Convoluted corporate financial reports are just as unreadable for professional stock analysts as they are for the average investor, according to a new study.
The study, published in the current issue of the American Accounting Association journal Accounting Review, tested the readability of tens of thousands of company filings over 12 years and found that analysts’ earnings forecasts for firms with less readable reports “have greater dispersion, are less accurate, and are associated with greater overall analyst uncertainty.”
Ironically, however, the syntactic and linguistic complexity of these reports generated greater demand from investors for analysts’ commentary and greater reliance on their forecasts.
“The less readable these documents are, the more pressure it creates from clients for analysts to provide commentary and forecasts,” said Feng Li of the University of Michigan, who carried out the study with colleagues Reuven Lehavy and Kenneth Merkley. “This increases the number of analysts who weigh in on these stocks and enhances their influence on investors, even while it reduces the analysts’ collective accuracy. In sum, corporate reports that are difficult to read stymie not just average investors, but analysts as well. And, by the same token, plain English is good for amateurs and professionals alike.”
Professor Li and his colleagues arrived at this conclusion through an analysis of 33,000 corporate 10-K filings, annual reports required by the SEC that provide a comprehensive summary of public companies’ performance. To gauge the reports’ readability, the professors employ the Fog Index, a standard instrument that captures the linguistic and synaptic complexity of a document as a function of the number of syllables per word and the number of words per sentence. The index provides an estimate of the number of years of formal education required for a person of average intelligence to read the document once and understand it.
The mean Fog score for the entire sample was 19.53. This ranged from 18.58 for 10-Ks at the 25th percentile of complexity to 20.29 for those at the 75th percentile. Health care and insurance companies had the highest mean scores (20.22 and 20.16 respectively), while precious metals had the lowest at 18.43. Possibly reflecting the folksiness of its CEO, Berkshire Hathaway weighed in at 17.23, putting it among firms with the lowest scores and highest readability.
Yet, even that is well above the level that would be considered plain English. The Fog score for Reader's Digest, for example, is 8, and that of The Wall Street Journal is 12.
Former SEC Chairman Christopher Cox, who is now a corporate partner at the international law firm of Bingham McCutchen LLP, tried during his tenure to make corporate filings more readable, and said the new study backed him up. “This study underscores that verbose and hyper-technical disclosure documents waste the time of investors and analysts alike,” he said. “Worse yet, unreadable SEC filings lead to less accurate analyst reports—illustrating anew George Orwell’s observation that uncertainty is both the result and the cause of muddled thought.”
In their study, the professors analyzed the relationship between Fog scores and a variety of analyst variables. They found that higher scores (meaning lower readability) translate into coverage by more analysts, but that the increase in coverage slows down at high levels of complexity, suggesting, according to Li, that “the cost outweighs the benefit of additional coverage.”
Analysts take 10 to 15 percent longer to issue reports on 10-Ks with above-average Fog scores than they do for those with low Fog scores, something, the study noted, that is “likely to be economically important given the rapid pace at which markets impound new information.”
Forecasts based on 10-Ks in the highest Fog quartile also have about 10 percent more influence on investors than those on documents in the lowest Fog quartile, as determined by movement of stocks on the days the forecasts are issued. This is the case notwithstanding the fact that “less readable disclosures are associated with more dispersed and less accurate analyst earnings forecasts.”
Despite this fall-off in accuracy, Li acknowledged that investors might benefit from the increased coverage that less readable reports elicit. “With more analysts reporting, investors may stand an increased chance of getting some useful nugget of information,” he said. “But, on the whole, low readability is the enemy of accuracy.”
