An investigation of sentiment analysis of information disclosure during Initial Coin Offering (ICO) on the token return

【Author】 Rasivisuth, Pornpanit; Fiaschetti, Maurizio; Medda, Francesca

【Source】INTERNATIONAL REVIEW OF FINANCIAL ANALYSIS

【影响因子】8.235

【Abstract】Initial Coin Offerings (ICOs) have emerged as vital sources of equity funding, yet there is mixed evidence so far about the relationship between ICO returns and non-financial information (e.g., ICO ratings, whitepapers, and sentiment). Our study, based on data from 391 tokens, reveals a mismatch between ICO ratings and actual token returns. We find that raw ICO characteristics and sentiment analysis offer limited insight into this discrepancy. Extracting sentiment and quantitative attributes from whitepapers proves impractical for token return analysis. Furthermore, we introduce a novel ICO index, combined with sentiment analysis of tweets, which significantly enhances the statistical analysis of factors driving six-month token returns. Additionally, our machine learning model offers a promising alternative to traditional token ratings, enabling transparent forecasting of post-ICO returns. These findings provide insights into leveraging technology to enhance capital raising for blockchain startups and the evolving landscape of transparent token assessments.

【Keywords】Initial coin offering; Sentiment analysis; Natural language processing; Machine learning

【发表时间】2024 OCT

【收录时间】2024-07-29

【文献类型】

【主题类别】

--