Abstract

One application of risk management models deals with the identification of fraudulent initial coin offerings and crypto-assets. Crypto-assets are the first application of blockchain technology and are considered one of the largest markets in the world remaining unregulated. Within the last decade, digital currencies, operating independently of a central bank have grown massively in popularity, price and volatility. Should we focus on Bitcoin, we would see that the market capitalization is close to USD173 billion as of May 2020. Bitcoin is the oldest, most popular and widely used digital currency and it offers low-cost, decentralized transfer of value anywhere in the world with the only constraint represented by the availability of an internet connection. Because of its construct and the value it provides, Bitcoin represents a new and significantly lower-cost alternative to traditional banking transfer systems. New generation start-up and existing businesses prefer its use to avoid rigid and long money-raising protocols imposed by classical channels like banks or venture capitalists over the inner value of their business by selling tokens, i.e. units of the chosen cryptocurrency, like a regular firm would do through an initial public offering. Indeed, fraudulent activities perpetrated by unscrupulous start-ups are frequent and the ability to highlight in advance clear signs of illegal money raising is crucial. In this chapter, we describe a statistical approach to detecting which characteristics of an ICO are related to fraudulent behaviours. We leverage several different variables like entrepreneurial skills, number of people chatting on Telegram on the given ICO and relative sentiment, type of business, country issuing, token pre-sale price. Through logistic regression, multinomial regression and Telegram-based sentiment analysis, we can shed a light on the riskiest ICOs.