Annual report readability variability: tests of the obfuscation hypothesis

The purpose of the paper is to investigate the presence of variability within the chairman’s address section of the annual report. The coefficiency of variation (V) is used as a simple statistical measure, with calculations based on three Flesch‐based Reading Ease scores from 100‐word passages of 120 annual reports of Hong Kong public companies for 1994/5. Vs range from 4.83 percent to 89.64 percent and 106 companies have Vs greater than 10 percent. Discernible reading ease patterns are shown to be present according to chi‐square tests. Management propensity to obfuscate is introduced in the form of hypotheses, namely, companies with high variability and low readability will be associated with “bad news” and high levels of press coverage, this last variable being introduced into the readability literature for the first time. The obfuscation was accepted for companies with high financial press coverage.