Panteion University of Social and Political Sciences
Προσδιοριστικοί παράγοντες επιτυχίας στις Αρχικές Προσφορές Νομισμάτων (Initial Coin Offerings-ICOs)
Abstract
dc:descriptionThis doctoral dissertation investigates the factors influencing the success of Initial Coin Offerings (ICOs) conducted through Decentralized Exchanges (DEX). Although ICOs have been the subject of extensive research over the past years, DEX ICOs remain an underexplored field despite their rapid growth. The majority of existing empirical studies focus primarily on traditional ICOs, while the field has now shifted towards DEX ICOs, which are expanding dynamically within the framework of Decentralized Finance (DeFi). The specific characteristics of this new issuance process and market have not yet been systematically examined using advanced analytical approaches, nor has sufficient attention been given to forecasting success based on project-level features. This dissertation addresses this research gap through a comprehensive and multi-level methodological approach. Initially, a bibliometric analysis and thematic mapping of the relevant international literature were conducted using Natural Language Processing (NLP) techniques to identify key research trends and existing gaps. Subsequently, an original dataset comprising 3.751 DEX ICOs conducted through the DxSale and Pinksale platforms during the study period was analyzed. Fundraising success was operationalized as the percentage of the Hardcap achieved (Success Ratio). The empirical analysis utilized predictive Machine Learning models (ML), including Logistic Regression, Random Forest, XGBoost, Support Vector Machines, k-nearest neighbor, and Bayesian Regularized Neural Networks, combined with a feature selection process, aiming to estimate success likelihood based on project characteristics. To assess the generalizability of the models, the ten-fold cross-validation technique was applied. The findings highlighted key success factors, including campaign duration (Duration), liquidity lock percentage and period (Lock Percentage, Lock Period), and the minimum and maximum contribution limits per investor. Moreover, shorter campaign durations were found to be positively associated with success, while higher liquidity lock percentages also correlated with increased success rates. The results of the analysis enable the development of a reliable predictive model for estimating the success probability of new DEX ICOs, thus offering a practical tool for issuers and investors. This dissertation contributes to the emerging literature on decentralized fundraising mechanisms by offering a large-scale empirical analysis of the determinants of DEX ICO success using Machine Learning techniques. The findings provide new evidence on how project-level design characteristics, including campaign duration, liquidity lock conditions, and investor contribution limits, are associated with fundraising performance. In doing so, the study enhances the theoretical understanding of decentralized token offerings and offers practical insights for issuers, investors, and potentially regulatory authorities.
Degree
thesis:*- Grantor dc:publisher
- Panteion University of Social and Political Sciences
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Georgitseas, Panagiotis
- Γεωργιτσέας, Παναγιώτης
Subjects
dc:subject × 18- Τεχνολογία Κατανεμημένου Καθολικού
- Κρυπτοστοιχεία
- Αποκεντρωμένη χρηματοοικονομική
- Αποκεντρωμένα Ανταλλακτήρια
- Αρχικές Προσφορές Κρυπτονομισμάτων
- Μηχανική μάθηση
- Blockchain
- Crypto-assets
- Decentralized Finance (DeFi)
- Decentralized Exchanges (DEXs)
- Initial Coin Offerings (ICOs)
- Machine learning (ML)
- Οικονομικά και Επιχειρήσεις
- Κοινωνικές Επιστήμες
- Χρηματοοικονομικά
- Economics and Business
- Social Sciences
- Finance
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/61471