1. Introduction
In a typical corporate report, the textual narrative represents the great majority of the disclosure— an average of 80% of an annual report, for instance—versus the remainder that consists of numbers and representations of quantitative data. The clarity of this large component of mandatory disclosure is crucial to understanding and to interpreting the information contained in the report. The U.S. Securities and Exchange Commission (SEC) has been very forthright about the overly complex corporate reports. Christopher Cox, Chairman of the SEC, suggested using direct measures of narrative clarity to enforce plain English communication. SEC seems to believe that “the jargon of lawyers has taken over” and the trend towards hard-to-read disclosures is due to the fact that “the main purpose of the drafting exercise has shifted from informing investors to insuring the issuer and the underwriter against potential claims” (SEC, 2007). In this paper, we examine whether managers use of complex disclosures goes beyond the presence of “legalese”, but also whether they use complex disclosure to hide information from investors.
The seminal work of Li (2008) explored the relationship between the readability of annual reports and financial performance. Borrowing the Fog Index from computation linguistics, where a higher reading on the Fog Index indicates disclosures that are more difficult to understand, Li finds a negative relationship between Fog and the level of earnings. It is unclear, however, whether this result is due to managers providing complex disclosures to obfuscate bad performance or that bad news is simply harder to be communicated (Bloomfield, 2008). To further explore these alternative explanations, obfuscation or ontology, and to better understand managers’ use of complex disclosures, we look at instances in which firms are more likely to have managed earnings upwards to meet or beat an earnings target (Burgstahler and Dichev, 1997). In these cases, although firms are releasing good news about meeting a benchmark, they have incentives to hide the tools used to achieve it, as suggested by Warren Buffett. In other words, when reported performance differs from underlying fundamentals, we expect managers to try to make it harder for investors to identify such earnings management behavior and the underlying performance. Our results suggest that the readability level of financial disclosures goes beyond the one derived from the ontological explanation of good vs. bad news being disclosed. Instead, we find that managers strategically use corporate disclosure to mislead or to influence investors’ understanding of firm's value.