What is the impact of lexical complexity in white papers on ICO performance and the likelihood of the project to be a scam?

Details

Supervisors

Thewissen, James

Faculty

Louvain School of Management

Degree label

Master [120] en sciences de gestion, à finalité spécialisée

Abstract

An initial coin offering (ICO) is a digital fundraising method which is known to be highly unregulated and thus a source of uncertainty and information asymmetry for investors. The latter is problematic as ICOs are often subject to fraudulent activities, called scams. Prior literature has therefore been applying textual analysis to ICO white papers in order to determine specific textual components, such as thematic content or linguistic errors, which could be used as signals for performance and for potential fraud. This master's thesis aims at completing the existing literature by investigating lexical complexity and its components (lexical diversity, lexical sophistication and lexical density) as such signals through regression analyses. We proceed to our analysis with a database composed of 2.503 initial coin offerings issued between 2015 and 2021. We first find that lexical complexity has a positive effect on the amount raised during the coin offering period, but that this positive marginal impact is getting smaller after reaching some specific level of lexical sophistication and/or density. Moreover, we deepen our analysis and find that the additional amount raised is superior for ICOs coming from English-speaking countries for a given level of lexical complexity in the white paper, except when we consider lexical density. Finally, we find that lexical complexity cannot be used as a signal for scams, as there does not seem to be any relationship between the level of lexical complexity in the white paper and the likelihood of the project to be a scam.