Textual fundamentals in earnings press releases. (June 2022)

• Record Type: Journal Article
• Authors: Li, Ken
• Is Part Of: Advances in accounting. Volume 57(2022)
• Publication Date: 2022-06
• Subjects: G10 -- M40 -- M41; Textual analysis -- Earnings announcements -- Tone -- Textual fundamentals -- Earnings press releases -- Neural network; Accounting -- Periodicals; 657.05
• DOI: 10.1016/j.adiac.2022.100591
• Languages: English
• ISSNs: 0882-6110

Abstract

Abstract: I re-examine the information content of earnings press releases (EPRs) by distinguishing textual fundamentals (TF), which captures information content in non-tone language, from tone, which captures optimism and pessimism and is the focus of prior research. I find after controlling for standard firm characteristics, TF positively predicts future operating performance, unlike tone, which negatively predicts future operating performance. I document a return continuation in TF, in contrast to the return reversal in tone documented in prior literature. My findings distinguish informative content in TF and misleading content in tone in EPRs.