Do cryptocurrencies feel the music?

The aim of this paper is to examine the influence of global mood on cryptocurrencies trading activity. Following an original way, we construct a real-time language-free proxy of sentiment on the basis of songs investors choose to listen to. We use a sample of daily data spanning a period of four years and we find that our music-based measure is a reliable proxy of sentiment being responsive to mood swings of individuals. Interestingly, we show that music sentiment is negatively and significantly correlated with contemporaneous trading volume and price volatility, does not suffer from a recall bias, is not driven by any specific cryptocurrency and is robust to listeners' preferences. The effect is significantly persistent to up to four lags for price volatility and reverts at day 3 with respect to trading volume.

Automated summary

The aim of this paper is to examine the influence of global mood on cryptocurrencies trading activity. By constructing a real-time language-free proxy of sentiment based on songs chosen by investors, the study finds that music sentiment is a reliable indicator of individuals' mood swings. Interestingly, the research shows a negative and significant correlation between music sentiment and trading volume as well as price volatility, which is not influenced by recall bias or specific cryptocurrencies, and is robust to listeners' preferences. The effect persists for up to four lags in price volatility and reverses on day 3 for trading volume.