Abstract:
In recent years, a novel alternative asset class called cryptocurrencies has captured significant attention from financial practitioners and academia. Cryptocurrencies generally lack quantifiable fundamentals that underpin their valuations, causing the asset class to become a prime target affected by sentiment and other behavioral factors. The purpose of this thesis is to investigate the relationship between the price dynamics of cryptocurrencies and investor sentiment. The study selected several measures of sentiment, which can be categorized into two groups: direct and indirect measures. While direct proxies are sentiment information extracted from social media and media platforms using sentiment analysis, indirect measures are other commonly-cited sentiment indicators. The initial analysis results show strong pairwise correlations between these measures, making it conceptually appealing to extract a common component that could be interpreted as an aggregated sentiment index. The index later proves to be a good predictor of cryptocurrency market returns, indicating that behavioral biases might play a significant role in the decision-making process of cryptocurrency investors.
