Objective: This study analyzed the relationship between the error in financial analysts' profit forecasts and the linguistic tones that make up their reports, considering different cultural environments (collectivist vs. individualist). Method: The sample consisted of companies from Brazil and the USA in 2019, totaling 979 observations for Brazil and 17,761 for the USA. The linguistic characteristics considered were: centrality, cooperation, rapport, diversity, exclusion, and liberation, all obtained via Diction®. The tests were performed using Ordinary Least Squares models. Results: The results indicate that, in the United States, the tones of cooperation, rapport and diversity are associated with a reduction in forecast error, suggesting that corporate language has incremental informational content in more mature institutional environments. In Brazil, only the tone of diversity showed a significant relationship, being associated with a higher forecast error. Theoretical/Methodological contributions: The findings suggest that the impact of corporate language on analyst accuracy depends on the cultural and institutional context in which information is produced and interpreted, contributing to the literature on behavioral finance, textual analysis, and cultural studies in accounting and capital markets.
Ribeiro et al. (Fri,) studied this question.