Linguistic errors and investment decisions: the case of Initial Coin Offering white papers

Thewissen, Jennifer [USL-B] Thewissen, James [UCL] Torsin, Wouter [UCL] Arslan-Ayaydin Özgür

Drawing on language expectancy theory, we predict that linguistic errors in ICO white papers negatively impact investors’ willingness to financially contribute to ICO projects. We manually annotate a sample of 546 ICO white papers according to 13 different error subcategories related to spelling and grammar. The error-annotated data are subsequently submitted to regression analyses which confirm that linguistic errors discourage potential investments in ICOs. Specifically, our analyses reveal the presence of “high penalty” vs. “low penalty” errors which result in higher vs. lower financial investment losses for the ICOs. The negative impact of language errors is stronger when ICO white papers are (1) written in native English-speaking countries and (2) from countries without cryptocurrency regulation. Results from an experiment confirm that this relationship is not driven by the entrepreneur- or investor-specific characteristics. Overall, we highlight that the reader identifies linguistic errors as a major ‘red flag’ that ultimately affects financial decision-making.