A Plain English Measure of Financial Reporting Readability
Posted: 6 Feb 2015 Last revised: 13 Apr 2017
There are 2 versions of this paper
A Plain English Measure of Financial Reporting Readability
A Plain English Measure of Financial Reporting Readability
Date Written: May 31, 2016
Abstract
Over the past decade, a burgeoning literature has emerged using textual analysis to examine the impact of financial reporting readability on capital markets. Despite the rapid growth of this literature, clear guidance on measuring readability in financial texts is rather limited. In this study, we introduce a new measure of readability, the Bog Index, which is designed to capture the plain English attributes of disclosure (e.g., active voice, fewer hidden verbs). We then validate this new measure based on a combination of regulatory guidance, experiments, and archival tests using a regulatory intervention related to the readability of prospectus filings. After validating this measure, we next examine the extent to which different readability measures are associated with future stock market volatility and equity analysts’ earnings forecast properties. We find that the Bog Index has nearly a twenty-five percent greater association with future stock return volatility than any other measure of readability. In contrast, only quantity-based measures of disclosure (e.g., total words, file size) are associated with analyst forecast accuracy. Combined, this evidence suggests that financial statement users may be affected by different attributes of readability. Finally, we use both archival and experimental evidence to demonstrate the importance of understanding the underlying drivers of quantity-based measures of readability. In particular, we caution researchers that a vast amount of the variation in the file size measure over time is driven by the inclusion of content unrelated to the underlying text in the 10-K (e.g., HTML, XML, PDFs).
Keywords: Voluntary disclosure; Plain English Reporting; Textual Analysis; Regulation
JEL Classification: G38; M41; L51
Suggested Citation: Suggested Citation