The study tested whether sentiment is a systematic risk factor affecting cryptocurrency returns. To explore this, cryptocurrencies are grouped by their sensitivity to sentiment, as measured by changes in the CryptoSent index. The goal was to determine if those with higher sensitivity have different average returns.
Regression analysis calculates sentiment loadings and sorts cryptocurrencies into quintiles based on these values.
The strategy applies a long-short approach, investing in the most sentiment-sensitive cryptocurrencies and shorting those with the least sensitivity.
The results show a significant negative return for the sentiment-related long-short strategy. Cryptocurrencies highly sensitive to sentiment experienced a notable negative return of 2.6% over the following week, while those less sensitive showed insignificant changes.
Over a longer period, the trend persisted, with high-sensitivity cryptocurrencies yielding a 5.1% negative return over the following month.
The findings confirm that cryptocurrencies with high sentiment sensitivity tend to perform worse in the short term, supporting the hypothesis that sentiment is a risk factor influencing returns.
