The relationship between analysts and investors can be tricky. An investor may grow fond of his or her analyst while receiving successful buy recommendations on stocks. However, at first sign of trouble, analyst research may become scarce, and the investor is left holding the bag.
The popular press offers infamous anecdotes about analysts who failed to downgrade their ratings in a timely manner on distressed or fraudulent firms such as Enron. Academic literature provides ample evidence of analysts’ reluctance to issue unfavorable reports. When analysts receive negative information on a firm, they prefer to discontinue coverage rather than issue a sell recommendation because they fear jeopardizing the access to the firm’s management. McNichols and O’Brien (1997) argue that analysts drop coverage after receiving negative information, which is not reflected in their last outstanding forecast.1
In 2003, regulators intended to eliminate the practice of dropping coverage rather than lowering the rating. The goal was to provide investors with the benefit of the guidance from the analysts that recommended the stocks in the first place. With this objective in mind, the NASD added paragraph (f)(5) to Rule 2711, while the NYSE added an almost identical paragraph (f)(6) to Rule 472. In August 2003, the SEC approved the added provisions requiring analysts to issue a final report when they terminate research coverage of a given firm. Eight years after the introduction of these provisions, the time has come to assess their effectiveness.
According to the final text of NASD Rule 2711 (f)(5): “If a member intends to terminate its research coverage of a subject company, notice of this termination must be made. The member must make available a final research report on the subject company using the means of dissemination equivalent to those it ordinarily uses to provide the customer with its research reports on the subject company. The report must be comparable in scope and detail to prior research reports and must include a final recommendation or rating, unless it is impracticable for the member to produce a comparable report (e.g., if the research analyst covering the subject company or sector has left the member or if the member terminates coverage of the industry or sector). If it is impracticable to produce a final recommendation or rating, the final research report must disclose the member’s rationale for the decision to terminate coverage.”
The provision requests that a termination notice must be “comparable in scope and detail to prior research reports” and must include a final rating. It is interesting to note that in the original text proposed by the NASD, the termination notice was required to be comparable to the report that initiated coverage. Now, in its final text, the provision requires the notice that terminates coverage of a firm to be as comprehensive as other reports previously issued on that firm by the analyst.
In addition, the provision makes an exception when “impracticable” circumstances prevent an analyst from producing a termination notice. When it is impossible, the analyst’s employer must disclose only the “rationale” for terminating coverage. However, when a final report is produced, the provision does not require an explicit rationale. Therefore, there might be a margin for interpretation. Not disclosing the rationale for terminating coverage may be perfectly adequate when an analyst issues a final report on a firm. Alternatively, not disclosing the rationale for terminating coverage may be inadequate considering the detail required in the final report.
To assess the effectiveness of the provision, I manually collected data from 7,038 termination notices issued between 2003 and 2008. The data source was Thomson Research, which stores the PDF files of the original analyst reports.2 The reports fall into four categories of rationale for the termination of coverage: analyst departure, stock delisting, redirection of research efforts, and no disclosed reason. Analyst departure and stock delisting are circumstances when it is impossible for final reports to be issued, while the redirection of research efforts and when the firm makes no disclosure are circumstances that require an analyst to produce a final report.
About 44 percent of termination notices disclose analyst departure as a rationale for terminating coverage, while 25 percent report stock delisting. These notices are expected to be succinct. The average length is less than three pages. However, one may wonder whether investors benefit from being notified that analyst coverage of a stock is terminated due to the known fact that the stock no longer trades publicly. In a cost-benefit analysis, termination notices due to stock delisting seem uninformative enough to be unnecessary. In the remaining cases, it would be practicable to provide investors with a final report. Specifically, about 29 percent of terminations (2,049) provide a rationale that I classify as redirection of research efforts, while 2 percent disclose no rationale.
An analyst’s decision to terminate coverage of a firm while redirecting research efforts elsewhere is likely to be driven by firm-specific reasons. However, analysts are reluctant to provide specific reasons. They typically use expressions such as, “the stock no longer fits the coverage universe.” The wording suggests that the termination is not motivated by something that the firm issuing that stock has done, rather by the need to focus research activity on other stocks. In other words, “It’s not you, it’s me.” In only 248 notices, the analysts specify why, in their view, the stock is unfit. In these cases, they explicitly mention poor operating performance or little institutional interest as main rationales for the decision to terminate coverage.
The termination notices that disclose a redirection of research efforts do not appear to be comparable to prior research reports, either in scope or detail. The scope is comparable to other termination notices, rather than to prior reports. The average length (2.7 pages) is much shorter than initiating or continuing reports. Not much detail is provided about the final rating. Notices generally mention and revise the next-to-last rating to a last rating of “Not Rated.” A distribution of ratings issued in the last 12 months rarely is included in the notices.
Overall, termination notices are not an effective reflection of the spirit of the rules, because investors are left without final reports that fully explain analysts’ decision to terminate coverage.
1. McNichols, Maureen F., and Patricia O’Brien, 1997. Self-selection and analyst coverage. Journal of Accounting Research 35, 167−199.
2. Thompson Research (formerly known as Investext) includes investment research, trade association research, and morning meeting notes. To isolate the termination notices, I search for the following keywords in the title of the report: terminat* OR drop* OR discontinu* AND coverage. The search is restricted to company-type reports on U.S. firms.