Financial reports can be opaque, complex, and difficult to understand. As far back as 1998, this was the premise behind the SEC’s Plain English Rule: an unsuccessful attempt to encourage firms to write more readable financial reports. In a new paper, we show that the turgid nature of financial reports could be an unintentional response to litigation risk. Indeed, firms’ concerns about satisfying disclosure requirements lead to more voluminous, and more complex, disclosures. We ensure that our results are well identified and take steps to mitigate endogeneity concerns.
What we investigate
We start with the observation that decreasing readability harms firms. It is associated, for example, with less access to credit, greater analyst forecast dispersion, and greater stock volatility. This raises the question of why firms might still choose to release opaque financial reports.
We hypothesize that opacity could be an inadvertent side-effect of litigation risk. We focus on securities class actions (SCAs) under SEC Rule 10b-5. SCAs typically arise where (inter alia) the firm makes a false or misleading statement (or omission), which inflates share prices and causes shareholders to incur losses by paying too much for those shares. While our analysis focuses on the U.S., most countries have similar rules.
SCAs typically harm firms and managers. For example, SCAs and SEC enforcement actions often harm firms’ relationships with customers and suppliers and reduce their access to debt. Further, SCAs can lead to pay cuts for managers and, in extreme cases, dismissals. As result, SCAs can deter managers from engaging in behavior that can lead to litigation.
We expect firms to take steps to reduce litigation risk by ensuring that disclosures are detailed and nuanced enough to make them not misleading. Practically, this would result in longer financial reports. Further, to ensure that concepts are explained fully enough to reduce litigation risk, firms would use more words per sentence and more complex words. They might also use more general and potentially vague language, strengthening claims that the disclosures were not technically inaccurate. Finally, firms could use more legalistic terms.
What we do
We analyze whether litigation experience influences how directors and officers write financial reports. We use a sample of over 50,000 firm-year observations from 1993 to 2013. We start with the full set of firms listed in the U.S. (for which we can obtain data from CRSP/Compustat). We then identify whether these firms have been hit with a securities class action in each year (as reported by the Stanford Securities Class action database). We also use metrics designed to determine a financial report’s readability. We obtain these data from the SEC Analytics database, which is a publicly available subscription source. We focus on firms’ 10-K reports, which allow us to compare firms with one another,
We assess several readability metrics, which we describe below. We strengthen our result by controlling for year and industry fixed effects, and by using multiple measures.
- Fog index aims to measure how many years of education it takes to read a document. A higher Fog index connotes less readability.
- Bog index captures similar ideas to the Fog index. It penalizes documents: (1) for using complex words, and (2) for using sentences that are more complex or are poorly constructed (i.e., use the passive voice). However, it “un-penalizes” documents where those complex words might merely represent names or places, or where they are conversational.
- PC index is the first principle component of six readability measures, which are analogous to the Fog index. All six measures capture similar concepts in different ways. The first principle component enables us to capture the common cause of less readability.
We also collect data on other linguistic characteristics. We capture disclosure volume by obtaining the number of words in the financial report and the file size of the financial report. We assess reports’ linguistic complexity by finding the number of words per sentence, and the ratio of words that are complex. This helps to identify whether firms add nuance to disclosures. We address whether firms try to avoid declarative statements and make reports vaguer by obtaining the proportion of “uncertain” words in the financial report. Finally, we obtain the number of litigious or legalistic words in the report. This captures firms’ increasing use of complex legalise – which can reduce readability – potentially to pre-empt lawsuits.
The empirical framework broadly involves regressing readability measures onto variables representing the firm’s litigation experience, controlling for myriad corporate factors that might influence readability metrics (i.e., firm size, complexity, analyst coverage, etc). We also ensure identification, and mitigate endogeneity concerns, by exploring how directors and officers respond to litigation shocks at other firms.
What we find
First, we explore whether financial report readability declines after a securities class action. We use a regression model in which we control for myriad factors that can influence readability. We find that readability declines after a class action. This is the case whether we look just at whether a class action is launched or whether a class action is resolved. This effect persists for several years after the class action, implying a clear shift in how executives structure disclosures.
We further explore the dimensions along which readability decreases. We explore matters such as whether disclosures become more voluminous, complex, or legalistic. We find that, after a class action, firms’ reports become longer (i.e., have more words) and larger (i.e., greater in file size). They are also, generally, more likely to use more words per sentence, more complex words, and more uncertain words. This implies that firms attempt to both disclose more information and to add nuance to those disclosures to improve disclosure accuracy. We also find that firms are more likely to use more legalistic words, implying efforts to deter lawsuits.
The results are robust to endogeneity concerns. We ensure identification in several ways. First, we explore whether CEOs’ experience with litigation, either as a non-CEO executive at the present firm or at other firms, influences their disclosure behavior at the focal firm. We find that it does. Importantly, when just isolating experience at other firms, we find supportive evidence, highlighting that such litigation experience shapes readability. Second, we analyze the effect of directors’ litigation experience at “interlocked” firms (i.e., other firms on which they sit as directors). The results highlight that experience at other firms also shapes disclosure behavior at the focal firm. This highlights that, after CEOs or directors become more cognizant of litigation risks, they shape disclosures even if that experience was outside the focal firm. We also use the introduction of the SEC’s 1998 Plain English disclosure rules as an instrument for readability in a two-stage regression and find that reducing readability reduces litigation likelihood.
We bolster these results by exploring how firms shape readability around various litigation flashpoints. These include the passage of the Sarbanes-Oxley Act, seasoned equity offerings, and bond downgrades. We find that readability decreases around all such points. This again suggests that firms make disclosures more complex and voluminous at such points, potentially in anticipation of increased litigation risk.
The results overall highlight that litigation risk affects how firms write financial reports. We highlight that directors and officers are cognizant of litigation risk. We show that readability declines around litigation. We demonstrate that this manifests in longer and more complex financial reports, indicating that firms attempt to disclose more information and add additional nuance to disclosure. The results indicate that litigation risk might be associated with decreasing readability, but this could be an unintended consequence of firms’ attempting to enhance their disclosures.
This post comes to us from professors Mark Humphery-Jenner at UNSW Business School, Yun Liu at Claremont Colleges – Keck Graduate Institute, Vikram K. Nanda at the University of Texas at Dallas, Sabatino Silveri at the University of Memphis, and Minxing Sun at Clemson University. It is based on their recent article, “Of Fogs and Bogs: Does Litigation Risk Make Financial Reports Less Readable?,” available here.