Influence of Investors' Sentiment on Bitcoin Returns

Abstract

While cryptocurrency trading is often based on various analyses and strategies, studies show that a significant portion of traders make decisions driven by emotions. This study evaluated the impact of investor sentiment on Bitcoin returns from 2019 to 2020 using daily time-series data. We employed Vector Auto-Regression (VAR), Impulse Response Functions (IRF), Variance Decomposition (VD), and Cointegration tests to analyze the relationship between sentiment indices, including the Fear and Greed index, Google Search index, Investor Happiness index, Bitcoin Trading Volume index, and American Association of Individual Investors (AAII) index, and the Euro-Dollar price, gold price, and S&P 500 index with Bitcoin returns. The findings indicated a negative relationship between the Fear and Greed index, Euro-Dollar price, gold price, Google Search index, and S&P 500 with Bitcoin returns. Conversely, there was a positive relationship between the AAII index, Investor Happiness index, and Bitcoin Trading Volume index with Bitcoin returns.Keywords: Investor Sentiment, Cryptocurrency, Bitcoin, Vector Auto-Regression (VAR) Model IntroductionIn recent years, cryptocurrencies have attracted widespread interest as an investment asset. While cryptocurrency trading often involves various analyses and strategies, studies have shown that a significant portion of traders make decisions based on emotions (Guler, 2021; Eom et al., 2018). Bitcoin sentiment analysis tools can be powerful in identifying market trends if used effectively. This study examined the role of investor sentiment in the cryptocurrency market, specifically evaluating the impact of sentiment on Bitcoin returns from 2019 to 2020. We utilized a wide range of sentiment indicators, including the Fear and Greed index, Google Search index, Investor Happiness index, Bitcoin Trading Volume index, and American Association of Individual Investors (AAII) index. Analyzing the relationship between these sentiment measures and Bitcoin returns can provide insights into the various dimensions of investor sentiment in the cryptocurrency market. Materials & MethodsThis study evaluated the impact of investor sentiment on Bitcoin returns from 2019 to 2020 using daily time-series data. We employed Vector Auto-Regression (VAR), Impulse Response Functions (IRF), Variance Decomposition (VD), and cointegration tests to examine the short-term and long-term relationships among the variables. The model for analyzing the impact of investor sentiment indices and other independent variables on Bitcoin returns was specified as: (1) where ​ represents Bitcoin returns, which are obtained through Eq. 2: (2)where Pt ​ is the price on day t and Pt-1 ​ is the price on the previous day. Forex represents the Euro-Dollar price in the Forex market, Gold is the price of an ounce of gold, and SP500 is the United States stock market index (S&P 500). The investor sentiment indices used in the model include FG (Fear and Greed index), Happy (Investor Happiness index), Vol (Bitcoin Trading Volume index), AAII (American Association of Individual Investors index), and Google (Google Search index). The Google Search index is derived from the data on the number of searches for the term "Bitcoin price" on Google Trends.This model allowed us to evaluate the relationship between the investor sentiment indices and Bitcoin returns based on the existing theories and prior research in this field. Research FindingsThe results indicated several significant long-term relationships: The American Association of Individual Investors (AAII) index had a positive long-term relationship with Bitcoin returns. According to the impulse response functions, increases in the AAII index led to higher Bitcoin returns. The Fear and Greed index had a negative long-term relationship with Bitcoin returns. Increases in this index resulted in lower Bitcoin returns. The Euro-Dollar exchange rate had a negative long-term relationship with Bitcoin returns. Higher Euro-Dollar prices led to decreased Bitcoin returns. The gold price had a negative long-term relationship with Bitcoin returns. Increases in gold prices corresponded with lower Bitcoin returns. The Google Search index had a negative long-term relationship with Bitcoin returns. Higher search volumes for "Bitcoin price" were associated with decreased Bitcoin returns. The Investor Happiness index had a positive long-term relationship with Bitcoin returns. Increases in this index led to higher Bitcoin returns. The S&P 500 index had a negative long-term relationship with Bitcoin returns. Higher S&P 500 levels corresponded with lower Bitcoin returns. The Bitcoin Trading Volume index had a positive long-term relationship with Bitcoin returns. Increased trading volume was associated with higher Bitcoin returns. These findings provided insights into the multi-dimensional relationships between investor sentiment, macroeconomic factors, and Bitcoin returns over the long term. Discussion & ConclusionThe findings of this study revealed several important insights about the relationship between investor sentiment and Bitcoin returns: 1) The Fear and Greed index, Euro-Dollar exchange rate, gold prices, Google search volume for "Bitcoin price", and S&P 500 index all had negative long-term relationships with Bitcoin returns. Increases in these variables corresponded with decreases in Bitcoin returns. 2) The American Association of Individual Investors (AAII) index, Investor Happiness index, and Bitcoin Trading Volume index had positive long-term relationships with Bitcoin returns. Increases in these sentiment indicators were associated with higher Bitcoin returns. These results demonstrated the significant influence of investor sentiment on the cryptocurrency market. Emotions and perceptions appeared to play a crucial role in driving Bitcoin returns over the long term. Investors reflected their sentiment in their trading decisions, which in turn impacted price dynamics in the Bitcoin market. This underscored the importance of understanding how various sentiment indicators affected financial markets, including cryptocurrencies. Investor sentiment analysis can provide valuable insights that inform investment decision-making. By monitoring sentiment metrics, investors can make more informed decisions and adjust their portfolios accordingly as sentiment shifts. In conclusion, this study highlighted the multifaceted relationship between investor sentiment and Bitcoin returns. The findings emphasized the need for investors to closely track sentiment indicators when participating in the cryptocurrency market. Further research in this area can yield additional insights into the behavioral aspects of cryptocurrency investment.

Keywords

• investor sentiment
• cryptocurrency
• bitcoin
• vector auto-regression (var) model