Abstract
We develop novel empirical measures of textual comparability that capture key topics of the entire 10-K text and incorporate the effect of economic events. For validation, we show that the likelihood of accounting information users (financial analysts, Internet-search users and corporate boards) selecting another firm as a comparable peer when analyzing a particular firm is positively associated with the pairwise textual comparability between these two firms. Furthermore, we document that firm-level textual comparability is positively associated with analyst coverage and forecast accuracy, and negatively associated with forecast dispersion, consistent with the notion that the comparability of 10-K text reduces information-processing costs. Additional analyses reveal that textual comparability increases with financial statement comparability in assisting analysts with their earnings forecasts. Overall, we contribute to the literature by proposing new measures of textual comparability and demonstrating the benefits of comparable 10-K text to accounting information users