The nature and sources of international variation in formal institutions related to initial coin offerings:preliminary findings and a research agenda

As prior researchers have suggested,a firm’s success in an international market depends on how well its strategy fits the nonmarket environment,such as formal institutions.This paper examines the determinants of formal institutions around new areas of economic activities.Specifically,we propose a framework for understanding how the quality of formal institutions in promoting entrepreneurship drives the focus of such institutions concerning initial coin offering(ICO),which is emerging as a popular fundraising method.The paper uses inductive analysis to examine how nonmarket factors—such as a jurisdiction’s tax haven nature,regulators’perceptions of ICOs as threats to national or political interests,and trade and industry associations—might moderate the relationship between the quality of institutions and the focus of such institutions regarding ICOs.One of this study’s key findings is that an economy’s quality of entrepreneurship-related institutions,perceived threats to national/political interests,and tax haven nature lead to different policy orientations.Consequently,regulators assign different importance when promoting crypto-entrepreneurship and dealing with associated risks.Regulators focusing mainly on promoting crypto-ventures have taken measures to enrich the blockchain ecosystem and provided tax and non-tax incentives to attract such ventures.Regulators focusing mainly on dealing with crypto-venture risks rely on a regulatory sandbox and close regulatory monitoring of such ventures.

Initial coin offerings(ICOs):Why do they succeed?

Recent literature has addressed initial coin offering(ICO)projects,which are an innovative form of venture financing through cryptocurrencies using blockchain technology.Many features of ICOs remain unexplored,leaving much room for additional research,including the success factors of ICO projects.We investigate the success of ICO projects,with our main purpose being to identify factors that influence a project’s outcome.Following a literature review,from which several potential variables were collected,we used a database comprising 428 ICO projects in the banking/financial sector to regress several econometric models.We confirmed the impacts of several variables and obtained particularly valuable results concerning project and campaign variables.We confirmed the importance of a well-structured and informative white-paper.The proximity to certain markets with high availability of financial and human capital is also an important determinant of the success of an ICO.We also confirm the strong dependency on cryptocurrency and the impact of cryptocurrency valuations on the success of a project.Furthermore,we confirm the importance of social media in ICO projects,as well as the importance of human capital characteristics.Our research contributes to the ICO literature by capturing most of the success factors previously identified and testing their impacts based on a large database.The current research contributes to the building of systems theory and signaling theory by adapting their frameworks to the ICO environment.Our results are also important for regulators,as ICOs are mainly unregulated and have vast future potential,and for investors,who can benefit from our analysis and use it in their due diligence.

Imbalanced Problem in Initial Coin Offering Fraud Detection

ICOs,the initial coin offerings,are a common way to raise funds for blockchain projects.Fraudulent ICO projects not only cause financial losses to investors but also cause a loss of confidence in the blockchain capital market.Whitepapers are usually the most important information source,so it is feasible to identify fraudulent ICO programs by analyzing whitepapers.However,the fraud samples are difficult to collect,and the classes are imbalanced.In this study,we attempt to solve this problem by extracting linguistic features from the ICO whitepaper and using a variety of cutting-edge machine learning and deep learning algorithms to train the prediction model and attempt to resample,modify the weight and modify the loss function for imbalanced samples.Our optimal method achieves an AUC of 0.94 and an accuracy of 82%,which is better than other traditional standard methods,and the results provide important implications for ICO fraud detection.