News Sentiment and Cryptocurrency Volatility

Abstract

The cryptocurrency market has shown remarkable growth in the last decade, resulting in heightened interest in research on several aspects of cryptocurrencies. The drastic price fluctuations have attracted attention from investors, but they have also raised concerns from national regulatory institutions. Several studies are conducted to understand the factors and the dynamics of its value formation. It is becoming more important to be able to value cryptocurrencies as an investor and as part of the process to legitimize them as a financial asset. This study aims to contribute to this field of research by examining the relationship between cryptocurrency’s volatile returns and the effects of different types of news on selected cryptocurrencies. This paper categorizes the news about cryptocurrencies and determines the effect of news from each category on the return structure of each cryptocurrency. By using 1054 news sources, 22 categories are created, and a clustering analysis is used to set these categories into six groups. These groups are modelized in proper ARCH family models, which are created for different cryptocurrencies to analyze the effect on volatility. The results show that different cryptocurrencies react differently to various news categories. News about regulations from national authorities exhibit a significant effect on all selected cryptocurrencies.