Are managers strategic in reporting non-earnings news? Evidence on timing and news bundling

Examples of reportable items include entry into a material agreement or its termination, bankruptcy or receivership, and material impairments, to name a few. See Sect. 2 for more detail.

Firms could try to strategically bundle mandatory items. However, the reporting requirements for mandatory items afford less flexibility (see Sect. 2). Hence we focus on the bundling of mandatory and voluntary items.

One could argue that reporting ATH is consistent with the SEC’s effort to provide investors equal access to the information. While plausible, note that the SEC’s effort applies to all information, positive or negative. Our finding that negative news is more likely to be reported ATH is therefore more consistent with strategic reporting (i.e., an attempt to exploit investor inattention).

In Re Comverse Tech., Inc. and In Re Browning-Ferris Indus. Also see the following law review articles: Beale (2009), Abril and Olazabal (2010), and Steinberg and Goldman (1987).

DeHaan et al. (2015) show that investor attention is indeed low ATH, on Fridays, and on busy days. Specifically, they provide evidence that, when earnings are released ATH or on busy reporting days, there are fewer news articles, EDGAR downloads, Google searches and analysts update their forecasts more slowly.

Although Gennotte and Trueman (1996) use earnings announcements in the discussion of their model, the model is generalizable to any mandatory public announcements (i.e., to non-earnings news as well).

Managers aim to maximize post-announcement share price for reasons related to compensation, job security, and litigation. Specifically, Jayaraman and Milbourn (2012) find that, as stock liquidity increases, the proportion of equity-based (cash) compensation in total compensation increases (decreases). Hence it is likely that managers of public firms receive a greater portion of their compensation in equity-based instruments, leading to greater compensation sensitivity to stock price. Defond and Hung (2004) and others document a link between CEO turnover and poor stock returns. Also, the likelihood of litigation is affected by stock reaction to negative events (see for example, Kim and Skinner 2012).

One could argue that managers have no reason to disclose negative voluntary news. However, voluntary news in the context of 8-K filings are material events that are eventually reported in the subsequent periodic 10-Q/K. Given that the news would eventually be released, the firm bears no real cost in disclosing negative voluntary news early. This is similar to management forecasts, which are used as the main proxy for voluntary disclosure. Once managers decide to issue forecasts, they typically do so regularly whether the news is positive or negative. In addition, the literature indicates that less timely disclosure increases firms’ litigation consequences (e.g., Skinner 1994; Kasznik and Lev 1995; Skinner 1997, Baginski et al. 2002). For example, Skinner (1994) argues that U.S. securities laws provide incentives for managers to disclose negative news voluntarily. This is because announcements of negative earnings surprises increase the likelihood of shareholder litigation. Early disclosures both reduce the plaintiffs’ ability to claim that managers failed to release material information promptly and limit the size of the plaintiff class by reducing the period of nondisclosure.

The SEC issued Release No. 34-49424, additional Form 8-K disclosure requirements and acceleration of filing date, in March 2004, and it became effective on August 23, 2004. The rule significantly increased the number of events to be reported in the 8-K report and shortened the period required to disclose these events to no more than four business days after the event.

Both the Exchange Act and Securities Act require a company to register its securities with the SEC if those are held by 500 (and in some situations, 300) or more persons and the company’s total assets exceed $10 million. Our definition of public firms is similar to that of Givoly et al. (2010).

Regular stock market hours are 9:30 a.m.–4 p.m. Eastern Time. About 88 % of the ATH cases were disclosed between 4 p.m. and 6 p.m. About 10 % of the observations were filed before market opens (BMO), and we treat them as filed during trading hours. Footnote 17 describes sensitivity analysis related to BMO disclosure.

Throughout our analysis, we assume that the news is disseminated to the market via 8-K report first. We expand on this issue in Sect. 4.5.1.

As a sensitivity analysis, we examine the mean market reaction by deciles formed on the basis of the news score. Untabulated results show that the mean market reaction generally increases with the news deciles, with the lowest (highest) mean return in the bottom (top) news deciles.

Kothari et al. (2009) show that litigation risk, distress risk, and information asymmetry moderate the decision of firms to delay the disclosure of negative news. Untabulated results indicate that these variables are not related to the decision to disclose when investor attention is low, and controlling for these variables does not affect the inferences.

The specification is based on negative news indicator in the equity return regressions because we also control for the absolute value of CAR. If we instead use equity returns as the proxy for news and omit absolute CAR from the regression, we find that the coefficient on the news variable is negative and significant, consistent with strategic reporting of negative news.

We do not control for governance in the reported results because data on the GIM and entrenchment variables is limited, resulting in a significantly smaller sample.

We also examine whether filing before market opens (BMO) is associated with strategic reporting of negative news. We find that there is no association between the type of news and BMO reporting; the likelihood of negative news reported BMO is similar to that of positive news.

We examine whether the results are sensitive to the measurement of the news variable. Specifically, we re-estimate the regressions defining as negative news all forms containing a priori negative news (see Sect. 4.1 for detail). In an additional analysis, we define negative news as one if the news score is in the lowest quartile. In both analyses, we observe that the likelihood of strategic reporting of negative news is higher for public firms than nonpublic ones.

We construct this cumulative measure, rather than an annual measure, because the filing of 8-K reports varies considerably over time, and the average number of 8-K reports filed annually is fairly small, with an even lower number of forms with negative news.

Since the regressions are estimated at the firm-year level (as opposed to firm-form level in previous regressions), we do not control for item fixed effect or the economic impact of the form.

One potential concern with our measure of strategic reporting is that it may be noisy in the early years of the data because of the smaller number of observations. To address this issue we measure strategic reporting using all of the data. While this measure is less noisy, the analysis involves look-ahead bias. Nevertheless, the results are similar to those reported—the likelihood of news bundling increases with strategic reporting through disclosure timing.

To ensure that the analysis is not contaminated by the presence of additional items reported in the form, we include only single-item forms. In a sensitivity analysis, we remove this restriction and examine all forms including multi-item filings. None of the inferences change.

For information on EDGAR FTP access and use of EDGAR Index files, see https://​www.​sec.​gov/​edgar/​searchedgar/​ftpusers.​htm.

Complete lists of positive and negative words can be downloaded from Bill McDonald’s website: http://​www3.​nd.​edu/​~mcdonald/​Word_​Lists.​html.

For a comprehensive review of papers using the LM measure and related literature, see Loughran and McDonald (2015).