Attention to Bitcoin

In this paper, we investigate different channels of information to predict Bitcoin's volatility and abnormal returns. We analyze the textual data in news about blockchain technology, major currencies and macroeconomic, and we... [ view full abstract ]

In this paper, we investigate different channels of information to predict Bitcoin's volatility and abnormal returns. We analyze the textual data in news about blockchain technology, major currencies and macroeconomic, and we investigate the predictive and causal power of extracted information to model the dynamic of bitcoin price. We apply Latent Dirichlet Allocation technique to classify and decompose news text into topics, we then show how the sentimental value and time dimensional of the topics can predict market characteristics of bitcoin. We find that the uncertainty of economy can shift the attention of traders to an unregular market like bitcoin.