Analyst Rating

An analyst rating is the summary recommendation a research analyst assigns to a security, conventionally expressed on a scale such as buy, hold and sell, or outperform, neutral and underperform. No rule defines the term, and no rule prescribes the scale. What the rules regulate is the document the rating appears in and the conduct behind it. FINRA Rule 2241 governs equity research and speaks of "a member that employs a rating system," which is the rule's acknowledgment that a firm need not have one. The federal layer is Regulation AC, 17 CFR 242.500 through 242.505, which defines a "research report" as "a written communication (including an electronic communication) that includes an analysis of a security or an issuer and provides information reasonably sufficient upon which to base an investment decision," and a "research analyst" as "any natural person who is primarily responsible for the preparation of the content of a research report." The rating is the part of that document a reader is most likely to act on and the part carrying the least information.

The plain-meaning constraint is the rule most worth knowing. FINRA Rule 2241(c)(2) provides that "[a] member that employs a rating system must clearly define in each research report the meaning of each rating in the system, including the time horizon and any benchmarks on which a rating is based. The definition of each rating must be consistent with its plain meaning." Two things follow. A rating has to come with its own definition, so the scale is never left to inference; and a firm cannot redefine a familiar word into something else, which is what closes the gap a reader would otherwise fall into when a firm's "hold" turns out to describe a stock it expects to fall.

The two distribution disclosures are the only calibration tool a reader gets, and they are in every report. Rule 2241(c)(2)(A) requires a member to include in each research report that includes a rating "the percentage of all securities rated by the member to which the member would assign a 'buy,' 'hold' or 'sell' rating." Rule 2241(c)(2)(B) then requires disclosure of "the percentage of subject companies within each of the 'buy,' 'hold' and 'sell' categories for which the member has provided investment banking services within the previous 12 months." Under (c)(2)(C) both must be current as of the end of the most recent calendar quarter, or the second most recent where the report is published less than 15 calendar days after the quarter end. The first figure tells a reader how selective the firm's top rating actually is. The second tells them how the firm's ratings distribute across companies it also does banking work for. Neither is a verdict, and both are checkable in the small print of the report itself.

The conflicts the report has to confess. Rule 2241(c)(4) requires disclosure, at the time of publication or distribution, of a list of specific facts: whether the analyst or a member of the analyst's household has a financial interest in the subject company's debt or equity securities and the nature of it; whether the analyst "has received compensation based upon (among other factors) the member's investment banking revenues"; whether the firm or its affiliates managed or co-managed a public offering for the company in the past 12 months, received investment banking compensation from it in the past 12 months, or "expects to receive or intends to seek compensation for investment banking services from the subject company in the next three months"; whether it received non-investment-banking compensation from the company in the previous 12 months; whether the company is or has been a client and the type of services provided; whether the firm or its affiliates "beneficially own 1% or more of any class of common equity securities of the subject company"; whether the firm was making a market in the securities; whether the analyst received any compensation from the company in the previous 12 months; and any other material conflict known to the analyst or to an associated person able to influence the report's content. Rule 2241(c)(6) requires these to appear on the report's front page, or the front page to point to the page where they are.

Two conduct rules sit behind the rating itself. A member's written policies must, under Rule 2241(b)(2)(K), "prohibit explicit or implicit promises of favorable research, a particular research rating or recommendation or specific research content as inducement for the receipt of business or compensation," and under (b)(2)(N) must "prohibit prepublication review of a research report by a subject company for purposes other than verification of facts." The second is narrower than it looks and more useful for it: a company may be shown a draft to check facts, and may not be shown it to influence the conclusion.

The analyst's own signature, under federal rather than FINRA rules. Regulation AC, 17 CFR 242.501(a), requires a broker, dealer or covered person distributing a research report to a US person to include "a clear and prominent certification by the research analyst" attesting "that all of the views expressed in the research report accurately reflect the research analyst's personal views about any and all of the subject securities or issuers," together with either a statement that no part of the analyst's compensation "was, is, or will be, directly or indirectly, related to the specific recommendations or views expressed," or, if part of it was, a statement identifying "the source, amount, and purpose of such compensation" and disclosing that it "could influence the recommendations or views expressed." Rule 242.502 extends a version of that certification to public appearances, requiring a quarterly record. Its enforcement mechanism is the interesting part: if the analyst does not provide the statement, the firm must promptly notify its examining authority in writing, and for 120 days afterward must disclose in every research report by that analyst that the certification was not provided. Rule 242.505 excludes bona fide news publishers from Regulation AC entirely.

One requirement is shared with the number beside it. Rule 2241(c)(1)(B) requires a member's policies to be reasonably designed to ensure that any recommendation, rating or price target has a reasonable basis and comes with an explanation of the valuation method and a fair presentation of the risks to achieving it. That obligation is joint, so it is not a target-specific rule, and the price target page carries it together with the compelled price chart that goes with a target or rating held for a year.

What a rating is not. It is not a valuation, which is what fundamental analysis produces, and it is not a claim the firm has beaten the market, which is the question the efficient market hypothesis is about. Nor is it independent of the sequence in which a reader meets it: a categorical label seen before the analysis is exactly the kind of first number that anchoring bias describes.

An analyst covers a company, builds a view, and expresses it as a rating on the firm's published scale together with the reasoning. The report defines each rating on that scale, states the time horizon, discloses the firm's distribution of ratings and its banking relationships, and carries the analyst's certification. The rating is then maintained, changed or withdrawn as the analyst's view changes, and a change is itself news.

A hypothetical, working the two distribution disclosures the way a reader should. A firm rates 200 companies. It assigns buy to 120, hold to 70, and sell to 10. The first required disclosure is therefore 120 divided by 200, or 60 percent buy; 70 divided by 200, or 35 percent hold; and 10 divided by 200, or 5 percent sell. The second disclosure looks inside each category. Suppose the firm provided investment banking services in the past 12 months to 48 of the 120 buy-rated companies, to 7 of the 70 hold-rated, and to none of the 10 sell-rated. That is 48 divided by 120, or 40 percent of the buy category; 7 divided by 70, or 10 percent of the hold category; and 0 percent of the sell category. Neither set of numbers proves anything about any single report. Together they tell a reader that this firm's buy rating is its default and that banking clients are concentrated in it, which is context the rating alone does not carry.

Pros

• The rating comes with its own definition in every report, including the time horizon, so the scale is never left to guesswork.
• The plain-meaning requirement stops a firm defining a familiar label into its opposite.
• The distribution disclosures let a reader see how selective the top rating is before treating it as selective.
• The banking-relationship disclosure is required per category, which is a more revealing cut than a single firm-wide figure.
• The analyst has to certify personally that the views are their own and to disclose compensation tied to specific recommendations.

Cons

• A rating compresses an entire analysis into one word, and the word is the part most likely to be quoted without the analysis.
• Scales are not standardized between firms, so two ratings with the same label can mean different things and two with different labels can mean the same thing.
• Analyst compensation may be based partly on the firm's investment banking revenues, which the rule addresses by disclosure rather than prohibition.
• The disclosures are usually in small print at the back or behind a front-page cross-reference.
• Ratings are relative to a stated horizon and benchmark, so a rating is not a statement about what will happen to the price.