Measuring Readability in Financial Disclosures

The following post comes to us from Tim Loughran and Bill McDonald, both of the Department of Finance at the University of Notre Dame.

The Fog Index has become a popular measure of financial disclosure readability in recent accounting and finance research. The SEC has even contemplated the use of the Fog Index to help identify poorly written financial documents. However, the measure has migrated to financial applications without its efficacy in the context of business disclosures having been determined.

In our forthcoming Journal of Finance paper, Measuring Readability in Financial Disclosures, we argue that traditional readability measures like the Fog Index are poorly specified in the realm of business writing. The Fog Index is based on two components: sentence length and word complexity. Although sentence length is a reasonable readability measure, it is difficult to accurately measure in financial documents. More importantly, we show that the count of multisyllabic words in 10-K filings is dominated by common business words that should be easily understood. Frequently used “complex” words like company, operations, and management are not going to confuse consumers of SEC filings. Additionally, the correlation of complex words with alternative measures of readability contradicts its traditional interpretation.

We find that the 10-K file size provides a better proxy for readability than traditional measures. As a measure of readability in financial disclosures, where readability is the ability to assimilate valuation-relevant information, we recommend that researchers use the file size of the “complete submission text file” available on the SEC’s EDGAR website. The measure does not require any parsing of the document and is readily replicated. Note, however, that we would not expect this measure to translate well to other types of documents like news articles and press releases.

In regression analysis, we report that, after controlling for other variables, larger 10-K file sizes have significantly higher post-filing date abnormal return volatility, higher absolute SUE, and higher analyst dispersion. This relation does not seem to be a simple artifact of firm complexity. The less material investors and analysts must digest to get valuation-relevant information from company managers, the better they are at predicting subsequent value-relevant events.

Our paper has a policy implication for the SEC. If a central purpose of the 10-K is effective communication of valuation-relevant information to investors, then the SEC should focus less on style—which is undifferentiated in 10-Ks—and instead encourage managers to write more concisely. Clearly, an SEC rule simply dictating page limitations (presumably conditional on factors such as firm size and industry) is not a reasonable solution. The SEC, however, should emphasize to filers that the benefit of exhaustive disclosure in the interest of litigation avoidance must be balanced with the costs of information overload and effective communication. Concisely written documents are more likely to be read, and the information from the 10-K is more likely to be effectively incorporated into stock prices and analyst forecasts.

The full paper is available for download here.

Both comments and trackbacks are currently closed.