The relationship between financial performance and readability is discussed in a recent issue of this journal (Subramanian, Insley, & Blackwell, 1993). Subramanian et al.'s paper analysed 60 annual reports from two groups of profitable and unprofitable corporations. They found a significant difference between mean readability levels of two groups of annual reports. In particular, their analysis showed that annual reports of good performers were significantly easier to read than those of poor performers. Narrative sections of annual corporate report are growing in importance and president's letter to stockholders and management discussion and analysis, in particular, represent a major opportunity for managements to discuss both financial and non-financial events. These narrative sections may be especially useful to unsophisticated investors because they can be used by managements to explain annual corporate performance in non-technical language. Subramanian et al.'s paper, given its concentration upon these important (and often neglected) narrative disclosures is, therefore, to be welcomed. It provides a useful contribution to debate on relationship between annual corporate performance and readability of financial statements. In order to fully appreciate findings of study, however, I feel that it is necessary to evaluate them within context of other readability research studies, and more broadly, within annual corporate reporting literature. The remainder of this paper addresses this contextualization. CONTEXTUALIZATION WITHIN THE READABILITY RESEARCH LITERATURE Subramanian et al.'s main finding is that mean readability levels of profitable corporations were significantly easier to read than those of unprofitable ones. They thus conclude that the results of this study contradict earlier findings by Courtis (1986) and Jones (1988). Both these studies failed to show a positive correlation between readability and performance. Subramanian et al. attribute Courtis' (1986) findings to fact that his study used four indirect measures of performance (current ratio, leverage, earnings variability, and rate of return on total assets), while their study used a direct measure, net income. This argument seems reasonable. In their interpretation of Jones' (1988) findings, however, they, in effect, unnecessarily weaken their own case. They focus solely upon which Jones (1988) found was negatively related to readability and point out that one explanation for this result is that turnover is quite distinct from net income. They, however, surprisingly omit to mention that Jones (1988) also tested net profit to sales and return on capital employed. These ratios are also negatively related to readability implying that increased profitability is associated with decreased readability. These results, which are at variance with Subramanian et al. study, may reflect fact that Jones (1988) used a relative measure of profit while Subramanian et al. used an absolute one. Alternatively, their suggestion, that turnover and profitability are poorly correlated, may be correct. Another plausible explanation is that Jones (1988) study was of a longitudinal nature on one firm from 1952-1985. This contrasts with cross-sectional, multi-firm approach used in Subramanian et al. The company used by Jones (1988) may have been atypical or time itself may have been an underlying explanatory variable for decline in readability. This latter explanation is consistent with findings of Jones (1993) who, after reviewing this issue, suggests that readability of corporate annual reports has declined over last thirty years. In fact, Jones' (1988) conclusion does indeed suggest possible causative effect of time: turnover increased, as time passed, so readability declined. When findings of other readability studies are evaluated, it is clear that Subramanian et al. …
Read full abstract