Managerial tone and linguistic features in earnings calls: Text-mining evidence on investor and analyst responses in Chinese firms

This study examines how investors and analysts respond to multiple linguistic features in earnings calls of Chinese A-share firms. Using a sample of 11,262 earnings calls from 2010 to 2023, we combine text-based linguistic analysis with event-study measures of short-window market reactions and analyst forecast outcomes. The results show that optimistic tone is positively associated with short-window market reactions and analyst forecast revisions, while linguistic vagueness and sentiment shifts are generally associated with less favorable market reactions and, in some cases, with less favorable or more dispersed analyst responses. Analyses distinguishing between scripted prepared remarks and unscripted Q&A sessions indicate that tone-related associations are present in both communication segments. We further find that tone–response associations are stronger when earnings performance is favorable and when governance quality is high. Overall, the findings provide additional evidence on how multiple dimensions of managerial language are perceived and priced by investors and analysts in a Mandarin disclosure environment, while underscoring differences in governance quality in shaping tone–response relationships.