Sentiment disagreement and bitcoin price fluctuations: a psycholinguistic approach

ABSTRACT

We investigate the extent to which Bitcoin price fluctuations are associated with investors’ sentiment disagreement. We employ three textual sentiment analysis techniques: 1) a Python library offered by the Computational Linguistics and Psycholinguistics Research Center; 2) Loughran and McDonald’s (2011) dictionary; and 3) semantic orientation by the point-wise mutual information method. The results show that investors’ attention and sentiment disagreement induce extremely high volatility and jumps in Bitcoin prices. These findings complement existing studies on how investors’ sentiment manifests in asset prices.

KEYWORDS:

• Bitcoin
• computational linguistics
• attention
• disagreement

JEL CLASSIFICATION:

• G12
• G14

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ We calculate the realised return volatility of Bitcoin using Bitstamp’s closing prices. Bitstamp is ranked first among the eligible exchanges in terms of trading volume.

² See Das and Chen (2007), Goh, Heng, and Lin (2013), Ma, Sun, and Kekre (2015), Zhang, Bhattacharyya, and Ram (2016), and Adamopoulos, Ghose, and Todri (2018).

³ Textual sentiment analysis can be substantially noisy (e.g. Loughran and McDonald (2016)). Thus, we follow Turney and Littman (2003) and apply a learning algorithm to online sentiment classification.

⁴ https://seekingalpha.com/.

⁵ Kearney and Liu (2014) and Loughran and McDonald (2016) offer a survey on textual sentiment analysis techniques in finance.

⁶ Refer to https://www.clips.uantwerpen.be.

⁷ For more details, we direct readers to Bill McDonald’s website, www3.nd.edu$/\sim$mcdonald.

⁸ The control variables are obtained from the Federal Reserve Economic Data of the Federal Reserve Bank of St. Louis.

⁹ Bitcoin is traded on a 24–7 basis whereas usual financial securities are traded only on business days. We are grateful to an anonymous referee for suggesting this exercise. The control variables are not available on non-business days.

¹⁰ Potential sentiment proxies in the stock market include investor surveys, investor mood, retail investor trades, mutual fund flows, closed-end fund discount, IPO first-day returns, IPO volume, equity issues over total new issues, and insider trading.