Robert C. Mendelson

April 18, 2002

Jonathan G. Katz
U.S. Securities and Exchange Commission
450 Fifth Street, NW
Washington, DC 20549

Re: National Association of Securities Dealers and New York Stock Exchange Proposals Relating to Research Analyst Conflicts of Interest, File Nos. SR-NASD-2002-21 and SR-NYSE-2002-09

Dear Mr. Katz,

Thank you very much for giving us the opportunity to comment on the above referenced proposals regarding research analyst conflicts of interest.1 The proposals by the National Association of Securities dealers ("NASD") and the New York Stock Exchange ("NYSE") are an attempt to implement greater regulatory oversight of securities analysts at U.S. broker-dealers, by increasing disclosure of research analyst conflicts and establishing substantive restrictions governing research analyst practices.

The proposals, among other things, seek to: (1) impose limits on how a firm's investment banking and research department may interact; (2) require disclosure of financial interests held by the firm and the analyst (including the analyst's household); and (3) require firms to describe their respective rating systems and price target methodology.

We agree with the overall goals of the proposals - to improve the objectivity of research and provide investors with more useful and reliable information when making investment decisions. Proposed NASD Rule 2711 ("Research Analysts and Research Reports") and proposed amendments to NYSE Rule 472 ("Communications with the Public") (collectively, the "Proposed Rules"), however, are unclear in several respects and would impose significant obligations and burdens on broker-dealer firms, especially smaller broker-dealer firms. We believe the Proposed Rules, if approved as submitted, would create difficult legal and compliance issues, and significant supervisory challenges for broker-dealer firms to address in a short time frame. Moreover, the Proposed Rules may have the unintended effect of limiting research coverage and diminishing the dissemination of important information to investors. The Proposed Rules may also result in a reduction of valuable investment banking services to issuers.

I. Summary and Overview

We support the attempt to limit analyst conflicts of interest, but have specific concerns and questions regarding the Proposed Rules. As noted in more detail below, we believe that the costs and burdens of complying with the rules as proposed will be significant, while the benefit to investors is far from clear and far from certain. The impact on small firms will be especially acute. We also believe that certain discrepancies between the NASD and NYSE's proposals will need to be addressed to prevent difficulties and ambiguities in, as well as added costs of, compliance.

We believe that some of the significant issues with the Proposed Rules include:

II. The Proposed Rules

The Proposed Rules attempt to address analyst conflicts of interest in a number of ways. Specifically, the Proposed Rules would prohibit:

The Proposed Rules also would impose several disclosure requirements on broker-dealer firms and research analysts concerning their relationship with, and financial interest in, a subject company and its securities, as well as disclosure related to a firm's rating system and identification of conflict of interests.

Finally, the Proposed Rules would require a firm to adopt written supervisory procedures "reasonably designed" to ensure that the firm and its employees comply with the Proposed Rules. Further, a firm's senior officer must attest annually to the NASD or NYSE that the firm has adopted and implemented such procedures.

III. Discussion

Certain aspects of the NYSE and NASD's proposals are substantially different. These disparities will result in difficult compliance and interpretive issues, as well as unnecessary added costs for firms. Moreover, these differences are likely to confuse investors. We respectfully suggest that the Proposed Rules be harmonized to a greater extent to avoid such problems.

A. Definitions

We respectfully suggest that certain of the definitions be modified as they are ambiguous and overly broad, and, as written, could have an enormous impact on the application of the Proposed Rules overall.

1. The NASD defines "Research Report" as "a written or electronic communication that the member has distributed or will distribute with reasonable regularity to its customers or the general public, which presents an opinion or recommendation concerning an equity security." We believe this term is defined too broadly. As currently written, the definition may capture documents including technical and quantitative research, general market commentary, reports by investment advisors, or company updates that do not contain a rating change or price target. In addition, it is unclear whether updates reporting on announcements from a company would constitute a "Research Report?" Would the answer change if it repeats what the firm's rating is for the subject company?

In addition, the proposed definition of "research report" focuses on whether a member "has distributed or will distribute" a research report, rather than the authorship of the research report. As a result, it is unclear whether this definition would capture third-party research that a firm may make available to customers, such as research reports prepared by independent research firms, other U.S. broker-dealers, or foreign broker-dealers whose research is distributed in the United States pursuant to Rule 15a-6(3) under the Securities Exchange Act of 1934 ("Exchange Act").

The phrase "has distributed or will distribute" also raises the question of to how many customers does a written or electronic communication have to be distributed for it to constitute a "Research Report." The inclusion of electronic communications in the definition of "Research Report" may also impede the efficient and swift dissemination of useful information in a fast moving market. For example, the transmission of e-mail alerts to wireless devices could be limited given the number and type of disclosures that will be required to accompany a "Research Report." One possible solution would be to permit the issuance of "Research Reports" that tell readers where to find the required disclosures.

We note that the NYSE and NASD have proposed very different definitions for "research report." Such a difference will have a substantial impact on the overall application of the Proposed Rules. Specifically, the NYSE's proposed definition states that "Research Reports" "are reports defined as, but not limited to, an analysis of equity securities of individual companies, or industries, which provide information reasonably sufficient upon which to base an investment decision." The NYSE's failure to limit the definition to written or electronically transmitted communication, or communications distributed with reasonable regularity, as well as its use of the phrase "but not limited to" raises questions regarding the scope of the definition.

Market commentary from a chief economist that suggests moving assets to or from equities or an industry sector and mentions specific "bell weather" or other securities could well be deemed a "research report" under either definition. We suggest that the scope of the rule be limited to "written" form, i.e., electronic or paper documents, or other readily reproducible media. We also suggest that it be limited to analyses and recommendations focused on one or several equity securities or an industry sector.

2. The term "Research Analyst" is defined as an "associated person who is principally responsible for, and any associated person who reports directly or indirectly to such a research analyst in connection with preparation of the substance of a research report, whether or not any such person has the job title of research analyst." This definition does not include the concept of preparation of a research report as part of a person's regular duties. As a result, it may be argued that someone may be deemed a "Research Analyst" even if their participation in preparing the report is only sporadic and not part of his regular duties.

Although the NASD states in the Proposing Release that the term "Research Analyst" would not include every associated person who may express an opinion on an equity security, it is not clear exactly who may be captured by this definition. For example, would a salesperson that periodically prepares his own views on companies and disseminates them to selected clients be deemed a "Research Analyst?" Moreover, would the dissemination of these views be deemed a "Research Report" for purposes of the rule?

Furthermore, although the NASD stated in the Proposing Release that most mutual fund portfolio managers would not be deemed to be "Research Analysts" when discussing the mutual fund's portfolio holdings in appearances, the proposed definitions of "Research Analyst" and "Research Report" do not address a portfolio manager's practice of distributing written or electronic reports. As a result, one may infer that the Proposed Rules may apply to a portfolio manager who distributes written or electronic communications that discuss equity securities. At a minimum, we believe that the Proposed Rules should be clarified so that they do not apply to a portfolio fund manager who distributes written or electronic communications. Our preferred approach would be to include a clear statement that mutual fund portfolio managers, registered or otherwise, are not subject to the Proposed Rules. Absent such clarification, the Proposed Rules could have the unintended and unhealthy effect of causing portfolio fund managers to terminate their registration.

3. "Member of a research analyst's household" is defined as "any individual whose principal residence is the same as the research analyst's principal residence." This definition is overly broad. The term is used in the definition of "Research Analyst Account." Presumably, this could require the research analyst or his firm to keep track of and disclose the trading and holdings of a household member of a research analyst, including a roommate. Such tracking will be extremely difficult and burdensome for firms.

Perhaps the definition could be modified to apply to particular family members of the research analyst or to persons that the research analyst materially supports (akin to the NASD's free riding and withholding interpretation). The NASD has proposed Rule 2790 relating to trading in hot equity offerings to replace the current Free-Riding and Withholding Interpretation, IM-2110-1. As proposed, the NASD would define "immediate family member" to "include a person's parents, mother-in-law or father-in-law, spouse, brother or sister, brother-in-law or sister-in-law, son-in-law or daughter-in-law, and children, and any other individual for whom the person, directly or indirectly, provides material support. The NASD also proposes to define "material support" to mean "providing more than 10% of a person's income or expenses." Material support will be presumed for members of the immediate family living in the same household.2

4. "Research Analyst Account" is defined as "any account in which a research analyst or member of the research analyst's household has a beneficial interest, or over which such analyst or household member has discretion or control, other than an investment company registered under the Investment Company Act of 1940." The meaning of the phrase "over which such analyst or household member has discretion or control" contained in the definition - if meant to suggest something more than beneficial ownership (which, in the text of the Proposing Release, the NASD suggests be defined by reference to Section 13(d) of the Exchange Act) - is not immediately apparent nor is it clear what it is intended to address. We note that the definition of "research analyst" is so broad that it may arguably capture brokers who prepare and distribute research to their retail clients. As a result, the proposed definition of "Research Analyst Account," as written, might then capture the accounts of such retail clients for which the broker exercises discretion. We assume that this is unintended, as it would result in the trading prohibitions and disclosure obligations applying to customer accounts over which an associated person had discretion. Furthermore, as noted above, the definition as currently written would seem to capture an account of an analyst's roommate. Perhaps the definition could be modified to apply to accounts of particular family members of the research analyst, or to accounts of persons that the research analyst materially supports, or to accounts over which the research analyst has a beneficial interest, discretion or control, other than firm customer accounts.

5. "Investment Banking Department" is defined as "any department or

division, whether or not identified as such, that performs any investment banking service on behalf of a member." It is unclear whether this would include investment banking activities or services performed on an occasional basis, and not as part of the regular course of business. Would persons who, on occasion, participate in investment banking services be part of an "Investment Banking Department"? If so, would this mean that they would be deemed to be a member of the investment banking department permanently or just with respect to the particular matter on which they assisted? Presumably, the breadth of the investment banking department could reach all the way up to the top of a firm. Consider the case of a CEO who, on occasion, participates in pitches for new business. Would the CEO be part of the "Investment Banking Department"? We believe it would be preferable to specify that the persons in the "Investment Banking Department" are those whose regular duties at the firm involve the provision of investment banking services.

Finally, this definition creates a very serious problem for small firms and mid-size firms. Small or mid-size firms may only have a few or even one person who is part of the Investment Banking Department. The restrictions laid out in the proposed rule may prevent these firms from performing investment banking activities altogether. Further, in these firms the president and CEO are more likely to be involved in obtaining investment banking assignments, exacerbating the problems identified in the prior paragraph.

6. The term "Investment Banking Services" includes, without limitation, acting as an underwriter in an offering for the issuer; acting as financial advisor in a merger or acquisition; providing venture capital, equity lines of credit, PIPEs or similar investments; or serving as placement agent for the issuer. First, this definition is ambiguous and difficult to interpret due to the "without limitation" language. Any number of activities could be swept up into the definition. Second, this definition should be modified so that it does not include merchant banking or providing venture capital. Although finding venture capital would normally be considered investment banking services, providing capital in a merchant banking or venture capital context is not normally thought of as an investment banking service.

7. The term "affiliate" is not defined. A definition seems necessary for purposes of the proposed provision requiring disclosure of ownership and material conflicts of interest. This provision requires disclosure if a firm or its affiliates own one percent or more of any class of common equity securities of the subject company five days prior to the issuance of a research report or public appearance. As described in greater detail below, we believe that the Proposed Rules should apply the same methodology, threshold, and timing provisions for calculating beneficial ownership as that used in Sections 13(d) and 13(g) of the Exchange Act.

B. Restrictions on Investment Banking Department Relationship with Research Department

The Proposed Rule addressing the conflicts of interest that may exist between a firm's research and investment banking departments is unduly burdensome and leaves certain aspects regarding the relationship between the two departments unclear.

The Proposed Rules would prohibit a research analyst from being under the supervision or control of any employee of the firm's Investment Banking Department. Would this preclude a member of the Investment Banking Department who serves on the firm's executive committee from participating in any manner in the indirect supervision or control of an employee who works in the research department?

Consider, once again, the example of the CEO who may participate in pitches for investment banking assignments. Does that mean the CEO may not supervise or control the employees of the Research Department? One possible resolution would be to insert the words "direct" or "immediate" before supervision or control. Perhaps a better resolution would be to clarify that participation in a committee of the firm does not constitute supervision or control of a research analyst if the activities of the committee are not controlled by members of the Investment Banking Department alone. Further refinement of the definition of who is a member of the Investment Banking Department would also help.

Subject to a limited exception to allow for reviewing of research reports by investment banking personnel, the Proposed Rules prohibit an employee of the Investment Banking Department from reviewing or approving a research report of the firm before its publication. As presently drafted, it is unclear what is meant by "approval." The Proposed Rules then describe the permissible review without any reference to the approval part. If this provision is intended, as we believe it should, to prohibit any approval of a research report by investment banking personnel, a direct statement to that effect should be made. A clear statement that review is permitted only in the following circumstances would be useful.

We think it is important to note the advisory functions that investment bankers perform for companies. Management of a company may look to the investment bankers for advice on steps to take in the event of a research downgrade. We believe that such an advisory role for investment bankers is appropriate as it provides a service that is of great value to an issuer. It is unclear whether the Proposed Rules would prohibit investment bankers from performing this function.

Under certain conditions, the Proposed Rules permit personnel of the Investment Banking Department to review a Research Report for factual accuracy or for any potential conflict of interest. It is unclear what is intended here with respect to the conflict of interest. Nor is guidance provided if a conflict is identified. What, if any, action should be taken in the event that a potential conflict of interest is identified?

The conditions that would allow for investment banking personnel to review a research report require that any written communication between investment banking and research department personnel concerning a research report must be made either through an authorized legal or compliance official of the firm or in a transmission copied to such an official, and any oral communication between investment banking and research department personnel concerning such a research report must be documented and made either through an authorized legal or compliance official acting as intermediary or in a conversation conducted in the presence of such an official. This is just one area where the Proposed Rules would require legal or compliance personnel to act as intermediaries. We suggest that this intermediary role is perhaps better performed by supervisors in the Investment Banking Department and the Research Analyst Department who understand the financial and market issues involved. Legal or compliance officers do not generally have the expertise or level of understanding necessary to make an informed judgment as to the purpose or intent of a comment suggested by an investment banker. Further, a legal or compliance officer likely also lacks the expertise to determine whether a change to a proposed rating or price target is appropriate.

Requiring that legal or compliance personnel authorize any change to a research report based on comments from the subject company, as described below, raises the same concerns.

C. Restrictions on Review of a Research Report by the Subject Company

The Proposed Rules also restrict communications between a firm and a subject company. Subject to certain exceptions, a firm may not submit a research report to the subject company before its publication. Specifically, a firm may submit sections of a research report to the subject company before its publication only to review for factual accuracy in those sections, provided that those sections do not contain the research summary, the research rating or the price target. In addition, a complete draft of the research report must be provided to the legal or compliance department before sections of the report are given to the subject company. If, after submitting the sections of the research report to the subject company, the research department intends to change the proposed rating or price target, it must first provide written justification to, and receive written authorization from, the legal or compliance department for the change. The firm must retain copies of any draft and the final version of such a research report for three years following its publication. Finally, the Proposed Rules permit a firm to notify a subject company that the firm intends to change its rating of the subject company's securities, provided that the notification occurs on the business day before the firms announces the rating change, after the close of trading in the principal market of the subject company's securities.

Again, the role assigned to legal and compliance personnel to review the justification for a proposed change to the research report and to provide approval is inappropriate. Legal and compliance staff are not necessarily equipped to make such judgments. A supervisor within the Research Department should sign off on whether the proposed change is legitimate. A record should be provided to the legal and compliance department if that gatekeeper function is thought to be necessary.

Finally, as noted above, the Proposed Rules would prohibit a research analyst's firm from notifying a subject company of a rating change until after the close of trading in the principal market of the subject security one day before the announcement of a change. It is unclear from the language of the Proposed Rules whether "the close of trading in the principal market" refers to the primary trading session or the extended trading session.

D. Prohibition on Research Analyst Compensation

We agree with the NASD and NYSE's proposals to prohibit firms from compensating a research analyst based on a specific banking services transaction insofar as the Proposed Rules are prospective. Such a prohibition should help to limit conflicts of interest. Our concern here is with contracts that currently provide a research analyst with rights to participate in the revenue from specific existing investment banking transactions. It is inappropriate to effectively void these contractual rights. We believe that such contracts should be "grandfathered" from the these prohibitions for a reasonable period of time, provided that a research analyst disclose the fact that he is receiving such compensation in public appearances and in research reports.

E. Prohibition of Promise of Favorable Research

The Proposed Rules would prohibit a firm from directly or indirectly offering favorable research, a specific rating or a specific target, to a company as consideration or inducement for the receipt of business or compensation. We agree with the Securities Industry Association's ("SIA") position that this prohibition should be extended to apply to affiliates of issuers as well. We also agree that the phrase "favorable research" may be overly restrictive because an issuer has a legitimate interest in learning if a potential underwriter will look kindly on the company and its stock. We note that responsible investment banking firms, including the research departments of such firms, do not take on an IPO for a company unless they have a favorable view of that company.

F. Quiet Periods

The Proposed Rules would prohibit the issuance of research reports by a firm that participates as an underwriting "manager" or "co-manager" during two newly imposed "quiet periods" - 40 calendar days following the date of an initial public offering ("IPO"), and 10 calendar days following the commencement of sales on behalf of selling security holder in a secondary offering.

The proposed "quiet periods" are very different from those recognized in various Securities Act provisions and related SEC rules. Specifically, this provision would effectively prohibit managers and co-managers from distributing post-IPO research reports preceded or accompanied by the final prospectus that otherwise constitutes permissible "free writing" under Sections 2(a)(10)(a) and 5(b)(1) of the Securities Act of 1933 (Securities Act) during the post-effective "quiet period." Under Rule 174(d) of the Securities Act, the post-effective "quiet period"3 with respect to an IPO security that is listed on a registered national securities exchange or quoted through Nasdaq is the 25 days (compared to 40 days under the Proposed Rules) following the effective date of the IPO. Moreover, the Proposed Rules limit the utility of the Rule 139 safe harbor for firms acting as a manager or co-manager that wish to disseminate research reports under the safe harbor. Although they do not define "manager" or "co-manager," the Proposed Rules suggest that other distribution participants would not be subject to the proposed "quiet periods."

We believe that the proposed quiet periods are unnecessary and excessive given the restrictions imposed by the applicable Securities Act provisions and related SEC rules. Such prohibitions would preclude analysts most familiar with a company from writing about the company. The ten day blackout following completion of a secondary offering is particularly oppressive because the completion of a secondary offering may often be viewed as a significant event for the subject company. Moreover, an issuer may not even have any direct economic benefit in a secondary offering where the offering is orchestrated by selling shareholders.

G. Restrictions on Personal Trading by Research Analysts

The Proposed Rules would restrict personal trading by research analysts and "research analyst accounts" (which includes the accounts of household members of a research analyst). The proposed restrictions would prohibit a "research analyst account" from:

A firm may allow an exception to the personal trading restrictions, but if it does so, the firm's legal or compliance department must pre-approve the report on the security in question and consider whether the research analyst knew or had a reason to know of the events leading to the new research report or change in the rating or price target.

Preliminarily, we believe that the proposed restrictions on personal trading seem to be premised on the belief that "Chinese Walls" or information barriers between a research department and investment banking department are no longer sufficient to prevent questionable or illegal trading. How are existing information barriers meant to mesh with the proposed trading restrictions, if at all?

Although we agree with the goal of trying to limit a research analyst's financial interest in a company he covers, we believe some of these provisions are excessively restrictive. First, we note that NYSE proposed rule 472(e) regarding research analyst's trading restrictions seems to restrict trading in all securities covered by the firm's research department, not just those securities covered by the research analyst. NASD proposed rule 2711(g) restricts trading only in those securities covered by the research analyst. We believe the NASD's approach is preferable.

Second, the Proposed Rules would seemingly prevent a research analyst from owning any securities he covers. We question whether it is appropriate for the NASD and NYSE to make this decision for broker-dealer firms. We believe it is more appropriate for competition and market forces to guide a firm in making this sort of business decision.

Third, it may be difficult or impossible for a research analyst to know whether he is purchasing securities of an issuer principally engaged in the same types of business that he covers prior to a public offering if the research analyst buys the securities before such issuer has even contemplated a public offering. Further, it is unclear how this prohibition would work with respect to rapid shelf takedowns. Perhaps a time period or safe harbor would better clarify the scope of this provision.

Fourth, it may be difficult or impossible for a research analyst to know whether he is purchasing securities of a company he follows outside the blackout period as required. With the possible exception of an initiation report, most analysts do not know precisely when a research report will be issued far enough in advance to comply with this provision of the Proposed Rule.

The Proposed Rules create certain exceptions to the personal trading restrictions. One such exception would allow a research analyst account to sell all of the securities held in the account that are issued by a company that the research analyst follows, within thirty days after the research analyst begins covering the company. We point out that this is another instance where the NASD and NYSE's proposals differ in substance. In contrast to the NASD's proposal, the NYSE's proposed rule 472(e)(4)(iii) and (iv) does not appear to require that all of the company's securities to be sold from the research analyst account after the research analyst begins covering the company.

Fifth, the NASD and NYSE's Proposed Rules also differ with respect to the "significant news" exception to the trading restrictions applicable to research analysts. The NYSE's proposed exception, Rule 472(e)(4)(ii), speaks in terms of allowing a firm to issue a research report (or to change the rating or price target) on a subject company upon significant news even where a research analyst has traded within the prior thirty days, provided that the report (or change in rating or price target) is cleared in writing through the firm's legal or compliance department. In contrast, the NASD's proposed rule 2711(g)(2)(B) speaks in terms of permitting a research analyst account to buy or sell a subject security within the thirty days prior to issuing a research report (or change in rating or price target) of the subject company's securities due to significant news regarding such company, provided that the research report (or any change in rating or price target) are cleared through the firm's legal or compliance department. Although both rules attempt to achieve the same objective, the NASD's proposed rule, as drafted, seems to be based on the incorrect premise that a research analyst knows when a research report is going to be issued as a result of significant news. Of course, this would not make sense because the obvious chronology of events would be for a research analyst to make a trade in a subject security and then for news to come out on a subject company. At that point, the firm would (perhaps) want to issue a research report, but would be prohibited from doing so absent an exception to the restrictions. We believe the NYSE's proposed language is preferable and much clearer.

Although we prefer the NYSE's proposed language in this exception, we believe that additional guidance is needed to establish exactly what is meant by the term "significant news or a significant event." As noted above, the completion of a secondary offering may often be viewed as a "significant event." Furthermore, the Proposed Rule requires that a firm's legal or compliance department pre-approve the research report and any change in the rating or price target. We believe that this is an inappropriate gate-keeping function for legal and compliance departments. Legal and compliance personnel generally do not have the necessary training or experience to make this determination. A Series 24 supervisory person (or a Series 16 registered person in the case of the NYSE) in the research department should be responsible for pre-approving the research report and any change in the rating or price target.

The Proposed Rules would, under certain conditions, permit a research analyst to trade in a subject security during the blackout period based upon a showing of significant personal financial circumstances of the beneficial owner of the research analyst account. "Significant personal financial circumstances" should either be defined, or the NASD and NYSE should at least offer guidance so that the firms will be in a position to make decisions that are not subject to unfair, after-the-fact criticism.

A firm should be permitted to adopt procedures that would permit selling in a manner that may be deemed inconsistent with a recommendation if, as an example, the analyst has held a stock for a number of years and adopts, what is in essence, a 10b5-1 plan type procedure for the disposition of the shares. This might be fair even though it would not constitute "significant personal financial circumstances" warranting the sale.

Sixth, the Proposed Rules carve out certain securities from the trading prohibitions. Specifically, trading restrictions do not apply to the purchase or sale of securities of a registered diversified investment company, or, subject to certain conditions, any other investment fund over which neither the research analyst nor a member of his household has any investment discretion or control if certain conditions are met. These conditions with respect to the other investment funds require that (1) the research analyst accounts collectively own interests representing no more than 1% of the assets of the fund, (2) the fund invests no more than 20% of its assets in securities of issuers principally engaged in the same types of business as companies that the research analyst follows; and (3) the investment fund does not distribute securities in kind to the research analyst or household member before the issuer's IPO.

The condition restricting research analyst accounts collectively from owning interests representing more than 1% of the assets of the fund is not clear in terms of when such a measurement must be taken. Must the 1% be measured monthly? Quarterly? Any time a research report is to be issued? Keeping track of the 1% threshold on a frequent basis will result in significant costs and compliance issues for firms. These same questions and concerns arise with respect to the 20% threshold. In addition, the 20% required measurement assumes that the fund has disclosed, relatively recently, in what particular securities it is invested. Such holdings information may not be sufficiently "fresh" assuming the Proposed Rules require this measurement to be taken on a more frequent basis than a fund makes it available.

We believe that the second condition's reference to the "same types of business as companies that the research analyst follows" could be construed rather broadly. For example, if an analyst covers technology stocks does that mean that all companies with businesses that may include technology are off limits? What if the analyst's focus is only on semiconductors? It is also unclear whether the prohibition would apply to a technical analyst or economic strategist.

Seventh, the proposed rule prohibits a "Research Analyst Account" from selling or buying in a manner inconsistent with the Research Analyst's recommendation. Prohibiting someone from buying or selling the securities of a Subject Company where the security has been given a "neutral" rating may be too restrictive.

Eighth, the definition of "Research Analyst Accounts" raises several concerns with respect to the proposed trading restrictions. Would a roommate of a Research Analyst be prohibited from buying or selling a security where such action is inconsistent with the research analyst's recommendation? "Research Analyst Accounts" are also prohibited from participating in IPOs of issuers principally engaged in the same types of business as companies that the research analyst follows. As proposed, this restriction would apply to a member of a research analyst's household regardless of whether the research analyst had any control over the household member's account. In addition, by virtue of the definition of "Research Analyst Account," a roommate of a research analyst would also be prohibited from trading in a subject security during the blackout period. All of these scenarios are unduly restrictive assuming the research analyst has no control over or beneficial interest in these accounts.

Finally, it is unclear what the repercussions will be if a research analyst should violate these trading restrictions. Will trades have to be broken?

H. Disclosure Requirements

The Proposed Rules would require numerous disclosures with respect to a firm's and its research analysts' relationship with, and interest in, a subject company and its securities. Specifically,

Preliminarily, we would like to point out that the NYSE's proposed rule 472(k)(2) would require disclosure where the person who prepared the research report received compensation that is based, to some degree, on the firm's overall investment banking revenues. The NASD has not proposed a similar requirement.

1. Disclosure Regarding Ownership and Material Conflicts of Interest

As mentioned briefly above, we are concerned about the one-percent beneficial ownership threshold standard included in this disclosure requirement. The proposal suggests that the "beneficial ownership" standard should be derived from Sections 13(d) and 13(g) of the Exchange Act. The proposal does not, however, elaborate on how the SEC's position with respect to disaggregation in the beneficial ownership context would apply under the Proposed Rules.4 Nor does the proposal apply the same ownership percentage threshold or timing provisions as Sections 13(d) and 13(g).

With respect to the ownership percentage threshold and timing provisions, Rule 13d-1 under the Exchange Act allows, among others, registered broker-dealers, investment advisers, and investment companies that have acquired securities in the ordinary course of business and not with the purpose or effect of changing or influencing the control of the issuer to disclose such ownership on Schedule 13G. Schedule 13G imposes a five-percent threshold and requires filing within 45 days of the end of the calendar year only if the five-percent threshold was triggered on the last day of the calendar year.

In contrast, the Proposed Rules require each firm to determine whether the firm or its affiliates exceed the one-percent threshold five days before either: (1) the distribution of the research report; or (2) any public appearance by a research analyst discussing the subject security. The one-percent threshold is significantly lower than that required by Sections 13(d) or 13(g). Moreover, the timing provisions of the proposed rules are drastically shorter than that required by Schedules 13D and 13G. It will be very difficult for firms to fulfill this nearly "real-time" disclosure obligation. This is especially true for large firms with many affiliates. We strongly suggest that the beneficial ownership disclosure requirement in the Proposed Rules incorporate the methodology, threshold, and timing provisions for currently required by Sections 13(d) and 13(g) of the Exchange Act.

Furthermore, there is also a practical problem with this disclosure requirement. Often a research analyst may not know exactly what he or she will speak about in a public forum. What will happen when a discussion on a television program turns to a security that the research analyst didn't plan on discussing? Will he be prohibited from talking about that security because he is unable to make the necessary disclosure as to whether his firm owned one percent of the security?

The Proposed Rules would also require a firm and a research analyst to disclose any other actual, material conflict of interest that the firm or the research analyst knows about or has reason to know about at the time of the research report or the public appearance. Use of the phrase "any other actual, material conflict of interest" is very vague. Further, is the "has a reason to know" language contained in this disclosure requirement intended to create a duty of inquiry for the analyst? What is it intended to require? Moreover, what if this information is on the other side of the Chinese Wall?

2. Disclosure Regarding Receipt of Compensation

We believe the provision regarding disclosure of trailing and future compensation is particularly problematic. It is not clear how exactly such broad and general disclosure will be helpful to inform the investing public. Further, it will be very expensive to track the receipt and expectation of future compensation. U.S. investment banking firms are often affiliated with many large entities that perform a variety of services, not all of which are necessarily financial in nature. Tracking past and possible forward-looking compensation for all of these entities and for all compensation will be extremely burdensome and expensive. Disclosure of compensation received from a non-public transaction may be a disservice to the client; not to mention, what if the client does not want that information to be disclosed? Even assuming disclosure of non-public information regarding compensation expected to be received within the next three months may be appropriate for large firms, it will be extremely problematic for smaller, more focused, boutique broker-dealers. Such disclosure raises concerns of tipping. Disclosing the receipt of compensation for non-public transactions may signal research analysts or their clients at these boutique firms of recent or upcoming significant events at an investment banking client. Certainly, this is not the result the Proposed Rules are trying to achieve.

Although we believe that the disclosure requirements are too broad in scope and would create unjustifiable burdens on broker-dealers, narrowing the scope of the disclosure requirement with respect to receipt of compensation may not be a superior alternative. We agree with the SIA's suggested approach that, instead of requiring this disclosure provision to be so broad, research reports should contain disclosure of the receipt of investment banking compensation from publicly-announced transactions within the 12 months preceding the issuance of a research report. Furthermore, with respect to the disclosure of compensation received from a non-public transaction at a subject company, it may be more appropriate to simply require a statement indicating whether the firm intends to seek to do business with the subject company in the three months following a research report.

As mentioned above, we are concerned that the Proposed Rules may discourage research analyst participation in public forums, including television. Such a result would not be in the public interest. Specifically, the Proposed Rules would require research analysts to make certain disclosures in public appearances related to financial interests in the subject company, beneficial ownership of one-percent or more of any class of the subject company's common equity securities by the firm or its affiliates, and any other actual, material conflict of interest exists or has reason to know exists. We respectfully suggest that, instead of requiring a research analyst to make these specific disclosures, a research analyst should be required to make a general statement at the beginning of his appearance referring persons to research reports about the subject company that are publicly available from the firm, or, if such reports are not public or are not widely or readily available, to NASD Regulation's web site for a discussion of research analysts' recommendations and potential conflicts of interest. (See

Referring potential investors to the relevant research report would provide specific information as to the research analyst's and his firm's potential conflict of interest, if any, in recommending a subject company. Moreover, referring persons to NASD Regulation's web site would provide more than adequate information to an investor regarding potential conflicts of interest. Both approaches would achieve the stated goal of providing investors with more useful and reliable information when making investment decisions.

Finally, we note that a particular aspect of this disclosure requirement is somewhat ambiguous. The Proposed Rules would require a research analyst to disclose in research reports or public appearances if he knows or has reason to know that the subject company is a client of the firm or its affiliates. It is unclear from the Proposed Rules how a research analyst can be required to disclose something he doesn't know but "has reason to know."

3. Other Disclosures

The Proposed Rules also call for an extensive set of disclosures required in research reports. These disclosures relate to a firm's rating system, pricing history for a security and other possible conflicts of interest. In particular, the Proposed Rules would require that a research report contain disclosures regarding:

We agree with the goal of better informing investors about rating and pricing systems used in research reports. We are concerned, however, that the proposed disclosures will impose significant costs and burdens on broker-dealers, without providing much of a benefit to investors. These disclosures may prove to be of little practical significance to investors given the volume of information and minutiae likely to be included. We note that the inclusion of all these disclosures - i.e., the meaning of ratings, distribution of ratings, price charts, price target methodologies - in each research report may often result in the disclosures being longer than the research report itself. The proposed disclosures may also impede the free flow of information given the associated costs and difficulty in using certain types of technology to transmit all of this information in the format required.

As an alternative to requiring the disclosures to be published in all research reports, we respectfully suggest that a firm be permitted to post the information to its web site and that the research report refer investors to the web site and information on how to access the disclosures.

Certain of the proposed disclosures are ambiguous in terms of what exactly will be required. With respect to price targets, the Proposed Rule would require a firm to disclose the valuation methods used to determine a price target, as well as the risks that may impede the achievement of the price target. Will such disclosure be required in each research report regardless of whether there is a change? We believe the disclosure of price target methodology, as proposed, would be more useful to investors and less burdensome to broker-dealer firms if it were included in research reports when coverage of the subject company is initiated or when the firm is changing the price target.

Finally, the Proposed Rules also would require firms to disclose in research reports and research analysts to disclose at public appearances if the research analyst or a member of his household is an officer, director or advisory board member of a subject company. Once again, we must point out that the definition of "member of a research analyst's household" is problematic. This definition is too broad in the sense that it includes more than immediate family members and does not incorporate the concepts of material support. In the context of this disclosure requirement, it could meant that a firm or research analyst would have to disclose the status of a roommate. Such disclosure could be an invasion privacy of a roommate, or anyone else living in the same residence for that matter.

I. Greater Obligations for Legal and Compliance Personnel

The Proposed Rules would entail significant new responsibilities on a firm's legal or compliance department. As noted above, legal/compliance departments under certain circumstances would be required to: (1) pre-approve research reports; (2) pre-approve analyst trades; and (3) act as an intermediary between the research and investment banking departments. This marks a substantial expansion of the traditional role that legal and compliance personnel perform. Many of these duties are typically performed by the firm's applicable supervisory principals (such as the head of each department or other designated supervisory principal).

J. Difficulty in Implementing a Supervisory System

The NASD states in the Proposing Release that it will announce the effective date of the proposed rule in a Notice to Members ("NTM") to be published no later than 60 days following Commission approval, and the effective date will be no less than 30 days following the NTM publication. Accordingly, firms may have to act quickly to implement a supervisory system to address these rules once approved by the Commission. We suggest that certain aspects of the Proposed Rule be phased-in over 18 months following Commission approval.

IV. Conclusion

Thank you for giving us the opportunity to comment on the NASD and NYSE Proposed Rules regarding research analysts. Notwithstanding our overall support for the stated goals, the Proposed Rules, as they stand now, are likely to impose significant burdens on research and investment banking departments of broker-dealers firms, and impose even greater supervisory responsibilities on such broker-dealers. Costs and burdens to small brokerage firms will be particularly acute. We believe, however, that with the suggested modifications the Proposed Rules could go a long way toward addressing research analyst conflicts of interest and benefiting investors.


Robert C. Mendelson

Steven W. Stone

John V. Ayanian

cc: The Hon. Harvey L. Pitt, Chairman
The Hon. Isaac C. Hunt, Jr., Commissioner
The Hon. Cynthia A. Glassman, Commissioner
Annette L. Nazareth, Director, Division of Market Regulation
Robert L.D. Colby, Deputy Director, Division of Market Regulation
Giovanni P. Prezioso, Director, Office of General Counsel
Christine M. Richardson - Morgan, Lewis & Bockius LLP

1 See Exchange Act Release No. 45526 (Mar. 8, 2002), 67 FR 11526 (Mar. 14, 2002) ("Proposing Release").
2 See Exchange Act Release No. 34-42325; File No. SR-NASD-99-60 (January 10, 2000).
3 A dealer's Section 4(3) exemption from Section 5 is not available during the "quiet period."
4 See Securities Exchange Act Release No. 39538 (Jan. 12, 1998), and 2001 SEC No-Act. LEXIS 458 (Mar. 30, 2001).