual, material conflicts of the firm be disclosed for any research report issued in response to significant news or a significant event, since research of that kind is unlikely to be motivated by inappropriate purposes and it is in precisely those circumstances that withholding research in order to preserve confidentiality is most likely to alert research clients that something is up.

             For the reasons above, we suggest that NASD rule 2711(h)(1)(C) and NYSE rule 472(k)(1)(i)(c) either be removed entirely, replaced with a requirement for the member to provide a general cautionary disclaimer regarding potential conflicts of interest or replaced with specified categories of relationships that the NASD and NYSE believe create an actual, material conflict of interest. A disclaimer would alert investors to the risks associated with relying on research yet would not lead to disclosures of confidential relationships or relationships that would only be of interest to competitors of the issuer and the member. A categorical approach to disclosure would avoid many of the problems inherent in an open-ended ‘materiality’ standard. However, if the new rules are to contain a requirement that firm conflicts be disclosed (even if limited to specified categories of relationships), we urge the SROs to modify the proposed rules to make clear that confidential investment banking relationships that are unknown to the research analyst, where the firm has maintained appropriate Chinese walls, are not "actual" conflicts and thus need not be disclosed.

             Finally, in both the above sections, as well as NASD rule 2711(h)(2)(B) and the last sentence of NYSE rule 472(k)(1)(ii), we would recommend deleting "or has reason to know." Given Chinese wall issues, it is not clear what this phrase means (i.e., does the analyst have reason to know matters that other areas of the firm knows). Most investment banks have strict controls over what analysts may know about other parts of the organization, and it is not appropriate or desirable to impute knowledge to them that they should not have or are otherwise forbidden from possessing by their firms.

             The proposed requirement that disclosure of compensation that is anticipated in the next three months is also troublesome for the reasons discussed above. Although disclosing compensation without revealing the source or amount is less likely to alert the market that a significant investment banking transaction involving the company is imminent, the risk remains and is real. Market participants are not naïve; they can be expected to put two and two together in the right circumstances, and to look for this kind of disclosure as evidence that something might be up. We urge the SROs to eliminate the requirement for disclosing the firm compensation or at least to require only a generic warning that such compensation could be received.

Personal Trading

             NASD rule 2711(g)(2) and NYSE rule 472(e)(2) and (3) would limit personal trading by research analysts and associated persons.

             The inclusion of the term "member" in the definition of "associated person" in the NYSE rule, as discussed in footnote 5 above, has serious negative implications. For example, NYSE rule 472(e)(1) would prohibit venture capital investments in companies by investment banks and their affiliates, NYSE rule 472(e)(2) would prohibit market making activities by investment banks during the blackout periods and NYSE rule 472(e)(3) would prohibit almost all market making activities. We trust this is an unintentional drafting error.

             The NYSE rule is also much broader than the NASD rule in that it prohibits trading with respect to any research report by the organization, not just those that the analyst follows. For an organization that has numerous research analysts who cover a large number of companies, the broader restriction of the NYSE rule would severely limit the investment opportunities for the analysts. Accordingly, we recommend that the NYSE rule be limited in the manner that the NASD rule is in this area.6

*     *     *

             We appreciate this opportunity to comment to the Commission on the proposed NASD and NYSE rules, and would be happy to discuss any questions the Commission may have with respect to this letter. Any questions about this letter may be directed to Steven B. Stokdyk (310-712-6624) in our Los Angeles office, or David B. Harms (212-558-3882) or John T. Bostelman (212-558-3840) in our New York office.

Very truly yours,


SULLIVAN & CROMWELL




Endnotes
1 The definition of secondary offering in NYSE rule 472.110 appears to exclude securities not subject to Regulation M, such as actively traded securities, investment grade non convertible and asset-backed securities and certain other securities specified in Rule 101(c) of Regulation M. This is helpful, and at a minimum should be added to the NASD rule.
2 We understand that this is the intention of the two regulatory organizations, and accordingly do not include any policy reasons regarding our support of the rules being limited in this manner.
3 This would be accomplished by (1) deleting the phrase ", but are not limited to," in the first sentence of NYSE rule 472.10(2), (2) adding the phrase "of equity securities" after "an initial public offering" in NASD rule 2711(f)(1) and after "a secondary offering" in NASD rule 2711(f)(2) and (3) adding the word "equity" before "securities" in the third line of NYSE rule 472.110.
4 This would be accomplished by (1) adding the phrase "by the issuer" after "a secondary offering" in NASD rule 2711(f)(2) and (2) deleting the phrase "or a registered offering by persons other than the issuer" in NYSE rule 472.110.
5 There is a very important drafting issue with the NYSE rule that makes it significantly more onerous than the NASD rule, and potentially unworkable. The NYSE rule includes the term "member" in the definition of "associated person" which means that an analyst must disclose in a public appearance any conflict about which the member knows of as opposed to conflicts about which he or she knows. This can be rectified by replacing the phrase "a member, allied member, or" with the word "any" in the definition of "associated person" in NYSE rule 472.40.
6 In addition, NYSE rule 472(e)(3) should be clarified to apply only to equity securities by inserting "in a research report" in the second line before the parenthetical.